Document VJnp8XzgrzjQz9bQ4Yweg3ZN

FILE NAME: Talc (TALC) DATE: 2010 DOC#: TALC126 DOCUMENT DESCRIPTION: IARC Report Vol 3 on Talc TALC NOT CONTAINING ASBESTIFORM FIBRES 1. Exposure Data Introduction Talc refers to both mineral talc and industrial mineral products that are marketed under the name talc and contain proportions of mineral talc that range from about 35% to almost 100%. The mineralogy of airborne particles in talc mines is restricted by that of the deposit and associated rocks. Therefore, mines and mills provide an opportunity to characterize exposure to one specific source of talc mineralogically. In contrast, the mineralogy of talc in an industrial setting where talc products are used may be difficult to characterize, because many different sources of talc are available for almost every application. Industrial talcs are quite variable in their talc content and in the identity and proportion of other minerals that they contain. In addition, talc is part of a complex mixture of materials in user industries. Talc particles are normally plate-like. When viewed under the microscope m bulk samples or on air filters, they may appear to be fibres and have been identified as such. Talc may also form as true mineral fibres that are asbestiform; asbestiform describes the pattern of growth of a mineral that is referred to as a `habit'. Asbestiform talc fibres are veiy long and thin and occur in parallel bundles that are easily separated from each other by hand pressure. Asbestos is a commercial term that describes six minerals that occur in the asbestiform habit: actinolite, anthophyllite, chrysotile, grunerite, riebeckite and tremolite (IARC, 1977). Similarly to talc, these six minerals occur more commonly in a nonasbestiform habit, and may also be elongated without being asbestiform. Actinolite, anthophyllite and tremolite may occur in some talc deposits; when asbestiform, they constitute asbestos and, when not asbestiform, they are referred to as mineral fragments or cleavage fragments. -277- 278 IARC MONOGRAPHS VOLUME 93 1.1 Chemical and physical data 1.1.1 Nomenclature CAS Registry No:. 14807-96-6 Chem. Abstr. Name: Talc Synonyms1: Soapstone; steatite; talcum Trade names1: Trade names of industrial, cosmetic and pharmaceutical talc include Agalite, Asbestine, Australian microcrystalline, Beaver White 200, CP 10-40, CP 38-33, Crystalite CR 6002, Desertalc 57, Emtal 500, Emtal 549, Emtal 596, Emtal 599, Ex-IT, Fibrene C 400, Finntalc, French Chalk, FW-XO, HSDB 830, IT Extra LMR 100 Microneeca K l, Micro White 5000A, Microtalco IT Extra, Mistron, Montana talc, MP 25-38, MP 40-27, MP 45-26, MST, MT 12-50, Mussolinite, NCI-CO6018, Nytal 200, Nytal 400, Pk-C, Pk-N, Plustalc, Polytal 4641, Polytal 4725, Snowgoose, Steawhite, Supreme, Supreme dense, Talcan PK-P, Talcron CP 44-31 and Westmin. Rocks or mineral composites that contain talc mineral include agalite, potstone, soapstone and talcite. Soapstone generally contains at least 25% of minerals other than talc while talcite is sometimes used to describe rock that contains at least 75/o talc (Harben & Kuzvart, 1996). Steatite originally referred to a rock that is relatively pure talc; today, it denotes a ceramic body with a high talc content that is used as an electrical insulator. The talc that is used in such applications is known as steatitic talc. French chalk is soft massive talc (Piniazkiewicz et al., 1994). Talc has also been referred to as snowgoose, agalite and kerolite. Industrial talc generally refers to products that contain abundant minerals other than talc; cosmetic talc now normally contains >98% talc (Zazenski et al, 1995) but the content may have been lower in the past (Rohl et al, 1976). Pharmaceutical talc contains >99% talc. Talcum powder is cosmetic-grade talc (Zazenski et a l, 1995). Pyrophyllite is similar to talc in atomic structure but contains aluminium instead of magnesium (Al2Si4Oio(OH)2 ) (Bish & Guthrie, 1993); the two minerals do not occur together in nature, although they have similar industrial applications. 1.1.2 Structure o f the typical mineral Chemicalformula-. Mg3Si4 O l0(OH)2 Molecular weight: 379.26 ., The original X-ray spectra of talc (Gruner, 1934; Hendricks, 1938) indicated that mineral talc had a monoclinic structure. Later investigations (Ross et al, 1968; Rayner & Brown, 1973) demonstrated that talc is triclinic (Table 1.1). The small deviations from 90 in angle a and angle y result in the triclinic symmetry. Indexing the X-ray diffraction ' These synonym s and trade nam es cover talc, m aterials that contain talc and talc that is contaminated with other minerals as admixtures. TALC 279 pattern as a monoclinic structure assumes that angles a and y are each 90 and doubles the magnitude of one of the lattice parameters (parameter `c' in Table 1.1). Table 1.1. Lattice parameters and crystallographic axes of talc Lattice parameters (nm) a b c Crystallographic axes a P y System References 0.5255 0.5293 0.9137 0.9179 0.9448 0.9496 9046' 9057' 9855' 989L 9000' 9003' Triclinic Triclinic Ross et al. (1968) Rayner & Brown ( 1973) The structure of talc is characterized by a hexagonal sheet arrangement of siliconoxygen tetrahedral groups linked in a common plane. Each silicon--oxygen tetrahedron shares three planar oxygen atoms with its neighbouring tetrahedra; the fourth oxygen, the apex of the tetrahedron, is not shared. Two such sheets are orientated so that unshared apical oxygen atoms face each other. The sheets are bonded by magnesium atoms that are coordinated octahedrally by two oxygen atoms from each tetrahedral sheet and two hydroxyl groups. This structural arrangement results in a double-sheet structure in which the valence demands of the constituent atoms are completely satisfied without interlayer cations; these double-sheet units are held together only by weak van der Waal's bonds. The double-sheet units are easily separated by slight forces that result in a perfect cleavage direction in the basal plane (Rohl et al, 1976; Pooley & Rowlands, 1975). The structure o f talc is depicted in Figure 1.1 (see cover photo of this Volume). 1.1.3 Chemical and physical properties o f mineral talc Hardness: 1 on M ohs' scale Density: 2.58-2.83 Cleavage: (001) perfect Colour. Pale to dark green or greenish grey to black; also white, silvery-white, grey, brownish Luster: Translucent; pearly, greasy or dull Indices o f refraction: Talc is biaxial with a = l.539--1.550, P=l.589--1.594 and 7=1.589-1.600. The indices of refraction increase with iron content. Because P and y are approximately equal, talc appears to be uniaxial (Deer et al, 1962). Description: Commonly thin tabular crystals, up to 1 pm in width; talc is usually massive, fine-grained and compact; it also occurs as foliated or fibrous masses or in globular stellate groups. Talc particles are normally thin and plate-like, but the size of the individual plates varies among different bodies of ore. When viewed under the microscope on end, talc platelets may appear as fibres (Cralley et al, 1968). These are not true fibres and should not be confused with asbestifonn talc. Asbestiform talc is IARC MONOGRAPHS VOLUME 93 K) OC Figure 1.1 Schematic structure o f talc From NIMSoffice, http://en.wikipedia.Org/wiki/File:Talc.GlF TALC 281 formed when talc plates elongate parallel to the a axis within the plate to form true ribbon-like fibres of talc. These fibres may occur in an asbestifonn habit consisting of bundles of narrow fibres randomly oriented around the axis of elongation (c axis). In some deposits, including those in the Gouvemeur District of New York State, a small proportion of talc fibres are intergrown on a nanoscale with amphiboles (Stemple & Brindley, 1960; Greenwood, 1998; Wylie et al, 1997). Chemical composition: The ideal formula is Mg3 Si4O10(OH)2 . When expressed in the standard oxide form, the ideal chemical composition is: 31.9% MgO, 63.4% Si02 and 4 8% H20 (Piniazkiewicz et al, 1994). No talc is ideal, and small amounts of aluminium and iron are common impurities. Aluminium may substitute for both magnesium and silicon; iron(II) and iron(lll) may substitute for magnesium. Talc that has almost all magnesium substituted by iron is called minnesotaite and is abundant in the iron formations of Minnesota, USA (Deer et al, 1962). Fluorine is the most common substitution for the hydroxy group (Petit, 2005). Small amounts of nickel, chromium, calcium, potassium, sodium and manganese are also found in the octahedral sites while titanium may substitute for silicon in the tetrahedral site. Table 1.2 provides examples of the variability in the composition of mineral talcs, talc ores and talc products. Solubility. The solubility of talc has been described in detail by Jurinski and Rimstidt (2001). On the sole basis of dissolution under pulmonary conditions,^ authors estimated that the maximum residence time in the lung of a 1-prn `spherical' particle of talc is approximately 8 years. The reader is referred to Section 4 for a detailed description o f the kinetics o f deposition and clearance. 1.1.4 Chemical and mineralogical characteristics oj talc deposits Talc ore deposits are formed from the hydrothermal metasomatism of pre-existing rocks by fluids that contain silicon and/or magnesium. Hydrothermal fluids may be derived from fluids that migrate during retrograde or prograde regional metamorphism or from contact metamorphism that is associated with nearby or distant intrusive igneous rocks. The chemical composition of talc and its associated minerals result from the original rock type, the nature of the hydrothermal alteration and metamorphic histoiy (Harben &Kuzvart, 1996). . . The chemical and mineral compositions of talc from various locations are shown in Tables 1.2 and 1.3, respectively. (a) Talc derivedfrom mafic and ultramajic rocks Talc deposits, the protoliths of which are ultramafic (or mafic) rocks, are abundant in number but small in total production. They are found in discontinuous bodies in erogenic belts, such as the Alps, the Appalachians and the Himalayas, and form during the regional metamorphism that accompanies orogenesis. They also occur in Canada (Ontario an Q u e b e c ), Egypt, Finland, Germany, Norway, the Russian Federation (Shabry and Miassy), K) t oO Table 1.2 Chemical composition (wt%) o f selected mineral talcs, talc ores and talc mineral products IARC MONOGRAPHS VOLUME 93 Component Mineral talc'1 1 2 3 4 5 6 7 8 9 Si02 TiO, ai2o , Fe20 , FeO MnO NgO CaO Na20 kiO Loss on ignition NiO Cr20-, H.CT HCT 62 61 2 46 001 30 22 - - - 62 67 0 38 0 68 0.65 29 95 1.35 - - 62 47 0 47 0 79 0 00 31 76 0 00 _ 62 16 - 0 88 1 41 30 86 - - - 60 06 1.60 1 74 30 83 0.40 - - 60 02 - I 88 1 51 30 39 1 00 - - 60 88 0 10 1 98 0 83 - - 31 18 0 14 - - 61 07 - 2 42 1 49 _ _ 29 13 0 75 - 51 29 0 04 0 61 2 00 33 66 0 12 6 26 0 00 0 08 0 03 4 72 5.05 4 70 4 92 5.02 5 37 4 98 4 82 5 54 - - 0 06 - _ 0 32 - - 0 24 Talc oresc ld 2d 3d 4d 5d 6C T 8C 9' 10s 70 8 49 8 44 6 44 8 35 98 59 15 62 65 59 80 54 92 60 0 07 0 03 0 03 0 06 0 02 _ _ _ 0 69 0 48 0 45 1 20 0 43 0 26 0 31 0.57 0.70 0 86 0 29 0 51 0 46 0 65 3 36 1 51 0.05 0 46 2 ? - - _ -- 5 96 _ _ 0 15 001 0 02 0 03 0 03 041 _ _ 0.39 23 2 19 9 23 2 25 0 32 95 31 34 30 23 27 45 27 20 31 0 07 10.4 147 9 98 0 00 0 15 Trace 6 80 5 76 <0 15 <0 15 <0 15 0.59 0 00 _ 0.15 <0 02 03 1 <0 02 0 93 0 00 -- 0.05 _ 3 99 18.1 160 16 1 23 18 6 06 5.14 5 93 10 76 5 80 -- -- _ _ 021 _ _ 0 18 ' From Deer el al (1962) b 1, Talc, altered periodotite (M uruhatten, northern Sweden), 2, Talc (Shabrov, Urals, USSR); 3, Talc (M urphy, NC, USA); 4, Light-green talc (Malangen, Norway), 5, Green talc, altered serpentine (Parma district, Apennines, Italy), 6, Black talc, with carbonaceous material derived from a bluish gray rock (Parma, Apennines, Italy), 7, Talc (Mount Fitton, South Australia); 8, Talc, altered tremolite (Yellandu Warangal district, Hyderabad, India), 9, Greenish gray iron talc (minnesotaite) (East Mesabi range, MN, USA) c I, Talc rock (Alliance Mine, CA, USA), 2. Talc ore (Pleasanton Mine, CA, USA), 3, Talc ore (Talc City, USA), 4, Talc ore (Acme Mine, CA, USA); 5, Vermont talc- magnesite ore (USA); 6, Flotation product (Johnson, VT, USA), 7, Steatite (Yellowstone Mine, MT, <USA); 8, Average ore (Talcville NY USA)- 9 Texas talc (USA) 10, FINNTALC M30 ' ' ' ' d From Van Gosen et al (2004) c From Cludester et al (1964) r From Pence (1955) e From Mondo Minerals (2005) TALC 283 southern Spain and the USA (Arkansas, California and Texas) (Piniazkiewicz et al., 1994; Harben & Kuzvart, 1996). These deposits may contain trace amounts o f nickel, cobalt and chromium that are derived from their ultramafic protolith. One major talc deposit in eastern USA contains substantial amounts of nickel (up to 0.2%; Rohl eta l, 1976). Nickel-substituted talc is also associated with serpentine bodies, at up to 0.5% by weight (Pooley & Rowlands, 1975); pentlandite has been reported in talc from Finland from which it is recovered by flotation (Harben & Kuzvart, 1996). Quartz is uncommon in talc that has mafic or ultramafic protoliths and the fluorine content is generally low (Ross et al., 1968). Chlorite and amphiboles are usually associated with this type of talc deposit although they are commonly separated in space from the talc ore (Vermont). The amphiboles may or may not be asbestiform, depending on the local geological history. A small amount of amphibole asbestos is associated with this type of talc deposit at Soapstone Ridge, GA (USA) and anthophyllite asbestos is abundant in the vicinity of the talc at Dadeville, AL (USA) (Van Gosen et al., 2004). In a few deposits, the parent was mafic rock (Virginia (Schuyler), Georgia and Egypt) (Harben & Kuzvart, 1996). Table 1.3. Mineral composition (wt%) of talc from various locations Mineral Montana Vermont North Carolina New York3 California Talc Tremolite Anthophylite Serpentine Quartz Chlorite Dolomite Calcite Magnesite 90-93 - - - <1 2-4 1-3 - 0-5 80-92 - - - <1 2-4 1-3 - 0-5 80-92 - 0-5 -- 1-3 5-7 2-4 ~ - 35-60 30-55 3-10 2-5 1-3 0-2 1-2 1-3 85-90 0-12 _ _ <1 _ 0-3 _ - From Harben & Kuzvart ( 1996) a Gouverneur District France 70-90 <1 10-30 - (b) Talc derivedfrom magnesium carbonates Talc deposits formed from the alteration o f carbonate and sandy carbonate such as dolomite and limestone are the most important in terms of world production. Two types are recognized: (i) those derived from hydrothermal alteration of unmetamorphosed or minimally metamorphosed dolomite (Australia (Mount Seabrook and Three Springs), China, India, Republic of Korea, the Russian Federation (Onot), northern Spain (Respina) and the USA (Alabama (Winterboro), California (Talc City), Montana (Yellowstone), Washington (Metaline Falls) and West Texas); and (ii) those derived from hydrothermal alteration (including retrograde metamorphism) of regionally metamorphosed siliceous dolomites and other magnesium-rich rocks (Austria (Leogen), Brazil (Brumado), Canada 284 IARC MONOGRAPHS VOLUME 93 (Madoc), France (Trimouns), Germany (Wunsiedel), Italy (Chisone Valley), the Russian Federation (Krasnoyarsk), Slovakia (Gemerska Poloma), Spain and the USA (Chatsworth, GA, Death Valley-Kingston Range, CA, Murphy Marble belt, NC, and New York). In a few of these deposits, including the large deposit at Trimouns, France, the talc may be classified as being derived from alumino-silicate rocks (Harben & Kuzvart, 1996; Luzenac, 2004). Talc derived from magnesium carbonate may contain quartz. Van Gosen etal. (2004) suggested that, among the first group, only those that are formed by hydrothermal alteration o f dolomites that are in direct contact with igneous bodies are probably accompanied by amphiboles (e.g. Death Valley, CA, USA) and that hydrothermal deposits in carbonates that are formed by relatively low-temperature fluids derived from distant igneous bodies contain no or only very minor amounts of amphibole (Talc City, CA, Southwestern Montana and Allamoore, TX, USA). In some deposits in the second group, amphiboles may be vety abundant, especially those formed during hightemperature regional metamorphism of impure dolomites. In the Gouvemeur District New York State, for example, non-asbestiform tremolite comprises between 30 and 70% o f the talc product (Harben & Kuzvart, 1996). Gouvemeur District New York State talc that is currently marketed under the trade name Nytal is a unique industrial mineral product that can readily be distinguished from all other commercially available industrial talcs based on its mineral content. Nytal 100, for example, contains 30-50% tremolite, 20-40% talc, 20-30% seipentine, 2-10% anthophyllite and 0.14% quartz (R.T. Vanderbilt Company, 2000). The tremolite, anthophyllite and serpentine occur as mineral fragments and not as asbestiform fibres. Tremolite from this deposit has been characterized in detail (Campbell et al, 1980). Nytal also contains asbestiform fibres of talc and talc intergrown on a nanoscale with amphibole (Wylie et al, 1997). Wylie et al. (1997) estimated that the abundance o f particles that are longer than 5 pm and have an aspect ratio of 3:1 or greater in sample FD14 (identified as a commercial talc product from New York State) is 0.8x103/pg; 62% of these particles were identified as talc, 24% as fragments o f tremolite plus a small amount of anthophyllite and 14% as talc intergrown with anthophyllite. Products from other mines in this district before 1964 contained different proportions of anthophyllite and tremolite, which may be asbestiform (Chidester et al, 1964). (c) Minerals associated with talc Because talc deposits are formed from different protoliths under many different geological conditions, each talc deposit has a combination of mineralogy and mineral habit that is distinctive and, in many cases, unique. The most common minerals found in talc products include chlorite, magnesite, dolomite, tremolite, anthophyllite, serpentine and quartz. However, many other minerals have been reported; these are given in Table 1.4 (Pooley & Rowlands, 1975; Piniazkiewicz et al, 1994; Harben & Kuzvart, 1996). Some o f these minerals are beneficial to certain applications such as tremolite in ceramics. TALC 285 Table 1.4. Minerals commonly associated with talc Mineral group Name Ideal formula Carbonate Phyllosilicates Amphibole3 Tectosilicates Oxides Sufides Other minerals Dolomite Magnesite Breunnerite Calcite Si derite Ankerite Chlorite Serpentine (lizardite and antigorite) Phlogopite (mica) Sepiolite Tremolite Anthophyllite Actinolite Quartz Feldspar Magnetite Ilmenite Manganese oxide Rutile Pyrite Pyrrhotite Pentlandite Tourmalineb Graphite (Ca,Mg)C03 MgC03 (Mg,Fe)C03 CaC03 FeC03 C a (F e,M g ,M n )(C 0 3)2 (Mg,Al,Fe)i2(Si,Al)80 2o (OH)l6 Mg3Si20 5(0H )4 K2(Mg,Fe)6Si6Al2O20(OH)4 M g 8S i12O30(OH)4(H2O)4 Ca2Mg5Si80 22(0H )2 (Mg,Fe)7 Si80 22(0H )2 Ca2(Mg,Fe)5 Si80 22(0fl)2 Si02 (K,Na)AlSi30 8 Fe30 4 FeTi 0 3 Mn02 Ti 0 2 FeS2 FeS (Fe,Ni)9S8 NaFe3Al6(B 03)3Si60 18(0H )3(0H) C Compiled by the Working Group from Pooley & Rowlands ( 1975); Piniazkiewicz et al. ( 1994); Harben & Kuzvart ( 1996) a See Leake et al (1997) for precise nomenclature and chemical composition of the amphibole group. This is the formula lor one member of the tourmaline group; chemistry is highly variable. (d) Chemical composition o f talc ore The variability in the chemical composition of talc ore, talc mineral products and talc rock primarily reflects their mineral composition (see Table 1.2). 286 IARC MONOGRAPHS VOLUME 93 1.1.5 Processing o f talc ores and composition o f talc products Talc ores may be processed by a variety of techniques that include selective mining, hand sorting and milling by roller mills, hammer mills, ball mills, fluid energy mills and jet mills and are classified and separated from other minerals by froth flotation or magnetic separation. Some may be treated with acid and calcined. The particle sizes of talc and the abundance of the associated minerals are determined by characteristics of the ore, methods of processing, and the duration of grinding. Grinding breaks the talc platelets along (001) and disaggregates the particles; prolonged grinding may destroy the crystallinity (Sanchez-Soto et al., 1997; Zbik & Smart, 2005). Roller mills tend to preserve the platy structure and different types of milling affect properties such as flatness, surface roughness, roundness, width and elongation (Yekeler et al., 2004). Talc particles are platy, and sizes reflect the dimension parallel to the plate; data are not available on the thinness of the plates. Talc products also vary in particle size; median sizes range from ~1 to >20 pm and top sizes range from <10 to >100 pm. The most common designations for fineness are based on US Sieve Series and Tyler equivalence and include 200 mesh (95-98% <74 pm), 325 mesh (95-99% <44 pm) and 400 mesh (95-99% <37 pm) (Zazenski et al., 1995). Talc products that contain >95% mineral talc are used in cosmetics, baby powder, pharmaceuticals, steatite ceramics, pitch control in the paper industry and as a filler in rubber. Today, the talc in baby powders is >99% 200 mesh (Zazenski et al., 1995). Talc products that contain between 75 and 95% mineral talc are used in paper fillers, reinforced plastics, paint, ceramics and dusting compounds for rubber. Lower-purity talc is used in roofing material, patching compounds, flooring and fertilizers (Piniazkiewicz etal., 1994). Particle sizes, colour and nature of associated minerals also vary among these applications. 1.1.6 Analysis (a) Analysis o f bulk samples Talc can be identified from its optical properties by polarized light microscopy and oil immersion, from its X-ray or electron diffraction pattern, from its chemical composition and from differential thermal analysis/thermal gravimetric analysis. Chlorite has similar optical properties. Talc platelets on end and talc intergrown with amphibole in fibrous talc have complex electron diffraction patterns that may resemble other silicates, including amphiboles (Stemple & Brindley, 1960) and sepiolite (Genuine, 1987), unless carefully indexed. Anthophyllite and sepiolite have chemical compositions that are very similar to talc and require quantitative chemical analysis to differentiate them, including the use of well characterized standards in the case of dispersive X-ray analysis used in conjunction with electron microscopy. Identification of mixed mineral assemblages by X-ray TALC 287 diffraction may be difficult because of pattern overlap (Krause, 1977) and X-ray diffraction cannot distinguish asbestiform minerals from other habits. Particle size distributions that are determined by settling underestimate the abundance of larger particles and overestimate the number of smaller particles because the platy structure results in longer settling times for talc compared with spherically shaped particles o f equivalent size. Computer-controlled scanning electron microscopy has been used to provide a more accurate size distribution. Determination of the respirable fraction of bulk materials by these two methods differs significantly (Zazenski et al., 1995). (b) Analysis o f exposure The standard methods for the analysis of airborne exposures in an occupational setting where asbestos is known to be present include those of the Health and Safety Executive (1995) and the Occupational Safety and Health Administration (2005). These methods were designed to provide an index of exposure since they count only particles longer than 5 pm with a length-to-width ratio of 3:1 or more that are visible by phasecontrast microscopy. They do not determine the mineral identity of the particles counted. In a mining environment where many minerals form elongated fragments, the results of fibre counts can be difficult to interpret. In bulk samples of talcum products, for example, Cralley et al. (1968) reported that particles longer than 5 pm with a 3:1 aspect ratio in 22 talcum products represented 19% of the particles, which were predominantly talc. Conversion of fibre counts to gravimetrically based exposure metrics is complicated as this will depend on the particle size. Oestenstad et al. (2002) adjusted million particles per cubic foot (mppcf) to milligrams per cubic metre (mg/m3) using the following regression equation: In (mg/m3)=ln (m ppcf>0.62-l .20 All gravimetric measurements to monitor exposure to talc in occupational settings are taken from samples of respirable dust particles. The reader is referred to the Glossary and the monograph on carbon black for further details. 1.2 Production and use 1.2.1 Production Talc deposits result from the transformation of existing rocks under hydrothermal activity and are classified according to the parent rock from which they derive. There are three broad types of talc deposit of commercial significance (Luzenac, 2004; EUROTALC, 2005; Industrial Minerals Association-Europe, 2005): (i) talc derived from mafic and ultramafic rocks, which provides about 40% of talc supplies; the crude ore is usually grey and, to be commercially viable, may be upgraded to improve the mineralogy and whiteness (generally by flotation); (ii) talc derived from magnesium carbonates, which provides >50% of world production; and (iii) talc derived from alumino-silicate 288 IARC MONOGRAPHS VOLUME 93 rocks, from which about 10% of world production is mined, and which is sometimes found in combination with deposits of magnesium carbonate; the crude ore is generally grey due to the presence of chlorite, but no upgrading is necessary as chlorite performs adequately in the applications of interest. This wide diversity of origins and types of deposit naturally gives rise to a wide variety of ores and product grades that differ according to their mineralogical composition, colour and crystalline structure (microcrystalline or lamellar) (Luzenac, 2004; EUROTALC, 2005; Industrial Minerals Association-Europe, 2005). World production of talc and pyrophyllite in both 2003 and 2004 was estimated to be 8.3 million tonnes. Of the total production, approximately 2.15 million tonnes were confirmed to be used for talc production in both 2003 and 2004. China was the leading producer of talc in the world, followed by the USA, India, Brazil (crude) and France (crude). The Republic of Korea was the leading producer of pyrophyllite, followed by Japan and Brazil. Brazil, China, France, India, Japan, the Republic of Korea and the USA produced 84% of talc and pyrophyllite in the world (Table 1.5) (Virta, 2004). Table 1.5. World production of talc (in tonnes unless otherwise specified)"b Country Argentina Australia1 Austria (crude+so)d Bhutan11 Brazil (crude) Brazil (marketable product)1 Canada (t+p+so) Chile China (unspecified)11 Colombia (t+p+so/ Egypt (t+p+so+st)d France (crude)11 Germany (marketable+st+t)d Hungary*1 India (st) Iran1*8 Italy (t+st)d Japan Macedonia Mexico Morocco Nepal11 Norway (t+so+st)d North Korea (unspecified)11 Paraauay (t+p+so)d Peru Portugal11 Republic of Korea Romania Russia" 2000 2001 6730 178 545 130 000e 3700 300 000 7049 86 000 2421 3 500 000 15 000 40 000 350 000 8000 500 545 000 25 000 140 000 50 000 562 20 569 12 522 5852 27 000 120 000 200 9668 8200 11 344 7850 100 000 1665 173 446 140 000 3800 397 000 6300 90 000 4177 3 500 000 15 000 40 000 350 000 10 000 500 546 000 25 000 140 000 45 000 557 77 650 27 246 3923 27 000 120 000 200 11 165 . 8200 47 712 7270 100 000 2002 1643 173 741 135 000 3900 348 000 5617 90 000 3537 2 500 000 15 000 40-000 350 000 10 000 500 550 000 25 000 140 000 40 000 550 111 621 39 612 2621 28 000 110 000 200 10 685 8200 37 863 7292 100 000 2003 1759 174 000 135 000 3900 365 000 5593 90 000 4374 3 000 000 15 000 40 000 350 000 10 000 500 552 000 25 000 140 000 40 000 550 114 870 1959 2500 28 000 110 000 200 10 791 8200 47 911 10 082 100 000 2004 1800 173 000 135 000 3900 370 000 5600 90 000 4400 3 000 000 15 000 40 000 350 000 10 000 500 550 000 30 000 140 000 35 000 600 115 000 2000 2400 28 000 110 000 200 10 000 8000 48 000 10 000 100 000 TALC 289 Table 1.5 (contd) Country Slovakia South Africa Spam (t+st)d Sweden (t+so) Taiwan Thailand United Kingdom (t+p+so)d USA Uruguay (t+p+so) Zimbabwe 2000 1800 5600 100 000 20 000 - 7390 5000 851 000 2903 989 2001 2600 3218 100 000 15 000 130 6838 5000 863 000 1694 1273 2002 2290 2511 100 000 15 000 27 1702 5000 828 000 1700 911 2003 1000 4472 100 000 15 000 466 8501 5000 840 000 1700 196 2004 1500 12 065e 100 000 14 000 411e 8500 5000 857 000e 1700 From Virta (2004) p, pyrophyllite; so, soapstone, st, steatite; t, talc aWorld totals, data from the USA and estimated data are rounded to no more than three significant digits, may not add to totals shown bTable includes data available through to April 19 2005. ` Data based on Australian fiscal year ending 30 June of the year stated. d estimated ` Reported figure ' Direct sales and/or beneficiated (marketable product) 8Data based on Iranian fiscal year beginning 21 March of the year stated hData based on Nepalese fiscal year beginning mid-July of the year stated 1.2.2 Use The properties of mineral talc (platyness, softness, hydrophobicity, organophilicity and inertness) and the mineralogical composition of talc products govern their specific applications in many industries and processes including paint, polymers, paper, ceramics, animal feed, rubber, roofing, fertilizers, cosmetics and pharmaceuticals. The principal technical applications of talc in commercial products are as an anti-sticking and anti caking agent, lubricant, carrier, thickener, strengthening and smoothing filler and absorbent (Industrial Minerals Association-Europe, 2005). (a) End-use categories (i) Agriculture andfood Talc is used as an anti-caking agent, dispersing agent and die lubricant in animal feed and fertilizers. In premixes and agricultural chemicals, it is used as an inert carrier. Talc is also used as an anti-stick coating agent in several foods and as a processing aid in the production of olive oil. (Luzenac, 2004; Industrial Minerals Association-Europe, 2005). Agricultural chemicals. Talc is a functional carrier in agricultural products that offers very low moisture equilibrium, relative hydrophobicity and chemical inertness. Costs are reduced by extending expensive chemicals and improving the dispersion and flow of 290 IARC MONOGRAPHS VOLUME 93 active ingredients. Talc is appropriate for garden dusts, flea and tick powders, seed treatments and biocides (Luzenac, 2004). Anti-caking and homogenization. Talc improves the flowability of difficult raw materials, e.g. oilseed meal and finished products, and feeds with high loads of sticky ingredients such as molasses, oil, fatty products, urea, milk powder and sugar. The smooth and flat lamellae of talc cover each particle and help them to flow freely. As they are naturally water-repellent, talc particles form a barrier when they envelop other particles and reduce the evaporation and uptake of water within the product mass. Talc platelets help different constituents to blend more easily and facilitate the dispersion of sticky ingredients (Luzenac, 2004). Die lubricant. Talc is a cost-effective die lubricant especially for high-fibre, highsugar and high-mineral formulations and pelleted feeds (Luzenac, 2004). Fertilizers. Talc is used as an anti-caking agent in both prilled (pelleted or granulated) ammonium nitrate and granular fertilizers. Talc particles reduce the absorption of moisture and prevent the formation of hydrate bridges, which enables longer storage periods. In Europe, amine-coated talcs are marketed with enhanced adhesion properties that enable the amine contents to be reduced and result in lower dust levels and less environmental impact (Luzenac, 2004). Foods. Talc is an effective anti-stick coating agent that is used in several foods, such as chewing gum, candies and cured meats (Luzenac, 2004). Processing o f olive oil. In the production of olive oil, talc acts as a natural processing aid that improves extraction and increases the yield of virgin olive oil (Luzenac, 2004). Premixes. Talc is used as an inert carrier for active premix ingredients. Certain talc grades have been specifically designed for dust-free, high-specification requirements (Luzenac, 2004). (ii) Ceramics Talc imparts a wide range of properties to floor and wall tiles and sanitary ware, tableware, refractory goods and technical ceramic products. In traditional building ceramics (tiles and sanitary ware), it is used essentially as a flux to enable firing temperatures and cycles to be reduced. In refractory applications, talcs that are rich in chlorite are used to improve thermal shock resistance. Talcs with a microcrystalline form are the most appropriate for steatite ceramics. During firing, the talc is transformed into enstatite, which possesses electro-insulating properties. Talcs with a very low iron content are particularly suitable for use in frit, engobe [underglaze] and glaze compositions (Luzenac, 2004; Industrial Minerals Association-Europe, 2005). (iii) Coatings Talcs confer several properties on coatings. In interior and exterior decorative paints, they act as extenders to improve hiding power and the efficiency of titanium dioxide. The lamellar platelets of talc make paint easier to apply and improve cracking resistance and sagging, and also enhance matting. In anti-corrosion primers, talcs are used to improve TALC 291 resistance to corrosion and adhesion of the paint. They are also used in inks, jointing compounds, putties and adhesives (Luzenac, 2004; Industrial Minerals AssociationEurope, 2005). (iv) Paper Talcs are used in both uncoated and coated rotogravure papers in which they improve printability, reduce surface friction and enhance handling characteristics. They also improve mattness and reduce ink scuff on offset papers. When used as pitch-control agents, talcs `clean' the papermaking process by adsorbing any sticky resinous particles in the pulp onto their platy surfaces, and thereby prevent the agglomeration and deposition of these on the felts and calenders. In contrast to chemical pitch-control products that pollute the process water, talc is removed with the pulp, which enables the papermaker to operate more easily in a closed circuit. In specialty papers such as coloured papers or labels, talcs help to improve quality and productivity (Luzenac, 2004; Industrial Minerals Association-Europe, 2005). (v) Personal care As it is soft to the touch and inert, talc has been valued for centuries as a body powder. Today, it also plays an important role in many cosmetic products, including products for feminine hygiene and baby powders, and provides the silkiness in blushes, powder compacts and eye shadows, the transparency o f foundations and the sheen of beauty creams. In pharmaceutical products, talc is an important excipient that is used as a glidant, lubricant and diluent. Soap manufacturers also use talc to enhance the performance of skin care products (Luzenac, 2004; Industrial Minerals AssociationEurope, 2005). Table 1.6 presents information on levels of talc in cosmetic products in the USA and Table 1.7 gives the composition of some examples of products that are used for body care. (vi) Plastics Talcs impart a variety of properties to polypropylene, such as greater stiffness and improved dimensional stability in automotive parts, household appliances and white goods. Advanced milling technology is required to obtain the finest talcs without diminishing the reinforcing power of their lamellar structure. Talcs are also used for the anti-blocking of linear low-density polyethylene and as a nucleating agent in semicrystalline polymers. In polypropylene that is used in food packaging applications, talc is a highly effective reinforcing filler. The grades of talc used for this purpose include calcined, surface-treated, ultrafine grind and high aspect ratio (Luzenac, 2004; Industrial Minerals Association-Europe, 2005). (vii) Roofing Talc is a high-performance product that is used to back surfacing asphalt shingles. The use of talc is even more important in the growing market for laminated shingles in 292 IARC MONOGRAPHS VOLUME 93 which handling is more complex, wear and tear on machinery is greater, cutting is doubled and adhesion o f the interlayer is critical (Luzenac, 2004). Table 1.6. The number of cosmetic products in the Cosmetics and Toiletries Formulations Database in the USA that contain talc or talcum Product categories No. of products Antiperspirants and deodorants 22 Baby products 6 Bath and shower products 2 Beauty aids3 184 Creams 14 Hair care products 1 Lipsticks 5 Lotions 1 Shampoos 1 Shaving products 2 Sun care products 3 Miscellaneous15 8 Compiled by the Working Group from Flick (2005) a Beauty aids includes aerosol talc products, face masks, foundations, body oils, make-up bases, concealers, blushes, body powders, rouge, make-up, compact powders, eye shadows, dusting powders, eyebrow pencils, pressed powder products, face powders, mascaras, liquid talc products and powder cleansers b Miscellaneous includes aerosol talc foams, wound ointments, foundations with extracts, foot powders, liquid foundations and sport tints (viii) Rubber Talcs reduce the viscosity of rubber compounds and thereby facilitate the processing of moulded parts. They also improve the quality of extrudates, which incieases production rates and enhances the resistance to ultraviolet (UV) radiation of exterior parts such as automotive profdes. In sealants and gaskets, they provide compression resistance, while in pharmaceutical stoppers, they create a barrier against liquids. Talcs are used as insulators in cables and as processing aids in tyre manufacture (Luzenac, 2004; Industrial Minerals Association-Europe, 2005). TALC 293 Table 1.7. Composition of some products used for body care Product Wt% talc Other components Wt% other components Dusting powder Dusting powder Velvety dusting powder Face and body powder Baby powder After-bath talc Body powder 97.7 91.6 77.4 89.30 72 92.5 4.0 Perfume oil GLUCAM P-20 Preservative Magnesium carbonate Zinc stearate Tnclosan Perfume oil GLUCAM P-20 Preservative Aluminum starch, Octenyl succinate Zinc stearate Methylparaben Propylparaben Germall II Fragrance Boron nitride Methylparaben Propylparaben Imidazolidinyl urea Iron oxide (yellow) Iron oxide (red) DYNASAN 114 Magnesium stearate Kaolin Perfume oil PPG-20 methyl glucose ether Macadamia nut oil Boron nitride Silica Starch Kaolin Magnesium stearate Bentone 38/Quaterniuml8. Hectorite Isopropyl myristate Perfume Pigments 0.8 1.5 q.s. 3.0 3.0 02 0.7 1.5 q.s. 20.0 2.0 0.10 0.10 0.20 0.20 10.00 0.15 0.20 0.05 0.20 0.10 2.0 8.0 18.0 5.0 1.50 1.00 5.0 2.5 30.2 10.0 1.00 1 0 6.0 1.8 q.s 294 IARC MONOGRAPHS VOLUME 93 Table 1.7 (contd) Product Wt% talc Other components Wt% other components Powder for babies and 20.0 children Dispersing bath powder 0 Body powder 0 Talc-free body powder 0 Kaolin Rice starch Zinc stearate Eutanol G Lanette O Kukui nut oil Phenyl trimethicone Cyclomethicone Fragrance Ethoxydiglycol Oleth-2 Oleamidopropyl PG, dimonium chlorite Topopheryl acetate (Vitamin E) Cornstarch Silica Zinc stearate Zinc oxide Magnesium carbonate Kaopolite TLC Cornstarch Kaolin Mica Titanium dioxide Red mica and titanium dioxide Tapioca starch Methylparaben Propylparaben Imidazolinidyl urea 20.0 51.0 5.0 2.0 2.0 1.0 1.0 2.0 1.5 2.0 2.50 2.0 0.50 86.00 1.50 5.0 5.0 15.0 75.0 88.45 5.0 2.0 2.0 0.25 2.0 0.10 0.05 0.15 From Flick (2005) The Working Group was aware that these data are not representative of all products q.s., quantum satis (sufficient quantity) (ix) Wastewater treatment Specialty talc can improve the performance of biological wastewater treatment plants. The talc particles ballast the floes of bacteria and accelerate their sedimentation (Industrial Minerals Association-Europe, 2005). TALC 295 (x) Other Talc is used as an anti-sticking agent to powder moulds in foundries and in the manufacture of pharmaceuticals and rubber or on conveyor belts that carry foodstuffs. It is also used in other products, such as condoms and surgery gloves. Particle-wood boards (chip boards) are powdered with talc to avoid sticking when stockpiled. Talcs are also used as smooth fillers, for example in the `lead' of colouring pencils and in putties (where it can be the major component) (Industrial Minerals Association-Europe, 2005). Talc had been used as a sclerosing agent in the pleural space for the treatment of spontaneous pneumothoraces. Talc is also used for pleurodesis in the treatment of malignant pleural effusions (Dresler et al, 2005). The products used for these purposes contain 95% talc and 5% chlorite and dolomite. (b) Use patterns The worldwide use pattern for talc in 2000 was: paper, 30%; ceramics, 28%; refractories, 11%; plastics, 6%; a filler or pigment in paints, 5%; roofing, 5%; cement, 3%; cosmetics, 2%; and other miscellaneous uses, 10% (art sculpture, asphalt filler, autobody filler, construction caulks, agriculture and food, flooring and joint compounds) (Roskill Information Services Ltd, 2003). The use pattern for talc in the USA in 2004 was: ceramics, 32%; paints, 19%; paper, 16%; roofing, 6%; plastics, 4%; rubber, 3%; cosmetics, 1%; and other, 19% (Virta, 2004). The use of talc in cosmetics in the USA decreased from 34 000 tonnes in 1993 to 5000 tonnes in 2004 (Virta, 2004). The estimated world consumption of talc by geographical region in 2000 was: Asia, 43%; western Europe, 19%; North and central America, 17%; South America, 8%; Indian subcontinent and Middle East, 8%; Africa, 2%; eastern Europe and Commonwealth of Independent States countries, 2%; and Australia and New Zealand, 1% (Roskill Information Services Ltd, 2003). 1.3 Occurrence and exposure 1.3.1 Natural occurrence Talc is found in small amounts in metamorphic mafic and ultramafic rocks and in carbonates. These metamorphic rocks crop out in mountain belts such as the Alps, the Appalachians and the Himalayas and in ancient continental shields such as the Canadian shield in New York and Canada. The occurrence of talc deposits of commercial importance is described extensively in Section 1.1.4. 1.3.2 Occupational exposure Exposure to talc dust occurs during its mining, crushing, separating, bagging and loading and in various industries that use talc (see Section 1.2.2). This section reviews 296 IARC MONOGRAPHS VOLUME 93 exposure to talc during its mining and milling, other than that from the Gouvemeur District New York State mines, and in user industries, whenever this information is available. Exposure to talc is also described, where possible, for those industries in which epidemiological studies have been carried out in relation to the occurrence of cancer. (a) Mining and milling Before the 1970s, exposure measurements were made by collecting particles in an impinger and counting them by optical microscopy. Concentrations were thus expressed as million particles per cubic foot of air (mppcf). More recent studies have described levels of exposure to dust that were assessed using gravimetric measurement techniques. Table 1.8 describes studies of exposure to talc in mines and mills. In Georgia, USA, average exposures to dust were 1440 mppcf (-50 854 particles/cm3) for miners who used jackhammer drills and 52 mppcf (--1836 particles/cm3) for millers. The talc was reported to contain 45% tremolite and 45% talc, with little or no quartz (Dreessen, 1933). Average dust concentrations in a talc mine were reported to range from 32 to 855 mppcf (-1130 to 30 195 particles/cm3; six samples), whereas those in mills ranged from 17 to 1672 mppcf (-600 to 59 000 particles/cm3; 14 samples). The dust was reported to contain 70% talc, 20-30% dolomite and 10% tremolite, and no quartz except for occasional fragments; its morphology was described as `bladed crystals'. Highest exposures to dust occurred during bagging operations (Dreessen & DallaValle, 1935). Concentrations of respirable dust in mass samples from three Vermont talc mines and mills surveyed in 1975-76 are given in Table 1.9. Geometric mean exposures to respirable dust ranged from 0.5 to 5.1 mg/m3 in the mines and from 0.5 to 2.9 mg/m3 in the mills; however, exposures in the mills were generally higher than those in the mines. Optical fibre counts as high as 60 fibres/cm3 were reported. Subsequent analyses of these samples by scanning electron microscopy showed that they consisted of rolled talc and elongated talc particles. X-Ray diffraction analyses of bulk samples from these mines and mills showed that talc and magnesite were the major (20-100%) mineral components, chlorite and dolomite were minor (5--20%) components and calcite, quartz, biotite, ankerite, chromite, phlogopite and oligoclase were present in small amounts (<5%). 3 race amounts of quartz were found in 15% of the samples (Boundy et al., 1979). Dust from one closed mine was reported to contain tremolite microinclusions, but its fibrosity was not documented (Selevan et al., 1979). A cross-sectional study of occupational exposures in talc mines and mills in the USA was conducted by the National Institute for Occupational Safety and Health; the results are summarized in Table 1.10. Bulk samples from each region were analysed by transmission electron microscopy: no fibre was found in any sample of Montana talc, fibrous tremolite and antigorite were reported in Texan talcs (0.5--3.0 pm in diameter, 4-- 30 pm in length); and talcs from North Carolina contained particles with length:diameter ratios as high as 100:1, with some <0.1 pm in diameter (Greife, 1980, Gamble et al., 1982). Van Gosen et al. (2004) recently reported that the Texan talc contained little or no amphibole. TALC Table 1.8. Studies of occupational exposures in talc mines and mills Reference Location of talc deposit Date of exposure measurements Method of measurement Dreessen (1933) Dreessen & DallaValle (1935) Rubino et al. (1976); Coggiola et al (2003) Rubino et al. (1976) Boundy et al. (1979) Georgia, USA Georgia. USA Piedmont, Italy Piedmont, Italy Vermont, USA Pre-1933 Pre-1935 1946-95 1920-75 1975-76 Impinger Impinger - Impinger Optical and electron microscopy fibre counts Greife (1980); Gamble et al. (1982) Wild et al. (1995, 2002) Montana, Texas and North Carolina, USA France, Austria 1977-80 1986-92 Gravimetric Gravimetric (CIP personal sampler) Other minerals present Tremolite Tremolite, dolomite Quartz (radon, diesel exhaust) Small amounts of tremolite Dolomite, calcite, magnesite, chlorite, traces of other minerals Varied by location studied Quartz: France, <3%; Austria, <4% CIP, capteur individuel de poussire [personal dust sampler] NJ N-JO 298 IARC MONOGRAPHS VOLUME 93 Table 1.9. Concentrations (mg/m3) of respirable dust in Vermont talc mines and mills Company Area Summer 1975 No. of samples Geometric mean (mg/m3) Winter 1976 No. of samples Geometric mean (mg/m3) A Underground mine 18 0.6 Mill (1st shift) 4 1.7 Mill (2nd shift) 6 0.5 B Underground mine 15 1.5 Mill (1st shift) 22 1.8 Mill (2nd shift) 12 2.9 C Underground mine 12 0.5 Walk-in mine 7 1.2 Walk-in mine Open-pit mine 2 5.1 Mill No. 1 (1st shift) 12 0.9 Mill No. 1 (3rd shift) 3 0.8 Mill No. 2 (1st shift) 11 1.0 Mill No. 2 (2nd shift) 13 0.8 16 0.5 13 1.7 3 1.5 23 0.9 42 1.8 16 1.9 19 07 6 1.7 - - 20 1.1 4 1.4 8 0.5 3 1.1 From Boundy et al. (1979) Table 1.10. Concentrations of respirable dust in 275 samples from talc mines and mills located in Montana, Texas and North Carolina, USA Samples Geometric mean (mg/m3) Montana Texas North Carolina From mines From mills Bulk talc samples (% free silica) 0.66 (0.47-0.92)" 1.1 (0.85-1.41) <0.8 0.45 (0.18-0.71) 1.56(0.96-2.54) 2.23 0.14 (0.07-0 31) 026(1.13-0.51) 1.45 Adapted from Greife (1980); Gamble el al. (1982) aIn parentheses, 95% frequency interval Analysis of 362 personal samples o f respirable dust collected over a full shift from talc mines and mills by the Mine Safety and Health Administration in the USA showed the median dust exposure to be 1.20 mg/m3; 90% of all exposures were <2.78 mg/m3 (National Institute for Occupational Safety and Health, 1979). TALC 299 Before the adoption of technical preventive measures in 1950, exposures in the talc operation in the Germanasca and Chisone Valley (Piedmont), Italy, were reported to be approximately 800 mppcf [-28 250 particles/cm3] in the mines and 25 mppcf [-883 particles/cm3] in the mills. Exposures in both areas were reduced to less than 10 mppcf [-353 particles/cm3] after 1965 when improved ventilation techniques and wet drilling procedures were introduced. Mineralogical analyses of the footwall rocks demonstrated that they contained quartz, muscovite, chlorite, garnet, calcite, magnesite and small quantities of other minerals. In a few specimens o f footwall rocks, a small amount of tremolite was detected, but no other type of amphibole or chrysotile. Talc specimens from these mines were found very commonly to contain chlorite, but no amphibole or chrysotile minerals. The quartz content of powdered talc specimens was generally below the detection limits of X-ray diffraction (Rubino et al., 1976). In recent years, the mean exposure to respirable dust was 1.1 mg/m3 (range, 0.5-2.5 mg/m3), while the mean exposure to talc alone was 1.0 mg/m3 (range, 0.3--2.0 mg/m3). The authors stated that there was a remarkable difference in the amount of quartz in air dust in mines and mills and within jobs in the mine between drilling and other occupations. This was mainly due to the high content of quartz in footwall rocks, rather than to the absence of quartz particles in talc minerals (Coggiola et al., 2003). [The Working Group noted that the analytical methods were not described in detail and the mineral habit of the tremolite was not documented.] Wild et al. (1995) reported on a survey of the respiratory health of workers in a French talc producing factory. At this quarry, crude talc was extracted and transported directly to the mill using an overhead cable. The extracted ore consisted of a mixture of talc, chlorite, some dolomite (<3%), occasionally quartz (<3%) and traces of calcite, apatite, pyrite and mica. Amphiboles were not detected. A total of 1440 personal samples were taken between 1986 and 1991. The mean levels of exposure to respirable dust ranged from 0.5 mg/m3for secretaries, managerial staff and outdoor workers who handled the railway wagons to 15 mg/m3 for site cleaning staff. In 1991, only one exposure group o f four maintenance workers was estimated to have a mean exposure in excess of 5 mg/m3. However, the probability of exceeding an exposure level of 5 mg/m3 was more than 10% for most maintenance and some production workers. This was explained by the high variability of exposure among maintenance workers; eight of 10 groups of workers in the maintenance workshop had geometric standard deviations >3. Exposure was found to be more homogeneous among the production workers. The authors claimed that the introduction of centralized aspiration devices and new working procedures had resulted in lower levels of exposure. Mean levels of exposure in the in the past were estimated to have been up to 60 mg/m3, especially for workers storing jute bags of talc in wagons. Before 1985, the highest levels of exposure to dust for site cleaning staff were estimated to be 30 mg/m3; for sacking and drying, exposure levels in the workplace before 1975 were estimated to be 20 mg/m3. Wild et al. (2002) also provided some additional exposure information for three Austrian mines and their respective mills in the Styrian Alps. The ore mined at one site 300 IARC MONOGRAPHS VOLUME 93 (site B) consisted of a talc-chlorite mixture with gange [dead rock] inclusions of about 25% (mainly alumino-silicate rock). The gange was dumped in the mine so that the milled product was talc-chlorite and contained between 0.5 and 4% quartz. At site C, the material mined was a talc-dolomite aggregation with a medium talc content of 25%. The amount of quartz in the end-product was below 1%. However, materials from certain parts of this mine that were rich in dolomite could have contained 2--3% quartz. At site D, a light greyish quartz-chlorite-mica schist (alumino-silicate rock that consisted of an aggregation of more or less equal proportions o f mica, chlorite and quartz) was mined and milled. Analyses of dust from the lungs and lymph nodes of employees in the Austrian talc industry confirmed the presence of quartz and the absence of amphibole and serpentine (Friedrichs, 1987). Table 1.11 summarizes the levels o f exposure reported in the French and Austrian talc mines and associated mills. Table 1.11. Levels (mg/m3) of exposure to respirable dust in one French and two Austrian talc mines and associated mills Exposure group Occupation Mine/mill No. of samples No exposure Office workers French talc quarry 168 Low exposure Maintenance French talc quarry 100 (<5 mg/m3) workers, garage Austrian mine B 173 mechanics, Austrian mine C 33 production workers with dust control/LEV Median exposure (5-30 mg/m3) Recent production workers French mine A 193 Austrian mines B 17 and C High Milling, Austria 3 exposure maintenance, (>30 mg/m3) cleaning Mean Range 0.2 0.5-2.6 0.11-17 0.02--4.61 0.02--4.1 3.5-25.6 0.21-134 6.5-19.6 73-159 Date 1986 1986 1988-92 1991-92 NR NR End of 1980s From Wild et al. (1995, 2002) LEV, local exhaust ventilation; NR. not reported Several samples were collected from a crushing, grinding and talcum powder packing unit at a plant in Pakistan to measure different particle sizes (Jehan, 1984). In total, seven 1-hour samples were collected, one for total suspended particles (concentration, 6.14 mg/m3), one for particulate matter (PM) <10 pm (1.12 mg/m3), one for PM <7 pm (1.93 mg/m3), one for PM <5 pm (0.40 mg/m3), one for PM <3 pm (0.26 mg/m3), one for TALC 301 PM <2 pm (0.05 mg/m3) and one for PM >1 pm (1.55 mg/m3). Further analyses of the samples with PM <10 pm and <2 pm by scanning electron microscopy showed that the fibre concentration was 0.25 fibres/cm3 and 0.12 fibres/cm3, respectively. Analyses by polarized light microscopy indicated the presence of asbestiform tremolite, chrysotile and anthophyllite in these samples. (b) User industries Only limited information is available on exposures in secondary industries in which talc is used or processed further. Results from some surveys are summarized in Table 1.12. Table 1.12. Mineral composition of talc used for dusting in the rubber industry in the USA Reference Location Date Mineral composition Method of analysis Hogue & Mallette (1949) Dement & Shuler (1972) Fine et al. (1976) Vermont Canton, MA Vermont 1943-48 1972 1972-74 Stated to be 'pure talc' 2-3% quartz Trace of quartz (<1%), <2 fibres/cm3 Impinger Gravimetric, optical fibre counts Gravimetric Personal air samples collected in a rubber band production plant, where housekeeping, ventilation and work practices were poor and talc was used as an anti sticking agent, had time-weighted average (TWA) concentrations of respirable dust of 2.5-7.8 mg/m3 (average, 4.8 mg/m3) for extruders, 5.3 and 6.1 mg/m3 for vulcanizers and 0.9 and 1.3 mg/m3 for cutters. Exposures to total dust were found to range from 5.4 to 199 mg/m3. The talc was reported to contain 2-3% quartz. Within these exposures, 4.7 19.2 fibres were >5 pm/cm3 as measured by phase-contrast optical microscopy (Dement & Shuler, 1972). [The Working Group noted that no electron microscopic analysis was conducted to confirm the identity of the fibres; however, most of these were probably not asbestos.] Concentrations of respirable dust in two rubber manufacturing plants where Vermont talc was used as an anti-sticking agent are shown in Table 1.13. Eighteen of 21 samples analysed for quartz contained less than 1% by weight. In 12 samples analysed for fibres, using phase-contrast microscopic techniques for asbestos, all concentrations were less than 2 fibres/cm3. No electron microscopic fibre analysis was reported (Fine et al., 1976). Hogue and Mallette (1949) found an average dust concentration of 15-50 mppcf [--530-- 1765 particles/cm3] talc in two rubber plants that used Vermont talc. Average exposures were 20 mppcf [-706 particles/cm3] for tube machine operators, 35 mppcf 302 IARC MONOGRAPHS VOLUME 93 [1236 particles/cm3] for tube `bookers', 15 mppcf [-530 particles/cm3] for tube cure men and 50 mppcf [-1765 particles/cm3] for `line rerollers'. Table 1.13. Concentrations of respirable dust in rubber processing plants that used talc Location No. of samples Average dust concentration (mg/m3) Plant A Lorry and bus inner tubes (splicer) 7 0.60 Lorry and bus inner tubes (cureman) 6 1.41 `Tuber operator' 3 0.47 `Booker' 3 0.74 Farm service inner tubes (splicer) 6 0.82 Farm service inner tubes (cureman) 2 0.91 Plant B Rubber band area Gum engraving room Hose extruding Curing heavy duty flaps `Dust room' 6 3.55 6 0.64 4 0.51 3 1.29 2 0.59 From Fine et al. (1976) In a mortality study of lung cancer and respiratory disease among pottery workers exposed to silica and talc, Thomas and Stewart (1987) estimated exposure to nonasbestiform talc and tremolitic talc. Exposure to talc occurred almost exclusively in the cast shop. Montana steatite talc that had been used to dust moulds since 1955 appeared to contain no asbestiform talc (Gamble et al., 1982; Grexa & Parmentier, 1979). However, before 1955, flint and ground clay had been used to dust the moulds. Up to 1976, tremolitic talc had been used in some glazes. No measurements of airborne talc or silica were available, and exposure estimates were based on detailed knowledge of industrial processes and job duties. All exposures to talc were associated with high exposure to quartz from the clays. Quartz particles from clay are smaller than approximately 4 pm. Kauppinen et al. (1997) developed an international database of exposure measurements in the pulp, paper and paper product industries. In total, 63 measurements for talc were included in this database-- four measurements in the pulp production and 59 in paper or paperboard production and recycling; 6% of the samples exceeded the 8-hour TWA threshold limit value (TLV) for talc o f 2 mg/m3 respirable dust (ACGIH Worldwide, 2005). [No information was provided on the methods of measurement, the time period when these measurements were taken or the actual processes and the materials used during these measurements. As only a limited number of measurements were available, it is improbable that these results are representative of exposure to talc in this industry.] TALC 303 Kauppinen et al. (2002) described the prevalence of exposure to talc among workers in the on-machine coating of paper. In total, 25 departments were assessed: in 60% of the departments, more than 5% of the workers were exposed to talc, with a median prevalence of exposure of 51-90%. The median level of exposure was assessed as medium (0.6-2 mg/m3) by a team of occupational hygienists. Pooley and Rowlands (1975) examined talc imported into the United Kingdom. These talcs were used in a variety of industries, including cosmetics. Only one of the samples examined contained tremolite (>30%). [The number of samples examined and their use were not given. The electron micrograph of the sample identified as tremolite and the concentration of tremolite are consistent with the Gouvemeur District New York State talc, which is unlikely to have been used in cosmetics.] All other elongated particles detected in the samples were identified as laths or rolled sheets of talc, chlorite or sepiolite (several samples). 1.3.3 Consumer exposure (a) Mineralogical characterization Two studies that were conducted between 1968 and 1977 examined the mineralogy of consumer talc in the USA. Cralley et al. (1968) examined 22 cosmetic talc products that were purchased off the shelf for particles >5 pm with a 3:1 or greater aspect ratio (diameter:length) and found that on average 19% of the particles met these dimensional criteria. [No additional information was provided on the source of the talc products, but the Working Group noted that the authors were located in Cincinnati, OH, USA.] The authors concluded that these `fibres' were predominantly talc, but suggested that some may have been anthophyllite, tremolite, pyrophyllite or chrysotile. [The Working Group noted that no data were provided to support this statement. The statement was based only on the fact that these minerals have been reported to occur in some talc deposits.] Using X-ray diffraction, quartz was found at a level of 0.2-53.4% in these samples. No limit of detection was given, but the lowest concentration reported was 0.2 wt%. Analysis for other minerals was not carried out. Rohl et al. (1976) examined 20 body powders, baby powders and facial talcums and one pharmaceutical talc, all of which were purchased at retail stores in New York City between 1971 and 1975. Based on X-ray diffraction, optical microscopy and transmission electron microscopy, the concentration o f tremolite, anthophyllite and quartz was estimated and the presence of several other minerals was established (see Tables 1.14 and 1.15). One of the 21 samples was composed entirely of cornstarch and one contained primarily pyrophyllite and only a small amount of talc. Quartz was present in nine of the 21 samples, tremolite was reported in nine, anthophyllite in seven and serpentine in two samples. Chrysotile was confirmed by transmission electron microscopy in these samples, but no estimates of the concentrations were provided. Krause (1977), in a review of this study, pointed out that the overlap of the X-ray diffraction patterns of tremolite and 304 IARC MONOGRAPHS VOLUME 93 anthophyllite makes accurate estimation of their concentration by this method impossible. A similar problem was pointed out for estimates of the concentration of quartz because of overlap with several talc peaks. [The Working Group believed that these criticisms were reasonable and that little reliance can be placed on the reported concentration of tremolite or anthophyllite. The Working Group also noted that Rohl et al. (1976) stated that their methodology did not distinguish between asbestos and non-asbestiform mineral fragments. In addition, the representativeness of these samples for other countries or for other areas o f the USA is unclear.] Table 1.14. Concentrations of minerals in 20 samples of body powders, baby powders and facial talcums and one sample of pharmaceutical talc Mineral No. of samples Concentration range (wt%) Quartz 9 Tremolite3 9 Anthophyllite3 7 Chrysotile 2 1 6-35.1 0.1-10.3 2.1-11.4 <0.5b From Rohl et al. (1976) 3 Six samples contained both minerals, which resulted in uncertainty about the absolute concentrations given for each mineral. b Visual estimates by transmission electron microscopy were given as 0.25-- 0.5%, but no methodology was provided. Table 1.15. Qualitative measurements of minerals other than anthophyllite, chrysotile, quartz or tremolite in 20 samples of body powders, baby powders and facial talcums and one sample of pharmaceutical talc Mineral No. of samples in which the mineral was present Talc 20" Chlorite 16b Calcite 8b Phlogopite 3b Pyrophillite 2b Dolomite T Kaolin lb From Rohl et al. (1976) a Talc was the major mineral in 19 of the 20 samples. b Present in quantities above trace amounts TALC 305 Paoletti et al. (1984) examined talc powders that were used in pharmaceutical and cosmetic preparations. Tremolite was identified in two of six cosmetic talcs on the Italian market. Six of 14 samples provide by the European Pharmacopoeia contained either tremolite, anthophyllite or chrysotile. [No information was provided on the concentration of minerals, including tremolite and quartz, or on the time of purchase.] Jehan (1984) reported on commercial cosmetic-grade talc (baby and body talcum powder) used in Pakistan between 2000 and 2004. Sixty samples were analysed using atomic absorption techniques, X-ray diffraction, polarized light microscopy and scanning electron microscopy, and the presence of asbestiform chrysotile, both asbestiform and non-asbestiform tremolite and anthophyllite was identified. Asbestiform varieties of tremolite and anthophyllite were uncommon, while chrysotile was common. Respirable quartz was also identified in most (80%) of the samples. Some products listed by the Cosmetic and Toiletries Formulations Database are shown in Table 1.7. Listing is voluntary and may not be representative of products that are on the market. Tables 1.16 and 1.17 present the average mineral composition of commercial products that were sold under the name of talc in North America and Europe, respectively, in the late 1980s. (b) Use o f talcfo r feminine hygiene The use of body powder for feminine hygiene can be estimated from the prevalence reported for controls in case-control studies that investigated the association between the use of cosmetic talc for feminine hygiene and the risk for ovarian cancer. The prevalence of ever use in these studies is summarized in Table 1.18. Higher prevalences were generally reported in studies from Canada, the United Kingdom and the USA (up to 59%), whereas the lowest prevalences were generally reported in studies conducted in other countries, including China, Greece and Israel (2.2-5.6%). Studies with high prevalences also reported doses in terms of frequency, duration of use, age at first use or cumulative doses. Frequency of use may vary from a few times per month to more than once a day, and a large proportion of use is more or less daily. Duration of use ranges up to more than 40 years. The cumulative exposure to talc by perineal dusting was over 10 000 days in 4% of the users in one study (Cook et al., 1997). The use of talcum powder for feminine hygiene is acquired in young adulthood, since 80% of women who use body powder start before the age of 25 years (Harlow & Weiss, 1989). The types of application also vary. Body powder can be applied perineally, on napkins or on underwear. Dusting of the perineum after bathing appears to be the most frequent single type of application, but simultaneous uses have also been reported. Alternatively, exposure may occur as a result of storing a diaphram in body powder or contamination from the male partner who has used body powder. One study in the USA reported that the use o f deodorant spray had a prevalence of 24% (Cook et al., 1997). In several of the studies in Table 1.18, the interviews on powder use occurred before 1988. O f these, all but one were conducted in the USA. Information on the composition O ON IARC MONOGRAPHS VOLUME 93 Table 1.16. Average mineralogical composition (%) of commercial products sold under the name of talc in North America Canada Vermont Talc production 40 floated 10 30 70 30 floated 200 (thousand tonnes) 12 floated California 10 Texas 307 Montana 326 20 New York 140 M ineral (%) Talc 92 5 64 5 60 5 55 90 52 5 94 5 54 Chlorite 3 11 5 105 7 7 9 1 5 5 Dolomite 1 4 8 2 0 5 2 0 5 9 Magnesite 1.5 17 18 34 2 33.5 0.5 16 Serpentine T Quartz Mica T T Calcite T Tremolite Anthophyllite 80 94 8 25 1 45 85 5 12 5 0 5 05 T T 25 T T T T T T 44 5 From Ferret & Moreau (1990) T, identified mineral that could not be measured by the methods of analysis used Table 1.18. Assessment of exposure to body powders in the perineal area by women Location No of controls Prevalence of Type of perineal use of powder by women ever use of talc Massachusetts, USA Washington DC, USA California. USA United Kingdom Washington. USA Massachusetts. USA China Maryland, USA Athens, Greece Israel Toronto, Canada Washington, USA New York, USA Montreal, Canada New England, USA New York, USA Delaware Valley, USA California, USA USA 215 171 539 451 158 239 224 46 193 408 564 422 50 170 523 693 1367 1122 78 630 cohort 28 4% 1 8% 45 8% 59 0% 40 5% 39 3% 2 2% 17 3% 3 6% 5 6% 35 6% 39 3% 26% 4 7% 18 2% 35% 40% 37 1% 40.4% Exposure to talc by dusting Body talc Use of talcum powder Use of talc Exposure to powder (cornstarch, baby powder, talc, deodorizing powder); detailed information on type of powder used Exposure to baby powder, deodorizing or scented powder Dusting powder Genital bath talc (also asked use on napkins or diaphragm) Local application of talc Use o f talc Regular application o f talc Dusting with cornstarch, talcum powder, baby or scented powder, and deodorizing spray Use of talc Use of talc Use of talc, baby or deodorizing powders or cornstarch Use of talc (on genital or thigh area and sanitary napkins) Use of talc (on gemtal/rectal area and feet, sanitary napkins, underwear, diaphragm/cervical cap, male partner user) Use of talcum powder Use o f talc Reference Cramer et a/ (1982) Hartge e t a l (1983) Whittemore et al (1988) Booth el al. (1989) Harlow & Weiss (1989) Harlow el al (1992) Chen etal. (1992) Rosenblatt et al ( 1992) Tzonou et al. (1993) Shushan et al (1996) Chang & Risch (1997) Cook e ta l (1997) Eltabbakh e ta l (1998) Godard e ta l (1998) Cramer et al (1999) Wong et al (1999) Ness e ta l (2000) Mills et a! (2004) Gertig el a! (2000) IARC MONOGRAPHS VOLUME 93 308 TALC 309 of baby powder, body powder, facial powder and pharmaceutical talcum powder on the market in New York City before 1976 suggests that many of these products were impure and contained anthophyllite, carbonate, chlorite, chrysotile, phlogopite, pyrophyllite quartz and tremolite (Cralley et al, 1968; Rohl et al, 1976). After 1976, these powdere probably did not contain anthophyllite, chrysotile or tremolite but may have contained up to 10% of other minerals including carbonate, chlorite and quartz (Grexa & Pannentier, 1979). In 1994, baby talcum powder available in the USA typically contained 99% talc- body powder typically contained 65--70% talc and the remaining material was cornstarch, sodium bicarbonate and fragrance (Zazenski et al., 1995). ' (c) Other uses o f cosmetic talc Russell et al. (1979) and Aylott et al. (1979) reported exposure to respirable dust during the use of talcum powders on the face, body and babies. Russell et al. (1979) took 48 measurements during baby dusting operations and 44 measurements during the application of powders to adult bodies. Adult exposure was assessed during normal face/body powdering practices by placing cyclone samplers on shelves at an appropriate height or by positioning a cyclone attached to a headband near the nose (i.e. in the breathing zone). Exposure to respirable dust was 2.031.48 mg/m3 during adult application and was estimated to be 0.19 mg/m3 for babies. The estimated duration of the application was 1.23 minute for adults and 0.52 minute for babies. Aylott et al. (1979) measured levels of exposure to respirable dust during the application of loose face powder (24 measurements), adult dusting powder (43 measurements) and baby dusting powder (32 measurements). In the study of baby dusting powder, a doll was used. The exposure to respirable dust during face powdering ranged from <0.1 to 1.7 mg/m (duration, 10--25 seconds), that for adult dusting powder ranged from 0.2 to 3.3 mg/m3 (duration, 15-80 seconds) and that for baby powders ranged from <0.1 to 0.9 mg/m3(duration, 15-60 seconds). (d) Other exposures Talc is used as a surface lubricant on the majority of condoms manufactured; contact with condoms may also represent a direct means of exposure of the female genital tract to talc (Kasper & Chandler, 1995). Exposure to talc can also occur during surgical procedures when using powdered gloves. Talc particles were observed in the navels o f small children, in the testes, on the vocal cords, in the urinary bladder tract and after removal of varicous veins (Ramelet, 1991; Simek et al., 1992). During breast implantations, it is possible that talc from surgical gloves can lead to unwanted encapsulation (Chandler & Kasper, 2003). 1.3.4 Environmental exposure Talc is often detected as a common anthropogenic contaminant in suspended sediment, even in remote snowfields in the Alps; this has been ascribed to its emission 310 IARC MONOGRAPHS VOLUME 93 into the atmosphere by industrial and agricultural process (Hillier, 2001). Talc had also been identified in the sediment of the River Don in Scotland (United Kingdom), although no obvious industrial or agricultural sources of the talc were apparent (Hillier, 2001). 1.4 Regulations and guidelines Occupational exposure regulations and guidelines for talc in several countries are presented in Table 1.19. Table 1.19. Occupational exposure standards and guidelines for talc Country or region Concentration (mg/m3) Interpretation Carcinogenicity' Australia Belgium China Canada Alberta British Columbia Ontario Quebec Czech Republic Denmark Finland Germany Hong Kong Ireland Japan Malaysia Mexico Netherlands New Zealand Norway 2.5 10(1) 2 3 CD 4 2 (R) 2 (R) 2 fibres/cm3 (R) 3(R) 10 (R) 2(R) 10 (T) 0.3 fibres/cm3 5 (R) 2 (R) 10(1) 0.8 (R) 0.5 (R) 2 CD 2(R) 2 (R) 1 (R) 2(R) 2 (R) 6 (T) TWA TWA TWA TWA STEL TWA TWA TWA; value is for particulate matter containing <1% crystalline silica TWA (talc-containing no mineral or asbestos fibres) TWA; fibres >5% TWA; fibres <5% TWA TWA; containing fibres K TWA MAK; without asbestos fibres 3B TWA A4 TWA TWA TWA TWA TWA TWA A4 TWA TWA TWA TWA TALC 311 Table 1.19 (contd) Country or region Concentration (mg/nT) Interpretation Carcinogenicity Poland South Africa Spain Switzerland United Kingdom USA ACGIH (TLV) NIOSH (REL) OSHA (PEL) 1 (R) 4(1) 1 (R) 10(1) 2(R) 2 1 (R) 2 (R) 2(R) ~3 (20 mppcf) TWA TWA TWA TWA Ceiling; containing no asbestos fibres and <1% crystalline silica TWA TWA TWA; containing no asbestos and <1% crystalline silica TWA(lO-h) TWA; containing <1% quartz From Direktoratet for Arbejdstilsynet (2002); Tysuojelusdksi (2002); SUVA (2003); ACGIH Worldwide (2005); Deutsche Forschungsgemeinschaft (2005); Health and Safety Executive (2005) ACGIH, American Conference of Governmental Industrial Hygienists; I, inhalable dust; MAK. maximum concentration in the workplace; mppcf, millions of particles per cubic foot; NIOSH, National Institute for Occupational Safety and Health; OSHA, Occupational Safety and Health Administration; PEL, permissible exposure limit; R, respirable dust; REL, recommended exposure limit; T, total dust; STEL, short-term exposure limit; TWA, 8-h time-weighted average (unless otherwise specified) "3B, substances for which in-vitro test, or animal studies have yielded evidence of carcinogenic effects that is not sufficient for classification of the substance in one of the other categories; K, included in the list of substances considered as carcinogenic; A4, not classifiable as a human carcinogen The Food and Drug Administration regulates talc in the USA, and states that it is generally recognized as safe for use in colour additives in foods, drugs and cosmetics, and in paper, paper products, cotton and cotton fabrics that come into contact with food. The Food and Drug Administration also states that talc is present in over-the-counter astringent drug products (National Toxicology Program, 2000). The Food Chemical Codex (2003) provides specifications for food-grade talc, including the statement that "talc derived from deposits that are known to contain associated asbestos is not food grade." Under the voluntary guidelines initiated in 1976, the Cosmetic, Toiletry, and Fragrances Association stated that all cosmetic talc should contain at least 90% platy talc (hydrated magnesium silicate) that is free from detectable amounts (0 .5 % ) of fibrous, asbestos minerals (Gilbertson, 1995; Zazenski et al., 1995; National Toxicology Program, 2000). 312 IARC MONOGRAPHS VOLUME 93 The current Occupational Safety and Health Administration (2005) permissible exposure level for non-asbestiform talc in the USA is ~3 mg/m3 (20 mppcf) measured as respirable dust. The current American Conference of Governmental Industrial Hygienists TLV-TWA is 2 mg/m3 (15 mppcf), which also is the proposed Occupational Safety and Health Administration limit. Levels of exposure of workers may exceed three times the TLV-TWA for no more than 30 minute during the workday (National Toxicology Program, 2000). 1.5 References ACGIH Worldwide (2005). 2005 Documentation of the TLVs and BEIs with Other Worldwide Occupational Exposure Values, Cincinnati, OH [CD-ROM] Aylott RI, Byrne GA, Middleton JD, Roberts ME (1979). Normal use levels of respirable cosmetic talc: preliminary study. Int J Cosmet Sei, 1:177-186. doi: 10.1111/j. 1467-2494.1979.tb00212.x. PMID: 19467066 Bish DL, Guthrie GD (1993). Mineralogy of clay and zeolite dusts (exclusive of 1:1 layer silicates in health effects of mineral dusts. In: Guthrie GD, Mossman BT, eds, Reviews in Mineralogy, Vol. 28, Chelsea, Ml, Mineralogical Society of America, Book Crafters, pp. 263 Booth M, Beral V, Smith P (1989). Risk factors for ovarian cancer: a case-control study. Br J Cancer, 60:592-598. PMID:2679848 Boundy MG, Gold K, Martin KP Jr e t al. (1979). Occupational exposure to non-asbestform talc in Vermont, ln: Lernen R, Dement JM, eds, Dusts and Disease, Park Forest South, 1L, Pathotox, pp. 365-378. Campbell WJ, Huggins CW, Wylie AG (1980). Chemical and Physical Characterization of Amosite, Chrysotile, Crocidolite, and Nonfibrous Tremolite for Oral Ingestion Studies by the National Institute of Environmental Health Sciences (Report of Investigations 8452), Washington DC, Department of the Interior, Bureau of Mines. Chandler PJ Jr, Kasper CS (2003). Frequency and distribution of talc contamination in patients with silicone gel-filled breast implants. Ann Plast Surg, 51:358-360. doi:10.1097/01 .sap.0000070642.91783.95. PMID: 14520061 Chang S, Risch HA (1997). Perineal talc exposure and risk of ovarian carcinoma. Cancer, 79:2396 2401. doi:10.1002/(SICI)1097-0142(19970615)79:12<2396::AID-CNCR15>3.0.CO,2-M. PMID:9191529 Chen Y, Wu PC, Lang JH e t al. (1992). Risk factors for epithelial ovarian cancer in Beijing, China. Int J Epidemiol, 21:23-29. doi:10.1093/ije/21.1.23. PMID: 1544753 Chidester AH, Engel AEJ, Wright LA (1964). Talc Resources of the United States (Geological Survey Bulletin 1167), Washington DC, US Government Printing Office, pp. 1-61. Coggiola M, Bosio D, Pira E e t al. (2003). An update of a mortality study of talc miners and millers in Italy. Am J Ind Med, 44:63-69. doi: 10.1002/ajim. 10240. PMID: 12822137 Cook LS, Kamb ML, Weiss NS (1997). Perineal powder exposure and the risk of ovarian cancer. Am J Epidemiol, 145:459-465. PMID:9048520 Cralley LJ, Key MM, Groth DH et al. (1968). Fibrous and mineral content of cosmetic talcum products. Am Ind Hyg Assoc J, 29:350-354. PMID:4300288 TALC 313 Cramer DW, Liberman RF, Titus-Ernstoff L e t al. ( 1999). Genital talc exposure and risk of ovarian cancer. Int J Cancer, 81:351-356. doi: 10.1002/(SICI) 1097-0215(19990505)81:3<351::AIDIJC7>3.0.CO;2-M. PM1D: 10209948 Cramer DW, Welch WR, Scully RE, Wojciechowski CA (1982). Ovarian cancer and talc: a casecontrol study. Cancer, 50:372-376. doi: 10.1002/1097-0142(19820715)50:2<372 ::AIDCNCR2820500235>3.0.CO;2-S. PMID:7083145 Deer WA, Howie RA, Zussman J (1962). Talc. In: Rock Forming Minerals, Vol. 3, Sheet Silicates, New York, John Wiley & Sons, pp. 121-130. Dement J, Shuler P (1972). Talc Dust and Industrial Hygiene Survey, Plymouth Rubber Company, Canton, MA (Report No. IWS-036.11A), Cincinnati, OH, National Institute for Occupational Safety and Health. Deutsche Forschungsgemeinschaft (2005). List of MAK and BAT Values 2005 (Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area Report No. 41), Weinheim, WILEY-VCH GmbH & Co., pp. 109,138 Direktoratet for Arbejdstilsynet (2002). WEA-Guide 2002-Limit Values for Substances and Materials, Copenhagen, p. 55. Dreessen WC (1933). Effects of certain silicate dusts on the lungs. J Ind Hyg, 15:66-78. Dreessen WC, DallaValle JM (1935). The effects of exposure to dust in two Georgia talc mills and mines. Public Health Rep, 50:131--143. PMID: 19315489 Dresler CM, Olak J, Herndon JE II e t air, Cooperative Groups Cancer and Leukemia Group B; Eastern Cooperative Oncology Group; North Central Cooperative Oncology Group; Radiation Therapy Oncology Group (2005). Phase III intergroup study of talc poudrage vs talc slurry sclerosis for malignant pleural effusion. Chest, 127:909-915. doi:10.1378/chest.l27.3.909. PMID: 15764775 Eltabbakh GH, Piver MS, Natarajan N, Mettlin CJ (1998). Epidemiologic differences between women with extraovarian primary peritoneal carcinoma and women with epithelial ovarian cancer. Obstet Gynecol, 91:254-259. doi:10.1016/S0029-7844(97)00650-9. PMID:9469285 EUROTALC (2005) Welcome to EUROTALC The Scientific Association of Talc Producers. Available at: www.ima-eu.org/eurotalc.htlm. Ferret J, Moreau P (1990). Mineralogy of talc deposits. In: Bignon J, ed, Health Related Effects of Phyllosilicates (NATO ASI Series, Vol. G21), Berlin, Springer-Verlag, pp. 147-158. Fine LJ, Peters JM, Burgess WA, Di Berardinis LJ (1976). Studies of respiratory morbidity in rubber workers. Part IV. Respiratory morbidity in talc workers. Arch Environ Health, 31:195 200. PMID:942261 Flick EW (2005). Cosmetics and Toiletries Formulations Database, William Andrew Publishing [CD-ROM], from Knovel Library. Available at: http://knovel.com. Food Chemical Codex (2003). Talc, Washington DC, National Academic Press. Friedrichs KH (1987). Electron microscopic analyses of dust from the lungs and the lymph nodes of talc-mine employees. Am Ind Hyg Assoc J, 48:626-633. PMID:3618475 Gamble J, Greife A, Hancock J (1982). An epidemiological-industrial hygiene study of talc workers. Ann Occup Hyg, 26:841--859. doi:10.1093/annhyg/26.8.841. PMID:7181311 Genuine M (1987). Sepiolite asbestos from Franklin, New Jersey: a case study in medical geology. Environ Res, 42:386-399. doi: 10.1016/S0013-9351 (87)80205-0. PMID:2952495 Gertig DM, Hunter DJ, Cramer DW e t al. (2000). Prospective study of talc use and ovarian cancer. JNatl Cancer Inst, 92:249-252. doi: 10.1093/jnci/92.3.249. PMID:10655442 314 IARC MONOGRAPHS VOLUME 93 Gilbertson WE (1995). The regulatory status of talc. Regul Toxicol Pharmacol, 21:230-232. doi: 10.1006/rtph. 1995.1033. PM1D:7644710 Godard B, Foulkes WD, Provencher D e t al. (1998). Risk factors for familial and sporadic ovarian cancer among French Canadians: a case-control study. Am J Obstet Gynecol, 179:403--410. doi: 10.1016/S0002-9378(98)70372-2. PM1D:9731846 Greenwood WS (1998). A Mineralogical Analysis of Fibrous Talc, MS Thesis, College Park, MD, University of Maryland. Greife A (1980). Preliminary findings of epidemiologic study of talc workers (industrial hygiene portion). In: Kraybill HF, Blackwood IC, Freas NB, eds, Proceedings of the First NCI/EPA/NIOSH Collaborative Workshop: Progress on Joint Environmental and Occupational Cancer Studies, Morgantown, WV, National Institute for Occupational Safety and Health, pp. 229-240. Grexa RW, Parmentier CJ (1979). Cosmetic talc properties and specifications. Cosmet Toilet, 84:29-33. Gruner JW (1934). The crystal structure of talc and pyrophyllite. Zeit Krist, 88:412. Harben PW, Kuzvart M (1996). Talc and soapstone. In: Harben PW, Kuzvart M, eds, Industrial Minerals: A Global Geology, London, Industrial Minerals Information Ltd, Metal Bulletin PLC, pp. 407-417. Harlow BL, Cramer DW, Bell DA, Welch WR (1992). Perineal exposure to talc and ovarian cancer risk. Obstet Gynecol, 80:19-26. PMID: 1603491 Harlow BL, Weiss NS (1989). A case-control study of borderline ovarian tumors: the influence of perineal exposure to talc. Am J Epidemiol, 130:390-394. PMID:2750733 Hartge P, Hoover R, Lesher LP, McGowan L (1983). Talc and ovarian cancer [Letter to the editor]. J Am Med Assoc, 250:1844. doi:10.1001/jama.250.14.1844. PMID:6620481 Health and Safety Executive (1995). Asbestos Fibres in Air (Methods for the Determination of Hazardous Substances 39/4), London, Her Majesty's Stationery Office. Health and Safety Executive (2005). Workplace Exposure Limits Containing the List of Workplace Exposure Limits for Use with the Control of Substances to Health Regulations 2002 (as amended) (EH40/2005), London, Her Majesty's Stationery Office, p. 23. Hendricks SB (1938). On the crystal structure o f talc and pyrophyllite. Zeit Krist, 99:264. Hillier S (2001). Particulate composition and origin of suspended sediment in the R. Don, Aberdeenshire, UK. Sci Total Environ, 265:281-293. doi:10.1016/S0048-9697(00)00664-1. PMID: 11227272 Hogue WL Jr, Mallette FS (1949). A study of workers exposed to talc and other dusting compounds in the rubber industry. J Ind Hyg, 31:359-364. IARC (1977). IARC monographs on the evaluation of the carcinogenic risk of chemicals to man: asbestos. IARC Monogr Eval Carcinog Risk Chem Man, 14:1-106. PMID:863456 Industrial Minerals Association-Europe (2005). Fact Sheet: Talc, Brussels. Jehan N (1984). Sustainable Management of Mineral Resources with Special Reference to Asbestos and Silica in northern Pakistan, PhD Thesis, Peshawar, National Centre of Excellence in Geology, University o f Peshawar. Jurinski JB, Rimstidt JD (2001). Biodurability of talc. Am Mineral, 86:392-399. Kasper CS, Chandler PJ Jr (1995). Possible morbidity in women from talc on condoms. J Am Med Assoc, 273:846-847. doi:10.1001/jama.273.11.846. PMID:7869551 TALC 315 Kauppinen T, Teschke K, Astrakianakis G e t al. (2002). Assessment of exposure in an international study on cancer risk among pulp, paper, and paper product workers. Am Ind Hyg Assoc J, 63:254-261. Kauppinen T, Teschke K, Savela A e t al. (1997). International data base of exposure measurements in the pulp, paper and paper product industries. Int Arch Occup Environ Health, 70:119-127. doi: 10.1007/s004200050195. PMID:9253640 Krause JB (1977). Mineralogical characterization of cosmetic talc products. J Toxicol Environ Health, 2:1223-1226. doi:10.1080/15287397709529521. PMID:864791 Leake BE, Woolley AR, Arps CES e t al. (1997). Nomenclature of amphiboles: report of the subcommittee on amphiboles of the International Mineralogical Association, Commission on new minerals and mineral names. Am Mineral, 82:1019-1037. Luzenac (2004). Talc for the World, available at: www.Iuzenac.com. Mills PK, Riordan DG, Cress RD, Young HA (2004). Perineal talc exposure and epithelial ovarian cancer risk in the Central Valley of California. Int J Cancer, 112:458- 464. doi: 10.1002/ijc.20434. PMID: 15382072 Mondo Minerals (2005). Technical Data Sheets National Institute for Occupational Safety and Health (1979). Mining Surveillance: Potentially Toxic Occupational Exposures, Morgantown, WV. National Toxicology Program (2000) Draft Background Document on Talc, Research Triangle Park, NC. Ness RB, Grisso JA, Cottreau C e t al. (2000). Factors related to inflammation of the ovarian epithelium and risk of ovarian cancer. Epidemiology, 11:111--117. doi :10.1097/00001648200003000-00006. PMID:11021606 Occupational Safety and Health Administration (2005). Detailed Procedure for Asbestos. Sampling and Analysis-Nonmandatory-1915-1001 App B (Standards 29CFR) Oestenstad K, Honda Y, Delzell E, Brill I (2002). Assessment of historical exposures to talc at a mining and milling facility. Ann Occup Hyg, 46:587-596. doi:10.1093/annhyg/mefD76. PMID: 12270883 Paoletti L, Caiazza S, Donelli G, Pocchiari F (1984). Evaluation by electron microscopy techniques of asbestos contamination in industrial, cosmetic, and pharmaceutical talcs. Regul Toxicol Pharmacol, 4:222-235. doi:10.1016/0273-2300(84)90022-9. PMID:6494497 Pence FK (1955). Commercially proven white firing talc occurring in West Texas. Bull Am Ceram Soc, 34:122-1235. Petit S (2005). Crystal-chemistry of talcs: a NIR and MIR spectroscopic approach. In: Kloproggs JT, ed, The Application of Vibrational Spectroscopy to Clay Minerals and Layered Double Hydroxides (CMS Workshop Lectures Vol. 13), Aurora, CO, The Clay Mineral Society, pp. 41-64. Piniazkiewicz RJ, McCarthy EF, Genco NA (1994). Talc. In: Carr DD, ed, Industrial Minerals and Rocks, Littleton, CO, Society for Mining, Metallurgy and Exploration, pp. 1049-1069. Pooley FD, Rowlands N (1975). Chemical and physical properties of British talc powders. Inhaled Part, 4:639-646. PMID: 1236242 Ramelet AA (1991). [An unusual complication of ambulatory phlebectomy. Talc granuloma], Phlebologie, 44:865-871 (in French). PMID:1805258 Rayner JH, Brown GT (1973). The crystal structure of talc. Clays Clay Miner, 21:103-114. doi: 10.1346/CCMN. 1973.0210206. 316 1ARC MONOGRAPHS VOLUME 93 Rohl AN, Langer AM, Selikoff 1J et al. (1976). Consumer talcums and powders: mineral and chemical characterization. J Toxicol Environ Health, 2:255-284. doi: 10.1080/15287397609529432. PMID: 1011287 Rosenblatt KA, Szklo M, Rosenshein NB (1992). Mineral fiber exposure and the development of ovarian cancer. Gynecol Oncol, 45:20-25. doi: 10.1016/0090-8258(92)90485-2. PMID: 1601331 Roskill Information Services Ltd (2003). The Economics of Talc and Pyrophyllite, 9th Ed., London, pp. 102-110. Ross M, Smith W, Ashton W (1968). Triclinic talc and associated amphiboles from Gouverneur mining district, New York. Am Mineral, 751: 10. R.T. Vanderbilt Company (2000) Material Safety Data Sheet: NYTAL 100, Norwalk, CT. Rubino GF, Scansetti G, Piolatto G, Romano CA (1976). Mortality study of talc miners and millers. J Occup Med, 18:187-193. doi: 10.1097/00043764-197603000-00013. PMID:1255280 Russell RS, Merz RD, Sherman WT, Sivertson JN (1979). The determination of respirable particles in talcum powder. Food Cosmet Toxicol, 17:117-122. doi: 10.1016/0015-6264(79)90208-6. PMID:478394 Sanchez-Soto PJ, Wiewiora A, Aviles MA e t al. (1997). Talc from Puebla de Lillo, Spain. II. Effect of dry grinding on particle size and shape. Appl Clay Sei, 12:297-312. doi: 10.1016/SO169-1317(97)00013-6. Selevan SG, Dement JM, Wagoner JK, Froines JR (1979). Mortality patterns among miners and millers of non-asbestiform talc: preliminary report. In: Lernen R, Dement JM, eds, Dusts and Diseases, Park Forest South, IL, Pathotox, pp. 379-388. Shushan A, Paltiel O, Iscovich J e t al. (1996). Human menopausal gonadotropin and the risk of epithelial ovarian cancer. Fertil Steril, 65:13-18. PMID:8557128 Sim^ek F, Trkeri L, liker Y e t al. (1992). Severe obstruction of the urinary tract due to talcum powder granuloma after surgery. A case report. Int Urol Nephrol, 24:31-34. doi: 10.1007/BF02552114. PMID: 1624242 Stemple IS, Brindley GW (1960). A structural study of talc and talc-tremolite relations. J Am Ceram Soc, 43:34-42. doi: 10.1111 /j. 1151-2916.1960.tb09149.x. SUVA (2003). [Limit Values in the Workplace 2003] (in German) Grenzwerte am Arbeitsplatz 2003, Luzern, Switzerland, p. 101. Thomas TL, Stewart PA (1987). Mortality from lung cancer and respiratory disease among pottery workers exposed to silica and talc. Am J Epidemiol, 125:35-43. PMID:3024482 Tysuojelusdksi (2002). HTP-Arvot 2002, Sosiaali- Ja Terveysministeri, Kemian tysuojeluneuvottelukunta, Tampere, Kirjapaino hrling, p. 22. Tzonou A, Polychronopoulou A, Hsieh CC e t al. (1993). Hair dyes, analgesics, tranquilizers and perineal talc application as risk factors for ovarian cancer. Int J Cancer, 55:408-410. doi: 10.1002/ijc.2910550313. PMID:8375924 Van Gosen BS, Lowers HA, Sutley SJ, Gent CA (2004). Using the geologic setting of talc deposits as an indicator of amphibole asbestos content. Environ Geol, 45:920-939. doi: 10.1007/s00254-003-0955-2. Virta RL (2004). Talc and pyrophyllite. In: US Geological Survey Minerals Yearbook, Reston, VA, pp. 75.1-75.3. TALC 317 Whittemore AS, Wu ML, Paffenbarger RS Jr e t al. (1988). Personal and environmental characteristics related to epithelial ovarian cancer. II. Exposures to talcum powder, tobacco, alcohol, and coffee. Am J Epidemiol, 128:1228-1240. PMID:3195564 ' Wild P, Refregier M, Auburtin G e t al. (1995). Survey of the respiratory health of the workers of a talc producing factory. Occup Environ Med, 52:470-477. doi:10.1136/oem 52 7 470 PMID:7670622 Wild P, Leodolter K, Refregier M et al. (2002). A cohort mortality and nested case-control study of French and Austrian talc workers. Occup Environ Med, 59:98-105. doi: 10.1136/oem.59.2.98. PMID:11850552 ' Wong C, Hempling RE, Piver MS et al. (1999). Perineal talc exposure and subsequent epithelial ovarian cancer: a case-control study. Obstet Gynecol, 93:372-376. doi: 10.1016/S0029- 7844(98)00439-6. PMID: 10074982 Wylie AG, Skinner HCW, Marsh J e t al. (1997). Mineralogical features associated with cytotoxic and proliferative effects of fibrous talc and asbestos on rodent tracheal epithelial and pleural mesothelial cells. Toxicol Appl Pharmacol, 147:143-150. doi: 10.1006/taap. 1997.8276. PMID:9356317 Yekeler M, Ulusoy U, Hicyilmaz C (2004). Effect of particle shape and roughness of talc mineral ground by different mills on the wetability and floatability. Powder Technol, 140:68-78. doi:10.1016/j.powtec.2003.12.012. Zazenski R, Ashton WH, Briggs D e t al. (1995). Talc: occurrence, characterization, and consumer applications. Regul Toxicol Pharmacol, 21:218-229. doi: 10.1006/rtph. 1995.1032. PMID:7644709 Zbik M, Smart R (2005). Influence of dry grinding on talc and kaolinite morphology. Miner Eng, 18:969-976. doi: 10.1016/j.mineng.2005.01.005. 318 IARC MONOGRAPHS VOLUME 93 2. Studies of Cancer in Humans 2.1 Occupational exposure 2.1.1 Talc miners and millers (Table 2.1) Rubino et al. (1976) conducted a study o f mortality among men who had begun work in the mines and mills of a talc operation in the Germanasca and Chisone valleys (Piedmont), Italy, between 1921 and 1950 and who had been employed for at least 1 year in a job that involved exposure to talc. A total o f 1514 miners and 478 millers were identified, of whom 168 miners (11.1%) and 40 millers (8.4%) were lost to follow-up before the end of the study in June 1974, yielding a combined cohort of 1784 men (89.6%) for analysis. The talc from these mines was described as pure and was reported to have been used in the pharmaceutical and cosmetics industries. However, due to the presence of `footwall contact rocks' and rock-type inclusions in the mines, drilling operations were associated with exposure to dusts that contained high levels of silica; such inclusions were removed before milling and talc products were reported to have a content o f free silica below 2%. [The Working Group understood that the term `silica' was in fact quartz.] In a few instances, talc samples from the area showed small amounts o f tremolite when examined by X-ray diffraction, but no amphibolic asbestos or chrysotile were detected. For each worker, cumulative exposure was estimated from regular measurements o f respirable dust content in the air o f mines and mills during the period 1948-74 and individual work histories were abstracted from files of the mining company. Periods of time during which the dust level was assumed to be uniform were first selected and cumulative exposure was then calculated as the summed product of the number of years in each specific working period (years) and the associated dust levels (million particles per cubic foot; mppcf), resulting in an overall measure of mppcf-years. Once individual cumulative exposures had been assigned, miners and millers were then classified separately into low, medium and high levels of exposure. Ranges of exposure (mppcf-years) for miners were 566--1699, 1700--5665 and 5666--12750, respectively; ranges o f exposure for millers were 25-141, 142^124 and 425-906, respectively. For each of the 1784 workers included (1346 miners and 438 millers), one unexposed control subject was chosen at random from among male inhabitants o f a nearby small, rural town. The control was matched to the talc worker on year of birth and vital status at date of entry into the study [date not specified]. Cause of death for 885 (95.1%) o f 931 deceased workers and 1067 (94.8%) of 1126 deceased controls was obtained from regional death certificate files supplemented with information from relatives, physicians and medical records. Observed numbers of deaths among talc workers were compared with expected numbers, calculated by the use of age-specific mortality rates experienced by the control cohort. The standardized mortality ratio (SMR) for all causes combined was 0.9 (95% TALC Table 2.1. Cohort studies of mortality from and incidence of cancer in populations occupationally exposed to non-asbestiform talc Reference, location Coiiort description Exposure assessment Organ site Exposure categories No. of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments Rubino et al (1976), Gerinanesca and Chisone valleys (Piedmont), Italy 1992 male talc workers (1514 miners, 478 millers) employed >1 year in talcexposed job during 19211974, hired 19211950, mortality follow-up, 192174, vital status. 90%, cause of death 95% ot exposed workers, 95% ot controls Occupational history from plant records, respirable dust measurements, 19481974, quantitative estimation ot cumulative exposure for individual workers, expressed as summed product of duration (years) and exposure (million particles per cubic foot, mppcf), classification of workers into 3 levels of exposure All cancers All miners All millers Miners (mppcf--years) Level 1 566-1699 Level 2' 1700-5665 Level 3 5666-12750 Millers (mppcf-years) Level 1 25-141 Level 2 142-424 Level 3 425-906 Lung, bronchus and trachea All miners All millers Miners (mppcf-years) Level 1 566-1699 Level 2 1700-5665 Level 3 5666-12750 Millers (mppcf-years) Level 1 25-141 Level 2 142-424 Level 3 425-906 SMR 100 0.8 (0.6-0.9) Adjusted for age. 42 0 9 (0 7-1 2) comparison with unexposed, age-matched controls from neighbouring rural town, 38 1 2 (0 8-1 6) controls matched on vital 28 1 0 (0.7-1 4) status at date o f entry into 34 0 9 (0 6-1 2) study, miners and millers exposed 18 1 1 (0 2-3 2) to a very pure form of talc, 13 I 3 (0-2.9) miners also exposed to 11 0 7 (0.4-2 7) inhalable silica. significantly elevated SMRs 9 0 5 (0.2-0 9) for silicosis with and 4 0 6 (0.2-1 6) without tuberculosis among miners; estimates increased with increasing cumulative 3 1.1 (0.6-1.7) exposure, no observed cases 1 0 5 (0 7-2 3) o f mesothelioma, 5 1 1 (0 4-1.3) no smoking data for exposed workers or 3 1.7(0 3-4 9) unexposed controls 1 1 25 (0-7 0) 0 -- 0-> "s O Table 2.1 (contd) Reference, location Cohort description Exposure assessment Organ site Exposure categories Rubino et al (1979), Germanesca and elusone valleys (Piedmont), Italy 1678 male talc workers (1260 miners, 418 millers). mortality followup, 1946-74 Same exposure categories as Rubino et al (1976) Lung Selevan et al (.1979), Vermont, USA 392 white male talc workers (163 miners, 225 millers) employed > 1 year between 1940 and 1969, mortality follow-up' date of first radiogram, 12-month employment anniversary or January 1940, whichever was later, follow-up through 1975, vital status 99%, cause of death 94% Historical insufficient information to calculate cumulative exposure histones, cohort classified into two work areas- mining and milling. All causes All cancers Respiratory cancer All miners All millers Miners (mppef-years) Level 1 566-1699 Level 2. 1700-5665 Level 3' 5666-12750 Millers (mppef--years) Level 1 25-141 Level 2- 142-424 Level 3 425-906 Total cohort Millers Miners Total cohort Millers Miners Total cohort Millers Miners OJ t-o o No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments SMR Re-analysis of cohort 8 0.5 (0 2-0.9) reported in Rubmo etal. 4 0 7 (0 2-1 7) (1976), SMRs recalculated using national death rates instead o f comparison with > 2 0 5 (0-1 9) neighbouring rural 73 n 1 0.2 (0 5-1 2) population, national death s. 5 0.6 (0 2-1 4) rates available only from 1951 onward; rates for 1951 O zo 3 2 0 (0.4-5 8) were applied for 1946-50 1 0.7 (1.7-3 7) a 73 0 > -o SMR Adjusted for age, sex, race, 32 o 90 1 2 [0.9-1 4] calendar year. US death 44 1 2 [0.9-1 6] rates. 1940-67, linear < 34 1.3 [0.9-1.8] extrapolation for all causes t-1 of death 1967-69 C 16 [1 3 (0 .7 -2 0)] Vermont death rates for cs 5 [0.8 (0.3-1 9)] specific causes of death' SO 7 [1.7 (0 7-3.5)] 1949-75, workers selected from annual radiographic 6 [1.6(0 6-3 5)] survey of dusty trades, no 2 [1.0(0 1-3 7)] data on smoking habits for 5 [4 3 (1 4-10 1)] millers or miners; exposure to radon daughters in mine, radiographic evidence of pneumoconiosis in most workers who died from non- malignant respiratory disease Table 2.1 (contd) Reference, location Cohort description Exposure assessment Organ site Exposure categories Wergeland et al (1990), northern and western Norway 389 male talcexposed workers (94 miners, 295 millers) employed > 1year in mine (1944-- 72) or >2 years in mill (1935-72), mortality and cancer incidence follow-up, 1953-- 87 Subjective assessment of exposure by experienced colleagues; workers classified by total duration of employment in jobs with low', medium, high and unknown exposure All causes All cancers All cancers Total cohort Miners Millers Total cohort Miners Millers Total cohort Miners Millers Years employed 1-4 5-19 >20 Years since first employment 1-19 20-29 >30 TALC No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments SMR Adjusted for age, smoking 117 0 8 (0.6-0 9) (miners only), national 27 [0.8 (0.5-1 2)] death rates. 1953-87, main 90 [0 7 (0 6-0.9)] minerals m mined talc deposit were talc and 26 0.8(0 5-1 1) magnesite, 90% of raw 9 [1.3 (0 6-2 5)] material for mill from mine, 17 [0 6 (0 4-1 0)] 10% from India, no information on smoking SIR habits for millers, smoking 46 0.9 (0 7-1 2) habits for miners above 15 [1.4 (0 8-2 3)] national average, low levels 31 [0.8 (0 5-1 1)] of exposure to radon daughters 11 [1 1 (0 6-2 1)] 19 [0 8 (0 5-1 2)] 16 [0.9 (0 5-1 5)] 6 [0 4 (0 2-0 9)] 18 [1 1 (0 7-1 8)] 22 [1 1 (0 7-1 6)] LO NJ Table 2.1 (contd) Reference, location Cohort description W ergelande/ al (1990) (contd) Exposure assessment Organ site Exposure categories Lung Stomach Total cohort Miners Millers Years employed 1-4 5-19 >20 Years since first employment 1-19 20-29 >30 Total cohort Miners Millers Years employed 1-4 5-19 >20 Years since first employment 1-19 20-29 >30 wIO IO No of cases/ deaths Relative risk (95% Cl) Adjustment factors; comments 6 0 9 (0 3-2 0) --* 2 [1.6 (0 2-5 7)] > 4 [0.8 (0 2-2 0)] n 0 - 2 3 [1 0 (0 2-3 0)] oz 3 [1.0 (0 2-3 0)] a73 > 2 [1 1 (0 1 -f 1)] ~0 1 [0 5 (1 3-2 8)] mX 3 [1 1 (0 2-3 2)] O< 6 1 1 (0 4-2 2) Cr~ 3 [2 5 (0 5-7 4)] 2 3 [0 7 (0 1-2 1)] mSO 2 [2 0 (0 2-7 2)] OJ 2 [0 8 (0 1-2 6)] 2 [1 2 (0 1^(3)] 1 [0.6 (1.4-3.1)] 2 [1.1 (0.1-4 0)] 3 [1.7 (0 3-4 8)] fable 2 .1 (contd) Reference, location Cohort description Exposure assessment Organ site Wild (2000). Luzenac, France I 160 talc workers (1070 men, 90 women) actively employed in 1945 or hired during 1945-94 and employed >1 year, mortality follow-up,, 194596, vital status' 97%; cause of death 74% pre1968 and 98% post-1968 Exposures assessed for case-control study, sennquantitative, site-specific job-exposure matrix based on personal dust measuiements (1986 onwards) and subjective assessments by experienced workers, workers assigned to four categories of exposure no exposure, ambient (<5 mg/m'), medium (530 mg/m') and high (>30 mg/m'), exposure prior to hiring also coded none, probable exposure to quartz, certain exposure to quartz, exposure to other carcinogens All causes All cancers Lung Stomach Exposure categories No of cases/ deaths Relative risk (95% Cl) Male (ale workers Pre-1968 (national rates) 101 Post-1968 (national rates) 294 Post-1968 (regional rates) 294 Post-1968 (regional rates) 80 Post-1968 (regional rates) 21 Post-1968 (national rates) 21 Men <60 years of age 7 Latency period <20 years 5 Duration of employment 8 <10 years Post-1968 (national rates) 5 SMR 0 8 (0 6-1 0) 0.8 (0 7-0 9) 0 9 (0 8-1.0) 1.0 (0 8-1 3) 1 2 (0 8-1 9) 0.9 (0 6-1 4) 2 0 [0 8-4.0] 2 4 [0 8-5.6] 2 1 [0 9-4 1] 1 2 (0 4 -2 8) Adjustment factors, comments Adjusted for age, sex, smoking, prior exposure to quartz (case-control study only), partial overlap of study population with Leophonte et al (1983) and Leophonte and Didier (1990). extent of overlap unknown, national mortality rates applied, pre- and post1968, regional mortality rates applied post-1968 excess mortality from lung cancer disappeared when national rates applied UN>J OJ Table 2.1 (contd) Reference, location Cohort description Exposure assessment Organ site Wild (2000) (contd) Wild et al (2002). Luzenac, France ( 1 site), and Styrian Alps. Austria (4 sites) Nested casecontrol study lung cancer, nonmalignant pulmonary disease and stomach cancer; three randomly selected controls per case, lung cancer' 23 cases, 67 controls Austrian cohort 542 male talc workers employed > I year during 1972-95, mortality followup, 1972-1995, vital status 97%, French cohort as described under Wild (2000) Cumulative exposure estimates (mg/m'-years) for individual workers Austrian cohort semiquantitative, site-specific job-exposure matrix based on personal dust measurements (1988-92) and descriptions of workplaces from management and long term workers, workers assigned to four categories of exposure no exposure, ambient (<5 mg/m1), medium (5-30 mg/m') and high (>30 mg/m'), other exposures coded quartz, other carcinogens, underground work Lung All causes All cancers Lung Stomach Exposure categories Unexposed <100 mg/m'-years 100-400 mg/m1--years 400-800 mg/m '-years >800 mg/m'-years Per 100 mg/m'-years French cohort Austrian cohort French cohort Austrian cohort French cohort Austrian cohort French cohort Austrian cohort UKJ) -U- No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments Odds ratio Unadjusted odds ratio, no 6 1 0 increasing trend with increasing 5 1 4 cumulative exposure, 6 22 information on smoking habits 3 0 7 3 09 available for 52% o f cases and 75% o f controls >P3 Assumes a linear trend n 23 1 0 (0 9-1.1) 2 oz oo 73 SMR Adjusted for age, calendar year, >-o 294 0 9 (0 8-1 0) smoking, exposure to quartz. Xcn o 67 0 8 (0 6-1 0) exposure to other carcinogens, < underground work (case- n 80 1 0 (0 8-1 3) control stud)'), study population C 17 0 7 (0.4-1 2) overlaps with that of Wild (2000), French SMRs calculated 2m 21 1 2 (0.8-1.9) by comparison with regional sOoJ 7 1 1 (0 4-2 2) rates, 1968-95, Austrian SMRs calculated by comparison with 5 1 2 (0 4 -2 8) regional rates, 1972-1995. 1 0 4 (0 -2 3) Austrian smoking information obtained from unpublished mortality studies on pneumoconiosis, from colleagues, from workers' compensation records, no missing information on smoking habits in Austrian cohort Table 2.1 (contd) Reference, location Cohort description Exposure assessment Organ site Wild et al (2002) (contd) Coggiola et al (2003), Piedmont, Italy Nested casecontrol study lung cancer, nonmalignant respiratory disease, three randomly selected controls per case, lung cancer 23 cases, 67 controls (France), 7 cases, 21 controls (Austria) Cohort of 1974 male talc workers employed >1 year in mine or mill during 1946-95, mortallty follow up, 1946-95, loss to follow-up, 9%, analysis based on 1244 miners, 55 1 millers Cumulative exposure estimates (mg/m'-years) assigned to individual workers by occupational physician using work histories abstracted from company records Lung Detailed job histories from plant records. workers classified on basis ofjob held (miner versus miller), duration of exposure (years) and time since first exposure (years) All causes All cancers Lung cancer Exposure categories Unexposed <100 mg/m'-years 101-400 m g/m '-years 401-800 mg/m3-years >801 mg/m3-years Per 100 mg/m'-years No of cases/ deaths 9 6 7 5 3 30 Relative risk (95% Cl) Odds ratio 1 0 09 1 1 06 07 1 0 (0 9-1 1) Adjustment factors, comments Unadjusted odds ratio, no trend observed with increasing cumulative exposure, trend not affected by adjusting for smoking, quartz exposure, underground work or by lagging the exposure estimate Assumes a linear trend Total cohort 880 Miners 590 Millers 290 Total cohort 185 Miners 130 Millers 55 Total cohort 44 Miners 33 Millers 11 Years since fir st exposure <20 6 20-30 10 >30 28 SMR 1 2(1 1-1.3) 1.3 (1 2-1.4) 1 1 (1 0 -1 2) 1 0 (0 9-1 1) 1 1 (1 0-1.3) 0.9 (0.6-1 1) 0 9 (0 7-1.3) 1.1 (0 7-1.5) 0.7(0 3-1.2) 1 1 (0 4-2.3) 1 0 (0 5-1 8) 0.9 (0.6-1 3) H > C* O Adjusted for age, calendar period, study population overlaps with that of Rubino et al (1976, 1979), national death rates used for pre-1970 period, rates for early 1950s used for 1946-49, regional rates used for 1970-95, except for cancers of oral cavity, oesophagus and suicide (regional rates unavailable, national rates used ), no information on smoking habits, no variation in lung cancer by duration of exposure UtOoSl NJ On IARC MONOGRAPHS VOLUME 93 Table 2.1 (contd) Reference, location Cohort description Exposure assessment Organ site Exposure categories No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments Coggiola et al (contd) Oral cavity Oesophagus Stomach Total cohort Miners Millers Total cohort Miners Millers Total cohort Miners Millers SMR 31 5 1 (3.5-7 3) 24 6 2 (3 .9 -9 1) 7 3 3 (1.3-6 9) 10 2.1 (1.1-3.9) 7 2.3 (0.9-4 8) 3 1 8 (0.4-5.2) 31 1.2 (0.8-1.6) 20 1 2 (0.7-1 8) 11 1 1 (0.5--2.0) Cl, confidence interval, mppcf, million parts per cubic foot, SIR, standardized incidence ratio, SMR, standardized mortality ratio TALC 327 confidence interval (Cl), 0.8-1.0) for miners and 0.9 (95% Cl, 0.8-1.0) for millers. No relationship was observed with increasing time between first exposure and death or with increasing cumulative exposure. Significant increases in specific cause of death among miners were found for silicosis (62 observed; SMR, 2.0; (95% Cl, 1.5--2.6) and for silicosis with superimposed tuberculosis (18 observed; SMR, 2.0; 95% Cl, 1.2-3.1). These estimates were found to increase with increasing cumulative exposure. A total of 100 deaths from cancers at all sites combined among miners (SMR, 0.8; 95% Cl, 0.6-0.9) and 42 deaths among millers (SMR, 0.9; 95% Cl, 0.7-1.2) were below those expected. Nine deaths among miners (SMR, 0.5; 95% Cl, 0.2-0.9) and four among millers (SMR, 0.6; 95% Cl, 0.2-1.6) were due to lung cancer. No excess risk for lung cancer was found in the highest exposure category among miners (cumulative exposure range, 5666 12750 mppcf-years; five observed; SMR, 1.1; 95% Cl, 0.4-2.7) or millers (cumulative exposure range, 425-906 mppcf-years; no observed deaths versus 1.3 expected). No cases of mesothelioma were found. [The Working Group noted that the lack of comparability between the workers and the comparison groups could influence the mortality ratio estimates of this study.] In a re-analysis of their 1976 study, Rubino et al. (1979) estimated relative mortality among talc workers using Italian national death rates for men instead of the control cohort. As national rates were available only for the period 1951-74 (end of the study), rates for 1951 were applied for the follow-up period 1946 through to 1950. The number of workers included in this analysis was 1260 miners and 418 millers. In contrast to the previous analysis, the age-standardized mortality for all causes combined was significantly increased for miners (560 observed; SMR, 1.3; 95% Cl, 1.2-1.4) as well as for millers (193 observed; SMR, 1.2; 95% Cl, 1.0-1.4). Eight observed cases of lung cancer in miners yielded an SMR of 0.5 (95% Cl, 0.2-0.9) and four cases in millers yielded an SMR of 0.7 (95% Cl, 0.2-1.7). No trend was obseived with increasing cumulative exposure for either group of workers [p-value for trend not provided]. Mortality from non-malignant respiratory diseases was significantly increased among miners (109 observed; SMR, 3.3; 95% Cl, 2.7-4.0), mainly due to 58 cases of pneumoconiosis and 23 cases of tuberculosis. The number of cases of pneumoconiosis and tuberculosis among millers was three and eight, respectively. Katsnelson and Mokronosova (1979) conducted a study of mortality among male and female workers [numbers not specified] in a talc mining and processing plant in the former USSR in 1949-75. The talc of the area was reported to contain no tremolite or fibrous materials and levels of quartz ranged from 0.2 to 1.6%. Very high mortality ratios were found for cancer at all sites combined (relative risks, 5.1 for men; 6.4 for women; P< 0.001) as well as for lung (relative risks, 4.5 for men; P< 0.02; 9.3 for women; / >>0.05) and stomach cancer (relative risks, 3.7 for men; jP < 0.02; 6.3 for women; P < 0.05) [observed numbers of deaths not specified], [The Working Group noted that the deaths observed among exposed workers included current and past workers but that the denominator comprised only currently employed persons.] 328 IARC MONOGRAPHS VOLUME 93 Selevan et al. (1979) used radiography records from the annual surveys of workers in dusty trades of the Vermont Health Department to identify all white male workers employed in the Vermont talc industry for at least 1 year between 1940 and 1969. The study covered three areas that had a total of five companies (two of which ceased operations in 1952 and 1960). The talc in this region is a mixture of pure talc, magnesite, chlorite and dolomite. Airborne dust samples and bulk materials were free o f asbestiform minerals, when examined by both X-ray diffraction and analytical electron microscopy. Levels of respirable crystalline silica were below 0.25% in nearly all ore and product samples, and free silica was only occasionally detectable in air samples. Insufficient information was available to estimate cumulative lifetime exposures, but the authors stated that historical data were sufficient to demonstrate past exposure levels for miners and millers far exceeded the standard for non-fibrous talc o f 20 mppcf that was in force at the time of the investigation. Due to the more continuous nature of the milling operation, it was considered probable that exposures to dust for millers were higher than those for miners. In one mine that had closed by the time of the study, `cobblestones' of highly tremolitic serpentine rock were present but were avoided or discarded as far as possible before milling. Miners were also exposed to radon daughters at mean levels ranging up to 0.12 working levels (WL), with single peaks of 1.0 WL. The study groups comprised 163 talc miners and 225 millers. Vital status of workers was ascertained through to 1975, and death certificates were obtained for 85 of 90 deceased cohort members. For non malignant respiratory disease and respiratoiy cancer, mortality rates for white men from Vermont were used for comparison, because they were considered to be more appropriate than national rates. For other causes of death, rates for the USA were used. Some increase was noted for all malignant neoplasms combined (16 observed [SMR, 1.3; 95% Cl, 0.7 2.0]) and specifically for respiratory cancer (six observed [SMR, 1.6; 95% Cl, 0.6-3.5]). [The Working Group noted that the results for respiratory cancer were not analysed by latency.] The excess mortality from respiratory cancer was statistically significant among the miners (five observed [SMR, 4.3; 95% Cl, 1.4--10.1]), but not among the millers (two observed [SMR, 1.0; 95% Cl, 0.1-3.7]). A significant excess o f mortality from non malignant respiratory disease was seen in millers (seven observed [SMR, 4.1; 95% Cl, 1.6-8.4]), but not in miners (two observed [SMR, 1.6; 95% Cl, 0.2-5.9]). Most workers who died from non-malignant respiratory disease had radiographic evidence of pneumoconiosis (rounded opacities). In two brief communications, Leophonte et al. (1983) and Leophonte and Didier (1990) reported on the mortality o f workers employed in a talc quarry in Luzenac in the South of France and in the associated talc processing plant. The cohort was composed of those who left employment between 1945 and 1981 and who had worked at the plant for more than 1 year. The talc in this region is a mixture of pure talc, chlorite and dolomite with no asbestos; levels of quartz vary from 0.5 to 3%. O f 470 workers available for study, 256 were alive, 209 had died and five were lost to follow-up. Of 204 workers with a known job history and date o f death, 192 had worked exclusively with talc at Luzenac. No significant excess of mortality from cancer in general or specifically from respiratoiy TALC 329 and digestive cancers was found. [Observed and expected numbers of cause-specific deaths and associated relative risks were not given.] A significant increase in mortality was found for non-mahgnant respiratoiy disease, especially for pneumoconiosis and obstructive lung disease. No cases of mesothelioma were observed. [The Working Group noted the unconventional definition of the cohort and that causes of death were obtained differently for cases (from local doctors, hospitals or families) and controls (from regional or national records).] Wergeland et al. (1990) studied 94 male workers at a talc mine in northern Norway who had been employed in talc-exposed jobs for at least 1 year during 1944-72 and 295 male workers at a talc mill in western Norway who had been employed for at least 2 years during 1935-72. Data on miners were gathered from the company pay rolls, lists of union memberships and the central registry o f workers exposed to silica in Norway; data on millers were collected from the company protocol and the local occupational health service. The information included name, date o f birth, first and last date of employment and number of periods o f employment. According to the authors, Norwegian talc contains only trace quantities of quartz, tremolite and anthophyllite as determined by optical microscopy and by electron microscopic analysis. The talc in the region where the mine was located is composed mainly of pure talc and magnesite. Approximately 90% of the raw material in the mill came from the mine and the rest was imported from India. In addition to talc, dolomite and mica were also processed at the mill. Personal air samples collected in the early 1980s showed that total dust levels varied greatly by job categoiy and workplace (mine, 0.9-97 mg/m3; mill, 1.4-54 mg/m3). Peak exposures occurred during drilling in the mine (319 mg/m3) and in the store house in the mill (109 mg/m3). X-Ray diffractometry indicated that dust samples from both operations contained less than 1% quartz. The mean value for concentrations o f radon daughters in the mine was 3.5 pCi/L [0.04 WL], with a range o f 1.5-7.5 pCi/L [0.02-0.08 WL], The majority of the 389 workers could be classified into one of three categories according to degree of dust exposure, based on measurements and qualified assessments o f dust level by experienced co-workers. Information on tobacco smoking habits, gathered during the study in 1981, was available for 63 of the 94 miners and showed that smoking rates among these workeis were above the national average. Follow-up for cancer incidence (through data linkage to the national cancer registry) and cause-specific mortality (through linkage to the national mortality files) was begun at the date of entry into the cohort or 1 Januaiy 1953, whichever came later, and ended at date of death or 31 December 1987, whichever came first. National rates were used to calculate expected numbers of cancers and deaths. The SMR for all causes for the total cohort was 0.8 (117 observed; 95% Cl, 0.6-0.9), which reflected a decrease among both miners (27 observed [SMR, 0.8; 95% Cl 0 5 1.2]) and millers (90 observed [SMR, 0.7; 95% Cl; 0.6-0.9]). An excess of deaths from all cancers was observed in miners (nine observed [SMR, 1.3; 95% Cl, 0.6-2.5]), but not in either the total cohort (26 observed [SMR, 0.8; 95% Cl; 0.5-1.1]) or in millers (17 observed; [SMR, 0.6; 95% Cl; 0.4-1.0]). Mortality from non-malignant respiratoiy diseases was decreased, with one observed death among miners [SMR, 0.4; 95% Cl, 0 - 330 IARC MONOGRAPHS VOLUME 93 2.2] and two observed deaths among millers [SMR, 0.2; 95% Cl, 0-0.9], No deaths from pneumoconiosis were reported. The standardized incidence ratio (SIR) for all types of cancer combined was [1.4 (15 observed; 95% Cl, 0.8-2.3)] among the miners and [0.8 (31 observed; 95% Cl, 0.5-1.1)] among the millers. Two cases of lung cancer were observed among miners [SIR, 1.6; 95% Cl, 0.2-5.7] and four cases among millers [SIR, 0.8, 95% Cl, 0.2-2.0]. The non-significant excess risk among the miners was confined to cancer o f the stomach (three observed [SIR, 2.5; 95% Cl, 0.5-7.4]) and cancer of the prostate (four observed [SIR, 2.0; 95% Cl, 0.6-5.2]). In the subgroup of 80 workers who belonged to the highest exposure category, a total of six cases of cancer were observed [SIR, 0.4; 95% Cl, 0.2-1.0], none o f which were cancer of the lung. There were no observed cases o f mesothelioma. Wild (2000) conducted a retrospective cohort mortality study, within a nested casecontrol study, at the same talc quarry and milling plant at Luzenac as that used by Leophonte et at. (1983) and Leophonte and Didier (1990). The cohort included employees who were active in 1945 or hired in the milling plant during the period 1945 94 and who had been employed continuously for at least 1 year. Employees, who were identified from the company files, comprised a total of 1070 men and 90 women. [The authors did not indicate the extent o f overlap of the study population with that investigated by Leophonte et al. (1983) and Leophonte and Didier (1990).] Dust levels in the 1960s and 1970s were generally high, ranging from below 5 mg/m3 to more than 30 mg/m . Average dust levels dropped to below 5 mg/m3 in the 1990s through process changes and installation o f engineering controls (e.g. installation of a central vacuum system). Overall mortality o f the cohort was evaluated from 1 January 1945 to 31 December 1996. Vital status was obtained from the local population register and national mortality files which also included information on cause o f death, in most cases for individuals who died after 1968. Overall, 32 (2.8%) employees were lost to follow-up! O f 106 individuals who died before 1968, cause of death was ascertained for 78 cases. SMRs were calculated using both regional mortality rates (pre- and post-1968) and national mortality rates (pre-1968). When regional mortality rates for 1968 and later were used, the SMR for all causes of death combined was 0.9 (294 observed; 95% Cl, 0.8-1 0) for men and 0.8 (11 observed; 95% Cl, 0.4-1.4) for women. Eighty men died from cancer at any she (SMR, 1.0; 95% Cl, 0.8-1.3) and 21 died from lung cancer specifically (SMR, 1.2, 95% Cl, 0.8-1.9). Mortality from lung cancer was non-significantly increased in subgroups of employees who were under 60 years of age (seven observed; SMR, 2.0 [95% Cl, 0.8-4.0]), had a latency period of less than 20 years (five observed; SMR^ 2.4 [95% Cl, 0.8-5.6]) or had a duration o f employment of less than 10 years (ewht observed, SMR, 2.1 [95% Cl, 0.9-4.1]). A slightly increased risk was seen for stomach cancer (five observed; SMR, 1.2; 95% Cl, 0.4-2.8). Twenty-six men died from nonmahgnant respiratoiy diseases (SMR, 1.1; 95% Cl, 0.7-1.6), three of which were pneumoconiosis (SMR, 5.6; 95% Cl, 1.1-16.2). When pre-1968 national reference rates were applied, the overall SMR for men was 0.8 (101 observed; 95% Cl, 0.6-1.0) and the excess mortality from lung cancer and non-malignant respiratory diseases disappeared. Of TALC 331 the 101 deaths observed during this period, one was caused by lung cancer (SMR, 0.3 [95% Cl, 0.7-1.5]) and five were caused by non-malignant respiratory diseases (SMR, 0.7 [95% Cl, 0.2-1.6]). A nested case-control study was performed to investigate further the risks for lung cancer, stomach cancer and non-malignant respiratory diseases in the men of the cohort. For the lung cancer case-control study, 67 controls were individually matched to the 22 cases by age and sex (approximately three controls per case). Information on job history at the plant and tobacco consumption was collected through interviews of subjects who were alive and/or from experienced co-workers. A semiquantitative site-specific job-exposure matrix for talc dust was established using dust levels measured from 1986 onwards and estimates o f levels before that year. Information on job history was then converted into estimates of cumulative exposure of the individual employees (expressed as mg/m3-years). Multiple logistic regression analysis with adjustment for tobacco smoking habits and exposure to quartz estimated the odds ratio for lung cancer to be 0.7 (three cases and 15 controls) and 0.9 (three cases and 10 controls) for employees with a cumulative exposure to talc dust of 400-800 mg/m3-years and more than 800 mg/m3-years, respectively, when compared with unexposed employees (six cases and 20 controls). [The Working Group noted that information on smoking habits was available for only 52% of cases and 75% of controls, and that no specific information was given on the proportion of subjects alive among cases and controls at the date of interview.] Wild et al. (2002) conducted a combined analysis of previously published cohort mortality studies among 1070 male employees at a talc quarry and milling plant in the south of France (Site A) (Wild, 2000) and 542 male employees at three talc mines and their respective mills in Austria (Sites B, C and D). The Austrian cohort comprised workers who had been employed for at least 1 year between 1 January 1972 and 31 December 1995. Complete work histories for the Austrian workers were abstracted from company registries and from the regional social insurance. Information on tobacco smoking habits was obtained from earlier unpublished studies of mortality and pneumoconiosis, from colleagues and from records of the compensation claim insurance. Talc from two of the three Austrian plants (Sites B and C) had a content of quartz that was less than 4%, while that of the third plant (Site D) had higher but unspecified levels. Vital status of workers was verified through to 1995, and cause of death for those who had died was obtained from national mortality files. Local mortality rates yielded an overall SMR for the Austrian cohort of 0.8 (67 observed; 95% Cl, 0.6-1.0;). A total of 17 deaths were due to cancer at any site (SMR, 0.7; 95% Cl, 0.4-1.2), seven of which were from cancer of the lung (SMR, 1.1; 95% Cl, 0.4--2.2). One death from stomach cancer (SMR, 0.4; 95% Cl, 0-2.3) and no deaths from mesothelioma (0.1 expected) occurred. On the basis of 23 lung cancer deaths observed in the French cohort in 1968-- 96 and seven in the Austrian cohort in 1972-95, a nested case-control study was conducted. A total of 88 control subjects were selected from the two cohorts, individually matched to cases on age, calendar period and company. All job tasks at the companies were categorized according to measured and estimated levels of talc dust into one of four 332 IARC MONOGRAPHS VOLUME 93 exposure groups (no exposure, < 5 mg/m3, 5-30 mg/m3 and > 30 mg/m3). Job histories of cases and controls were converted into cumulative exposure to talc dust by summing the products of duration and level of exposure for each of the tasks held by the subject (mg/m3-years). Subjects were also categorized according to tobacco smoking habits, exposure to quartz or a history of underground work on a yes/no basis. Information on smoking habits was available for approximately 50% of the cases and 75% of the controls in the French cohort and for 100% o f the Austrian cohort. When the no-exposure category was used as the standard (nine cases, 23 controls), the unadjusted odds ratios for lung cancer were as follows: 0.9 (exposure category, 1-100 mg/m3-years; six cases, 18 controls); 1.1 (exposure category, 101-400 mg/m3-years; seven cases, 15 controls), 0.6 (exposure category, 401-800 mg/m3-years; five cases, 21 controls) and 0.7 (exposure category, > 801 mg/m3-years; three cases, 10 controls). Assuming a linear trend, the odds ratio was 1.0 (95% Cl, 0.9-1.1) per unit o f 100 mg/m3-years. Adjustment for tobacco smoking, exposure to quartz or underground work or any two of these variables did not change the results. Coggiola et al. (2003) updated the cohort of Rubino et al. (1976, 1979) to include 1974 men who had worked for at least 1 year in the mine and/or in the factory during the period 1946-95. The mortality analysis included 1795 subjects (90.9% of the total cohort; 1244 miners and 551 millers), after excluding 179 workers who were lost to follow-up. No data on smoking habits were available. Follow-up began on 1 Januaiy 1946 or the date of first employment and ended at the date o f death or 31 December 1995, during which time a total of 880 deaths occurred. The expected number of deaths was calculated from national rates for 1950-69 and regional mortality rates for 1970 onwards (with the exception of cancers of the oral cavity and oesophagus for which regional rates were unavailable; national rates were therefore used). Rates for the early 1950s were applied for the period 1946-49. Total mortality among workers was higher than expected (880 observed; SMR, 1.2; 95% Cl, 1.1-1.3), mainly due to excess mortality from non malignant respiratory tract diseases among the subgroup of miners (105 observed; SMR, 3.1; 95% Cl, 2.5-3.7). O f the 105 deaths in this category, 58 were from silicosis. In the combined cohort of workers, there was no excess mortality for all cancers (185 observed; SMR, 1.0; 95% Cl, 0.9-1.1) or for lung cancer, in particular (44 observed; SMR, 0.9; 95% Cl, 0.7-1.3). No deaths from pleural or peritoneal mesothelioma were found. A significantly elevated risk was seen for cancers of the oral cavity (31 observed; SMR, 5.1; 95% Cl, 3.5-7.3) and the oesophagus (10 observed; SMR, 2.1; 95% Cl, 1.1-3.9). When the analysis was stratified by job, the SMR for lung cancer was 1.1 (33 observed; 95% Cl, 0.7-1.5) among miners and 0.7 (11 observed; 95% Cl, 0.3-1.2) among millers. The slight excess found among miners seemed to be due to a slightly elevated risk in workers with less than 20 years since first exposure (latency) (six observed; SMR, 1.1; 95% Cl, 0.4 2.3) compared to that o f workers with 20-30 years (10 observed; SMR, 1.0; 95% Cl, 0.5 1.8) and more than 30 years (28 observed; SMR, 0.9; 95% Cl, 0.6-1.3) since first exposure. There was no variation in lung cancer mortality by duration of exposure. Cancer of the oral cavity caused the death of 24 miners (SMR, 6.2; 95% Cl, 3.9-9.1) and TALC 333 seven millers (SMR, 3.3; 95% Cl, 1.3-6.9) and oesophageal caused the death o f seven miners (SMR, 2.3; 95% Cl, 0.9-4.8) and three millers (SMR, 1.8; 95% Cl, 0.4-5.2). Excess mortality was seen in miners for non-malignant respiratory tract diseases (105 observed; SMR, 3.1; 95% Cl, 2.5-3.7), non-malignant digestive tract diseases (50 observed; SMR, 1.4; 95% Cl, 1.0-1.8) and liver cirrhosis (37 observed; SMR, 1.8; 95% Cl, 1.3-2.5). An increased risk for liver cirrhosis was also observed in millers (18 observed; SMR, 1.7; 95% Cl, 1.0-2.7). Meta-analysis o f riskfo r lung cancer Wild (2006) performed a meta-analysis of lung cancer mortality among miners and millers from industries that produced non-asbestiform talc in Vermont, USA (Selevan et al, 1979), Norway (Wergeland et al., 1990), Italy (Coggiola et al., 2003), France (Wild, 2000) and Austria (Wild et al., 2002). The purpose of the analysis was to compute risk estimates separately for talc miners, who usually have some co-exposure to silica and/or radon daughters, and talc millers, who normally have no such co-exposure. Previously unpublished risk estimates for the subgroup of millers in the French and Austrian cohorts were used and additional information on smoking habits was obtained for Italian, French and Austrian workers. Data indicated that the prevalence of smoking was higher than that in the reference populations [figures not specified]. In the estimation of the overall risk for millers, data from all five countries were used, while only data from the USA, Norway and Italy were included in that for miners. Based on SMRs for lung cancer of 1.0 (USA; two cases; 95% Cl, 0.1-3.7), 0.7 (Italy; 11 cases; 95% Cl, 0.3-1.2), 1.2 (France; 21 cases; 95% Cl, 0.8-1.9), 0.7 (Austria, Site B; three cases; 95% Cl, 0.1-2.0) and 1.1 (Austria, Site C; one case; 95% Cl, 0-6.2) and an SIR o f 0.8 (Norway; four cases; 95% Cl, 0.2 2.0) for talc millers, a summary SMR of 0.92 (42 cases; 95% Cl, 0.7-1.3) was obtained. No heterogeneity between studies was detected. Similarly, based on mortality ratios for lung cancer o f 4.4 (USA; five cases; 95% Cl, 1.4-10.2) and 1.1 (Italy; 33 cases; 95% Cl, 0.7-1.5) and an incidence ratio of 1.6 (Norway; two cases; 95% Cl, 0.2-5.7) for talc miners, a summary SMR of 1.2 (40 cases; 95% Cl, 0.9-1.6) was found. Due to a significant heterogeneity o f the latter data set, a random effect estimate of the overall SMR was also calculated (40 cases; SMR, 1.9; 95% Cl, 0.7-5.1). 2.1.2 User industries (Table 2.2) Information on risk for cancer among workers exposed to talc is available from studies that were conducted in user industries. However, they are less informative than those conducted in talc miners and millers because the potential contamination of talc was not addressed. In addition, these studies provided no details about the type of talc used. (a) Manufacture o f ceramic plumbingfixtures Thomas and Stewart (1987) conducted a cohort mortality study of 2055 white men employed for at least 1 year between 1939 and 1966 at three plants o f a single company in OCOJ ^useHndu'stries StUd'eS of mortality from and incidence of cancer in workers occupationally exposed to non-asbestiform talc IARC MONOGRAPHS VOLUME 93 Reference, location Cohort description Exposure assessment Organ site Exposure categories M anufacture of ceramic plumbing fixtures Thomas & Stewart (1987), USA. 5 plants in 1 company 2055 white men employed >1 yearm 1939-66. mortality followup through to 1 Jan 1981, vital status, 96% Exposure to silica and talc assessed qualitatively by job title-department by industrial hygienist All causes Lung cancer Manufacture of pulp and paper Langseth & Andersen (1999), Norway, 10 paper nulls 4247 women employed >1 year, 1920-93, follow-up of cancer incidence, 1953-93 All cancers Ovarian cancer Exposure >3 years Age 25-35 years Ovarian cancer Total cohort Total cohort High silica High sihca+non-fibrous talc High silica+non-fibrous talc+fibrous talc High sihca+no talc Total cohort Paper mill workers No, of cases/ deaths Relative risk (95% Cl) SMR 587 0 9 [0 8-1 0] 52 1 4 [1 1-1 9] 44 1.8 [1.3-2.4] 21 2.5 (1 6-3.9] 5 1 7 [0 6-4.0] 18 1 4 [0 8-2 2] SIR 380 1.2 (1 1-1 3) 37 1.5 (1 1-1 2) 31 1 6 (1 1-2.3) 6 8 0 (2 9-17 4) 18 2 1 (1 3-3.4) Adjustment factors, comments Crystalline silica was the major exposure, also exposure to non-fibrous and fibrous talc Comparison with 5-year age-specific rates m Norwegian women, cancer incidence from National Cancer Registry Table 2.2 (contd) Reference, location Cohort description Exposure assessment Organ site Langseth & Kjaerheim (2004), Norway, IO paper nulls Nested casecontrol study in cohort of Langseth & Andersen (1999); 46 cases, 179 matched controls, 100% histologically confirmed Exposure to asbestos, talc and total dust from work histories, questionnaires by industrial hygienists/ senior employees and international database, personal use o f talc, 76% of cases, 57% of controls; personal interviews Rubber m anufacturing industries Blum et al (1979), USA, 2 rubber companies Nested casecontrol study; 100 cases, 4 controls per case, matched on age, race, sex, company, 196473 Exposure to polycyclic hydrocarbons, mtrosamines, carbon black, talc (high, moderate, low', none) from job histories Ovarian cancer Stomach cancer Exposure categories No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments Total dust Ever talc Ever asbestos Asbestos according to interview Odds ratio Parity, breastfeeding, 0.8 (0 4-1.7) tobacco smoking habits, 1.1 (0 6 -2 2) family history of breast or 2.0 (0.7-5 7) ovarian cancer, conditional 2.2 (0 5-9 1) logistic regression, odds ratios unchanged after adjustment for confounders H >r* n Company A High+moderate talc High talc No information on 27 2.4(1 4 -4 1)* composition or purity o f 13 13 (0 9-2 5)* talc, no increase in risk in Company B *90% Cl OJ CO Table 2.2 (contd) Reference, location Cohort description Exposure assessment Organ site Exposure categories Straifet n/ (1999), Germany, 5 rubber production plants 8933 male bluecollar workers hired after 1 Jan. 1950 and alive 1 Jan 1981, follow-up, 1 Jan 1981 to end of 1991 ; cause o f death known for 97% of 1521 deceased Work histones recontructed from cost centre codes Lung cancer Stomach cancer Straif el al (2000), Germany, 5 rubber production plants Same as that of Straif eta! (1999) Same as Straif el a1 (1999) plus semiquantitative cumulative exposure (low, medium, high) to asbestos, talc, nitrosamines, carbon black for 95% o f cohort Lung cancer Stomach cancer Laryngeal cancer High talc Medium talc High talc Medium talc High talc Medium talc Cl, confidence interval, SIR, standardized incidence ratio; SMR, standardized mortality ratio CO U-> OS No of cases/ deaths Relative risk (95% Cl) Adjustment factors, comments SMR SMRs calculated from 154 1 2 (1 0-1.4) national death rates > 44 1 2 (0.8-1.6) PO O ooz o >v X CO < o 21 1.9 (I 1-3 I) Unadjusted, reference low 41 1 1 (0 8-1 6) exposure to talc t-- c 11 4.3 (2 1-9.0) 12 1 2 (0 6-2 4) 2m 3 5.4 (1 1-27 0) SCOO 2 2 8 (0 5-16 7) TALC 337 the USA that manufactured ceramic plumbing fixtures. Crystalline silica was said to be the major occupational exposure of these workers, but, in some parts o f the plant, exposure to fibrous [tremolitic] and non-fibrous [tremolite-free] talc had also occurred. Vital status was ascertained for 96% of the cohort through to 1 January 1981 and observed numbers of deaths were compared with numbers expected from cause-specific mortality rates for white men in the USA. For each job title-department combination, exposure to silica and talc were qualitatively assessed by an experienced industrial hygienist. Silica exposure was categorized as none, low or high; high exposure to silica was further categorized on the basis o f no exposure to talc, exposure to fibrous talc and exposure to non-fibrous talc. The SMR for all causes combined was 0.9 (578 observed [95% Cl, 0.8-1.0]) and that for lung cancer was 1.4 (52 observed [95% Cl, 1.1-1.9]). The excess mortality from lung cancer was seen exclusively among workers who had been exposed to high levels o f silica dust (44 observed; SMR, 1.8 [95% Cl, 1.3-2.4]) and, to a greater extent, in the subgroup with additional exposure to non-fibrous talc (21 observed; SMR, 2.5 [95% Cl, 1.6-3.9]) than in subgroups with additional exposure to fibrous talc (five observed; SMR, 1.7 [95% Cl, 0.6-4.0]) or no exposure to talc (18 observed; SMR, 1.4 [95 /o Cl, 0.8-2.2]). [The Working Group noted that all jobs that involved exposure to talc also involved high exposure to respirable silica.] (b) Manufacture ofpulp andpaper Langseth and Andersen (1999) examined cancer incidence among a cohort of 4247 women who had been employed for at least 1 year between 1920 and 1993 in the Norwegian pulp and paper industry. The women had worked mainly in paper sorting and packing departments in 10 paper mills or in administration (85% o f the cohort). Production was judged to involve occupational exposures that included paper dusts, microbes, formaldehyde, talc and asbestos (the latter was used as insulation material in boilers and in the breaks o f various rolling machines), but no measurement data were available. Women were followed for cancer incidence between 1953 and 1993 and SIRs were calculated by comparing the observed incidence to the 5-year age-specific incidence rates foi the female population o f Norway. Information on cancer incidence was obtained by linkage with the National Cancer Registry and information on dates of death and emigration was obtained from the Central Bureau of Statistics of Norway. Records of women who died between 1953 and 1960 were identified manually. Between 1953 and 1993, 535 women in the cohort had died, 65 women had emigrated and 380 new cases of cancer had been diagnosed. The SIR for all cancers was 1.2 (380 observed; 95% Cl, 1.1 1.3). An excess of ovarian cancer diagnoses was observed (37 observed; SIR, 1.5; 95% Cl, 1.1-2.1). In the analyses, workers were also stratified by exposure into the following categories, short-term (< 3 years) versus long-term (> 3 years); period of first exposure (1920-39, 1940-59, 1960-74, 1975-93); and time since first exposure (3-14 years, 15 29 years, > j 0 years). The excess risk was predominantly seen among women who had been employed in the industry for 3 years or more (31 observed; SIR, 1.6; 95% Cl, 1.1 2.3). The excess risk for ovarian cancer was also highest for women under the age of 338 IARC MONOGRAPHS VOLUME 93 55 years at diagnosis, with an SIR of 8.0 (six observed; 95% Cl, 2.9-17.4) for women aged 25-35 years at diagnosis. Among women who worked in the paper mills, the SIR for ovarian cancer was 2.1 (18 observed; 95% Cl, 1.3-3.4). In the discussion, the authors noted that talc is added as a filler in paper mills and may contribute to the excess risk for ovarian cancer observed. On the basis of an extended follow-up of cohort members for cancer incidence to the end of 1999, Langseth and Kjaerheim (2004) conducted a nested case-control study that included 46 employees who had ovarian cancer and 179 controls individually matched to cases by incidence density sampling. An experienced oncologist reviewed the pathology for all cases. Work histories were obtained from personnel records at each mill. Exposure to asbestos, talc and total dust was assessed on the basis of the work histories, questionnaires on production processes completed by industrial hygienists and senior employees, as well as semiquantitative exposure assessments for the 10 mills extracted from an international database of exposure in the pulp and paper industry. Information on possible confounders (including use of talc on sanitary napkins, underwear or diapers) was obtained for 76% o f cases and 57% of controls through a personal interview with the study subject or next of kin. Odds ratios for ovarian cancer were derived by conditional logistic regression. Ever exposure to asbestos was associated with a non-significantly increased odds ratio for ovarian cancer o f 2.0 (95% Cl, 0.7--5.7), while ever exposure to talc (odds ratio, 1.1; 95% Cl, 0.6-2.2) or to total dust (odds ratio, 0.8; 95% Cl, 0.4-1.7) was associated with risks that were close to unity. Among women who were interviewed, the odds ratio for exposure to asbestos was 2.2 (95% Cl, 0.5-9.1). This estimate was unchanged after adjustment for multiple potential confounders, including parity, breastfeeding, tobacco smoking habits and family history of breast or ovarian cancer. The odds ratios for occupational exposure to talc and total dust were similarly unchanged after adjustment for confounding. (c) Rubber manufacturing industries Following the finding of an excess risk for stomach cancer in a cohort of rubber workers in the USA, Blum et al. (1979) carried out a nested case-control study of stomach cancer. Cases were defined as deaths from stomach cancer in two of the rubber companies from 1 Januaiy 1964 to 31 December 1973 (100 deaths in total). Four controls were matched to each case on age, race, sex and company. Using the recorded job history of each worker, the investigators and a group of environmental scientists assessed the potential for exposure (high, moderate, low or none) in each job to the following substances: polycyclic hydrocarbons, nitrosamines, carbon black and detackifiers (anti sticking agents which were mainly talc). No information was available on the purity or composition of the talc (i.e. whether it contained asbestiform materials or other fibrous or non-fibrous carcinogens). While no clear elevation of odds ratio was reported in Company B, a significantly increased relative risk of 2.4 (27 observed; 90% Cl, 1.4-4.1) was found in Company A when workers with moderate and high exposure to talc were TALC 339 pooled into one group High exposure in the latter company was associated with a modest increase in relative risk of 1.3 (13 observed; 90% Cl 0 7--2 5) al n T te7 l0ym ev fileS f fiVe mbber p;duCtion P'ants in Germany, Straif*/ h ' H ft Cr r dUCted 3 m0rt3llty C h0rt Study of 8933 male blue-collar workers who were r: 1 e: 1J; nU7 1950andWhOWerealiveOn 1 Januai^ 1981. Follow-up was started on the date o f completion of 1year of employment or 1 Januaiy 1981, whichever came 1 De ende on at death, at 85 years of age, at the date of loss to follow-up or 31 December 1991, whichever came first. Cause of death was obtained for 97% of wer! ,eCT l W rkerS' W?rk hlSt0neS W6re reconstructed from cost centre codes and were classified into six work areas. SMRs were calculated from national death rates and 7 o r7 eSu m3t 3t 12 (154 observedi 95% Cl, 1.0-1.4) for lung cancer and aall, Z-000t),Zinfeodrm; 9aftion^on 0wZorkVhisOtoriSytwmaasChcoCmanbCinere`dInw3itShUsbesem9iUqeunatnatnitaaItyivseis l(eSvterlasifo,*f exposure to asbestos, talc mtrosamines and carbon black that were estimated by industrial ygienis s o yie d overall estimates o f cumulative exposure (low, medium high) for approx,rnately 95% of ,he cohod. Talc is widely used in rubber p U lc io u and,' a c S h ig to the authors, asbestos was used in all five plants at least until the early 1980s In risk c o n f o r m nW? UnadjUSi d ^ eXp0SUre t0 asbestos or ^ h e r potential workplace confounders high and medium occupational exposure to talc were associated with relative risks for lung cancer of 1.9 (21 observed; 95% Cl, 1.1-3.1) and 1.1 (41 observed; yV U> ^ L6)' resPectlvely, when workers with low exposure were used as the reference group Equivalent risk estimates were 4.3 (11 observed; 95% Cl, 2.1-9.0) and 1'? 9 7 ( ? Se7 ? i 9r5% CI' '6~2 '4) f r St mach C3nCer and 5'4 (three observed; 95% Cl n a T v iy m rt (f f rVCd; 95% C l -5~ 16'7) f r 'aiyngeaI Cancer' SeParate risk yses with adjustment for potential confounders were not performed. [The Workins presented] ` nSk analySCS ^ 3djUSted f r 6StimateS of exP0Sure to asbestos were not 2.1.3 Community-based studies Chen et al. (1992) conducted a case-control study in Beijing, China, of several risk actors for ovarian cancer that included occupational exposure to talc. A total of 220 cases ot newly diagnosed epithelial ovarian cancer were identified between 1984 and 1986 3 7 r r i f o / y Beijing CanCf 1Regist^ - 0 f 1those, 67 [30.5%] were excluded due to death, LI6 .8 / 0 ] due to unavailability of current contact information and four [ 1 .8 %] due to patient refusal. The analysis was carried out on 112 cases and 224 community controls with two age-matched controls per case. Potential controls were excluded if they had a is ory o f serious illness, although the percentage of those excluded for this reason was not specified. In addition, 15 o f the 224 eligible controls initially selected [6.7%] refused to participate m the study and were therefore replaced by other eligible controls No information was provided on the age range of the cases and controls, although the mean age at the time of interview was similar for cases (48.5 years) and controls (4 9 . 0 years). 340 IARC MONOGRAPHS VOLUME 93 All cases were confirmed by laparotomy and pathological review. Data were collected in person by trained interviewers. Odds ratios were estimated using conditional logistic X a T o d d JUStr f ' edUCatl" ^ ^ CCUPational CXP0S^ to talc was associated with an odds ratio for ovarian cancer of 0.9 (95% Cl, 0.3-2.9). [The Working Group noted the incomplete ascertainment of cases o f ovarian cancer due to the nature o f the cancer-reportmg system in China, the large number of cases who were excluded due to death and the exclusion of controls who had a histoiy of serious health problems which may have resulted in selection bias.] ' oo -S*1186 and Stewart C " 4) analysed the occupational histories of 296 women aged ,A years who were diagnosed with ovarian cancer between 1978 and 1981 in the ashington DC area of the USA and 343 hospital-based controls matched to cases on age and race. Pathology was confirmed for all cases. Trained interviewers used a standardized questionnaire to obtain information from each participant on their lifetime job histoiy and occupational exposure to talc. An industrial hygienist blinded to the case status of each participant evaluated each industiy and occupation for potential exposure to talc, ionizing radiation, polycyclic aromatic hydrocarbons and solvents, using a scale of 0 (definitely not exposed) to 4 (definitely exposed). Women were considered to be exposed if they had an exposure rating of 2-4 (possibly, probably or definitely exposed). Logistic regression adjusted for race, age, parity, gynaecological surgery and duration o f employment in jobs with the exposure of interest was used for the analyses. Controlling for additional known and potential risk factors for ovarian cancer, including parity, oral contraceptive use and cigarette smoking, did not change these estimates. Women who were classified as having been occupationally exposed to talc had odds ratios below the null, although the confidence limits were wide due to the small number of exposed women (12 cases 31 controls). For women with 10 or more years of employment in an occupation with o w ' r i ' P' bable r definite exPsure to talc, the odds ratio was 0.5 (five exposed cases; 95/o Cl, 0.2-1.5). The risk for ovarian cancer was not significantly elevated for any exposure or duration of employment assessed. [Limitations o f this analysis include the small number o f women occupationally exposed to talc.] `Industrial talc' was one of the substances evaluated by the exposure assessment team in the com m une-based case-control study carried out in Montral, Canada (Siemiatycki, 991) and described in detail in the monograph on carbon black. About 5% of the 4263 study subjects was considered to be exposed to industrial talc, mostly in the following occupations: painters, motor vehicle mechanics and farmers. Exposure to talc was analysed in relation to 11 different types of cancer, at two levels o f exposure (any or substantial). No statistically significant increases in risk were observed. The odds ratios for lung cancer were 0.9 (35 exposed cases; 90% Cl, 0.6-1.4) for `any exposure' and 0.9 (nine exposed cases; 90% Cl, 0.5-1.9) for `substantial exposure'. Prostate cancer was the only site with a borderline significant increased risk, with an odds ratio of i ^ Ter Sed C3SeS; 90% CI' 1-0 -2 ' 1} f r `any exPsure' and 11 0*ven exposed cases; 90% CI, 0.5-2.3) for `substantial exposure'. [The main limitation of the study was the reliance on expert opinions of exposure rather than measurements for exposure TALC 341 assessment Also, exposure levels tend to be lower in such community-based studies than " t that are selected for cohort studies. The main advantages t m th" avadaluhty o f histologically confirmed incident cases and detailed information on tobacco smoking habits and other characteristics o f the subjects.] 2.2 Cosmetic use of talc This evaluation was limited to ovarian cancer because the Working Group was unaware of studies o f other cancers associated with the cosmetic use o f talc. The content of body powders used by women varies by product and has changed over ime although data that document this are limited. Before the mid-1970s, body powders may have contained vaiying but usually small quantities o f amphiboles. After that time amphibole was voluntarily reduced to less than detectable levels, at least in western Europe and the USA Other non-talc minerals that include chlorite, quartz, carbonates and pyrophyllite may also be found in body powders in vaiying and occasionally not insignificant quantities m the past and currently. Other added ingredients, which depend on the product, could include cornstarch and perfumes. 2.2.1 Cohort studies Gertig et al. (2000) earned out the only prospective cohort analysis that reported an association between perineal use o f talcum, baby or deodorant powder and the risk for ovarian cancer. This analysis was conducted among participants in the Nurses' Health Study, a cohort of 121 700 female registered nurses who had been followed since 1976 All participants were between the ages of 30 and 55 years and lived in one of 11 states of the USA at study enrolment. Questionnaires were mailed to participants eveiy 2 years beginning m 1976 to obtain information on the medical histoiy of each woman and potential risk factors for cancer, heart disease and other conditions The 1982 questionnaire requested information on histoiy and frequency of application of powder to t e perineal area (none, daily, one to six times a week, less than once a week) and histoiy of application o f powder to sanitaiy napkins (no/yes). `Ever talc use' was classified as ever use on either the perineal area or on sanitaiy napkins. The study population included 630 women who responded to the questions on powder use in 1982 and who were not excluded from the analysis for another reason (cancer other than non-melanoma skin cancer before 1982, bilateral oophorectomy, surgeiy with unknown number of ovaries removed or radiation therapy) and entailed 984 212 person-years of follow-up. Between 982 and June 1996, j 07 incident cases of epithelial ovarian cancer were identified by self-reporting in a biennial questionnaire, by deaths that were reported by relatives or postal authorities or through the National Death Index. Physicians blinded with respect to exposure status reviewed pathology reports to confinn each case and to detenu ine the istological subtype for each tumour as reported by the woman's pathologist. Pooled logistic regression was used to model the incidence rate ratio of ovarian cancer for the 342 IARC MONOGRAPHS VOLUME 93 exposed versus unexposed participants. The reported results were adjusted for aee in years, parity (defined as the number o f pregnancies lasting 6 months or more) duration of oral contraceptive use, body mass index, history o f tubal ligation, tobacco smoking status and postmenopausal use of hormones. Additional covariates considered as potential F a m r hiltow o f ^ * me" arChe' dUrati'n o f 1^ f e e d i n g and age at menopause, amity history o f ovarian cancer was not considered to be a confounder since information on this covariate was not collected until 1992. In 1982 40 4% of the cohort Z t n h4 n 7 n 1pe1 r ' ,alC "se (" = 31 789) and l 4 M rePrtcd " of *ily use ( 1141 ). Overall, no association between `ever use' o f talcum powder and total risk for epithelial ovarian cancer (relative risk, 1.1; 95% Cl, 0.9-1.4) and no trend of increased risk for ovarian cancer with increasing frequency of talc use were observed. owever a modest increase in risk for serous invasive cancers was associated with anv history o f talc use (relative risk, 1.4; 95% Cl, 1.0-1.9) and a borderline significant trend was found w i^ increasing frequency of use ^ for trend = 0 05) WQmen without a istory o f tubal ligation, no association was observed between history o f talc use and total risk for epithelial ovarian cancer (relative risk, 1.0; 95% Cl, 0.7-1.3). Similarly, history of tubal hgation did not modify the association between the use of talc and risk for serous invasive cancers. [Limitations of this analysis include the availability o f exposure information at a single time-point only, the relatively short follow-up period after exposure assessment and the lack of information on age at first use of talc, duration o f use o f talc, current use o f talc in 1982 and use o f talc before tubal ligation or pregnancy, all of which are potentially important parameters based on previous studies.] 2.2.2 Case-control studies (Table 2.3) Cramer et al. (1982) reported the first epidemiological study of genital talc use and the risk for ovarian cancer. The analysis included 215 cases of epithelial ovarian cancer and 215 population-based controls matched to cases by age (within 2 years), race and residence. All cases were Caucasian, English-speaking residents of Massachusetts, USA aged 18-80 years, who had been diagnosed with epithelial ovarian cancer between November 1978 and September 1981. Cases were identified through pathology logs or tumour boards of 12 participating Boston hospitals. Among 297 eligible cases identified du" ng, time period o f interest, 41 were excluded from the study due to: physician lefusal (13), patient refusal (14) or death/change of address (14). An additional 41 cases were excluded because they had a non-ovarian primaty (18) or a non-epithelial ovarian tumour based on a review of pathology specimens by the authors. Controls were identified though annual listings of the names, addresses and ages o f all Massachusetts residents. Among 475 women identified as potential controls, 11.8% (56) could not be reached, 6.1% (29) were ineligible due to previous bilateral oophorectomy, 4.2% (20) were the wrong age, not Caucasian or did not speak English and 32.6% (155) refused to participate. All cases and controls were interviewed in person to obtain information on their medical history, menstrual and reproductive histories, as well as potential for exposure Table 2.3. Case control studies of epithelial ovarian cancer (invasive or borderline) and cosmetic use of talc Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories No. of exposed cases Odds ratio (95% Cl) Cramer et aI (1982) Boston, MA, USA, 1978-81 215 Caucasian, English-speaking women, aged 18-80 years, identified through pathology logs or tumour boards of 12 Boston hospitals, histological confirmation of diagnosis, 215 population-based controls identified through annual listings of names, ages and addresses of all Massachusetts residents, matched by age (2 years), race, residence In-person interviews, information collected on medical history, menstrual and reproductive history, potential or definite exposure to talc `Any' perineal 92 exposure to talc As dusting 32 powder on perineum and sanitary' napkins 1 6 ( 1 0 - 2 5) 3.3 (1 7-6.4) Hartge eta I (1983) Washington DC, USA, 1974-77 135 incident cases Interviews to collect `Any' use of 67 treated at participating information on reproductive talc hospitals, and sexual history', medical `Genital' 7 171 population-based history, drug use and other exposure to controls, frequency- exposures, exposure to talc talc matched by age, race, categorized as `any' or hospital `genital' (includes use on genitals, on sanitary napkins or on underwear) 0.7 (0.4-1 1) 2.5 (0.7-10.0) Adjustment for potential confounders Comments Parity, menopausal status, religion, marital status, educational level, weight, age at menarche, exact parity, oral contraceptive use, postmenopausal use o f hormones, tobacco smoking Distribution of tumour histologies similar for exposed and unexposed cases, potential for talc exposure by way of contraceptives, pelvic surgery or perineal hygiene considered, no information on duration or frequency of talc use; low participation rates among controls (56% o f cases matched with no refusals; 27% matched after 1 refusal; 17% matched after 2 or more refusals) Age, race, pregnancy Questions on talc added after study began: no information on duration or frequency o f exposure, no controlling for other potential confounders, potential for selection bias TALC UJ CtO*. Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Whittemore et al (1988), San Francisco, CA, USA 198385 188 incident cases diagnosed at 8 hospitals, aged 18-74 years, histological verification o f diagnosis; 539 controls selected from women hospitalized for noncancerous conditions (=280) or from the population using random digit-dialling (=259), matched by age (5 years), race, hospital/date of admission (hospital controls) or telephone area code/prefix (population controls) Structured m-person interview's, information collected on medical history, menstrual and reproductive history, family history, environmental exposures (talc, coffee, alcohol. tobacco), talc exposure categorized by type of application, duration o f use prior to tubal ligation or hysterectomy, frequency of use Type o f application Perineum only Sanitary pads only Diaphragm only Any two All three Duration o f use (years) None 1-9 >10 Frequency o f use Never 1-20 times/month >20 times/month 30 times/month p for trend UJ ti. -fe- No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments No trend of increasing risk with increasing duration of 22 1.5 (0 8-2 6) Parity, oral exposure, as measured m 5 0.6 (0 2-1 8) contraceptive use years o f talcum powder use > 73 on the perineum prior to n 9 1.5 (0 6-3 6) 67 1.4 (0.9-2.0) 1 0.4 (0.0-2 9) Parity 103 1.0 34 1 6 ( 1.0- 2 .6) 50 1 1 (0.7-1 7) Parity tubal ligation or o hysterectomy, non-statistically significant z: o trend o f increasing risk o with increasing frequency 73 of exposure, as measured in >-O number of applications of X talc to the perineum per (/i month < or* C 97 1 0 m 41 1 3 (0.8-2 0) NO 44 1 5 (0 9-2.2) 1 3 (0 9-1.9) 0 19 Table 2.3 (contd) TALC Reference, study location, study period Characteristics o f cases Exposure assessment and controls Exposure categories No. o f exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments Booth et al (1989), London and Oxford, United Kingdom, I 978-83 235 incident cases from 15 hospitals, aged 65 years or under at diagnosis, diagnosed within 2 years of interview, histological confirmation of diagnosis, 451 hospital-based controls selected from same 15 hospitals, same age distribution as the cases Interviewer- administered standard questionnaire, information obtained on reproductive and menstrual history, on exposure to exogenous estrogens, cigarettes, talc, talc exposure categorized by frequency of use on perineum and whether it was used to store a diaphragm Frequency o f use Never 76 Rarely 6 Monthly 7 Weekly 57 Daily 71 p for trend 1.0 0.9 (0 3-2 4) 0.7(0 3-1.8) 2 0 (1.3-3 4) 1 3 (0 8-1 9) 0.05 Age, socioeconomic status Participation rates not provided; questions on talc use added 3 months after start of study, data on talc exposure missing for 18 cases and 17 controls Harlow & Weiss (1989), western Washington State, USA. 1980-85 116 Caucasian women from 3 urban counties captured in Seattle- In-person interviews, information obtained on reproductive, sexual and `Any' perineal 49 use Puget Sound Cancer medical histories, as well as Type o fpowder Surveillance System, perineal exposure to talc, talc used aged 20-79 years, exposure categorized as `any' Cornstarch only 4 independent perineal use, by method of Baby powder 18 pathological review use, and by type of powder only 73% of total, used Baby powder, 22 histological agreement combined 94% of reviewed cases, Talc, unspecified 13 158 white population- Deodorizing 10 based controls selected powder only by random-digit Deodorizing, 14 dialling, matched by combined age, county of residence 1 1 (0 7-2.1) 0 8 (0 2-3 8) 0 8 (0.4-1.9) 0 9 (0 5-2 0) 1 0 (0 4-2 4) 3 5 (I 2-28 7) Age, parity, use f ral contraceptives Cases diagnosed with borderline (serous or mucinous) tumours, study limited by incomplete information on powder use and small size, no significant association between method o f powder use and risk for borderline tumours 2 8 (1 1-11 7) U4^J czi Table 2.3 (contd) Reference, study location, study period Characteristics o f cases Exposure assessment and controls Exposure categories Chen et al (1992), Beijing, China, 1984-86 112 women from Beijing Cancer Registry, with a mean age of 48 5 years, confirmation of diagnosis by laparotomy and pathological examination in all cases, 224 population-based controls selected first on basis of area of residence of cases and then randomly from census lists of all women within 1 year of age of identified case, matched by age, mean age, 49 0 years Interviewer-administered questionnaire, information obtained on menstrual, obstetric, marital, medical. family and dietary histories as well as exposure to talc (perineally and occupationally), perineal exposure reported as yes/no Use on perineum or lower abdomen O4^J Os No. of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments 7 3.9 (0 9-10 6) Education, parity Age range o f cases and controls not reported T>O n i2 8TO >-o oXo < Or* C sO Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Harlow et a! (1992), Boston, MA, Massachusetts, USA, 1984-87 235 white women from 10 hospitals in metropolitan Boston area, aged 18-76 years, independent pathological confirmation of diagnosis, 239 population-based controls randomly selected from town registers, matched by age (2 years), race, precinct o f residence, no history of bilateral oopherectomy In-person interviews, information collected on occupational history, medical and reproductive history, dietary history, tobacco smoking, hygienic practices including perineal exposure to talc, exposure to talc categorized by type of application, brand of powders, duration and frequency of use Any' perineal use of talc Method o f application Sanitary napkins or underwear only Partner or applications to diaphragm Dusting on perineum Frequency (no per month) None <5 5-29 >30 p for trend Years o f use None <10 10-29 >30 p for trend Total applications N<1o0n0e0 1000-10 000 >10 000 p for trend No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments 114 1 5 (1.0-2.1) Parity, Odds ratio for women with education, >10 000 lifetime marital status, applications unchanged religion, use of after excluding applications 9 1 1 (0.4-2 8) sanitary napkins, that occurred after tubal douching, age, ligation or hysterectomy weight (odds ratio, 1 7, 95% Cl, 20 1 2 (0.6-2 4) 1.0-3 0), significant increase in odds ratio for women with >10 000 85 1 7 (1 .1 -2 7) lifetime applications observed after excluding use o f talc during non ovulatory periods and after 121 1 0 surgical sterilization (odds 32 1 5 (0.8-2 7) ratio, 2 8, 95% Cl, 1 4 -5 4) 24 1 2 (0.6-2.2) 58 1 8(1 1-3 0) 0 046 121 1 0 14 1 2 (0 5-2 6) 49 1 6 (1 0-2 7) 5! 1 6(1 0-2 7) 0 07 TALC 121 1 0 18 1 3 (0 7-2 7) 54 1 5 (0 9-2 4) 42 1 8 (1 0-3 0) 0 09 Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Rosenblatt et al (1992), Baltimore, MD, USA, 1981-85 77 women admitted to Johns Hopkins Hospital as in-patients for treatment or diagnosis, diagnosed within 6 months of admission, residents of the USA, pathological confirmation of diagnosis, 46 hospital-based controls selected from female in-patients with no gynaecological or malignant conditions, matched a posteriori by age (5 years), race, closest date of diagnostic admission Questionnaire administered by telephone and in the hospital, information collected on genital and respiratory exposure to fibrecontainmg substances, such as talc, sources o f genital exposure included contraceptive methods (diaphragm, condoms), dusting of perineum and sanitary products, sources of respiratory' exposure included use of face and/or body powders, residential or occupational exposure to fibre-containing substances, such as talc, asbestos, fiberglass, estimation of `dose' by adding number of years of exposure from all sources Genital fibre use Method o f application Diaphragm use with powder Genital bath talc Sanitary napkin with talc exposure -ouo No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments ozo 67 1 0 (0 2-4 0) Parity Investigators encountered > difficulty finding controls who met all of the n?0 14 3 0 (0 8-10 8) Parity, education matching criteria For 2 No adjustment analysis, 46 matched sets, 22 1 7 (0 7-3 9) Highest weight, of which 31 sets had 2 21 4 8 (1 3-17 8) 1 year prior to cases and I control, o diagnosis limitations include small 7S > study size, broad definition -o o f fibre exposure, limited X in information available on < perineal exposure to talc On C 2 m vOOJ Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Tzonou et al (1993), Athens, Greece, 1989-91 189 women hospitalized for ovarian cancer surgery in 2 major cancer hospitals in Greater Athens, aged 75 years or under, histological confirmation of diagnosis, 200 hospital visitor controls (selected from visitors to patients hospitalized in the same wards as cases), not matched to cases by age Questionnaire administered m hospital by medical residents; information collected on medical and leproductive histories, as well as personal, demographic and socioeconomic variables, qualitative assessment of talc exposure (yes/no use in the perineal region) Talc application in perineum No Yes No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments Age, education, Study limited by very low weight, age at prevalence of perineal talc 1836 1.0 1 1 (0 3-4.0) menarche, menopausal use status, age at menopause, parity, age at first birth, smoking status, H alcohol use, > n coffee o consumption, use of analgesics, use of tranquilizers or hypnotics, use o f hair dyes 0-1 O Table 2.3 (contd) Reference, study location, study period Characteristics o f cases Exposure assessment and controls Exposure categories Purdie el al ( 1995). Queensland, New South Wales, Victoria, Australia, 199093 824 incident cases diagnosed and registered in all major gynaecologicaloncology treatment centres in 3 states, aged 18-79 years, independent pathological confirmation of diagnosis, 860 population-based controls selected randomly from electoral rolls, stratified by age and geographical region Interviewer-administered standardized questionnaire in clinic (cases) or home (some cases, all controls). information collected on medical, reproductive, family and occupational histories, as well as dietary factors and history of talc use Use of talc around the abdomen or perineum Shushan el al (1996), Israel, 1990-93 200 incident cases (164 invasive, 36 borderline) diagnosed and reported to Israel Cancer Registry, aged 36-64 years, histological confirmation of diagnosis, 408 population-based controls selected by random-digit dialing, matched by eograplucal aiea Interviewer-administered standard questionnaire. information collected on reproductive history', use of oral contraceptives and fertility drugs, exposure to talc, exposure to talc stratified into "never/seldomf 'moderate/a lot' Use o f talc Moderate/a lot CoOOn No. of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments [467] 1 .3 (1 .0 -1 5 ) Parity, other 56.7% potential confounders, e g contraceptive use, also __ considered ^7i n o o >-o Xon < Or c No control for Study limited by the very 21 [197] confounding sparse information on talc (p = 0 04) use and the unavailability VOOJ of adjusted results for the association between use of talc and the risk for ovarian cancer Table 2.3 (contd) Reference, study location, study period Chang & Risch (1997), metropolitan Toronto and southern Ontario, Canada, 198992 Characteristics of cases Exposure assessment and controls 450 incident cases (primary, invasive and borderline), aged 35-- 79 years, histological confirmation of diagnosis, 564 population-based controls identified through provincial records ot all homeowners, tenants and family members; randomly selected from same residential area, matched by age within 15-year age groups Interviewer-administered questionnaire, information collected on menstrual and reproductive history', use of hormones and oral contraceptives, and use of talc, exposure to talc categorized on basis of 'any' exposure, type of exposure, frequency and duration of perineal application Exposure categories No. of exposed cases Odds ratio (95% Cl) `Any' exposure 198 to talc Type o f exposure Sanitary' napkins 51 After bathing 172 Frequency o f after-bath use (times/month) None <10 76 10-25 54 >25 41 Per 10 applications per month Duration o f after-bath use (years) None <30 60 30-40 71 >40 41 Per 10 years of use 1 4 (1.1-1 9) 1.3 (0 9-2 0) 1 3 (1 0 -1 7) 1.0 1 8 (1 2 -2 7) 1 1 (0 7-1 7) 1 0 (0.6-1 5) 0 9 (0 7-1 1) 1.0 1 7 (1 .1 -2 6) 1 4 (1 0-2 2) 0 9 (0 5-1 4) I 1 (1 0-1.2) Adjustment for potential confounders Comments Age at interview, duration o f oral contraceptive use, parity (number o f fullterm pregnancies), duration of lactation per pregnancy, history' o f tubal ligation or hysterectomy, family history of breast or ovarian cancer Authors do not specify whether cases were identified through a cancer registry or some other reporting mechanism Borderline significant trend observed with increasing duration o f exposure to talc, but not with increasing frequency o f exposure TALC u> LLO/i NJ Table 2.3 (contd) IARC MONOGRAPHS VOLUME 93 Reference, study location, study period Cook el al (1997) Western Washington State, USA, 1986-1988 Characteristics of cases Exposure assessment and controls Exposure categories No of Odds ratio exposed (95% Cl) cases 313 incident cases (234 invasive, 79 borderline) identified from records of Cancer Surveillance System o f western Washington; white residents of three counties (King, Pierce. Snohomish), aged 20 79 years, no information on whether diagnosis was histologically confirmed, 422 W'hite populationbased controls selected by random digit dialling (part of a larger control pool for several studies of cancer m women), matched by age Structured m-person interviews, information collected on medical and reproductive histories, smoking habits, birth control methods and use of genital powders and deodorant sprays, exposure to genital powders assessed on the basis o f `any' lifetime exposure, method of use and cumulative lifetime exposure (days, months or lifetime applications) Lifetime perineal application None Any Exclusive use n f powderfo r Perineal dusting Diaphragm storage Dusting sanitary napkins Deodorant spray Anv use o f poM'derfo r Perineal dusting Diaphragm storage Dusting sanitary napkins Deodorant spray Cumulative lifetime perineal dusting (days) None <2000 2001-5000 5001-10 000 >10 000 154 159 55 22 12 18 95 46 38 40 154 20 24 21 28 1 0 1 5 (1 1-2.0) 1 8 (1 2-2 9) 0.8 (0 4-1 4) 1 5 (0 6-3.6) 1 5 (0 8-3 0) 1 6 (1 1-2 3) 1 0 (0 6-1 6) 0 9 (0.5-1 5) 1 9 (1.1-3 1) 1 0 1 8 (0.9-3 5) 1 6 (0 9-2 9) 1 2 (0 6-2.4) I 8 (0.9-3 4) Adjustment for potential confounders Adjusted for age Adjusted for age Adjusted for age and other methods of genital powder application Adjusted for age and other methods of genital powder application Comments Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Eltabbakh et al (1998), Buffalo, NY, USA, 1982-96 `Study' group150 women admitted for treatment of primary extra-ovarian peritoneal cancer to Roswell Park Cancer Institute, histological confirmation of diagnosis, `control' group 466 women treated for primary ovarian cancer at same centre, pathological review of diagnosis Self-administered, 44-item questionnaire completed at hospital admission Exposure categories Perineal use of talc No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments 224 p=Q 003 (48 1%) No control for `Cases for this study were confounding women diagnosed with primary peritoneal cancers. Case definition excluded patients with diagnoses of peritoneal mesothelioma, borderline tumours of peritoneum or invasive --j ovarian cancer, no healthy ^ controls enrolled m this q study 'Controls' were women diagnosed with primary epithelial ovarian cancer Control definition excluded patients with diagnoses of non-epithelial ovarian cancer and ovarian cancer secondary to mtastass front other sites Lo.nJ CO UL*J -U IARC MONOGRAPHS VOLUME 93 Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Godard et al (1998), Montreal. Quebec, Canada, 199596 170 incident cases with primary invasive or borderline epithelial tumours, identified at two gynaecological clinics, aged 20-84 years; histological confirmation of diagnosis, 170 population-based controls selected by a modified random-digit dialling method, frequency-matched by age (1 year), French Canadian ethnicity Standardized 57-itein questionnaire; telephone or in-person interviews conducted with cases, no information on how controls were interviewed, qualitative assessment of perineal talc exposure (ever/never) `Ever' use o f talc on perineum No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders [18] 2 5 (0 9-6 6) (10 6%) Age at menarche, age at menopause, parity, age at first and last childbirth, duration of oral contraceptive use, age at last oral contraceptive use, tubal ligation, alcohol use, previous breast or abdominal surgery Comments Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Cramer el al (1999). eastern Massachusetts and New Hampshire, USA, 1992-97 563 incident cases (including borderline tumours) identified through hospital tumour boards or statewide cancer registries, age range not provided: histological confirmation of diagnosis for all cases, 523 population-based controls selected by random-digit dialling and through annual listings of names, ages and addresses of all Massachusetts residents (women over the age of 60 years). frequency-matched by age (4 years), location of residence In-person interviews using standardized questionnaire. information collected on medical and reproductive histories, family history' and personal habits, multiple questions on potential routes of talc exposure (non-genital, genital, husband's use), brands used, age at first use. duration and frequency of use No genital exposure Any genital exposure M ethod o f use No use Non-genital areas Dusting perineum Dusting sanitary napkins Dusting underwear More than one method Frequency (uses/month) None <30 30-39 >40 Duration o f use (years) None <20 20-30 >30 No of exposed cases Odds ratio (95% Cl) Adjustment lor potential confounders Comments 411 1 0 Age, study site, parity, oral 152 1 6(1 2-2 1) contraceptive use, body mass index, family 312 1 0 history o f breast 99 1 1 (0 8-1.5) or ovarian cancer, history of 71 1.5 (1 0-2 2) tubal ligation H 20 1 5 (0 7-3 1) > r* n 8 1 2 (0 4-3.6) 53 2 2 (1 3-3.6) 312 1.0 54 2.2 (1.4-3 6) 59 1 7 (0.8-1 8) 23 1 7 (0.8-3 1) 312 1 0 55 19(1 2-3 0) 32 1.3 (0 8-2 3) 59 1.4 (0 9-2 3) UenJ CO Table 2.3 (contd) Reference, study location, study period Characteristics o f cases Exposure assessment and controls Cramer et al (1999) (contd) Exposure categories Total no o f applications None <3000 3000-10 000 >10 000 p for trend Total no. o f applications (censored analysis) None <3000 3000-10 000 >10 000 p for trend LO CO OS No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments 312 10 51 1.8(1.1-3 0) 36 1 4 (0.8-2 4) 59 1 4 (0 9-2 2) 0.16 312 1.0 59 1 5 (1 0 -2 .4 ) 51 1 7 (1 1-2 8) 36 1 S (1 0-3 2) 0 02 > n o2 oo JO Censored analysis excludes talc applications that ^ occurred during non- ovulatory years or after 2 hysterectomy or tubal Js ligation. Includes non- C genitally exposed women 2 OoJ Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories Wong et al (1999) Buffalo, NY, USA, 1982-92 462 incident cases admitted foi treatment of primary extraovarian peritoneal cancer to Roswell Park Cancer Institute, mean age, 54 9 years, histological confirmation of diagnosis, 693 hospital-based controls treated for non-gynaecological malignancies at same cancer centre; mean age, 54 9 years, frequency-matched to cases by age at diagnosis (5 years) Self-administered, 44-item questionnaire completed at hospital admission. information collected on medical, social, family, dietary and occupational histones, method o f talc use (never, sanitary napkin, genital/tlugh area, both) assessed and duration of use Method o f use Never Sanitary napkin Genital or thigh area Both Duration o f use (years) None 1-9 10-19 >20 No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments Age, parity, oral Case population largely 241 1 0 contraceptive that reported by Eltabbakh 13 0.9 (0 4-2 0) use, smoking, el a/ (1998), 32 cases, 157 1 0 (0 8-1 3) family history of 39 controls did not recall ovarian cancer, duration o f use 51 1 1 (0 7-1.7) age at menarche, menopausal status, income, 241 1 0 education. 39 0 9 (0 6-1 5) geographical 49 1 4 (0.9-2 2) location, history 101 0 9 (0 6-1.2) of tubal ligation or hysterectomy LJ L /l Table 2.3 (contd) Reference, study location, study period Characteristics o f cases Exposure assessment and controls Exposure categories Mess et al (2000), eastern Pennsylvania, southern New Jersey, Delaware, USA, 1994-1998 767 incident cases identified at 39 hospitals in the Delaware Valley region, aged 20-69, diagnosis within 6 months prior to interview, pathological review o f a random subset of cases (n = 120) 1367 population-based controls identified through random digit dialing (<65 years of age) and Health Care Financing Administration lists (65-69 years o f age), frequency matched by age and location of residence Standardized in-person interviews, information collected on sexual activity, use of contraceptives. menstrual and reproductive history, and history and duration of talc use (genital, non-genital applications, exposure via male sexual partners) Method o f use Never Feet, arms, breasts Genital/rectal Sanitary' napkin Underwear Diaphragm/ cervical cap Male partner Duration o f use (years) Never <1 1-4 5-9 >10 oo No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments 349 1 0 Age, parity, race, Risk for ovarian cancer family history ot compared with 50 women > 335 1 4 (1.1-1 6) ovarian cancers, with primary peritoneal O oral cancers, no control for 2 161 1 5 (1 2-2.0) contraceptive confounding, analysis of oz 77 I 6 (1 1-2.3) use, tubal duration examined risk for o 70 1 7 (1 2-2.4) ligation, cases reporting use of talc QTO 10 0.6 (0 3-1.2) hysterectomy, on the feet, genital and > lactation rectal areas 56 1 0 (0 7-1.4) X oW< 401 1 0 Cr 1 17 2.0 (1 0--4.0) 2 76 1 6 (1 1-2.3) m 40 1.2 (0 8-1 9) 233 1 2 (1 0-1.5) Table 2.3 (contd) Reference, study location, study period Characteristics of cases Exposure assessment and controls Exposure categories No of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments TALC Langseth & 35 (invasive and In-person interviews `Ever' use of talc 12 Kjaerheim borderline tumours) conducted at mills or by for personal (2004), selected from cohort of telephone, information hygiene Norway, 1953- 4247 female pulp and collected on occupational 99 paper workers, cohort history, household exposure follow-up, 1953-99. to asbestos, menstrual and histological review and reproductive history, confirmation of hereditary risk o f cancer, as diagnosis, well as talc use on sanitary 121 selected from the napkins, underwear or diapers cohort by incidence or by husband in genital area density sampling, matched by birth (year 2 years), controls had no ovarian cancer and had intact ovaries 1 2 (0 4-3 2) Adjusted for possible confounders, but not explicitly stated Nested case-control study conducted in a cohort study o f 10 pulp and paper mills, many missing values among proxy respondents 359 GO O Table 2.3 (contd) Reference, study location, study period Characteristics o f cases Exposure assessment and controls Exposure categories No. of exposed cases Odds ratio (95% Cl) Adjustment for potential confounders Comments IARC MONOGRAPHS VOLUME 93 Mills et a i (2004), central California, USA, 2000-01 249 incident cases Telephone interview to obtain Perineal use o f from 22 counties information on medical talc diagnosed in two history, menstrual and Never 143 regional cancer reproductive history, family Ever 106 registries, using rapid history of cancer, history of Frequency o f use case ascertainment perineal talc exposure Never 143 procedures. (frequency, duration and cl/w eek 34 histological calendar years of use), 1-3/week 31 confirmation of 'cumulative' use calculated 4-7/week 41 diagnosis for a subset by multiplying frequency p for trend of cases. (categorical variable) by Duration o f use 1 105 population-based duration in months (years) controls identified by Never 143 random-digit dialling, <3 18 frequency-matched by 4-12 32 age, race, ethnicity 13-30 29 >30 21 p for trend Cumulative use Never 143 1st quartile 18 (lowest) 2nd quartile 28 3rd quartile 34 4th quartile 20 (highest) p for trend 1 0 1.4 (1 0-1 9) 1 0 1 3 (0.9-2 I) 1 6 (0 7-1 8) 1 7 (1 1-2 6) 0 015 1.0 1 0 (0.6-1 8) 1 9 (1 .2 -3 0) 1.5 (0.9-2.3) 1 2 (0.7-2 1) 0 045 1.0 1.0 (0.6-1 8) 1 8 (1 1-3 0) 1 7 (1 1-2 7) 1 1 (0.6-1 8) Age, race/ethnicitv, duration of oral contraceptive use, breastfeeding Additional co variates considered to be potential confounders included family history of breast or ovarian cancer, parity, history of pregnancy, body mass index, hysterectomy, tubal ligation, duration of postmenopausal use of hormones Cumulative use calculated as frequency (categorical weighting from 0-3) multiplied by duration 0 051 Cl, confidence interval TALC 361 to talc by way of contraceptives, perineal hygiene or surgery. Ninety-two cases (42.8%) and 61 controls (28.4%) reported a history o f regular use of talc as a dusting powder to the perineum, on sanitary napkins or on both. After adjustment for parity (yes/no) and menopausal status (pre-/post-), a significant association was found between `any perineal use' of talcum powder and the risk for ovarian cancer (odds ratio, 1.9; 95% Cl, 1.3-2.9). This association was attenuated but still significant after adjustment for additional potential confounders, including religion, marital status, level of education, weight, age at menarche, parity (number of children), oral contraceptive use, menopausal use of hormones and tobacco smoking (adjusted odds ratio, 1.6; 95% Cl, 1.0-2.5). A single type of perineal exposure to talc (either as a dusting powder to the perineum or on sanitaiy napkins) was associated with a borderline significantly increased risk for ovarian cancer (odds ratio, 1.6; 95% Cl, 1.0-2.5) after adjustment for parity and menopausal status, while a history of both types of perineal exposure was associated with a significant increase in risk (adjusted odds ratio, 3.3; 95% Cl, 1.7-6.4). No association was seen between other potential sources of exposure to talc (pelvic surgery, use of condoms, use of diaphragm or using talc for diaphragm storage) and the risk for ovarian cancer. In addition, the results were essentially unchanged after excluding women who had had a tubal ligation or hysterectomy (odds ratio, 2.8; P < 0.003), although the authors noted that these surgical procedures are usually performed at mid-life when substantial exposure to talc may already have occurred. The distribution of tumour histologies was similar for exposed and unexposed cases; 53.7% of tumours were classified as serous among the unexposed cases and 48.9% among the exposed cases with `any' perineal use of talc. [Limitations of this report include the lack of information on duration and frequency of talc use. In addition, participation rates among the controls were quite low (50%), although the authors noted in a secondary analysis that, when cases were matched to the first control selected (i.e. 100% participation), a positive association was also found (odds ratio, 2.44; P < 0.05).] Hartge et al. (1983) published a brief report o f a study conducted between 1974 and 1977 in the Washington DC (USA) area. The study included 197 cases treated for pathologically confirmed epithelial ovarian cancer at participating hospitals and 197 controls treated at the same hospitals for conditions other than pregnancy, malignancies and gynaecological or psychiatric diseases. Controls were frequencymatched to cases by age, race and hospital. Interviews were conducted in the hospital for controls and at home for most cases to collect information on reproductive and sexual histoiy, medical history, drug use and other exposures. Questions on exposure to talc were added after the study began. As a result, the analysis included only 135 cases and 171 controls with information on exposure to talc. Sixty-seven cases [49.6%] and 100 controls [58.5%] reported `any' use of talc (including non-genital uses), while seven cases [5.2%] and three controls [1.8%] reported genital use of talc (including use on genitals, on sanitaiy napkins or on underwear). No association was observed between `any' use of talc and the risk for ovarian cancer (odds ratio, 0.7; 95% Cl, 0.4-1.1). This estimate was unchanged after adjustment for race, age and pregnancy. A non-significant positive association was found between genital use of talc and the risk for ovarian cancer (odds 362 IARC MONOGRAPHS VOLUME 93 ratio, 2.5; 95% Cl, 0.7-10.0). [Limitations o f this study included its small size and the low prevalence of genital use of talc, the lack of information on its duration and frequency and age at first use, the lack of control for other potential confounders and the increased potential for selection bias due to different interviewing protocols for cases and controls. In addition, no information was given in this brief report on the methods used in the analysis to control for confounding.] Whittemore et al. (1988) analysed the association between perineal use of talc and the risk for invasive epithelial ovarian cancer among 188 cases and 539 controls in the San Francisco Bay area (CA, USA). Cases were residents of northern California, aged 18-- 74 years, who had been diagnosed with an invasive ovarian tumour between January 1983 and December 1985 at one of eight hospitals. Controls were either selected from among women who had been hospitalized for a non-cancerous condition at one of these eight hospitals or were identified from the population using random-digit dialling. Women in each control group were matched to each case by age (within 5 years) and race (white, black, other), plus hospital and date of admission (within 3 months) for the hospital controls (n = 280) and telephone area code and prefix for the population-based controls (n = 259). Structured interviews were conducted in the homes of participants to obtain information on the history, frequency and duration of perineal use of talc, medical history and additional covariates of interest (menstrual and reproductive histories, family history and environmental exposures, such as consumption o f alcohol, coffee and tobacco). O f 317 eligible cases, eight (2.5%) were excluded due to physician refusal, 30(9.5%) due to patient refusal, 44 (13.9%) due to death or incapacitating illness and 47(14.8%) due to non-invasive tumours, which left 188 (59.3%) for inclusion in the analysis. Among the controls, 68% of the women identified as eligible hospital controls ( = 354) and 71% of the women identified by telephone as eligible population-based controls (n = 329) agreed to participate. After excluding controls matched to cases with bordeline tumours, 280 hospital controls and 259 population controls were included in the analysis (Wu et al., 1988). Exposure to talc was categorized by type of application (perineum only, sanitary pads only, diaphragm only, any two types of application or all three types of application), duration of use before tubal ligation (none, 1-9 years, > 10 years, unknown) and frequency of use (none, 1-20 applications per month, > 20 applications per month, unknown). Conditional logistic regression was used to calculate the odds ratio for each exposure and to test for trend. Ninety-seven cases (51.6%) and 247 controls (45.8%) reported previous use of talcum powder on the perineum to yield an odds ratio o f 1.40 (P = 0.06) after adjustment for parity. Since the odds ratios were similar when hospital-based and population-based controls were analysed separately, analyses using the combined group of controls were reported. After adjustment for parity and oral contraceptive use, the odds ratio for use of talc on the perineum only was 1.5 (95% Cl, 0.8-2.6). No significant associations were observed with either individual or multiple types of perineal talc use, including the combination of use on the perineum, sanitary napkins and a diaphragm (odds ratio, 1.4; 95% Cl, 0.9-2.0 for any two types of use versus 0.4; 95% Cl, 0.0-2.9 for all three types combined). No TALC 363 significant trend was observed with duration of talc use on the perineum before tubal ligation or hysterectomy. Odds ratios were 1.6 (95% Cl, 1.0-2.6) for 1-9 years of exposure and 1.1 (95% Cl, 0.7-1.7) for more than 10 years of exposure. A non-significant trend of increased risk with increasing frequency of perineal use o f talc was observed, with an overall odds ratio of 1.3 (95% Cl, 0.9-1.9; T5= 0.19) for 30 applications per month. When stratified by history of perineal use of talc (yes/no) and history of tubal ligation or hysterectomy (yes/no), women who had used talc perineally and but had not undergone surgery for sterilization had the highest risk for ovarian cancer (odds ratio, 1.3; 95% Cl, 0.9-2.0). [Limitations of this study included the lack o f information on talc use.] Booth et al. (1989) reported results of a hospital-based case-control study of the risk for ovarian cancer conducted in 15 hospitals in London and Oxford (United Kingdom) from October 1978 to February 1983. Women aged 65 years or under at diagnosis and who were diagnosed within 2 years of the study interview were eligible for inclusion. A total o f 280 potential cases were identified, interviewed and classified with respect to tumour histology. After excluding 45 women, 235 cases were included in the analysis. A total o f 451 controls with the same age distribution as the cases were selected from the same 15 hospitals. Controls had a range of admission diagnoses; gastrointestinal disease (n = 105) and bone or joint disease (n = 70) were the most common. Women were excluded as controls if they had a history of bilateral oophorectomy or if they had a condition related to oral contraceptive use or other reproductive factors. Participation rates were not provided. Interviewers used a standard questionnaire to obtain information on reproductive and menstrual history, as well as exposure to exogenous estrogens, cigarettes and talc. Talc exposure was categorized according to the frequency o f perineal use (never, rarely, monthly, weekly or daily) and whether it was used for storage of a diaphragm. Multiple logistic regression adjusted for age and socioeconomic status was conducted. Fifty-seven cases [24.3%] and 77 controls [17.1%] reported a history of weekly use of talc in the genital area, while 71 cases [30.2%] and 139 controls [30.8%] reported daily use. Weekly genital use of talc was associated with a significantly increased risk for ovarian cancer (odds ratio, 2.0; 95% Cl, 1.3--3.4), while daily use was associated with a non-significant increase in risk (odds ratio, 1.3; 95% Cl, 0.8-1.9), after adjustment for age and socioeconomic status. The /7-value for trend with increasing frequency of use was of borderline significance (P = 0.05). The percentage of diaphragm users who reported storing their diaphragm in talc was not significantly different between the cases (86%) and controls (81%). [Limitations of this hospital-based study included the limited information on talc use. As participation rates were not provided, the possibility of selection bias is difficult to evaluate. Although covariates such as oral contraceptive use or parity were available, it was not explicitly stated if they were evaluated.] Harlow and Weiss (1989) conducted a study of perineal use of powder and the risk for borderline ovarian cancer in western Washington State, USA. Cases were 116 Caucasian women aged 20-79 years who had been diagnosed with borderline serous or mucinous epithelial ovarian cancer between 1980 and 1985, and who were identified by International Classification of Diseases-0 codes obtained from a population-based 364 IARC MONOGRAPHS VOLUME 93 cancer-reporting system. Controls were identified from the same counties of residence by random-digit dialling. A total of 158 women with a similar age distribution to the cases and who had not undergone a bilateral oophorectomy were included in the analysis. Cases and controls were interviewed in-person to obtain information on reproductive, sexual and medical histories, as well as on perineal exposure to talc (through multiple openended questions about the history of powder use of the participant). Among all eligible cases and controls identified for the study, 68% o f the cases and 74% of the controls were interviewed. The authors controlled for age (20-39, 40-59 or 60-79 years), parity (nulliparous or parous) and oral contraceptive use (ever/never). Exposure to talc was broadly categorized as `any perineal use of dusting powders' (after bathing, on sanitary napkins or for diaphragm storage) and further subcategorized according to method of use (diaphragm storage only, after bathing only, sanitary napkins only, after bathing and on sanitary napkins and specific combinations of the various methods) and type of powder used (cornstarch only, baby powder only, talc unspecified (no combined use), deodorizing powder only or combinations o f powders). Forty-nine cases [42.2%] and 64 controls [40.5%] reported a history of `any perineal exposure to powder' to yield an odds ratio of 1.1 (95% Cl, 0.7-2.1). When analysed by the type of powder used, the risk for borderline ovarian cancer was elevated only for perineal use of deodorizing powder alone (odds ratio, 3.5; 95% Cl, 1.2-28.7) or in combination with other powders (odds ratio, 2.8; 95% Cl, 1.1-11.7). No association was noted for the use of baby powder alone (odds ratio, 0.8; 95% Cl, 0.4-1.9) or for combined use (odds ratio, 0.9; 95% Cl, 0.5-2.0) or for other unspecified use of talc (odds ratio, 1.0; 95% Cl, 0.4--2.4). No significant association was found between risk for borderline tumours and any individual method of powder use, including use after bathing, on sanitary napkins or for diaphragm storage. The authors reported no increase in risk with increasing number of days of powder use, although the data were not provided in the paper. [Limitations of this study included the incomplete information on powder use and its small size.] Chen et al. (1992) (described in detail in Section 2.1.2) conducted a case-control study in Beijing, China, of several risk factors for epithelial ovarian cancer that included perineal exposure to talc (yes/no use of dusting powder to the lower abdomen or perineum for 3 or more months). The analysis was carried out on 112 newly diagnosed cases identified between 1984 and 1986 through the Beijing Cancer Registry and 224 age-matched community controls (two controls per case). Seven cases [6.3%] and five controls [2.2%] reported use o f talc-containing powders which resulted in an odds ratio o f 3.9 (95% Cl, 0.9-10.6) after adjustment for education and parity. [The Working Group noted the incomplete ascertainment of cases of ovarian cancer due to the nature of the cancer-reporting system in China, the large number of cases that were excluded due to death and the exclusion of controls who had a histoiy of serious health problems (which may have resulted in selection bias), the limited information on perineal use of talc, the lack o f adjustment for other potential confounding variables, the small number of cases and the low prevalence of talc use.] TALC 365 Harlow et al. (1992) analysed perineal exposure to talc and the risk for ovarian cancer among 235 cases and 239 controls in the Boston, MA metropolitan area (USA). Cases were diagnosed with ovarian cancer between June 1984 and September 1987 at one of 10 Boston hospitals and controls were identified from town registers listing the name, age and address o f all residents in Massachusetts. All cases were Caucasian women aged 18-- 76 years at diagnosis and were similar to the controls with respect to race, age and area of residence. O f 397 cases identified during the study period, 31 % were not interviewed due to physician and/or patient refusal, death or change of address. After excluding women whose cancer diagnosis was not confirmed by an independent pathology review [9.4% of eligible cases], 235 women were included in the analysis. A total of 526 women were contacted as potential controls. O f these, 239 [45.4%] were interviewed, 25% could not be reached, 10% reported a previous bilateral oophorectomy and 19% did not wish to participate in the study. In-person interviews were conducted with cases and controls to obtain information on occupational history, medical and reproductive histories, dietary history, cigarette smoking and hygienic practices (use of douches, types of sanitary protection used, perineal exposure to talc). Exposure to talc was categorized on the basis of `any' exposure, the method of application (dusting on sanitary napkins and/or underwear, via partner or application to diaphragm, dusting on perineum), the brand used, age at first use, duration and frequency of use. Total lifetime exposure to talc was estimated by cumulating the frequency of exposure and years of use to arrive at a summary measure of the total number of applications (< 1000, 1000-10 000, > 10 000). Covariates evaluated as potential confounders included age, education, marital status, religion, weight, use of oral contraceptives and parity; of these, age, education (< 12 years, > 12 years), marital status (never/ever), religion (Jewish, non-Jewish), weight (< 140 lb, > 140 lb) and parity (0, 1-2, > 2) were included in all multivariable models. A history of `any' perineal exposure to talc-containing powders was reported by 48.5% of cases and 39.3% of controls to yield an odds ratio of 1.5 (95% Cl, 1.0-2.1). When the method of application was examined, only direct application to the perineum as a dusting powder was associated with a significant increase in risk (odds ratio, 1.7; 95% Cl, 1.1-- 2.7). Women who reported at least 30 applications of talcum powder per month had a significant increase in risk (odds ratio, 1.8; 95% Cl, 1.1-3.0), while women with fewer applications per month did not. A significant positive trend was seen with number of monthly applications (P --0.046). Women with at least 10 years of perineal exposure had a borderline significant increase in risk (odds ratio, 1.6; 95% Cl, 1.0-2.7) and the /-value for trend was also of borderline significance (P = 0.07). Analyses stratified by age at first use indicated that women who first used talc genitally before the age of 20 years had the highest risk (odds ratio, 1.7; 95% Cl, 1.1-2.7); those stratified by years since last use suggested that women with the most recent perineal use of talc (within the previous 6 months) had the highest risk (odds ratio, 2.3; 95% Cl, 1.3-4.0). In an analysis stratified by use before versus after 1960, women who reported some perineal use of talc before 1960 had a significantly elevated risk for ovarian cancer (odds ratio, 1.7; 95% Cl, 1.1 2.7), while women with exclusive genital use of talc after 1960 did not (odds ratio, 1.1; 366 IARC MONOGRAPHS VOLUME 93 95% Cl, 0.6-2.1). Women who had used more than 10 000 lifetime applications had a borderline significant increase in risk (odds ratio, 1.8; 95% Cl, 1.0-3.0). This was unchanged after excluding applications that occurred after tubal ligation or hysterectomy (odds ratio, 1.7; 95% Cl, 1.0-3.0). However, when use of talc during non-ovulatory periods and after surgical sterilization was excluded, the increase in risk associated with more than 10 000 lifetime applications was significant (odds ratio, 2.8; 95% Cl, 1.4--5.4). In analyses of each histological type and grade, the strongest associations were seen for endometrioid tumours (odds ratio, 2.8; 95% Cl, 1.2-6.4) and tumours of borderline invasiveness (odds ratio, 2.4; 95% Cl, 1.2-4.5) (Table 2.4). Rosenblatt et al. (1992) conducted a hospital-based case-control study among 77 women who were hospitalized at Johns Hopkins Hospital in Baltimore, MD (USA) for ovarian cancer (cases) and 46 who were hospitalized for non-gynaecological, non malignant conditions (controls). The cases were newly diagnosed with pathologically confirmed epithelial ovarian cancer between 1981 and 1985, the majority of whom were aged 40-69 years. O f 140 eligible cases, 108 (77.1%) were interviewed. Thirteen were subsequently excluded because no control was identified and 18 were excluded for an unspecified reason. Controls were matched to cases by age, race and date of diagnostic admission. Information on genital and respiratory exposure to fibre-containing substances (talc, asbestos and fibreglass), as well as potential confounders, was collected using a structured questionnaire which was administered in the hospital and by telephone. Covariates that were considered to be potential confounders included tobacco use, `ovulatory time period', parity, family history of cancer, obesity, education, education of husband, previous history of cancer, marital status, religion and the use of oral contraceptives and other methods of contraception. Sources of genital fibre exposure (yes/no) included diaphragm use and dusting of either the perineum or sanitary napkins with talcum powder. Potential sources of respiratory fibre exposure (yes/no) included use of face or body powders containing talc, insulation installed at residence and living in the vicinity of or employment in a fibre-emitting industry (such as shipyard, asbestos or talc mine, asbestos/talc/fibreglass processing plant). A large percentage of both the cases (87%) and controls (88%) reported exposure to genital fibre, with an odds ratio of 1.0 (95% Cl, 0.2-^LO) after adjustment for parity. A long duration of genital fibre use (median duration, > 37.4 years) was associated with a borderline significant increase in the risk for ovarian cancer (odds ratio, 2.4; 95% Cl, 1.0-5.8) after adjustment for religion. Odds ratios were also calculated for genital use of bath talc (odds ratio, 1.7; 95% Cl, 0.7--3.9), use of talc on sanitary napkins (odds ratio, 4.8; 95% Cl, 1.3-17.8) and use of talc on a diaphragm (odds ratio, 3.0; 95% Cl, 0.8-10.8). No association was observed between risk for ovarian cancer and history of previous gynaecological or abdominal surgery that may have resulted in peritoneal exposure to talc. [Limitations of this study included the very small number of cases and controls, the broad definition of fibre exposure used in certain exposure variables and the limited information on perineal exposure to talc.] Tzonou et al. (1993) conducted a hospital-based case-control study of risk factors for epithelial ovarian cancer in the Greater Athens region o f Greece. The cases were 189 women Table 2.4. Perineal talc use and ovarian cancer risk: by tum our histology References No. of cases Histology Relative risk3 (95% Cl) TALC Harlow e t al. (1992) 60 17 18 Chang & Risch (1997) 254 80 74 Cook e t al. (1997) 131 43 36 Cramer e t al. (1999) 229 83 130 Wong e t at. (1999) 136 11 21 Gertig e t al. (2000) 76 Mills e ta l. (2004) 42 10 14 Serousb Mucinous Endometrioid Serousb Mucinous Endometrioid Serous Mucinous Endometrioid Serous invasive Mucinous Endometrioid/clear cell Serous Mucinous Endometrioid Serous invasive Serous invasive Mucinous invasive Endometrioid 1.4 (0.9--2.2) 1.2 (0.6--2.5) 2.8(1.2-6.4) 1.3 (1.0-1.9) 1.6(1.0-2.6) 1.7(1.0-2.8) 1.7(1.1-2.5) 0.7 (0.4-1.4) 1.2 (0.6--2.3) 1.7(1.2-2.4) 0.8 (0.4-1.4) 1.0 (0.7-1.6) 1.2 (0.7-2.1) 1.5 (0.6-4.0) 1.4 (0.7--2.7) 1.4 (1.0-1.9) 1.8(1.1-2.8) 2.6 (0.9--7.4) 1.3 (0.6--2.6) Cl, confidence interval aAny or ever use of talc bIncludes borderline and invasive serous tumours OJ >-0 368 IARC MONOGRAPHS VOLUME 93 under 75 years of age who underwent surgery for ovarian cancer at one of two cancer hospitals in Athens between June 1989 and March 1991. The controls were 200 women under 75 years of age who were residents of Greater Athens and who visited patients hospitalized in the same wards as the cases during the study period. Ninety per cent of the eligible cases and 94% of the eligible controls agreed to participate. In-hospital interviews were conducted to collect information on a range of demographic, socioeconomic and reproductive factors, as well as information on exposure to hair dyes, analgesics, tranquilizers and talc. Exposure to talc was assessed qualitatively as `yes/no' application of talc in the perineal region. In multivariable analyses, models were adjusted for age in 5-year groups, education, weight, age at menarche, menopausal status, age at menopause, parity, age at first birth, tobacco smoking status, alcohol use, coffee consumption and the other exposures of interest (use of analgesics, tranquilizers and hair dyes). Application of talc to the perineal region was reported by six cases [3.2%] and seven controls [3.5%] to yield an odds ratio of 1.1 (95% Cl, 0.3^1.0) after adjustment for the potential confounders. [Limitations of this hospital-based case-control study included the very low prevalence of perineal use of talc.] Purdie et al. (1995) conducted a case-control study among women in the three most populous Australian states-- Queensland, New South Wales and Victoria. Cases were women, aged 18-79 years, who had been diagnosed with epithelial ovarian cancer between August 1990 and December 1993 at gynaecological oncology treatment centres in one o f these three regions. Women were excluded if they had a metastatic tumour, were outside the eligible age range, could not be contacted, were too ill or were incapable of completing the questionnaire in conjunction with a trained interviewer (because of language difficulties or psychiatric conditions). Each case was confirmed by an independent pathological review of tissue specimens. O f 1116 cases identified during the study period, 201 (18%) were ineligible (e.g. due to a non-ovarian primary cancer or age at diagnosis). Among the 915 eligible cases, 824 (90%) agreed to participate and were interviewed. Reasons for non-participation included death before interview (50 cases), patient refusal (34 cases) and physician refusal (seven cases). Controls were identified from the electoral roll and were similar to the cases in age distribution and area of residence. Women were excluded as a control if they had a history of ovarian cancer or bilateral oophorectomy, could not be reached or could not complete the questionnaire. Among 1527 potential controls identified from the electoral roll, 1178 were located and found to be eligible (77%). Of these, 860 agreed to participate in the study (73% of the eligible controls). Reasons for ineligibility among the controls included failure to locate the individual (192), inability to complete the questionnaire due to language difficulties, a psychiatric condition, illness or death (105), previous bilateral oophorectomy (48) and age (four). Trained interviewers used a standardized questionnaire to collect information on medical, reproductive, family and occupational histories, as well as data on dietary factors and history of talc use. Questionnaires were administered face-to-face either in the clinic (for cases) or in the home of participant (for some cases and all controls). Covariates evaluated as potential confounders included parity, hysterectomy, tubal ligation, duration TALC 369 of oral contraceptive use, age, education, body mass index, tobacco smoking status, family history of cancer and multiple menstrual and reproductive factors. Talc use around the abdomen or perineum was reported by 56.7% of cases and 52% of controls to yield an odds ratio of 1.3 (95% Cl, 1.0-1.5) after adjustment for parity. Although enrolment in the electoral roll is mandatory in Australia, the authors determined that 28 cases [3.4%] had never enrolled and the enrolment status could not be confirmed for 46 cases [5.6%]. The results did not change when the analyses were limited to cases with confirmed enrolment in the electoral role. Green et al. (1997) evaluated the association between tubal ligation or hysterectomy and the risk for ovarian cancer using the Australian study population described by Purdie et al. (1995). [The analysis by Green et al. (1997) used the same number of cases but five fewer controls than Purdie et al. (1995).] Duration o f talc use was calculated as age at first reported use until age at occurrence of the earliest of any of the following events: surgical sterilization, reported last use of talc, diagnosis or interview. A modest increase in risk for ovarian cancer was observed with peritoneal use of talc (odds ratio, 1.3; 95% Cl, 1.1-1.6). Neither duration of talc use nor age at first use were associated with risk for ovarian cancer, although the relative risks (95% Cl) were not provided and the duration categories evaluated were not specified. When compared with women with no history of genital exposure to talc and patent fallopian tubes, women with a history o f talc use and no history of surgical sterilization had the highest risk for ovarian cancer (odds ratio, 1.3; 95% Cl, 1.0-1.7), while women with a history of tubal ligation or hysterectomy and no talc use had the lowest risk (odds ratio, 0.6; 95% Cl, 0.5-0.8). [The primary limitation of this study was the restricted information on perineal use of talc.] Shushan et al. (1996) examined the association between exposure to fertility drugs and the risk for ovarian cancer among 200 cases of epithelial ovarian cancer (164 invasive and 36 borderline) and 408 controls. All participants were living in Israel and were 36-- 64 years of age at enrolment into the study. Cases were identified through the Israel Cancer Registry from January 1990 to September 1993. Among 287 women who met the eligibility criteria (histologically confirmed diagnosis, cancer diagnosed and reported during study period, bom between 1929 and 1957 and alive at time o f interview), 87 (30.3%) were excluded because of inability to locate the patient or physician (25%), illness (1%), refusal by the physician (1%) or refusal by the patient (3%). Controls were identified by random-digit dialling and were matched to the cases by geographical area. Women were eligible to be included as a control if they were bom in the same period as the cases. Potential controls were excluded if they had a history o f bilateral oophorectomy (1%). O f 2072 telephone calls that successfully reached a household member, approximately half of the households [47.8%] contacted had a potentially eligible woman who was at home. O f these, 16.2% refused to participate and 10.7% were excluded because the woman did not speak Hebrew. Trained interviewers administered a standard questionnaire to all cases and controls. The questionnaire collected detailed information on reproductive history, use of oral contraceptives and fertility drugs, as well as exposure to talc (never/seldom, moderate/a lot). Although the main association of interest was use 370 IARC MONOGRAPHS VOLUME 93 of fertility drugs and the risk for ovarian cancer, the authors reported that 21 cases (10.5%) and 23 controls (5.6%) had a history o f moderate or frequent use of talc, which yielded an unadjusted odds ratio of [1.97] (P = 0.04). [Limitations of this study included the very sparse information on talc use and the unavailability of adjusted results for the association between use of talc and the risk for ovarian cancer.] Chang and Risch (1997) analysed the association between perineal use of powder and the risk for ovarian cancer among 450 cases and 564 population controls from metropolitan Toronto and southern Ontario, Canada. Cases were diagnosed between November 1989 and October 1992 and were between the ages of 35 and 79 years at entry into the study. O f 631 cases identified during the study period, 71.3% (450) were interviewed and included in the analysis. Reasons for non-participation included death (8.7%), physician refusal (4.6%), severe illness (4.8%), loss to follow-up (2.7%) and patient refusal (7.9%). Potential controls were identified through records of the Ontario Ministry of Finance based on their residence and age, were matched to cases within 15-year age groups and were excluded from the study if they had a history of bilateral oophorectomy more than 1 year before entry into the study. Among 873 eligible controls identified, 309 [35.4%] did not participate. Reasons included participant refusal (30.2%), illness (1.9%) or loss to follow-up (3.2%). Interviewers administered a standard questionnaire during an in-home interview to obtain information on the histoiy, frequency and duration of use of talcum and cornstarch powder, as well as multiple medical and reproductive covariates of interest. Talc exposure was categorized on the basis of `any' exposure in the perineal area, on the method o f application (directly to the perineum after bathing or showering, dusting on sanitary napkins), on the frequency of application (< 10, 10-25, > 25 applications per month) and on the duration of exposure (<30, 30-40, > 40 years of use). Multiple logistic regression was used in the analyses, with adjustment for age, duration of oral contraceptive use, parity (defined as the number of full-term pregnancies), duration of lactation for each pregnancy, histoiy of tubal ligation or hysterectomy and family history of breast or ovarian cancer. Forty-four per cent of cases and 36% of controls reported `any' talc use in the perineal area to yield an odds ratio of 1.4 (95% Cl, 1.1-1.9). Among the specific types of talc exposure, application to the perineum after bathing was associated with a borderline significant increase in risk (odds ratio, 1.3; 95% Cl, 1.0-1.7), while application on sanitary napkins (a less common use in this study population) was associated with an elevated but non-significant increase in risk (odds ratio, 1.3; 95% Cl, 0.9-2.0). A borderline significant trend was seen with increasing duration of exposure to talc (odds ratio per 10 years of exposure, 1.1; 95% Cl, 1.0-1.2), but not with increasing frequency of exposure. An analysis of duration by categoiy (< 30, 30-40, > 40 years) did not suggest a dose--response relationship (odds ratios of 1.0; 1.7; 95% Cl, 1.1-2.6; 1.4; 95% Cl, 1.0-2.2 and 0.9; 95% Cl, 0.5-1.4, respectively). Use of cornstarch in the perineal area, either alone or in conjunction with occasional talc, was not associated with the risk for ovarian cancer, although prevalence of use was low (less than 2% of subjects). To evaluate exposure pre- and post-1970, as well as exposure pre- and post-tubal ligation or hysterectomy, the authors assumed that participants initiated TALC 371 penneal use of after-bath talc at the age o f 20 years. A similar, non-significantly elevated, risk for ovarian cancer was seen for use pre- and post-1970. A higher odds ratio was seen for use of after-bath talc before tubal ligation or hysterectomy (odds ratio, 1.1; 95% Cl 1.0-1.2) than for use after these surgical procedures (odds ratio, 1.0; 95% Cl, 0.8-1.3). These estimates did not change when different starting ages, between 15 and 24 years, were used in the analysis. The authors also evaluated the association between perineal use of talc and invasive and borderline cancers separately, and found that the risk was elevated for both tumour types but was significant only for invasive tumours. In addition risk was similar across the major histological subtypes o f ovarian cancer (serous' mucinous, endometrioid) (see Table 2.4). [Limitations of this study included the lack of information on use o f talc.] Cook et al (1997) evaluated the association between use of genital powders or deodorants and the risk for ovarian cancer in a case-control study conducted in three counties of western Washington State, USA. Cases were aged 20-79 years at diagnosis, were diagnosed with borderline or invasive epithelial ovarian cancer between 1986 and 1988 and were identified using the population-based Cancer Surveillance System of western Washington. Controls were identified using random-digit dialling, were residents of the three counties of interest and were similar in age to the cases. Among 512 eligible cases^ identified, 329 were interviewed (64.3%) and 313 were included in the analysis [61.1%]. A total of 183 eligible cases were not interviewed due to death (104), physician or patient refusal (73) or loss to follow-up (six). An additional 16 cases who were interviewed were excluded from the analysis because of non-white race (seven) and unknown genital use of powder (nine). Among 721 women identified as potential controls, 521 were interviewed (72.3%) and 422 were included in the analysis [58.5%]. Reasons for excluding interviewed controls from the analysis included: non-white race (28), age greater than 79 years (five), history of bilateral oophorectomy (58), unknown oophorectomy status (four) and unknown genital use of powder (four). Information on powder use, including the type, method, frequency and duration o f use, and the covariates of interest was collected during in-person interviews. Covariates considered to be potential confounders in multivariable analyses included age, education, income, marital status, body mass index, oral contraceptive use and parity. A history o f `any' lifetime genital powder use (perineal dusting, diaphragm storage, use on sanitary napkins or use of deodorant spray) was reported by 50.8% of cases and 39.3% o f controls to yield an odds ratio of 1.5 (95% Cl, 1.1--2.0) after adjustment for age. Among the individual methods of genital use of powder, risk was significantly elevated only for exclusive perineal dusting (odds ratio, 1.8, 95% Cl, 1.2--2.9) after adjustment for age. In analyses adjusted for age and other types of genital use of powder, both perineal dusting (odds ratio, 1.6; 95% Cl, 1.1-2.3) and genital deodorant spray (odds ratio, 1.9; 95% Cl, 1.1-3.1) were associated with risk for ovarian cancer, while use of powder on a diaphragm or on sanitary napkins was not associated with an increased risk. There was no evidence of an increasing trend in risk with greater duration of perineal dusting, but a significant positive trend was noted for both duration (odds ratio, 2.7; 95% Cl, 1.1-6.6 for> 12 cumulative lifetime months;/! for 372 IARC MONOGRAPHS VOLUME 93 trend < 0.05) and number of lifetime applications (odds ratio, 2.6; 95% CI, 0.9-7.6 for > 500 lifetime applications; p for trend < 0.05) o f genital deodorant spray. The effect estimates did not change materially when perineal use of dusting powder after the date of tubal ligation or hysterectomy was excluded. Risk was significantly elevated among women with any history of perineal dusting before 1976 (odds ratio, 1.8; 95% Cl, 1.1 2.9), but the authors were unable to evaluate risk for use exclusively after 1976 due to the small number of women (four cases and 10 controls) who had had this exposure. Among the individual types of powder evaluated (cornstarch, talcum powder, baby powder, deodorant powder, scented body/bath powder), risk for ovarian cancer was non significantly elevated for `any' use of talcum powder (odds ratio, 1.6; 95% CI, 0.9-2.8) and bath/body powder use (odds ratio, 1.5; 95% CI, 0.9-2.4) after adjustment for age and other types of powder use (yes/no). The authors also evaluated the association between any genital use of powder and the risk for the major histological subtypes of ovarian cancer (see Table 2.4). Risk was significantly elevated for serous tumours (odds ratio, 1.7; 95% CI, 1.1-2.5) and all other tumour types (odds ratio, 1.8; 95% CI, 1.1-2.8) but not for mucinous or endometrioid tumours. [Limitations of this study included the relatively low participation rates among the cases and controls.] Eltabbakh et al. (1998) compared risk factors among 50 cases of primary extra ovarian peritoneal carcinoma (the `study' group) and 503 cases of primary epithelial ovarian cancer (the `control' group) treated at Roswell Park Cancer Institute in Buffalo, NY (USA), between October 1982 and October 1996. No healthy controls were enrolled in this study. Diagnoses were reviewed by staff in the Division of Pathology (study and control groups) and were confirmed by a single pathologist as part o f another study (study group only). Information on reproductive history, menstrual history, use of hormones and contraceptives and personal hygiene was collected through a self-administered, 44-item questionnaire which all patients were asked to complete during the hospital admission process. All women who returned a questionnaire were eligible to be included in the study. Among these patients, the overall questionnaire response rate was 60%. Response was inversely correlated with severity of disease and response rates were similar for the two diagnoses included in this study. Because data on perineal talc use was missing for 37 patients in the `control' group, only 466 ovarian cancer patients were included in the analysis. Women who had primary ovarian cancer were significantly more likely to report a history of perineal use of talc compared with women who had primary peritoneal cancer (48.1% versus 26.0%; [crude odds ratio = 2.6] /> = 0.003). Among the other characteristics examined, only age and age at menarche differed significantly in the two groups. [Limitations of this study included the minimal information on talc use, the low questionnaire response rate among study participants, particularly among the patients with more advanced disease, the use of a self-administered questionnaire completed during the admissions process, which may have limited the quality of the responses, and the lack of a `healthy' comparison group.] Godard et al. (1998) evaluated risk factors for familial and sporadic ovarian cancer in a population of French Canadian women in Montral, Quebec (Canada). O f 231 cases TALC 373 who were identified between 1995 and 1996 at two gynaecological oncology clinics in Montral, 183 (79.2%) were interviewed and 170 (73.6%) were included in the analysis. Reasons for non-inclusion were death ( = 21), refusal/unavailability to participate (,n = 12), loss to follow-up (n = 15) and tumours were non-epithelial in origin (n = 13). All cases were between the ages of 20 and 84 years at diagnosis, with a mean age at diagnosis of 53.7 years and a mean age at interview of 55.9 years. Controls were identified using a modified random-digit dialling method and were frequency-matched to cases by age (within 1 year) and French Canadian ethnicity. The mean age at interview for the controls was 56.7 years. Among 750 households contacted regarding participation in the study, 66.7% ( = 500) either did not have an eligible female resident or did not reply to the researchers' inquiries and 10.7% refused to participate. A total of 170 women were interviewed and included in the analysis as controls. A standardized 57-item questionnaire was used to obtain information on the family, medical and reproductive history of each participant. Cases were interviewed either by telephone (30%) or in the study clinics (70%). No information was given on the methods of interview for control subjects. Information on family history of cancer was collected to determine whether risk factors differed for the sporadic and familial cases of ovarian cancer. Familial cases were those patients who had one or more family members (first, second or third degree relatives) with breast cancer diagnosed before 55 years of age or ovarian cancer diagnosed at any age. Sporadic cases were those patients who had no family members with breast cancer diagnosed before 55 years of age or with ovarian cancer diagnosed at any age. Perineal exposure to talc was assessed qualitatively (ever/never, with `never' as the baseline). Covariates that were considered to be potential confounding variables were age at menarche, age at menopause, parity, age at first and last childbirth, duration of oral contraceptive use, age at last oral contraceptive use, tubal ligation, alcohol use and previous breast or abdominal surgery. Talc exposure was more common in cases than controls, with 10.6% of the cases and 4.7% of the controls reported perineal use of talc (P = 0.06). No difference between perineal use of talc was reported in the familial and sporadic cases (P = 0.79). Multivariate analyses were performed comparing all cases, (all, sporadic, familial) with controls. In these analyses, perineal use of talc was associated with a non-significant increase in the total risk for ovarian cancer (odds ratio, 2.5; 95% CI, 0.9-6.6; P = 0.07). Risk was similarly non-significantly elevated for sporadic (odds ratio, 2.5; 95% Cl, 0.9-7.1) and familial cases (odds ratio, 3.3; 95% Cl, 0.9-12.4) compared with the controls. [Limitations of this study included its small size and the lack of any detailed information on perineal use of talc. The control participation rates may have been low (although this is not clear) and it is not certain how representative the controls were.] Cramer et al. (1999) analysed the association between genital exposure to talc and the risk for primaiy epithelial ovarian cancer among 563 cases and 523 controls residing in eastern Massachusetts and New Hampshire, USA. Cases were identified between May 1992 and March 1997 through hospital tumour boards or statewide cancer registries. Among 1080 cases diagnosed in this period (including borderline tumours), 203 (18.8%) 374 IARC MONOGRAPHS VOLUME 93 were excluded due to death, change of address, inability to speak English, no telephone in residence or a non-ovarian primary cancer. O f the 877 eligible cases remaining after these exclusions, 563 (64%) were included in the analysis. The remaining 314 cases were excluded because of physician refusal (n = 126) and patient refusal (n = 136). Pathology reports were reviewed to confirm the diagnoses for all cases, and slides were requested and reviewed in the case of discrepancies between the reported histology and the histology assigned based on the pathology report review. Controls were identified by random-digit dialling and town resident books (to identify additional women over the age o f 60 years who lived in Massachusetts) and were frequency-matched to cases by age (within 4 years) and location of residence. O f the potentially eligible controls, 72% of those identified by random-digit dialling and 49% of those identified through town books agreed to participate. All study participants were interviewed in-person using a standardized questionnaire to obtain information on their medical and reproductive histories, family history and personal habits. The questionnaire also asked multiple questions on powder use, including route of exposure (application to non-genital areas, application to perineum, sanitary napkins or underwear, husband's use of powders in his genital area), brand o f powder used (talc, cornstarch), age at first use, duration and frequency of use (< 30, 30-39, > 40 uses per month). Participants were asked about exposures that occurred at least 1 year before the date of diagnosis (cases) or the date of interview (controls). The results were adjusted for the following potential confounding variables: age, state of residence, body mass index, parity, oral contraceptive use, family history o f breast or ovarian cancer and history of tubal ligation. The prevalence of talc use was higher among cases than controls; 44.6% of cases and 36.1% of controls reported `any' use o f talc (included use in both genital and non-genital areas) and 27.0% of cases and 18.2% o f controls reported `genital' use o f talc (included dusting o f perineum/sanitaiy napkins/underwear, either exclusively or in combination). Talc use in non-genital areas was not associated with risk when compared with women who did not use personal powder (odds ratio, 1.1; 95% Cl, 0.8-1.5). However, genital use of talc was associated with a significant 60% increase in risk (odds ratio, 1.6; 95% Cl, 1.2-2.2). Women who reported more than one method of talc use in the genital area had an even greater risk for ovarian cancer (odds ratio, 2.2; 95% Cl, 1.3-3.6). No association was observed between genital use of talc and risk for ovarian cancer among women who had undergone tubal ligation after adjustment for age (odds ratio, 1.0; 95% Cl, 0.5-2.1). Because of the low prevalence of use (< 1% of the study population) of cornstarch, evaluation of this product was uninformative. When women who had been exposed to powder only in non-genital areas were excluded from the analysis, no linear trend was observed between risk for ovarian cancer and age at first genital use of talc, duration of use, frequency of use or total number of lifetime applications. However, when non-genitally exposed women were included in the analysis, a significant linear trend was observed with increasing number of lifetime applications, after talc applications that occurred during non-ovulatory years or after tubal ligation or hysterectomy were excluded (P = 0.02). Additional findings of interest included: a non-significant increase in risk among married women with no TALC 375 personal talc use whose husbands had used talc for genital hygiene (odds ratio, 1.5; 95% Cl, 0.9-2.5); and a stronger association between genital use of talc and risk for ovarian cancer among women who had used talc before their first live birth (odds ratio, 1.6; 95% Cl, 1.1-2.3) than for women who had used it exclusively after their first live birth (odds ratio, 1.0; 95% Cl, 0.4-2.5). The association with genital use of talc was strongest for serous invasive tumours (odds ratio, 1.7; 95% Cl, 1.2-2.4). No association was observed for endometrioid/clear-cell (odds ratio, 1.0; 95% Cl, 0.7-1.6) or mucinous tumours (odds ratio, 0.79; 95% Cl, 0.4-1.4) (see Table 2.4). Wong et al. (1999) reported the results of a case-control study conducted at Roswell Park Cancer Institute, Buffalo, NY (USA) o f 499 cases treated between October 1982 and October 1992 (largely those reported by Eltabbakh et al, 1998) and 755 hospital-based controls. The controls were randomly selected from a registry of patients who were being treated for non-gynaecological malignancies and were frequency-matched to cases by age at diagnosis (within 5 years). The most common diagnoses among controls were colorectal (43.3%) and skin cancers (34.5%) and leukaemia (17.7%). All participants completed the self-administered, 44-item questionnaire that all patients were asked to complete during the hospital admission process. All analyses were adjusted for age at diagnosis, parity, oral contraceptive use, tobacco smoking, family history of ovarian cancer, age at menarche, menopausal status, income, education, geographical location and history of tubal ligation or hysterectomy. The analysis was restricted to 462 cases and 693 controls with information on perineal use of talc. `Ever' use of talc (genital or non genital) was reported by 47.8% o f the cases and 44.9% o f the controls, while use o f talc in the genital or thigh area was reported by 34.0% of the cases and 32.2% of the controls. There was no association between any method of talc use and the risk for ovarian cancer after adjusting for several potentially confounding variables. The adjusted odds ratio for talc use in the genital or thigh area was 1.0 (95% Cl, 0.8-1.3). Duration o f talc use was similar in the cases and controls, and no association between talc use and the risk for ovarian cancer was found for any duration category. No significant association was observed between talc use and any of the major histological subtypes of ovarian cancer (see Table 2.4); the odds ratio for serous cystadenocarcinoma was 1.2 (95% Cl, 0.7-2.1). No evidence was found of effect modification by history of tubal ligation or hysterectomy. Among women who had not undergone tubal ligation or hysterectomy, the odds ratio for the association between talc use and risk for ovarian cancer was 1.2 (95% Cl, 0.8-1.6) while among women who had undergone tubal ligation or hysterectomy, the odds ratio was 0.8 (95% Cl, 0.5-1.2). [Limitations of the study included the sparse information on talc use. In addition, the use of hospital controls with non-gynaecological malignancies may have caused selection bias. As noted in the earlier report by Eltabbakh et al. (1998), the response rate to the questionnaire was low in this study population, particularly among the patients with more advanced disease.] Ness et al. (2000) examined whether factors related to an inflammatory response of the ovarian epithelium (such as exposure to talc, endometriosis, cysts and hyperthyroidism) played a role in the risk for ovarian cancer. The study was conducted 376 IARC MONOGRAPHS VOLUME 93 among 767 recently diagnosed cases o f epithelial ovarian cancer and 1367 population- based controls. Cases were aged 20-69 years and were identified between 1994 and 1998 a 39 hospitals in the Delaware Valley region (USA). O f 1253 potentially eligible cases .2 /o were lnterv.ewed and included in the analysis. Reasons for excluding women from the study included^!,agnosis more than 6 months before the interview ( = 296) severe illness or death (// - 69), unavailability of contact information (n = 15), physician refusal (n 14) or patient refusal (n = 92). Controls were identified through random-digit dialling ( o r controls <65 years of age) and Health Care Financing Administration lists (for ontiols 65 69 years of age) and were frequency-matched to cases by age and location of residence. Overall, 72/ of the eligible potential controls agreed to participate in the study A pathological review was conducted for a subset o f the cases ( = 120). When compared with the original diagnosis, the central review was 95% concordant for invasiveness and 82 /o concordant for cell type. The original pathological diagnosis was used in the analysis or all cases. A standardized, 1.5-hour interview was conducted in the homes of the participants to collect information on menstrual and reproductive histoiy, sexual activity use of contraceptives, histoiy and duration of talc use (genital and non-genital applications and exposure via male sexual partners). Talc use was categorized according to the method of application (never, feet, genital/rectal, sanitaiy napkins, underwear diaphragm or cervical cap, or male partner) and duration o f exposure (< 1 year, \-$ years' - 9 years, > 10 years). Unconditional logistic regression adjusted for age, parity, race' amily histoiy of ovarian cancer, oral contraceptive use, tubal ligation, hysterectomy and lactation was used in all analyses. A histoiy of talc use in the genital/rectal area was reported by 161 cases [21.0%] and 219 controls [16.0%] to yield an adjusted odds ratio of U ' U ~2 0 )- Significant associations were also observed for the use o f talc on . 0' . L6; 95% CI' L1~2-3) and u" ^ rw e a r (odds ratio, use of talc on the feet, arms or breasts was associated with a significant 40% increase in risk; however, women may also have used talc on more than one area of the body, including the genital and/or rectal area. Use o f talc on diaphragms or cervical caps and use by a male sexual partner were not associated with the risk for ovarian cancer. There was no clear trend between risk for ovarian cancer and increasing duration o f use of talc on the genital and/or rectal area or feet. Adjusted odds ratios of 2.0(95% CI, 1.0-4.0), 1.6 (95% CI, 1.1-2.3), 1.2 (95% CI, 0.8-1.9) and 1.2 (95% CI l.O -l.5) were observed for < 1year, 1 -4 years, 5 - 9 years and > 10 years of use' respectively. [Limitations o f this analysis included the sparse information on talc use In' analyses o f duration, the use of talc on the feet was also included as an exposure. The relatively low participation rates among cases was also a limitation of the study.] Langseth and Kjaerheim (2004) (described in detail in Section 2A.2(b)) evaluated the association between employment in the pulp and paper industry in Norway and the risk for ovarian cancer. In addition to the assessment of occupational exposure, information was collected on hygienic use of talc and potential confounders for a subset of the cases and controls during a personal interview conducted at the mills or by telephone Exposure to hygienic talc products was categorized as ever/never for personal use on diapers, TALC 377 sanitary napkins, underwear or husband's use in the genital area. Thirty-five cases and 102 of the eligible controls or their next of kin agreed to an interview and an additional 19 women who were not cases were interviewed and included in secondary analyses as supplementary controls. A family member completed the interview (due to the death of the case or control) for 25 of the cases and 31 of the controls. Use of talc on the genital area was reported by 12 cases and 53 controls to yield an odds ratio of 1.2 (95% Cl, 0.4 3.2). [The primary limitations o f this analysis were the small number of cases, the small percentage of cases and controls who were interviewed to obtain information on the covariates of interest and use of surrogate respondents to obtain information on covariates for the deceased cases and controls. The Working Group noted that hygienic exposure to talc was assessed retrospectively in the nested case-control study.] Mills et al. (2004) evaluated the association between perineal exposure to talc and the risk for ovarian cancer in an ethnically diverse population from 22 counties of central California, USA. The study included 256 incident cases diagnosed between 1 January 2000 and 31 December 2001 and identified through two regional cancer registries using rapid case ascertainment procedures and 1122 controls identified by random-digit dialling. Controls were frequency-matched to the cases by age and ethnicity. Pathology reports were reviewed centrally for a subset of the cases to confirm the diagnosis, subtype and invasiveness o f each cancer. Potential controls were ineligible for inclusion in the study if they were under 18 years o f age, were not a resident of the counties of interest or if they had a histoiy of epithelial ovarian cancer or bilateral oophorectomy. Among 652 cases identified during the study period, 263 (40.3%) were excluded due to: language or hearing difficulties ( = 17), death ( n - 76), physician refusal ( = 10), severe illness (n = 41) or unavailability of current contact information (n = 119). O f the 389 eligible cases who were contacted regarding participation in the study, 256 (65.8%) agreed to participate and were interviewed. O f a total of 2327 potential controls, 740 (31.8%) were excluded from the study due to: age ( = 80), location of residence ( = 21), language difficulties ( = 10), previous bilateral oophorectomy ( = 252), severe illness ( = 19) or change of address or telephone number or inability to contact the woman after repeated attempts (n = 358). O f the 1587 potential controls who were contacted and found to be eligible, 1122 (70.7%) agreed to participate and were interviewed. All cases and controls were interviewed by telephone to obtain information on their medical history, covariates of interest and history o f perineal exposure to talc, including the frequency, duration and calendar years of use. Information on talc use was unavailable for seven cases and 17 controls; thus, the final study population for this analysis included 249 cases and 1105 controls. For the final models, unconditional logistic regression adjusted for age, race/ethnicity, duration of oral contraceptive use and breastfeeding was used. Additional covariates considered to be potential confounders included family history of breast cancer or ovarian cancer, parity, history of pregnancy, body mass index, hysterectomy, tubal ligation and duration of postmenopausal use of hormones. A history of perineal talc use was reported by 42.6% of the cases and 37.1% of the controls to yield an adjusted odds ratio o f 1.4 (95% Cl, 1.0-1.9). A significant trend (P = 0.015) with increasing frequency 378 IARC MONOGRAPHS VOLUME 93 of talc use was observed. The greatest risk for ovarian cancer was observed among women with the highest frequency of use (odds ratio, 1.7 for use 4-7 times per week; 95% Cl, 1.1-2.6). There was a borderline significant trend with increasing duration of use (P = 0.045). The highest risk was observed among women with 4-12 years of use (odds ratio, 1.9; 95% Cl, 1.2-3.0) and elevated but non-significant risks were seen among women with longer durations of use with odds ratios of 1.5 (95% Cl, 0.9-2.3) and 1.2 (95% Cl, 0.7-2.1) for 13-30 and > 3 0 years of use, respectively. A borderline significant trend was noted for cumulative talc use (frequency times duration o f use), although this was also not clear-cut { P - 0.051). The highest risks were observed in the second and third quartiles of cumulative talc use. When examined according to the time of use, the risk was higher among women who had first used talc after 1975 (odds ratio, 1.9; 95% Cl, 1.3-2.9) than among those who had first used talc before or during 1975 (odds ratio, 1.2; 95% Cl, 0.8-1.8). Risk was also higher among women who were aged 20 years or more at first talc use than among those who were under 20 years of age and among women who initiated talc use after their first birth than among those who had some use before their first birth. When time since last use was examined, women who had last used talc 1 2 years previously had the highest risk (odds ratio, 2.4; 95% Cl, 1.4-4.1); women who had last used it 3-20 years previously had an elevated but non-significant risk for ovarian cancer (odds ratio, 1.6; 95% Cl, 0.9-2.7). Modification of the association between perineal use o f talc and risk for ovarian cancer by tubal ligation, hysterectomy, parity, oral contraceptive use, postmenopausal use of hormones and body mass index was also evaluated. Risk was higher among women who had not had tubal ligation (odds ratio, 1.5; 95% Cl, 1.1-2.2) than among those who had (odds ratio, 0.9; 95% Cl, 0.5-1.7), although the interaction was not statistically significant. Risk was also higher among women who had ever been pregnant (odds ratio, 1.4; 95% Cl, 1.1-2.0) than among those who had never been pregnant (odds ratio, 0.9; 95% Cl, 0.4-2.3) and among women who had no history of oral contraceptive use (odds ratio, 1.6; 95% Cl, 1.0-2.6) than among those who had used oral contraceptives (odds ratio, 1.3; 95% Cl, 0.9-1.8). No evidence was found of a modification of effect by hysterectomy status, body mass index or postmenopausal use of hormones. [Limitations of this study included the low participation rate and relatively small number of cases. In addition, pathology was not confirmed for all cases, which may have resulted in some misclassification of histological subtype.] 2.3 Use of talc in pleurodesis The use of talc or iodized talc to produce pleurodesis began in the 1930s as a treatment for recurrent spontaneous pneumothorax or pleural effusions. The therapy involves the introduction of 0.5-10 g talc directly into the pleura using intrapleural injection. In recent decades, the therapy has most commonly been restricted to use for the treatment of malignant pleural effusions. An individual case report described a lung adenocarcinoma that was diagnosed 2 years after pleurodesis with iodized talc (Jackson & Bennett, 1973). TALC 379 the Med l l R h r Committee o f the Bri^ h Thoracic Association and the Medical Research Council Pneumoconiosis Unit, 1979) of the long-term effects of pUurodesis with talc and kaolin among a series o f British patients who were followed for \4-40 yeais The one talc mentioned (BP Indian Finex) was reported not to contain fibrous amphiboles, but it was unclear if that was true of all the talcs used Three lung cancers were observed (2.14 expected, P > 0.3) among 210 talc pleurodesis patients Two of the lung cancer patients developed tumours on the opposite side from where treatment had occurred (18-month and 19-year intervals between treatment and death) The third patient had an oat cell carcinoma (site unknown) and died 32 years after treatment. No cases of mesothelioma were reported. Viskum e t al. (1989) reported on 99 Danish patients who had been treated in 1954-64 by pleurodesis with talc at doses that ranged from 0.5 to 4.9 g and who were followed for at least 20 years. Three deaths from lung cancer occurred [expected number of cases not provided], one on the side opposite from where treatment had occurred and two with no origin reported. No cases o f mesothelioma were reported. [The Working Group noted that these reports are difficult to interpret because of the high prevalence o f lung disease in the patient groups, which could be related to risk factors such as tobacco smoking. The type or source o f talc used was not clear, although it was assumed to be pharmaceutical grade. No case o f mesothelioma was observed but the number of expected cases would probablv be veiy low.] 2.4 References Blum S, Arp EW Jr, Smith AH, Tyroler HA (1979). Stomach cancer among rubber workers- an epidemiologic investigation. In: Lemen, R,, Dement, JM, eds. D u sts a n d D isease, P ro ceed in g s o j the C onference on O ccupational E xposures to F ibrous a n d P articulate D ust a n d Their E xtension into the E nvironm ent. Park Forest South, IL, Pathotox Publisher, Inc., pp. 325-334. Booth M, Beral V, Smith P (1989). Risk factors for ovarian cancer: a case-control study B r J C an cer, 60:592-598. PMID:2679848 Chang S, Risch HA (1997).Perineal talc exposure and risk of ovarian carcinoma C ancer M79.:P2M3I9D6-:2941091.15d2o9i:10.1002/(SICI)1097-0142(I9970615)79:12<239'6-AID-CNCR15>30C(>^2' Chen Y, Wu P C, Lang J-H e t al. (1992). Risk factors for epithelial ovarian cancer in Beijing China. Int J E pidem iol, 21:23-29. doi: 10.1093/ije/21.1.23. PMID: 1544753 ' Coggiola M, Bosio D, Pira E e t al. (2003). An update of a mortality study of talc miners and millers in Italy. A m J I n d M e d , 44:63-69. doi: 10.1002/ajim. 10240. PMID: 12822137 Cook LS, Kamb ML, Weiss NS (1997). Perineal powder exposure and the risk of ovarian cancer A m J E pidem iol, 145:459-465. PMID:9048520 Cramer DW, Liberman RF, Titus-Emstoff L e t al. (1999). Genital talc exposure and risk of ovarian cancer. Int J C ancer, 81:351-356. doi: 10.1002/(SICI) 1097-0215( 19990505)81 3<351 AID IJC7>3.0.CO;2-M. PMID: 10209948 ' 380 IARC MONOGRAPHS VOLUME 93 Cramer DW, Welch WR, Scully RE, Wojciechowski CA (1982). Ovarian cancer and talc: a case- ontrol study. C ancer, 50:372-376. doi:10.1002/1097-0142( 198207i 5)50 2<372 ATD CNCR2820500235>3.0.CO;2-S. PMID:7083145 U w u /n p u .2< 372..A ID - Eltaw l e G\ PiVT MS' .Natarajan N' Mettlin CJ (1998). Epidemiologic differences between women with extraovanan pnmaiy peritoneal carcinoma and women with epithelial ovarian r rtCanf"' G ynecoL 91;254-259. dot:10.1016/S0029-7844(97)00650-9. PMID9469285 Gertig DM Hunter DJ, Cramer DW e t al. (2000). Prospective study of talc use and ovarian cancer J N a tl C an cer Inst, 92:249-252. doi:10.1093/jnci/92.3.249. PMID-10655442 Godard B, Foulkes WD, Provencher D e t al. (1998). Risk factors for familial and sporadic ovarian cancer among French Canadians: a case-control study. A m J O b stet G yn eco l 179-403-410 doi: 10.1016/S0002-9378(98)70372-2. PMID:9731846 ' Green A, Purdie D, Bain C e t al.; Survey of Women's Health Study Group (1997). Tubal sterilisation, hysterectomy and decreased risk of ovarian cancer. Int J C an cer 71:948-951 doi: 10.1002/(SICI)l097-0215( 19970611)71:6<948::AID-IJC6>3.0.CO;2-Y PMID'9185694 Harlow BL, Cramer DW, Bell DA, Welch WR (1992). Perineal exposure to talc and ovarian cancer risk. O b stet G yn ecol, 80:19-26. PMID: 1603491 Harlow BL Weiss NS (1989). A case-control study of borderline ovarian tumors: the influence of perineal exposure to talc. A m J E pidem iol, 130:390-394. PMID:2750733 Hartge P, Hoover R, Lesher LP, McGowan L (1983). Talc and ovarian cancer. JAM A 250 1844 doi: 10.1001 /jama.250.14.1844. PMID:6620481 ' ' Hartge P, Stewart P (1994). Occupation and ovarian cancer: a case-control study in the ashington, DC, metropolitan area, 1978-1981. J O c c itp M ed, 36:924-927. PMID:7807277 JaCk^ Penaett, r H (1973)' Chest Wa" tUm0Ur following iodized talc pleurodesis. Thorax, 28.788-793. doi:10.1136/thx.28.6.788. PMID:4787992 Katsnelson BA, Mokronosova KA (1979). Non-fibrous mineral dusts and malignant tumors: an epidemiological study of mortality. J O ccu p M ed, 21:15-20. PMID:215733 Langseth H, Andersen A (1999). Cancer incidence among women in the Norwegian pulp and lndustly' A m J In d M e d 36:108-113. doi:10.1002/(SICI) 10970274( 199907)36:1 <108::AID-AJIM 15>3.0.CO;2-N. PMID: 10361594 Langseth H, Kjaerheim K (2004). Ovarian cancer and occupational exposure among pulp and paper employees in Norway. S ca n d J W ork E nviron H ealth, 30:356-361 PMID-15529799 Leophonte P, Basset MF, Pincemin J e t al. (1983). [Mortality of talc workers in France- a retrospective epidemiological study.] R ev F r M ai R espir, 11:489-490. Leophonte P Didier A (1990) French talc pneumoconiosis. In: Bignon, J, ed., H ealth E ffects o f rh y llo silic a te s, Berlin Heidelberg, Springer-Verlag, pp. 203-209. Mills PK, Riordan DG, Cress RD, Young HA (2004). Perineal talc exposure and epithelial ovarian cancer risk m the Central Valley of California. Int J C an cer doi:10.1002/ijc.20434. PMID:15382072 112-458-464 ' Ness RB Grisso JA, Cottreau C et al. (2000). Factors related to inflammation of the ovarian epithelium and risk of ovarian cancer. E pidem iology, 11:111-117. doi lO 1097/00001648200003000-00006. PMID:11021606 Purdie D, Green A, Bam C e t a l.; Survey of Women's Health Study Group (1995). Reproductive and other factors and risk of epithelial ovarian cancer: an Australian case-control study Int J C an cer, 62:678-684. doi: 10.1002/ijc.2910620606. PMID:7558414 TALC 381 Research Committee of the British Thoracic Association and the Medical Research Council Pneumoconiosis Unit; Research Council Pneumoconiosis U (1979). A survey of the long-term effects of talc and kaolin pleurodesis. B r J D is C h est, 73:285-288. doi lO 1016/00070971(79)90054-8. PM1D:553661 Rosenblatt KA, Szklo M, Rosenshein NB (1992). Mineral fiber exposure and the development of ovarian cancer. PMID: 1601331 G ynecol O ncol, 45:20-25. doi:10.1016/0090-825 8(92)90485-2 ' Rubino GF, Scansetti G, Piolatto G (1979) Mortality and morbidity among talc miners and millers in Italy. In: Lemen, R., Dement, J.M., eds. D u sts a n d D isease, P ro ceed in g s o f the C on feren ce on O ccupational E xposures to F ibrous a n d P articulate D ust a n d Their Extension into the Environm ent. Park Forest South, IL: Pathotox Publisher, Inc.,pp. 357-363. Rubino GF, Scansetti G, Piolatto G, Romano CA (1976). Mortality study of talc miners and millers. J O c c u p M ed, 18:187-193. doi:10.1097/00043764-197603000-00013. PMID:1255280 Selevan SG, Dement JM, Wagoner JK, Froines JR (1979). Mortality patterns among miners and millers of non-asbestiform talc: preliminary report. J E nviron P a th o l T oxicol 2 273-284 PMID:512559 ' ' Shushan A, Paltiel O, Iscovich J e t al. (1996). Human menopausal gonadotropin and the risk of epithelial ovarian cancer. F ertilS teril, 65:13-18. PMID:8557128 Siemiatycki J, editor (1991). R isk F actors f o r C a n cer in the W orkplace. Boca Raton, FL: CRC Press Straif K, Chambless L, Weiland SK e t al. (1999). Occupational risk factors for mortality from stomach and lung cancer among rubber workers: an analysis using internal controls and refined exposure assessment, fnt J E pidem iol, 28:1037-1043. doi lO 1093/ije/28 6 1037 PMID: 10661645 Straif K, Keil U, Taeger D el al. (2000). Exposure to nitrosamines, carbon black, asbestos, and talc and mortality from stomach, lung, and laryngeal cancer in a cohort of rubber workers. A m J E pidem iol, 152:297-306. doi: 10.1093/aje/l 52.4.297. PMID: 10968374 Thomas TL, Stewart PA (1987). Mortality from lung cancer and respiratory disease among pottery workers exposed to silica and talc. A m J E pidem iol, 125:35^13. PMID:3024482 Tzonou A, Polychi onopoulou A, Hsieh C-C e t al. (1993). Hair dyes, analgesics, tranquilizers and perineal talc application as risk factors for ovarian cancer. Int J C an cer, 55:408-410 doi: 10.1002/ijc.2910550313. PMID:8375924 ' Viskum K, Lange P, Mortensen J (1989). Long term sequelae after talc pleurodesis for spontaneous pneumothorax. P neum ologie, 43:105-106. PMID:2717548 Wergeland E, Andersen A, Baerheim A (1990). Morbidity and mortality in talc-exposed workers. A m J I n d M e d , 17:505-513. doi:10.1002/ajim.4700170408. PMID:2327417 Whittemore AS, Wu ML, Paffenbarger RS Jr e t al. (1988). Personal and environmental characteristics related to epithelial ovarian cancer. II. Exposures to talcum powder, tobacco, alcohol, and coffee. A m J E pidem iol, 128:1228-1240. PMID:3195564 ' Wild, P. (2000) [An epidemiolgical mortality study in the Talc-producing industry: Study Report] (INRS/EE Report TMT), Paris, Institut National de la Recherche Scientifique (in French) Wild P (2006). Lung cancer risk and talc not containing asbestiform fibres: a review of the epidemiological evidence. O ccu p E nviron M ed, 63:4-9. doi: 10.1136/oem 2005 020750 PMID: 16361399 ' 382 IARC MONOGRAPHS VOLUME 93 Wild P, Leodolter K, Reffegier M et al. (2002). A cohort mortality and nested case-control study of French and Austrian talc workers. O ccu p E nviron M ed ., 59:98-105. dor 10 1136/oem 59 2 98 PMID: 11850552 Wong C, Hempling RE, Piver MS e t al. (1999). Perineal talc exposure and subsequent epithelial ovarian cancer: a case-control study. O b stet G yn ecol, 93:372-376. doi lO 1016/S00297844(98)00439-6. PMID: 10074982 Wu ML, Whittemore AS, Paffenbarger RS Jr e t al. (1988). Personal and environmental characteristics related to epithelial ovarian cancer. I. Reproductive and menstrual events and oral contraceptive use. A m J E p id e m io l, 128:1216-1227. PMID:3195563 TALC 383 3. Studies of Cancer in Experimental Animals The Working Group identified an issue that relates to the interpretation of several of the inhalation and intratracheal instillation studies of talc. A lesion that is frequently seen in rats that have been exposed by inhalation to a range of poorly soluble particles such as talc has been described variously as `proliferating squamous cyst', `proliferative keratinizing cyst', `proliferating squamous epithelioma', `benign cystic keratinizing squamous-cell tumour' or `cystic keratinizing squamous-cell tumour'. Various authors have included this lesion in tumour counts, but the neoplastic nature of this lesion has been debated (Kittel et al., 1993; Carlton, 1994; Mauderly et al., 1994; Boorman & Seely, 1995; Rittinghausen et al., 1997; Rittinghausen & Kaspareit, 1998); its relationship to pulmonary neoplasia is uncertain. The Working Group noted that, in many of the studies of `talc' described below, no or limited characterization of the mineralogy o f the sample employed was given, and, in particular, that there was a lack of information on fibre content or particle size. 3.1 Oral administration Rat Groups of 25 male and 25 female Wistar rats, 10 weeks of age, received about 50 mg/kg body weight (bw) per day of commercial talc [characteristics unspecified] in the diet (average survival, 649 days) or standard diet alone for life (average survival, 702 days). No significant difference in tumour incidence was found in the treated animals compared with control animals (Gibel et al., 1976). Groups of 16 male and 16 female Wistar-derived rats, 21-26 weeks of age, were fed 100 mg Italian talc (grade 00000; ready milled; mean particle size, 25 pm; containing 92% talc, 3% chlorite, 1% carbonate minerals and 0.5-1% quartz) per day per rat in the diet for 5 months (talc-containing diet was actually given for 101 days) and were then maintained on basal diet for life (average survival, 614 days). No differences in tumour incidence were noted between treated animals and eight male and eight female control animals fed basal diet throughout (average survival, 641 days) (Wagner et al., 1977). [The Working Group noted the limited exposure period and the advanced age of the animals at the start o f the study.] 384 IARC MONOGRAPHS VOLUME 93 3.2 Inhalation exposure 3.2.1 Mouse Groups of 47-49 male and 48--50 female B6C3F, mice, 7 weeks of age, that were fed an NIH-07 diet, were exposed by inhalation to aerosols containing 0, 6 or 18 mg/m3 MP 10-52 grade talc for 6 hours per day on 5 days per week for up to 104 weeks (dose equivalent, 0, 2 or 6 mg/kg bw per day for male mice and 0, 1.3 or 3.9 mg/kg bw per day for female mice). MP 10-52 grade is a high-purity microtalc (from a strip mine located in Missouri State, USA) that has a maximal particle size of 10 pm and is reported to contain no tremolite or any asbestiform minerals. After analysis, the talc was found to be free of asbestos and almost free of silica. The average mass mean aerodynamic diameter (MMAD) and the geometric standard deviation (GSD) of the talc aerosols were calculated to be 3.3 1.9 pm and 3.6 2.0 pm for the 6- and 18-mg/m3 chambers, respectively. At approximately week 70, difficulties were experienced in generating the talc aerosol, and the chamber concentrations were substantially lower than the target concentrations over a period of 12 weeks. Survival and final mean body weights of male and female mice exposed to talc were similar to those of the controls, and no clinical findings were attributed to exposure to talc. No significant increases in the incidence of neoplasms were observed. The incidence of pulmonary neoplasms (males: 27%, 11% and 23%; females: 11%, 12% and 6%) was similar between exposed and control groups of mice. [The Working Group noted that the incidence of alveolar/bronchiolar adenoma or carcinoma combined in historical control B6C3Fi mice fed an NIH-07 diet in National Toxicology Program inhalation studies was 26.8% for males and 10.1% for females] (National Toxicology Program, 1993). 3.2.2 Rat Two groups of 12 male and 12 female Wistar-derived rats, 6-8 weeks of age, were exposed by inhalation to a mean respirable dust concentration of 10.8 mg/m3 Italian talc (grade 0000; ready milled; mean particle size, 25 pm in diameter; containing 92% talc, 3% chlorite, 1% carbonate minerals and 0.5-1% quartz) for 7.5 hours per day on 5 days a week for 6 or 12 months (cumulative exposures, 8200 and 16 400 mg/m3 x h, respectively). Ten days after the end of each exposure period, six rats per group were killed; 12 rats per group died and two rats per group were unaccounted for; the remaining four rats per group were killed 1 year after the end of the exposure period. No differences were noted in the incidence of lung tumours compared with 24 male and 24 female untreated controls (Wagner et al., 1977). [The Working Group noted the limited number of animals allowed to survive longer than 12 months after the end of each exposure period.] Groups of 49 or 50 male and 50 female Fischer 344/N rats, 6--7 weeks of age. were exposed by inhalation to aerosols of 0, 6 or 18 mg/m3 MP 10-52 grade talc (see Section 3.2.1) for 6 hours per day on 5 days per week until mortality in any exposure group TALC 385 8.4 mg/kg bw per f o /L T i G S D ^ o T ^ S ^ 9 ^ equivalent' > 2-8 or The average MMAD and the GSD of the nil ' ' i bW pe' day for females)- and 3.2 1.9 um for foe 6 and 18 3T " S^ to be 2 7 ] 9 ^ chamber concentration for the 18 mJ f o *;esPectlvely- At week 11, the 40 mg/m3 for a nerind n f 7 v u ^ 8r Up Vaned prom approximately 30 to s L r : : of 1* * * * * * * * * - d * yTM * ; experienced in generating ,he ,a . T TM "6* Week 70' dMc,,lties were 1Absoi T T St ` 18 mg/mi t a " l t ^ r o l S" fte" w eei lifetime J d y f e ^ n r e 1 c n 'm ' Val" atio" s a" d at the end of the significantly greater (P<() n o n ;n m u- L J (carcinoma, 5/50) and was P = 0 028'i ThP in a i i thC hlh_dose grouP than in controls (carcinoma s = i! S S S ; -ban dra, 1 ! ^ f T T * jf" * o M Z yT 1 in r - was: amn ' S )- ^ hUgh adrenal meduUa hyperplasia occurred with similar freciuencv he inc.dence o f these tumours was also observed in several other National Toxicology r e o f i^ that this increase may not be related to talc.] sy e , d hence telt 386 3.2.3 Hamster IARC MONOGRAPHS VOLUME 93 ^ , ' ^ z ^ : z z ,, z z u : ma!eand 50 fema,e syrian *>" that was prepared from Vemraut talc by flotmion" (95% w/w"olaiv'"! ' ! i T " 1" ' chlorite and m,iie) f* 3- 3 0 ^ Z "z measurable respirator fra,on of p T ^ X n 'd T M M A D 0" 4TM m a T ' Wi'h d group comprised 25 males and 25 females Two fi,rthp 47 A sham-exposed age, ,,ere exposed folc aerosol for 30 or Im T ^ f mg/m3 aerosol concentration was 274 me/m3 with m P . 7 * 3 d ys` The mean 8.1 and a MMAD nf fin a ', measurable respiratory fraction of and 25g f l a , e S A- oth-^ a m -e x p o s e d group comprised 25 males (Wehner et al 1977 | 0 7 0 3 n~u T o , 25/o compared Wlth 10% m the control group the 8 "P " ed "K Sh" daily eXp0SUre 3.3 Intratracheal administration Hamster J z szs m : d. S i S : 63/o; magnesium oxide, 32-34%; other dusts, 0.85-1.06%- 9 3 3 0 / < 7 t ^ de' 61~ un,rea,rdLThe7 " " l ' " "T 3 ^ or' 0.2 mL sa lin e'a lo " e'were untreated TJe an,reals were allowed to live out their lifespan (average 50% sureival T 3.4 Subcutaneous administration Mouse nf n 9 r r a,e R3 m`Ce' 3-6 m0nths of age' were iven single subcutaneous injections aaboonut 8g0t rmeg)t a^nd were fotbseerBve'Cd ^for li"fenS(Pa"veir,!a'gde] 5a"0%d 20su8rvPiveaalnuStQofdi hdQx,cOe Sa t di osei tumour was observed (Neukomm & de Trey, 1961). ' T ) No local TALC 387 Singie sUC L r " t r : ,, tsc u sp ;a,Mai mice' i mom,,s 0f *<* No tumour developed at the W e c t l , a"d 7 " ' bserVed for 18- 21 months, controls (Bischoff & Bryson, 1976). ' rCated animals or m 24 saline-injected 3.5 Intraperitoneal administration 3.5.1 Mouse mr ri,r * ^ " < intraperitoneally with a 0.5-inL suspension (50"/) oflaT^-age unsp" lfled] were injected ^ ^ In ^ s lL id ^ n m ^^^neo^lasm s were reporteif(Jagatic^^^ single i n t r p e r i . r a C X t T m g ^ s t Kt a n ,,dmiCe' ^ " 'd f ^ reCe'Ved " Intraperitoneal lymphoid t o p o ^ e^r Tu , fOT mon,bs treated controls (Blschoff& Bryson 1976) ^ an`ma S a"d 6/28 salme' = ^ ; sk " normal lifespan three peritonei I m " a -^ SUrV1VOrs all Wed t0 llve out their 3/46 saline-treated - K i" Z m? T * TM*b occurrence of mesotheliomas in sahd-treattl animais.] W rkmg " 3.5.2 Rat injections of 25 s injr ed wi 2 r - state. A ^ With none in 72 controls (Pott et al 1974 1976a b) r tT T rit0,,|!rI * "" 'b""d dayS comPared 388 IARC MONOGRAPHS VOLUME 93 3.6 Intrapleural and intrathoracic administration 3.6.1 Mouse In a study reported as an abstract, male Marsh mice, 3 months of age, received a single intrathoracic injection of 10 mg USP talc. After 18-21 months, 5/47 treated mice had tumours (two adenocarcinomas and three lymphoid tumours of the lung) compared with none o f 48 saline-injected controls (Bischoff & Bryson, 1976). 3.6.2 Rat In a study reported as an abstract, female Evans rats, 3 months of age, received single intrathoracic injections of 50 mg USP talc. After 18-21 months, intrathoracic reticulum cell sarcomas or lymphomas were observed in 7/30 talc-treated rats, 8/32 saline-treated rats and 7/28 untreated controls (Bischoff & Bryson, 1976). In a lifetime study, a group of 24 male and 24 female Wistar-derived rats, 8-14 weeks of age, received a single intrapleural injections o f 20 mg Italian talc (grade 00000; ready milled; mean particle size, 25 pm; containing 92% talc, 3% chlorite, 1% carbonate minerals and 0.5-1% quartz) in 0.4 mL saline. The mean survival time of the treated rats (655 days) was similar to that o f 24 male and 24 female controls (691 days) that were injected with saline. No mesothelioma was detected in either group; one small pulmonary adenoma was found in one treated rat that died 25 months after injection (Wagner et al., 1977). Following thoracotomy, groups of 30-50 female Osbome-Mendel rats, 12-20 weeks of age, received intrapleural implantations of 40 mg of one of seven grades of refined commercial talc from separate sources in hardened gelatin. The rats were followed for 2 years, at which time survivors were killed. The incidence of pleural sarcomas was: talc 1, 1/26; talc 2, 1/30; talc 3, 1/29; talc 4, 1/29; talc 5, 0/30; talc 6, 0/30; talc 7, 0/29; untreated controls, 3/488 (0.6%); and controls that received implants of `non-fibrous' materials described by the authors as `non-carcinogenic', 17/598 (3%) (Stanton et al., 1981). 3.7 Ovary implantation Rat In a study that investigated the effect of implanted talc on the rat ovary, a group of 10 female Sprague-Dawley rats, 10-15 weeks of age, received implants of 100 pL of a talc suspension in saline (100 mg/mL) onto the surface of the ovary by intrabursal injection. The talc was described as Italian 00000 (particle size, 0.3-14 pm) and contained no asbestos. Three sham-operated and three sham-treated control animals were included. Animals were killed after 12 months and histopathological examination of the ovaries was performed. Small focal areas of papillary change that were considered to be TALC 389 preneoplastic changes were seen in the surface epithelium of 4/10 treated animals (0/6 controls). No neoplasms were reported (Hamilton e t a l , 1984). [The Working Group noted that groups of animals implanted for 1,3, 6 or 18 months were also included, but no results were reported for any of these groups.]. 3.8 References Bischoff F, Bryson G (1976). Talc at the rodent intrathoracic, intraperitoneal, and subcutaneous sites (Abstract No. 1). P ro c A m A sso c C a n cer R es, 17:1. Boorman GA, Seely JC (1995). The lack of an ovarian effect of lifetime talc exposure in F344/N rats and B6C3F1 mice. R egu l T oxicol P h arm acol, 21:242-243. doi:10.1006/rtph. 1995.1035. PMID:7644712 Carlton WW (1994). "Proliferative keratin cyst," a lesion in the lungs of rats following chronic exposure to para-aramid fibrils. F undam A p p l Toxicol, 23:304-307. doi: 10.1006/faat. 1994.1108. PMID:7526997 Gibel W, Lohs K, Horn KH et al. (1976). [Experimental study on cancerogenic activity of asbestos filters. A rch G eschw ulstforsch, 46:437-442 (in German). PMID:999453 Hamilton TC, Fox H, Buckley CH e t al. (1984). Effects of talc on the rat ovary. B r J E xp P athol, 65:101-106. PM1D:6696826 Jagatic J, Rubnitz ME, Godwin MC, Weiskopf RW (1967). Tissue response to intraperitoneal asbestos with preliminary report of acute toxicity of heat-treated asbestos in mice. E nviron Res, 1:217-230. doi:10.1016/0013-9351 (67)90014-X. PM1D:4303313 Kittel B, Ernst H, Dungworth DL et al. (1993). Morphological comparison between benign keratinizing cystic squamous cell tumours of the lung and squamous lesions of the skin in rats. E xp T oxicol P ath ol, 45:257-267. PMID:7508775 Mauderly JL, Snipes MB, Barr EB et al. (1994). Pulmonary toxicity of inhaled diesel exhaust and carbon black in chronically exposed rats. Part I: Neoplastic and nonneoplastic lung lesions. R es R ep H ealth E ffln st, 68:1-75, discussion 77-97. PMID:7530965 Melnick RL, Bucher JR, Roycroft JH et a l (2003). Carcinogenic and toxic effects of inhaled, non fibrous, poorly soluble particulates in rats and mice contradict threshold lung cancer hypotheses that are dependent on chronic pulmonary inflammation.. Eur J O ncol., 8:177-186. National Toxicology Program (1993). T o x ic o lo g y a n d C a rc in o g e n e s is S tu d ie s o f T a lc (C A S N o. 1 4 8 0 7 --96--6) in F 3 4 4 /N R a ts a n d B 6 C 3 F i M ic e (In h a la tio n S tu d ie s). (Tech Rep Ser 421), Research Triangle Park, NC. Available at: http://ntp.niehs.nih.gov/ntp/htdocs/LT_rpts/tr421 .pdf Neukomm S, de Trey M (1961) [Study of possible carcinogenic and/or co-carcinogenic brightening agents.] M e d Exp, 4:298-306 (in French). Ozaki K, Haseman JK, Hailey JR et al. (2002). Association of adrenal pheochromocytoma and lung pathology in inhalation studies with particulate compounds in the male F344 ratthe National Toxicology Program experience. T o x ico l P a th o l, 30:263-270. doi: 10.1080/019262302753559605. PM1D: 11950170 Ozesmi M, Patiroglu TE, Hillerdal G, Ozesmi C (1985). Peritoneal mesothelioma and malignant lymphoma in mice caused by fibrous zeolite. B r J I n d M e d , 42:746-749. PMID:2998433 390 IARC MONOGRAPHS VOLUME 93 Pott F, Dolgner R, Friedrichs K-H, Huth F (1976b). [The oncogenic effect of fibrous dust Animal Pott F, Friedrichs K-H, Huth F (1976a). [Results of animal experiments concerning the carcinogenic effect of fibrous dusts and their interpretation with regard to the carcinogenesis in P tt FUH T p i T ? /Wu5flA/erto/0^ 5' 162:467- 505 (in German). PMID485852 Pott F, Huth F, Friedrichs KH (1974). Tumorigenic effect of fibrous dusts in experimental animals E n viron H ealth P ersp ect, 9:313-315. doi: 10.2307/3428305. PMID:4377876 ' lttmghausen S, Kaspareit J (1998). Spontaneous cystic keratinizing epithelioma in the lung of a S tctT r3t T0Xkl PathU 26:298' 300' doi: 10.1177/019262339802600218. Rittinghausen S Mohr UDungworth DL (1997). Pulmonar cystic keratinizing squamous cell lhalatl0n/inSti,,ati0n 0f different Particles. E xp T oxicol P athol, 49:433- Stanton MF Layard M, Tegeris A e, al. (1981). Relation of particle dimension to carcinogenicity PMIDP6h9460253aSbeStOSeS ^ ^ fibrUS minera!s' '' ^ C m Inst> 67:965-975 Stenbck F, Rowlands J (1978). Role of talc and benzo(a)pyrene in respirator tumor formation An experimental study. S ca n d J R esp ir D is, 59:130-140. PMID:684384 Waif f r,JC' Ber^ G' Coke TJ et al ( 1977). Animal experiments with talc. In: Walton WH McGovern B, eds, In h a le d P a rtic le s, Vol. IV, Part 2, Oxford, Pergamon Press, pp. 647- Wehner AP, Stuart BO, Sanders CL (1979). Inhalation studies with Syrian golden hamsters P ro ? Exp Tum or Res, 24:177-198. PMID:538242 S Wehner AP. Zwicker GM Cannon WC (1977). Inhalation oftalc baby powder by hamsters F ood C o sm et Toxicol, 15:121-129. doi: 10.1016/S0015-6264(77)80317-9. PMID:873404 TALC 391 4. Mechanistic and Other Relevant Data s o l u i t e p f ^ atnhiveS|ow toxfc'^Tr' in .his volume. * * * d'SC"SSSed ^ re,Mi" ( " * M"8">Pll on carbon black 4.1 Humans 4- U Deposition, retention and clearance pne,,mlP,, t s i\X e r / Z L : : ^ r s e e ,"6 P`T T TM ^ X T s tfS T r 'rs c r s . : " ; h7i ' x c o Z ;^ mineral fibres and non-fibrous narhVl twofold higher respectively in rtf The aveiage concentrations of y fourfold and approximately m . i af lb ^ r t ie a |d W'gf <l987) " amine<l diS,ribu,ran "f 392 IARC MONOGRAPHS VOLUME 93 exposed subjects six o f whom were talc millers. In the latter group two workers had 4 o T e h lXCrUSir ^ I3'0 `n the"' 'aVage 1Ud' Wh' le the 0thers had about 60% talc and 40/o chlonte In other workers, tale generally accounted for <3% o f the p a r t i e s in avage fluid. It was noted that, although the exposure o f one of the millers had ceased years before the examination, talc particles were still present in his lavage fluid ST7 (HenTM w T s T : , C- bee" " S,0m"d ' aTM TM JaPa ,, .Se men P SS y d"e ,nest*on of talc-treated rice (Merlisi su b se rta l o m T S T ' appare" tly " oth,ir insol"ble P ^ lc le s , were found in the ubserosal stroma of hernia sacs, possibly due to ingestion of medications in which c ':s present as a fller (Pratt et al., 1985). Anani et al. (1987) reported the presence of ale fibres in the intestinal wall o f a 46-year-old patient who hadSevere S " and was diagnosed with intestinal talcosis. A possible source o f exposure was the talc contained m oral medications against tuberculosis, which the patient had taken nearly 20 yea s earlier over a period o f 22 months (total intake o f tale, 183 g). Y a c is often present as a filler in some materials used by drug addicts which results m wide dissemination o f talc particles to the lungs (Groth et al 1972 Lamh ^ Roberts ,, 7 2 , Farber et al., 1981; Crouch & Churg, I983X Spleen Icidn'ey lfv'e ^ b r a i f 19724 f T 1 1 thyr ld gl3ndS (Gr th 6t a l> ]912) and even the retina (AtLee 972). In the lungs most o f the talc particles are found within the vessels o f the alveolar wa Is, and are almost invariably associated with marked foreign body granulomas (Crouch & Churg, 1983). The tale particles found in theTun ^ l a r g e r Section 4ai " a tha" after inhalation (Abraham & Brambilla, 1980) (see Section 4.1.2 for a discussion of the associated toxic effects) n e r i n J T .f 'p e n o l o g i c a l evidence o f a possible association between talc use for P , ygl^ ne and an increased risk for ovarian cancer (see Section 2) several studres have been conducted in women to determine potential re,4 de i v I Z o f particles through the reproductive trae, to the ovaries These StudiesTnvCved cTmmp h c a l0,,sWoThlb " 1d " nder8 8ynaeCOl 8 i<:al most|y for diseases or Therefore' bioad t w o ^ ' t ^ NeT n (I961) Und that inert carbn particles deposited in the vagina in nQTn aP3 ir trave ed t0 the fadopian tubes in about 30 minutes De Boer ( 72) concluded that Indian ink deposited below the level of the cervix is unlikely to rave Iquicklj/ through the reproductive tract. In contrast, the findings of Venter and hutralde ,979) and Mostafa a,. ,,9 8 5 ) suggested that retrograj^transport m opian tubes is possible. Henderson et al (1971) reported the actual presence o f talc Tnd f i v e T n 2 Snonmer ^ 10 f 13 Vanan tumours' 12 o f 21 cervical tumours ti 11 , nQ 113 ovanan tissues. Subsequently, Henderson et al (1979) and Her et al (1996) provided further evidence o f the presence o f talc in the ovaries of women who had purportedly had perineal exposure to talc. However in the latter u y, no ielation was found between talc-particle counts and reported perineal use of TALC 393 4.1.2 Toxic effects The toxic effects o f talc in humans are dependent on the route and dose of administration and the physicochemical properties of the talc. In addition talc products commonly contain other potentially toxic minerals (see Section 1) ' P Talc pneumoconiosis is somewhat more prevalent and severe among people who are exposed to talc that contains asbestiform minerals than among those wlfo^re exposed to vader'widdvUf lmp"m ies<Kleinfdd ' < 19). The form of this pneumoconiosis ,1, . ' ' , " STMP e a;'-VI,T'om!l"c W * (Vallyathan & Craighead 1981) to saWmg conglomerate pneumoconiosis (Hunt, 1956; Graham & Gaensler, 1965; Miller . 7 1971). Mixed-dust pneumoconiosis is frequently seen, including silicosis asbestosis and occasionally other forms (Kleinfeld et al.,1963; Mark et al. 19 7 9 ) ' Several early reports described `talcum powder granuloma' that arose from the use of on surgical gloves (reviewed m Eiseman et al., 1947). Subsequent reports of cases have documented a variety of surgical complications, L lu d in g adhesions pseudotumours and sinus tracts that were attributable to exposure to talc (Lichtman et al. ' 946, Eiseman etal, 1947; reviewed by Hollinger, 1990). Both skin granulomas and talc pneumoconiosis have been reported after liberal use o f talc on the body (Tye et al 1966- Nani & Gracey, 1972; Wells /al, 1979; Tukiainen etal, 1984; Wehner 1994 ' Respn-atoty distress syndrome, which can be fatal, has been described in children e al T9g6 r i7 n d a& FCMt Rnha'atl0n f '^ P Wder (CleSS & AnSer' 1954^Molnar n t \ * L, d. & Feldt-Rasmussen> '969; Gould & Bamardo, 1972) and in adult patients after talc pleurodesis (Rehse et al., 1999). A variety of pathological effects arise from'the intravenous use by drug addicts of pm ucis that contain tale. These include micronuclear pulmonaiy opadtles (Hopk ns & Taylot. 1970; Arnett e, a l, 1976; Waller e, a l , 1980), a n g ilr lb o t ic puLon,TM hypertension (Wendt el al., 1964; Pate e t a l , 1979; Waller et a l, 1980) and couglojrate pulmonary lesions (Siemewicz & Nidecker, 1980; Crouch & Churg, 1983). In addition etmopa hy, cerebral microembolization and granulomas of the loer, lymph nodes and kidneys have been reported (Min et a l, 1974; Pare et a l, 1979; Caiman, 1985), I S7 v f -i ^ oss' sf ctlonal studies sported from the New York State Department of m aT l ,955' ,% 3 ' ] 9 H j973) have documented talc pneumoconioses aassssoocci,aUtedd wwhi hhannplleural^ plaeqSuPeesC, lar"eystraimctiv"eg otrremobslittriCucttiavlCe WbrOeraktherisn'gThdeisoCradseerss waenrde ecreased vital capacity o f the lungs. The prevalence of disease was lower among those Am M u s t e k 3tlVe CXP0SUre t0 dUSt 3nd am ng th Se Wh Pr CeSSed granular rather A series of cross-sectional studies that described talc pneumoconiosis in workers in talc mining milling and manufacture in Italy (Rubino et a l, 1963; Tronzano et al 1965) found that the prevalence was related to extent and duration of exposure and that talcs pneuTdconiosir ^ Se'Pentine ^ qUa,1Z WCre aSSOdated with signiflcant 394 IARC MONOGRAPHS VOLUME 93 One representative, well-controlled study among 80 workers exposed in the rubber industry to Vermont talc, which is reported to have a low content of silica and fibres, showed significantly increased respiratory symptoms, impaired ventilatory function and increased respiratory morbidity, but no radiographic abnormality (Fine et al., 1976). There has been some concern that talc may cause adult respiratoiy distress syndrome when instilled into the pleural space for pleurodesis (Rinaldo et al., 1983; Bouchama et al., 1984; Kennedy et al., 1994; Rehse et al., 1999; Light, 2000). Relatively recent cases were observed when talc was both insufflated and used as a slurry (Brant & Eaton, 2001; Scalzetti, 2001). However, other case series did not report the development of this disease (Weissberg & Ben-Zeev, 1993; Rodriguez-Panadero & Antony, 1997; Sahn, 2000; Ferrer et al., 2001, 2002; Cardillo et al., 2006). Many o f the patients in the case reports had co morbid conditions. [The Working Group noted that the talc used in these reports was not always characterized mineralogically and may have contained contaminants.] The role of exposure to talc in the development of ovarian cancer has raised concerns (see Section 2). The normal ovarian epithelium is known to express several mucins that are protective against epithelial inflammation and injuty (Lalani et al., 1991; Gipson etal., 1997; Ness & Cottreau, 1999; Taylor-Papadimitriou et al., 1999; Ness et al., 2000; La Vecchia, 2001). Several epithelial cancers, such as breast and ovarian cancer, express mucin (MUC-1) which is upregulated and aberrantly glycosylated in many carcinomas (Taylor-Papadimitriou etal., 1999). Cramer et al. (2005) examined the association between the characteristics of women with no previous diagnosis o f ovarian cancer and levels o f antibodies to MUC-1, a protein that is expressed by normal epithelial cells and overexpressed by ovarian cancer cells. The study participants were 705 controls from a case-control study of ovarian cancer conducted in Massachusetts and New Hampshire (USA) between 1998 and 2003. Plasma specimens collected from participants at enrolment into the study were analysed for antiMUC-1 antibody levels using an enzyme-linked immunosorbent assay. Forty-eight cases of ovarian cancer with pre-operative blood specimens were also included in additional analyses; further 668 cases of ovarian cancer were included in the analyses to evaluate risk factors for ovarian cancer. Multivariable logistic regression, Spearman rank correlations and generalized linear models were used in the statistical analyses to determine which characteristics were associated with anti-MUC-1 antibody production and which were associated with the risk for ovarian cancer. Women who reported no previous genital use of talc were more likely to have antibodies to MUC-1 than women who had a histoiy of regular genital exposure to talc (38.1% versus 28.6%; P = 0.04). In addition, there was a borderline significant trend between frequency of talc use and lower anti-MUC-1 antibody levels (P = 0.11), after adjustment for other characteristics that affect antibody levels. Several conditions associated with increased antibody production were associated with a decreased risk for ovarian cancer. The authors concluded that these findings suggest that the presence of anti-MUCl antibodies is inversely correlated with risk for ovarian cancer. [Limitations of this study included the potential for bias in the participants' recollection of their genital use of talc, due to the case-control study design. TALC 395 In addition, antibody levels in the cases and controls may not be comparable, since the presence of a cancer may affect anti-MUC-1 antibody levels.] ' 4.2 Experimental systems 4.2.1 Deposition, retention and clearance The deposition, translocation and clearance of talc was investigated in 44 female golden Syrian hamsters (10 weeks o f age) that were exposed by nose-only inhalation for 2 hours to 40-75 mg/m3 neutron-activated talc (Johnsons's Baby Powder, lot 228p; median aerodynamic diameter, 6.4-6.9 pm). The powder was high-grade cosmetic talc and consisted o f 95% (w/w) platy talc mineral (Wehner et al., 1977a). Alveolar deposition was approximately 20-80 pg, which represented 6-8% o f the inhaled amount. The retention half-time of the talc deposited in the alveoli was 7-10 days, and alveolar clearance was reported to be essentially complete 4 months after exposure. No translocation of talc to liver, kidneys, ovaries or other parts of the body was found (Wehner et al, 1977b). [The Working Group noted that the unusually short clearance time may be related to limitations in the sensitivity o f the detection methods and the large size o f the particles used.] In rats exposed for 7.5 h per day on 5 days a week to aerosols of Italian talc (mean concentration of respirable dust [not further defined], 10.8 mg/m3), the mean amounts of talc retained in the lung were 2.5, 4.7 and 12.2 mg per animal following exposures for 3, 6 and 12 months, respectively. These levels were approximately proportional to the cumulative exposures (Wagner et al., 1977). In rats exposed for 6 hours per day on 5 days a week for 4 weeks to 2.3, 4.3 and 17 mg/m3 respirable talc, the amounts retained in the lung at the end of exposure were 77, 187 and 806 pg talc/g lung, respectively (Hanson et al, 1985). Lung burdens of talc were determined in groups of 10 male and 10 female Fischer 344 rats and B6C3F, mice following exposure to asbestos-free talc for 6 hours per day on 5 days a week for 4 weeks. In rats exposed to 0, 2.3, 4.3 and 17 mg/m3, average lung burdens were 0, 0.07, 0.17 and 0.72 mg talc/g lung, respectively. In mice exposed to 0, 2.2, 5.7 and 20.4 mg/m3, average lung burdens of 0, 0.10, 0.29 and 1.0 mg talc/g lung' respectively, were observed. When normalized to the exposure concentration, the lung burden in mice was greater than that in rats and the normalized burden in rats increased with increasing exposure concentration (Pickrell et al, 1989). Conflicting data exist on systemic distribution following intrapleural instillation of talc (i.e. talc pleurodesis) in rats. Following administration of 10 or 20 mg talc [particle size unspecified] to rats (20 per group), talc was identified in the chest wall, lungs, heart, brain, spleen and kidneys. The authors concluded that talc is rapidly absorbed through the pleura and reaches the systemic circulation and organs 24 hours after administration (Werebe et a l, 1999). However, following instillation of 40 mg talc (median particle size, 31 pm) into 33 rats randomly assigned to autopsy 24 or 72 hours later, talc particles were 396 IARC MONOGRAPHS VOLUME 93 observed in only a few extrapulmonary organs, i.e. the brain, spleen and liver, but not the kidneys (Fraticelli etal., 2002). The systemic distribution of talc was investigated in rabbits following talc pieurodesis in two studies. In one study (Ferrer et al., 2002), 10 rabbits received 200 mg/kg bw 8.4-pm asbestos-free talc particles and 10 received 200 mg/kg bw 12-pm talc particles. Five animals from each group were killed after 24 hours and five at 7 days after instillation. A tendency was seen for increased extrapulmonaiy distribution of the smaller particles, which were identified in the pericardium o f 0/5 and 3/5 rabbits at 24 hours and 7 days, respectively. For the larger particles, one of five animals had talc in the pericardium at each time-point. Particles were identified in the liver of three of five animals exposed to the smaller particles 7 days after instillation; other groups had no particles in the liver. Small particles were found in the kidney of only 1/5 animals 24 hours after instillation. Both particle types were found in the spleen of 1/5 animals 24 hours after instillation. The results indicate that talc reached the lung parenchyma by breaking the mesothelial and elastic layer and that mobility was greater for the smaller particles. In the other study, Montes et al. (2003) performed talc pieurodesis in rabbits (20 per group) at doses o f 50 and 200 mg/kg bw of the small-particle talc used in the study by Ferrer et al. (2002). Doses were chosen to simulate treatment o f a 60-kg patient with amounts of 3 and 12 g talc. The lung parenchyma o f two and 14 rabbits o f the low-dose and high-dose groups, respectively, contained talc. In the high-dose group, six of the animals had talc in the pericardium and five had talc in the liver; talc was not detected in these organs in the low-dose group. The results show that the systemic distribution of talc was dose-dependent. In studies in rats, mice, guinea-pigs and hamsters that used radioactive tracer techniques, no intestinal absorption or translocation of ingested talc to the liver or kidneys was detected (Wehner et a l, 1977b; Phillips et a l, 1978). No translocation of talc into the ovaries was detected after single or multiple intravaginal applications of talc to rabbits (Phillips etal, 1978) or monkeys (Wehner etal., 1985, 1986). 4.2.2 Toxic effects Reviews of the literature on the biological effects of talc in experimental animals are available (Lord, 1978; Wehner, 1994). [The Working Group noted that in most o f the studies o f `talc' described below, no or limited characterization of the mineralogy of the sample employed was given, and, in particular, information on fibre content or particle size was lacking.] ' (a) Chronic toxicity Mild to marked arterial endothelial cell proliferation with cellular encroachment into the lumen, the occurrence of occasional foreign-body giant cells within the endothelial masses and model ate thickening of the intra-alveolar septa of the lungs were observed TALC 397 after intravenous injections of talc in rabbits and guinea-pigs (Puro et al., 1966; Dogra et al., 1977). No effect on the rat lung was observed after intravenous injection of talc (Schepers & Durkan, 1955b) but talc granulomas were seen in rats following intrasplenic injection o f talc (Eger & Canaliss, 1964). No chronic pathological effect was associated with oral administration of Italian talc (92% pure; 100 mg per day on 101 days over 5 months) to rats (Wagner et al., 1977). Intratracheal injections of talc (total dose, 150 mg) into guinea-pigs induced perivascular and peribronchiolar focal accumulations of histiocytes, fibrocytes, plasma cells and eosinophils within 1 month. After 2 years, the dominant effects were bronchiolectasia, bronchiolitis and marked fibrosis (Schepers & Durkan 1955b). Rats exposed to dust clouds o f 30-383 mg/m3 `industrial'- or `pharmaceutical'-grade talc for 9 months developed chronic inflammatoiy changes including thickening of the walls of the pulmonary arteries and, eventually, emphysema (Bethge-Iwanska, 1971). In rats exposed by inhalation to 10.8 mg/m3 Italian talc (grade 00000; ready milled; mean particle size, 25 pm) for 3 months, minimal fibrosis was obseived, the degree of which did not change during the observation period after exposure. Animals that were exposed for 1 year had minimal to slight fibrosis, the degree of which had increased to moderate within 1 year after cessation of exposure (Wagner et al., 1977). In contrast, Syrian golden hamsters exposed to 8-mg/m3 aerosols of cosmetic-grade talc for up to 150 minutes per day on 5 days a week for 30 days showed no histopathological change in the lungs, heart, liver, renal tissues, stomach or uterus (Wehner et al., 1977c). Two years after injection of 20 mg Italian talc (see above) into the right pleural cavity of rats, granulomas at the injection site were common, and one small pulmonaiy adenoma was observed, but no other relevant pathology was seen in the lungs (Wagner et al 1977). ' Groups of male and female rats, 6-7 weeks old, were exposed to aerosols of 0, 6 or 18 mg/m3 talc until mortality in any exposure group reached 80% (113 weeks for males and 122 weeks for females). These exposure concentrations provided a dose equivalent of 0, 2.8 or 8.4 mg/kg bw per day for male rats and 0, 3.2 or 9.6 mg/kg bw per day for female rats. The talc used for this study was MP 10-52 Grade (see Section 3.2.1) and was found to be free from asbestos by polarized light microscopy and transmission electron microscopy. Survival of male and female rats was similar to that of the controls. Mean body weights of rats exposed to 18 mg/m3were slightly lower than those of controls after week 65. No clinical findings were attributed to exposure to talc. Absolute and relative lung weights of male rats exposed to 18 mg/m3 were significantly greater than those of controls at the 6-, 11- and 18-month interim evaluations and at the end of the lifetime study, while those o f female rats exposed to 18 mg/m3 were significantly greater at the 11-, 18- and 24-month interim evaluations and at the end of the study. Talc produced a spectrum o f inflammatory, reparative and proliferative processes in the lungs. The principal toxic lesions observed included chronic granulomatous inflammation, alveolar epithelial hyperplasia, squamous metaplasia, squamous cysts and interstitial fibrosis o f the lung. These lesions were accompanied by impaired pulmonary function characterized 398 IARC MONOGRAPHS VOLUME 93 primarily by reduced lung volumes, reduced dynamic and/or quasistatic lung compliance, reduced gas-exchange efficiency and non-uniform intrapulmonary gas distribution (National Toxicology Program, 1993). Groups of male and female B6C3Fi mice, 7 weeks of age, were exposed by inhalation to aerosols that contained 0, 6 or 18 mg/m3 MP 10-52 grade talc (see Section 3.2.1) for up to 104 weeks (dose equivalents, 0, 2 or 6 mg/kg bw per day for male mice and 0, 1.3 or 3.9 mg/kg bw per day for female mice). Survival and final mean body weights of male and female mice exposed to talc were similar to those of the controls. No clinical findings were attributed to exposure to talc. Inhalation exposure to talc was associated with chronic inflammation and accumulation of macrophages in the lung. Accumulations of macrophages (histiocytes) containing talc particles were also observed in the bronchial lymph nodes (National Toxicology Program, 1993). (b) In-vitro toxicity A concentration >50 pg/mL Italian talc caused a 50% reduction in the colony-forming efficiency of cultured Chinese hamster V79-4 lung cells (Chamberlain & Brown, 1978). The concentration of talc (99% pure) required to cause 50% haemolysis of red blood cells was 6.5 mg/mL, which is more than 50-fold that of chrysotile. A concentration of 0.1 mg/mL talc caused 35% release of MCr from Syrian hamster tracheal epithelial cells labelled with radioactive sodium chromate; the concentration was twofold that required for chrysotile (Woodworth et al., 1982). Davies et al. (1983) examined the effect of different types of talc on mouse peritoneal macrophages in vitro. Macrophages were exposed to seven specimens of high-purity talcs and the release of lactate dehydrogenase and P-glucuronidase was measured. These enzymes are produced by macrophages after they digest materials that can induce fibrosis and chronic inflammation. Enzyme release after exposure of macrophages to quartz, a known fibrogenic dust, and magnetite, a non-fibrogenic dust, was also measured. Quartz caused the greatest cytotoxic reaction in vitro: the amount of enzyme released increased with the dose. Magnetite had no effect. All seven talc specimens were cytotoxic to the macrophages: the levels of enzymes released were dose-related but were lower than those observed after exposure to quartz. The results show that talc is cytotoxic to macrophages and may be able to induce fibrosis and chronic inflammation in animals. However, the macrophage response to talc appears to be weaker than that for other fibrogenic dusts such as quartz, and the response of macrophages to talc may be different in vivo. Talc caused the release of several cytokines including C-X-C and C-C chemokines from normal human pleural mesothelial cells (Nasreen et al., 1998). Pleural mesothelial cells exposed to talc did not undergo apoptosis, whereas malignant mesothelioma cell lines (ATTC CRL-2081, CRL-5820, CRL-5915) exposed to the same dose did (Nasreen et al., 2000). Talc also caused the release of basic fibroblast growth factor in pleural mesothelial cells (Antony et al, 2004). In bone marrow-derived macrophages from mice, talc was found to stimulate DNA synthesis ([3H]thymidine incorporation) (Hamilton et al, 2001). TALC 399 4.2.3 Genetic and related ejfects Three samples of respirable talc failed to elicit significant unscheduled DNA synthesis (10, 20 and 50 pg/cm2, 24 hours), sister chromatid exchange or aneuploidy (2, 5, 10 and 15 pg/cm , 48 hours) in rat pleural mesothelial cells, in contrast to various positive controls. The three samples, i.e Spanish talc (No. 5725), Italian talc (No. 5726) and French talc (No. 7841), contained 90-95% talc; the remaining contents were chlorite and dolomite. Electron microscopy analysis revealed that talc particles were taken up by the rat pleural mesothelial cells, but no aneuploidy was observed in metaphases (EndoCapron et al., 1993). 4.3 References Abraham JL, Brambilla C (1980). Particle size for differentiation between inhalation and injection pulmonary talcosis. Environ Res, 21:94-96. doi: 10.1016/0013-9351 (80)90011-0 PM1D:7389708 Anani PA, Ribaux C, Gardiol D (1987). Unusual intestinal talcosis. Am J Surg Pathol, 11:890 894. doi:10.1097/00000478-198711000-00007. PMID:3674285 Antony VB, Nasreen N, Mohammed KA et al. (2004). Talc pleurodesis: basic fibroblast growth factor mediates pleural fibrosis. Chest, 126:1522-1528. doi:10.1378/chest. 126.5.1522 PMID: 15539722 Arnett EN, Battle WE, Russo JV, Roberts WC (1976). Intravenous injection of talc-containing drugs intended for oral use. A cause of pulmonary granulomatosis and pulmonary hypertension. Am JMed, 60:711-718. doi: 10.1016/0002-9343(76)90508-8. PMID: 1020758 AtLee WE Jr (1972). Talc and cornstarch emboli in eyes of drug abusers. J Am Med Assoc. 219:49-51. doi: 10.100l/jama.219.1.49. PM1D:5066587 ' Berner A, Gylseth B, Levy F (1981). Talc dust pneumoconiosis. Acta Pathol Microbiol Scand A, 89:17-21. PMID:7223423 Bethge-Iwanska J (1971). [Pathomorphological changes of respiratory system in experimental talcosis.] Med Pr, 22:45-57 (in Czech). Bouchama A, Chastre J, Gaudichet A et al. (1984). Acute pneumonitis with bilateral pleural effusion after talc pleurodesis. Chest, 86:795-797. doi: 10.1378/chest.86.5.795. PMID:6488927 Brant A, Eaton T (2001). Serious complications with talc slurry pleurodesis. Respirology, 6:181 185. doi: 10.1046/j. 1440-1843.2001,00327.x. PMID: 11555375 ' Cardillo G, Carleo F, Giunti R et al. (2006). Videothoracoscopic talc poudrage in primary spontaneous pneumothorax: a single-institution experience in 861 cases. J Thorac Cardiovasc Surg, 131:322--328. doi: 10.1016/j.jtcvs.2005.10.025. PMID: 16434260 Carman CR(1985). Talc retinopathy. J Am Optom Assoc, 56:129-130. PMID:3980906 Chamberlain M, Brown RC (1978). The cytotoxic effects of asbestos and other mineral dust in tissue culture cell lines. BrJExp Pathol, 59:183-189. PMID:656318 Churg A, Wiggs B (1985). Mineral particles, mineral fibers, and lung cancer. Environ Res, 37:364-372. doi:10.1016/0013-9351 (85)90117-3. PMID:4017991 400 IARC MONOGRAPHS VOLUME 93 Churg A, Wiggs B (1987). Types, numbers, sizes, and distribution of mineral particles in the lungs of urban male cigarette smokers. Environ Res, 42:121-129. doi:10.1016/S00139351(87)80013-0. PM1D:3803330 Cless D, Anger R (1954). [Fatal asphyxia caused by aspiration of baby powder.]. Kinderarztl Prax, 22:506-508 (in German). PMID: 14354807 Cramer DW, Titus-Emstoff L, McKolanis JR el al. (2005). Conditions associated with antibodies against the tumor-associated antigen MUC1 and their relationship to risk for ovarian cancer. Cancer Epidemiol Biomarkers Prev, 14:1125-1131. doi: 10.1158/1055-9965.EPI-05-0035. PMID: 15894662 Crouch E, Churg A (1983). Progressive massive fibrosis of the lung secondary to intravenous injection of talc. A pathologic and mineralogic analysis. Am J Clin Pathol, 80:520-526. PMID:6624719 Davies R, Skidmore JW, Griffiths DM, Moncrieff CB (1983). Cytotoxicity of talc for macrophages in vitro. Food Chem Toxicol, 21:201-207. doi: 10.1016/0278-6915(83)90237-5. PMID:6682083 De Boer CH (1972). Transport of particulate matter through the human female genital tract. J ReprodFertil, 28:295-297. doi:10.1530/jrf.0.0280295. PMID:5061985 Dogra RKS, Iyer PKR, Shanker R, Zaidi SH (1977). Effect of talc injected intravenously in guinea pigs. Toxicology, 7:197-206. doi:10.1016/0300-483X(77)90065-8. PMID:857344 Dumortier P, De Vuyst P, Yemault JC (1989). Non-fibrous inorganic particles in human bronchoalveolar lavage fluids. Scanning Microsc, 3:1207-1216, discussion 1217-1218. PMID:2561220 Eger W, Canaliss DA (1964). [On organ-, especially liver changes after a single quartz-, asbestosor talc injection into the portal circulation of rats.]. Beitr Silikoseforsch Pneumokoniose, 81:11M2 (in German). PMID: 14233950 Egli GE, Newton M (1961). The transport of carbon particles in the human female reproductive tract. FertilSteril, 12:151-155. PMID: 13725928 Eiseman B, Seelig MG, Womack NA (1947). Talcum powder granuloma: a frequent and serious postoperative complication. Ann Surg, 126:820-832. doi:10.1097/00000658-19471100000015. PMID47859035 Endo-Capron S, Renier A, Janson X et al. (1993). In vitro response of rat pleural mesothelial cells to talc samples in genotoxicity assays (sister chromatid exchanges and DNA repair). Toxicol In Vitro, 7:7-14. doi: 10.1016/0887-2333(93)90107-G. PMID:20732166 Farber HW, Fairman RP, Glauser FL (1981). Bronchoalveolar lavage: a new technique for the diagnosis oftalc granulomatosis [Abstract]. Chest, 80:342. Ferrer J, Villarino MA, Tura JM et al. (2001). Talc preparations used for pleurodesis vary markedly from one preparation to another. Chest, 119:1901-1905. doi: 10.1378/chest. 119.6.1901. PMID: 11399721 Ferrer J, Montes JF, Villarino MA et al. (2002). Influence of particle size on extrapleural talc dissemination after talc slurry pleurodesis. Chest, 122:1018-1027. doi: 10.1378/chest. 122.3.1018. PMID: 12226049 Fine LJ, Peters JM, Burgess WA, Di Berardinis LJ (1976). Studies of respiratory morbidity in rubber workers. Part IV. Respiratory morbidity in talc workers. Arch Environ Health, 31:195 200. PMID:942261 TALC 401 Fraticelli A, Robaglia-Schlupp A, Riera H el al. (2002). Distribution of calibrated talc after intrapleural administration: an experimental study in rats. Chest, 122:1737-1741. doi: 10.1378/chest. 122.5.1737. PMTD: 12426279 Gipson IK, Ho SB, Spuir-Michaud SJ et al. (1997). Mucin genes expressed by human female reproductive tract epithelia. Biol Reprod, 56:999-1011. doi:10.1095/biolreprod56.4.999. PMID:9096884 Gould SR, Bamardo DE (1972). Respiratory distress after talc inhalation. BrJ Dis Chest, 66:230 233. doi: 10.1016/0007-0971 (72)90034-4. PMID:5044100 Graham WGB, Gaensler EA (1965). Talco-silicosis in a rubber worker. Med Thorac, 22:590-604. PMID:4955721 Groth DH, Mackay GR, Crable JV, Cochran TH (1972). Intravenous injection of talc in a narcotics addict. Arch Pathol, 94:171-178. PMID:5046803 Gylseth B, Stettier L, Mow G et al. ( 1984). A striking deposition of mineral particles in the lungs of a farmer: a case report. Am J Ind Med, 6:231-240. doi:10.1002/ajim.4700060306. PM1D:6475967 Hamilton JA, McCarthy G, Whitty G (2001). Inflammatory microcrystals induce murine macrophage survival and DNA synthesis. Arthritis Res, 3:242-246. doi: 10.1186/ar308. PMID: 11438042 Hanson RL, Benson JM, Henderson TR et al. (1985). Method for determining the lung burden of talc in rats and mice after inhalation exposure to talc aerosols. J Appl Toxicol, 5:283-287. doi: 10.1002/jat.2550050504. PMID:4056305 Heller DS, Westhoff C, Gordon RE, Katz N ( 1996). The relationship between perineal cosmetic talc usage and ovarian talc particle burden. Am J Obstet Gynecol, 174:1507-1510. doi:10.1016/S0002-9378(96)70597-5. PMID:9065120 Henderson WJ, Joslin CA, Turnbull AC, Griffiths K (1971). Talc and carcinoma of the ovary and cervix. J Obstet Gynaecol Br Commonw, 78:266-272. PMID:5558843 Henderson WJ, Evans DMD, Davies JD, Griffiths K (1975). Analysis of particles in stomach tumours from Japanese males. Environ Res, 9:240-249. doi: 10.1016/0013-9351 (75)90004-3. PMID: 1157802 Henderson WJ, Hamilton TC, Griffiths K (1979). Talc in normal and malignant ovarian tissue. Lancet, 1:499. doi: 10.1016/SO140-6736(79)90860-2. PMID:85089 Hollinger MA (1990). Pulmonary toxicity of inhaled and intravenous talc. Toxicol Lett, 52:121 127, discussion 117-119. doi: 10.1016/0378-4274(90)90145-C. PMID:2198684 Hopkins GB, Taylor DG (1970). Pulmonary talc granulomatosis. A complication of drug abuse Am Rev Respir Dis, 101:101-104. PMID:5410600 Hunt AC (1956). Massive pulmonary fibrosis from the inhalation of talc. Thorax, 11:287-294. doi: 10.1136/thx.l 1.4.287. PMID: 13391835 Kennedy L, Rusch VW, Strange C et al. (1994). Pleurodesis using talc slurry. Chest, 106:342-- 346. doi: 10.1378/chest. 106.2.342. PMID:7774299 Kleinfeld M, Messite J, Tabershaw IR (1955). Talc pneumoconiosis. Arch Ind Health, 12:66-72. Kleinfeld M, Giel CP, Majeranowski JF, Messite J (1963). Talc pneumoconiosis. Arch Environ Health, 7:101-115. PMID:14047558 Kleinfeld M, Messite J, Shapiro J et al. (1964). Lung function in talc workers. A comparative physiologic study of workers exposed to fibrous and granular talc dusts. Arch Environ Health, 9:559-566. PMID:14195257 402 IARC MONOGRAPHS VOLUME 93 Kleinfeld M, Messite J, Langer AM (1973). A study of workers exposed to asbestifonn minerals in commercial talc manufacture. Environ Res, 6:132-143. doi: 10.1016/0013-9351(73)90026l.PMID:4713673 La Vecchia C (2001). Epidemiology of ovarian cancer: a summary review. Eur J Cancer Prev, 10:125-129. doi: 10.1097/00008469-200104000-00002. PMID:11330452 Lalani EN, Berdichevsky F, Boshell M et al. (1991). Expression of the gene coding for a human mucin in mouse mammary tumor cells can affect their tumorigenicity. J Biol Chem 266:15420-15426. PM1D:1714457 ' Lamb D, Roberts G (1972). Starch and talc emboli in drug addicts' lungs. J Clin Pathol, 25:876 881. doi: 10.1136/jcp.25.10.876. PMID:4566961 ' Langer AM, SelikofFlJ, Sastre A (1971). Chrysotile asbestos in the lungs of persons in New York City. Arch Environ Health, 22:348-361. PMID:5100107 Lichtman AL, McDonald JR, Dixon CF, Mann FC (1946). Talc granuloma. Surg Gynecol Obstet 83:531-546. ' Light RW (2000). Talc should not be used for pleurodesis. AmJRespir Crit Care Med 162 2024 2026. PIVQDT1112104 Lord GH (1978). The biological effects of talc in the experimental animal: a literature review. Food Cosmet Toxicol, 16:51-57. doi: 10.1016/S0015-6264(78)80328-9. PMID:631664 Lund JS, Feldt-Rasmussen M (1969). Accidental aspiration of talc. Report of a case in a two-year- old child. Acta Paediatr Scand, 58:295-296. doklO.l 111/j. 1651-2227 1969*04721 x PMID:5783418 Mark GJ, Monroe CB, Kazemi H (1979). Mixed pneumoconiosis: silicosis, asbestosis, talcosis, and berylliosis. Chest, 75:726-728. doi:10.1378/chest.75.6.726. PMID:436529 ' Merliss RR (1971a). Talc-treated rice and Japanese stomach cancer. Science, 173:1141-1142. doi: 10.1126/science. 173.4002.1141. PMID:5098957 Merliss RR (1971b). Talc and asbestos contaminant of rice. J Am Med Assoc, 216:2144. doi: 10.1001/jama.216.13.2144d. PMTD:5108683 ' Miller A, Teirstein AS, Bader ME et al. (1971). Talc pneumoconiosis. Significance of sublight microscopic mineral particles. Am J Med, 50:395-402. doi:10.1016/0002-9343(71)90229-4 PMID:5553956 Min K-W, Gyorkey F, Cain GD (1974). Talc granulomata in liver disease in narcotic addicts. Arch Pathol, 98:331-335. PMID:4416043 Molnar JJ, Nathenson G, Edberg S (1962). Fatal aspiration of talcum powder by a child. Report of a case. N Engl J Med, 266:36-37. doi: 10.1056/NEJM196201042660110. PMID44475255 Montes JF, Ferrer J, Villarino MA et al. (2003). Influence of talc dose on extrapleural talc dissemination after talc pleurodesis. Am J Respir Crit Care Med, 168:348-355. doi: 10.1164/rccm.200207-7670C. PM1D: 12773332 Mostafa SA, Bargeron CB, Flower RW et al. (1985). Foreign body granulomas in normal ovaries. Obstet Gynecol, 66:701-702. PM1D:3903583 Nam K, Gracey DR (1972). Pulmonary talcosis from cosmetic talcum powder. J Am Med Assoc, 221:492-493. doi:10.1001/jama.221.5.492. PMID:5067955 ' Nasreen N, Hartman DL, Mohammed KA, Antony VB (1998). Talc-induced expression of C-C and C-X-C chemokines and intercellular adhesion molecule-1 in msothlial cells. Am J Respir Crit Care Med, 158:971-978. PMID:9731033 TALC 403 Nasreen N, Mohammed KA, Dowling PA et al. (2000). Talc induces apoptosis in human malignant mesothelioma cells in vitro. Am J Respir Crit Care Med, 161:595-600. PMID: 10673205 National Toxicology Program (1993). Toxicology'and Carcinogenesis Studies of Talc (CAS No. 14807--96--6) in F344/N Rats and B6C3F/ Mice (Inhalation Studies). (Tech Rep Ser 421), Research Triangle Park, NC. Available at: http://ntp.niehs.nih.gov/ntp/htdocs/LT_ipts/tr421 .pdf Ness RB, Cottreau C (1999). Possible role of ovarian epithelial inflammation in ovarian cancer. J Natl Cancer Inst, 91:1459-1467. doi: 10.1093/jnci/91.17.1459. PMID: 10469746 Ness RB, Grisso JA, Cottreau C et al. (2000). Factors related to inflammation of the ovarian epithelium and risk of ovarian cancer. Epidemiology, 11:111-117. doi: 10.1097/00001648200003000-00006. PMID: 11021606 Par JAP, Fraser RG, Flogg JC et al. (1979). Pulmonary `mainline' granulomatosis: talcosis of intravenous methadone abuse. Medicine (Baltimore), 58:229-239. PMID:449659 Phillips JC, Young PJ, Flardy K, Gangolli SD (1978). Studies on the absorption and disposition of 3H-labelled talc in the rat, mouse, guinea-pig and rabbit. Food Cosmet Toxicol, 16:161-163. doi:10.1016/S0015-6264(78)80197-7. PMID:669513 Pickrell JA, Snipes MB, Benson JM et al. (1989). Talc deposition and effects after 20 days of repeated inhalation exposure of rats and mice to talc. Environ Res, 49:233-245. doi: 10.1016/S0013-9351 (89)80069-6. PM1D:2753008 Pooley FD (1976). An examination of the fibrous mineral content of asbestos lung tissue from the Canadian chrysotile mining industry. Environ Res, 12:281-298. doi: 10.1016/00139351(76)90038-4. PMID: 1001300 Pratt PC, George MH, Mastin JP, Roggli VL (1985). Crystalline foreign particulate material in hernia sacs. Hum Pathol, 16:1141-1146. doi:10.1016/S0046-8177(85)80183-0. PMID:4054893 Puro HE, Wolf PL, Skirgaudas J, Vazquez J (1966). Experimental production of human `blue velvet' and `red devil' lesions. J Am MedAssoc, 197:1100-1102. doi: 10.1001/jama. 197.13.1100. PM1D:5953153 Rehse DH, Aye RW, Florence MG (1999). Respiratory failure following talc pleurodesis. Am J Surg, 177:437-440. doi: 10.1016/S0002-9610(99)00075-6. PMID:10365887 Rinaldo JE, Owens GR, Rogers RM (1983). Adult respiratory distress syndrome following intrapleural instillation of talc. J Thorac Cardiovasc Surg, 85:523-526. PM1D:6834872 Rodriguez-Panadero F, Antony VB (1997). Pleurodesis: state of the ait. Eur Respir J, 10:1648 1654. doi: 10.1183/09031936.97.10071648. PMID:9230261 Rubino GF, Maranzana P, Pettinati L, Scansetti G (1963). [Aetio-pathological and clinical aspects of talc pneumoconiosis.] Med Lav, 54:496-506 (in Italian). Sahn SA (2000). Talc should be used for pleurodesis. Am J Respir Crit Care Med, 162:2023-- 2024, discussion 2026. PMID:11112103 Scalzetti EM (2001 ). Unilateral pulmonary edema after talc pleurodesis. J Thorac Imaging, 16:99 102. doi: 10.1097/00005382-200104000-00006. PMID: 11292212 Schepers GWH, Durkan TM (1955a). The effects of inhaled talc-mining dust on the human lung. Arch Ind Health, 12:182-197. Schepers GWH, Durkan TM (1955b). An experimental study of the effects of talc dust on animal tissue. Arch Ind Health, 12:317-328. 404 IARC MONOGRAPHS VOLUME 93 Seeler AO (1959). Talc pneumoconiosis. N Engl J Med, 261:1084-1085. doi: 10.1056/NEJM195911192612115. PMID:14444496 Sieniewicz DJ, Nidecker AC (1980). Conglomerate pulmonary disease: a form of talcosis in intravenous methadone abusers. Am J Roentgenol, 135:697-702. PMID:6778101 Taylor-Papadimitriou J, Burchell J, Miles DW, Dalziel M (1999). MUC1 and cancer. Biochim BiophysActa, 1455:301-313. PMID:10571020 Tronzano L, Coscia GC, Capellaro F (1965). [Exposure and risk in the process of grinding talc.] Med Lav, 54:744--745 (in Italian). Tukiainen P, Nickels J, Taskinen E, Nyberg M (1984). Pulmonary granulomatous reaction: talc pneumoconiosis or chronic sarcoidosis? BrJIndMed, 41:84-87. PMID:6691939 Tye MJ, Hashimoto K, Fox F (1966). Talc granulomas of the skin. J Am Med Assoc, 198:1370-- 1372. doi: 10.1001/jama. 198.13.1370. PMID:5953727 Vallyathan NV, Craighead JE (1981). Pulmonary pathology in workers exposed to nonasbestiform talc. Hum Pathol, 12:28-35. doi: 10.1016/S0046-8177(81)80239-0. PMID:7203452 Venter PF, Iturralde M ( 1979). Migration of a particulate radioactive tracer from the vagina to the peritoneal cavity and ovaries. SAfr MedJ, 55:917-919. PMID:472930 Wagner JC, Berry G, Cooke TJ et al (1977). Animal experiments with talc. In: Walton WH, McGovern B, eds, Inhaled Particles, Vol. IV, Part 2, Oxford, Pergamon Press, pp. 647-- 654. Waller BF, Brownlee WJ, Roberts WC (1980). Self-induced pulmonary granulomatosis. A consequence of intravenous injection of drugs intended for oral use. Chest, 78:90-94. doi: 10.1378/chest.78.1.90. PMID:7471850 Wehner AP (1994). Biological effects of cosmetic talc. Food Chem Toxicol, 32:1173-1184. doi: 10.1016/0278-6915(94)90135-X. PMID:7813991 Wehner AP, Wilkerson CL, Cannon WC et al. (1977a). Pulmonary deposition, translocation and clearance of inhaled neutron-activated talc in hamsters. Food Cosmet Toxicol, 15:213-224. doi: 10.1016/S0015-6264(77)80392-1. PMID:892677 Wehner AP, Tanner TM, Buschbom RL (1977b). Absorption of ingested talc by hamsters. Food Cosmet Toxicol, 15:453^155. doi: 10.1016/S0015-6264(77)80013-8. PMID:598798 Wehner AP, Zwicker GM, Cannon WC (1977c). Inhalation of talc baby powder by hamsters. FoodCosmet Toxicol, 15:121-129. doi: 10.1016/S0015-6264(77)80317-9. PMID:873404 Wehner AP, Hall AS, Weller RE et al. (1985). Do particles translocate from the vagina to the oviducts and beyond? Food Chem Toxicol, 23:367-372. doi:10.1016/0278-6915(85)90073-0. PM1D:4040089 Wehner AP, Weller RE, Lepel EA (1986). On talc translocation from the vagina to the oviducts and beyond. Food Chem Toxicol, 24:329-338. doi: 10.1016/0278-6915(86)90011-6. PM1D:3525355 Weissberg D, Ben-Zeev I (1993). Talc pleurodesis. Experience with 360 patients. J Thorac Cardiovasc Surg, 106:689-695. PMID:8412264 Wells IP, Dubbins PA, Whimster WF (1979). Pulmonary disease caused by the inhalation of cosmetic talcum powder. Br J Radiol, 52:586-588. doi: 10.1259/0007-1285-52-619-586. PMID:465949 Wendt VE, Puro HE, Shapiro J et al. (1964). Angiothrombotic pulmonary hypertension in addicts. `Blue velvet' addiction. J Am MedAssoc, 188:755-757. PMID:14122687 TALC 405 Werebe EC, Pazetti R, Milanez de Campos JR et al (1999). Systemic distribution of talc after intrapleural administration in rats. Chest, 115:190-193. doi: 10.1378/chest. 115.1.190. PMID:9925083 Woodworth CD, Mossman BT, Craighead JE (1982). Comparative effects of fibrous and nonfibrous minerals on cells and liposomes. Environ Res, 27:190-205. doi: 10.1016/00139351(82)90070-6. PM1D:6279387 Yao Y-T, Wang N-S, Michel RP, Poulsen RS (1984). Mineral dusts in lungs with scar or scar cancer. Cancer, 54:1814-1823. doi:10.1002/1097-0142( 19841101)54:9<1814::AIDCNCR2820540909>3.0.CO;2-V. PMID:6478417 406 IARC MONOGRAPHS VOLUME 93 5. Summary of Data Reported 5.1 Exposure data The term `talc' refers to both mineral talc and industrial mineral products that contain mineral talc in proportions that range from about 35% to almost 100% and are marketed under the name talc. Mineral talc occurs naturally in many regions of the world where metamorphosed mafic and ultramafic rocks or magnesium carbonates occur. Mineral talc is usually platy but may also occur as asbestiform fibres. (Asbestiform refers to a habit (pattern) of mineral growth and not to the presence o f other minerals. Asbestiform talc must not be confused with talc that contains asbestos.) Together with platy talc, asbestiform talc is found in the Gouverneur District of New York State, USA, and occasionally elsewhere; it may be associated with other minerals as observed by transmission electron microscopy. Talc products vary in their particle size, associated minerals and talc content depending on their source and application. Minerals commonly found in talc products include chlorite and carbonate. Less commonly, talc products contain tremolite, anthophyllite and serpentine. Mineral talc is valued for its softness, platyness, inertness and ability to absorb organic matter. It is used in agricultural products, ceramics, paint and other coatings, paper, plastics, roofing, rubber, cosmetics and pharmaceuticals and for waste treatment. Cosmetic talc, which contains more than 90% mineral talc, is present in many cosmetic products and is used for many purposes, including baby powders and feminine hygiene products. The type of talc that is currently used for cosmetic purposes in the USA does not contain detectable levels of amphibole, including asbestos. It is not known whether this is true in other countries. Workers are exposed to talc during its mining and milling. Reported geometric mean exposure levels to respirable dust are typically in the range of 1-5 mg/m3. Workers may also be exposed in user industries, primarily in the rubber, pulp and paper and ceramics industries. Due to the presence of other particulates, exposure levels may be difficult to measure accurately. Consumer exposure by inhalation could occur during the use of loose powders that contain talc. Accurate estimates of prevalence are not available. However, in some series of controls from epidemiological studies of ovarian cancer, the prevalence of use for feminine hygiene o f body powders, baby powders, talcum powders and deodorizing powders, most of which contain cosmetic talc in varying amounts, has been reported to be as high as 50% in some countries. Perineal use for such purposes seems to have been a common practice in Australia, Canada, the United Kingdom, the USA and other countries, including Pakistan. Use of cosmetic talc in the USA has declined steadily since the late 1970s. TALC 407 5.2 Human carcinogenicity data The carcinogenic effect of exposure to talc not contaminated by asbestos fibres has been investigated in five independent but relatively small cohort studies of talc miners and millers in Austria, France, Italy, Norway and the USA. The miners and to a lesser extent the millers in these cohorts were also exposed to quartz. In a case-control study nested in the combined cohorts of talc workers from Austria and France, there was no tendency of higher risks for lung cancer by increasing cumulative exposure of workers to talc dust. In four of five studies, it was explicitly stated that no case of mesothelioma was observed. In the two studies from Italy and Norway, which included an estimate of cumulative exposure of the cohort to talc dust, the risk for lung cancer in the highest category was found to be close to or below unity. In the subgroup of miners in the study in the USA, an excess risk for lung cancer was found, which may be have been due to exposure in the workplace to radon daughters and quartz. In all the other groups o f workers studied, there was no increased risk for lung cancer. Female workers in the Norwegian pulp and paper industry had an increased risk for ovarian cancer, which, however, was attributed to exposure to asbestos. A communitybased case-control study did not find an increased risk for ovarian cancer associated with occupational exposure to talc, but the prevalence o f exposure was low. Body powder containing talc has been used by women on the perineum (or genital area), on sanitary napkins and on diaphragms. In total, data from one prospective cohort study and 19 case-control studies were reviewed in the evaluation of the association of cosmetic talc use and the risk for ovarian cancer. The information collected on perineal talc use varied substantially by study (e.g. ever use versus regular use, and whether information on the mode of application, frequency or duration of use was available). The cohort study was conducted among nurses in the USA and included 307 cases of ovarian cancer that occurred over 900 000 person-years of observation and a maximum of 14 years of follow-up. Information was collected on the frequency but not duration of regular use. Perineal use o f talc was not associated with a risk for ovarian cancer. The 20 case-control studies were conducted in Australia, Canada, China, Greece, Israel, Norway, the United Kingdom and the USA (nested case-control study), and included between 77 and 824 cases and 46 and 1367 controls. Five were hospital-based designs and the others were population-based studies. The Working Group designated a subset o f these studies as being more informative based on the following characteristics: the study was population-based, was of a reasonable size, had acceptable participation rates and included information to allow control for potentially important confounders. Eight population-based case-control studies from Australia, Canada (Ontario) and the USA (two non-overlapping studies in Boston, MA, and one each in California, Delaware Valley, eastern Massachusetts and New Hampshire and Washington State) were thereby identified as being more informative. The selected studies included at least 188 cases and had participation rates that generally ranged from 60 to 75%. Among these eight studies, the prevalence of use of body powder among controls ranged from 16 to 52%; however, 408 IARC MONOGRAPHS VOLUME 93 information on exposure was not collected in a comparable manner across studies. In addition, the frequency and duration of use or total lifetime applications were investigated in several studies as well as consideration of prior tubal ligation or simple hysterectomy. Only sparse data were available on whether women had used body powder before or after the mid-1970s. The relative risks for ovarian cancer among users of body powder (versus non-users) were homogenous across this relatively diverse set of eight studies, each of which indicated a 30-60% increase in risk. Among the other 11 case-control studies, most also reported relative risks o f this magnitude or higher. The subset of studies that assessed use of talc on a diaphragm were relatively uninformative due to their lack of precision. Results on exposure-response relationships were presented in the cohort study and in seven o f the more informative case-control studies. In the cohort study, no exposureresponse trend was apparent. Positive exposure-response trends were apparent in the two Boston-based studies that presented the most comprehensive analysis. In the Canadian and Californian studies, a non-significant, weakly positive trend was observed for either duration or frequency of use, but not for both. In the other three case-control studies, no consistent trend was observed and the strongest associations tended to be seen among the shorter-term or less frequent talc users. The cohort study and four of the eight more informative case-control studies presented results on histological type of ovarian cancer. When the analysis of the cohort study was restricted to the 160 serous invasive cases, a statistically significant increase in risk of about 40% was observed. The risk increased with increasing frequency of body powder use. Risks for serous ovarian cancer were somewhat greater than those for other histological types in two of the four case-control studies in which the contrast was reported. Results for other histological types were inconclusive. The Working Group carefully weighed the various limitations and biases that could have influenced these findings. Non-differential misclassification of talc use, given the relatively crude definitions available, would have attenuated any true association. Although the available information on potential confounders varied by study, most investigators accounted for age, oral contraceptive use and parity. In most studies, only the adjusted relative risks were presented; however, in the three studies in which both ageadjusted and fully adjusted estimates were provided, relative risks did not differ materially, suggesting minimal residual confounding. It is possible that confounding by unrecognized risk factors may have distorted the results. One or more such factors, if they are causes of ovarian cancer and also associated in the population with perineal use of talc, could induce the appearance of an association between the use of talc and ovarian cancer where there is none. In order for such an unrecognized risk factor to induce the consistent pattern of excess risks in all of the casecontrol studies, it would be necessary for the factor to be associated with perineal talc use across different countries and different decades. While the range of countries and decades covered by the more informative case-control studies is not very broad, it provides some TALC 409 diversity of social and cultural context and thereby reduces the likelihood of a hidden confounder. There was a distinct pattern of excess risk discernible in all of the case-control studies when users were compared with non-users; however, methodological factors needed to be considered. First, while chance cannot be ruled out as an explanation, it seemed very unlikely to be responsible for the consistent pattern of excess risks. A second possible explanation would be recall bias, to which case-control studies may be particularly susceptible. This may have been the case if there had been widespread publicity about the possible association between the use o f body powder and cancer. In such circumstances, it is possible that women who had ovarian cancer could be more likely than women who did not to remember or over-report a habit, such as body powder use, if they thought that it may have played a role in their illness..There was a flurry of publicity in the USA in the mid-1970s concerning the possible risks for cancer posed by the use of talc-based body powders. Following an industry decision to market talc powders with no asbestos, it was the opinion of the Working Group that there had not been widespread public concern about this issue, at least until veiy recently. Therefore, the Working Group considered it unlikely that such a bias could explain the set of consistent findings that stretch over two decades. The Working Group believed that recall bias was a possibility inherent in the case-control studies and could not be ruled out. The Working Group also considered publication and selection biases and these were not judged to have substantially influenced the pattern of findings. The Working Group searched for documentation on the presence of known hazardous minerals in talc-based body powders. There were strong indications that these products contained quartz in the mid-1970s and still do. There were also indications that occasional small concentrations of asbestos were present in these products before the mid-1970s, but the available information was sparse, sampling methods and detection limits were not described, and the range of locations where data were available was extremely limited. As a result, the Working Group found it difficult to identify a date before which talc-based body powders contained other hazardous minerals and after which they did not, or to have confidence that this would be applicable worldwide. In addition, the epidemiological studies generally do not provide information about the years during which the female subjects were exposed. Consequently, the Working Group could not identify studies in which an uncontaminated form of talc was the only one used by study subjects. Nevertheless, the Working Group noted that, even in the most recent studies in the USA, where exposure histories may have been much less affected by hazardous contaminants of talc, the risk estimates were not different from the early studies in which the possibility of such exposure was more likely. To evaluate the evidence on whether perineal use of talc causes an increased risk for ovarian cancer, the Working Group noted the following: The eight more informative case-control studies, as well as most of the less informative ones, provided overall estimates of excess risk that were remarkably consistent; seven o f these eight case-control studies examined exposure-response 410 IARC MONOGRAPHS VOLUME 93 relationships; two provided evidence supporting such a relationship, two provided mixed evidence and three did not support an association. The cohort study neither supports nor strongly refutes the evidence from the casecontrol studies. Case-control studies were susceptible to recall bias which could tend to inflate risk estimates but to an unknown degree. All of the studies were susceptible to other potential biases which could either increase or decrease the association. All of the studies involved some degree o f non-differential misclassification of exposure that would tend to underestimate any true underlying association. 5.3 Animal carcinogenicity data Talc of different grades was tested for carcinogenicity in mice by inhalation exposure, intrathoracic, intraperitoneal and subcutaneous injection, in rats by inhalation exposure, intrathoracic injection, intraperitoneal injection, oral administration and intrapleural and ovarian implantation, and in hamsters by inhalation exposure and intratracheal injection. In male and female rats exposed by inhalation to a well-defined talc, the incidence of alveolar/bronchiolar carcinoma or adenoma and carcinoma (combined) was significantly increased in female rats. The incidence o f adrenal medulla pheochromocytomas (benign, malignant or complex (combined)) showed a significant positive trend and the incidence in high-dose males and females was significantly greater than that in controls. The incidence of malignant pheochromocytomas was also increased in high-dose females. The Working Group did not consider it probable that the increased incidence of pheochromocytomas was causally related to talc but, based on the experimental data available, neither could talc-related effects be excluded. Tumour incidence was not increased following the intrapleural or intrathoracic administration of a single dose of various talcs to rats. In two studies of intraperitoneal administration in rats, no increase in the incidence of mesotheliomas was observed. No increased incidence of tumours was produced in rats in two studies of talc administered in the diet or in another study of the implantation of talc on to the ovary. Tumour incidence was not increased in mice following the inhalation of talc in one study, the intrathoracic administration of a single dose of various talcs in another study or the administration of talc by intraperitoneal injections in three studies. A single subcutaneous injection of talc into mice did not produce local tumours. Tumour incidence was not increased following inhalation or intratracheal administration of talc to hamsters. 5.4 Mechanistic considerations and other relevant data Different mechanisms are probably operative in the effects of talc on the lung and pleura, depending on the route o f exposure. TALC 411 In humans, deposition, retention and clearance of talc have been insufficiently studied, although talc particles have been found at autopsy in the lungs of talc workers. In humans and experimental animals, the effects of talc are dependent on the route of exposure, and the dose and properties of the talc. Talc pneumoconiosis was somewhat more prevalent and severe among miners exposed to talc containing asbestiform minerals and/or asbestos than among those exposed to talc without such contaminants. However, the role of quartz and asbestos in the observed pneumoconiosis could not be ruled out. Among drug users, intravenous injection of talc present as a filler in the drugs resulted in microembolization in a variety of organs and alterations in pulmonary function. In animal studies, talc has been shown to cause granulomas and mild inflammation when inhaled. Observations of the effects that occurred in the lungs of rats exposed by inhalation to talc suggested that the operative mechanisms may be similar to those identified for carbon black, and talc is known to cause the release of cytokines, chemokines and growth factors from pleural msothlial cells. In humans, intrapleural administration of talc as a therapeutic procedure results in pleural inflammation which leads to pleural fibrosis and symphysis. Pleural fibrosis is the intended effect of intrapleural administration of talc in patients with malignant pleural effusions or pneumothorax. Animal studies suggested that extrapulmonaiy transport of talc following pleurodesis increases with decreasing particle size and increasing administered dose. Talc has been shown to cause apoptosis o f malignant cells in vitro. Perineal exposure to cosmetic talc in women is o f concern because of its possible association with ovarian cancer. Several studies have been conducted in women to assess potential retrograde movement of particles through the reproductive tract to the ovaries. These have been conducted in women who were about to undergo gynaecological surgery, most of whom had diseases or complications o f the reproductive tract and organs that required surgery. The findings reported in these studies may be confounded by the various levels of dysfunction in clearance from the female reproductive tract due to underlying pathologies. In addition, most of the studies had little or no further information on the use of talc products for perineal hygiene or changes in habits that may have preceded surgery. On balance, the Working Group believed that the evidence for retrograde transport of talc to the ovaries in normal women is weak. In women with impaired clearance function, some evidence of retrograde transport was found. Studies in animals (rodents, langomorphs and non-human primates) showed no evidence of retrograde transport of talc to the ovaries. In one study, predictors of the presence of antibodies to mucin protein were inversely related to the risk for ovarian cancer and exposure to powder containing talc. No data were available on the genotoxic effects of exposure to talc in humans. The limited number of studies available on the genetic toxicology of talc in vitro gave negative results. 412 IARC MONOGRAPHS VOLUME 93 6. Evaluation and Rationale 6.1 Cancer in humans There is inadequate evidence in humans for the carcinogenicity of inhaled talc not containing asbestos or asbestiform fibres. There is limited evidence in humans for the carcinogenicity of perineal use of talcbased body powder. 6.2 Cancer in experimental animals There is limited evidence in experimental animals for the carcinogenicity of talc not containing asbestos or asbestiform fibres. 6.3 Overall evaluation Perineal use of talc-based body powder is possibly carcinogenic to humans (Group 2B). Inhaled talc not containing asbestos or asbestiform fibres is not classifiable as to its carcinogenicity (Group 3). 6.4 Rationale In making this evaluation the Working Group considered the human and animal evidence as well as evidence regarding the potential mechanisms through which talc might cause cancer in humans. The Working Group found little or inconsistent evidence of an increased risk for cancer in the studies of workers occupationally exposed to talc. The studies of talc miners and millers were considered to provide the best source of evidence, but no consistent pattern was seen. One study observed an excess risk for lung cancer among miners, but confounding from exposure to other carcinogens made it difficult to attribute this to talc and no excess risk was seen in millers. Other studies also found no increased cancer risk or no higher risk with increasing cumulative exposure. Overall, these results led the Working Group to conclude that there was inadequate evidence from epidemiological studies to assess whether inhaled talc not containing asbestos or asbestiform fibres causes cancer in humans. For perineal use of talc-based body powder, many case-control studies of ovarian cancer found a modest, but unusually consistent, excess in risk, although the impact of bias and potential confounding could not be ruled out. In addition, the evidence regarding exposure-response was inconsistent and the one cohort study did not provide support for an association between talc use and ovarian cancer. Concern was also expressed that TALC 413 exposure was defined in a variety of ways and that some substances culled talc may have contained quartz and other potentially carcinogenic materials. A small number of Working Group members considered the evidence to be inadequate. Despite these reservations, the Working Group concluded that the epidemiological studies taken together provide limited evidence of an association between perineal use of talc-based body powder and an increased risk for ovarian cancer. In one study of rats that inhaled talc, an excess incidence of malignant lung tumours was seen in females. The same study observed an excess incidence of pheochromocytomas in the adrenal medulla in both sexes, but the Working Group was divided as to whether these rare tumours could be attributed to exposure to talc. Other studies in rats and mice using different routes of administration did not find an excess of cancer, and two studies in rats were considered to be inadequate for evaluation. Based on the one positive study, the Working Group found that there was limited evidence of carcinogenicity of inhaled talc in experimental animals. There was no agreement within the Working Group as to whether the evidence on pheochromocytomas should be taken into account in the evaluation of animal data.