1 U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES PUBLIC HEALTH SERVICE FOOD AND DRUG ADMINISTRATION CENTER FOR BIOLOGICS EVALUATION AND RESEARCH INTERNATIONAL ASSOCIATION FOR BIOLOGICALS NATIONAL INSTITUTE OF ALLERGY AND INFECTIOUS DISEASES NATIONAL VACCINE PROGRAM OFFICE WORLD HEALTH ORGANIZATION EVOLVING SCIENTIFIC AND REGULATORY PERSPECTIVES ON CELL SUBSTRATES FOR VACCINE DEVELOPMENT WORKSHOP TUESDAY, SEPTEMBER 7, 1999 The workshop took place in the Plaza Ballroom, DoubleTree Hotel, 1750 Rockville Pike, Rockville, MD 20852 at 7:45 p.m., William Egan, Ph.D., Chair, presiding. PRESENT: WILLIAM EGAN, Ph.D., Chair ELWYN GRIFFITHS, Ph.D., D.Sc., Co-Chair REGINA RABINOVICH, M.D., Co-Chair LEONARD HAYFLICK, Ph.D., Speaker ANDREW LEWIS, Jr., M.D., Speaker PHILIP KRAUSE, M.D., Speaker 2 AGENDA Welcoming Remarks Regina Rabinovich, M.D. . . . . . . . . . . . . 3 Institute of Allergy and Infectious Diseases, NIH Elwyn Griffiths, Ph.D., D.Sc. . . . . . . . . . 5 WHO SESSION I - Evolution of Cell Substrates Used in the Manufacture of Biologicals Review of workshop purpose and goals William Egan, Ph.D. . . . . . . . . . . . . . . 7 CBER/FDA Evolution of cell substrates used for the production of human biologicals Leonard Hayflick, Ph.D. . . . . . . . . . . . . 11 University of California, San Francisco Major issues associated with cell substrates Elwyn Griffiths, Ph.D., D.Sc. . . . . . . . . . 26 WHO Developing an approach to evaluate the use of neoplastic cells as vaccine substrates Andrew Lewis, Jr., M.D. . . . . . . . . . . . . 58 CBER/FDA Conceptual and experimental approaches to address product safety issues raised by novel cell substrates Philip Krause, M.D. . . . . . . . . . . . . . . 73 CBER/FDA 3 1 P-R-O-C-E-E-D-I-N-G-S 2 (7:54 p.m.) 3 DR. RABINOVICH: Good evening. I would 4 like to ask those who would like to participate in 5 this evening's session to please have a seat so we can 6 get started. My name is Regina Rabinovich. I am here 7 from the Institute of Allergy and Infectious Diseases 8 from the NIH in the U.S., and on behalf of my co- 9 sponsoring agencies and colleagues, the Center for 10 Biologics Evaluation and Research, the International 11 Association for Biologicals, the National Vaccine 12 Program Office, which includes Marty Myers, who is 13 supposed to be a co-chair for this session and is 14 currently either grounded or wandering around Newark, 15 New Jersey, and the World Health Organization, I 16 welcome you to the meeting on Evolving Scientific and 17 Regulatory Perspectives on Cell Substrates for Vaccine 18 Development. 19 I have a couple of housekeeping comments. 20 I won't bother to tell you where the bathrooms are. I 21 am sure you can find those. There are two documents 22 that I think would be very useful, if you haven't 23 noticed them already, because they are key to the 24 workshop. The first is the document prepared by our 25 colleagues from the Food and Drug Administration, 4 1 which is under your tab entitled, Draft CBER document. 2 Because I think that the creation and comment upon 3 this is central to the scope of this meeting, as will 4 be elucidated further by Dr. William Egan. 5 Because the sponsoring groups as well as 6 the FDA particularly are so interested in the concepts 7 within the document and the discussion that will take 8 place, this meeting is being taped. And for that 9 reason, we must ask that speakers as well as those 10 that are asking questions or getting answers speak 11 into the microphones, and that we will work with the 12 moderators to make sure that everyone that wants to 13 ask a question, regardless of where they are in the 14 room, have a microphone available to them because it 15 is key that it be part of the written record. 16 The steering committee would like to thank 17 a couple of individuals that have worked very hard to 18 make the meeting possible. Two of our colleagues at 19 FDA, Dr. Keith Peden and Dr. Phil Krause, and Dr. John 20 Daugherty at NIAID. 21 In thinking about eight months ago about 22 this workshop, the scope and what it should cover, we 23 never in our wildest dreams considered that in 24 September of 1999 we would have gone through a summer 25 where there would be no question about how to spell 5 1 thimerosal, and that early information coming out of 2 sort of a group analysis of all clinical trial data 3 for the rotavirus vaccine would give us the first 4 inkling to the potential connection to 5 intussusception. Indeed, the issues regarding this 6 topic, vaccine safety, which never makes sense to 7 anyone, what we are talking about in terms of vaccine 8 safety, are elucidated by the kinds of thoughts and 9 strategies that will evolve from this meeting. And if 10 ever there was a connection between basic research and 11 what we do with human subjects as an outgrowth of your 12 deliberations, I think this is it. 13 So on behalf of my colleagues at the 14 National Institute of Allergy and Infectious Diseases, 15 I thank you for your participation. I would like to 16 go ahead and introduce Dr. Elwyn Griffiths from the 17 World Health Organization, who is my co-chair in my 18 role as Dr. Marty Myers, and will introduce the first 19 speaker. Thank you. 20 DR. GRIFFITHS: Can you hear? Oh, yes, 21 okay. (Speaking in Welsh.) It is the only 22 opportunity I have ever had to introduce a meeting 23 such as this in Welsh. I know there is one colleague 24 of mine, Merfyn Williams, who used to be at FDA, and 25 he is in the audience. So at least he understood that. 6 1 It is really an honor and a pleasure for 2 me to co-chair this session, the first session of this 3 important meeting with Regina here. Let me also say 4 that the World Health Organization is very pleased 5 indeed to be a co-sponsor of this workshop. Because 6 I think it considers the subject to be vitally 7 important from a global perspective, and we take that 8 view of course, for future development and production 9 of vaccines in general. 10 I see this evening's activities really as 11 a sort of an hors d'oeuvre, I guess you might say, or 12 an appetizer to the entrees that come during the next 13 two days or so. So it is an overview of where we are 14 going and where we have been. I think perhaps we 15 should, because of the time scale -- I think we should 16 move forward with the program, taking particular pity 17 on any colleagues who have just flown in from Europe, 18 because they would be suffering from some jet lag or 19 biological clock problem at the moment. So I think we 20 should move forward fairly quickly. 21 So it is a pleasure then really to call 22 upon Dr. Bill Egan, who is now the Acting Director of 23 the Office of Vaccines at the Center for Biologics 24 Research and Evaluation, CBER, who will actually give 25 us a review of the works or purpose and the goals of 7 1 this. So he will set the scene for where we are going 2 to go during the next few days. So, Bill? 3 DR. EGAN: Thank you, Elwyn. Can I have 4 the first slide, please? I too would like to thank 5 everybody who has been able to come to the meeting and 6 everybody who helped in the organization of the 7 meeting and all of the participants in the meeting. 8 This is a workshop and everyone will be working 9 together to try and address some very interesting, 10 some practical, and some very difficult questions. 11 Every now and then I think we need to 12 remind ourselves that viruses can propagate only in 13 live cells, and this of course holds true for whole 14 viral vaccines. They can only be produced in cells. 15 The choice of suitable cell substrates for the 16 manufacture of viral vaccines has over the years 17 engendered considerable discussion. The primary focus 18 in these discussions on cell substrates have been 19 safety, in particular the potential safety concerns 20 from residual cellular DNA and from adventitious viral 21 agents. 22 As history has shown, the need for concern 23 about cell substrate issues was real. We have only to 24 think back to the finding of SV40 in polio virus 25 vaccines to realize the extent of the risk that any 8 1 cell substrate may pose, and there is still great need 2 for concern. 3 Early discussions on cell substrates 4 allowed only for the use of "normal" cells, that is, 5 cell populations that were derived from normal tissues 6 and that had undergone no subculture. And right now, 7 I am referring primarily to the 1954 decision of the 8 Armed Forces Epidemiological Board. 9 I can't read this -- it is hard to read 10 from this perspective back here. The use of only 11 normal cells has prevailed for many years. For 12 example, in 1971, the regulations of the Public Health 13 Service for the production of OPV stated that only 14 primary monkey kidney tissue cultures may be used in 15 the manufacture of polio virus vaccines. There were 16 similar regulations in place for the production of 17 inactivated polio virus vaccine, live attenuated and 18 inactivated measles vaccines, mumps vaccine, rubella 19 vaccine, and the adenovirus vaccine. 20 This was taken from the regulations in 21 1971, and it was in 1972 that the Federal Register 22 notice was placed that contemplated the use of other 23 cell lines other than these primary animal tissue 24 cells for use in viral vaccines. 25 Now our thinking and our practices with 9 1 regard to cell substrates has evolved over time. At 2 present, for example, in addition to primary animal 3 cell culture, diploid cell lines, both of human and 4 animal origin, are in routine use, as, for example, 5 the mumps, rubella, IPV vaccines. And in addition to 6 being manufactured in the human diploid cell, live IPV 7 is also produced in viro cells, a continuous cell 8 line. 9 We now wish to consider additional cell 10 substrates or additional cell substrate uses. In 11 particular, we would like to consider the use of 12 continuous cell lines for live attenuated viral 13 vaccines, and the use of other neoplastic cell lines 14 for both live attenuated and inactivated viral 15 vaccines. Indeed, the question that the Armed Forces 16 Epidemiology Board met to consider 45 years ago, 17 namely the use of tumor cells, specifically HeLa 18 cells, as a cell substrate is a question that we are 19 again raising. 20 The goals of the workshop are simple. 21 They are two. The first is to identify the various 22 concerns that are associated with using neoplastic 23 cell lines, to catalog them as completely as possible. 24 The second is to determine approaches and to identify 25 approaches to determining the level of the risk that 10 1 are associated with these concerns. In the decisions 2 that we will make with regard to the use of neoplastic 3 cells as potential substrates for viral vaccines, we 4 need to be guided by data; either data that already 5 exists in the literature or the data that we need to 6 generate. We should not be guided by intuition or by 7 opinion. If I may quote Dr. Hilleman from a 8 discussion about the use of tumor cell lines from a 9 meeting that was held in 1968, he said, "If the debate 10 has taken on an ecclesiastical or philosophical 11 overtone, it should not be unexpected, since it is an 12 area in which the definitive information for making a 13 judgment is clearly missing. In fact, I feel that the 14 greatest problem of all may lie in ascertaining 15 whether there is a problem in the first place." 16 The workshop in which we will be 17 participating over the next four days is designed to 18 see how far we have progressed and what we still need 19 to do. Thank you. 20 DR. GRIFFITHS: Thank you very much, Bill. 21 I think we have been given the task, then, for the 22 next few days, that is, to identify these concerns and 23 issues associated specifically with neoplastic cells 24 for production of vaccines. We are focusing on 25 vaccines. And to identify some approaches in 11 1 determining this risk associated with their use and 2 really how to move forward on these issues. I think 3 those are going to be our main concerns over the next 4 four days. 5 So we now move on to the second talk and 6 we are really looking at historical perspective to a 7 certain extent here, the evolution of cell substrates 8 used for the production of human biologicals. Because 9 as Bill already mentioned, they have been in use for 10 a long time and there has been a tremendous amount of 11 discussion over the years. And who better to give us 12 this talk and to take us through from the beginning of 13 the evolution of these cell substrates than someone 14 who has been involved right from the very beginning, 15 and the name is synonymous really with diploid cells, 16 Dr. Hayflick. So would he come forward and give his 17 talk on the evolution of cell substrates for the 18 production of human biologicals. Dr. Hayflick? 19 DR. HAYFLICK: Although the first cell 20 culture is generally attributable to Ross Harrison in 21 1907, it was not until 1949 that contrary to the then 22 existing dogma, which held that the polio virus could 23 replicate only in neural cells, that Enders, Weller 24 and Robbins showed that the virus could replicate in 25 cultured human fibroblasts. This set the stage in the 12 1 late 1950's for the development by Salk and his 2 colleagues of an inactivated polio vaccine prepared in 3 primary monkey kidney cells. 4 I should define primary cells, because 5 neophytes are using it to mean something entirely 6 different and are oblivious to the fact that its 7 original meaning still has a legal FDA definition. A 8 primary culture is simply a population of cells 9 derived from impact tissue that has not undergone any 10 subcultivations. After a few years of use of primary 11 monkey kidney cells, it became apparent that these 12 cells contained many unwanted viruses, some of which 13 were lethal to humans. At about this time in 1961, we 14 described the development of normal human diploid cell 15 strains that were free of contaminating viruses, and 16 we suggested that they may be useful for the 17 preparation of human virus vaccines. It took about 10 18 years before our suggestion became generally accepted 19 in the scientific community. 20 Normal human diploid cell strains have 21 several interesting properties. If I can have the 22 first slide? The properties that we described are 23 listed here. Very briefly, point number one, if 24 derived from human embryos, they undergo about 50 25 population doublings. Contrary to the belief that 13 1 this is a serious limitation, the limit is actually 2 about 50 million metric tons after 50 doublings. 3 Human diploid cells undergo a number of population 4 doublings inversely proportional to donor age. This 5 suggested to us that the finite replicative capacity 6 of cultured normal cells is an expression of aging at 7 the cell level. This notion received considerable 8 experimental support in subsequent years and is now 9 the basis for the field of cell senescence or 10 cytogerontology. 11 We also found, of course, that the cells 12 derived from this tissue had the diploid karyotype and 13 are incapable of replication in suspension culture. 14 The next point, number four, human cell strains will 15 not produce tumors when inoculated into the hamster 16 cheek pouch or even when we directly introduced them 17 into terminal human cancer patients, something that 18 was relatively easy to do in the 1960's. 19 Human diploid cell strains can be 20 cryogenically preserved. When, for example, WI38, 21 which was the first highly characterized human diploid 22 cell strain which we developed in 1962, is preserved 23 at a particular doubling level and then reconstituted, 24 the number of doublings remaining is equivalent to 50 25 minus the number of doublings spent prior to 14 1 preservation. The cells have an extraordinary memory 2 and remember at what doubling level they are 3 preserved, even after 37 years of storage in liquid 4 nitrogen. WI38 has been cryogenically stored longer 5 than any other normal human or animal cell population. 6 We also reported that human diploid cell strains had 7 the broadest virus spectrum of any cell population 8 known, and even heretofore unknown common cold viruses 9 were isolated with these cells at that time. 10 As a result of this characterization, we 11 suggested that cell populations in culture could be 12 classified into three groups: primary cultures, which 13 I have already described: cell strains, which are 14 populations that have a finite capacity to replicate 15 and don't produce tumors when inoculated into 16 experimental animals, have the karyology of the tissue 17 of origin and are anchorage-dependent; and the third 18 category, cell lines, which are populations consisting 19 of immortal cells that may produce tumors when 20 inoculated into lab animals, don't have the karyology 21 of the tissue of origin, and are usually anchorage 22 independent. 23 Our description of these three fundamental 24 cell classes in which we define normal cell strains to 25 be mortal overturned a dogma that existed from the 15 1 turn of the century. The dogma was that all cells 2 placed into culture have the capacity to replicate 3 indefinitely. And if they do die, you simply do not 4 have the proper culture conditions. We upset this 60- 5 year-old dogma by proving that normal cells do have a 6 finite capacity for replication and showed further 7 that normal cells have an intracellular counting 8 mechanism and that the replicometer is located in the 9 nucleus. Without our having proven that there exists 10 mortal normal cells, then the field of cancer research 11 rooted in the concept of immortalization of normal 12 cells clearly could not have developed. 13 Our efforts to define these classes of 14 cells were not universally accepted, and 15 terminological chaos has reigned for the last 40 16 years. A good recent example can be found in the very 17 first paragraph of the major document that you all 18 received in the package of materials for this 19 workshop. It is titled, "A Defined Risks Approach to 20 the Regulatory Assessment of the Use of Neoplastic 21 Cells as Substrates for Viral Vaccine Manufacture." 22 Reference is made to a 1954 U.S. Armed Forces 23 Epidemiological Board recommendation that "normal 24 cells, rather than cell lines established from human 25 tumors be used for the development of adenovirus 16 1 vaccines." By normal cells, this Board meant immortal 2 cell lines as I have just defined them but derived 3 from normal tissue. The cells lines referred to by 4 the Board are not normal by any criterion. In fact, 5 the normal human diploid number was not known in 1954, 6 and no normal cell had yet to be described. 7 This confusion led Maurice Hilleman to 8 prepare a live adenovirus vaccine in the Henley 9 intestine cell line, which was thought to have been 10 derived from normal human tissue, but was in fact a 11 grossly abnormal cell line later identified as HeLa. 12 The six human recipients of this vaccine fortunately 13 did not suffer any ill effects after 25 years of 14 follow-up. It was not until 1961 that we showed how 15 to make normal human diploid cell strains that normal 16 cells became available. 17 The advantages of these cells are shown on 18 the next slide. In 1961, we suggested that these 19 normal diploid cells have definite advantages over 20 primary cell populations. I won't go into detail on 21 each of these items, but simply give you some details 22 on two of them, the first and the last actually. In 23 respect to latent viruses, primary adult monkey kidney 24 cells, as we all know now, harbor quite a number of 25 serologically distinct latent viruses, several of 17 1 which were demonstrably lethal for humans. The B 2 virus and the Marburg agent both were lethal for 23 3 people after accidental percutaneous inoculation of 4 handlers of monkeys and their primary kidney cultures. 5 SV40, as was mentioned earlier, known to produce 6 tumors in rodents and with the unknown potential to do 7 so with humans was also an early contaminant of 8 primary monkey kidney cells. SV40 was also capable of 9 transforming normal human cell strains into cancer 10 cell lines in vitro. SV40 was in fact inadvertently 11 administered to several million recipients of early 12 lots of inactivated and attenuated polio virus 13 vaccines. Latent viruses have never been found in 14 WI38. To this day, no evidence exists for an 15 endogenous human oncogenic agent, latent virus, 16 transforming principle, slow virus or endogenous 17 intact retro in any human diploid cell derived human 18 virus vaccine. 19 I won't discuss cost of cell procurement, 20 available cells, cell storage. These are all fairly 21 obvious to most of us by now. But the last point I 22 think is worth emphasizing, and that is that the 23 essential point that we tried to make was that a 24 diploid cell strain was more attractive than a primary 25 cell population in its ability to be thoroughly tested 18 1 before use. In a word, standardization. Thus, the 2 concept of cell standardization was first introduced. 3 Once standardization is achieved, cryogenic storage 4 permits the acceptable strain to be held indefinitely 5 for subsequent use. It is not possible, practical or 6 economic for primary cells to be similarly handled. 7 The argument holds equally well now for 8 cell lines that all of you now take as perfectly 9 obvious. But it was not so obvious from 1962 to 1972. 10 In the late 1950's and early 1960's, cell lines were 11 generally believed to be a forbidden candidate for the 12 production of human biologicals because they shared 13 properties with cancer cells. Our suggestion that 14 human diploid cells could provide a safe and superior 15 substrate for the production of human virus vaccines 16 was met with considerable resistance for the next 17 decade. The major reasons for this resistance were 18 the belief that hypothetical cancer viruses might be 19 present and that spontaneous transformation might 20 occur resulting in the presence of human cancer cells. 21 Latent viruses were such a problem with primary monkey 22 kidney cells that a worldwide moratorium on the 23 licensing of all polio virus vaccines was called in 24 1967 because of death and illnesses that occurred in 25 monkey kidney workers and vaccine manufacturing 19 1 facilities in Germany and in Yugoslavia. The 2 responsible virus, of course, was the Marburg agent. 3 The arguments against the use of human 4 diploid cells for the preparation of human virus 5 vaccines were used as the basis for the philosophy of 6 the Division of Biologics standards, the precursor of 7 what is today CBER, that "the devil you know is better 8 than the devil you don't know." 9 In 1961, together with Dr. Stanley 10 Plotkin, we prepared the first human biological 11 produced in a human diploid cell. It was an oral 12 polio vaccine which we showed to be both safe and 13 efficacious. By 1963, 7,000 people received a similar 14 vaccine produced in WI38 with no safety problems. 15 Many of you use the concept of the master working cell 16 bank in the production of biologicals, but few know of 17 its origins. In 1963, I suggested the concept of the 18 two-tiered system of master and working cell banks for 19 WI38, although these terms were not used. The terms 20 I used were master and working cell seeds. I based 21 this reasoning and the terms on the identical way in 22 which virus seeds were then utilized. The first 23 publication of these procedures was made at the 24 landmark Opatia, Yugoslavia conference in 1963. 25 By the 1980's, when heteroploid and mortal 20 1 cell lines like CHO lines were used for the 2 manufacture of human biologicals, the procedures for 3 characterizing and handling WI38 were lifted over for 4 use with cell lines with very few changes even to this 5 day. In the mid-1960's, several WI38 users expressed 6 difficulties growing the cells that were traceable to 7 variations in the medium reagents used by various 8 laboratories. As a consequence of this, we developed 9 a practical standardized cell culture powdered media 10 in kilogram lots, that unlike wet media was easily 11 transported worldwide and used by many labs 12 simultaneously and for which only two variables 13 existed, water and serum. As a result of our 14 development of powdered media, the WI38 growth problem 15 was quickly resolved and the use of powdered media has 16 now become commonplace throughout the world. 17 In the years following our development of 18 WI38 in 1962, as a matter of national pride other 19 countries produced their own strains. The Medical 20 Research Council in the UK produced MRC5. In 1970, 21 the Japanese made TIG1 and the Chinese made 2BS, et 22 cetera. Today, most of the world's virus vaccines are 23 made on WI38 or similar strains. It has been used to 24 provide more individual doses of a biological than any 25 other cell substrate ever used. This slide following 21 1 gives you a summary of or a rough idea of the extent 2 of the use of these populations. 3 I am frankly astonished to observe that 4 after 10 years of debate on the risks of moving from 5 frequently contaminated primary monkey kidney cells to 6 virus free normal human cells that the time required 7 to make the final leap to abnormal cell lines for 8 producing human biologicals occurred in the 1980's in 9 a matter of weeks. Today, the molecular mechanism for 10 our finding that normal cells have a replicometer has 11 been substantially described as a result of a 12 fascinating story that time constraints prevent me 13 from describing. The explanation is briefly described 14 on the next slide and consists, as most of you here 15 know by now I am sure, of the shortening of telomeres 16 in normal cells each time DNA replicates. And for 17 those cells that are immortal, they escape the 18 inevitability of telomere shortening by switching on 19 an enzyme called telomerase that adds on the 20 nucleotides at each round of DNA replication, thus 21 maintaining the telomere length constant. 22 The remaining question, as I just 23 described, in respect to telomere shortening was the 24 discovery of this enzyme called telomerase, an enzyme 25 that contains an RNA template and a catalytic moiety, 22 1 among other components. This is the only enzyme that 2 contains a reverse transcriptase and is part of normal 3 cell biology. 90 percent of cancer cells express the 4 enzyme. Only normal stem cell populations do so. 5 As for the use of abnormal and mortal cell 6 lines for the production of human vaccines, that has 7 already been done. The Vero cell line derived from 8 the kidney of an African green monkey kidney has been 9 licensed in France, Belgium, the United States and 10 Thailand for the production of both live and 11 attenuated polio and rabies vaccines. There are now 12 more than 20 million vaccinees. The Chinese have made 13 a Japanese encephalitis vaccine in the abnormal cell 14 line, BHK21. Recombinant vaccines have been licensed 15 for hepatitis B grown in a CHO cell line, and phase 16 III trials are currently underway for an HIV vaccine 17 in these same cells. 18 There are other immortal cell lines that 19 bear consideration as substrates for the preparation 20 of human virus vaccines. About two years ago, I 21 initiated a culture of fibroblasts from a skin biopsy 22 that I took from my knee. This culture was the first 23 to be immortalized by transpection with human 24 telomerase reverse transcriptase. Three other human 25 cell strains were similarly treated and these results 23 1 were published in Science in January of 1998. These 2 immortalized normal cell populations have now 3 undergone well over 400 population doublings. This 4 opens up enormous opportunities, not only to 5 immortalize WI38, but many other normal cell types for 6 applications like -- for exploitation actually as 7 vaccine substrates among many other potential uses. 8 Another category of immortal cell lines 9 that merits consideration are those transformed by 10 agents other than viruses. SUSM1 is a cell line 11 derived from normal human lung that we transformed 12 with a chemical carcinogen. KMSD6 is also an immortal 13 cell line derived from normal human fibroblasts and 14 transformed by exposure to cobalt 60 gamma radiation. 15 I might add that contrary to the belief of some, early 16 passage WI38 does exist in ample quantities, and 17 anyone wishing to obtain these cells should feel free 18 to contact me. 19 Thus, there exists several methods for 20 transforming well-characterized normal human and 21 animal cell populations to immortal cell lines using 22 at least three proven methods. What should not be 23 overlooked, however, in our zeal to condemn any 24 immortal abnormal cell line from consideration as a 25 human virus vaccine substrate is the fact that 24 1 downstream processing has reached such a level of 2 efficiency that the preparation of a human virus 3 vaccine containing only the purified moieties capable 4 of eliciting an immune response is possible now. 5 Affinity technology in particular is clearly capable 6 of meeting this exacting goal. Nevertheless, as we 7 all know, economic considerations will surely guide 8 most of these decisions. Thank you very much. 9 DR. GRIFFITHS: Thank you very much 10 indeed. We do have time for one or two questions if 11 people want to start the ball rolling and ask Dr. 12 Hayflick any questions. I don't see any immediate -- 13 oh okay. Sorry, I can't see your name from here. Can 14 you introduce yourself? Is there a microphone? 15 PARTICIPANT: I am Jim Cook from Chicago. 16 What all types of animals have WI38's been inoculated 17 into for tumorigenicity testing besides the hamster 18 cheek pouch? 19 DR. HAYFLICK: Oh my. 20 PARTICIPANT: I mean, have they been put 21 into newts. Have they been put into other kinds of 22 immunocompromised animals? 23 DR. HAYFLICK: All immune compromised lab 24 animals -- mice, rats -- 25 PARTICIPANT: What about things like the 25 1 odd observation that Balb C3T3's when hooked onto 2 plastic disks will make tumors in newts? Has anything 3 unusual like that been done with WI38's? 4 DR. HAYFLICK: Yes. And to the best of my 5 knowledge, there has been no tumor formation. There 6 have been occasional reports of nodules where the 7 cells will replicate to a point where they reach a 8 macroscopic size. But usually what happens is that 9 they fail to grow beyond that state. 10 DR. GRIFFITHS: Johannes? 11 PARTICIPANT: Johannes Loewer. You 12 mentioned that the human diploid cell lines have a 13 very broad host range for many different viruses. Is 14 it known whether neoplastic cells have even a broader 15 host range or is it more or less similar to diploid 16 cells? 17 DR. HAYFLICK: You are talking about cell 18 -- immortal cell lines? 19 PARTICIPANT: Yes, immortalized cell lines 20 or neoplastic cell lines. 21 DR. HAYFLICK: From human material? 22 PARTICIPANT: Yes. 23 DR. HAYFLICK: To the best of my 24 knowledge, the human diploid cell strains have a far 25 broader virus range than the immortal cell 26 1 populations. Most or many of the rhinoviruses were 2 first discovered on WI38 and later on MRC5. But the 3 fact that they could ultimately be adapted to growth 4 in immortal cell populations was also revealed. 5 However, I think it is still fair to say that normal 6 human diploid cells are the most sensitive cell 7 population to most of the viruses that are now well 8 characterized. 9 DR. GRIFFITHS: Any other points that we 10 should take? Because we are right on time, so we are 11 really doing quite well this evening. If not, then 12 thank you very much indeed. 13 DR. HAYFLICK: Sure. 14 DR. RABINOVICH: Our next speaker is Dr. 15 Elwyn Griffiths from the World Health Organization, 16 who has also worked in this field for -- I won't even 17 venture to guess how many years. He is going to talk 18 about major issues associated with cell substrates. 19 DR. GRIFFITHS: Thank you very much. I 20 must say I was very impressed when I got this badge 21 actually from the organizers. I mean, I don't know if 22 they get a prize for the number of flags that is on 23 it. It is very impressive. I mean, I was only asked 24 to do this particular talk and stand in just a couple 25 of weeks ago. I think it should have been Dr. 27 1 Petricciani who would have done this and done it far 2 better than I am sure. But in standing in, perhaps I 3 should have stood in for the patron committee or 4 something just to have another badge on here. 5 Anyway, for this evening to a certain 6 extent this talk really is something of a trailer to 7 the more detailed discussion we are going to have over 8 the next few days, and really an extension of the 9 points already touched upon by Dr. Hayflick. 10 The way I thought I would like to do it is 11 to take a different stance in the beginning. Before 12 dealing with the major issues which really revolve 13 around these different cell substrates, to start with 14 a reminder. And I think I would like this reminder to 15 be borne in mind throughout the meeting that what we 16 are dealing with has great benefits for mankind. I 17 mean, the benefits of the vaccines produced on cell 18 substrates. There has been enormous success in cell 19 substrates for producing vaccines. These particular 20 vaccines have had a tremendous influence on human 21 health. 22 I would just want to pick up on polio, for 23 example, because that is something which the WHO 24 clearly has a -- and everybody else, I guess -- has a 25 major interest in. The last case of polio in the 28 1 whole of the Americas, that is PAHO, was in 1991. In 2 the Western Pacific region, it was in March of 1997. 3 The world is moving towards global eradication of 4 polio in roughly about -- it is expected in 2000. I 5 am not sure if 2000 is realistic -- 2000, 2002 or 2003 6 or so. You can see the difference from -- the top one 7 is 1988 really to 1998. There are patches left in 8 1998, and this is gradually getting smaller and 9 smaller. 10 So essentially, global eradication of 11 polio, which will have enormous benefits for mankind, 12 has been achieved primarily through live polio 13 vaccines. These vaccines are produced on a number of 14 cell lines -- primary monkey kidney cells, diploid 15 cells, and also in vero cells. I think it is true to 16 say on a global perspective that the majority of the 17 producers are producing on primary monkey kidney 18 cells. 19 Now having said that, I just wanted to 20 have that borne in mind. When we deal with the 21 issues, we really do have some major benefits from 22 these products. So it really is how much of a 23 risk/benefit equation which we have to deal with here 24 and not forget the actual benefits by focusing too 25 much on the issues. That isn't to say the issue is 29 1 not important. 2 The major issues associated with the use 3 of cell substrates really relate to safety. I mean, 4 that is our main problem here. This is something 5 which has been really extensively debated over the 6 last 40 years or so, with the acceptance of cells for 7 production really being a hot topic as they come up 8 from time to time -- primary cells, diploid cells and 9 continuous cell lines, as Dr. Hayflick already 10 mentioned. The perceived risks being attributed to 11 contaminants coming from the cells. The main ones of 12 interest really being viral contamination, DNA, and 13 transforming proteins. 14 The whole of these issues have been 15 discussed -- I wouldn't say ad nauseam, but they have 16 been extensively discussed over -- what is it, from 17 1967 -- over 30 years or so. And what is interesting 18 is when you look back over this, this is a selection 19 of meetings. I am sure there are many, many more 20 meetings. You see, I am only just taking this from 21 1967. It is the same players essentially or the same 22 cast as we have for this meeting. We have the FDA. 23 The IABS features strongly in this, the WHO, the NIH. 24 So they are all there really right from the very 25 beginning. 30 1 The main point here is that some of the 2 issues -- I mean, each meeting will have a special hot 3 topic, if you like. In 1967, it was diploid cells. 4 Then in 1978, that meeting was organized essentially 5 to see whether we could move forward with the use of 6 continuous cell lines, primarily because namalva cells 7 have already been produced by the Wellcome Institute 8 for producing interferon. And I suppose the 1978 Lake 9 Placid conference gave us a cautious yes to the use of 10 continuous cell lines for the production of 11 biologicals. And I am broadening this now not just 12 for vaccines, but for the production of biologicals. 13 But there was still a lot of "yes, but" at 14 that meeting. I wasn't there, and I am sure that John 15 and other people who would have been there would say 16 the same thing. It seems that it was, yes, we can go 17 forward, but, but. And certainly not in the USA, I 18 guess. That was the sort of general feeling at that 19 time. It was really the meeting in 1986, the WHO 20 meeting, a study group in Geneva, which really allowed 21 the field to move forward in so far as use of 22 continuous cell lines were concerned for the 23 production of biologicals. 24 At that time, there was an agreement from 25 this particular study group that continuous cell lines 31 1 could be used for the production of biologicals 2 provided that the level of residual DNA or residual 3 host cell DNA was kept below 100 picograms of DNA per 4 dose. That is the origin of the magic figure which has 5 been used for the last 10 years or so. And this 6 really allowed -- this decision really allowed the 7 field to move forward in the expansion of the 8 biotechnology industry to produce recombinant DNA 9 biologicals in continuous cell lines. This has been 10 the situation for the last 10 years or so or more now. 11 At the last meeting of the WHO, which was 12 1996, a WHO and NIABS consultation at Merieux 13 Fondation in Annecy, that is the time where the field 14 was reviewed, if you like. Where are we after 10 15 years or so of using this figure of 100 picograms? 16 Incidentally, that figure of 100 picograms was used as 17 the guidance for all the national regulatory 18 authorities. The FDA went a little bit further and 19 went down to 50 at one time, less than 50 picograms. 20 Industry was pushing toward the limited detection. 21 And it is interesting to see the European 22 Pharmacopoeia has a limit for recombinant hepatitis B 23 vaccine produced in continuous cell lines of 20 24 picograms per dose, which is very low and very 25 stringent. This meeting in 1996 was to review that, 32 1 and things have changed a little since then and I will 2 come to that later on. 3 The present meeting, of course, really 4 focuses on cell substrate for vaccine production, not 5 for biologicals in general. I think there is this 6 distinction which I would like to make later on. We 7 are now moving to potential use of overtly neoplastic 8 cells for vaccine production, and I think this is a 9 very timely meeting. 10 The sort of landmarks -- and I am really 11 going over a little bit the ground which has already 12 been covered by Dr. Hayflick. These were sort of, I 13 think, the sort of landmarks as you might say. The 14 gradual overcoming of the hurdles -- the use of 15 primary cells for vaccine production in polio really 16 exemplifies that in the 1950's. In the 1970's, 17 diploid cells were accepted for vaccine production, 18 but not without problems as we have already heard. 19 Then in the 1980's, after the WHO meeting, continuous 20 cell lines then became acceptable for vaccine 21 production -- namalva cells for interferon, vero cells 22 and so on -- hybridomas for monoclonal antibodies 23 before then of course, and recombinant DNA 24 derivatives. 25 Now the progress during this period of the 33 1 1960's to the 1990's has been really on a number of 2 fronts. I think it is fair to say -- I can say this 3 because the WHO isn't really a regulatory agency. But 4 I think one can now say that regulatory decisions are 5 made more on a scientific basis and not on opinions as 6 such. And the progress that has been made in 7 scientific knowledge really over this period is 8 enormous. We have to remember that. I just put a few 9 up here to flag them -- new diagnostic procedures, 10 let's say, such as PCR; better understanding of 11 molecular mechanisms of pathogenicity; and new 12 concepts, the concept of validation has been 13 introduced. 14 So we are now in a position -- it is 15 perhaps unfair to say that the regulatory decisions 16 can be made more on a scientific basis. We have got 17 more scientific basis to make the decisions. I think 18 that is perhaps the main point to make there. We can 19 make decisions, better informed decisions, and I think 20 this is an important part of the gradual move to 21 acceptance of different cell substrates. 22 Part of this process as well has been co- 23 defining, if you look -- perhaps co-define is too 24 strict. Having regulations and guidelines and 25 recommendations for production and quality control of 34 1 these products using cell substrates has also been a 2 major development since that time, both on a national 3 basis, on a regional basis, such as in the European 4 Union and the European Commission, and the WHO, of 5 course, on a global perspective. And I think what we 6 are doing is really learning how to manage the risks 7 to take advantage of the benefits. And I think that 8 is the important -- what we want to do to take the 9 field forward. 10 Very briefly then I will go through just 11 to remind ourselves and really to reiterate what Dr. 12 Hayflick has said. The primary cells -- I mean, I 13 will just mention the disadvantages here in the sense 14 that many people tend to use them because they tend to 15 be easy to prepare and so on. But they really do have 16 this major disadvantage of microbial contamination and 17 especially of viruses, of course. 18 Can we go to diploid, the next one, and 19 just go through these very quickly. Again, and it has 20 already been eluded to, the great advantage of using 21 diploid cells is that they can be well characterized 22 and standardized. And so production then becomes based 23 on well characterized cell banks, the master cell bank 24 and a working cell bank. And this is true also for 25 the continuous cell line. 35 1 But they all have their advantages and 2 disadvantages. I just want to focus on the 3 advantages. An infinite life span for the continuous 4 cell lines, so you can keep them going forever I guess 5 really. And again, production on well characterized 6 and standardized cells for an indefinite time basis 7 really. 8 Now when you look at the disadvantages 9 amongst these cell lines, the primary cells had the 10 problem with the contamination. But the main 11 disadvantage of the continuous cell line is that many 12 do express endogenous viruses and there has always 13 been this concern over tumorigenic potential, should 14 we say, associated with cellular DNA. 15 The main three risks then with these 16 different cell lines for producing biologicals are 17 contaminating viruses, and we must include here, the 18 TSEs, the transmissible spongiform encephalitis 19 agents, whatever that may be in the end, residual host 20 cell DNA, and growth-promoting proteins. And as I 21 said, the trick really is how to manage these. And by 22 developing procedures, we can actually move forward 23 onto guidelines and so on. 24 In fact, there are three principles 25 essentially really to production of biologicals. And 36 1 these are really the basis for moving forward on all 2 these cell lines. Control of the starting material, 3 such as the baseline data, characterization of the 4 host cell, and so on, if you are able to do that 5 depending on what sort of cell line you're using. 6 Then you control the manufacturing process itself, 7 this is for adventitious agents and so on, and also 8 control of the final product. And that includes tests 9 performed during development. Because not all tests 10 will be carried out on a routine lot-to-lot basis. 11 There will be some testing on development. And then 12 there is a subset of those tests, which will be 13 carried out on a routine basis. These are the lot-to- 14 lot releases. 15 Now historically, assuring the quality and 16 safety of what you might call traditional biologicals 17 has been what I call disaster-led. And you can make 18 a nice list of all the problems we have had over the 19 years and you find that control procedures are really 20 established or altered following some major mishap, 21 let's say. For example, Creutzfeldt Jakob disease 22 from the bottom there in recipients of pituitary- 23 derived growth hormone. I mean, that meant that we 24 weren't now going to use the pituitary-derived growth 25 hormone and move over to recombinant. So it is these 37 1 sorts of issues which really trigger the traditional. 2 But if you look at the situation which I 3 call applied to biotechnology products -- I mean novel 4 biotechnology products. I mean, it is all 5 biotechnology I guess really, but I am thinking of 6 recombinant DNA, monoclonal antibodies, cell cultures 7 and so on. The approach has been somewhat different. 8 Guidelines have been established early-on in the 9 development of the new field to try and pre-empt any 10 disaster. I mean, the idea here is that we think of 11 what the problems might be and try to make sure that 12 we don't land ourselves in these sort of situations. 13 And developing guidelines and points to consider, 14 whatever you might want to call them -- and the 15 different countries and different agencies call them 16 different things -- I think these have been 17 instrumental in establishing safety and quality of 18 these newer type of biologicals, the biotech products 19 if you like. Because they are by far the best 20 characterized and best purified biological medicines 21 in clinical use. 22 You see during this period in the 1970's 23 and 1980's, guidelines on production and control or 24 you might call them points to consider have been 25 derived by many agencies. The main ones being the CBER 38 1 FDA points to consider, Japan has guidelines, the 2 European CPMP biotech working group, those are the 3 European Commission guidelines, and WHO. Now what is 4 interesting is that the majority of these guidelines 5 do have the benefit of mutual consultation during 6 their development. If you look at them in some 7 detail, you find that essentially they all say the 8 same thing, but they are actually viewing it from a 9 slightly different perspective with slightly different 10 words. The ICH guidelines, for example, on cell 11 substrates, that actually deals with more of the 12 characterization of the cell lines and so on. So we 13 find that these different type of guidelines do all 14 have the same thrust, but they take it slightly in 15 different perspectives. But they all say essentially 16 the same thing. 17 Now as far as WHO is concerned, the first 18 requirements for cell cultures for production of 19 biologicals was the 1959 requirements for primary 20 monkey kidney cell production of polio vaccine. Now 21 all the other major agencies -- FDA and in Europe and 22 the individual national states -- would also have 23 their requirements as well. But WHO is the embodiment 24 of a global situation. And using the primary cells, 25 monkey kidney cells, for polio vaccine really did lead 39 1 to some confidence in developing other primary cell 2 lines for producing a range of vaccines. And these 3 are still in use. We mustn't forget that. The 4 primary cell is still a major source -- for example, 5 the chick embryo cells for measles vaccine and mumps 6 vaccine as well. So the development of these at this 7 point was a sort of major move forward. 8 Now I am not going to try and sort of 9 compare and contrast the different guidelines, because 10 I don't think that is particularly helpful. But what 11 I will try and do is just to scan briefly through the 12 guidelines and points to consider which have been 13 developed in relation to the cell cultures. These, 14 again, reflect considerable discussion. I mean, I have 15 just put the WHO ones. The first one there was the one 16 for oral OPV, and these are updating. As time goes on, 17 there is always a need for updating because some 18 technology changes and things move forward. 19 I have already mentioned the meeting of 20 the study group in 1986. It was published in 1987, 21 which allowed the field to move forward, and that was 22 the acceptability of cell substrate for production of 23 biologicals. And then in 1987 as well, there was the 24 continuous cell line for the production of 25 biologicals. The FDA, the ICH, the CPMP also have 40 1 their guidance documents as well specifically on cell 2 substrates. 3 Now I just want to say something about the 4 situation of 1998 as far as the WHO cell substrate 5 document is concerned. This document actually 6 reflects considerable international consultation. I 7 think this is probably the latest one or major 8 international one. It was published last year, but the 9 actual meeting -- there was a meeting in 1996 with 10 WHO, IBS, and we had the benefit of input from the 11 ICH, which went into great detail about how we should 12 control the quality of the biologicals produced on 13 primary, diploid and continuous cell lines. It covers 14 the whole three for production of viral vaccines and 15 the biologicals. It is not restricted to viral 16 vaccines. And it is interesting that this document is 17 very clear in that it encourages the move away from 18 primary cells to cell substrates which can be 19 generated from well characterized cell lines. As we 20 have already mentioned -- Dr. Hayflick mentioned this 21 idea of the well characterized cell line goes back a 22 long time, probably to the meeting -- I think there 23 was a meeting in Zagreb where this was developed of 24 having the master cell bank and so on. 25 Now in this particular document, there has 41 1 been a major change in certain parts of the way we 2 handle the control of some of these cell lines. There 3 is, for example, for well characterized diploid cells, 4 a relaxation of the karyology and tumorigenicity 5 testing and so on. But I just want to touch on one 6 particular point which is important for the next part 7 of the issues. And that is related to the DNA and the 8 viruses. The main risks -- in this document, we still 9 see that the main risks still are for the continuous 10 cell lines residual DNA and maybe growth promoting 11 proteins, and I will come to that in a moment, and for 12 all cell lines, contaminating viruses. 13 Now as far as the DNA is concerned, there 14 has been a major departure in the previous 15 recommendation regarding residual DNA. And this is 16 really due to a reevaluation of the situation and new 17 data over this decade from 1986. But also new 18 technologies have appeared on the scene -- gene 19 therapy and DNA vaccines, which actually are involving 20 injection or administration of in the case of DNA 21 vaccines quite large quantities of DNA. So one has to 22 be consistent here. Either DNA is really a big 23 problem or it isn't. Or is it the type of DNA? And 24 this is the issue. 25 If you look at the 1986 summary, I guess 42 1 really, the cause of concern for that time was 2 transmission of oncogenes. And the 1986 consultation 3 provided this way forward by saying that if you can 4 get your DNA down to less than 100 picograms per dose, 5 then there is no realistic risk, and this is what 6 everybody went forward on. The perceived problem was 7 really not DNA post-op, but the specific sequences of 8 DNA which may be encoding oncogenes and the possible 9 insertion of endogenous or host cell protooncogenes or 10 inactivation of suppressor genes by the DNA. The DNA 11 itself was largely ignored, I think, and that is were 12 the move forward is on DNA vaccines. We can go forward 13 there because although there is a risk, the acceptance 14 of this minimum risk is something which has been 15 actively discussed and debated for the last two or 16 three years. 17 So the figure now is that we now think -- 18 by we, the current opinion. I think this is 19 everybody. As I say, everybody has partaken into 20 these general discussions here. They are based on 21 current state of knowledge. The suggestion is that 22 DNA from continuous cell lines can be considered -- 23 the DNA itself now -- as really a cellular contaminant 24 rather than itself as a significant risk factor 25 requiring removal to extremely low levels which needed 43 1 validation and testing of each lot and so on. And the 2 relaxation now is that we think that a figure of 10 3 nanograms of residual DNA from a continuous cell line 4 will apply. And this applies primarily, of course, to 5 purified recombinant DNA biologicals. 6 This is something which I want to touch on 7 here. What we are talking about is removal of DNA to 8 reasonably low levels. But the guidelines do make the 9 point that residual DNA from continuous cell line may 10 pose a higher risk if it might include infectious 11 retroviral or provirion sequences, and then we may 12 well still have to go down to lower levels. Now this 13 is something I am sure that is going to come up in 14 this meeting. What is the actual level which is 15 applicable to vaccines. Can you actually go to that 16 lower level. 17 I don't want to say much about the growth 18 of multi-proteins. These are sort of more or less 19 dismissed, I suppose, in the WHO guidelines. Growth 20 factors that are produced by cells, of course, the 21 risk is very limited. Yes, they do affect cell 22 growth. It is usually transient and reversible. They 23 don't replicate and many are inactivated in vivo. So 24 the general thrust of the argument there was that they 25 are not such a problem as we originally thought. 44 1 Now that might be different today, but the 2 viruses remain the major and real concern for all cell 3 types. The test for viral contamination really is a 4 major part of cell bank characterization. In the 5 guidelines, in the WHO ones, it covers the primary 6 cells here. Now, here we are taking cells directly 7 from animals. Sometimes in the wild animals for polio 8 -- from monkeys for polio. And quite clearly now, we 9 have to move into a situation if you are going to use 10 animals that they must be healthy animals subject to 11 veterinary and laboratory monitoring, very closed 12 colonies wherever possible, not from the wild if 13 possible. And animals, if they do come in from the 14 wild, then they really have to be quarantined and 15 appropriately monitored in the time before you 16 actually take the cells for preparation of vaccines. 17 And serological screening of donor animals for 18 relevant pathogens. 19 Now you might say what is a relevant 20 pathogen. Well that really is changing as time goes 21 on, of course. The problem with doing this for a 22 primary cell is that you have to do it for every 23 vaccine batch. I mean, it is not the case of doing it 24 once. You need to test the cells vigorously. It is 25 an awful lot of work. And of course you can only test 45 1 for a certain number of known viruses. You can have 2 a sort of catch-all, but there are mishaps. As we 3 know, SV40 was missed, and this is an issue that comes 4 back to haunt us as time goes on. 5 With a cell bank, we can actually do this 6 testing, extensive testing for exogenous and 7 endogenous agents, once only. You can do it and then 8 you can make sure that this is the master cell and 9 then you move on to the working cell bank. The 10 further testing really is on the working cell bank or 11 production cells really to detecting common 12 adventitious agents, and that could be sort of a 13 catch-all. It is not such a major issue. 14 It is interesting that the working group, 15 the WHO study group, in 1986 valued the well 16 characterized cell banks so much that it recommended 17 the establishment of a well characterized cell line 18 that would be available to national control 19 authorities and manufacturers of biologicals globally. 20 And, in fact, the WHO master cell bank for vero cells 21 was established following the recommendation by this 22 group. It comes from African green monkeys and it was 23 selected -- the vero cell was apparently selected 24 because of the immediate prospect for improving the 25 quality and quantity of several vaccines. Now it has 46 1 been used but perhaps not used as widely as was 2 thought at the time. 3 Now the strategy for preventing viral 4 contamination then is basically what all the documents 5 go through. Tests on starting materials -- not just 6 the cells. We have to remember that some of the 7 agents will come in from the reagents -- the serum and 8 so on. We have to be very aware of excipients as 9 well. 10 Evaluation of production processes. That 11 is the ability to remove or inactivate potential viral 12 contaminants. That is your validation essentially. 13 This is the concept where validation has come in. And 14 then tests on final product or appropriate 15 intermediates. And essentially what people think is 16 that it is not a single approach. One approach isn't 17 sufficient. It is the combination of approaches which 18 gives you this degree of security. 19 Now if you look at the WHO guideline, and 20 I think this is true of most of the points to consider 21 in the guidelines, they are really focusing on biotech 22 products, that is, highly purified recombinant DNA 23 products essentially. Now if you look down at the 24 bottom, the purified biotechnology product -- you can 25 have a very robust virus clearance system that you can 47 1 check for validation of virus clearance downstream and 2 so on. So even if you are -- I mean, I don't think 3 there are any biotech products produced or recombinant 4 DNA products produced in primary cells, but even if 5 you were doing it that way, you could be sure that -- 6 you could actually be sure that everything was going 7 to be cleared up by the system. When you come to virus 8 vaccines, you have much less purification, and the 9 possibility of virus/virus interaction during 10 production needs to be borne in mind, and this is an 11 issue which will come up later in the meeting. 12 So virus vaccines, though, are a different 13 ballpark really from the general biologicals, which 14 everybody -- you know, most of these guidelines are 15 focusing on clearance of DNA and clearance of viruses, 16 looking at the characterization of cells. When you 17 move to virus vaccines, there is less you can do. 18 Now we mustn't think, of course, that we 19 know everything at this stage. I mentioned that we 20 have had enormous progress in the last 30 years or so. 21 We must be prepared for surprises. Two recent events 22 I think have raised awareness of the challenge there 23 is in dealing with viral contamination of cell lines 24 and vaccines which are used in their preparations and 25 the consequences. One was the detection of very low 48 1 levels of reverse transcriptase in chicken cell- 2 derived vaccines using this newly developed assay -- 3 we call it PERT assay. This is from -- like measles 4 vaccine is an example and mumps is another example. 5 All the vaccines produced in eggs will contain very, 6 very low levels of RTAs. And this, of course, when it 7 first was discovered raised a lot of alarm bells. 8 People were thinking, yes, there must be retroviruses 9 in here and so on. And this really led to quite a lot 10 of activity trying to show where this activity was 11 coming from. Is it real retrovirus or what do we have 12 there? What is the problem? Or do we just have some 13 non-specific activity which really looks like reverse 14 transcriptase? And much work -- and WHO was very much 15 interested in this, of course, because if national 16 control authorities of one country ban a vaccine 17 because it is considered to be potentially unsafe, 18 this has ripples throughout the whole of the vaccine 19 community. I mean, what is not safe for one set of 20 kids in one country must be unsafe for other kids of 21 course. 22 So there was a lot of work to try and make 23 sure that this activity -- where did it come from and 24 so on. And manufacturers and the national control 25 authorities were very active here. It has now been 49 1 shown to be particle associated, but extensive studies 2 have shown no transmission or productive infection, 3 and that of course is good news. But as we will hear 4 later on in the meeting, what about the potential for 5 interaction between these particles and some of the 6 viral vaccines -- the actual virus which is used for 7 producing the vaccine during growth. We need to be 8 sure -- I am thinking of pseudotypes and so on here. 9 This is an issue which we need to address in this 10 meeting. 11 The other surprise was the detection of 12 SV40 genome in rare human tumors. This is something 13 which has come back to haunt us after 30 years or so. 14 I am sure you all know that SV40 was a contaminant of 15 some of the early batches of primary rhesus monkey 16 kidney cells used to produce polio vaccines. This is 17 no surprise. During the 1950's, these were actually 18 used in a large number of people -- in the millions. 19 There was follow-up with that to see whether they 20 actually caused any problems, and nothing much 21 materialized. And then suddenly about three or four 22 years ago, the SV40 sequences were picked up in 23 various rare human tumors. That raised the issue of 24 was the vaccine -- was the polio vaccine made in 25 primary kidney cells actually still transmitting SV40 50 1 or SV40 sequences. Because right in the beginning 2 when SV40 was discovered, measures were introduced 3 very quickly by national regulatory authorities to 4 exclude SV40 from polio vaccines. For example, I am 5 thinking here of the WHO. The 1959 requirements had 6 to be modified and they were updated and the 7 regulations in all the major national authorities were 8 also changed to exclude SV40. And for more than 30 9 years then, polio vaccine made in primary monkey 10 kidney cells have been shown to be free of SV40. Now 11 by shown to be free, it was shown to be free by 12 technology of the 1960's, I guess. I mean, it is old 13 technology. And the question which everybody was 14 rather anxious about was were we actually missing 15 something here? And then the application of new 16 highly sensitive PCR techniques for detecting SV40 17 genome was then introduced to see whether batches of 18 vaccines were actually carrying any SV40 sequences. 19 This is work carried out by CBER and also by NIBSC, 20 and we are looking back on historical samples as being 21 quite clear that the methods introduced in the 1960's 22 were effectively excluding SV40. 23 Now if I can sort of come back to where I 24 started from in the beginning in relation to polio, 25 the polio eradication is going to be with us in about 51 1 two or three years, and this is really an end-game and 2 how do we move forward here. Once polio has been 3 eradicated, transmission of wild polio would be 4 stopped. And the global certification of polio 5 eradication will come in around 2005, I guess. But 6 there will be continued use of the OPV expected for 7 maybe 5 or 10 years. It is unclear. The end-game here 8 is unclear really and how we move forward. But WHO is 9 looking towards cessation of immunization with OPV 10 around 2005 or 2010. USA will be moving to IPV in 11 2000 -- in January of the year 2000. So the question 12 is do we need to do anything about the vaccine, the 13 OPV. And we need now, because -- I think now as we 14 move to eradication, I think we do need to be 15 absolutely sure of the vaccine and introduce an 16 additional level of security. This is provided by 17 ensuring that SV40 sequences are absent from polio 18 virus seed. Because during the actual survey of the 19 vaccines, one seed I think from a manufacture was 20 found to be positive. The seed was positive. There 21 was no actual live SV40 there. There were some SV40 22 sequences there. It is considered that it is much 23 better to have the seed free of SV40. And we are now 24 actually updating the original requirements -- this 25 will come up this year -- making sure that all seed 52 1 will be tested for absence of SV40 sequences. The 2 primary monkey kidney cells for production will now 3 have to come, I think, from closed, intensely 4 monitored colonies of animals and not from the wild. 5 As time goes on, of course, new viruses 6 are discovered and new problems arise. The foamy 7 virus has been identified as one that we should be 8 really sure is absent from these vaccines. So 9 updating is something which we are moving forward 10 towards in this end-game for on the polio. 11 What is the next step then? What are the 12 challenges for the future? The use of neoplastic 13 cells or novel cell lines for vaccine production I 14 think is really a challenge. And I think one could 15 say as with any new technology, new sets of safety 16 issues are generated for consideration. Not just by 17 regulatory agents but by industry and regulatory 18 agencies together. I mean, everybody is on a learning 19 curve here. It is not one or the other dictating. 20 Everybody is together here. We need to consider 21 again, I think, some of the issues of residual DNA. 22 Is it oncogenic? What is the issue there? Is there 23 an infectious DNA in relation to what sorts of cell 24 line you've got and what is in the cell? It is really 25 timely to review and assess the risks in light of a 53 1 much better understanding of the molecular mechanisms 2 of geo-originicity and of viral/viral interaction. As 3 I mentioned, I think we will touch on that during the 4 meeting. 5 What I think we have to do is to come to 6 some realistic and scientifically sound decisions 7 concerning the use of these substrates. Now, a point 8 which I from sort of a WHO and a global perspective 9 think is very important -- I think science and 10 commerce in biologicals -- I mean, vaccines are made 11 in one country and travel the world. They are not 12 necessarily just used in one country. It really is an 13 international situation. And so too are the public 14 health questions which the use of these products 15 raise. So if there is an issue in relation to a 16 problem in one country, it is a global issue. And the 17 international dimension of the discussion of these 18 issues is vital. And I am glad that we do have an 19 international group here, and I think it is important 20 that WHO is part of this and we are very pleased to be 21 co-sponsoring this meeting. Thank you very much. 22 DR. RABINOVICH: Are there any questions 23 specifically for Dr. Griffiths? Dr. Rubin? 24 PARTICIPANT: I was not quite clear on 25 what you were saying about the presence of 54 1 retroviruses in chick embryo cell cultures. Whether 2 you thought that in fact it was present or not. And 3 the reason I am asking is we had a lot of experience 4 with chicken leukosis viruses in chick embryo cells 5 beginning back in 1960. And the thing about them is 6 they are not easy to detect because they don't produce 7 any pathogenic effect and they have to be detected in 8 indirect means. The indirect means in those days when 9 they were first found by us, at least, was 10 interference with RAS sarcoma virus preventing. And 11 then much later on came -- at least 10 years later 12 came reverse transcriptase. The thing about these 13 leukemia viruses or leukosis viruses, these 14 retroviruses, other than the really pathogenic ones, 15 is that they are highly species specific for one 16 thing. So chicken viruses in general, with some 17 exceptions, will not infect other species. So if we go 18 up the scale, I think, to the mouse leukemia viruses, 19 I think it is generally true of them also that they 20 are not cytopathic in culture. They can't be detected. 21 And probably they are of no concern to other species 22 unless they are very closely related. The concern I 23 would have of vaccines made in higher species, monkeys 24 or humans, is that probably there are -- or not 25 probably, there may be some stealth viruses like these 55 1 that don't produce any obvious effect and that we 2 don't even know about their presence to even detect 3 them in animals. Let's say the chicken virus requires 4 a flock of chickens that is not infected with them 5 that has to be infected at a very early age and you 6 have to practically wait for half a lifetime of the 7 animal to produce any statistical effect among them. 8 So I wonder if there is any concern about these points 9 with the retroviruses. 10 DR. GRIFFITHS: Can I just clarify the 11 point about the ALV, the avian leukosis virus. 12 PARTICIPANT: Yes. 13 DR. GRIFFITHS: The vaccine production 14 stipulates that the eggs are free from the -- the 15 chickens are free from the avian leukosis virus. 16 PARTICIPANT: Yes. 17 DR. GRIFFITHS: So the reverse 18 transcriptase activity wasn't -- it wasn't expected. 19 There are very low levels of RTAs. I mean, it has 20 been tested by the ordinary methodologies -- the sort 21 of standard methods. It was only when you went down 22 to these very low levels. And the origin is now known. 23 Where it comes from is endogenous viruses, retros. 24 And they do form particles, but as I said they are 25 non-infectious particles. But on a global basis, 56 1 yellow fever for example -- the yellow fever vaccine 2 -- this is some sort of a risk/benefit which the WHO 3 has to sort of wrestle with which a national authority 4 has the benefit of not necessarily -- like the USA, 5 for example, doesn't have to. It can make a decision 6 right away. But, for example, WHO requirements -- 7 they will be changing, but they have allowed 8 production in ALV positive eggs because you cannot get 9 ALV-free flocks in many -- in places where you are 10 actually making yellow fever vaccine, and it was 11 considered that the lack of yellow fever vaccine was 12 much more of a risk than the presence of the ALV. 13 Nevertheless, Western Europe and USA making yellow 14 fever vaccine is all in ALV-free -- from eggs from 15 ALV-free chickens. And that will now be introduced 16 into WHO requirements as well, I think, in the future 17 because these countries which were producing in ALV- 18 positives can now actually get the eggs free. 19 PARTICIPANT: Okay. 20 DR. RABINOVICH: One last question? 21 PARTICIPANT: I was really going to 22 comment on actually the last comment from Dr. Rubin. 23 And reiterate your point that the sequence that appear 24 to be responsible for the RTA activity appear to be 25 the AEV gene and not the ALV gene. 57 1 DR. GRIFFITHS: That is right. Yes. 2 PARTICIPANT: But actually both the AEV 3 and ALV genomes can actually undergo recombination. 4 One of the subgroups of ALV is actually based on the 5 AEV envelope gene sequence. But I really wanted to 6 challenge the assumption that we should not be 7 concerned about the species specificity of the 8 retroviruses. In fact, you can induce tumors quite 9 readily in certain species quite diverse from any 10 species with RAS sarcoma virus that contains the ALV 11 envelope gene in subgroups like Subgroup D. So it is, 12 I don't think, a correct assumption to state that 13 viruses like the avian retroviruses are not of 14 concern. They could be. 15 DR. GRIFFITHS: I think David's point -- 16 did you introduce yourself David? That is David Onions 17 for the record. 18 DR. RABINOVICH: Thank you. 19 DR. GRIFFITHS: But I think the point made 20 is very important that although we assume that some of 21 these things are innocuous, I think it is much better 22 if you can get rid of them to have them out of the 23 way, and I think this is the point really. Because 24 there are a lot of interactions going on and there is 25 a lot of new information coming through, which I am 58 1 sure will be discussed over the next few days. 2 DR. RABINOVICH: Thank you, Dr. Griffiths. 3 I would like to introduce our next speaker, who every 4 single presenter and panel chair at this conference 5 knows well because he has been contacted individually 6 by him. It is Dr. Andrew Lewis from the Center of 7 Biologics and Review at the FDA. 8 DR. LEWIS: Thank you, Regina. I think 9 pestering is a better word perhaps. 10 DR. RABINOVICH: Persistence, leadership. 11 DR. LEWIS: We have had a lot of 12 conversations with everybody. As you've heard from 13 Dr. Hayflick and Dr. Griffiths, until the end of the 14 1980's, the use of all types of neoplastic cells for 15 the production of biologicals was controversial. With 16 the development of the defined 100 picogram limits of 17 residual DNA, the development of the concept of viral 18 clearance, and the World Health Organization's 19 acceptance in 1987 of interferons and monoclonal 20 antibodies that were produced in tumor cells, the 21 issues regarding the use of neoplastic cells for the 22 production of purified biological products were 23 resolved. However, the use of neoplastic cells as 24 substrates for live virus vaccines continue to be 25 controversial. 59 1 A number of factors are motivating the 2 need to reconsider using neoplastic cells for vaccine 3 development, and I have listed these factors on the 4 next slide. These factors include the development of 5 the whole virus or traditional vaccines to HIV, 6 bioengineering approaches to viral attenuation to 7 vaccine development, the rapid development of vaccines 8 to emerging viruses such as the H5N1 and the H9N2 9 influenza viruses, significant progress in 10 understanding carcinogenesis and in detecting 11 adventitious agents, and finally the very successful 12 experience with the highly purified biologicals that 13 were derived from neoplastic cells. 14 Now conflicting with the need to use 15 neoplastic cells as vaccine substrates are the 16 regulatory concerns over the possible risks associated 17 with vaccine manufacturers in these cell types. I 18 think it is no secret that issues regarding vaccine 19 safety receive a very high level of public attention. 20 It is also no secret that public confidence in vaccine 21 safety is a critical component for the success of 22 immunization programs that are important to the public 23 health. 24 The scientific and regulatory challenge is 25 to develop the concepts and the technologies that can 60 1 be used to assess the risks perceived to be associated 2 with neoplastic cell substrates in a manner which 3 sustains public confidence in vaccine safety. To 4 achieve this goal, we at CBER believe that it is 5 essential that an approach be developed which can be 6 used to objectively and critically assess using state 7 of the art technology those issues associated with the 8 use of neoplastic cells as substrates for vaccine 9 manufacture. 10 The purpose of my talk tonight is to 11 outline an approach that CBER is proposing to 12 establish an objective and systematic means of 13 managing the regulatory concerns that are associated 14 with the development of vaccines in neoplastic cells. 15 To begin the discussion of this approach, I would like 16 to define exactly what we mean by the term neoplastic 17 cells. 18 For my talk, the term neoplastic cells 19 refers to immortalized cells that are derived from 20 either tumors or from transformation of instant tissue 21 culture, and these cells can either be tumorigenic or 22 non-tumorigenic when injected into animals. 23 The approach that CBER has followed to 24 consider neoplastic cells as possible vaccine 25 substrates is based on the approach worked out in the 61 1 1980's to evaluate the use of continuous cell lines to 2 manufacture biologicals. Now as shown in the next 3 slide, this approach consisted of five components. 4 These components included identifying the issues, 5 developing theoretical and experimental models to 6 evaluate each issue, validating the models for issue- 7 associated risk evaluation, developing criteria to 8 evaluate the levels of risk, and finally discussing 9 these issues and the approaches in public meetings, 10 which we are doing over the next several days. 11 The process of developing the CBER 12 approach began with organizing and presenting a 13 discussion of the use of neoplastic cells as vaccine 14 substrates before CBER's Vaccines and Related 15 Biological Products Advisory Committee in November of 16 1998. Following this meeting, CBER drafted a proposal 17 which evolved from the Advisory Committee 18 presentations. This proposal represents the first 19 attempt in CBER to formally address the regulatory 20 issues associated with the use of neoplastic cells as 21 substrates for vaccine development. 22 Perhaps I should also point out that the 23 draft proposal which has been distributed in your 24 meeting package, as Dr. Rabinovich mentioned in the 25 introductory remarks, is not -- and I will emphasis 62 1 the word not -- an official, approved FDA document 2 offering guidance on the use of neoplastic cell 3 substrates. The issues and concerns that were 4 presented to the Advisory Committee were developed in 5 greater detail in this CBER draft proposal, and these 6 issues and concerns were use to develop the agenda for 7 this meeting. And I am going to review them for you in 8 the next series of slides. 9 Now in this series of slides, I have 10 attempted to organize the issues in a common format. 11 The title of the slide represents the issue of 12 concern. The first bullet represents the cell 13 substrates that are involved. The second bullet 14 illustrates the issues that are responsible for the 15 concern. And the third bullet is a very brief summary 16 of data that documents why the concern or why the 17 issue is in fact a concern. 18 Now concern 1 represents tumor cell 19 contamination. With regard to the presence of tumor 20 cells as contaminants, at first glance this concern 21 appears to be trivial, as viable cell substrates are 22 almost always or always removed from viral vaccines. 23 However, if cells from human tumors are used as a 24 manufacturing substrate, the issue becomes a bit more 25 complex. As noted in bullet 3, the inoculation of 63 1 human tumor cells into humans has in fact produced 2 tumor allografts on some occasions. Thus, for 3 regulatory purposes, not only will it be necessary to 4 remove or eliminate the cell substrate, it will 5 probably be necessary to document by validated 6 procedures the efficiency with which the manufacturing 7 process removes tumor cells from the final product. 8 The second concern is a concern with 9 adventitious agent contamination. All cell 10 substrates, as Dr. Griffiths has pointed out, are 11 subject to contamination with adventitious agents. 12 Such agents might infect vaccine recipients, and the 13 ample history of substrate contamination with viruses 14 and with other agents has provided reason enough for 15 this concern. Without saying much more about this, 16 Dr. Phil Minor is going to review this topic in 17 considerable detail during the Friday meeting. 18 Concern 3 is the concern with cell 19 substrate DNA contamination. Most vaccines consisting 20 of whole varions contain some level of residual DNA 21 from the cell substrate in which the vaccine was 22 manufactured. There has been considerable discussion 23 over the years as to the possibility that residual DNA 24 could be the source of activated oncogenes or other 25 manifestations of a neoplastic process and might 64 1 result in the transfer of neoplastic activity to 2 vaccine recipients. Even though some live virus 3 vaccines contain microgram quantities of DNA from 4 their diploid cell substrates, there is no evidence 5 that any adverse event has resulted in an exposure to 6 residual cell DNA. While diploid cell DNA is thought 7 to be free of all the accoutrements of neoplasia, 8 immortalized cells and tumor cells contain various 9 combinations of genetic alterations that contribute to 10 their neoplastic state. For this reason, we believe 11 it is necessary to revisit the possible risk 12 associated with the residual DNA from neoplastic cell 13 substrates. 14 Now in addition, tumor cells can carry 15 part of or the entire genome of oncogenic viruses. 16 Furthermore, the need to produce or develop retrovirus 17 vectors and retrovirus vaccines means, as Dr. 18 Griffiths has pointed out, that retrovirus proviruses 19 can also be present in any substrate DNA that remains 20 in retrovirus vaccines that are manufactured in 21 neoplastic cells. These possibilities generate the 22 concerns produced by the presence of residual cell DNA 23 in viral vaccines manufactured in neoplastic cells and 24 are the reasons why this topic is covered in the 25 session on Thursday. 65 1 Concern 4 represents cell protein 2 contamination. Tumor cells produce a variety of 3 cytokines and lymphokines and other biologically 4 active proteins. Some of these proteins have the 5 capacity to produce local and systemic reactions. In 6 addition, random selection of cells from the human 7 population as well as alterations that occur with some 8 frequency in the genomes of neoplastic cells might 9 result in the presence of abnormal PRP proteins, and 10 whether such proteins might be able to produce 11 spongiform encephalopathy so far as I am aware is not 12 really known, but the theoretical possibility that 13 such events can occur means that we should be 14 addressing these concerns and discussing them at this 15 meeting. 16 Concern 5 is the concern with viral/viral 17 and viral/cellular interactions. In cells infected 18 with more than one virus, a variety of interactions 19 are known to occur between the replicating viral 20 genomes. These interactions include genetic 21 recombination, genome reassortments, pseudotyping or 22 sequestering of the genomes of one of the viruses 23 within the capsid of another virus, and finally the 24 parasitism of the genome of large viruses by the 25 entire genomes of the smaller viruses. If vaccines 66 1 are manufactured in cells containing adventitious 2 agents, each of these processes could be associated 3 with the transfer of non-viral components to vaccine 4 recipients or their non-vaccine viral components. 5 Evaluating the possibility that these events could 6 occur and developing methods to search for novel virus 7 derivatives and establishing baselines or threshold 8 levels considered to be necessary to document the 9 absence of such novel agents will occupy two sessions 10 of the workshop. 11 The final and sixth concern is the concern 12 with genomic instability. In contrast to diploid 13 cells, immortalized cells and especially tumor cells 14 evolve by changing the configuration and expression of 15 their genomes. The question, and therefore the 16 concern, raised by this characteristic of neoplastic 17 cells is whether such changes could over time mobilize 18 occult viruses or proviruses or result in the 19 production of biologically active proteins resulting 20 in adverse events. Although we are not aware of any 21 evidence that such events have occurred, the ability 22 of tumor cells to evolve over time is well established 23 and we suggest that the consequences of neoplastic 24 evolution for the safety of vaccines manufactured in 25 neoplastic cells needs to be considered. 67 1 Now once the issues and concerns 2 associated with the possible use of neoplastic cells 3 as vaccine substrates were identified, it was 4 necessary to consider an approach that would allow the 5 possible risk posed by these issues to be assessed. 6 As issues of risk assessment are best evaluated using 7 quantitative data and as current technology provides 8 the opportunity to measure most biological events, we 9 are attempting to develop a systematic, quantitative 10 approach to assessing the concerns associated with 11 neoplastic cell substrates using what we have 12 designated as a defined risk approach algorithm. This 13 algorithm is presented in the next slide. 14 The basic features of this algorithm 15 include assessing the levels of risk posed by each 16 issue quantitatively, establishing the probability of 17 a worst case scenario for each issue, using the data 18 to evaluate the risk individually and cumulatively, 19 and then using the data to assess the relative risk of 20 the product. In the next talk, Dr. Phil Krause is 21 going to discuss risk factors and the use of 22 experimental data and the application of this 23 algorithm to specific concerns. 24 I would like to end my talk with sort of 25 a short story. In June of 1998, I was visiting with 68 1 two friends from medical school. Both of these 2 friends are practicing physicians and have been in the 3 business of delivering health care for many years. 4 During one of the many conversations about 5 professional issues that always come up when we get 6 together, I asked my friends if they would consider 7 giving vaccines produced in cancer cells to their 8 family and their patients. There was a slight pause 9 in the conversation and then one of them answered with 10 a seven-word remark that was both an expression of 11 confidence as well as a reminder of responsibility. 12 When this remark is committed to paper, it also turns 13 out that it can be somewhat poetic as well as 14 prophetic. And since it has been following me around 15 for the past 14 months, I thought it would be 16 appropriate to share it with you as my last slide. 17 In response to the question about using 18 vaccines made in cancer cells, my friend simply 19 replied, "If the FDA says they are okay." That 20 question is being addressed at this meeting. The 21 question that is actually being addressed at this 22 meeting is how to find out if they are okay. 23 Beginning with the controversy over the distribution 24 of non-defective adeno SV40 hybrids that were 25 developed in my laboratory in the late 1960's, with 69 1 the Oscillimar (Phonetic) Conference on recombinant 2 DNA in 1975, and with xenoscience transplantation 3 issues since I have been with CBER, I have observed 4 that when technological advances make it necessary to 5 address precedent setting issues, it is essential to 6 first bring together knowledgeable scientists and 7 interested individuals from other disciplines to 8 discuss the issues and the data that are available for 9 informed decision making. Following these initial 10 instructions, consultations, collaborations and 11 cooperation among the interested parties, we have 12 usually been able to resolve many or most of these 13 issues. 14 Technological advances are once again 15 asking us to address a precedent setting issue in the 16 field of vaccinology. With this meeting on cell 17 substrates, we are in the process of reviewing and 18 evaluating the data that are available for informed 19 decision making. And I would like to conclude by 20 saying that those of us at CBER who have been working 21 on cell substrate issues believe that through 22 consultation, collaboration and cooperation with our 23 colleagues in academia, industry and the regulatory 24 community that it should be possible to develop an 25 effective approach to evaluate and manage the use of 70 1 neoplastic cell substrates for vaccine manufacture. 2 Thank you. 3 DR. RABINOVICH: Are there any questions 4 specifically for Dr. Lewis? Please identify yourself. 5 PARTICIPANT: Dr. Tevethia from Penn 6 State. In the study that you cited where the 7 allografts were accepted in the patients, you didn't 8 mention, did they progress to the tumor stage or did 9 they regress or disappear? 10 DR. LEWIS: I am sorry? 11 PARTICIPANT: Did they progress? 12 DR. LEWIS: Yes, they did. These studies 13 -- if you look at that literature, Tev, it goes back 14 to the late 1940's and early 1950's most of it, and 15 the cell line that was particularly problematic was 16 called hep-3. And most of these studies were done in 17 patients with terminal cancer. And so it is very hard 18 to assess the implications of that for a more modern 19 problem. But preceding that information, there were 20 several reported cases in the literature, and also 21 cited by Dr. Luci Gross in his book "Oncogenic 22 Viruses". So reviews -- this is where most of his 23 information came from. But there were several cases 24 in the literature in which surgeons had actually 25 grafted themselves with a tumor cell from patients who 71 1 they were operating on in the 1920's and 1930's. So 2 it is an unusual phenomenon, but it is not unheard of. 3 And I think when we are talking about vaccines that 4 are going to go into possibly millions of people, we 5 have to consider the breadth of the experience that 6 the population would be subjected to. 7 PARTICIPANT: Dr. Fried from ICRF. How 8 many cells were injected to get these nodules? 9 DR. LEWIS: Mike, a lot of those things 10 were done not with cells but with explants. That was 11 before the days when trypsin was used to break up 12 cells. So they would just take a little piece of the 13 tumor with a trocar and stick it in. The tumors were 14 generally several mm, but it was not -- that is just 15 a rough estimate because this is a fragment. Of 16 course with the surgical -- when the surgeons 17 inoculated themselves, there is no way to know. 18 DR. HAYFLICK: Actually, many of those 19 studies were done by Chester Savin. 20 DR. LEWIS: Yes, and Alice Moore. 21 DR. HAYFLICK: And Alice Moore. And they 22 were done with prisoner volunteers at the state 23 penitentiary in Ohio. So they were not terminal 24 cancer patients. 25 DR. LEWIS: Well, yes. I went back and 72 1 looked at that data, Dr. Hayflick, and the prisoners 2 uniformly rejected the tumors. 3 DR. HAYFLICK: Correct. 4 DR. LEWIS: Yes. 5 DR. HAYFLICK: Correct. And I think that 6 is important to know. 7 DR. LEWIS: Yes. 8 DR. HAYFLICK: However, when we inoculated 9 HeLa into terminal human cancer patients, the HeLa 10 cells did indeed grow to the point that they had to be 11 extirpated. 12 DR. LEWIS: Yes. 13 DR. HAYFLICK: I also should point out 14 that hep-3 later turned out to be HeLa. 15 DR. RABINOVICH: Thank you very much, Dr. 16 Lewis. Last but not least, Dr. Phil Krause has the 17 unenviable position of being between you and bed 18 before tomorrow morning. So I welcome him and his 19 presentation of setting the stage for conceptual and 20 experimental approaches to address product safety 21 issues raised by novel cell substrates. And he 22 becomes our guinea pig for testing of the computer 23 system. 24 While he is doing that, let me remind all 25 speakers that the manuscripts are to go to our 73 1 colleagues at IQ Solutions, and that there is, I 2 believe, still space at the late breaker session for 3 tomorrow evening which was thought up by Dr. Lewis, 4 but we welcome comments, arguments and particularly 5 data that people may want to share at the late breaker 6 session. 7 DR. KRAUSE: What I want to talk to you 8 about tonight is the defined risks approach that Andy 9 just alluded to. And the idea is that we would like 10 to develop a quantitative approach that would enable 11 us to use laboratory data to assess risks associated 12 with cell substrates. The change in precedent against 13 the use of neoplastic cells as vaccine substrates 14 needs to be based on objective scientific data that 15 can be used to evaluate levels of risk. 16 This approach is an example of science- 17 based regulation in which FDA attempts to gather the 18 relevant scientific data before making a decision. 19 And as Andy pointed out or as Dr. Lewis pointed out, 20 this could be implemented by assessing the level of 21 risk posed by each issue quantitatively, establishing 22 the probability of a worst case scenario for each 23 issue, using data to evaluate the risks individually 24 and cumulatively, and using the data to assess the 25 relative risk of the product. So I am going to go 74 1 through these four points in this talk. 2 The first of which is to quantitatively 3 assess the level of risk posed by each issue. I am 4 also going to go through a simple example of how the 5 defined risks approach can be applied to an issue. In 6 order to quantitatively assess the level of risk posed 7 by each issue, it is necessary to first assess what 8 are the issues associated with the plausible risk to 9 vaccine recipients, and Andy Lewis went through some 10 of these, evaluate the need for and availability of 11 quantitative data to assess the risks in a regulatory 12 setting, to evaluate factors that could increase or 13 decrease this possible risk, and to consider 14 development of assays that could be used in evaluating 15 specific products and the general issue. 16 As it happens, these are four of the major 17 points which are going to come up in a recurrent way 18 in the panel discussions later on during this meeting. 19 So much of the discussion in this meeting is going to 20 focus on this first part of the defined risks approach 21 algorithm. 22 Now I am going to briefly discuss how this 23 algorithm could be applied to the issue of infectious 24 residual DNA in biological products. And I don't mean 25 to prejudice the discussions that will occur later in 75 1 the meeting, but I think it would be useful to go 2 through these steps as we have considered them so far 3 for this issue to illustrate the kind of analysis that 4 we think will be valuable for all of the issues to be 5 discussed at the meeting. 6 You have already heard some of this from 7 Dr. Hayflick and from Dr. Griffiths, but according to 8 the recently revised WHO requirements for the use of 9 animal cells as in vitro substrates for the production 10 of biologics, the amount of cellular DNA in biological 11 products should be limited to 10 nanograms per dose. 12 This is an increase from a previously established 13 limit of 100 picograms per dose. This limit is meant 14 to apply to continuous cell lines but not to products 15 given orally or to products derived from microbial, 16 diploid or primary cell culture systems. The 10 17 nanogram figure was derived by considering data and 18 theoretical calculations regarding the tumorigenicity 19 of injected DNA. 20 While CBER is attentive to WHO guidelines, 21 CBER evaluates products on a case-by-case basis in 22 determining appropriate limits for cell substrate DNA. 23 Moreover, it should be noted that for live viral 24 vaccines and other less purified products, it may not 25 be possible to limit the total DNA to 10 nanograms. 76 1 Thus the question is raised, what data would be 2 required to provide assurances that this or any 3 specific limit would be appropriate for vaccines 4 produced in novel cell substrates, including 5 neoplastic cells? 6 The first question is whether there is a 7 plausible risk or not. Viral genomic DNA is 8 infectious when injected into animals. Moreover, 9 tumor cells and primary cells may contain virus 10 genomes and may harbor latent viruses. And DNA from 11 cells used to produce retroviral or DNA-viral vaccines 12 may contain viral genomes as a by-product of vaccine 13 production. Thus, based on the data, it would appear 14 that under some circumstances there could be an 15 infectious risk from residual DNA. 16 Several factors may influence an 17 assessment of the tumorigenicity or infectivity risk 18 associated with residual DNA. These include the total 19 quantity of DNA in the vaccine, the number of doses to 20 be given, the size of the DNA, sequence-related 21 properties of the DNA, for example whether it encodes 22 a virus, the number of copies of potentially 23 infectious sequences per cell, and the state of the 24 DNA, which I take to mean such factors as whether it 25 is chromatin-associated, whether it is integrated into 77 1 a cellular genome, whether it is linearized or 2 circular, et cetera. For many of these kinds of 3 considerations raised on the slide, although it is 4 likely that they have an effect on the ultimate 5 infectivity or tumorigenicity of the DNA, these 6 studies have not in general been performed in a 7 quantitative fashion that would enable us to apply a 8 quantitative risk assessment model. 9 This slide presents the quantitative data 10 that we are aware of regarding the potential 11 infectivity of cloned or purified viral genomic DNA. 12 Information on tumorigenicity is also presented based 13 on the assumption that viral genomic DNA is at least 14 as infectious as it is tumorigenic. In these 15 experiments, cloned or purified genomic DNA was 16 injected directly into various animal models. On the 17 right-hand side of the table, I have calculated the 18 theoretical risk associated with a product that 19 contains one microgram per dose of cellular DNA that 20 contains a single genome per cell. This calculation 21 accounts for the dilution of the viral genome and the 22 cellular genome, and assumes that the viral genomic 23 DNA is as infectious or tumorigenic when incorporated 24 in cell substrate DNA as it is when it is linearized 25 and injected directly. 78 1 The estimated risk of infection associated 2 with the theoretical product would thus range from as 3 high as 1 in 8,000 for polyoma virus DNA to 1 in tens 4 of millions for other DNAs. An improved understanding 5 of the relative infectivity of different types of DNAs 6 might assist in developing tests to ensure that 7 unacceptable quantities of infectious DNAs are not in 8 biological products produced in novel cell substrates. 9 Another key point for discussion at the 10 meeting will be the development of appropriate assays 11 or methods to better evaluate these risks. It should 12 be pointed out that direct assessment of the 13 tumorigenicity and infectivity of residual DNA 14 associated with a particular product may be 15 impractical. This is because in order to achieve 16 meaningful safety margins, very large quantities of 17 DNA would need to be purified and tested in multiple 18 animal models for infectivity and tumorigenicity. The 19 example on this slide, for example, if a dose of a 20 product contained one microgram of residual cellular 21 DNA, assessment of a million doses in a single type of 22 assay would require testing of an entire gram of DNA. 23 This is why I believe it is important to take a more 24 general approach and understand the potential 25 underlying infectivity and tumorigenicity of different 79 1 types of DNA. 2 Based on the data that I have presented, 3 one might conclude that additional work could be done 4 to achieve a more complete understanding of this 5 issue. This might include studies of how to detect 6 latent genomes, including for unknown viruses, how to 7 quantitatively assess the effect of various factors on 8 DNA infectivity, and improved assessment of DNA 9 infectivity for different types of viruses. 10 Now it may be helpful to think about this 11 issue in the context of a hypothetical example. 12 Suppose an unpurified, live, attenuated viral vaccine 13 is grown in a cell line that contains 50 copies per 14 cell of a latent virus. This is the approximate 15 number of copies of HPV DNA that are in HeLa cells. 16 If there were 10 nanograms of residual DNA per dose, 17 which is the current WHO recommendation, and if two 18 doses were recommended per child, as for instance is 19 the case with MMR vaccine, and the infectivity of 20 viral DNA in the vaccine were comparable to that of a 21 purified polyoma virus DNA, we can calculate the 22 theoretical infectivity risk from the DNA in this 23 product, which I proceeded to do on this slide. 24 Multiplying these numbers together with the 25 infectivity of polyoma virus DNA yields a theoretical 80 1 infectivity risk of 1.2 infections per 10,000 2 children. For a vaccine that is universally 3 administered to the U.S. birth cohort of 4 million 4 children, this would represent about 500 infections 5 per year, clearly an unacceptable rate. Of course, 6 this assumes that the DNA is as infectious as polyoma 7 virus DNA. It assumes that this DNA is as infectious 8 as residual DNA in a vaccine as it is when it is 9 cloned and purified. And it doesn't account for the 10 potential effect of DNA fractionation or other 11 measures that might reduce its infectivity. Based on 12 the limited data I showed you, retrovirus DNA may be 13 about 1000 to 10,000-fold less infectious than polyoma 14 virus DNA. Thus, this calculation is very dependent 15 on a clear understanding of the infectivity of the 16 viral DNA and a clear understanding of what type of 17 viral genomes a neoplastic cell could harbor. 18 The example I just went through could be 19 thought of as a worst case scenario. And in fact, the 20 second element of the defined risk approach algorithm 21 is to establish the probability of a worst case 22 scenario for each issue. Thinking of these kinds of 23 issues in the context of their worst case is justified 24 by the need for public confidence in vaccines and 25 their safety, which requires that we take a very 81 1 conservative approach. This includes identification 2 and discussion of all of the issues, determining which 3 can be dismissed, which are relevant, and which are 4 amenable to a resolution with additional data. 5 The third element of the defined risk 6 approach algorithm is to use data to evaluate the 7 risks individually and cumulatively. In this meeting, 8 we will be individually considering a broad variety of 9 potential risks posed by neoplastic cell substrates. 10 Each of the meeting sessions is devoted to discussion 11 of these potential risks. The meeting starts tomorrow 12 morning with the consideration of mechanisms of 13 oncogenicity because this is a final common pathway 14 for many of the potential concerns with neoplastic 15 cells. We then consider virus/virus and virus/cell 16 interactions, residual DNA, and adventitious agents. 17 We then turn to a discussion of how cell substrates 18 might be designed or chosen to minimize these types of 19 concerns. 20 In lieu of a poster session, there will be 21 a session on miscellaneous topics on Thursday evening 22 to permit participants to present additional relevant 23 data. The final panel discussion on Friday afternoon 24 will then summarize the results of the meeting. It is 25 our hope that a comprehensive discussion of these 82 1 individual potential risks will help CBER also 2 consider them cumulatively. 3 The fourth element of the defined risks 4 approach algorithm is to use the data to assess the 5 relative risk of the product. Potential risks must be 6 placed in the context of the potential benefits of the 7 product. This is something which may need to be 8 considered by CBER on a product-by-product basis. 9 These risks also must be placed in the context of 10 risks already found acceptable and unacceptable by 11 society, and this can be a changing target. For 12 example, society has been willing to accept the one in 13 a few million or so risks of a vaccine associated with 14 paralytic polio for many years, but more recently it 15 appears less willing to do so. 16 We also hope the discussion will include 17 some consideration of the relative risks of using 18 different types of cell substrates, including the role 19 of species in cell type, whether it be primary, 20 diploid or neoplastic, and for neoplastic cells, what 21 the role of the transforming event and the known 22 history of the cell plays in the safety assessment. 23 I would like to close by putting this 24 discussion into the context of the real world. CBER 25 currently has applications to study the use of various 83 1 types of neoplastic cells in vaccine development. 2 Some of these products could be amenable to production 3 in other types of cells, but others might not. By the 4 time CBER gets an application, the manufacturer 5 usually has already made a commitment to one cell 6 substrate or another. It is our hope that this 7 meeting will both raise awareness of these issues 8 among manufacturers in the academic community and will 9 provide CBER with a sound scientific foundation to 10 answer these critical regulatory questions, which 11 include what types of data regarding these cells are 12 necessary to ensure that vaccines, including 13 unpurified ones that are produced in them, are safe. 14 And are there some vaccine cell substrate combinations 15 for which current technology simply precludes an 16 adequate safety assessment. I thank you very much and 17 good night. 18 DR. RABINOVICH: I think we should 19 entertain some questions. I think your approach will 20 be provocative and will raise points for discussion 21 throughout the meeting. Are there any questions at 22 this time? Dr. Rubin? 23 PARTICIPANT: Well, I just thought this 24 was an opportunity to raise a question about DNA that 25 you, yourself, raised in a recent paper, of which you 84 1 were a co-author and which we were just talking about 2 and which I haven't seen proposed for discussion at 3 any point, which is that DNA are double-stranded RNA 4 of as small as 25 base pairs when incorporated into 5 the DNA of the cell and actually not incorporated even 6 in the cytoplasm, I take it, of the cell can cause the 7 production of protein molecules -- what are they MLH-1 8 and 2 or something like that -- that are supposed to 9 be very active in autoimmune reactions. And I wonder 10 how much you have thought about that and how much do 11 you think that is relevant to the considerations that 12 we have here. It is a very surprising development to 13 me that at least as little as 25 base pairs can be 14 inserted and produce a fairly large biological effect. 15 DR. KRAUSE: Yes. Of course, that was a 16 study in which these small molecules were transfected 17 into cells and the cells started to display some of 18 the characteristics of antigen presenting cells. And 19 then the question then is, is that something which 20 could subsequently be related to autoimmune responses 21 or things like that. I think those are legitimate 22 kinds of questions. I know that the world has a 23 fairly long experience with unpurified viral vaccines 24 which do contain eukaryotic DNA as residual DNA in 25 quantities greater than 10 nanograms. In particular 85 1 I am referring to WI38 or MRC5 DNA, which appear to be 2 safe from that perspective. I was on the committee 3 that licensed the -- or actually I chaired the 4 committee that licensed the chickenpox vaccine, and 5 that was actually an issue that we considered at that 6 time as well, and actually looked among recipients of 7 the vaccine for evidence of an autoimmune response 8 associated with the DNA that was included in that 9 vaccine. Actually, we didn't look, we asked the 10 company to look and they did not find one. Of course, 11 in the context of DNA vaccines, we are talking about 12 injecting even larger quantities of DNA into people 13 which then -- and one might imagine that this response 14 is at least related to some of the responses 15 associated with the DNA vaccine. 16 I think in the context of the meeting, we 17 had envisioned the discussion covering mostly issues 18 that are specific to neoplastic cells, and at least in 19 our initial thinking about the issue didn't think that 20 this was an issue, given the small sizes of the 21 molecules that were involved, that would be likely to 22 be more of a problem with neoplastic than with other 23 cells. So we decided not to emphasize it at this 24 time, in part because the plate is already very full. 25 But you raise a very interesting point. 86 1 DR. RABINOVICH: Please identify yourself. 2 PARTICIPANT: Fried from ICRF London. Do 3 you think it is fair in your polyoma example to use 4 linear DNA which can recircularize and then go into 5 the mouse and replicate and amplify and compare it to 6 integrated, where it can't come out and amplify? 7 DR. KRAUSE: You know, I guess the 8 question isn't necessarily is it fair, but the 9 question is what data does one need to know or to have 10 in order to be certain that a given level of residual 11 DNA from a certain cell type is safe. Now the 12 experiment that I -- the trouble is that not a lot of 13 information is available on this. The only study that 14 I am aware of where these data come from is that which 15 was published by Marc Israel and Mel Martin's lab some 16 20 or more years ago I think. And what they did is 17 they looked at simply circularized DNA, and that was 18 more infectious than linearized DNA. And then 19 depending on where you linearized the DNA, you then 20 became more likely to either get infection or tumor. 21 But I think these kinds of experiments could certainly 22 legitimately be done in other ways, whether 23 linearizing with an enzyme which leaves overhangs to 24 yield a different result than linearizing with an 25 enzyme which leaves blunt ends and thereby potentially 87 1 precluding recircularization. But I think -- 2 PARTICIPANT: It doesn't preclude it. I 3 mean, it will recircularize even when you cut -- 4 DR. KRAUSE: Okay, reduce the likelihood. 5 PARTICIPANT: I mean, you could put it in 6 a plasmid, so you have a small piece of DNA and it is 7 integrated in that, and then you can compare that to 8 say cellular DNA, where the virus can't get out and 9 replicate itself. 10 DR. KRAUSE: You are raising a very 11 legitimate point. To my way of thinking, these are 12 the kinds of questions which need to be answered. The 13 question is do they need to be answered on an 14 intuitive basis or do they need to be answered on an 15 experimental basis. 16 DR. RABINOVICH: Thank you. I think at 17 this point, for those of us who have to coerce our 18 children to go to school an hour earlier than usual, 19 we will go ahead and close the session tonight and 20 expect everyone to be here awake and coherent at 8:00 21 in the morning to start. Thank you very much. 22 (Whereupon, at 9:53 p.m., the session was 23 concluded.) 24 25