Document wgDoKXX8XDMBB4yyY2V7QnebE

Yr cue n t * /7 7 ; (7 1 -7 3 , A * ..insider impio. .nvestigate a at, re ignored ir r time difTeru,. here seems to l. r arguments ah . indicate the sat-, trip, for we fir J. C. Hafeip 030 flURD E. K caH i, m . md Notes F Keating. .Science 17? Hall. D . B. P c r a v v'Ui. A G Mun;2 1 .. refer herein to the ] AT(A.I M E A N USV> SOI Mockler, J . M Ricinn! 7). 114 (1965) 1 N alt. H ewlett-P ackard J ') J Phys. 40. 61 (19?: ,i of the T h ir d Aunt, .h m ? I n t e n a l Strati n, - to 18 N o v e m b e r ly* Technical Officer. 1 s ishington, D C 2(>39i ' rate-change" is me.: aKt ot tio ss corren ~ns identification of il ' ^ te , a.id n o t (hat tin nMcally co rrela te d c change in the tip' hy an a p p a re n t co m ! i m relative ra te s for i i uncorrclated period. ` long a g o to sug^>' <h's ro ta tio n a l motion l Set. A r ts 8, 161 (18'J *'>ron In s t N e th . 4. 21 P<ctfdmgs o j the I.hii on F req u e n cy Control ` Library of Congre" '9). p. 316; ---------- j r j ` l">gt 5th A n n u a l Su>,- Control (D oc V Technical Clearinghouse ''* P 168. son d a ta show ed th>' *:tc three tim es, 4l'8 changed twice, a* J " once d u rin g the ci-*cstard tnpT, clock 13 changed fo u r tinu^ m ra te w ere fon 1 <4> during th e wcsiwj > 'ft o f th e T h ird <' Iuue I n te n a l Strati- *' 18 N o v e m b e r I'* 1 Technical Officer, L s ' 'hington, D .C 2010' c of N a v al Rc*-ia v for evaluating > " or o f ato m ic cl1*-- 1 Wc appreciate > *!cr. R . G H a!!, ar * th a n k R St'-1'' r for helping with ,hank th e O S N " ne the ato m ic cl-' 1 ' "'ut-P ackard Co ' ` 'tk s, p artic u la rK 1 CK fo r th e ir coop*-1' , . b D epartm ent. Ca 1 Illin o is 61602 * June 1972 SCIENCE. V O t 1` ! Asbestos Contamination of Parenteral Drugs Abstract. Chrysolite asbestos has been found in approximately one third of ;/,v samples from two sets of 7 widely used parenteral drugs. Appropriate filtration procedures for Another reason for chrysotile's effec .yrilization and removal of foreign mat tiveness as a filter is the possibility of its Kr from parenteral solutions became fragmentation in solution. The fibers are necessary several decades ago ( / ) . The not unit structures, but rather bundles pharmaceutical industry found asbestos of ultramicioscopic fibrils, each 200 to fillers useful and effective for this pur 400 A in diameter and 1000 A to pose, and they have become widely used. several micrometers ip length (3). Un Unfortunately, we have found that treated. the fibcis in a filter can fragment such filters add a contamination of their into a much larger number of fibrils, own to the filtrate- and that, at least in the number of efiective filtering elements the past 3 years, parenteral solutions being vastly increased and their surface used for intravenous, intramuscular, and area concomitantly extended. infraperitoneal therapy in the United Serious health hazards have been Slates have often contained measurable identified with asbestos exposure (4); amounts of asbestos. these include pulmonary fibrosis and a The physicochemical properties of variety of neoplasms (such as broncho asbestos make it highly suitable for fil- genic carcinoma, pleural and peritoneal 'ration, and it is therefore widely used mesothelioma, and perhaps other neo in chemical, food, drug, and other in plasms). On?e thought to be limited to dustries. The individual fibers and fibrils heavily exposed asbestos workers, these are both fine, and strong, ranging in neoplasms are now known to occur diameter from 200 A to tenths of with much less exposure, such as that micrometers and in tensile slrergth experienced by those living in the house from 20,000 to 60,000 kg/cm 2. hold of an asbestos worker (5 ). working Hydrated magnesium silicate minpr. ,n bip.^rU, is uscu -Is, the;, tend be eiieiiiic.tiiy inert m some aieas (6 ), or merely living and resist attack by most chemicals (2). within S00 m of an asbestos plant (5, 7) Theie are several varieties of asbestos or in the vicinity of asbestos operations (chrysotile, amosite, crocidolitc, antho- () phyllite, tremolite), and special proper Such human disease has been the ties of each can be utilized for specific result of inhalation of asbestos fibers. It filtration purposes. Thus, crocidolitc has been used for gas mask and cigarette . filters. Chrysotile is widely used in food processing industries (sugar, lard, beer, and so forth). For pharmaceutical use. chrysotile again has been the asbestos variety of choice. These filters are used at many points m the manufacturing process, from the preparation of raw materials, through 'he intermediate steps, to the final filtra b\ n.V.'.ilV-'. fi tion of the product. The latter is im portant: in order to minimize or prev'ent bacterial growth (and the presence of bacterial toxins), it is often desirable to complete production procedures within .! 8 hours or less. Speed of final filtration may be critical in such production schedules. Chrysotile allows rapid filtra tion without reduction of efficacy. f'C 1. (a) Chrysotile asbestos observed by rolarized light microscopy from a singlc''`l sample o f tetracycline. This particular fiber was 1 mm in total length imagmlica,!on x 1300). (b and c ) Chrysotile obceived by electron microscopy in ampicilbn and vancomycin, respectively (magnifi cation X 25,000). 14 JULY 1972 is not now known whether i. occurs as the iesult of other routes of administra tion, although ingestion may be suspect ir. view of some evidence of increased incidence of gastrointestinal cancer among asbestos workers ( 9). In the ex perimental animal, it has been demon strated that parenteral administration (typically, of 10 to 30 mg) of all vari eties of asbestos can produce neoplasms (10). Such neoplasms have appeared at the site of injection of fibers. Other studies have also shown that adminis tered fibers are readily disseminated from the injection site, both hematog enously and along draining lymphatic channels ( I I ) . An investigation of a number of par enteral drugs has been undertaken. Sin gle-dose vials of widely used drugs for parenteral injection, taken from the pharmacy stock of Mount Sinai Hospi tal during 1969 and again about 1 year later, were examined for their asbestos content. Analysis for chrysotile as bestos was made by using optical and electron microscopy. Samples for analysis were obtained by filtering the contents of each vial uiiousn ivmlmnre a a membrane filters (diameter 13 mm, pore size 0.8 71m). Drugs received from manufacturers as powders were prepared with sterile water, which in turn had been filtered through membrane filters (pore size 0 8 (im); drugs received as solutions r; r were filtered with no additional dilu tion. While the effective pore size of the membrane filters used is larger than the diameter of many of the asbestos fibers of interest, it was found that the surfacecharge properties of the filters and the asbestos fibers, as well as the circuitous path through a filter, allow virtually complete collection of all asbestos ma terial. At the time of preparation of the drug samples, control samples were prepared with filtered sterile water. One such control was processed with each of four drug samples. AH were prepared in a laminar-flow, filtered air hood, and strict clean-room procedures were fol lowed as normal room air may provide a ready source of contamination. Samples were prepared for optical microscopic examination by mounting one quarter of the membrane filter with collected material on a microscope slide. The filter was cleared for viewing with of polarized light microscopy at a mag nification of x 400. All birfringent fibers longer than 5 iin and with a length-to-width ratio greater than 5 were counted. Whenever possible, identification of chrysotile asbestos was made on the basis o f ex tinction angle and morphology. This identification could be made with rea sonable assurance only for those fibers longer than about 100 /im. For identification of asbestos by elec tron microscopy, one half of the filter and collected material was ashed in a low-temperature activated oxygen asher (12). The residue was dispersed by grinding for 2 to 5 minutes in a 1 per cent solution of nitrocellulose dissolved in amyl acetate, and allowed to dry. A known fraction of the dried film was transferred to a 200-mesh, Formvarcoated electron microscope grid. The prepared grids were then scanned at a magnification of X 42,000, and the tially ashed membrane filter along witfi not recorded collected residue was mounted directlj on the electron microscope grid. In all cases, the identification of chrysotile .hows a portion ..her found by To confirm tl asbestos was made on the basis of in 0 of the 16 sam unique morphology (3). Other tubular j duplicate of minerals, such as halloysite, exist a ..imples were sc nature, but their presence in these .rescopy. In five samples can be ruled out on the basis dialing all tho of size, geologic rarity, and absence fibers were ident from manufacturing processes for par ir.itions of asbes enteral drugs. m background During 1969, 16 samples were Evamplcs of as scanned optically, and all fibers longer imples by mean than 5 /tm were recorded. Fibers longer iron microscopy than 00 /m that could be identified b and c. with reasonable certainty as asbestos During fall 197 were noted. The results of this analysis second set of dru are given in Table 1 (13). Many fibers , 'gated to ascertt and particles, in addition to asbestos, 1 might still be fc were present in the filtered samples. 1 enteral drug prep, They included,cellulose fibers, starch end study, only u granules, and inorganic and organic isis was undeitakei fibers of unidentified composition filtered drugs. Th a 10 percent solution of membrane fil ter material in a 1 : 1 mixture of di ethyl oxalate and dimethyl phthalate. The cleared filter was scanned by means number of asbestos fibrils and their size were determined in two to four grid squares (100 by 100 /ami. In some selected cases, the residue from a par Moreover, in optical scanning, numer- ! Table 1 indicate tb ous fibers were seen that were consisteni I parenteral drugs by with asbestos, but identification could j common occurrence not be made reliably and they were < of asbestos in 6 of o.mtiy exceeded ba> The amounts of a Table 1. Fibrous particulates seen in samples of parenteral drugs obtained during two periods of time. Observations were uj*uc by optical mi 1 . .. ,T ijiese samp croscope (O) or electron mlc-c^ccpc (E ). The opucal v.c.c nauc on one quarter of each sample, the clectron-microscope deter \ -rogram and were mmauons on 10-' M1 ju.nple. /\ii drugs were labeled for intravenous use except those identified by IM for intramuscular or IP fo: < ihose reported for o mtraperitoneal. Amounts of samples are in grams unless specified otherwise. "Yes" after optical data indicates that some fibeis in the sam pie were identified as chrysotile by optical microscopy; while many fibers seen in all samples could have been asbestos, the majority were not j ! - rcumstances. Asbe All values of 27 x 10"*g or less for the estimated mass of asbestos could have resulted from the analysis background; N, number. ` measured in ambien Drug Amount (g) Fibers by O, 1969 Chrysotile fibrils by E Asbestos, estimated mass in total sample (!0-*g) 5-20 tin (<V) 20-100 // m (N ) >100 [itn (N) Asbestos 1969' (V) 1970-71 (tV) 1970-71 Sodium acetazolamide Sodium ampicillin Cephaloridine Sodium cephalothin Chloramphenicol Cortisone (IM ) Lincomycin HC1 Sodium methiciilin Neomycin sulfate* (IM ) Neomycin sulfate* (IP) Sodium oxacillin Streptomycin sulfate* (IM ) Streptomycin sulfate* (IM ) Sulfisoxazole diolamine Tetracycline HClf Tetracycline HCH Vancomycin HC1 1 1 2 2 2 0.25 1.2 2 1 5 2 I 2 4 1 0.5 1 52 35 10 13 38 28 44 it 62 24 91 51 28 6 101 57 24 35 15 0 15 8 24. 7 26 16 16 7 7 3 4 5 1 0 10 1 38 18 0 1 1 7 i >100 1100 Yes 32 2 76 85 210 8 6 7.7 8 14 2 43 4 10 6 22 82 2 17 Yes 79 i 39 >100 410 4 75 Yes 50 15 26 8 15 35 86 30 78 Potassium penicillin G Sodium penicillin G Water control 10* units 14 4 1 2 X 10*units 20 cm* 600 12 14 9 27 3 7 101 "D uplicate analyses of samples from different manufacturers of the same drug. f Duplicate analyses of samples from the same lot of a manufacturer. Average of seven samples ranging from 1 2 X 1 0 - to 2t X 10- g. -re typically in the r. per cubic meter. C. r.inograms or micro "'ay be evaluated w, hat 10- g of asbesti b' fibrils of a size mgs (400 A in dia; a length), it should bt -'e or indefinite resul mmple do not guara ef asbestos in the dru he sample was taken. a inadequate sample c n run. Moreover, th '-'present only a small -n the market. On the iding of asbestos it ;1c single vials that Wi investigation over ociicates significant asl stion of some parentei hate. W. c Sinai School of A "l University of New York 10020 172 SCIENCE, VOL I' | 1 JULY 1972 -Mite filter along with , not recorded as asbestos. Figure la . js mounted direct!) .hows a portion of a chrysotile asbestos ,-roscope grid. In all liber found by optical microscopy. icjtion of chrysotilc To confirm the presence of asbestos, jc on the basis of it, 1 b of the 16 samples taken in 1969 and .-v (J). Other tubular j a duplicate of one of the positive . halloysite, exist u camples were scanned by electron mi ,, r presence in these croscopy. In five of the seven cases, in ruled out on the basis cluding all those in which asbestos . rarity, and absence fibers were identified optically, concen ng processes for par trations of asbestos several times those m background controls were found. i 16 samples were Examples of asbestos found in two and all libers longer | omples by means of transmission elec- recorded. Fibers longer uon microscopy are shown in Fig. 1, st could be identified . b and c. certainty as asbestos j During fall 1970 and spring 1971, a results of this analysis . second set of drug samples was inves 'e 1 (Id). Many fibers | tigated to ascertain whether asbestos addition to asbestos, j might still be found in typical par i the filtered samples ' enteral drug preparations. In this sec ,1'llulose fibers, starch I ond study, only ultramicroscopic anal organic and organic , ysts was undertaken of the residue from .'.ntified composition, filtered drugs. The results shown in seal scanning, r.umer- Table 1 indicate that contamination of -.cn that were consistent parenteral drugs by asbestos was still a .1 identification could common occurrence. Here, the quality i.ably and they were of asbestos in 6 of 17 samples signifi cantly exceeded background levels. The amounts of asbestos estimated in . .-r. e'adc mi- i cmy. of thca- aompies exceeded a rni- !.J,,,ii-ijiicroscope deter- ciograin and were much higher than 'rlramuscular o r IP for those ieported for other environmental some fibers in the sam s. the m ajority were not I, M, num ber. circumstances. Asbestos concentrations measured in ambient air, for example, are typically in the range of nanograms Asbestos, estimated per cubic meter. Contamination with 4>!j mass in nanograms or micrograms of asbestos t)H total may be evaluated with the knowledge sample that 10 -9 g of asbestos might represent o o -* g) 10 fibrils of a size typically seen in 1970-71 (N) 1970-71 drugs (400 A in diameter by 1000 A in length). It should be noted that nega >100 2 85 6 1100 1.6 210 7.7 tive or indefinite results for a particular sample do not guarantee the absence of asbestos in the drug lot from which the sample was taken. A single vial is 8 * 14 an inadequate sample of a large produc 2 4.3 tion run. Moreover, the drugs sampled 4 22 10 82 | represent only a small fraction of those 2 1.7 1 on the market. On the other hand, the 1 3.9 finding of asbestos in one third of >100 4 410 the single vials that were examined in 7.5 this investigation over a 1-year period 8 15 , indicates significant asbestos contami 35 86 ' nation of some parenteral drugs at this time. 30 78 W. J. N icholson 12 14 9 27 7 lo t ! ** tame lot of a smglr , C . J. M aggiore I. J. Se l ik o ff Mount Sinai School of Medicine of the City University of New York, New York 10029 . science, vol. in 14 JU L Y 1972 References and Notes 1. See, for example, L. C. Miller, in Safety of Large Volume Parenteral Solutions, National Symposium Proceedings (Government Printing Office, Washington, D C., 1967), p. 6. 2. S. Spiel and 3 P. Leineweber, Environ. Res. 2, 166 (1969). 3. A. M. Langer, I. J. Selikoff, A. Sastre, Arch. Environ. Health 22, 348 (1971). 4. I. J. Selikoff, 3. Churg, E. C. Hammond, J. Amer. Med. Ass. 188, 22 (1964). 5. M. L. Newhouse and H. Thompson, Brit. J. Ind. Med. 22, 261 (1965). 6. P. G. Harries, Ann. Occup, Hyg. 11, 135 (1968); J. Stumphius, Brit. J. Ind. Med. 28, 59 (1971). 7. 3. Lieben and H. Pistawka, Arch. Environ. Health 14, 559 (1967). 8. J. C. Wagner, C. A. Sleggs, P. Marchand, Bril. J. Ind. Med. 17, 260 (1960). 9. I. J. Selikoff, E. C. Hammond, J. Churg, t. Amer. Med. Ass. 204, 106 (1968). 10. W, E. Smith, L. Miller, R. E. Elssser, B D. Hubert, Ann. N .Y . Acad. Sei. 132, 456 (1965); J. C. Wagner, Nature 196, 180 (1962). 11. F. J. C. Roe, R. L. Carter, M. A. Walters. 3. S. Harrington, Int. J. Cancer 2, 628 (1967); K. Kanazawa, M S. C. Birbeck, R L. Carter, F. J . C. Roe, B n t. J. Cancer U , 96 (1970); A. Holmes and A. Morgan. Harwell Report AERE-5289 (Atomic Energy Research Estab lishment, Harwell, Berkshire, England, 1967) 12. C. Berkley, 3. Churg. I. J. Selikoff, W. E Smith, Ann. N .Y . Acad Sei. 132, 48 (1965). 13. These data were presented at a seminar of the Food and Drug Administration on 7 October 1969 and at the annual meeting of the Parenteral Drug Association in New York City on 30 October 1969. 30 March 1972; revised 15 May 1972 Crustacean Color-Change Hormone: Amino Acid Sequence and Chemical Synthesis Abstract. The blanching hormone of the prawn, Pandalus borealis, is pGlu-LeuAsn-Phe-Ser-Pro-G ly-Trp-NH Its structure was settled by a combination of mass spectrometry and Edman-dansyl analysis of a thermolysin fragment. Confirmation of the structure war obtained by chemical synthesis from amino acids. This neurosecreted hormone is active in picogram amounts when tested in shrimps. Many crustaceans have the ability to change their body color in order to match their background These colui changes are brought about by hypoder mal chromatophores, that is, specialized cells containing movable pigment granules and having richly ramified cell processes. Crustacean chromatophores are controlled via neurosecreted hor mones, which are released from nerve endings in the sinus gland, a neuro hemal organ located in the eyestalks of most decapod crustaceans ( I ) . One color-change hormone, the blanching (red-pigment-concentrating) hormone, has been isolated frim eyestalks of the prawn, Pandalus borealis (2 ). This hormone, which is active in shrimps in picogram amounts (J ), is present in very minute amounts in the crustacean eyestalks (J) and, therefore, only about 90 fig of the pure blanching hormone has been available for our studies of its structure. The hormone was found to be a small peptide with a blocked NHj-terminus (2 ), and we now report the deduction of its com plete structure, which we have con firmed by chemical synthesis. Quantitative amino acid analysis (4) on 8 pg of the isolated hormone (after hydrolysis in 6M HC1 at 110C for 24 hours at reduced pressure) and de termination of its tryptophan con tent by ultraviolet spectroscopy (2) .gave the following composition: Asp123- Glu0.8sGly0.98Leu1 14Phe0 Pro, 01Ser i igTrpi oo ( 5), which account, fo>- op percent of the weight uf the hormone. The hormone is electrophoreticaiiy immobile at acid, neutral, and alkaline pH (6 ), in accordance with its blocked NlF-tcrminus, and proving that none of its carboxyl groups is free. Digestion of the hormone (23 pg) with thermolysin and fractionation of the digest on a column of Sephadex G-25 gave two major peptide fragments. Upon acid hydrolysis, one o f these gave aspartic acid, glutamic acid, and leucine in about equimolar proportions. Since no free NH 2-terniinal group was obtained by the DNS-C1 method (7), this peptide was an N H 2-tcrminal frag ment of the hormone. Acid hydrolysis of the other peptide yielded about equi molar amounts of glycine, phenylala nine, proline, and serine. In addition, it contained tryptophan, which was de termined by ultraviolet spectroscopy. It had a free NH 2-terminus, and analysis by the Edman-dansyl method, essenti ally as described by Gray and Smith (8), established its sequence as PheSer-Pro-Gly-Trp-NH2. The COOHterminus was identified by omitting the hydrolysis-after the DNS-CI treatment which followed a four-cycle Edman degradation. The DN S product so ob tained was identical to a reference DNStryptophan amide when compared in polyamide thin-layer chromatography m