Tobacco Documents Online

L C THE JACKSON LABORATORY •AA NAIiBOR'. R7AIFfv- MEMORANDUM J 10 March 1978 l ( ( C ( To: DR. W. U. GARPNER, SCIENTIs'IC DIRECTOR Sub;: CONFE3Z*ICE ON HUMAN CARCINOGEN METABOLISM: ARE H K01094002 I should like to make a few observations on the human A i con:erence before too much time has elapsed. You may or may not wish to consider -_-7 evaluations and conclusions when certain of the present grantees are up ior renewal or continuation application: This conference was timely for it allowed both interested bca rdl- and staff inerabcrs to gain a perspective of the work in pgogrese. uost of tre first day's presentations were disappointing, disturbing and even vaseed. uost or all of the highlights surfaced the second day: 1. I am disappointed that the situation with respect to AZE :,a man is considerably more complex thanin inbred strains of mice. No doubt, the prerequisite of mitogen sti=lation of lymphocytes in culture is the rajog draw-back. Additional factors influencing results are the culta.e sysgem and the donor histories; these factors have been identified and esumerated. They make assays in field or populat-ion studies not only unreliable but alr-cse imtiossible. 2. Although some level of standardization of assay c9nditiots has been achieved it is disturbing to me that cause and effect of AFr-R acti^=t7 often seem to have been ignored i.e. do AFD1 levels indicate a risk for t:-or'-genesis or are they a reflection.of the t=or condition itself. Thus, the detailed histopathological diagnoses and types of tunors and possible associated disorders (infections etc.) have been largely replaced by the generality ef cancer i.e. lung cancers, neck cancers, etc. rather than specific tcmror types. 3. I realize that standardization o: an assay as variable as A-1-i takes time but fancy instr-ser.tation is wasted if the assay cannot be interpreted anc is undependable. While 3-OH-3? appears to be the relevant metabolite j=- aice associating with cancer risk it is far from clear whether or not the enti:e metabolic profile needs measurement in man in order to derive a meaning_'_1 relationship to either a cancer risk or a cancer state itself. There appear to be too many redundancies atsong the individual projects and people d-x~licating similar efforts. Thus, although many procedural problems have been doc:~ented I am not sure whether or how aany of them can be solved despite a considerable expense already. I CTR HN 043952

I -1 2 H 01094G03 4. It is clear now that AH8 is neither an absolute risk msrker for cancer nor a diagnostic =arker in cancer as a whole or lung cancer specifically. Although it is no better than other biomarkefs applied in screening procedures for cancer, within kindreds it may be a most usEfu discriminant of relatives at f•iYk. Thus, 1RR and other biomarkpss, as diScussed, are best studied within the context of a clinical prografa of h=an caaeer genetics. This setting (Lynch-Guirgis-Harris, Rasco) undoubtedly deserves CTVs continued support while most of the others (Koufi, Contrell, Martin, Alfred, Busbee) should be reduced and some (Fisher,T~rrer) perhaps even eliminar~ed. ~gs is most likely the best way for support money to be spent most effectively in the future. Obviously a great deal has been learned and some progress has been made, the likelihood of AKi being a reliable cancer screen outside families or kindreds is small. 5. Aside from technical problems of AHA assays for ;isk or cancer dEterminations one must remember.that a particular phenotype is generally an expression far removed from its Aenobvwic co i.e. usually many events =st happen or come about to express a particular phcaotype. These events m3y or may not involve mutations; indeed a tendency to ma.lignant tPansformation is fv=d to be arv-.rLiated with a variety of mutatinns in mice and man. Ttbe possibility of geaeri.c variaLilit7 can bcst be assessed by either or both pirent-offspri-ng and sib-sib correlations. If none are found, genetic factors are not segregating. Since t=ors for the nost part oceux fairly late in life the influence of environnental factors increases. Stus, the msjor €rtctian of phenomic variability is probabiy almest all noh-aenetic. Tiais nbn^genetic vagiabil.oe; ray be so large as to obscure expected genotypic Elasses in segregating generatioc.s. On the other hand phcnotypic variations do not necessarily connote ganornic dif.terences because a genotype defines a range rather than a single phenot,ye. Thus, different cancer phenotypES may occur despite genetic usiformitw. =vrt7eY- more it is unlikely that co-ailete ptacnotypic estpreasio€a of transformation occurs in a single step. Indeed t:_--Pgs usually develop froca presaeplastic lesions. Thus, there is a large sequence of events leading to a tumor phenotype. Also, it is iaprobable that various cell types can be initiated to transform with ecual ease or probability and trans:ormation may be dependent upon the target cell =n a specific cell cycle or proliferative phase. For these reasons it is improbable that AHi may represent an absolute marker for the generality of cancer-or even classes of cancer (e.g. lung cancer) but nay serve a useful purpose in conjunction with other biomarkers as risk assessment parameter ::::::n families or kindreds. : look forward to the conference's proceedings publica:ion for it will no doubt contain valuable inforr..ation but may also be the concl_?:n¢ chapter for increased A::ri studies in man. FM:kn C T R I I N 0 4 3 9. '% 5 tP 1 I