[CANCER RESEARCH 36. 2703-2705, July 1976]
Putative Precursor Lesions: Summary and Some Analytical Considerations' Emmanuel Farber Department of Pathology. University of Toronto, Faculty of Medicine, Toronto, Ontario, M5G 1L5 Canada
it is readily
apparent
from the presentations
at this Con
fenence that several generalizations can be made concern ing the nature of possible precursor populations during the development of cancer in several organs and tissues. Some of these can be summarized as follows. 1. A multiplicity of cellular and tissue structural changes can be seen in every organ system discussed, long before the appearance of cancer can be recognized by any criteria, including morphological. This overall pattern is seen in many human systems including skin, breast, gastrointes tinal tract, urinary bladder, and tracheobmonchial tree. It is certainly seen in most, if not all, animal models useful in the study of carcinoma, including cancer of the liven. This oven riding generalization must be included in any valid hypothe sis of cancer development that purports to be realistic and meaningful. 2. The patterns
of cellular
alterations
seen in any organ
or tissue during cancer development are remarkably simi !ar, even with different chemical carcinogens. This is valid even with carcinogens that are widely diverse in their chem ical structure. This seems to be true also in comparing the human with the animal models, although the data are very spotty concerning the response to specific carcinogens in humans. 3. The majority of such early lesions are proliferative. Doubts have been expressed periodically about whether the early lesions seen in the liver during carcinogenesis are proliferative. Alterations in preexisting hepatocytes without hyperplasia have been considered by some as playing the dominant role during cancendevelopment. Cleanly, from the presentations at this Conference, it is evident that the changes in surface epithelium during carcinogenesis are proliferative. This consideration, plus the data presented by Williams (11), would suggest that the liver is not exceptional with respect to the type of precursor lesion seen. No evidence was presented to support an old suggestion that chemical carcinogens rapidly induce cancer and that the relatively long period of cancer development is con cenned entirely with the control of the growth of these hypothetical so-called dormant cancer cells. 4. Several tissue or organ systems require only a rela tively brief exposure
(hours to days) to a chemical
gen to initiate cancer development. I Presented
at the Conference
“EarlyLesions
carcino
These observations and the Development
of
Epithelial dancer,―October 21 to 23, 1975, Bethesda, Md. The author's research included in this presentation was supported by research grants from the National Cancer Institute (dA-10439, CA-12218), the American Can cer Society (Bd-7P), and the National Cancer Institute of Canada and by Contract N01-CP-33262 from the National dancer Institute.
would suggest that many if not all carcinogenic processes induced by chemical carcinogens have an initial short, more or less irreversible, phase (“initiation―) followed by a period of cancer development (so-called promotion) of much longer duration. Thus the basic pattern of initiation and promotion (on better “development―) studied most in tensively in the skin seems to apply in principle to many other tissues.
5. The cellular proliferative response to chemical carcin ogens is focal in nature. This seems to be obvious in many tissues, even though many biochemical effects or conse quences of exposure to carcinogens seem to involve the majority if not all of the cells of any 1 type. The focal pattern of early cell proliferation suggests at least 2 possibilities: the varying state of receptivity of any population of cells at any time period in their life-span or the relatively uncommon nature of those induced biochemical events that have rele vance to neoplastic development. There is a growing meali zation that similar cells in any organ may show gradients of many biochemical and physiological properties, e.g. , hepa tocytes in different zones of the liven. Conceivably, the carcinogen may have different quantitative or even qualita tive effects on cells in different regions of the gradient. Thus, an element of selection may play a role in the early responses to carcinogens. The response of the surface epithelium of the urinary bladder (4) as a mosaic in the human conceivably could reflect in part such gradients. Major subjects for study would seem to be the essential nature of the molecular and metabolic alterations that char actemize the focal responsive cells and the nature of the selection pressures on environment that favor their growth oven the surrounding cells. 6. The focal proliferative cells are frequently organized or arranged in patterns
differently
than are the norma! cells in
themature organ.Thisisseen to a striking degree inthe urinary bladder (6) but is also observed in skin (9) and in liven (1, 2). The patterns seem to indicate a new type of differentiation (urinary bladder) on a reversion to a more primitive or less mature type (skin, liven). Conceivably, such reversions on deviations may be essential to allow more biological options to the cancer cell that may ultimately evolve. For example, invasion or metastatic spread may involve properties that were needed during embryological and fetal development but not after maturation. The genetic availability of such options may be possible only if the cells are in a more undeveloped physiological state. Comparative studies of suitable precursor populations with the ‘ ‘appropriate' ‘ stage of development might be very profitable in this context.
2703
JULY 1976
Downloaded from cancerres.aacrjournals.org on December 30, 2018. © 1976 American Association for Cancer Research.
E. Farber
Particularly provocative is the observation that a “pro of cancer research is whether themeis truly a preneoplastic moting agent―in skin cancinogenesis induces a reversion to or premalignant lesion. Given the chemical reactivity of a more primitive pattern of differentiation. Conceivably, ultimate carcinogens and the multiplicity of their intemac such an effect might be characteristic of many agents that tions with target cell components, perhaps the lesions we can enhance neoplastic development in a system already see during the formative stages of malignant neoplasms are initiated. independent end-stage manifestations of such a ‘ ‘shot gun― 7. Interruption or disturbance in maturation or differen effect of a carcinogen (Chart 1). What evidence do we have tiation seems to be a common characteristic of many pre that any lesion appearing in the time-span between the cursor cell populations. This is seen in the urinary bladder, initial interaction of a carcinogen with the target cells and skin, colon, stomach, liver, pancreas, and probably many the appearance of cancer has a precursor-product relation other sites. In at least some sites (liver, bladder, skin), the ship to cancer (Chart 2)? The close correspondence be early focal lesions manifesting such changes are ostensibly tween the occurrence of putative lesions and cancer could reversible, in that removal of exposure often leads to matu simply reflect the multiplicity of effects of carcinogens, as nation of the cell population. Later in time, similar lesions already mentioned. appear that seem to have lost the ability to mature and these Two groups of data offer support for the precursor-prod appear to be sites of further evolution or transformation to uct hypothesis. The 1st is the well-known observations in malignancy. humans and the less frequent findings in animal models (e.g. , liven) that histological and cytological earmarks of 8. The time lag between the initial appearance of a cellu malignancy are not infrequently seen in putative focal pro lar reaction to a carcinogen and the appearance of recog nizable cancer by morphological or biological criteria is liferative precursor lesions, such as polyps, hypenplastic on often relatively long — months or years in animals and many benign neoplastic nodules, areas of dysplasia, etc. This body of observation is impressive and highly suggestive. years in humans. Despite the rapidity with which cancino gens can induce many changes in the original target cell The 2nd relates to the occurrence of common markers in presumptive pneneoplastic or premalignant and malignant population, the subsequent history of most carcinogenic lesions. Themeare many examples of negative ‘ ‘ markers,― processes is very protracted. This period is often shortened by continued or repeated exposure to carcinogens on so i.e. , decrease or loss of some enzyme activity or other functional or structural component of cells. More convinc called promoting agents on cocarcinogens. ing are positive markers, e.g., a-fetopnotein, preneoplastic This slow development pattern was attributed commonly to host immunological control (surveillance). However, antigen (2, 3), and enzyme activities (a-glutamyltmanspepti themeis mounting evidence against the essential validity of dase) in liven cancinogenesis. Obviously, the need for many more markers, especially highly specific ones, is evident. this concept. Another aspect to be greatly expanded is the need for When viewed in perspective, the frequency of such long periods of development of malignancy, lasting probably 15 additional functional handles for a more mechanistic analy sis of cancer development. To date, the major emphasis has to 25 years in the human, coupled with the infrequent oc cummenceof effective long-term host control of frankly ma lignant cell populations, suggests the existence of as yet NEW CtLL unknown host factors that have delayed the rate of cellular POPULATION A evolution to malignancy. Such a delay could have an impor / tant survival value for the host and the species. The peak of appearance of malignancy is past the reproductive period. N@WCELL Most cancers, therefore, have little or no influence on the survival of the species. On the other hand, prolongation of CELL — the period of evolution for cell malignancy during carcino genesis beyond the reproductive period could have a signif ‘ S @1.POPULATION CARCINOGENIC “ C icant effect in preserving the species. It is to be anticipated STIMULUS that host protective factors, operating to delay the appear ance of cancer by prolonging the carcinogenic process, POPULATION might well have developed during the long period of evolu tion. A search for such factors and the formulation of the Chart 1. Diagrammatic representation of possible effects of an activated possible specific nature of such factors could lead to new carcinogen on target cells generating a variety of independent results, among which is included neoplastic transformation. ways to analyze an important segment of the carcinogenic process. , I
1*16(1.'
@
POPULATION
:
—CANCER
I ‘ — — S__ _ NEW CELL .4
TARGET—.@
@
CELL
@—-
NEW CELL
NEWCELL NEWCELL NEWCELL —@, POPULATION POPULATION - - - - - - -@. . —@ A
- .
POPULATIONS
N
CANCER
N
General Considerations CANONOGENIC
The presentations during this Conference naturally have raised more questions than they have answered. This is 1 index of the profitable nature of the interchange of ideas. One of the key questions that dominates this whole area
2704
STiMULUS
Chart 2. Diagrammatic representation of a cellular chain reaction induced by an activated carcinogen. In this representation the various noncancerous cell populations are considered to be related to each other and to cancer as precancer-prod uct.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on December 30, 2018. © 1976 American Association for Cancer Research.
VOL. 36
Putative Precursor
been morphological. In my view, this must remain as a point of reference for studies in the future. Morphology is still one of the most useful approaches to cell biology, since it ena bles us to observe easily a host of modulations that often can be correlated with biological behavior. However, from this base must emanate a wide variety of functional and molecular approaches that almost certainly will help clarify many aspects of altered cell behavior. In this context, we may draw again on the liver as a model. Recent research is pointing strongly to the development of resistance to the cytotoxic action of carcinogens and other hepatotoxins as an early manifestation of new hepatocyte populations in liver carcinogenesis (1-3). I would expect that each system would have its own functional base for the development of the appropriate selection pressure, de pendent upon the normal physiology of that tissue. Natu rally, liven plays a major role in drug metabolism and in so called “detoxification.― An environment based on this would be expected. On the other hand, other organs or tissues would be expected to have other types of selection pressures, based on their normal physiology. Consistent with such an expectation is the recent work with mammary tumors (7) indicating a possible major selection pressure in the endocrine area. The identification and study of the selection environment would seem to be an especially profitable area for research that could lead to novel ways to interrupt the carcinogenic process. Although the emphasis in carcinogenesis is shifting rap idly to in vitro studies, the knowledge to be gained must be related to intact organisms if it is to become knowledge and not simply the accumulation of trivial data. Models must be developed in which an easy flow of information from in vivo to in vitro and vice versa can take place. The in vitro must include organ as well as cell studies. The profitability of such an approach is already evident at this early stage of development, as demonstrated at this Conference (5, 10). Another obvious lesson is the urgent need to study purer and purer cell populations during carcinogenesis. Methods of cell isolation and purification and manipulation of more complex systems in vitro and in vivo to obtain cleaner and cleaner cells would seem to be obvious requirements for the development of in-depth studies of carcinogenesis. The challenge is to develop simpler models in which we can
JULY
Lesions
pose more penetrating questions of mechanism, at both the biological and the biochemical levels. Cell culture of repro ducible lesions induced in vivo would seem to have great potential in the immediate future (8). In the liven, we are now able to ask a few more penetrating questions because of the increasing knowledge of carcino genesis in this organ (e.g. , Refs. 1 to 3). This is approaching the point where we can begin to develop testable hy potheses that relate to functional attributes of new cell populations on a quantitative basis. The presentations at this Conference make one hopeful that we shall be able to do this to an increasing degree in several other systems. These results almost certainly fore shadow the development of a functionally meaningful merger of information at the molecular and the cellular levels, from which knowledge leading to new practical de velopments in diagnosis and therapy is virtually inevitable.
References 1. Farber, E. Hyperplastic
Liver Nodules.
In: H. Busch (ed).
Methods
in
Cancer Research, Vol. 7, pp. 345-375. New York: Academic Press, Inc., 1973. 2. Farber, E. The Biochemical Pathology of Experimental Liver Carcino genesis. In: G. P. Warwick, H. M. Cameron, and C. A. Linsell (eds.), Liven Cell Cancer. Amsterdam: Elsevier Publishing Corp. , in press. 3. Farber, E., Hartman, S. P., SoIt, D., and Cameron,
A. Studies on Precan
cerous Liven Cell Populations and Their Identification. In: T. Sugumura (ed), Fundamentals in Prevention of dancer. Proceedings of the Sixth Princess Takamatsu Symposium, Tokyo, 1975, in press. 4. Farrow, G. M. , Utz, D. C., and Rife, C. C. Morphological and Clinical Observations of Patients with Early Bladder Cancer Treated with Total dystectomy. dancer Res., 36: 2495-2501 , 1976. 5. Gniesemer, A. A., Nettesheim, P., and Marchok, A. C. Fate of Early Carcinogen-induced Lesions in Tracheal Epithelium. dancer Res., 36: 0000-0000, 1976.
6. Hicks, R. M. , and Wakefield, J. St. J. Membrane Changes during Urothe hal Hyperplasia and Neoplasia. dancer Res., 36: 0000-0000, 1976. 7. Kim, U., and Depowski,
M. J. Progression
from Hormone Dependence
to
Autonomy in Mammary Tumors as an in Vivo Manifestation of Sequential Clonal Selection. dancer Res., 35: 2068-2077, 1975. 8. Laenum, 0. D., and Rajewsky, M. F. Neoplastic Transformation of Fetal Rat Brain Cells in Culture after Exposure to Ethylnitnosounea in vivo. J. NatI.dancerInst., 55:1171-1188,1975. 9. Raick, A. N. Cell Differentiation and Tumor-promoting Action in Skin Carcinogenesis. Cancer Res., 34: 2915-2925, 1974. 10. Reese, D. H., Friedman,
A. D., Smith,
J. M., and Sporn,
M. B. Organ
Culture of Normal and Carcinogen-treated Rat Bladder. dancer Res., 36: 0000-0000,
1976.
11. Williams, G. M. Functional Markers and Growth Behavior of Preneoplas tic Hepatocytes.
Cancer Res., 36: 0000-0000,
1976.
2705
1976
Downloaded from cancerres.aacrjournals.org on December 30, 2018. © 1976 American Association for Cancer Research.
Putative Precursor Lesions: Summary and Some Analytical Considerations Emmanuel Farber Cancer Res 1976;36:2703-2705.
Updated version
Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/36/7_Part_2/2703.citation
E-mail alerts
Sign up to receive free email-alerts related to this article or journal.
Reprints and Subscriptions Permissions
To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected]. To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/36/7_Part_2/2703.citation. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.
Downloaded from cancerres.aacrjournals.org on December 30, 2018. © 1976 American Association for Cancer Research.