Document nRNkjwxwr3geE9v51aVmLqvm

JOSEPH E. KELLER JEROME H. HECKMAN CRADLES M. MEEHAN WILLIAM H. BOnOHESAXI, JR. ROBERT S. TtERNAN W4TSI V. SLACK DAV7D L. HILL MABTIN W. BEDCOVTCl PETER M. NEMNQV JOSEPH E. UADLET, JB. cabole c. Harris PETER THOMAS SMITH MICHAEL F, MOHHONE LARBT S. SOLOMON JOHN B. DCBECK. CHRISTINE A. MF.AOHEB LAW OFFICES Keller and Heckman MSO ITTM STREET, W. W. SUITE JOOO WASHINOTOy, I>, C. 20000 April 8, 1977 TELEPHONE 209 -L3T- UOO cable addbess "kiuiax" WBITEB'S DIRECT DIAL NUMBER 202-457-1110 Dr. Howard R. Roberts Acting Director Bureau of roods Food and Drug Administration 200 C Street, S.W. Washington, D.C. 20204 Re: Docket No. 75N-0190; Vinyl Chloride Polymers in Contact With Food? Notice of Proposed Rule Making, 40 Fed. Reg. 40529, September 3, 1975 Dear Dr. Roberts: The purpose of this letter is to follow up .on the January 5, 1977 conference between your staff and representa tives of the Vinyl Chloride/Polyvinyl Chloride Producers Group of The Society of the Plastics Industry, Inc. (SPI). Responsive to the three requests made by members of your staff during that conference, we are herewith submitting tho following: (a) "raw data" to confirm and sub stantiate the submitted reports of the impressive achievements in reducing residual vinyl chloride monomer (RVCM) to insigni ficantly low levels in commercially available polyvinyl chloride products intended to contact food, (b) descriptions of the analytical procedures employed to determine not only RVCM levels but also the levels of vinyl chloride (VCM)in food simulating solvents. and 22568001 BFG21587 Dr. Howard R. Roberts April 8, 1977 Page Two (c) a memorandum that provides a full rationale for the conclusion that no vinyl chloride can reasonably be expected to mi grate from rigid and semirigid polyvinyl chloride food-contact products under in tended conditions of use, provided said products are made in accordance with good manufacturing practices that limit RVCM to a level not exceeding 0.1 parts per million. 1/ It is respectfully submitted that the record, in its entirety, compels the determination that the public interest warrants the use of rigid and semirigid PVC food packaging materials produced in accordance with appropriate good manufacturing practices. Such PVC food packaging materials have been found environmentally preferred over many competing materials by the Bureau of Alcohol, Tobacco and Firearms (BATF)in its assessment of PVC bottles for liquor, 2/ and that determination is confirmed by FDA's findings in its Environmental Impact Statement relating to plastic containers for Beverages. 3/ In particular, the PVC containers provide important benefits with respect to freedom from breakability and the absence of problems with regard to safety. We submit that it is fatuous to claim a risk to health exists when no vinyl chloride can be measured in food simulating solvents after exaggerated exposures using analytical procedures sensitive to 1 or 2 parts per billion. This is especially true when one considers that under no conceivable circumstances will as much as 10% of the diet be packaged in rigid and semirigid PVC products. The pre sent level of analytical sensitivity is so low that the risk is less than approximately one ten-thousandth the risk deemed by FDA to be "safe" in an analogous context. 4/ This 1/ Due to manufacturing tolerances, a maximum limit of ZT.l ppm requires the average product to be well below that level. 2/ BATF, "Final Environmental Impact Statement, Polyvinyl Chloride Liquor Bottles," March 9, 1973. 3/ FDA, "Final Environmental Impact Statement, Plastic Bottles for Carbonated Beverages and Beer," September, 1976. 4/ 42 Fed. Reg. 10412 BFG21588 Z0089S8S Dr. Howard R. Roberts April 8, 1977 Page Three level is at least three orders of magnitude lower than the level the Occupational Safety and Health Administration (OSHA) determined to be safe when it wrote its standard for occupational exposure to .vinyl chloride. 5/ The data submitted to the Food and Drug Administra tion {FDA) during the pendency of the subject rule making has demonstrated that the polyvinyl chloride industry has reduced the residual vinyl chloride levels in rigid and semirigid products from levels in the range of approximately 500 parts per million, which were common before the possi bility of migration of vinyl chloride was recognized, to levels well below 0.5 parts per million for many products, a thousand-fold reduction. Furthermore, some products can now be produced with levels of RVCM no greater than 0.1 part per million? and in some cases, particularly for sheet materials used in so-called blister packs, the levels do not exceed 0.05 parts per million. In other words, poly vinyl chloride rigid and semirigid products are now avail able to the food packaging industry with vanishing small levels of residual vinyl chloride and assure that there will be "no migration" 6/ of vinyl chloride to food packaged in such products when tHe foods are packed and otherwise handled in accordance with good manufacturing practices for the production and marketing of food products. 7/ 5/ 29 C.F.R. 1910.1017 6/ The expression "no migration" is used as a convenient shorthand to mean "no reasonable expectation of becoming a component of food under the intended conditions of use." 7/ It is not claimed nor is it necessary that all rigid and semirigid polyvinyl chloride products be capable of showing no migration of vinyl chloride to food packaged therein; it is only necessary that those products which are offered as, and represented to be suitable for use as, food packaging materials comply with such requirements. Thus, the fact that all manufacturers may not now be offering materials which will provide a "no migration" package should not lead the FDA to ban all PVC products any more than the FDA would ban all paper from food packaging uses because some grades of paper are unsuitable for that purpose. See Natick Paperboard Corp. v. Weinberger, 525 F.2d 1103 (1st Cir. 191^)~. Accordingly, we urge, as we have in the past, that the FDA set appropriate standards in its good manu facturing practices regulations to assure, in light of the intended uses of such material, that there will be no migrar tion of vinyl chloride to packaged food. BFG21589 CO099S22 Dr. Howard R. Roberts April 8, 1977 Page Four The raw data and descriptions of analytical pro cedures are presented in Appendices I through IV. Appendix I, supplied by the American Hoechst Corporation, Film Divison, summarizes the residual vinyl chloride content of PVC resins received by American Hoechst during 1974 through 1976, pro vides a typical detailed report of lot-by-lot results for RVCM for the month of October 1976, provides typical chromato grams for a sampling of the individual lots reported for October 1976, demonstrates the reduction in RVCM achieved during the blending process wherein resin as purchased is converted to a product ready for formation into sheet product, provides typical chromatograms for material before and after blending, and, finally, sets forth the detailed analytical procedures that were employed. Appendix II was submitted by the B. F. Goodrich Company. It includes an update of the RVCM levels in both resin and finished bottle compounds supplied by B.F. Good rich during the last quarter of 1976. The gas chromato graphic procedure employed by B. F. Goodrich for the determina tion of the residual vinyl chloride monomer content of- poly vinyl chloride resins and wet cake samples is provided, as well as a report by J. A. Nikora and E. G. DeCapita entitled "Confirming the Presence of Vinyl Chloride in Food Simulating Solvents." Appendix III supplied by Ethyl Corporation consists of information already supplied separately to FDA. It includes a document dated January 10, 1977, addressed to the Hearing Clerk, containing Addenda A through E as direct responses to the January 5, 1977 request for raw data and analytical methodology. In addition, there is a separate submission also addressed to the Hearing Clerk by means of a letter dated January 13, 1977, entitled "VCM Migration Studies with PVC Bottles Containing Low VCM Concentrations" dated January 10, 1977 written by Cannaday, Daniels, and Gaeke. This paper is particularly significant because it demonstrates that the chromatographic determination of vinyl chloride in food simulating solvents used to test low RVCM containers is subject to significant interferences and the consequent appearance of false positives. This same ob servation was reported by Tenneco in its independent sub missions to FDA, and we understand similar effects have been observed by FDA scientists as well. BFG21590 J \f H )l)9 9 S Z Z Dr. Howard R. Roberts April 8, 1977 Page Five In order to provide more definitive evidence of interferences and to provide direct experimental evidence for setting a residual vinyl chloride level in bottles or bottle compounds that will assure "no migration," a protocol for a study was developed in consultation with FDA scientists; and samples of the experimental bottles have already been provided to the FDA staff. The test is now under way; and when the exposure time has been completed, samples of the extract will be supplied to the FDA scientists for their own independent analysis. The final results are expected soon, and will be reported by Ethyl Corporation as soon as they are available. Appendix IV has been supplied by the Ruco Division of Hooker. This includes analytical procedures for the determination of RVCM in resins by either a solution pro cedure or by head space analysis, typical quality control chromatagrams including standardizations, blank determina tions and typical analyses of bottle compounds using both procedures, and the analysis of a compound deliberately "spiked"* to show a large quantity of vinyl chloride. The final attachment, Appendix V, is a memorandum that discusses in considerable detail why there is no reason able expectation of migration of vinyl chloride into food under intended conditions of use from rigid and semirigid PVC food-contact materials that contain very low levels of residual vinyl chloride. Taken together, these appendices confirm current industry capability of producing PVC resins with residual monomer levels well below 0.5 parts per million, that pro ducing compounds from these resins results in a loss of from 50% to more than 90% of the RVCM originally contained in the resin and that processing into finished products still further lowers the residual vinyl chloride level. In other words, rigid and semirigid PVC containers can be produced for food-contact purposes with residual monomer BFG21591 0089522 Dr. Howard R. Roberts April 8, 1977 Page Six levels not exceeding 0.1 pacts per million; and when such materials are used in food packaging, there is no reasonable expectation that vinyl chloride will become a component of food under the intended conditions of use. Accordingly, we urge the FDA to adopt final Regulations consistent with these facts. Cordially yours, K BFG21592 2566006 Appendix V MEMORANDUM THERE IS NO REASONABLE EXPECTATION THAT VINYL CHLORIDE WILL BECOME A COMPONENT OP FOOD IF THE RESIDUAL VINYL CHLORIDE CONTENT OP THE PACKAGING MATERIAL DOES NOT EXCEED ONE-TENTH PART PER MILLION Background In its Notice of September 3, 1975 (Docket No. 75N-0190, Vinyl Chloride Polymers In Contact With Pood) the Food and Drug Administration proposed to deal separately with polyvinyl chloride (PVC) food packaging materials in three categories: 1. Uses to be affirmed as prior sanctioned including coatings, gaskets, cap liners, flexible tubing, plasticized film and the like; 2. Potable water pipe; and 3. Rigid and semi-rigid materials. SPI filed extensive Comments which included a comprehensive legal analysis of the meaning of the term "food additive" as defined in Section 201(s) of the Federal Food, Drug and Cosmetic Act, voluminous scientific data relating to residual monomer levels in rigid and semi-rigid PVC food contact materials, and the results of the migration studies conducted with such materials. Subsequent to the filing of these Comments, SPI and member companies thereof continued to provide to the Food and Drug Administration additional scientific data as they were being developed bearing on the possible migration of vinyl chloride (VCM) from rigid and semi-rigid PVC food packaging materials. BFG21593 / 2 In anticipation of the promulgation of final regulations in this proceeding the industry requested a meeting to bring the Food and Drug Administration completely up-to-date regard ing the major improvements made by the PVC industry in reducing residual vinyl chloride (RVCM) levels in food contact materials to vanishingly low levels. This information was presented at a conference held in Washington, D.C. on January 5, 1977. At the conclusion of the conference the Food and Drug Administration representatives requested (1) that the informa tion supplied be supplemented with "raw data" bearing on the results that were reported, (2) that the analytical procedures employed be supplied, and (3) that explanatory information be furnished to provide a rationale for the industry conclusion that, from currently available rigid and semi-rigid materials, there would be no reasonable expectation of migration of vinyl chloride to food. The technical information requested is being supplied in separate documents which are being submitted at this time. The requested rationale forms the substance of this document. 1/ As a matter of internal management procedures it is under stood the Food and Drug Administration attempts to promulgate final regulations or withdraw a proposed regulation within two years of the publication of a Notice of Proposed Rulemaking. Q im S Q Z Z BFG21594 3 Scientific--Technical Considerations 1. Clearly, the most significant factor leading to the conclusion of "no migration" 2/ is that no vinyl chloride could be detected in the appropriate test solvents using analytical procedures of extreme sensitivity, i.e., capable of quantifying vinyl chloride at a level of 1 or 2 parts per billion (ppb) when exaggerated extraction tests were conducted. In other words, rigid and semi-rigid PVC packaging materials containing the very low levels of residual VCM now being achieved in commerce were tested under conditions designed to reasonably exaggerate the likelihood of migration and no such migration could be detected. This demonstrable fact alone should be sufficient to confirm the validity of the conclusion reached by the in dustry scientists that there will be "no migration" of vinyl chloride into food under intended conditions of use of rigid and semi-rigid PVC packaging materials of the quality now being supplied. It should be borne in mind that conclusions in the physical sciences must be based on empirical evidence. Theory and speculation guide the choice and design of experiments but the experimental results are the only conclusions. 2/ it should be appreciated that the phrase "no migration" Ts used in this document as a convenient shorthand for the more lengthy statutory criteria embodied in 201(s) of the Act of "no reasonable expectation of becoming a component of food." BFG21595 60093S2Z 3 2. The experimental data cited above are supported and reinforced by theoretical considerations. The "driving force" that makes a substance migrate from the interior of a container wall to the surface and, thereafter, into the container contents --^ is the difference between the effective concentration of the substance in the container wall and its concentration in the food. When the level in the container wall is relatively high, a large differential in concentration exists; the resistance of the plastic to migration is overcome, and migration is rapid and extensive. As the level in the container wall is reduced, the concentration differential is reduced; and both the rate of migration and the quantity migrating diminish. Finally, when the concentration in the walls becomes vanishingly small, as is the case in the presently available PVC containers, the driving force cannot readily overcome the resistance; and both the rate of migration and the quantity migrating likewise become vanishingly small to the point where the substance cannot be detected even at extremely low levels. Indeed, there is no way of knowing that the resistance can 3/ It is recognized that migration from the container wall to the exterior will also occur. Data in the file show that of the RVCM actually migrating from "high" RVCM containers approximately one-third migrates into the contents and twothirds migrates to the exterior. BFG21596 [ CT089S2Z 4 really be overcome at all and that any migration will occur, in short, when the concentration in the walls of a container becomes as small as it is now, there is no realistic expec tation of migration. 3. This explanation is still further reinforced by the following considerations: If one were to consider a typical rigid container wall (0.020 inches thick) that contained 0.1 ppm of residual vinyl chloride, and make the totally unrealistic assumption that all the residual vinyl chloride available for migration into the contents did in fact migrate to food packag ed in such a container, the maximum calculable concentration of vinyl chloride would be approximately 2 parts per billion (ppb), a quantity capable of analytically quantifiable measurement. The evidence cited above demonstrates that such total migration does not occur. Much scientific evidence already provided to the Pood and Drug Administration has demonstrated with con tainers having high levels of RVCM that far less than the maxi mum potentially migratable quantity of vinyl chloride does in fact migrate even under exaggerated exposure conditions. It is apparent, therefore, that with containers made with the extremely low level of RVCM discussed here, there can be no realistic expectation of migration of vinyl chloride. 4. Summarizing then, scientific-technical considerations based upon available information and established scientific BFG21597 TT089SZZ 5 principles lead to the conclusion that there is no reasonable expectation of migration of vinyl chloride from container walls when the level of residual vinyl chloride in such walls is adequately low. The data presented in this docket indicate that 0.1 ppm is an "adequately low" level. Practical Considerations 1. Since the chemical tests show "no migration" of vinyl chloride, the major practical consideration is whether the analytical procedures used to determine whether vinyl chloride migrates are sufficiently sensitive to assure the public health and safety when the "no migration" conclusion is relied upon. The commissioner has answered this question in the affirmative in an analogous context, and we submit those same principles are clearly applicable here. In his rulemaking on chemical compounds in foodproducing animals, published in the Federal Register on February 22, 1977 (42 Fed, Reg. 10412), the Commissioner clearly embraced the concept that the Mantel-Bryan biostatistical procedure is suitable for specifying a level of analytical sensitivity for resolving the question as to whether or not carcinogenic drug residues^ are present in edible tissues of an animal to which that drug was fed. If no residue of the drug in issue can be detected in the edible tissues of animals fed such drug BFG21598 Z fO Q S Z Z Z 6 by means of an analytical procedure with a capability of de tection at levels specified by the Mantel-Bryan procedure as adequate to assure the public health and safety, the Commissioner has determined that no drug residue is present. The same biostatistical extrapolation procedure, applied to vinyl chloride, will specify a detection sensitivity that will likewise assure that a finding that no vinyl chloride is present in food or foodsimulating solvents will assure the safety of the general public. Such a finding should likewise lead to the conclusion that no vinyl chloride is present--that there has been "no migra tion" of vinyl chloride. 2. The file in this proceeding demonstrates that the application of the Mantel-Bryan procedure to the data available at the time our original Comments were filed shows that a mini mum detection capability orders of magnitude higher than that actually available would be sufficient. In other words, a finding of no vinyl chloride in food-simulating solvents after suitably exaggerative exposures provides far greater assurance of no significant risk of harm than the Commissioner indicated would be sufficient in the case of animal drug residues, hence, a far sounder basis for reaching a "no migration" conclusion. n BFG21599 ET099SZ 7 3. In the section on Scientific-Technical Considera tions/ calculations and conclusions were presented for a typical bottle wall thickness of 0.020 inches and a residual VCM con tent of 0.1 ppm. However, evidence has been presented showing that for many applications thinner walled containers and mate rials of much lower RVCM are used. Furthermore, many appli cations exist involving refrigerated shipment and storage, as well as products with limited shelf life. To the extent that any of these considerations apply, there will be still greater assurance of "no migration"; to the extent that several are simultaneously effective, the degree of assurance will be again compounded. 4. One final practical consideration should be mentioned. Vinyl chloride is relatively insoluble in aqueous or even fatty foods: it tends to evaporate from food to the air which is in contact with that food. This phenomenon is the basis for the most sensitive analytical procedures now available since the vinyl chloride in a relatively large volume of food or solvent is readily transferred into and thus concentrated in a smaller quantity of air. Thus, even if a measurable level of vinyl chloride were present in the contents of an actual food container, the vinyl chloride would tend to concentrate in the head space leaving only a very low level in the food itself. When the BFG2^600 frT 089S ^ 8 container is opened, the vinyl chloride in the head space will escape immediately. Upon reclosure the same redistribution phenomenon from food to head space would reoccur and upon reopening the second increment of vinyl chloride would escape. Likewise, food removed from such a container and exposed to air, stirred, heated or otherwise prepared would experience a rapid and substantially complete loss of any vinyl chloride it might have contained, in other words, where formerly used containers may have permitted the migration of vinyl chloride to a limited extent to the contents of such containers, the concentration of vinyl chloride that actually became a component of the diet was tremendously reduced. The application of this same principle to food packaged in the low RVCM containers adds still another major safety factor to assure there will be no vinyl chloride in food as consumed. Legal 1. Comprehensive and well documented evidence has been submitted to the Commissioner demonstrating that there are no detectable extractives of vinyl chloride from food-contact articles into food-simulating solvents at a detection limit of approximately 0.001 part per million. It is respectfully submitted that this showing demands the conclusion there is no basis for regulating the residual vinyl chloride content of PVC food-contact articles as a Food Additive as defined BFG21601 ST0S9S2Z 9 in Section 201<s) of the Act. Failure to reach this conclusion would require the Commissioner to find that the mere presence of residual vinyl chloride monomer in the food-contact article is sufficient to determine that VCM in fact becomes or reasonably may be expected to become a component or otherwise affect the characteristics of any food irrespective of a factual basis for such a conclusion. It is respectfully submitted that such a determination is inconsistent with the governing statutory standard and the Commissioner's authority. (2) The Food Additive definition in Section 201(s) of the Act entails essentially four (4) criteria: (a) That the substance becomes a component of food, (b) That the substance otherwise affects the charac teristics of food, (c) That the substance may reasonably be expected to become a component of food, or (d) That the component may reasonably be expected to affect the characteristics of food. The credible evidence before the Commissioner indicates that such residual vinyl chloride as may remain in PVC food contact articles neither becomes a component of food (i.e., is not a detectable component) nor affects the characteristic of any food. Furthermore, in that the known behavior of vinyl chloride in small but measurable quantities evidences that its presence in food does not affect the characteristics of such food, a BFG21602 22568016 10 fortiori, its possible presence in immeasurable quantities cannot affect the characteristics of food. Thus, the first, second and fourth criteria of Section 201 (s) as stated above do not provide a basis for regulation as a food additive. Accordingly, the issue presented is whether vinyl chloride present in PVc food-contact articles at a level not exceeding 0.1 ppm may reasonably be expected to become a compo nent of any food. Obviously to reach an affirmative conclusion relies on a priori reasoning that necessarily ignores the only relevant scientific data, namely that there are no detectable extractives of vinyl chloride in the food-simulating solvents. Recognizing that there are few, if any, absolutes with respect to food additive chemistry, the Congress estab lished a standard based upon reasonable expectation rather than upon an absolute accounting for all potential contingencies under all undefined conditions. As discussed in SPl's December 19, 1975 Comments, hypothetical abstractions which cannot be proven do not give rise to reasonable expectations; in the absence of a clearly expressed congressional intent to the contrary, such hypothetical abstractions do not form the basis for the exercise of delegated regulatory authority. Moreover, such a conclusion is not permissible under Section 201(s) which defines food additives in terms of "becoming a component or otherwise affecting the characteristics of any BFG21603 11 food. ..." in that the phrase "o otherwise affecting the characteristics of any food" refers back to the "component" criteria, it thereby indicates a statutory intent that a compo nent affect the characteristics of the food. Thus, to the extent that a component, in the statutory rather than theoretical sense, must affect the characteristics of any food, it must be measurable and perceptible in and of itself or in its effect. Otherwise stated, the effect of being a component is one of discernable physical presence. It is established, however, that vinyl chloride at levels less than 0.001 part per million does not have either an ascertainable physical presence or any other effect. Thus, it cannot reasonably be considered to be a component of any food. The Mantel-Bryan procedure adopted by the Commissioner in 42 Fed. Reg. 10412 confirms that FDA concurs with this position in an applicable analogous Section 409 situation. tO BFG21604 8708952 i Josi;hii e ki.u.ek JEROME II HECKMAN CJIAHLKS M MEEHAN WILLIAM II nORtillESAM JH ROBERT R TIF.RNAX VAYN'E V DLACK DAVID L HILL MARTIN' W BERCOVICI PETER M XEMKOV Joseph e Hadley, jr. CAROLE C. HARRIS PETER THOMAS SMITH MICHAEL F MOKHON'E LAFRT S. SOLOMON Law oPncr.s Kellet* and Heckman* USO ITT" STREET, NT. W. Sl'TTE IOOO WASHINGTON, D C. 20030 January 3, 1977. TELEPHONE 208 4ST-IIOO CABLE ADDRESS"KELMaN" WRITERS DIRECT DIAL NUMBER Howard R. Roberts, Ph.D. Acting Director Bureau of Foods Food and Drug Administration 200 "C" Street, S.W. Washington, D.C. 20204 Re: Vinyl Chloride Polymers in Contact With Food, Notice Of Proposed Rulemaking? Docket No. 75N-0190 Dear Dr. Roberts: The purpose of this letter, sent in anticipation of our meeting with you and your associates on January 5, is to provide up-to-date information which we have elicited and received from key members of the VCM/PVC Producers Group of our client. The Society of the Plastics Industry, Inc. In light of the long pendency of the ref erenced rulemaking proceeding, we requested the meeting and have obtained these additional data because we con sider it essential that the Food and Drug Administration be as fully advised as possible about the current state of the art vis-a-vis potential polyvinyl chloride food contact surface technology. In our opinion, and we are most hopeful that FDA's scientists will agree, the in dustry has now progressed to a point where it cannot be legitimately stated that polyvinyl chloride food contact surfaces, including such packages as PVC bottles and "blister pack," may reasonably be expected to result in vinyl chlo ride monomer becoming a component of food unless one gives the word "reasonable" a type of meaning that would amount to a reductio ad absurdum. Before briefly summarizing and discussing the attached reports received from the members of the Society, it occurred to us that it might be best to restate a little of the background relevant to this situation al though we are certainly aware that you and your associates are fully cognizant of where the matter stands. TO BFG21tf05 3 of data. One reports the residual monomer content of the PVC homopolymer resins it has been purchasing in the period from 1974 through 1976? the other provides an in dication of the reduction of residual monomer content achieved by its new blending process in going from resin to finished compound before sheet formation. Supplemental to the data on residual monomer levels supplied by each of the participants in this sur vey, we are also including a copy of an internal report prepared by C.A. Daniels of the B.F. Goodrich Company which provides predicted vinyl chloride migration data obtained from a predictive model developed by Goodrich. The data were developed for migration into 50% ethanol/ water, a solvent known to be extremely aggressive in extracting vinyl chloride from PVC containers. These calculated results are considered to be conservative in that they are based on the assumption that the diffusion coefficient is constant and does not decrease as the concentration of residual monomer decreases (i.e. that the system does not show the so-called `'Gilbert'' effect.) Further conservatism is introduced into the assumptions of this model in that no allowance is made for loss through the seal, a phenomenon which appears to be real and ubiqui tous and which tends to reduce the actual monomer level within a container below those calculated on the assump tion that "leakage" does not occur. In light of the fact that the information pre sented here has been developed independently and reports the discrete experience of the different reporting com panies, it is our view that it is best to let the reports speak for themselves rather than attempt to provide any detailed characterizations of our own. Nevertheless, we believe it is fair to say the data demonstrate that the PVC industry can now provide, and is providing, bottle and sheet compounds with residual monomer levels suffi ciently low as to warrant the conclusion of any ordinary, reasonable and prudent man that the likelihood of migra tion to foods, drugs or cosmetics is non-existent. Cer tainly, in light of data already provided to the Food and Drug Administration by Ethyl Corporation regarding samples which had been stored for longer than two years, data submitted in the Comments which showed no detectable VCM migration from bottles with approximately 0.35 ppm ~ RVCM in the walls, and the predictive model data which is being provided herewith, we believe it would be * if 5 BFG21607 J. R. Legs QiflCCTOe or Plastcs Gaou* Ethyl Corporation December 23, 1976 Mr. Jerry Heckman Keller St Heckman 1150 - 17th Street NW Suite 1000 Washington, D. C. 20036 Dear Jerry: Attached is our position paper a little in advance, but we think we can handle what it says. We will be available for conversa tion by phone all next week and are planning to attend the meeting on January 5. We look forward to receiving your compendium. Very truly yours. mar attachment i ' ', . ) J. R. -Lees 22569013 Ethyl Tower / 451 Florida / Baton Rouge, Louisiana 70801 / 504--3B8-7S77 BfG2l609 2- of compound and bottles, the final VCM levels possible in bottle walls do not exceed 80% of the above compound values. The compounding resin from which Ethyl bottle compound is produced is also rigidly controlled in regard to reduced VCM level. Our production specification for bottle grade resin is less than 10 ppm VCM content. To maintain this level in production also requires averages substantially lower than this number. Average bottle resin level during the latest 431 lots was 4.33 ppm. It cannot be over emphasized that to maintain a rigid specifica tion of 10 ppm maximum in bottle resin results in actual lot analyses covering a distribution of values whose mean must of necessity fall significantly below the control specification. The identical considera tion applies to bottle compound produced from this resin. Thus, Ethyl starts with bottle resin produced with a distribution of values centering about a mean of about 4. 0 ppm to. maintain a maximum control specifi cation of 10 ppm and converts this resin to bottle compound having in turn a distribution of values centering about a mean of 0. 2 ppm (to maintain a maximum control specification in compound of < 1 ppm VCM). BFG21611 ST0o92Z - 4submitted to the FDA extension long-term storage data representing today's commercial bottle compounds stored under exaggerated conditions. The bottles employed were small (2 oz.) and contained ^1 ppm VCM in the wall. Results showed no detectable vinyl chloride monomer migration, after 27 months' room temperature storage, in standard FDA foodsimulating solvents. Reference Ethyl Corporation submitted to Hearing Clerk, Food and Dru'g Administration, dated Sept. 27, 1976, BFG21613 T069SZZ STATEMENT BY UNION CARBIDE CORPORATION REGARDING VINYL hKqINS' TOR RIGID AND SSMf-RICID SHEET rNTh'NDEn FOR VGZ IX FOOD CONTACT*APPLICATIONS This- Is Essentially a condensation of that potion of our submission dated November 10, 1375, (Docket No, 75-N-019Q) which deals with the use of Union Carbide Corporation solution fcs.d non-solvent resins in food contact applications. Solution polymerised vtnyl chloride-vinyl acetate resins are produced from monomers, solvent, Anti peroxide catalyst in a stirred autoclave under controlled conditions of temperature and pressure. Polymer formed ts soluble in the solvent.. Dissolved polymer Is withdrawn, residual vinyl chloride monomer removed by a stripping process, and the polymer ts precipitated by the addition of water and alcohol. Polymer is recovered by centrifugation, washed, and vacuum stripped. Wist resin is then ground, dried in a multi-stage dryer and packaged. This unique solvent polymerization precipitation recovery system results In relatively porous particles highly favorable to the efficient removal of residual vinyl chloride monomer, BAKELTTE solvent polymerized resins contain no detectable vinyl chloride monomer when Analyzed bye method sensitive to 0.2 ppm- Non-solvent polymerised resins are produced by a continuous process, in which fresh monomer Is fed continuously to a pressurized autoclove and a mixture of polymer and mcnomor is continuously withdrawn. Other than catalyst and its comer, no other materials are added. The monomer/rasin slurry is discharged into hot water where monomer is boiled off and recovered. Suspended resin Is continuously recovered by centrifugation, dried in a two stage flash drier, and packaged. This low conversion bulk polymerisation process yields a small, vary porous particle. These resins, as produced, contain less than 3 ppm of residual monomer hut due to its small size and porous nature, within three days of manufacture the residual monomer content is reduced to 10% of its original value. Samples of resin as recisved by customers and analyzed by Union Carbide by a method sensitive to 0.2 ppm normally show nodetfcctabls vinyl chloride monomer. Both solution and non-solvent vinyl resins are suitable for use jn rigid and semi-rigid sheets for food packaging applications* the single-service ieity and jam contaitver, which holds approximately 15 g of food and weighs approximately 600 mg, represents the moat severe surface to volume ratio to be expected m a food packaging application. Assuming a residual vinyl chloride monomer content of <0.2 ppm, the 6C0 mg container would- provide less than 0.00012 mg of monomer as a potential migrant to food. Assuming that half of this amount would migrate inward to the food, the monomer concentration in the 15 g of contained food would be leas than 4.0 ppb. BFG21615 k 22566019 To: Mr, Jerome H0 Heckman Page 2. 4) Based on the experimental test results which were submitted in our December, 1975 presentation to the Food and Drug Administration, as well as continuing work, it is established that PVC sheet o the kind currently being produced by Tenneco Chemicals, as well as the sheet that has been produced in the last twelve months, cannot reasonably be expected to cause VCM to become a component of food. We have further assured this on a continuing basis by the regular analysis o production samples of food grade rigid sheet. It should be further noted that it has been established by the Department of Food Science at Futgers University, that when the residual monomer in a rigid vinyl sheet is at 10 parts per billion or less reading on the gas chromatograph, subsequent analysis by G, C. mass fragmentography has been shown to be unable to identify this very minute amount as being vinyl chloride. Our laboratories and manufacturing departments are continuing to work on ways and means o virtual elimination of vinyl chloride while continuing to use more and more sensitive tests. While this objective is currently under way, it now becomes evident as to the impracticality of the endeavor. We are speaking of such very low levels that we can find these levels in materials other than rigid vinyl. We, therefore, conclude that establishment of a level of 30 parts per billion very adequately protects the consumer from ary VCM ingestion by virtue of migration from the sheet, and based on our work and data we would suggest that a regulation be established at the 30 parts per billion level. However, we also feel that there is sufficient safety at the 50 parts per billion level residual monomer in the sheet. I trust the foregoing adequately outlines our progress to date and our position re garding this matter. Should ary further information or details be needed, we shall be glad to supply same. Sincerely, TENNECO CHEMICALS, INC. // ' / Michael F. Saggese Business Director J ^ or 22569021 BFG21617 TABLE I-A Residual VCM Content of Resin Blends Before and After New Blending Process (American Hoechst Corporation Data) Control No. ^ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Residual VCM Level - ppm Before Blending After Blending(^) 37.0 21. 0 13.0 7.7 7.7 6.5 5.7 4.9 4.2 4.0 3.0 2.0 1.0 0. 5 0.25 0. 15 0. 10 2.5 <1.5 41.5 <1.5 0.50 0,34 0. 19 0. 16 0. 11 ND (40.10) ND (40.10) ND (4 0.10) ND (40.10) 0.020 ND 0.01) ND (4 0.01) ND (<0. 01) NOTES; ^ Samples no, 1-4 run by method with LDL of < 1,5 ppm; samples no. 5-13 run with new method with LDL of < 0,10 ppm; samples no, 14-17 run with revised method with LDL of < 0. 01 ppm (2) After high intensity devolatilization blending technique which can be licensed from American Hoechst Corporation if desired. RHjvel 12/10/76 BFG21619 Ol c\ r.0 o N U> (CAD-95-76) 12/15/76 2. Another method of expressing the data is to choose a level of RVCM in the bottle compound, and to calculate the level of vinyl chloride contained in the^bp.tle contents. For a one milligram per kilogram (ppm) pipe compound-,'^for an 8 and 64 oz. bottle, 75F, these tabulations are given in Table I. This distinctly shows the relationship between the size (volume) of the bottle and the migrating VCM levels, that is, the surface to volume ratio effect. Naturally, the smaller the volume of the bottle, the higher the VCM in the contents at a given time and RVCM in the bottle wall. The graph also shows the expected effect of temperature, namely that the higher temperatures produce higher vinyl chloride migration levels. The calculation capabilities are not limited to these sizes, temperatures or residence times. However, the diffusivity data are those published for specific, but typical, bottle compound compositions. Thus, we feel with certainty that these calcula tions accurately represent the expected migration levels from PVC bottles into 50% ethanol-water solutions. CAD/dms cc: J. R. Goots A. P. Metzger F. E. Krause A. L. Schultz E. A. Collins C. A. Daniels SZ0S9 to to BFG21621 (CAD-95-76) 1 2 /1 5 /^J BFG21623 J v b -n ni 3om i3ni aid oz x os 03 N3QZ3ia 3H30H3 M3d*d HHV-Hfl N 3DZJ.3IO DZ-Ovr t)M PREPRINT (Accepted for May 1977 issue of JAOAC) Gas-Liquid Chromatographic Headspace Procedure for Datermination of Vinyl Chloride in Vegetable Oil and Three Food Simulating Solvents. By Gregory W. Diachenko, Charles V. Breder, Margaret Brown, and J. Lawrence Dennison Division of Chemical Technology, Bureau of Foods, Food and Drug Administration, 200 Street, S.W., Washington, D.C. 20204 22370006 BFG21625 -1- In early 1973, alcoholic liquors packaged in polyvinyl chloride (PVC) bottles were found to be contaminated with vinyl chloride (VC) migrating from the plastic (1)- These findings, coupled with the demonstrated car cinogenicity of VC by inhalation (2), led to a great deal of concern about the possible migration of VC into foods. Several investigators subsequently detected VC in other foods such as vegetable oils, edible fats, and vinegars packaged in PVC bottles (2,3,4). These, findings have increased significance in view of research by Maltoni et-.al. (2), indicating the possibility that VC is also carcinogenic by ingestion. PVC producers and converters have recently Improved their manufacturing technology resulting in lower amounts of residual VC monomer in PVC resins and products. Several manufacturers have used new stripping technology to produce resin and finished rigid PVC articles with residual VC levels in the low ppm to low ppb range. Direct solution injection gas chromatographic (GLC) techniques generally do not have sufficient sensitivity to detect the small quantities of VC migrating from these materials. In view of this, it is desirable to have a more sensitive analytical procedure to detect and measure low levels of VC which may migrate from PVC Into foods or food simulating solvents. Building on other GLC-headspace techniques (3,4,5) and the procedure of Breder et.al. (6), our laboratory has refined and developed procedures which permit the quantitative determination of VC at the 1 ppb level in food simulating solvents such as 50Z ethanol, 3Z acetic acid, n-heptane, and vegetable oil. BFG21627 22570008 -3- (d) Gas Chromatograph.-Hewlett-Packard Model 7620A, or equivalent, equipped with temperature programmer and flame ionization detector. Operating conditions: temperatures (C)-detector 260, injection port 200; flows (^-/min)-hydrogen 70, oxygen 480; electrometer setting, 1 or 2X10"^ amps full scale. Detector sensitivity: 0.7 ng VC gave 50% full scale deflection at retention time of 2.5 rain and electrometer setting of 1X10"^ amps full scale. (e) Chromatographic column.-Coiled stainless steel column 5'X2 ram id, packed with 60-80 mesh Chromosorb 104; temperature programmed (C), 95 for 5 min, 95-250 at 30/min, hold for 6 min; helium carrier flow adjusted to give VC retention time of 2.0-3.5 min (ca 50 rol/rain). BFG21629 Preparation of Standards Prepare stock VC solutions of ca 1 and 5 ppm concentrations in each of desired food simulating solvents as previously described (6). Check VC concentration by injecting solution into gas chromatograph and comparing area response with that obtained from equivalent amount of 1.00 ppm VC standard gas. Area measurements should agree within +5%. Prepare headspace standards by injecting calculated volumes of VC stock solution into 35 ml (1 oz) septum sealed bottles containing measured volumes of food simulating solvents. Volumes of sample and standard 35 ml septum sealed bottles should be matched within +0.5 ml. Preparation of Sample and Analysis: Invert solutions to be analyzed for VC several times to Insure solution homogenity. Analyze as described below. Ethanol (50%) and acetic acid (3%)-Thoroughly purge 35 ml bottle with Nitrogen and seal with septum and cap. Using a 10 ml gas & S o o -5- Discussion and Results The high vapor pressure of VC (B.P.-14C) makes it Ideally suited for determination by a headspace technique. The favorable partitioning of VC into the headspace over a sample and the relatively clean matrix in the headspace enables GLC injections of larger percentages of the total VC in samples than could be obtained by direct solution injection. Detection levels by this headspace technique ranged from approximately 0.1 ppb (n ^/ml) in 50% ethanol to 0.5 ppb in corn oil. These detec tion levels are 10 to 50 times lower than those obtainable using a direct solution injection technique for food simulating solvents (6). Equilibration of the headspace/solvent system at 90*C is rapid. This was shown by the vinyl chloride concentration in the headspace remaining constant when analyzed at periods from -30 min to 2 hr. The experimentally determined concentrations of VC in the headspace over spiked solutions were used to calculate partition coefficients. By dividing the ng ^/ml of headspace by the ng ^c/ml remaining in solution (obtained by difference), the following partition coefficients at 90 were calculated; 0.5 for 50% ethanol, 1 for 3Z acetic acid, 0.09-for heptane, and *0.09 for.corn oil. Food simulating solvents spiked at concentrations ranging from 0.1 to 100 ppb with VC stock solutions in the same solvent gave linear calibration curves. The reproducibility in preparing and checking VC standards by this headspace technique was determined for corn oil and 50% ethanol. Three sets of standards, each consisting of 3 or 4 units, were prepared in each solvent , and quantitated using previously prepared BFG21631 ZXQOLSZZ -7- vere analyzed as previously described. Previous work in our laboratory showed that equilibrium was reached for similar PVC materials within those two :tiroe periods. Results of the headspace analyses of these solutions, shown in Table 2*, gave average concentrations of 1.6 and 2.4 ppb VC in 50% ethanol and 2.5 and 4.8 ppb in corn oil for lots 1 and 2 respectively. These concen trations represent total migrations of approximately one third and one half of the residual VC from the PVC sheets into-50% ethanol and corn oil respectively. Consistent losses of VC were obtained for the four 1 ppb standard solutions in each solvent subjected to the same conditions as the sample migration solutions. Average losses of VC were 64% (60, 63, 63, 68) from 50% ethanol standards and 37% (35, 35, 37, 41) from corn oil standards. These losses are suspected to be due to diffusion of VC through the Teflon lined septa used to seal the migration bottles. If one assumes that similar losses occurred from the PVC migration solutions, then the migration of much larger amounts of VC actually took place. The measured values might therefore be corrected for losses by applying correction factors derived from the standard solutions. These loss-corrected values would give percentages of maximum VC migration into corn oil and 50% ethanol ranging from about 2/3 to near complete migration. Although the GLC chromatograms of these solutions, shown in Figs. 5 and 6, are less than ideal, the data in Table 2 indicate they were adequate for reasonably reproducible quantitation. The chromatogram in Fig. 7, obtained in a previous VC migration study in our l.ihorutoiy, illustrates the type of chromatogram obtained for migration Mentions containing higher concent, rut ions of VC and smaller amounts of extraneous chromatographable compounds. The presence and approximate amounts of VC in the 50% ethanol migration solutions were confirmed by GLC-MS. BGH21633 frTOGLSZZ Acknowledgment The authors thank Virgil Warren, Division of Chemical Technology} Food and Drug Administration, Washington, D.C. for his GLC-MS confirmation of VC in food simulating solvents* BFG21635 9T0GS?; T a b le 2. VC M ig r a tio n fro m PVC Sheets in t o 507. E th a n o l and Corn O il 22570018 FIG. 1 - Gas-liquid chromatograms of 3.0 ml headspace aliquots from over 50% ethanol solutions: A, Blank; B, 0.5 ppb VC standard solution, ca 0.5 ng VC injected; 23UO~12 afs, *: i ,< * FIG. 2 - Gas-liquid chromatogram of 3.0 ml headspace aliquot from over 50% ethanol solution; 5.8 ppb. VC standard^ ca 6.2 ng VC injected; 8X10"^2 afs. FIG 3 - Gas-liquid chromatograms of',4.0 ml headspace aliquots from over corn oil solutions: A, Blank; .B, 1.0 ppb VQ standard solution; ca 0.35 ng VC injected; 2Xi0"*2 afs.; FIG. 4 - Gas-liquid chromatogram of 4.0 ml headspace ali'qiiot from over corn oil; 5.7 ppb VC standard solution; ca 1.9 ng VC injected; 2X10~12 afs. FIG. 5 - Gas-liquid chromatogram of 3.0 ml headspace aliquot from over 50% ethanol migration solution; PVC lot lb, representing ca 1.4 ng VC; 2X10"12 afs. FIG. 6 - Gas-liquid chromatogram of 4.0 ml headspace aliquot from over corn oil migration solution; PVC lot la, representing ca 1.1 ng VC; 2X10"12 afs. 22570020 FIG. 7 - Gas-liquid chromatogram of 3,0 ml headspace aliquot from over 50% ethanol migratiqn solution; PVC sheet contained 0.54 ppm residual VC monomer before migration; concentration in solution ca 10.7 ppb; 16X10~12 afs. " BFG21639 n 0 MIN RESPONSE vzm uszz BFG21643 RESPONSE FIG. 6 BFG21645 22570026 JObU'JI r.. KELLER JEnOME 11. 11ECKMAN' CHARLES M. MEEHAN WILLIAM H BORGHESANI.JR RQUERT R. TIERS'AN WAYNE V BLACK DAVID L HILL MARTIN W. BEHCOVICI FETCH M- NEMKOV JOSEPH E- HADLEY. JR. CAROLE C. HARRIS PETER THOMAS SMITH MICHAEL F. MORHONE LARRY S. SOLOMON law orricES Kelleh akd Heckman U30 IT!1 STREET. M. W. SUITE lOOO Washington, d.c. 20030 December 13, 1976 Mr, Richard J. Ronk Director Division of Food and Color Additives Food and Drug Administration 200 HC" Street, S.W. Washington, D.C, 20204 TELEPHONE aoa -437-noo CABLE ADDRESS "KELMaN" WRITERS DIRECT DIAL NIMUER 457-1110 Re: PVC; 90-Week Interim Report of Two-Year Oral Toxicity Study Conducted by CIVO-TNO Dear Mr. Ronk Following up on my telephone report to you last week, we are enclosing herewith a copy of the 90-week interim report concerning the chronic oral toxicity study of vinyl chloride now underway at the Central Institute for Nutrition and Food Research (CIV0-TN0) in Holland. Although we received this copy on a confidential basis through informal channels, we felt it necessary and de sirable to advise you about it immediately. Despite the fact that we can see no immediate relevance of a report on a feeding study to the interest of the Environmental Protection Agency or the National Institute of Occupational Safety and Health insofar as their regulatory responsibilities with respect to vinyl chloride monomer are concerned, rather than take any risk that a reporting omission might somehow be misconstrued as constituting some attempt to withhold pertinent data, you will note that copies of this letter and the enclosure have today been sent to the Administrator of the Environ mental Protection Agency, Mr. Russell Train,and the Director of the National Institute of Occupational Safety and Health, Dr. John Finklea. 9ZOOA9ZZ BFG21647 wew.iiAAL INSTITUUT VOOfl VOGOiNGSONDSftZOSK Utrchtswsg 48 Zeltt CENTRAL INSTITUTE FOR NUTRITION AND FOOD RESEARCH INJTTTUT CENTRAL OE LA NUTRITION ET Of L*AUMENTATION ZENTRAUNSTTFUT FOR ERNAHRUNCSFOR5CHUNG Chronic (r)_ora|_toxicit^_stud^_with^vin^l^cMoride_in_>rats (Interim information III) 1. Conduct of the study Sec Interim Information of 15th December 1975, given in Enclosure 1. Vinyl chloride-treatment by gavage was discontinued in week 84 be cause the condition of the rats given 300 mg vinylchloride/kg body weight was rapidly declining and mortality in this group was high (55 X). More over, most of !the rats of this group that died or were killed in extremis showed extensive liver damage (haemorrhages, focal necrosis), liver tumours or tumours at other sites. 2. Results available after a test period of 90 weeks 2.1. fody^weights Not affected. 2.2. Food^ronsumgtion Not distinctly affected. 2.1.Mortality VCM (mg/kg body wt) Total number of deaths*) at week 90 males females 0 1 3 9 300 (by gavage) 2 3 6 20 46 5 4 16 40 44 ) Initial number of rats:: 60 animal s/sex/group 2.4. Haematology Slightly decreased haemoglobin values in males of the 9 mg/kg group after 78 weeks. 2.5. Hlond biochemistry BFG21649 No indications of an adverse effect of vinyl chloride. .. \ D200LQZZ CENTRAAL INST1TUUT VOOR VOEOINCSONDGRZOEK m 9 mg/kg group: 300 mg/kg group: l.ivcr nodules (tumours ?) Pancreas tumour Thymic tumour Lymphoreticular tumour Mammary tumour Pituitary tumour Subcutaneous tumour Liver nodules/haemorrhagic cysts (tumours ??) Pancreas tumour Pituitary tumour Lymphoreticular tumour Pulmonary nodules/haemorrhagic lesions (cumou Nodules aer~duct area Abdominal tumour Mammary tumour Kidney mass (tumour ?) 3. Preliminary conclusion From the results so far available it appears that the Administration vinyl chloride monomer to rats at levels of 9 (as PVC-powder in diet) and 300 mg/kg body weight (in soyaoil by gavage) resulted in obvious liver chv which in many cases seem to be neoplastic in nature. Moreover, there are i dicaeions that similar lesions also occur at the 3 mg/kg level, but so far are absent at the I mg/kg level. 22570032 BFG21651 The Society of the Plastics Industry, Inc. 355 Lexington Avenue New York. New York 10017 (212)573 9400 a February 23, 1976 Mr. Don R. Goodwin Emission Standards & Engineering Division Environmental Protection Agency Research Triangle Park, North Carolina 27711 Dear Mr. Goodwin: Enclosed herewith, on behalf of the Vinyl Chloride and Polyvinyl Chloride Producers Group of The Society of the Plastics Industry, Inc. (SPI), are comments on the Proposed Standard for Vinyl Chloride published in 40 Fed. Reg. 59532 on December 4, 1975. As we testified at the ' public hearing on the Proposed Standard on February 3, 1976, we commend the Agency for adopting an approach which allowed an open exchange of information between the techni cal people of the Agency, the industry, and other interested parties. We also support the Agency's utilization of a Quantitative Risk Assessment Analysis in determining the ^ appropriate control technology, although as I also testified we believe the Agency did not weigh finely enough the costs against the benefits likely to be achieved in protecting public health. Although there is much with which we agree, there are also a number of individual statements and assertions in the Proposed Standard and supporting documents which we consider erroneous. The two documents I an enclosing to spell out our differences are as follows: 1. Comments on the Standard and its supporting documents prepared by the Technical Subcommittee of the Vinyl Chloride and Polyvinyl Chloride Producers Group; and w U1 2. Comments by the Health Subcommittee of the Vinyl Chloride and Polyvinyl Chloride Producers Group. BFG21652 2 We urge the Agency to take these comments into account before the final Standard is promulgated; to correct the errors noted in the attached statements; and to weigh more carefully some of the control requirements whose costs we believe far exceed any potential benefit. There were four specific matters raised when the EPA panel was questioning me during the public hearing on February 3. To complete the record, the following comments are respectfully submitted. 1. We were asked by Mr. Fanner (Transcript of Proceedings, hereinafter: "Tr." page 68) to comment on "...the uses of PVC and then the possible substitutes for PVC in all these end use applications." EPA's own reports on this question have already covered the widespread and diversified uses of PVC in the medical, construction, applicance, electrical, transportation, packaging and many other industries. As to the concept itself, we respectfully submit that substitutability per se is not a valid issue. First, our legal research confirms Dr. Kuzmack's observation that there is no authority in the Clean Air Act for EPA to ban selected uses of PVC (Tr. 52). Secondly, under the proposed Standard, the level of VCM emissions would be reduced by 95 percent; at such low levels there is no clearly established health risk and, hence, no danger to the public health which must be further obviated. Finally, even given exhaustive investigation of questions interrelating safety, costs, availability of raw materials, production and technological capability, capital investment and other related issues, as a practical matter substituta bility remains an imponderable. Thus, under the circum stances of this case, we do not believe it would be fruitful for EPA to explore these issues further. 2. In response to Dr. Knelson's question (Tr. about dialysis machines, the principal usage of PVC is in the extensive tubing which circulates the patient's blood through the machine. 71) 2572004 BFG21653 3 3. To Dr. Marcus (Tr. 76) I promised to clarify the dates "... when Maltoni first informed industry and industry first informed EPA ..." about the preliminary results of Maltoni*s inhalation studies with rats, in-? dustry was first informed at a technical symposium in Bologna, Italy held from April 9 to April 12, 1973. This can be confirmed by Dr. Saffiotti of NIOSH who was in the audience. The USA VCM/PVC industry representatives did not report to EPA; this information was reported to NIOSH as required by regulations promulgated under 20 of the Occupational Safety and Health Act. 4. Dr. Marcus also asked (Tr. 82) for a response on SPI's position regarding toxic substances legislation. We generally support the moderate, responsible approach of the McCollister Bill (H.R. 7664). Finally, we want to re-emphasize our belief that EPA leadership is needed to develop a vinyl chloride health research program under the joint sponsorship of expert representatives from government, industry, labor and other interested groups. A further reason for this approach was provided on the very day of the EPA hearing on the Vinyl Chloride Standard by way of the sensationalized disclosure of a new study relating to whether vinyl chloride is a potential mutagen. The report in question was not available to any of the industry representatives who testi fied at the hearings, nor was it made a part of the record. Instead, it was released directly to the press by the Health Research Group, under circumstances which did not permit rational discussion of the information or informed industry comment. We comment in more detail on this subject in the attached Comihents. - Thus, while the vinyl chloride and polyvinyl chloride producers obviously believe that the concerns raised in the study should be explored further, they do not believe that a press release provides the best means to evaluate the risks involved nor the further precau tions, if any, which should be observed. The producers have expressed their willingness to provide their full share of the costs of jointly-sponsored research in the best interests of the general public and 22572005 BFG21654 -4- of the workers directly involved. The questions involved are too important to be dealt with by acrimonious and sensationalized debate. Unfortunately, recent history indicates that until the government exercises leadership in bringing all interested parties together, we are all too likely to see more time and money wasted on disputed research and incomplete or ever, biased discussion of scientific evidence. In closing our written submission to 5?A, we again want to express appreciation for the careful manner in which E?A is handling the development of this Standard. V?e would be pleased to respond to any further questions you might have or which might arise as a result ci the comments of others. Respectfully submitted. Ralph L. Harcinc, Jr. President BFG21655 22572006 I* U.S. DEPARTMENT OF LABOR Occupational Safety an;J Health Afsisittration WASHINGTON*. D.C. ttZM FEB 1 S 197$ Howard L. Watson, Administrative Manager Manufacturing Air Products and Chemicals, Inc. Chemicals Group Five Executive Mall Swedesford Road Wayne, Pennsylvania 19087 Dear Mr. Watson: In response to your letter of September 2, 1975, this Office has reviewed your interoffice memorandum dated April 1, 1975 to R. Fleming at Valle;: Forge. The subject of the memorandum is: VCK KZ1ICAL EXAMINATIONS Persons required to be covered by annual or semi-annual physicals. In regard to the program proposed by you, Mr, John Barr, and Mr. John Novak for application tc the occasional visitor from outside your plants, the Occupational Safety and Health Administration (OSKA) agrees you need not provide medical surveillance as mandated by CFR 1910.101700 if the duration of the visitor's exposure is so limited that he will not be exposed to vinyl chloride in excess of the action level. However, a few words of caution are in order: (1) Stationary monitors may not provide a very accurate indication of actual exposures. (2) Should a visitor remain in the regulated area for a duration of ^5 minutes, sotting the alarm at 5 ppm leaves very`little margin for error. (3) Exposure in excess of 5 ppm. averaged over a 15 minute time span is a violation of the permissible exposure iir.it. to to 3 to o BFG21656 2 The rest of your proposal appears to be satisfactory. Thank you for taking the initiative in this important matter. Sincerely, Barry J. White Associate Assistant Secretary for Regional Programs BFG21657 JOSEPH E. SELLER JEROME H. HECKMAN CHAHLES M. MEEHAN WILLIAM B.BOBGHESANI. JH BOSEBT B.TIEHNAN WAYNE V. BLACK DAVID L. BILL MARTIN W. BEBCOVICX PBTEfi M- NEMKOV JOSEPH E. HADLEY CAROLE C. HARRIS PETER THOMAS SMITH MICHAEL T. MORRONE URBY S- SOLOMON LAW OFFICES Kelleh and Heckman UOO 17TB STREET, N. w. SUITE lOOO WASHINGTON, D. C. 80036 January 19, 1976 Mr. Ellis Rottman, Editor FDA Consumer HFI-20 Food and Drug Administration 5600 Fishers Lane Rockville, Maryland 20852 1 TELEPHONE aoo aoe-szoo CABLE ADDRESS"KELMAN" Dear Mr. Rottman: While we certainly have no quarrel with your printing an article which recites the chronology and rationale for FDA action on polyvinyl chloride, we strongly question the timing and content of "Polyvinyl Chloride: Why FDA Acted" which appeared in the December/ January issue of FDA Consumer. Should the official magazine of FDA publish an article on- the vinyl chloride regulations prior to the completion of that process? We think not. In fact, we believe that the published article could suggest that FDA has prejudged the case and decided to ban several important types of polyvinyl chloride packaging prior to reviewing new industry data which shows that these products present no hazard to public health. We also find it difficult to understand why author Wayne Pines' otherwise excellent chronology fails to mention: 1) Commissioner Schmidt's finding that polyvinyl chloride packaging presents no immediate hazard. 2) The December 19 filing of 900 pages of Comment by The Society of the Plastics Industry documenting in dustry's contention that there is no reasonable expectation of vinyl chloride monomer migration from any PVC product. sooe^ssz BFG21658