Tobacco Documents Online

"Cancer Research The Official Organ of the AMERICAN ASSO IATiON FOR CANCER RESEARCH, INC. Contents Volume 25 • No. 8 • September 1965 1199 Anthony J. Sbarra, William Shirley, Ratnam J. 1238 Selvaraj, Ronald J. McRip]ey, and Ernest Rosenbaum. The Role of the Phagocyte in Host-Parasite Interactions. III. The Phagocytic Capabilities of Leukocytes from ~Iyeloproliferative and Other Neoplastic Disorders. 1207 1213 1219 1225 1232 Estelle R. Lefkowitz, William A. Creasey, Paul Calabresi, and Alan C. Sartorelli. Clinical and Pharmacologic Effects of Combinations of 6-Thioguanine and Duazomyein A in Patients with Neoplas'tie Disease. F. Molnar and R. Daoust. Nueleocytopla, mic Ratios in Different Populations of Rat Liver ParenchymuI Ceils during Azo Dye Carcino- genesis. Thomas E. Webb, Giinter Blobe], Van R. Potter, and Harold P. Morris. Polyribosomes in Rat Tissue~ II. The Polyribosome Distribution in the Minimal Deviation Hep~toma.~. Y. T. Li, S. C. Li, and M. R. Shetlar. Carbohydrate Content in Subcellular Fractions of Rat I.iver and of Walker 256 Carcinosarcoma. N. Bruchovsky, A. A. Owen, A. J. Becker, and J. E. 'Fill. Effects of Vinblastine on the Proliferative Capacity of L CelLs and Their Pr¢~res~ Through the Divi~i~m Cycle. T. J. Bardos, J. L. Ambrus, Z. F. Chmielewicz, A. G. Penny, and C. M. Ambrus. Biochemical Parameters of Neoplasia. I. "Priming" Ac- tivity of Deoxyribonucleic Acid from Normal.and Neo- plastic Human Tissues in a Deoxyribonucleic Acid Polymerase System. 1244 R. J. Fiei, T. J. Bardos, Z. F. Chmielewiez, and J. L. Ambrus. Biochemical Parameters of Neoplasfa. II. Some Macro- molecular Properties of ]~eoxyribonucleie Acid from Normal and Neoplastic Hu~, tan Tissue. 1254 G. E. Foley, A. H. Handler, P. M. Lynch, S. R. Wolman, C. S. Stulberg, and 1I. Eagle. I,o.~ of Neopla..t~c Properties in Vitro. II. Observations on KB Sublines. 1262 Fernando Marroquin and Emmanuel Farber. The Binding of 2-Aeetylaminofluorene to Rat Liver Ribonucleic Acid in Vivo. 1271 Symposium: Conference on Epidemiologic Approat~hes to Cancer Etiology 1392 Axmouncemeats T154351310

CANCER RESEARCH MIC~.~. B. SHmxm, Edi!or CHAI~ES T. A~HWOSTH NATH~NIEL I. BERLIN OSCAR BODANSKY M~CH~EL J. BaP.NN~N IRWIN Do J. BROSS K~LLY H. C~rroN Editorial Advisory Board Eu~.~,E D. Dxr K~N~NT~ B. DEO~m LEON DMOCHOWSKI I-L~Ns L. F,~" JACOB FUR~ ~aav V. G~L~Om ~ED GELDHORN CANCER RESEARCH is sponsored by The American Association for Cancer Re- search, Inc.; The Anna Fuller Fund; The Jane Coffin Childs Memorial Fund for Medi- calResearch; The Elsa U. Pardee Foundation; The American Cancer Society; the Na- tional Cancer Institute, Public Health Service; and the Damon Runyon Memorial Fund for Cancer Research, Inc. NOTICE TO MEMBERS OF THE AMERICAN ASSOCIATION FOR CANCER RESEARCH O~cers .for 19~1~-66 President: Jos~.PH H. BUaCH~.~rAL, Sloan-Kettering Institute for Cancer Re,arch, New York~ N. Y. 1~21 V~c~-Pr~$: H~s~v S. K~L~, S~anford Medical ~nter, Pale Alto, Calif. ~¢~lary-Tre~ur~r: Huu~ J. Ca~cs, Institu~ for Cancer Research, Fox Chic, Philadelphia, Pa., 19111 T~e annual dues of a~tlve members of the American A~sociation for Cancer ~sea~ch include subscription t~ the journal CANff~R RE~EAR~H. Payment of dues and change~ of addre~ mem~rs of the Association should ~ sent promptly to Dr. Creech. NOTICE TO SUBSCRIBERS CANCER RE,~EARCH is published monthly (except March) for Cancer Research, Inc., b~' The Williams & Wilkins Co. Subscriptions are payable in advance. Subscriptions include Proceedmg~ ~f American Association for Cancer Research, published in March each year. Subscription rates in U.S.A. and pocscssions: 1 year, $15.00; 2 years, $30.00; 3 years, $45.00. Canada: 1 year~ $15.50; years, $31.00; 3 years, ~.50. All other countries and Pan American Postal Union: 1 year, $16.00; 2 years, $32.00; 3 years, ~,8.00. Single copies: $2.50. Subscriptiane will be entered on a volume basis only. Business communications, remittances (in United States currency or its equivalent), and sub- scriptions should be addressed to The Williams & Wilkins Co., 428 E. Preston St., Baltimore, 21~2. All other communications should be addressed to: Dr. Michael B. Shimkin, Cancer Research Editorial Office, Fels Research Institute, Temple University Mcdieal School, Philadelphia, Pa., 19140. Claims for missing numbem should be made within the month following the regular month of pu.biic.a.tion. Th.e publishers .expect to supply missing numbers free only when losses have been sus- tg. med,z,n, transit an.d w.hen the reserve stoc_k will permit. Changes of address should be reported to the_ pumisher and meat postmaster immediately. All changes of address should provide ~o~h the old and the new address. Allow ~iz t~ee~'.~ for t~ change. No responsibility is accepted by the Editors, by Cancer Research, Inc., or by the publishers of CANCER RESEARCH for opinions expressed by contributors. Second-cla~s postage pakl at Baltimore, hlaryland. C~pyright 1~ by Cancer Re~earch, Inc. T154351311

SYMPOSIUM SPONSORED BY THE AMERICAN CANCER SOCIETY with the assistance of THE WHITING FOUNDATION OF FLINT,/VHGHIGAN Epidemiologic Approaches to Cancer Etiology Held at the MAIN GRILL, WESTCHESTER COUNTRY CLUB, P.YE, NEW YORK MAY 20-22, 1964 PROGRAM L~sT oF I~ARTIC~PAN~S ................................................................ 1273 L~s~Ea Bar.stow. Introductory Rettmrks ............................................... 1275 Vmus F~croas Thom~ Francis, ~r., WAlLaCE P. ROWE. A Su~ey of %he Tumor Vies Problem from an Epidemiologic Standpoint ........................................................................ 1277 W. R~Y B~xAN. Focal Disc~sion of: A Su~ey of the T~or V~ Prob]em from an Epi- demiolog[c Standpoint .............................................................. 12~ EDWIN H. L~N~E. Formal D~cussion of: A Su~ey of the Tumor Vires Problem from an Epidemiologic Standpoint ........................................................... 1286 Bm~N h~cM.<~oN. Fomal Discussion of: A S~ey of the T~or V~s Problem from an Epide~olo~c Standpoint, ........................................................... 1289 WAL~C~ P. ROWE. Summa~ of %he I~ormal Discussion on Vires Factors ................ 1291 Paul KotM, U~i~n ~.~s A. ~I~ER AND ELiZABetH O. ~LER. Natural and Synthetic Ohem~c~ in the Efiolo~ of Cancer ........................................................... 1~92 How~ J. Cu~T~s. Fo~l D~c~s~on of: Somatic ~[uta~o~ and Caro~ogenes~s ......... 1305 ~OBERT E. EC~T. Focal D~s~s~on of: Nat~a] and Synthetic Che~cal Carcinogens in the Efiolo~ of Cancer ........................................................... 1311 NORTON NELSON. Fom~a~ D~scu~on of: Chemical and Pkys~cal C~c~o~ens ............. 1314 P~ KOTm. Summa~ of %he [nforma~ D~so~s~on on Phys~c~-Che~c~ Factors .......... 1317 1271 T154351312

1272 Cancer Re2zarch Vol. 25, September 1£65 Rupert. E. Billingh~m, C2.a~rrnar~ W.~t,XZR E. l-lr-szor~. Genetic Factors in the Etiology of Cancer ........................... 1320 ~-c~ P. OLdeR. Fo~ D~cu~ion of: Cancer ~ Man .............................. 1327 ~r~ ~I. Lz~. Fo~nal Discussion of: Genetic Fac~m ~ the Etiolo~ of Cancer: ~a Epidemiolo~ ~iew ............................................................. 13~ R~E~r E. B~G~. S~y of the Info~l Discussion of Gene~ie Fac~rs in the Etiol- ogy of Cancer ..................................................................... 1336 ~~ Fxc~o~ Ernst Stebb~s, Chair~n L~s~m Fo~Ds. Multiple Etiologic FaeMrs ~ Neoplastic Development ..................... 1~39 K~a B. DsO~t~. Fo~ul Dise~ion of: M~tiple Fae~m ~ Mouse Mamma~ Tumori- genes~ ............................................................................ 1348 ft. E. D~. Formal D~e~sion of: M~tiple Etiologic Faeto~ in Neopl~tie Developmen¢ ..... 1352 W~,~r~t M. HA~sz~. M~tiple Faetom: D~e~sion of S~n¢~¢ioal and Epide~ologie Aspeots.. 1356 E~s~ L. Sr~ms. 8u~a~ of I~ormal Diseu~ion on M~iple Fae~m ................. 1361 Robe~ ~. MambO, ~c~ B. S~N. Epide~olo~ of Cancer: Spatial-Temporal Aggr~a~ion .............. MI~SUO S~I. (Deliver~ by Minom Kurih~r~.) Focal Discussion of: Epidemiolo~ of Cancer: Spatial-Temporal A~r~ation ....................................................... 1375 ~o~s C~mnms~. Formal Discu~ion of: The Statistical Approach ~ the Etiolo~ of Mali~ant N~pl~ms .............................................................. 13~ A~a D. L~tvm. Formal Discussion of: Epidemiolo~ of C~eer: Sp~tial-TemporM Aggregatioa ....................................................................... 1384 Ro~Ea~ R. M~x. Smma~ of I~omaal Dis~u~ion on Spatial-Tem~ral Aggregation... 1387 Lxsw~ B~s~ow. Closing Remrks .................................................... 13~ T154351313

List of Participants in Epidemiology Conference RUPERT E. Br~LmQ~I, g,lember, The Wistar Institute, 36th Street~ at Spruce, Philadelphia 4, Pennsylvania L~sz~ BP~SLOW, Chief, Division of Preventive Medical Services, State of Callfomia Department of Public Health, Berkeley 4, California RAY BRY,~, Chief, Laboratory of Viral Oncology, Na- tional Cancer Institute, Nil=I, Bethesda, Maryland 20O14 JOH,,~N~S C~.~-~msE~, Director, Danish Cancer Registry and Department of Pathology, Finsen Institute, Strandboulevard 49, Copenhagen, Denmark HOWAP, D J. CURTIS, Chairman, Department of Biology, Brookhaven National Laboratory, Upton, Long Island, New York K~N~r~ B. D~O~m, Director and Professor of Zoology, Cancer Research Genetics Laboratory, University of California, Berkeley 4, California JOH~ E. DUNN, JR., Cancer Consultant, Bureau of Chronic Diseases, State of California Department of Public Health, Berkeley, California 94704 R. E. ECKAP~UT, Director, Medical Research Division, Esso Research and Engineering Company, P.O. Box 45, Linden, New Jersey 07036 Lss~. Fo~J~DS, Chester Beatty Research Institute, Royal Cancer Hospital, Fu]ham Road, London, S.W. England Tno~tAs FRANC~s, JR., Professor of Epidemiology and Chairman, Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan S~tXON Gm~HA~I, Associate Cancer Research Scientist, Roswell Park Memorial Institute, Buffalo 3, New York WII~m~ I~.NSZ~.L, Chief, Biometry Branch, National Cancer Institute, NIH, Bethesda, Maryland 20014 E. CUYL~a HA~i~O.~-D, D/rector, Statistical Research Section, Medical Affairs Department, American Cancer Society, 219 Eas~ 42nd Street, New York City, New York W. E. H~.s~o~¢, Chief, Laboratory of Biology, National Cancer Institute, NIH, Bethesda, Maryland 20014 Jom~ HmQINso~, Professor of Pathology, University of Kansas School of Medicine, Rainbow Boulevard at 39th Street, Kansas City 3, Kansas PAul. Ko~N, Associate Director for Field Studies, tional Cancer Institute, NIH, Bethesda, Maryland 20O14 MINORU KVamaP~ (representing Dr. M. Segi), Assistant Professor, Department of Public Health, Tohok~ Uni- versity School of Medicine, Kitayobancho, Sendal, Japan ALr, XAm)r~R D. I~,~rura, Chief, Epidemiology Branch, Communicable Disease Center, USPHS, Atlanta, Georgia 30333 Enwn~ H. L~.m¢~.~% Chief, Viral and Rickettsial Disease Laboratory, California State Department of Public Health, 2151 Berkeley Way, Berkeley, California 94702 Mo~o~¢ L Lm~n~, Professor of Epiderniology, Depart- ment of Epidemiology, Roswell Park Memorial Insti- tute, Buffalo 3, New York ABRAHA~ M. LmI~.~F~.~.D, Professor of Chronic Diseases, The Johns Hopkins University, School of Hygiene and Public Health, 615 North Wolfe Street, Baltimore, ~Iaryland 21205 BmA~ MAcMAHOZ¢, Professor, Department of Epidemiol- ogy, Harvard School of Public Health, Boston, Massa- chusetts 02114 ROB~R~ R. MAP~SHAE, Professor of Medicine and Chair- man, Department of Clinical Studies, School of Vet- erinary Medicine, University of Pennsylvania, Phil- adelphia, Pennsylvania J~.s A. Mizen.R, Professor of Oncology, McArdle Memorial Laboratory for Cancer Research, University of Wisconsin Medical School, Madison 6, Wisconsin ROBV.R~ W. M]~, Chief, Epidemiology Branch, Na- tional Cancer Institute, NIH, Bethesda, Maryland 20014 No~tTo~ Nv.~.so~, Director, Institute of Environmental Medicine, New York University Medical Center, .New York City, New York CL~P~C~. P. O~.Iv~, Ashbel Smith Professor, Depart-ment of Zoology, University of Texas, Austin, Texas JOH~ R. P,~w., Professor Emeritus of Preventive Medicine, Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut 06511 W.~J~I~C~ P. Row~, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, Nil=I, Bethesda, Maryland 20014 M~CHXV.L B. SHL~mIN, Professor of Medicine, Fels search Institute, Temple University School of Medicine, Philadelphia, Pennsylvania 19140 P~I~.~PP~. SHUBIK, Professor of 0neology, The Chicago t~.Iedical School, Chicago 12, Illinois J~.Rzy J. Swszmvsxi, Instytut 0nkologii, ul. Czerw. Armii 15, CIiwice, Poland ERm~s~ L. STSBSn~S, Dean, School of Hygiene and Public 1273 T!54351314

1275 Caner Research Vok 25, ~eptember 1965 Health, The Jo~ Hop~fir_z Y~niver-:ity, 615 North Weh'e Street, BalSraore, Mm-~.-land 21205 I~IO:I~D A. TJ.~J~% Chief, E]:dzoo~ology Section, Na- tional Cancer Institute, USPHS, Bethezda, hlaD'land 20014 Ea~n:sv I,. Wa~-o~a, Head, Ep~cten~o:o~- ~ect~on, Divi- sion of EnvironxaentaI Caucerigen~:i~, Slo~a-Kettering Ir~-fitute for C~ncer Re~.~rch, 410 E~-t. ~$th Street, New York 21, New York T!54951315

Introductory Remarks--Conference on Epidemiologic Approaches to Cancer Etiology LESTER ~BRESLO~V Ditqsion of Preventive Medics! Services, State of California Department of Public Health, BerI;eley, California I'd like to extend ray greetings on behalf of the Com- mittees on Etiology and Epidemiology and Statistics of the American C~ncer Society, through whose joint efforts this conference has come into being. The current rapid expansion of knowledge and clues concerning etiologic factors in leukemia may serve to illustrate briefly the reason for this conference. The association of abnormalities in chromosome 21 with leu- kemia; the high frequency of the disease among pe~ous with Down's syndrome; the increased risk with exposure to radiation; the relationship of radiation to chromosomal abnormalities; the fa~t that benzene can produce leu- kemia; the demonstration of the virus etiology of leu- kemia in animals, including delineation of the pattern of spread of the virus in flocks of chickens; the spatial- temporM aggregation of cases of a leukemia-like disease in cattle, and po~ibly aggregation of eases of human leukemia--most of these fragments have come into focus during the past very few years. You'll notice that the very words used in describing these events may serve as descriptive terms for the several sessions of this eo~ffer- ence. Meanwhile in vital statistics, dat~ have been accumu- lating on trends in the occurrence of the disease, shifts in peak ~e of childhood leukemia, and other bits that fit into a gradually emerging picture of the naturM history of the disease in matt and other species. Not only leukemia, of course, but other forms of neo- plastic disease seem to be yielding to the worldwide attack on cancer. At this time, therefore, ~t seemed appropriate to the Committees on :Research in Etiology and Epidemiology and Statistics of the Americav~ Cancer Society that we assemble this group to review critically the present epi- demiologie approaches to cmacer, and I'd like to emphasize that word critically. It has become fashionable now in American scientific circles to reflect on the "politeness" of American scientific discussion as compared with t.hat in some other countries of the world, where the same care may be exercised and even greater elegance of speech, but much more forceful discussions take place. I hope that we, in this meeting, can begin to emulate some of lhat. Probably only about half the persons here would, if asked to, classify them_~elvcs as primarily epidemiologists, but it is clear that all of us here are concerned with the natural histoD, of cancer. Many of us would use the latter te~Ta in defining epidemiology. Epidemio~o~ts are inclined to ~peak pridefully of their shoe-leather tradition. The fi~t eouNe in epi- demioto~" u~ually begha~ with a lecture on ~.hoe-lea~er tradition. It points to Goldberger's work, which led to the discovery of the means for controlling pellagra. Start- ing with the geographic concentration of ca~e,~ in the Southeastern part of the United States and a temporal peak each Spring, Goldberger found an aggregation of cases in institutions. The disease was limited to those confined there and did not affect the attendants. He proceeded to analyze the conditions of life which char- acterized those most frequently affected and thus came to the hypothesis that the disease was due to a dietary deficiency. Finally he proved this hypothesis correct by therapeutic triM. Goldberger accomplished his work--and epi- demiologists point this out over and over again--quite apart from the laboratory, until the very end stage re- quired identification of the vitamin. While this tradition is well worth continuing as one aspect of epidemiologie work, it appears that cancer should be approached in perhaps a different way, and that is through the dose collaboration of the shoe-leather set with those in the laboratory. To us, the epidemiol- ogists concerned with cancer as it occurs naturally in populations, some of the most effective models are those already worked out by the virologist, the chemist, and the geneticist. On the other hand, ~ virologist who is at this conference had a great deal to do with stimulating it. I-Is insisted i~hat the virologist could enhance his work by learning from the epidemiologist what is known about the out- standing p~tterns of cancer occurrence in human and ani- mal populations. A simple description of geographic, ethnic, temporal, and other factors in the disease--the first, stage of epidemiologic knowledge~m~_y serve as clues to those working primarily in the laboratory. Yet heretofore in research on cancer we've had relatively little opportunitT for personal interchange between those engaged principally in labor~tory studies mad those in field studies, and that is the reason for this conference. Now, a few words as to expectation. First of all, it must be admitted that a good deal of the epidemiologie work on cancer up to the presen~ time has been purely descriptive. While this may be necessary and useful, such data become exciting only when related to some model that may explain how cancer occurs. The more stimulating models in cancer, I believe, are now coming out of laboratoD' investigations, which sometimes are linked to field studies. Thus, some of the most promising t-Ands of work for the epidemiologist require tha~ he be- come much more tmowledge~ble than he is about the method, findings, and problems of the laboratoD'. 1275 TI54351316

1276 Cancer Re~zarch Vol. 25, September 1965 It iz time fer the ep[demiol.~gi~t to ~o beyoncl d~cr~p- t~on come,ted from p~ents co~g to e~eal eeater~, and other ~ch b~e~ of data. He must b~n to fortunate etiolo~c h3To~es b~ed upon pre~ent developmcn~ in the laboratoo, that e~ he t~ted in field ~tu~. A move ~ ~h~s direction req~ that the ep~demiolo- ~st in e~eer overcome his pre~ent n~vet~ abou~ IaboratoD, and acquire a much better undemtanding of its pron~ze ~d i~ ~t~tious, ~ well ~ of the ne~ to de- velop better tools. In t~, the l~boratory investigator m~v gain no~ only ~om the fi~d~gs about the occurrence of cancer ~ populations, which ~e the ~oek ~d tr~e of the epide~olo~s~, but a~ from ~ome of the more techni- cal ~pec~ of the field, espceia~y statistical design. ~le Che laborato~ scientist may rega~ as hopelessly naive some of the fumb~ng ~th broad desefip~ve d~ta about cancer toward some fu~ etiolo~c hypo~esis, so the field ~vestigator m~v be sharply critical of the d~- regard shown by some laborato~ workers ~ward prob- lems of statistical desi~. Both need to inere~o the level of sophistientioa with respect to the other's technics, approaches~ and problems, and ne~ to work in much closer collaboration. The ve~ power of 20th cent~ man to alter ~he bio- lo~c and che~cal ~ct of his ecolo~ heightens the importance of our work. In 1952, I heard Chewer Bz~rd, then the Prc~-~deat of the Rockefeller £Qundz- ticn, reraark, as ~ layman, that the thing that interacted him most about virology and seeing it zdvanced was the po-=~bility of some emergent virus which might wipe out ma~'Jnd if we did not have technolo~- to deal with it. Now, just. 12 years later, v;e h~ve to add ~ome notes to this idea. Man's manipulation of food has resulted in a tremendous outbreak of cancer, hepatoma, fortunately among trout. His manipulation of tobacco for smoking has lead to a vast increase in lung cancer. His m~nipula- tion of ~us to make a vaccine has aroused at least stantial eoneem about iatrogenie cancer on a big scale. The chemist, and the virologist h~ve already done these things. Will the geneticist be next to give uncomfortable substance to the notion thnt the scientist, while he may be useful, can also be a dangerous fellow? While each of us tony be modest about his individual work, I think we're all conscious of the tremendous po- tential represented by the scientific disciplines from which this group comes. It is this sense of responsibility, to- gether with the excitement of being appaxently close to at least some of the secrets of cancer, and the competence of those here, which I am certain will assure a good con- ference. TI54351317

A Sulwey of the Tumor Virus Problem Epidemiologic Standpoint from an W~J~L.~.CE P. ROWE Laboratory o] Infeetio~ Dictates, National Institute of Allergy and [nfec~io~ Di~e~, U~PHE, Betide, Maryland SU~RY This presentation represents an attempt to ~mthesize cu~ent ~owledge of onco- ge~c ~mses from the ~e~oint of the epiden~olo~t faced Mth the problem of test- ing hypotheses on the relationship of vims~ to h~an c~cer. In p~tlcular, the neM for co~der~g modeN of the pathogenesis of the infection is strewed. In some respects the history of tumor virology reads like the case history of a manic-depre~sive psychosis: great cycles of energy and productivity alternating with withdrawal and discouragement. However, to many persons, this case history may resemble more that of a schizophrenic, where out of a massive flow of words and ideas it is difficult to ~jnthesize a coherent picture and to sift out the meaningful from the imaginary. Regardless of the diagnosis, it is apparent that our patient is to be reckoned with; at all levels of cancer research, from the biochemical to the epidemiologic, it is necessary con- tinually to keep in mind the possible participation of viral agents in the biologic system under study. There are several reasons for this: the subtlety and variety of viral influences on cells; the virtual impossibility of prov- ing the absence of a viral influence; and the increasing number and diversity of known viruses detectable only by indirect and specialized technics, which can be extra- polated to mean that there must exist many more even subtler agents not detectable by present procedures. The purpose of this presentation is to try to do a little sifting, to give a concise summary of those aspects of tumor virology which seem most pertinent to the problems of the epidemiologist, and to discuss these problems in terms of several questions facing the epidemiologist who is attempting to elucidate the etiology of tumors. Our thinking on the over-all approach to studying viral etiology of cancer has changed markedly in the past few years. The traditional goal has been to isolate a virus from a tumor and to let the etiologic implication depend on the frequency with which the virus can be recovered from tumors, as compared with control tissues, and on the sinfilarity of tumor ts, pc produced in laboratory systems to that from which the ~irus was recovered. Several observations have forced re~sion of this line of thought-- particularly the findings that polyoma virus acts in many respects like nontumorigenic cytolytic viruses (8), the diszociation between time of maximal virus growth and time of tumor response in polyoma infection (30), and the oncogenicity of several adenoviruzes (22, 33). These findings indicated for the first tinm that there is no sharp distinction in general properties between tumori- genic and nontumorigenic viruses~that~ indeed~ viruses of acute infectious disease may be tumorigenic under certain circumstances, and that a tumor may be a late sequel of an acute infectious process and may not contain the virus. Thus, the floodgates have been opened; we no longer lack candidate viruses to consider as possible tumorigenic agents but must at least keep an open mind towards every virus. It is only too clear that oncogenicity for an experimental animal does not prove that a virus is oncogenic in its natural occurrence; such proof can only be obtained by epidemiologic technics, which involve testing hypotheses of the relation of particular viruses to particular tumors. What etiologic hypotheses should the epidemiologist consider testing? This is a most difficult question to answer in realistic terms. There is no a priori basis for rejecting the possibility that any conceivable virus is the causative agent of some particular tumor~ provided there is a possibility for the virus to have gained access to that tissue. Thus, the number of conceivable hypotheses of the relation of particular viruses to particular types of tumor is so immense that we must somehow make judg- ments of most likely hypotheses based on whatever gener- alizatious we can extract from known systems in lower animals. Stated in another way, we should like to know what characteristics of a virus give us information as to whether it is a likely or an unlikely candidate for tumori- genicity in its natural occurrence. First, let us list the model systems from which we must seek our generalizations. In Table 1 the most important known tumor viruses are listed, along ~th a tentative classification of their known tumorigenicity--whether it occurs in nature or is only known as an artificial labora- tory plaenomenoa, and whether virus replication can occur independent of tumor response (facultative tumorigenic- ity) as opposed to the situations in which vires is liber- ated only by tumor cells (obligate productive tumor- igenicity) or in which tumor development is the only 1277 T154351318

1278 Cancer Re~earch Vol. 25, September 1965 TABLE 1 P~.povavir~ Adenovirus Po.xTirus Myxovirus- like Polyoma Bovine v:art Rabbit wart SV40 Adenoviruses 12, 18, 7 Hamster, P~t, etc. EIor~e, hamster Cottontail rabbit Domestic rabbit Monkey Hamster Human ~lamster, mouse, (Natural), facultative Artificial, cblig~t e-nvnprocluctive Natural, obligate-productive ArtificiM, obligate-nonproductive Natural, obligate-productive Artificial, obligat e- (nonproductive) Artificial, obligat e- (nonproductive) Artificial, obligate-nonproductive Rabbit fibroma Squirrel fibrom~ Yaba Mouse leukemia Fowl leukosis Mouse mammary Rabbit Squirrel l~Ionkey Mouse Rat Chicken Hamster, guinea pig, rat Mouse Natural, obligate-productive (?) Natural, obligate-productive (?) Natural, obligate-productive (?) Natural, facultative (?) Artificial, facultative Natural, facultative Artificial, obligate- (nonproductive) Natural, facultative response and infectious virus is not liberated at any time (obligate nonproductive ~umorigenicity). This type of clarification may be useful in stressing how much of our knowledge is bascd on artificial ~stems which often do not reflect the epidemiologically important characteristics of the natural situations and in emphasizing the fact that, in its present usage, the term "tumor virus" means only an agent that under some particular unique set of conditions may induce tumors, regardless of whether it does so in its natural occurrence. Is there any distinguishing characteristic of all, or even some, of these viruses which sets them apart from "non- tumorigenic viruses"? Such common denon~nators could be looked for in the structure, chemical composition, laboratory characteristics, epidemiology, etc. First, even membership in a particular virus group does not guarantee that a virus is or is not tumorigenic. Known tumorigenic viruses fall into at least 4 major virus families, but in none of these groups is every member known or suspected to be tumorigenic. For example, the papovavirus group (26), which includes polyoma, SV40, and the papilloma viruses, also includes 2 apparently nontumorigenic viruses --mouse K virus and the rabbit kidney-vacuolating virus. Among the adenoviru~es, tumorigenicity appears not only to be restricted to certain serotypes, but in the ease of Type 7, to be limited to certain strains within the type (12). To date, no tumorigenicity has been detected with any picornavirus, arbox4rus, nitavirus, or reovirus, al- though the posgbility of tumorigenic potential of reovi- ru-~es has been suggested by their clo~e similarity in structure and chemical compositinn to the wound tumor virus of plants (18). It is apparent that, as a whole, there is no common denominator with re~pect to structure or cher~icaI position. The papova- and adenoviruses are DNA- containing/nonmembranous, non-lipid-containing viruses with cubic symmetry; the poxviruses contain DNA; and the myxovirus-like agents are RNA viruses consisting of a helically wound internal nucleoprotein component covered by a llpid-containing protein envelope. Papovaviruses and adenoviruses replicate in the nucleus and poxviruses, in the cytoplasm; the myxovirus-like viruses assemble at cell surfaces. At a finer level, however, there may be important and useful markers, i.e., in the composition and structure of the viral nucleic acid, at least in the case of the DNA viruses. The DNA of tumorigenic papova- viruses and adenoviruses shows base composition closer to that of mammalian cells than does DNA from non- tumorigenic viruses, including nontumorigenic adeno- viruses (14). While this type of study has been done with only a very limited number of viruses and cannot be regarded as a valid generalization, the implication of homologous regions in viral and host DNA fits so well with current thinking about the molecular basis of viral neo- plasia that there is ~zeat hope for important predictive information from this approach. Another possibly im- portaut and unique feature is the ring structure of polyoma virus DNA (6); however, so little information is available on the tertiaD, structure of other viral nucleic acids that this finding can be considered only a possible lead. The question of general laboratory characteristics as an indication of possible tumorigenicity of a x~us is ono that has undergone immense change in the last decade. Ten years a~o one might have felt that if a ~4rus is not able to be haud/ed by standard laboratory procedures, if it cannot be propagated in t~!ro, if i~ v:on't do anything * The abbreviatir, n~ u~ed are: DNA, d~o~-ribonucleic aci~l; P~NA, rlb~nucleic ~cid; CF, c~mplement fitting. T15435~3~9