Document baXJXkqpOG6Benrv5b0G3xnrk

Review f@-K LIBRARY OF THE HEALTH 8 C I c N Q t S UNIVERSITY O F PIT?SBURGn 'QL'TlfSSlURW, PF(NSYLVANU IS261 Multiple Myeloma: Clusters, Clues, and Dioxins \htrJct ,,ultjple m e l o m a (Sfkl)is a B-cell neoplasm of unknown diulog,. ive searched for etiological clues by examining ,bliterature on geographic clusters of MM. We searched ,hc X I E D L I ~dEatabase from 1966 to 1996 for spatial m u r r e n ~ eo~f V.Cl that were significantly greater than ted tspatial .'clusters"). Eight clusters with verified Clpec diagnoses of M\I were identified. All of the eight clusters *.urred in locations that were proximate to a body of rater. Six of these bodies of water are known to have kncontaminated with dioxins. tve h! pothesize that the observed association h t w e n SISI and proximity to bodies of water is caused b%exposure to dioxins in individuals who consume local &h and seafood. This hypothesis is consistent with the ~ifificantly elevated risks for bfM?riin groups with high M u m p t i o n of dioxin-contaminated fish, e.g., Baltic Sea frhermen and Alaskan Indians. and among persons accidentally exposed to dioxins in Seveso, Italy. Dioxins UT irnmunotoxic and inhibit the differentiation of B cells. mus. dioxins are plausible myelomagens. A dioxin hypothesis could illuminate many epidemiological features \IS1 and may suggest new avenues for analytic research. introduction S n l : I\ ;1hematological cancer characterized by the malignant proliferailon of plama cell3 and plasma cell precursors (immuno~i~,bulin-prlucincgells of the B-cell lineage). It is the mnrnone\i clinical \yndrome of a spectrum of disorders. the plasma id1 dyxra3ias or monwlonal gammopathies. Common io rhe-rt Jiwrden ih the proliferation of a \ingle clone of :mmu:i~~gl~)bulin-~ccretcienlgl\ and the resultant increase in rerum kvcI\ ot' J \ingle immunoglobulin I ). \lJl xcounts for approximately I4 of all malignancies in the United States and affects Blacks twice 3s often as 'A%ites. The annual age-adjusted I1970 U.S. standard) incidence rates for M M are 1.7 and 3.2 per 1OO.OOO for White men and \\-omen (respectively)and 10.2 and 6.7 for Blacks. The racial difference in incidence rates does not appear to result from differences in socioeconomic factors (2). The etiology of M M is obscure. Exposure to ionizing radiation is an established risk factor. However. because few people have significant exposure to ionizing radiation. most cases of %lMmust have other causes. Other risk factors for XtM are less well established and include exposure to farming, pesticides. and petroleum producrs (3). Because .MAWis a disorder of proliferating immune cells. "chronic antigenic stimulation" has been proposed as an etiological factor (4,5 ) . HOWever. the results of most investigations of that hypothesis have been negative (2. 3). We recently investigated a cluster of M M that was located near a large body of water." Review of the literature revealed that several other clusters of .Mb! have been reported; these also were located near bodies of water. The association between MM and bodies of water has not been widely recognized. however. and suggests a role for environmental pollutants. This suggestion is strengthened by the observation that many of these bodies of water have been contaminated with dioxins. The purposes of this paper are to: ((1) review the literature on clusters of MM: f b )review the evidence of dioxin contamina- tion of these bodies of water: and ( c )develop the hypothesis that dioxins are a biologically plausible cause of MM.Before reviewing these clusters. we briefly review some toxicological properties of dioxins. Dioxins. In general usage. the term "dioxin" refers t o the members of three chemical classe\ with similar biological and toxicological effects. These are the PCDDs. PCDFs. and coplaner PCBs (6). Dioxins are ubiquitous environmental pollutants that are formed 1s by-products in many industrial activities that use chlorine. The best studied and most toxic ofthe dioxin5 is TCDD. a contaminant in the \ynthehis of ~.l.j-trichlorciphcnox?acetic acid. an herbicide and defoliant ( 7. 8 I. TCDD produces a variety of to.xic effects in laboratory animal\ at very low daws. It is cmbryotoyic. teratogenic. and carcinogenic (9).The immune systems in m m a l s are panicularly sensitive to TCDD. Impairments in both T-cell- and Bcell-mediated immune functions occur at doses of TCDD that do not produce overt signs of general toxicity I 101. Although TCDD is il known carcinogen in animals. the carcinogcnic effects of TCDD in humans have been more difficult to rvaluate. IARC has classified TCDD and occupational exposure to phenoxy acid herbicides as a possible human carcinogen (group 29: Ref. 1 1 1. Over 9 5 8 of human exposures to dioxins comes from food. Approximately 1% of dioxin exposure comes from breathing contaminated air, and a smaller amount comes from 'G. G. Schuanz. R. D. Rdnguez. P Gomet.. R. Duncan. and E. T T r ~ p l d ~ . unpublkhed Jaia. 50 Review: .Multiple .Myeloma and the Dioxin Hypothesis drinking contaminated water (6). Dioxins are lipophilic and accumulate in fatty tissue such as milk. meat, and fish. Con- taminated fish and shellfish are an important source of dioxins in people with a diet high in fish ( 12). Sport-caught fish is the major source of dioxins for the general population ( 13) ( 11). Materials and Methods We searched the MEDLINE data base (in all languages) from Januap 1966 to January 1996 in an attempt to identify all spatial clusters of MM. We did not restrict the search to reports that used the word "cluster"in the title or abstract. Each cluster was examined for references to other clusters. The definition of a spatial disease "cluster" and of a cluster's geographic boundaries is the subject of debate (IS). We detined a "cluster" as a statistically significant excess of disease in a definable location. We accepted the geographic bounds of the clusters as those given by the authors of the individual reports. To be included, a cluster must have occurred within a defined geographic area. such as a city. county. or administrative district. We excluded clusters that lacked verification of MM. either via laboratory tests or death certificates. We evaluated each cluster with respect to its proximity to a body of water. A cluster was considered "proximate" to a body of water if the geognphic bounds of the location. as seen on our atlas (Today's World Ref. 16). included a coastline. river. or lake. For clusters within the continental U.S.. proximity to bodies of water was evaluated using MapQuest (GeoSysterns Global Corp.). an intenctive atlas available on the Internet. This permitted resolution to a scale as large as I inch = I .7 miles. We used MEDLINE to search for possible contamination of these bodies of water using the text words "water pollution," "fishes." "organochlorines." "PCBs." and "dioxins." Because carnivorous marine mammals are at the apex of the marine food pyramid. dioxin levels in these mammals. where available, were used as an indicator of marine pollution ( 17). To evaluate bias in the selection of locations near water, we estimated an "expected" prevalence of proximity to bodies of water using "control" locations. Using the index of our atlas ( 16). we selected 50 control locations as follows. Two locations were randomly selected from each letter of the alphabet. A-Y. A table of random numbers was used to select control locations. A random number. r. indicated the rth location listed under the designated letter in the index. For example. under letter "A." the random number "IO" indicated the 10th location. "Aba. Nigeria." Number "21" corresponded to "Abagner Qi. China." We repeated this procedure for each succeeding letter. We excluded locations with known clusten of MM and locations that were purely aquatic ie.g.. the Pacific Ocean). as well as names indicating topographical features ( e . g , .Mt. Everest,. Proximity to water in the control locations was evaluated using the same methods described above. Results Eleven clusters were identified using the criteria described above. Four were excluded: one due to lack of geographic information (18): two due to lack of disease verification (19. 10);and one because it was a preliminary report of a cluster that was published subsequentl? after longer follow-up (21). Thus. seven published clusters were available for review. We included our recent cluster. for a total of eight. Geographic Clusters of M.M and Dioxin Contamination Thief River Falls. blN. The tint reported cluster of MM occurred i n Thief River Fulls. M N ( 2 2 ) .Thief River Falls ( 1970 population of 7151) is located 50 miles northeast of Grand Forks. ND. The Thief River Falls cluster came to attention when six cases of MM were diagnosed in the year 1968 (0.2.; expected). When incidence rates were evaluated for the time period 1960 through 1968. the crude incidence rate remained -elevated at 10.6 per 100,OOO per year (3.0expected, based on cases). Of the seven cases. all were White: five were female. Two patients were husband and wife. None of the patients had a medical histop of significant radiation exposure or unusu;d histon of cancer. None shared a common association save for their residence in Thief River Fltlls. All patients lived in widel! %cartered Ioccutions throuphout the town. The authors had t i t i explanation for thi5 cluster. Dioxin LPveIs. Thief River Falls is proximate to the Thief a d Red Lake Rivers. Historical data on dioxin levels were not available in these locations. Aberdeen. Scotland. The second cluster was reported in 197.;. when Dawson and Ogston drew attention to the high rates 11t "myelomatosis" in Northeast Scotland ( 1 3 ) .("Myelomatosi~" includes MM and plasmacytoma. Le.. a solitary plasma 4 1 tumor without evidence of disseminated disease.) The authors studied incidence and mortality from myelomatosis and usual place of residence for the period 1960-1969. Compared to the whole of Scotland. crude incidence and mortality rates were significantly higher in Aberdeen (3.4 versus 1.8 per 1OO.OOO per year. incidence; 2.65 versus I .67 per 1OO.OOO per year. mortality). Eight % of the incident cases of myelomatosis ( I 3 ot 153) were plasmacytomas. and their geographic distribution paralleled the distribution of MM. There was no evidence of clustering of myelomatosis in time. The authors had no explanation for the elevated rates in Aberdeen. Dioxin Levek. Aberdeen is a coastal city proximate to the Dee and D o n Rivers and to the North Sea. Organochlorine residues were reported from prey seals (Halichnerusgnpus). hiup seals (Phoca groenlundicu). and porpoises (Phocucna phocoerld) taken from varying locations of Scotland's coastal waters. Pesticide contamination was greater in the seals and porpoises o n the east coast of Scotland (taken from an area from Aberdeen to Tay estuary) than in specimens from the west and nnnh coasts. Total residue concentrations of pesticides in seal bluh- ber were as high as 73.3 ppm pesticides (dieldrin + total DDT: Ref. 21). Western Ireland. In the third cluster, Greally er ul. ( 2 5 )n o d an increased incidence of monoclonal gammopathy (the presence of a monoclonal protein in serum or Urine, and MM in the West of Ireland. The incidence of monoclonal gammopathlej was determined by examining laboratory records from hospirds that performed serum protein electrophoresis. Patients d I monoclonal gammopathy were followed-up using hospltd, records. From Januap 1976 to August 1983. 196 case5 of monoclonal gammopathy and I 17 cases of MM were obsen 2d. The incidence of MM was highest in hospitals of the Western Health Board. serving Counties Galway. Mayo. and Ros~ommon. The crude incidence rate for MM in the Western Heal& Board was 1.5 per I oO.o(X) per year. it=rsus 1 . I per 1OO.OOo Fr year for the South of Ireland (Counties Cork and Ken?',. though these rates were not age adjusted. the age structure.' O' West and South Ireland populations reportedly do not differ. The authors speculated that the higher rate!, for MM in n'r9 Ireland could be due to higher levels of background ndiu:lon. However. they acknowledged that studies of cancer m o d t r ) ' and background radiation levels in Ireland have been nqarl''' ( 2 61. Canrcr EpidemiolopLI.. H i o m a r t e n 1 Prcvcnlion 51 dtion. Djur;n Levels. The Nest coast of Ireland borders the Atlantic OrgancKhlorine pesticide residues and K B s were ex- aJimchilnoreoIn~di:?pheonyttIedriscthlluotrrouetlhuarnrterI)DfrDoEm).IraenladndPC12B7s). Dieldrin. were de- tected in all samples and were highest in otters from harbor ye3p, Elevated dioxin levels were observed in two of the three ,-ollected in an area corresponding to the Western Health Dl,tdct {total organochlorine Ieveh of 109 m g k g and 170 m p ~ ; P j . Levels of mercury. another common pollutant. were i.1evated. petcrsburg, VA. The fourth cluster was described by Ende in peter,burp. Virginia (28).Dunng 1971-1975. 21 cases of MM Jiagnoscd in the city of Petersburg. Fifteen of these ,-cu~ed in a smaller zone of the city. termed the "target area." pc,pulation I5.OOO- 17.000. The crude incidence rate was calculated as 11-12.5 per 100.O00 per year. The target area was Black. Of the 15 cases, 7 were Black males. 7 uere Black females. and I was a White male. Ende reported [hat "an effort was made to see if there could be an association betueen multiple myeloma and Kepone. the chemical reported to ha\e deleteriously affected workers recently in the nearby community o f Hopewell. Virginia." Ende questioned seven >p,1 patients in his care but found no association with Kepone below i . He noted that four incident cases of MM were seen during 1971-1978 in the Hopewell Hospital. Dioxin Levels. Petersburg is located in eastern Virginia proximate to the James md hppomattox Rivers. an area that was histoflcdly contaminated with the pesticide. Kepone. The his- tory of the Kepone disaster has been comprehensively reviewed (29. 30). Brietly. Kepone (chlordecone) is a persistent poly- chlorinated hydrocarbon pesticide that is no longer manufac- tured in the U.S. From 1966-1974. 1.6 million pounds of Kepone were produced by Allied Chemical Corporation in its Hopewell. V h plants. The last operating plant was closed as ;1 public health hazard in 1975 (31). Kcpone produces a range of acute toxicities in laboratory animals and is a carcinogen in rodents (32). Although the first cases of Kcpone toxicity in humans were recognized in 1975, subsequent investigations revealed that m v i n e life in the Tide- water region of Virginia had been contaminated with Kepone for a tiecade. Kepone from the Hopewell sewage treatment plant had contaminated local u a t e n s a y . Air. water. and especiail? tish. were g r o 4 y contaminated. Kepone levels measured 0.3 ppb in the James River. These values were biomagnitied in f i h h Jnd shelltish. Thus. clams and oysters from the James Riier Jt locations 8-64 miles from Hopewell contained 0.2-0.8 pan\ ;)er m i [ / i o f ilppm 1 o f Kepone: liver and entrails of bass and brem had concentr3tiona up 10 I 4 ppin. Concern about widespread human exposure through the conbumption o f contaminated ~eatoodled to a government ban on taking shellfish and finfiah from [he James River. Peteraburp is Itxated 8.2. miles southwest of the Kepone plant at Hopewell. V A . Although none of the patients inter\-ieaed by Ende had worked at the Kepone plant. they may have been expowJ to Kepone unknowingly via the consumption of local fish and \hellIish. Island of Walney. The fifth .WM cluster occurred on the Island of Walney (population I1.ooO). a sUip of land separated by a narrow channel from the coastal town of Barrow in Furness, Cumbria. United Kingdom (33).Between 1974 and 1980.seven c a w of MIM were observed in Walney ver.w.7 2.8 expected IP < 0.03,). An epidemiological investigation was undertaken in response to residents' concerns about the etiological role of a nuclear reprocessing plant. located 15 miles up the coast at Sellafield. A case-control study was performed in South Cumbria (the health district that includes Walney Island) uhing 34 cases of X l M and 69 community controls. Cases and controls did not differ u ith respect to previous medical history. occupational exposures to radiation. or other subpected risk factors. However. the study uncovered one signi ticant difference: cases had resided longer in the seaside town of Barrow in Furness. Fortyeight 'Ir of the cases i w w s 259 of the controls had lived in Barrow for more than 20 years ( P = 0.05). The authors noted. "Barrow constitutes the Irish Sea coast of the disuict and almost all Barrow residences are within I O kilometers of the sea" (33). The authors considered that seaside residents might be exposed to higher doses of radioactive material via inhalation of sea spray but considered this hypothesis implausible. They concluded that the high incidence of MM in Walney Island was due to chance. Dioxin Levels. The Island of Wulney borders the Irish Sea. PCB levels uere reported for seals from the Liverpool Bay area of the Irish Sea (34). In three grey seals (H.gnpirs) examined in 1988. blubber values of PCBs ranged from 34 to 101 mgkg wet weight total PCBs. The author concluded. "These values are considered to be high and approach those found in common seals in [he polluted Dutch Wadden Sea" (34). Commenf. The case-control study described within this cluster is Important because it represents an association between MM and proximity to bodies of water at the level of the individual. Because coastal dwellers are known to consume more fish than inland residents (35. 36). the association of MM with proximity to the sea may reflect increased fish consumption. Baglan Bay, Wales. The sixth cluster of ,MMoccurred in the vicinity of Baglan Bay, Wales. Concerns about clusters of cancer in the vicinity of the Baglan Bay petrochemical works led to an evaluation of cancer incidence and mortality in the 7.5-km area surrounding the plant. Although there was no excess of cancers overall. the investigators found a significant excess of MM (42 deaths observed: 28.3 expected), predominantly among women. The authors were unable to suggest a possible cause of this increased mortality. Dioxin Levels. Baglan Bay. Wales is located near Baglan Bay and the Bnstol Channel. Reports of dioxin levels in these bodies oi water were not available. However. the petrochemical works in the vicinity would be expected to be a source of dioxins in the aquatic environment (37) (38). T Town. Japan. The seventh M M slubter was reported by Kosaka e! ( I / . 139)in T Town, Japan. T Town i\ a small seaside community t population %No) located at the mouth of a large nvrr uithin the houndaries of Tokushima City. During the period lY78-1988. seven incident cases of M X 1 rfour males. three females) were observed in T Toun for an annual ageadjusted incidence m e of 7.03 per 1o().OM)O (8.3I in males and 4.46 in temales). w r . ~ u s1.20 in Tokushima City and 1.53 for Japan as 3 w hole. None of the patients had a history of exposure to signiticant radiation. toxic chemicals. or chronic infection. Five of the seven patients had been fishermen since youth. and six had lived in a small area close to the tishing port near the river. The authors noted. "We were unable to find any etiologic factors in this group of tishermen, except for their abundant intake of sea food which could be polluted with some chemical agents.'' They observed. "all three communities which had myeloma clusters. namely Thief River h l l s . Petersburg and T Town. are located close to large rivers." and suggested that "additional studies of myeloma clusters throughout the world 52 Review. Multiple .Myeloma and Ihe Dioxin Hypothesis - Tublr I Location of geographic clusters of multiple myeloma. adjacent hodies of water. and evldence of dioxin contamination Location of cluster Year bod! of uater" Bod\ o i uater Dioxin contamination Thief River Falls, MS Yes Aherdeen. Scotlmd Yes We\tern Ireldnd Ye\ Pelenburg. V A Yes South Cumhria. Cli YC\ Baplan Ea). Wale\ Yes T Toun. Japan YC\ Oheeuhokc Count?, FL Y c.5 -' I SD.No data Slrcnglh 01 c\ldence- -. \trong. -. mcxleraie Thief Riier: Red h k e Riber . Dee Ri\er: Don Riicr: Nonh Sea Atlantic Ocean Jame\ Ri\er: ..\ppomattox River Insh Sea B;IgIan Ea!. Bnktoi Channel Ywhino Riier: O d a Bar Ki\\immec Rner: Lake Oheechtike ~ may provide further information on the etiology of this disease" (39,. Dioxin Levels. T Town. Japan is located near the Yoshino River and Osaka Ba!. .Measurements of dioxins in blue mussels (Myrilus t'dtthJ demonstrate that a wide coastal area of Osaka Bay is polluted with dioxins. The average concentration of PCBs was 65 ppb (40).The main source of these dioxins is fly ash from municipal waste incinerators (41). Studies of dietary dioxin intake of Japanese women from Osaka indicate that 176 of 177 of the daily diets analyzed were contaminated with PCBs. The daily intake of PCBs ranged from 0-334.0 pg/day and was highest among those who consumed fish (47,).Thus. these data indicate that individuals who consume fish from Osaka Bay have high exposures to dioxins. Okeechobee County. FL. The eighth MM cluster took place in Okeechobee County. Florida." During 1981-1993. 37 incident cases of MM were reported to the state cancer registry, for an annual age-adjusted incidence rate of 7.91 per 100.000, versus 3.46 for Florida. We confirmed both the diagnoses and the residential addresses for 30 of the cases. Twenty-eight of the cases were White. Fourteen cases were female. Using these verified cases only, the annual age-adjusted incidence rate for MM in Okeechobee County is 6.52 per 1OO.ooO. versus 3.44 for the rest of Florida (excluding Okeechobee County: P < 0.001). Examination of patients' medical records yielded no evidence of unusual prior radiation or history of cancer. Examination of m t e agricultural reports reveals that Okeechobee County is remarkable in two respects. It is the largest source of daip cattle in Florida. and it is the largest source of revenue from the sale of commercial fishing licenses (43).The latter reflects its proximity to Lake Okeechobee. the state's largest freshwater lake. Dioxin Lpvefs. Okeechobee. Florida is located in southkentral Florida along the northern shore of Lake Okeechobee. Chlorinated hydrocarbon insecticides have been heavily applied in the agricultural economy of southern Florida. Surface runoff. including sediment and absorbed chlorinated hydrocarbons. is transported by a system of levees. canals, and water pumping stations from the Everglades into Lake Okeechobee. This has resulted in high dioxin levels in lake sediments 144)M. aximum concentrations of dioxins in sediments of southern Florida were reported to be 6 pgAg for dichlorodiphenyldichloroethane (DDD) and 9 pglkg DDD (45). In summary. we reviewed eight clusters of MM from diverse locations around the world (Table 1). ,411 of the eight clusters were located proximate to a body of water. At least six of these bodies of water have k e n contaminated with dioxins. Data on dioxin contamination were not available for two sites. However. one of these bites (BagIan Bay) is proximate to an industrial activity knoNn to be a source of dioxins. For comparison. we evaluated proximity to water in 3) randomly selected "control"locations not associated with clubters of .MM. Thirty-two of the 50 locations (64%)were aswciated with water. 18 (36%) were not. Thus. the probability thJr eisht of eight sites would be associated with a body of w l s r solely by chance is low ( P = 0.04. Fisher's exact test). The Dioxin Hypothesis We suggest that a plausible explanation for the association of MM with proximity to bodies of water is the following: ( a )Dioxin contamination of water results in the accumulation of dioxins in fish and seafood: ( bH~igh consumers of fish and seafood have elevatcd body levels of dioxins; and ( c )Humans with high exposures to dioxins arc at increased risk for MM. The evidence for ( a )has already been addressed (12-14). The evidence for ( b )and ( c ) is summarized below. (b) High Consumers of Fish and Seafood Have Elevated Bod-v Lovels of Dioxins Data from several populations indicate that the consumption of dioxin-contaminated fish is associated with elevated levels dioxins in blood. For example. serum PCB levels in Japanc,c men increased almost ?-fold (e.g.. from 2.5 to 4.0ppb) sevcd hours after consuming fish containing 181 pg of PCBs r&J. Additionally. mean serum PCB and DDT levels were signtfi- cantly higher in consumers of Great Lakes sports fish than in controls I 19.0 ivrsus 6.8 ppb for PCB: 15.6 versus 6.8 ppb DDT: 112 fish-eaters versus 95 controls: Ref. 47). Similar findings have been observed among fish-eaters in Sweden I&), Even greater differences in serum dioxin levels between t i h eaters and non fish-eaters were reported among the Inuit. Dewailly et al. (49)found that the total dioxin toxic equivalens among 10 fishermen with high consumption of fish from [he contaminated St. Lawrence River were 900 ngkg compared t@ 36 n:Ag for controls. Svensson er a/. (50) measured reponed fish consumptll'n and exposure to persistent organochlorine compounds in t\\O cohorts of Swedish fishermen. The fishermen as a group almost twice as much fish as non-fishermen referents from [he general Swedish population. Fishermen from the eastc('Jsl (Baltic Seal. which is heavily polluted with dioxins. ate more fatty fish (salmon and herring) than fishermen from the rdatively unpolluted westcoast. Eastcoast fishermen had 2-6 tilllCs higher total dioxin toxic equivalents of P C D D K D F and PCBP in their plasma than both westcopst fishermen and non-fi*hcrmen referents from the east and wtfstcoasts. Thus. blood k \ S l Z of dioxins in tishernlen were proportional to the estin1JIed consurnpiion of dioxin-contaminated fish. (c) Hum Increast The diol sumptior risk for ' that ha\: in indiv: groups t: Alaskan Alaskan mammai of the cii eludes Ir dietvy :. although to the hi; The rine cont. concenn among .4 Compare phom ar -Wlerea. similar tC from the than the High rate: ing in th: waage I1 -hl Additions of organc mcse dat:! a.iatec hIndia: Swedish (58) corn1 coast (i.e. hhenner, the rrpior. fish (Standard ; Moreover m o n g the -en that b o , 3.1 for th m n g su W y in 5 mxyacetic (60). Thus diouns Inl %so, It; mental Numerous pten kwmim MtbeSe st; thbias c: Of exwsur ~ v e s o .It. Cancer Epidemiology. Biomnrkers 81 Prevention 53 ,c, Hun,ans with High Exposures to Dioxins Are at ,ncrrasrd Risk for .VM J,o'cln hypothesis predicts that individuals with high con- [,,,,, (, dioxin-contaminated fish should be at increased ,"n.u.tnilP,or blb1. We are not aware evaluated the risk of of my epidemiolo,~ blkl with respect to i c asltudies this factor ih31Indi\,jUa]s. However. epidemiological data exist for two rn,up, [hat have high exposures to dioxin-contaminated fish: ;la>kan yatives and Swedish fishermen. \,*kan Natives. Because of their diet high in fish and sea I j I I. Alaskan Natives comprise a natural experiment ,,( [he dioxin hypothesis. The Alaskan Native population in- ilu~e, [nuits tEskmos). Indians. and Aleuts. The majority of JIL.lary information on Alaskan Natives concerns the Inuit. Jlth,,ugh the diets of Indians and Aleuts are similar with respect ((, [he high consumption Of fish and marine mammals ( 5 2 ) . The Inuit diet is known to have high levels of organochlorine iontaminants. and these contaminants are present in high ,oncentrations in blood and other tissues (53).Cancer incidence Jmung .ilaskan Natives was studied by Alberts and Lanier (54). C<)mparedto U.S. Whites. incidence rates were tow for lym- phoma and leukemia but not for MM. These authors noted, -Whereas the rate for all Alaskan Native males combined was >imiI;lrto U.S. and European rates, the rates for Indian males (run1 the southeastern Alaska tribes appear to be even higher than the high incidence reported for black Americans" (54). High rates for MM were not found among women. It is intriguing in this regard that dietary studies have shown that the average Inuit woman's diet contains less than one-third of the orpanochlorine contaminants present in the diets of males ( 5 5 ) . ,\dditionally. women may significantly lower their body burden of organochlorine compounds via breastfeeding (56). Thus. these data suggest that high dietary intake of dioxins may be ssociated with an increased risk of ,MM.at least among Alaskan Indian males. Swedish Fishermen. Svensson er al. (57) and Hagmar er al. ( 5 8 ) compared cancer incidence and mortality between eastcoast lie.. dioxin exposed) and westcoast cohorts of Swedish fishermen and the general Swedish population. Compared with the regional population and the westcoast fishermen. the eastcoaht fishermen showed "suggestive increases of mydomas (Standardized Incidence Ratio. 2.08: 9 5 8 CI. 0.76-4 jj):' Sloreover. mortality from Mh.1 was significantly increabed among the eastcoast fishermen as compared with the westcoast hhermen (Incidence Rate Ratio. 3.2: 95% CI. 12-8.7) and uith that of the general population iStandardized Mortality Rmo. 3.1: 95T CI. 1.2-6.4,. The authors offered no explanation tor this finding but noted a >imilur increased risk for blM among Swedish P m e r s ( 5 9 ) .A > u b q u e n t population-based study in Sweden found that exposure to two pesticides. phe- nouyacetic acid and DDT. was a bignificant risk factor for MM (601. Thus. these data support the hypothesis that exposure to diwins increases the risk of klM (61), Seteso. Italy. The dioxin hypothesis predicts that individuals c~po4edto high levels of dioxins through other forms of environmental pollution also would have an increased risk of MM. Numerous epidemiological studies have focused on persons w t h potential exposure to dioxins through manufacturing, during wartime. or following industrial accidents. However, many of these studies are subject to potentially severe misclassification bias caused by the use of indirect historical measurements of exposure (61-63). Follow-up on the population living near Seveso. Italy is relevant in this regard. In 1976, an accidental explosion in a chemical plant exposed the local population to the highest documented human exposure to TCDD. The incidence of cancer in Seveso was evaluated by Bertazzi er ul. (65) for the decade following the accident. Among males. the authors found a significantly increased risk for lymphoreticulosarcoma. a form of NHL (ICD9 Code 200: RR. 5.7: 95% CI. 1.7-19).and an elevated risk for MM t RR. 3.1: CI. 0.8-13.3. based on two cases). Women showed an increased risk for cancers of the hematopoietic system and I significantly increased risk for MM (RR. 5.3: CI. 1.1-22.6. based on two cases). Recently, a large occupational cohort study in Germany among male workers known to be exposed to TCDD also found a significantly increased risk of M M (SMR. 543: CI. 111-1587. based on three deaths: Ref. 66). Thus, these data provide additional support for the dioxin hypothesis. Biological Plausibility The evidence summarized above demonstrates that occupa- tional and environmental exposures to dioxins are associated with an increased risk of MM.We now address the question. "A.reDdiiooxxiinnss biologically plausible myelornagens'? have been associated with NHL (67-69).Because NHL and 4lM are both neoplasms of B cells, these malignan- cies may have risk factors in common. Moreover. extensive laboratory evidence indicates that dioxins are immunotoxic and that B cells are especially sensitive to their effects (70-77,). This evidence includes the following. Exposure of murine B cells to TCDD causes an increase in the activation of resting B cells and an increase in the secretion of IgM in the absence of antigen (73). In addition. TCDD inhibits the differentiation of murine B cells into antibody secreting cells in vitro (74)and in vivo (7% Similar findings have been observed in human B lymphocytes (76). Furthermore, TCDD exposure in adult mice causes a depression in B cell-mediated immunity that results in decreased host resistance to the non-lethal parasite. Plasmo- dium voelli (77). This observation resembles the increased risk of infection seen in patients with 4lIM caused by decreased levels of normal immunoglobulins (78). In s u m m q . dioxins can cause many of the B-cell abnormalities that are characteristic of MM. These abnormalities often have been interpreted as evidence of "chronic antigenic stimulation." In contrasf. dioxin5 can produce the apparent effects of "chronic antigenic stimulation" and can do so without antigens. Discussion The goals of this paper xe to: ( ( 1 1 draw attention to an associAtion between clusters of 4lM and bodies of water: and ( h ) construct an etiological hypothe.sis for klkl that could explain this association. We hypothesize that these clusters reflect exposure to dioxins. especially via the consumption of dioxincontaminated fish and seafood. Other possible explanations for this association include chance and confounding. Because bodies of water are common. the association of M M with bodies of water may be coincidental. However. our estimate of the prevalence of proximity to bodies of water in "control" locations ( 3 2 of 50) suggests that the observed prevalence in the MM clusters (8 of 8)is unlikely to have occurred solely by chance. Moreover, in addition to the association between MM and proximity to bodies of water at the level of the cluster. the case-control study in South Cumbria found a significant association at the level of the individual (33). Although this association also may be due to chance, a hypothesis 54 -R. eview: Multiple .\lycloma and the Dioxin Hypothesis Dioxins contaminate water c Dioxins accumulate in fish and seafood High consumers of fish and seafood have elevated body levels of dioxins B cell dysregulation c Increased risk of multiple myeloma linking risk of MM to proximity to bodies of water offers a more consistent explanation of these findings. Two clusters were located in proximity to nuclear establishments. Because ionizing radiation is a risk factor for MM. it is possible that these clusters are a direct consequence of increased radiation near these installations. This hypothesis is plausible but was not supported by estimates of radiation made during these investigations. Moreover. to our knowledge. the other six sites are not located in the vicinity of similar installations. Thus. a hypothesis involving some exposure to water provides a more coherent explanation for the similarities among the clusters. There are many pollutants in water other than dioxins that could be potential myelomagens. For example. mercury is a common aquatic pollutant and is known to have immunosuppressive effects in animals (79).However. mercury contamination was not a feature shared by the MM clusters we reviewed. Moreover. unlike blood levels of dioxins. measurements of blood mercury levels did not differ between the cohorts of Swedish fishermen with significantly different risks for MM ( 5 0 ) .Thus. although there may be other possible causes of MM that are associated with water. dioxins appear to provide the conceptual "best fit" to these observations. The dioxin hypothesis is summarized in Fig. I . The dioxin hypothesis is consistent with many other observations concerning MM. For example. Friedman and Hemnton (80)reported an unanticipated association between obesity and the development of MM. This finding. they noted. could result from "the increased storage and subsequent mobilization of fat-soluble pesticides. pesticide metabolites. or other potential carcinogens in the adipose tissue of obese persons" (80). The dioxin hypothesis also makes predictions about place and time. Thus. dioxin levels are known to be elevated in fish and shellfish collected at sites near discharges from pulplpaper manufacturing ( 5 7 ) .I t is intriguing in this regard that ecological studies have shown elevated mortality rates from MM in these areas ( 8I ). In addition. the rapid increase in MM incidence and mortality in industrial societies since 1950. especially in Japan 182).is consistent with temporal trends in the manufacture of chlorinated organic chemicals. Finally. the dioxin hypothesis suggests new opponunities for analytic htudies of the moni~lonalpammopathies. For ex- ample. cases with MM should report greater consumption of fish from contaminated waters than non-MM controls. Similarly. the incidence of MM should be higher in individuals with high consumption of dioxin-contaminated fish. Because dimins have been shown Io inhibit the differentiation of immunoglobulin-producing cells ( 7 3). other monoclonal gammopathics also may be more common among dioxin-exposed individual\. For example. MGUS denotes the presence of increased le\ cis of iminunoglohulins in abymptomatic persons and is the mo\t common monoclonal gammopathy I 831. Because the incidence of MGUS is greater than the incidence of .MM. the dioxin hypothesi5 may be more easily tested among person5 u i t h MGL'S. Becauw dioxinz are .;table and have long half-lives. [lie dioxin hypothesis is amenable to molecular epidemiological study. For example. Hardell and colleagues compared PCDD and PCDF concentrations in adipohe tissue from 7 patients u ilh "malignant lymphoproliferative diseases" ( 6 NHL and I pia- macytoma) to 12 surgical patients without malignancy tS11. The concentrations of dioxins in the patients with lymphoprolitrrative malignancy were xignificantly higher. Because patients with lymphoproliferative disease were selected to panicipate because of their greater dioxin exposure. this was not a case-control study. However. this study illustrates that molscular epidemiological studies of MM are feasible and sugge>ts that the observed differences may be etiologically important. Acknowledgments We thank Richard Rodnguez Tor in\aluablc lib- ~ s i s u ~ a~nd~ cDrcs Lord E Fleming. Stephen Kntchevsk) and Jim Schlesvlman for thoughtful discuwons on the manuscnpi References I . Gautier. M.. and Cohen. H. J. Multiple myeloma in che elderly. J. Am. Genalr Scc.. 42: 653-hW. 1994. 2. Reidel. D. A , . and Pottern. L. M. The epidemiolog) of multiple m)eloma. HematoL-Oncol. Clin. N. Am.. 0: 425-247. 1992. 3 Hemnton. L J.. Weiss. N. S.. and Olshan. A. F Epidemiology of myeloina. hi. J. S Ma1pa.s. D E. Bergsagel. and R. A. Kyle led%]M. yeloma: Biolog! Management. pp. 117-168. Oxiord: Oxford Untvenit! Rcss. 1995. 4 Penn): R.. and Hughes. S Repeated stimulation o i h e rcttculoendothcdi~ system and the de>elopment o i plasma-cell dyscrasiu. Lancet. I : 77-18. ;Io' 5. Antigenic sttmulation and myeloma. L n c e t . I - 3 2 . 1978. 6. Fnrs. G F .A revieu o i the \ignilicance of animal i d products as potcntrd pathuays of human exposure to dioxins. J. .Animal Sci. -3: 1639-1650. luui 7 . Diclson. L. C.. and L'zth. S. C. Heallh nsks of "dioxinr" a re\icu environmenwl and toxicological considemtioris. Vet Hum. Toxicol.. 35- hh -': I9Y.7 -:'8. Huff. 1.. Lucier. G.. and Tntscher. A Cminogentcti) o i T C D D : expcnmc.lltd. mechanistic. and epidemiologic e\ idence. Annu. Re\ P h m a c o l . Toxicol . 343-372. 1991 9 Poland. A.. and Knutson. J. C. 2.3.i.S-Teuachlomfibnz~~-dioxinand Rlated halogenated aromatic hydrocarbons: exmination of the mechaniznl ot toxicit). Annu. Re\ P h m a c o l . Toxicoi.. 22: 517-5.54. 1982. IO. Chastain. J E.. Jr. and Pazdernik. T L. 2 . 3 . 7 . 8 - T e t n c h l o r o d i ~ n ~ ~ ~ - d i ~ ~ \ ~ " ITCDDbinduced immunotoxicit?. h i . J. Immunophannacol.. 7: 849-8.56. Ius' I I . IARC M o n o p p h s on the Ekaluation of Carcinogenic Risks to HuiWn'. Supplement 7. pp. 150-3fiJ. Lynn. France- IARC. 1987 12. Siensson B-G.. Nilsson. A.. Hansron. M.. Rapy. C.. ,&sson. B.. Jnd SLertvtng. S.Exposure to dioxins and diknzofurans through the consumpti*'" . ''fish. N. Engl. J Med.. 324. S-12. 1991 I3 Fiore. B.J.. Anderson. H. A.. Hanrahan. L. P Olwn. L. I.. and S o n m p l . C. Span, fish consumption and hod) hurden levcls oi chlonnated hydnrdh"" a \tud) of Wisconsin mglen. Arch En\irnn.. Health. U 82-88. 1989. '14. Schwartz. P M.. J x o h w n . S.W . Fcin. G.. Jacohwn. J. L.. and h C e . H 'Lake Michigan fish conrumptiim ~s a wurce ol po1)chlonnated hiphcn!!' human cord w u i n . matcrnill w u m . dnd milk. .Am. J Puhlic Hcalth. 73: 2Q'-:w. IYX3. Cancer Epiderniologv. Biomarkcrs & Prevention 55 . a ,,I,cnburg. D . md Greenberg. M. Solving the h a t e r pu.~4e:clue\ to t!llL)u ,Ili~p l r i ~ l lto~ Jkoid. SUI. Med.. I :1763-177Y. IY93. K J n ~\I~N.III) T&y'\ World Chicago: Rand McNally. IYY?. I _ TJliJ5k..,\.Ion. T . Tatwkaua. R.. and Miyarakt. N. Glohal polluiion o l , l l J l ~ ~ m ahl)r PCB.i. DDT,. Jnd HCHs iBHC\i. Chcmo\phcrc. I:1269- IW K:id, R A.. md Cnepp. P. R. Multiple myeloma: houses and y ~ ~ u r c \ j ~ c L . r,phlia.~?/: 73.5-739. 1983. K~~,,, 1. Tajima. K.. Himse. K.. Nakagawa. N.. Kuroisht. T.. and Tominaga. s. Je.Lnpt~\e epidemiologic m d v o i hematopoietic neoplasms in J a p . Jpn. I Clln 1)nioI.. 13. 437-.W. 1985. ." jlr,,n5l. S.. Spitz. \l.R.. Cech. I. M...Anncgen. J. F . and Downs. T. D. i,ie,.Icuiemia md multiple myeloma in a mining county in Nonheasi Texas. \led w): 55-59, IY9.I. .pJnE. I.. O'Beim. D P md Gnmes. H. Multiple myeloma in the West o i .,;Kland. ;Irc.entation md frequency I . Ir. .Wed. .As.soc.. 17: 23-26. IY78. iL'.R. .A,. Herbcr. L.. Evatt. B L.. Heath. C. W.. Jr .Muliiple myeloma: a _-. iommunlt! cluster. 1. .Am. Mcd. ;\ssoc.. 213; 1334-1341. 1970. ,q Dauson. .A. .A,. md Ogston. D High incidence of myelomatosis in Nonh- SLotland.Scot. bled. J.. 18: 75-77. 1473. -J H,,iJcn. .A V . Jnd M-den. K. Oryanochlonnc pe3ticides in seals and ,,,,rp,,.c, Saiurc 1Lond.i. 2/15: 1274-1276. 1967. --.G ~ ~ e ~ l Jl. !$.1.. O.Toole. G.. McCmn. H.. Gnmes. H.. O'Connor. B.. Solan. G,, W4i.h. V.. m d Epan. E. Monoclonal gammopaihiea in the Wesi o i Ireland: a pllniin;u\ >tudy. lr. 1. \led. Sci.. 154: '44-1 101. 1985 :a. uli\cght. S. P A.. Cogan. P. h..McAulay. 1. R.. Jnd Mullins. E. Natural bJcl;ground r;rdiation md cancer mortality in the Republic of Ireland. hi. J. Epldemiol.. I:414-418. 1'393. 1' \h.o,n. C. F . and O'Sullivan. W M.Organochlorine pesticide res~duesand p c ~ ,tn otter+I Lirrru lurrui from Ireland. Bull. Environ. Conim. Toxicol.. 48: ;g:-.x!.:. 1992. 28. En&. 1.1 Multiple myeloma: a cluster in Virginia? Virginia .%led.. IU6: 115-ll7. I979 29 Ep,ictn. S. S. Kepone-huud evaluation. Sct. Tocll Environ.. Y: 1-15?. 19:s. 30. Retch. M. R.. m d Spong. 1. K. Kepone: a chemical disaster in Hopcwell. VirginiJ. Int. 1. Health Sew. 13: 227-246. IY8.3. 31. Cmnon. S. B.. V e u e y . J. M..Jr. Jackson. R. S.. Curse. V. W.. Hayes. C.. Smuh. W E.. Lmdngw. P. J. and Liddle. I.A. Epidemic Kepone poisoning in ihemiial uorken. Am. I.Epidemiol.. 107: 529-537. 1978. 32. Gurelian. P S. The clinical toxicology of chlordccone ;LC an example of toxiiolopcal nsk ucssmcnt lor man. Toxicol. Lett.. M6S: 589-596. IW2. 33. h w p . E. G.. ;md Honley. S. D. Multiple myeloma in South Cumbna I 9 - L - W problem, of health analyhis in small ;ommuniiies. 1. Epidemiol. Cmmunii) Health. .W 131-236. 1'485. 3 i.cr. J R. Polluii,)n-a*~rriated utennc Ic\ionb tn grey wals irom L i \ e r p u l B.I\ xi ot the In\h *ea Vet Rcc . 125. .:I):, IYXY . .1; 'I xiion. Y S u a . Y , Inianl. V Wc,rkm;m. R 1.. and Innqami. T Compu- i w n ,,I .mum phwpholipid iatty acid5 among h h i n p and hrming Japane>e p>pui.,iiiinr ~ n .dAmencan inlanden. J. Nutr Sci. Vitaminol.. 31: 417-42:. IY8.S h t k m k . R I- \Iulliplc \cIcro\i\ J conzlation 0 1 11, incidence with dieiar). iat \ E 1 \Id Sci.. 2 3 ) 41I-4?0. 19.50 .:- hddhl. D W . Buttcruonh. B C . McBride. A. Kroner. S.. md Bahnick. D C,m..:iiination < ) I I b h h) 2.'.:.~-TL'tnchl,iri*ihcn/l,~,,-di,)xin: .I Xurvc) 01 li\h itdior uatcr\hd\ In the L'nrl~xSl late* Chciiii~\phr~rJeX. I W7-21)14. I q X Y .:\ '..cnning. K J.. Ham\. %.AI... Pauwnhach. D 1 . md Bedhu?. H. Potential \c'ur..'\ o i l I. 1 . Y . ' J - I e r r a c h h ) r i ~ i h c n ~ i ) ~ ~ J ~ind it~h~exaiqnualie en\ ironmeni. Ecotn\,.Gd. En\iron. S a l . :.t. 133-141, I W ~ . . .14 K t i d ~ \. I..O l i ~ g a ~ iKl .. Miyamoio. Y Goto. T md Saito. S . Ccogrlphic du.:mng 1 ~ 1myeloma in ToLu\hims. Int. 1. Hcmatol.. 3J: 40s-4lN. I W I 11' \ l i > d a . H.. TaLa!ma. K.. Ogaki. J. Ka\hinwio. T . and Fukushima. S P~~i-..hiilonnatcdihcnm-p-diuxins in hluc m u z d irom milrinc coastal hater in J+:1 Bull Emiron. C i ~ n u mT. oxicol.. 39: X77-S8?. 1487. 11 \Ii!~ta. H.. Tak+am;L K.. \limur;l. M . Kahhiinoto. T.. and Fuku5hima. S. spci:lic iongcner profile5 of poljchlorinafed dibcnro-,>-dioxins and dibcnrc! u r m in blue m u r d in Owka Bay in Japan: queous dubilitles of K D D S and K D R i Bull. Emiron Contam. Toxicol.. 13: 342-i4Y. 148Y. ' 2 &atmabe. 1.. Ykushiji. T.. Kuwabm. K.. Yoshida. S.. .Marda. K.. Kahnoto. T..Koyma. K.. and Kunitn. N. Survcillwce of the dailv PCB intake from dirt of 1apane.w women from 1972 through 1976. Arch. Environ. Contamin. T*~\icoI..(1. 67-75. 1979. 1; Fernald. E. A. .Aih5 of Ronda. Tallahaswe. R. Ronda State Univenii) Foundation. Inc.. 1981 U Wan?. T C.. Kriban. J. P . and Johnson. R. S. Rr*idues of polychlonnated hiphcnyls nnd D D T in uaier and wdinicni ofthe SI. Lucic Estuary. Flonda. I977 Pc\tic. Monitonng J.. / 3 : 136'4-1371. IY7Y .45 blattnw. H. C Jr. Occurrence o i chlonnated hydrocarbon in%cttcrder. Southern Flonda-1168-72. Pewc. Slonitonng I.9.: 106-1 14. 1975. l6. Watanabc. I.. Yakuahiji. T.. KuuabilFa. K.. Yo\hida. S.. Made& K.. Ka\himoto. T.. Ko)ama. K.. m d Kunita. N. Sunelllance of the daily pCB intake from diet of Japanese women trom 1472 through 1976. .Arch. Environ. Conram. To<ICOI.. ,Y: b7-75. 1979. 47. Hovinga. \I.E.. Sowers. M..and Humphrey. H. E. 8 . Environmental expo- sure Jnd life,tyle predictors of lead. cadmium. pCB. md DOT levels in Grror U e i fish cater%.Arch. En\iron. Hrnlth. 48; '48-IUJ. IW?. 48. .Asplund. L.. Svennson. B-G.. Nilsron. .A,. Enkkson. U..Jannson. B.. Jenxn. S.. Wideqvist. U..and Skerfving. S. Polycholonnated biphenyls. I.I.I-mchloro2.2-hi~tp-chlorophenylleih~ipc.p-DDT) and I.I-dichloro-2.2-bislp-uhlorophe- nyl)ith>lene tp.p-DDEi in human p l w n a related 10 tihh consumpilon. Arch. Environ. Health. 4Y: 477-4815, 1Y94 49. Dewailly. E.. Ryan. I . J.. Ldibcnc. C.. Bnmrau. S.. Ebcr. 1. P. Gingns. S.. and C m e r . G. Exposure o i remote mantime populaiions tu coplanar PCBs. Environ. Health Perspect.. 1 0 ~SupplI):105-109. 199.4 50. Skenzon. B-G.. Nilsson. .A,. Junssun. E.. Schutz. .A. Akesson. E.. Hagman. L. Fish consumption Jnd esposure to persisient organochlonnc uompounds. mercury. denium. and methylamine\ m o n g Swedish lishermen. Scand. J Work Environ. Health. 21: 96-10.5. I Y V 5 I. Feikenr. E. I.. md Kromhour. D. Epidemiologic mxiics on Eskimo5 and lish tnmkr irevieui. Ann. N Y .%cad.Sei.. 083: 9-15, lYY3 5 2 . Nobman. E. D . Eyers. T . Lamer. .A. P. Hankin. J. H.. and Jackson. .M. Y. The die1 of .Alaskan native adults. Am. J. Clin. Nulr.. 55. 1024-1032. 1992. 53. Dewailly. E.. Ayottc. P . Brunrau. S.. LdibeRe. C.. Muir. D. C.. and Nonuom. R. 1. Inuit exposure IO organochlonnes throueh the aquatic food chain in Arctic Quebec. Envtron H r i l t h Perspect.. 101.- 618-620. lY93. 54. A l k n s . S. R.. and h n i e r . .A. P.Leukemia. lymphoma. and multiple my. eloma in Alaskan Natives. J. Nail. Cmcer Insi.. 78: 831-837. 1987. 55. Kuhnlein. H. V Nutntion of the Inuit: 3 bncf OVCNICW. Arch. .Mcd. Res.. ISuppl. 11: 728-730. 1991. 36. Dcwailly. E.. Ayotte. P.. Brunneau. S.. Lalibme. C.. Muir. D.C. G. and Norstrom. R. I.Inuit cxposure to organochlonnes through rhe 3quaiic food cham in .Arctic Quebec. Environ. Hcallh Penpeci. 101: 618-610. I W 3 . 57 Svensson. B-G.. Miloczy. Z.. Sudmbrrg. U..and Hogmu. L. Monality and uwcer incidence among Swedibh fishermen with a high diewry intake of pcrsistent organa-hlonne compounds. S c a d . I . Work Environ. Health. ? I : 106-1 IS. 1995. 38. Hagmar. L.. Lindin. K.. Nilsson. .A,. Noming. B.. ikebson. B.. Schutz. 4.. and Moller. T Cancer incidence and monality among Swedish Bal~tcSen fishermen. Scand. J. Work Envtron. Heallh. IX: 217-224. 1'392. 3s. Stcincck. G.. and Wiklund. K. $lultiplc m)eloma in Swcdirh agricultural workers. h i . J. Epidcniiol.. 15 ?21-:1?. 19x6. .IMI EriL\u)n. M..And Karl\,am. \l Occupational and lrther en\ ironmenial iacioru m d multiple rn)elnina: p ~ p u l a i i mh a d caw-conirol m d ) Br J. Ind. Med.. 4Y. Y5-10?. lYY2. hl EriLwm. M.. H.irdell. L.. .W:ilLcr. H . md Wciner. J Malignam lymphopro- Iiierati\e direax\ in ~rrcupati~~un.irth p w n u a l expmurc t u phenc~x>scctic~ c i d \ .Ior dit)xi,i\. rc.gi\t?-hared .rud> ,Am. J Ind. \led.. 22; W.5-.ll2. I W 2 . 02 Fingcrhut. M 4..S w c n e ? . \I.H.. Halpcnn. W b.. Jnd Schnon. T M The cpidciiiiolog? yx)pulation\ c x p ~ tdu dimin. 111 IARC' SLicntilii Puhlicaiion\. 'in11 I O X . pp I 1 - W L > m .France I..WC'. I ' W I h l SLenc. 5 .A,. Dcuhur\t. I C . and Greenher:. M. PiiIychlt~nnaiedd i k n m - p-dioxinh Jnd p)l)chlorinaicd dihentoiuran\ the nrk3 IO human health. A review Hum. Tc)xiiol.. 177-103. I%+) tA Lnrwn. J C Lt\el\ oipdlu!ant\ m d ihcir mctsh)litc\. r.xpowre\ 11) orpnic \uh\iance\ Toxicolog). / I . 11-27, I W S . b.5 BcnJ/ll. .A,. Pc~atori..A. C.. C8)nwnni. D . Tironi. .A . Lmdi. $1. T.. md zm.c.hctli. C Cnnccr incidence in J p,pulaiion ~cc.idcntall>csporcd tu 2.3.7.Y- t r t n c h l o r i * l ~ n ~ ~ ~ ~ ~ ~ r r ~ rE-pdiidoexiiiino.logj. 4 . !YX -106. 1 Y Y i &. Bruhcr. H.. Fle>ch-Jan)\. D. KJuppincn. T . Kogcrinar. .M.Sleindod K.. Slanz. A,. md Wahrendort. J. CJncer monalip in Cennan male workers exposed to phenoxy herbicide\ and dioxinr. Cancer CJUW\Control. 7. 311-321. 19%. 67. Fingerhut, M.A. Halpmn. W. E.. Mulow. D. A. Piacitelli. L. A,. Honchar. P. A.. Sweeney. M.H.. Greife. ,A. L.. Dill. P. A,. Stemland. K.. and Suruda. A. 1. Cancer monality in workers expowd to 2.3.7.8-tru;lchlorocibenzo-p-Jioxin. N. Engl. J, bled.. J24: 212-218. 1 9 3 1 , 68. Coggon. D . Pannett. B.. and Winter. P %lonalityand incidence of cancer at four factories making phenoxy herbicides. Br J lnd. Med.. 48: 173-178. 1991. 56 Review: Multiple Myeloma a d the Dioxin Hypothesir -V d 6. 57-0 69. Woods. 1. S.. Polissar. L..Scucrson. R. K.. Heuwr. L.S.. and Kulander. B. 0.Soft tissue sarcoma and non-Hodgkin's lymphoma in relation to phcnoxyher- bicidc and chlorinate phenol exposure in Western Washington. 1. Natl. Cancer Inst.. 3: 899-910. 1987. 70. Luster. M. I.. Tucker. A. N..Hong. L.. Boorman. G.A.. and Palteaon. R. In w w and in rirm effects of TCDD on sem cell and B cell differentiation. In: A. Poland and R. Kimhrough i c d s . ~ .Banhur) R e p n 18. Biological Mechanism of Dioxin Action. pp. 41 1-419. Cold Spnng Harbor. NY: Cold Spnng Harbor Lhoratory. 1984. 71. K m s . J. G..and Holrapplc. .M. P. .Mechanisms of 2.3.7.8-tetrachloro- dihmro-p-dioxin (TCDDI-induced disruplion of B-lymphocyte signaling in the mouse: a current perspective. Exp. Clin. Immunogenet. / I : 110-1 18. 1994. 72. Holwpplc. M. P..%>der. N. K..W d .S.C.. and Morns. D. L.A review o i H-tetrachlornliknrop~ioxin-induced changes in immunocompetence: 1991 update. Toxicology. 69: 219-255. 1991. 73. Morns. D. L.. and Holsapple. M. P. Effects of 2.3.7.8-tctnchlorodibenzo-pdtoxin ITCDD, on humoral immunity II. B cell activation. Immunophmacolopy. ' I ; 171-182. 1991. 74. Luster. M. I. Gcnnolec. D. R. Clark. G..Wicgand. G..and Rosenthal. G.J. Selective effects of 2..1.7.8-te~rachlomdibenzo-p-dioxiannd corticosteroid on in w r o lymphocyte maturation. 1. Immunol.. /JO:928-935. 1988. 75 Dnoley. R. K . and Holsapplc. M. P. Elucidation of cellular targets rcspon- sihlc for tcmchlorndibenm-p-dioxm t TCDDI-induced suppression of anubody responses. 1. The role of the B lymphocyle. Immunopharmacology. 16: 167-180. 1988. 76. Takenaka. H.. Zhanp. D.. Diaz-Sanchez. D.. Tsicn. A.. and Saxon. A. t n - hanced human IgE pmfuction resulr~from exposure to Ihc yomatic hyJrocuhwi\ from diesel exhaust: direct effects on B-cell lgE pmluction. J . .4llerpy Clin Immunol.. 95: 103-1IS. 199s. 77. Tucker. A. N.. Vore. S. J.. and Luster. M.I. Suppresvon of B sell dilfcrm. tiation hy 2.3.7.8-tctnchlordibnrop-dioxin.Mol. f'hmacol.. 29. 372-3". 1986. 78. Husxin. M. Muluple myeloma: an o~en'ic'wof diagnosls and managcniwr. Clcvel. Clin. J. .Med.. 01: 285-298. 1994. 79. Svensson. E-G.. Schutr A.. Nilsson. A.. Akesson. I.. ALesson. B . d i d SLerfving. S. Fish ss a bource oi exposure lo mercup and \elenium. Scicii.d IWashington. DCI. 126: 61-74. 1990. HO. Friedman. G.D.. and Hemnton. L. J. 0 k s i l ) i and multiple myelcrma. Cm.r Causes Conuol. 5: 179-4113. 19Y-J 81. Blamer. W. A.. Blair. A.. and Mason. T. 1. Multiple myeloma in the l'niid States. Cancer (Phila.). J8: 1547-2554. 1981. 82. Cuzick. 1.. Velez. R.. Doll. R. Internationalvariations and temporal trend. in monality from multiple myeloma. Int. 1. Cancer. J2; 13-19. 19x3. 83. Kyle. R. A. Current concepts on monoclonal gammopathics. hust \Z 1. Wed.. 12: 291-302. 1992. 84. Hadell. L.. Frcdrikson. M.. Enksson. M.. Hansson. M.. and Rappc. C Adipose tissuc conccnuauons of dioxins and dibenzofunns in patients n i t h malignant lymphoproliferauvc diseases and in patients wlhoul a malignant dicase. Eur. J. Cancer Rcven. 4: 23-229. 1995. - Dolores R Elena Cai and Jose Depanment Connecticut P S I h O l ~ >[R Epldemolog: A. 0.1. and t d'ODoologIa F.X B.] Abstract geogr ha -Ik diver, kLetogent W set! mtior hnor gen -F Kkar gen .ad epide! W g . Mu dctstedb PdYmOrpk mion observed c tumors ~i dierent c density mc heenn Our s -nsa -ted of car mu, * in mutate -Pin: