Document dnM91ZkDmOd6pra4ya9VM5yrb

The Cosmetic, Toiletry and Fragrance Association, Inc. 1625 EVE STREET, N.W., WASHINGTON, D. C. 20006 202/393-2070 December 26, 1973 Jam es H. M e rritt P re s id e n t N o rm a n F. E strin, Ph.D. Vice President-- Science Hearing Clerk Food & Drug Administration Department of Health, Education and Welfare Room 6-86 5600 Fishers Lane Rockville, MD 20852 Re: Proposed Order Method for Asbestos in Talc Published in the Federal Register (38 F.R., No. 188, pgs. 27076-27081) September 28, 1973 Dear Sir: This letter and attachment is being submitted in response to the proposal published on September 28, 1973 (Federal Register, Vol. 38, No. 188, pages 27076-27031) in which a method is proposed for the determination of asbestos particles in food and drugs. The comments are being submitted on behalf of The Cosmetic, Toiletry and Fragrance Association, Inc., a national association representing manufacturers and packers of cosmetic products as well as the manufacturers and vendors of cosmetic product ingredients and supplies. Members of the Association manufacture approximately 85% of the cosmetic products distributed in the United States. While the Association is not directly affected by the Federal Register Proposal cited above, one of its scientific subcommittees, the CTFA Talc Subcommittee, has completed a critical review of the opti cal microscopic merhcd published in this proposal. In the belief that tiis review is rvlant for your consideration of this proposal, the report of the CTFA Talc Subcommittee, dated December 10, 1973. is attached as part of this submission. As a result of round robin testing of five different types of talc, by members of the CTFA Talc Subcommittee, the subcommittee concluded that the proposed method does not provide a reliable means for the detection of asbestos in talc and that it results in both (Over) -2- false-positive and false-negative findings. The procedure is also tedious and may consume as much time as one-half day per sample. The subcommittee has also reviewed several alternative methods to detect chrysotile and tremolite in talc including modifications of the optical microscopic method. As a result of its review the subcommittee feels confident that a feasible method can be developed, and recommends a collaborative effort between FDA and industry to develop such a method. The questions concerning asbestos detection and the relationship of asbestos to human health are indeed complex ones. At a confer ence last month, sponsored by the National Institute of Environ mental Health Sciences and the Environmental Protection Agency, three serious difficulties in determining the effects of asbestos on human health were presented,: 1. "The term asbestos itself is misleading, for it encom passes a whole group of minerals that are distinct from each other in chemical composition, particle size, and v probably biological effect. 2. Many of these particles are so small that detection is extremely difficult and time consuming, placing restraints on the scope of research that can be reasonably undertaken. 3. Perhaps the most serious difficulty in asbestos research, however, is the very long period between initial exposure and evidence of biological effects.M (!) Other points made at this Government sponsored meeting that merit consideration were: 1. Chrysotile fibers are too small to be detected by conven tional optical microscopy. 2. Amphibole asbestos fibers are larger. But they also require sublight microscopes for careful examination and cannot be positively identified without additional microchemical or selective area diffraction analysis. 3. A question yet to be answered is the effect of low-level exposure to asbestos on the general population. 4. Dr. Ian Webster of the National Institute for Occupational Diseases in Johannesburg, Republic of South Africa, reported at the conference that very little ingested asbestos penetrates the walls of the stomach and colon and that almost all that does is of the smaller chrysotile type. n r Chemical & Engineering H e w s , December 10, 1973, pages 18-19. -3- Similarly, studies with inhaled nsbonbon also show that the fiber is absorbed is predoninn tc*ly the ehrysorilo type. ', 5. Dr. William Eisenberg of the Food and Drug Jidmina ctrBtion was quoted as stating that "at present, analysis by polarized light microscopy is used to determine purity of food and drugs, even though this method probably misses many of the smaller and more easily absorbed asbestos fibers that may be present in these materials." In light of the perplexing questions raised at this conference, the Association considers it premature for FDA to impose its flopesca optical method and place a limit on asbestos in talc for food u.fJG . The CTFA urges deferment of promulgation of the proposal until: 1. Determination of the standard deviation of the proposed compliance test method. 2. Evaluation of the valrcitv of the assum.otion that a sincle milligram sample can be taken as statistically representa tive of any lot as well as the limit size of any lot so sampled. 3. A re-evaluation of the economic cost and practicality of th proposed Comoliance Test Methods-. ' * 4. An intensive' search is undertaken for an alternate prccc-cur- that is more reliable and mors practical than the proposed test methods. The Cosmetic, Toiletry and Fragrance Association, Inc., stands z to assist the Feed and Drug Administration in method develocr.er.: research as well as to provide assistance in other areas discuss above where as its scientific expertise ray be of value. Cordially, Norman F. Estrin, Pli.D. Vice President - Science N F E :mp ,. Attachment (b)(5) REPORT OF CTFA TALC SUBCOMMITTEE ON METHOD TO DETECT CHRYSOT1LE AND TREMOLITE IN TALC. ' .* December 10, 1973 The Scientific 'Advisory Committee of the Cosmetic, Toiletry & Fragrance Association, 1625 Eye Street, N.W. , Washington, D.C. 20006 at a meeting held on September 2, 1973, designated that a subcommittee be set up to study the optical microscopic method proposed by the Food and Drug Administration. This method was published in the Federal Register of September 28, 1973. The present membership of the CTFA Talc Subcommittee is shown on an attached sheet. At the first meeting of the Subcommittee, it was agreed to submit several samples of talc to members who volunteered to apply the FDA proposed optical microscopic method. The following talc samples were distributed without identification - except for randomly applied and non-duplicated code numbers: Italian Talc Montana Talc Montana Talc Source I Vermont Talc Source II Alabama Talc No. Carolina Talc ); In addition, a sample of Italian Talc spiked w/v with 1% chrvsotile and 0.15% tremoli.te was sent to each participant. The participants were aware that the sample was spiked, but did not know the percentage of spiking. The attached table summarizes all reports submitted: Final Report (Over) Sample , Montana-^ Montana-^ Alabama Vermont ' No, Carolina No. of Fibers No. of Fibers With R.I. Greater With R.T. Less Than 1590 Than 1574 Status With Regard to Froposed FDA Method Dr. John A. Refiner (U. of Conn.) for Avon Cont'd 0 13 26 41 14 5,844 9,281 15,125 165 11,852 Fails chrysotile Passes tremolito Fails chrysotile Passes tremolite Fails chrysotile Passes tremolite Falls chrysotile Passes trcTT.olite Fails chrysotile YPNa_s_ se,,s~ tremo.lIin te Dr. Trypr.ve Baak of United Sierra Spiked Italian Montana-^ Montana-H Alabama Vermont > Mo. Carolina 3,000 1,000 0 0 0 0 500 . 2,000 0 0 0 0 0 0 (b) (4) Italian Montana-^ Montana-^ Alabama Vermont No. Carolina 8,250 18,000 29,250 21,825 i 225 20,850 1,350 450 5,200 6,075 675 6,000 Spiked sample was not run. Fails Passes Passes Passes Passes Passes Passes for Whittaker C U Fails chrysotile & Fails chrysotile & Fails chrysotile ts Fails chrysctilc Ci Fai Is chrysotile If Fails chrysotile & 0t o (Over) / Sample Spiked Italian Montana- I Alabama Vermon t No-Carolina Montana- u .C-'>' No. of Fibers With R.I. Greater Tiian 1590 it-1'" No. of Fibers With R.I. Less Than 157A Status With Regard to Pro posed FDA Method Mrs. Lucy McCrone, McCrone Associates for Chesebrouph 320 16 Passes 120 0 Passes 0 0 Passes 0 0 Passes 0 0 Passes 0 0` Passes 0 *0 Passes Mr. Harold Stanley of Pfizer Italian Montana- I Montana- 11^. Alabana Vermon t No. Carolina Spiked 100 -- 100 -- 100 ~100 100 ^ 100 <1000 > 100 >100 >100 >100 >100 >100 >100 Mr. David Hamer of Johnson & Johnscn Italian Montana- I Montana-* II Alabama Vermont No. Carolina 0 1 25 * 0 0 0 0 0 0 0 0 0 Fails Chrysotile Passes trer.olite Fails Chrysotile Passes tremolite Fails Chrysotile Passes trer.olite Fails Chrysotile Passes trer.olite Fails Chrysotile Passes tremolite Fails Chrysotile Passes trer.olite Fails Chrysotile Passes tremolite Passes Passes Passes Passes Passes Passes Spiked Italian Dr. John A. Reffner (U. of Conn.) for Avon 632 12,930 Fails Chrysctile Passes Tremolite 220 1,114 ..-.Fails Chrysotile Passes Tremoli tc -4- Sample No, of Fibers With R. I. Greater Than 1590 No. of Fibers With R.I. Less Tlian 1574 < (b) (4) Italian Mont ana-I Montana-U Alabama Vermont *No. Carolina Spiked Sample 53 0 0 0 0 0 488 0 0. 0 0 0 0 16 Status With Regard to Froposed FDA Method WNJfor Kolmar Passes Passes Passes Passes Passes Passes Passes *Sample noted to contain some fiber bundles which are roiled up talc pi some non-fibrous tremolile The following participants examined their samples but were unable conscientiously to count particles vi th confidence, therefore, did not report numbers: Miss Marie Jones for Dr. G. Cohen - Bristol-Myers Products Reseat' -v Hr Salvatore D1 B^anra - The Mcnnen Company (b) (4) t , Research Center *- for Klner Laboratories Mr. John Facq - The Colgate Palmolive Company> Research Center The table reveals strong inconsistency between results obtained by the different scientists applying the method to the same group of coded talc samples. This inconsistency is the result of problems encountered in the methodology, 1. The alph i refractive index of talc - one of the two indices exhibite [ by talc plates standing on edee - varies in value between 1.539 and 1.550-T This in combination with item. 2, below, ni ly easily result in the mis-clnssificstion of such plates a; "fibrous, less than 1.574" and therefore chrysotile. Dr. Reffner of the University of Connecticut, in his report to Avon, admits that this factor is paramount - "counts for chrysotile are high since (talc) edges will often be ni St alien for fibr ms particles". ' -2. The immersion method for the determination of-relative refractive index becomes unreliable when applied to extremely small (ca. 1 jjm)par tide's . This is due in part to the fact that rarticlcs around IJW1 in width approach the limit of resolution of the human eye When both size and contrast ore considered, especially when applying the Becke line technique prescribed. This will - J- cause eye fatigue and resulting errors. Thus, a fine talc shard or plate on edge nay be easily mistaken for an asbestos fiber. Ibis applies to both the chrysotile and ampbibole portions of the test. i It is quite likely that chlorite - a mineral of varied compo sition -commonly found in pharmaceutical grade talc - would be incorrectly identiticd as trcnolite. Refractive indices for chlorite vary in the range 1.57-1.662 . Hie examination of a sample of one milligram dispersed on a single microscope slide presents the investigator with a preparation which is too dense for a petrographic study. The method is laborious and time consuming. Dr. Refiner quotes an analysis time of about five hours per sample and at JOHNS OR {, JOHNSON1 the investigator (D. H. Knmer) estimated at least two or three hours. . ! in an earlier letter to the FDA suggested tnat there were relatively few scientists in the country who could use the method effectively. Scientists at Sr is tul -'.'vers , The Mermen Company, Liberty Mutual and Coi tate-HaInolive had great difficulty in following the method to conclusion. The proposed FDA method implies that all fibrous particles; in talc having optical properties similar to those of asbestos minerals are in fact asbestos. This is probably an unwarranted assumption. REFERENCES: ' 1. Deer, W.A., Howie, R.A,, and Zusstnan, J., Rock-Form.inc Minerals, Vol. 3, p. 121 (1962). 2. Ibid, p. 131. * * * * * * * * * * * -k i. * i A -m J*Je have concluded that the method published in the Federal Register does not provide a. truly reliable means for the detection of asbestos in talc. It results in both false-positive and false-negative findings. It is also tedious and may consume as much as one half day per samnle. The subcommittee urges that the Food and Drug Administration defer finalizing the proposed optical microscopic method and proceed into a Phase II program which would combine FDA and Industry in a strong, effort to develop a truly reliable method for measuring chrysotile content in talc. We are confident that the. detection and estimation of fibrous trenolite can also be done, if necessary, as a fringe benefit, at the sane time. ' f Htrriv- -6- We estimate that a satisfactory method will take at least six months to a year to develop. We suggest a close liaison between FDA and industry in this all-out effort with updating reviews to take place as frequently as necessary. Review of Alternate Methods The CTFA Talc Subcommittee has reviewed several methods to detect chrysotile and tremolite in talc, including modifications of the optical microscopic method. These are listed with comments. 1. Optical Microscopic Method - may be used with some modification. Possibly better selection of R.I. liquids and reduction of sample size per slide. 2. X-ray Step Scanning - A good method lor detecting tremolite if we could accept about 0.2% as the threshold of detection. It does ^ yi not distinguish fibrous from non-fibrous trer.oli te. The simultaneous i-1 a detection of chlorite makes method impractical for chrysotile. 3. X-Ray Step Scanning + Optical Micros cony - Acceptable as in t12 above but the optical method continues to present the dilemma in reliable identification of chrysotile. However, this docs not rule out a major revision of the optical method to do the chrysotile identification and counting, i- 4. X-Ray Scanning - The problem of chlorite interference in the detection of chrysotile is present in all x-ray procedures thus far available. It is reliable for the detection of 1% tremolite (both fibrous and non- fibrous) . 5. Differential Thermal Analysis - This procedure is capable of detecting chrysotile at the 1% level, however, it will not detect tremolite. 6. Scanning Elect ronMicrosconv - This procedure is not capable of identi fying asbestos. Even with energy dispersive analysis, completely satisfactory identification is not possible. 7, Transmission Electron Microsconv 4- Electron Diffraction - This appears to offer the best, most reliable method and is probably capable of detecting chrysotile and tremolite (fibrous), both at a level, of 0.1%. It is estimated that an installation would cost about $130M +, and is obviously prohibitive for the small manufacturer who uses talc. The amount of talc sample examined by this procedure is miniscule. -7- Summary . The CTFA Talc Subcommittee Has completed its review of the Optical Microscopic Method for the Estimation of Chrysotile and Fibrous Trenolite in Talc, which appeared in the Federal Register, vol 38 (No. 188), (p. 27076 , September 28, 1973). It recommends postponement of the finalization of the proposed regulation on talc. In addition, a collaborative: effort between FDA & industry to resolve a satisfactory method of estimating chrysotile and tremolite in talc, is recommended, with periodic liaison reviews to monitor the project status. A task force of industry experts has been designated to pursue alternate methods. Ve invite FDA to participate in collaboration with the task force. v t f ' George V, Sandland / ... r . ___ Chairman CTFA Talc Subcommittee (b) (5)