Znt.

J . Cancer: 18, 122-129 (1976) SPECIAL REPORT

DEFINITIONS AND TERMINOLOGY ’*’ I N CANCER (TUMOR) ETIOLOGY-AN ANALYSIS AIMING AT PROPOSALS FOR A CURRENT TNTERNATIONALLY STANDARDIZED TERMTNOLOGY Erich HECKER Deutsches Krebsforschiitigszertrmm, itn Neurnheirrier Feld 280, 0-6900 Heidelberg, Germany

In the course of its development every field of science passes at least once through a period of confusion, resulting from discrepant use of terminology. At present in cancer (tumor) etiology such confusion is considerable, because some of the key terms describing the generation of neoplasia are being used with divergent and even with opposite meanings. This situation calls for clarification: a clearly defined and generally accepted system of terminology in cancer (tumor) etiology is essential for a more eKicient worldwide collaboration in research and in teaching as well as in documentation, environmental hygiene and legislation. During recent years considerable progress has been made in understanding the contributions of causative agents, of patterns of exposure and of the host or target tissue to the etiology of cancers (tumors). This permits in depth analyses of the meaning of some of the key terms used in this field to provide a sound background as well as a starting point and a stimulus for the development of a proposed concise system of descriptive terminology. Such a system has to be purely descriptive and should be as simple and flexible as possible. Its development must not be delayed until mechanistic details of the generation of neoplasia have been explored. It will be shown that the system required can be based upon well defined terms that describe reproducible (experimental) facts and that terms involving hypothetical e.g. mechanistic interpretations can be excluded. In most cases it is considered preferable to clarify, and if necessary, to redefine existing terms, rather than seek to replace them by new ones. An informal trial to define a limited number of terms used in cancer (tumor) etiology was made by a group of temporary advisors to WHO ( I 97 I ). A glossary of the definitions of terms proposed in this paper is given in Table IV. BASIC CONSIDERATIONS

Carcinogenesis For “generation of neoplasia” the use of the term ‘‘ carcinogenesis ” may be preferred to “ tumorigenesis ”, ‘‘ oncogenesis ” or “ blastornogenesis ”. ‘’ Carcinogenesis ’’ (from the latin curcirm, the crab and genere, to make, to create) is generally understood to

include all phases of the process of generation of neoplasia, including in particular the invasive phase. The terms “ tumorigenesis ”, ‘‘ oncogenesis ” or “ blastomogenesis ” do not necessarily include the invasive phase. Also “carcinogenesis ” is the most adequate term for use in connection with the more complex (multifactorial) patterns of exposure of the host or target tissue to causative agents (see below “ Fundamental variables ” and the following paragraphs). If “ carcinogenesis ” is accepted as the standard term to describe generation of neoplasia in the broadest possible sense it will be necessary t o state by definition and to agree by convention that it is meant t o imply also the generation of, for example, sarcomata and leukemia. Such a generalization of the more specific meaning attributed to the term “ carcinogenesis ” in pathohistology appears to be a more acceptable alternative (Epstein, 1974; Cairns, 1975) than the introduction of a new term

The fundamental variables of carcinogenesis The coniplexity of any process of carcinogenesis requires that the contributions of its fundamental variables be distinguished clearly. These are: causative agents, patterns of exposure and host o r target tissue predispostition. One of the main reasons for the present confusion i n terminology in cancer etiology is that frequently in Received: March 18, 1976. The Committee on International Collaborative Activities (CICA) of the lnternational Union against Cancer (UICC) has decided to attempt to provide a current, internationally standardized terminology in cancer (tumor) etiology, in co-operation with the International Agency for Research on Cancer (IARC) and the World Health Organization (WHO). As a first stage of this attempt and to encourage international discussion, the CICA has requested one of its members, the author, to formulate and publish under his responsibility the present analysis and proposals. This paper, to be published in a number of cancer journals, outdates all of its previous versions or drafts (see acknowledgements). Readers who might care to comment on these proposals by letter are requested to direct their remarks to the office of the UICC, 3, rue du Conseil-General, Geneva, Switzerland, within 6 months of publication of this article.

123

DEFINITIONS AND TERMINOLOGY IN CANCER

carcinogenesis a clear-cut distinction between causative agents, patterns of exposure and host or target tissue predisposition is not or cannot be made. For classification of such cases detailed analysis of the factual information available and/or carefully devised further (experimental) study should help to clarify the situation (see below and “ Carcinogenic factors and unspecified carcinogenesis ”1.

Causative agents. Carcinogenesis is the result of exposure (s) of the host t o causative agents of exogenous or endogenous origin and of physical, chemical or viral nature. By definition and convention the term “ carcinogen ” may be used for such agents if their origin, nature and identity are unequivocally clarified. Carcinogens can induce carcinogenesis in one or more tissues of the host: they exhibit ‘‘ tissue specificity ”. The general term “carcinogen” may be specified further with respect to the kind of carcinogenic process caused (see below and ‘‘ Solitary carcinogenesis ”, ‘‘ Syncarcinogenesis ”). For the sake of consistency terms such as tumorigen, oncogen and blastomogen should not be used

(see also under “ Carcinogenesis ” above and Epstein, 1974). The term “ tissue specificity ” is proposed in preference to the term “ organotropy ”, because within a certain organ of the host, carcinogenesis may be induced in a more specific target tissue. According to present knowledge, the majority of all known carcinogens of physical, chemical or viral nature are of exogenous origin (e.g. Hecker, 1972; Epstein, 1974). Carcinogens of endogenous origin may be of either chemical or viral nature. Examples of some well known carcinogens are given in Table I.

Pattern of exposure. Carcinogenesis may be the result of unifactorial or multifactorial exposure of the host or target tissue to carcinogens. The multiplicity of possible patterns of exposure of the host or target tissue according to kindCs), number@) a n d sequence(s) of carcinogens may be called collectively “ carcinogenic processes ”. By definition and convention any carcinogenic process is initiated by exposure of the host or target tissue to a “ solitary (or to an “ incomplete ”) carcinogen ” (see above and

TABLE I EXAMPLES OF CARCINOGENS: SOLITARY CARCINOGENS A N D COCARCINOGENS

Origin of carcinogens

Exogenous

Identity of carcinogens

Nature of carcinogens

Sources of carcinogens

Physical

Solitary carcinogens

Sun, X-ray apparatus, nuclear fission

UV-light, X-rays, rays of nuclear fission

Unknown

Asbestos, talcum, chromate, aromatic hydrocarbons and amines, nitrosamines and nitrosamides, chloromethyl methyl ether, vinyl chloride, plastic film

Iron oxide, phenols, aliphatic hydrocarbons, detergents

Pharmaceuticals

Isonicotinic acid, hydrazide, alkylating cytostatics, diethyl stilbestrol

Anthralin, immunosuppressive drugs

Food additives and contaminants

Dimethylaminoazobenzene, DDT,

Ethanol, certain autoxidation products of unsaturated fatty acids

Chemical Chemical raw materials or products

~

~~

nitrosamines and nitrosamides

Fungal and higher plant Aflatoxins, pyrrolizidine alkaloids, products cycasin

~~

Viral

Cell particles with latent virus

Limonene, phorbol and similar diterpenes and their esters, cyclopropenoic fatty acids

~

Biological environment

Endogenous Chemical Anabolic and catabolic metabolism Viral

Cocarcinogens

Informational macromolecules

Shope papilloma and fibroma virus

3-hydroxy-anthranilic acid

Estrogens, androgens, bile acids

Informational macromolecules

Helper virus (?)

’This presentation of identified carcinogens classifies them primarily according to their origin, nature and sources as indicated in the first three columns and separated by horizontal lines. Synergistic actions of solitary carcinogens and cocarcinogens of different origin, nature and sources, e.g. aromatic hydrocarbons with phorbol esters, are by no niedns excluded.

124

HECKER TABLE 11 SURVEY OF CARCINOGENIC PROCESSES: SOLITARY CARCINOGENESIS AND SYNCARCINOGENESIS

Pattern of exposure of host or target tiswe

Carcinogenic processes

Uni factorial

Solitary carcinogenesis Exposure of host or target tissue to one carcinogen Causative agents. Solitary carcinogen, specific event(s) of solitary carcinogen considered irreversible, but not lethal Host or target tissue predisposition. Inborn or acquired

Multifactorial

Syncarcinogenesis Exposure of the host or target tissue to more than one carcinogen Pluricarcinogenesis

Cocarcinogenesis

Causative agents. Solitary (or incomplete)

Causative agents. Solitary (or incomplete) carcinogen simultaneously with or fol-

carcinogen simultaneously with or followed by at least one other solitary carcinogen; specific event($ of both solitary carcinogens considered irreversible Host or turget tissue predisposition. Inborn or

acquired

the following chapter). A survey of known classes of carcinogenic processes is given in Table TI. The term “ carcinogenic process ” is proposed to be used in a purely descriptive (i.e. toxicologic) manner without mechanistic implications. To investigate mechanistic aspects of carcinogenesis at the cellular and the molecular levels, certain carcinogenic processes in certain target tissues may be more useful experimental model systems than others (see below). Host or target tissue predisposition. As the third

fundamental variable a disposition of the host or target tissue may contribute to its neoplastic response. Such a disposition may be either inborn (genetically determined) or acquired. In either case, by definition, such a disposition exists prior to the start of any carcinogenic process. Hence it may be called “ host or target tissue predisposition ”. Any exposure of the host or target tissue to predisposing, non-carcinogenic agents prior to the initiation of a carcinogenic process (see above) is a conditioning exposure of normal cells. Exposure of the host or target tissue toper se non-carcinogenic agents (e.g.cocarcinogens, see below) ufier exposure to a solitary carcinogen (see below) hits upon (what may be called neoplasiogenically) altered cells and hence is basically different from any conditioning exposure. The contrary of host or target tissue predisposition is resistance of the host or target tissue (see also below, “ Remark on anticarcinogenesis ”.) Some examples of host or target tissue predisposition have been assembled in Table 111.

lowed by cocarcinogen; specific event($ of solitary (or incomplete) carcinogen considered irreversible, specific event(s) of cocarcinogen considered reversible Host or target tissue predisposition. Inborn

or acquired

TABLE IJI EXAMPLES OF INBORN OR ACQUIRED HOST OR TARGET TISSUE PREDISPOSITION

Congenital genetic defect predisposing to the development of skin cancer following exposure to sunlight (Xeroderrna pigmentosum). Partial hepatectomy prior to exposure to a hepatocar-

cinogen. Exposure to certain per se non-carcinogenic agents, such as cocarcinogens (Table I), prior to exposure to solitary carcinogens. Etiologically undefined non-carcinogenic ‘‘ influences ”

such as irritation, wound healing, age, sex, diet and strain prior to exposure to solitary carcinogens.

TERMS FOR SPECIFIC CARCINOGENIC FACTORS AND PROCESSES

The following are proposed definitions of standard terms for specific carcinogens and carcinogenic processes. To leave no doubt that all three fundamental variables must be considered for classification of any carcinogenic process, host or target tissue predisposition was included throughout in Table 11. Solitary carcinogenesis Carcinogenesis may be the result of single o r of repeated (chronic) exposure of the host or target tissue t o a single carcinogen: such exposure is unifactorial. It is proposed that the carcinogen

DEFINITIONS AND TERMINOLOGY IN CANCER

involved be called a “ solitary carcinogen ” and the corresponding process “ solitary carcinogenesis ” (Tables I and 11). After the etiology of certain occupational cancers had become obvious primarily by clinical observations, solitary carcinogenesis by exogenous physical (Marie el a!., 1910), chemical (Yamagiwa and Ischikawa, 1914) and viral factors (Rous, 1910) was demonstrated experimentally. Subsequent experimental investigations revealed the tissue specificities of solitary carcinogens : the wide spectrum of solitary carcinogens known today covers specificity for practically any tissue (Schmahl, 1970). In solitary carcinogenesis the carcinogenic response of the host or target tissue as identified qualitatively

125

by well-established histologic means is measured quantitatively by the following criteria: (1) the latent period (the time needed before cancers, or tumors, appear; see above); (2) the cancer (or tumor) rate (the number of individuals carrying at least one cancer, or tumor); and (3) the cancer (or tumor) yield (the number of cancers, or tumors, per individual), o r by a combination of criteria 1-3. The specific carcinogenic effect(s) of a solitary carcinogen is (are) considered irreversible. Chronic exposure of the host to a solitary carcinogen results in a cumulation of its specific carcinogenic effects to cause cancer of the tissue(s) involved (summation hypothesis, Druckrey and Kiipfmiiller, 1949). Apart from its specific carcinogenic effect(s) a solitary carcinogen may cause nonspecific epiphenomena in the host or target tissue(s) due, for example, to toxicity.

TABLE 1V GLOSSARY OF THE MAIN TERMS AS PROPOSED AND DEFINED OR REDEFINED IN THE TEXT, I N ALPHABETICAL ORDER ~

Carcinogenesis is generation of benign and malignant

neoplasia in the broadest possible sense, including generation of sarcomata and leukemia. Carcinogens are agents causative for generation of neoplasia whose origin, nature and identity are unequivocally clarified. Carcinogens may exhibit tissue specificity. Carcinogenic factors are agents causative for generation of neoplasia whose origin, nature and identity are not unequivocally clarified. Carcinogenic processes are all patterns of exposure of the host or target tissue to carcinogens and carcinogenic factors. By definition carcinogenic processes start with the exposure of the host or target tissue to a solitary (or an incomplete) carcinogen. Cocarcinogenesis is one of the prototype process of syncarcinogenesis resulting from simultaneous or sequential exposure of the host or target tissue to at first one solitary (or one incomplete) carcinogen and then to an unequivocally identified causative agent non-carcinogenic and non-initiating per se. The latter factor is called a cocarcinogen and may exhibit tissue specificity. Host or target tissue predisposition may be inborn or acquired and exists prior to the start of any carcinogenic processes. Pluricarcinogenesis is one of the prototype processes of syncarcinogenesis resulting from the simultaneous or sequential exposure of the host or target tissue to a solitary (or and incomplete) carcinogen and to at least one other solitary carcinogen. Solitary carcinogenesis is the process of carcinogenesis resulting from unifactorial exposure to the host or target tissue to a single carcinogen. Such carcinogen is called a solitary carcinogen. Syncarcinogenesis covers a11 processes of carcinogenesis resulting from multifactorial exposure of the host or

target tissue to one solitary (or one incomplete) carcinogen simultaneously with or followed by other carcinogens or carcinogenic factors. Unspecified carcinogenesis is any process of carcinogenesis for which the role of the kind, number and sequence of exposure to carcinogens (or carcinogenic factors) has not been clarified unequivocally.

It is preferable to use the term “ solitary carcinogen ” (Hecker 1972, 1975) referring to the carcinogenic process involved, rather than terms such as ‘‘ carcinogen ”, “ procarcinogen ”, “ complete carcinogen ” or “ full carcinogen ”. The term ‘‘ carcinogen ” would not be specific enough because it covers also cocarcinogens (see above). The other (composite) terms listed above are already associated with special meanings: “ procarcinogen ” refers to the metabolism of carcinogens and is used to indicate that the solitary carcinogen in question requires metabolic activation (to a “ proximate ” or an ‘‘ ultimate ” carcinogen). The terms ‘‘ complete carcinogen” or .‘full carcinogen ” are used to differentiate certain carcinogenic factors (i.e. solitary carcinogens) from “ incomplete carcinogens ” (see below “ Pluricarcinogenesis ” and “ Cocarcinogenesis ”).

Of all possible processes of carcinogenesis solitary carcinogenesis is of least complexity. In the etiology of human cancer, solitary carcinogens are considered as first order carcinogenic risk factors (Hecker, 1972, 1975). As an experimental model, solitary carcinogenesis may be used to investigate the etiology of human cancers and the biology and biochemistry of cancer cells.

Syncarcinogenesis (with a remark on anticarcinogenesis) In most reaI-life instances, carcinogenesis may result from exposure of the host or target tissue to more than one carcinogen: the exposure is multifactorial. More specifically, neoplasia may arise because of an augmentational or synergistic action of subcarcinogenic doses of two or more carcinogens. The corresponding processes may be termed “ syncarcinogenesis ”. The origin (exogenous or endogenous) and nature (physical, chemical, viral) of the synergistically acting carcinogenic factors is unimportant as also is their number. In processes of syncarcinogenesis it is particularly important to distinguish clearly the carcinogens and

126

HECKER

processes involved from the host or target tissue predisposition (see above, ‘‘ Fundamental variables ”). For example, in viral carcinogenesis, if the generation of neoplasia is a direct consequence of an acute infection of the host, the virus clearly is an exogenous solitary carcinogen (Table I). If, alternatively, the generation of neoplasia is due to the activation of an integrated viral genome (oncogen hypothesis : Huebner and Todaro, 1969; protovirus hypothesis: Temin, 1971), by convention such a genome may be considered an endogenous solitary carcinogen or else an “ incomplete carcinogen ” (Hecker, 1972). The activation of such a genome or provirus leading to neoplasia by an additional causative agent (of exogenous or endogenous origin and of a physical, chemical or viral nature) consequently would be classified as a process of syncarcinogenesis (see below and Table 1). Clearly, in this case predisposition of the host or target tissue will have to be considered with the (early) infection by the virus leading to the incorporation of the viral genome. The term “ syncarcinogcnesis ” was proposed in 1949 (Bauer, 1949; see also Bauer, 1963) and redefined in I972 (Hecker 1972; see also Hecker, 1975). The preposition “ syn- ” is used in the broadest possible sense excluding, however, any conditioning exposure of the host or target tissue. Also, it does not mean any specific mode of pharmacological action (additive vs multiplicative). The Greek prefix syn- means the contrary of ‘‘ anti- ”. Remark on anticarcinoggenesis. Exposure of the host or target tissue t o a solitary carcinogen followed by one or more agents may result, not only in a n augmentation (“ syncarcinogenesis ”) but also in a n inhibition (“ anticarcinogenesis ”), of the carcinogenic response (for a review see Falk, 1971). Such an inhibition is caused by interference of the inhibitor (anticarcinogen) with a n ongoing process of carcinogenesis and may result in a lengthening of the latent period or in a reduction of the cancer (or tumor) rates and yields, or by a Combination of all three of these criteria.

In anticarcinogenesis, as in syncarcinogenesis, resistance of the host or target tissue (see above “ Fundamental variables ”) is to be distinguished from inhibition due to interference with an ongoing carcinogenic process. Resistance may be inborn or acquired by conditioning of the host or target tissue, c.g. by stimulation of its immune surveillance to protect against cancer virus infection or by prior adrenalectomy protecting against butter yellow hepatocarcinogenesis. Conditioning treatments to acquire resistance may be a means of cancer prophylaxis. Inhibition by interference with an ongoing carcinogenic process is the basis of cancer therapy. Due to the large number of carcinogens already known (Table I) theoretically a n almost infinite number of multifactorial patterns of exposure of the host o r target tissue, and hence of processes of syncarcinogenesis, may be visualized. In the following some prototype processes of syncarcinogenesis will be analysed.

Pluricarcinogenesis

Carcinogenesis by simultaneous or sequential exposure of the host or target tissue to one solitary (or one incomplete) carcinogen followed by another solitary carcinogen may be called ‘’ pluricarcinogenesis ” (Table 11). Carcinogenesis through activation of the latent genome of a tumor virus by a solitary carcinogen of whatever

origin, nature and identity may be classified as pluricarcinogenesis considering the latent viral genome as an ” incomplete carcinogen ”. Such a process of pluricarcinogenesis corresponds to that type of cocarcinogenesis in which the host or target tissue is exposed to an ” incomplete chernicd carcinogen ” at first, followed by exposure to a cocarcinogen (see below). Experimental investigations into the biology of pluricarcinogenesis and its possible role in the etiology of human tumors have been initiated in recent years ( e . g . physical: Emmett, 1973; chemical: Likhachev, 1968; Schmahl, 1970; Mohr, 1973; viral: Roe, 1968). The role of incomplete carcinogens (of any origin, nature and identity) as well as that of single and/or niultiple tissue specificity of the solitary (or incomplete) carcinogens involved requires clarification (e.g. Mohr, 1973). I n pluricarcinogenesis the response of the host or target tissue as identified qualitatively by usual histological means may be measured quantitatively by the latent period of cancer (or tumors), by the cancer (or tumor) rate, by the cancer (or tumor) yield, or by a Combination of these criteria. Pluricarcinogenesis is the consequence of the assumed irreversibility of the specific carcinogenic effect(s) caused by the solitary or by the incomplete carcinogens involved. The term pluricarcinogenesis ’’ was proposed in 1972 (Hecker, 1972, 1975) in an attempt to subdivide the yynergistic processes of carcinogenesis according to the kind, number and sequence of the carcinogens involved (Table 11). .‘ Pluri ” is an abbreviation of the latin pbrrs, meaning several. Clearly it would be meaningless to use the term “ pluricarcinogenic factor ‘I

’I.

In environmental hygiene the possibility of pluricarcinogenesis is of the utmost practical relevance (Cairns, 1975). However, as a n experiniental model for mechanistic investigations at the cellular and the molecular level it is obviously impractical: it superimposes the specific cellular and molecular carcinogenic effects evoked by the carcinogens involved. Even for just one solitary carcinogen these effects are unknown as yet (Hecker, 1972, 1975). Cocarcinogenesis

Carcinogenesis by simultaneous or sequential exposure of the host or target tissue to at first a subcarcinogenic dose of a solitary (or one ‘‘ incomplete ”)

DEFINITIONS AND TERMINOLOGY IN CANCER

carcinogen and another per se non-carcinogenic and non-initiating factor may be called cocarcinogenesis. The latter factor of any origin (exogenous o r endogenous) or nature (physical, chemical or viral) and of known identity is called a “ cocarcinogen ” (Tables I and 11). ‘‘ lncomplete carcinogens ” do not cause neoplasia in a certain tissue even on repeated application, although a single dose of them may be capable of initiating that same tissue, e.g. urethane in skin of mice. An “ incomplete carcinogen ” for one tissue might be a solitary carcinogen for another tissue (urethane for lung of mice). It is proposed therefore to maintain the term “ incomplete “ carcinogen and consider “ solitary carcinogen ” as the alternative (see also ‘‘ Pluricarcinogenesis ”).

127

Syncarcinogenesis ”). Etymologically and literally co- ” has the same meaning as syn- : it is an abbreviation of the latin cum, meaning “together with ”. Most probably these inherently identical meanings of these prefixes have contributed much to the ambivalent usage of the term “cocarcinogenesis” causing much of the present terminological confusion. To overcome this situation it is proposed that the term “ cocarcinogenesis ” be reserved and redefined as above and according to its original meaning (Hecker, 1972, 1975). “ “

I n cocarcinogenesis the response of the host o r target tissue as identified qualitatively by wellestablished histological means is measured quantitatively by the latent period of cancers (or tumors), by the cancer (or tumor) rate, by the cancer (or tumor) yield, or by a combination of these. The The basic elements of cocarcinogenesis were (specific) biological effect($ caused by cocarcinogens established in the now classical experiments with is (are) considered to be reversible (Berenblum, 1964; Boutwell, 1964; Graffi, 1964; Hecker, 1968, 1972, mouse skin, using, after exposure to subthreshold 1975). Cocarcinogenesis is the consequence of a doses of carcinogenic aromatic hydrocarbons as solitary carcinogens, the skin irritant croton oil as a synergism o f t he assumed irreversible specific carcinococarcinogen (Berenblum, 1 9 4 1 ~ ;Mottram, 1944; genic effect(s) caused by exposure to a subcarcinoBerenblum and Shubik, 1947). Later on it was genic dose of a solitary carcinogen (or to a n inestablished that the active principles of croton oil are complete carcinogen) and of the assumed reversible esters of the polyfunctional tetracyclic diterpene biological effects caused by subsequent exposure to phorbol such as -1 2-O-tetradecanoylphorbol-13- a cocarcinogen. Apart from its specific cocarcinoacetate (TPA) (Hecker, 1968; van Duuren, 1969; genic effect(s), a cocarcinogen may cause non-specific Hecker, 1971; Hecker and Schmidt, 1974). With epiphenomena in the host or target tissue(s) due, for cocarcinogenic doses of the order of magnitude of example, to its toxicity. hormone doses the phorbol esters are the most active It is important to note that reversible biological exogenous cocarcinogens presently known (Hecker, 1966, 1968). Other cocarcinogens of exogenous and effects can also cumulate by summation (Druckrey and endogenous origin (see Table I), including viruses, Kiipfmiiller, 1949). Hence cumulation of the underlying biological effect(s) may not be misinterpreted as indicating have been identified in experiments with hosts other irreversibility (Hecker, 1972). The detection of a definite than mice and in target tissues other than skin tumorigenic and even a weak carcinogenic activity of (Salaman and Roe, 1964; Roe, 1968; Roe and croton oil and of the phorbol esters (Hecker, 1956, 1968) Rowson, 1968; Berenblum, 1974). The role of tissue caused some investigators to question the existence of a specificity of the solitary (or incomplete) carcinogens defined cocarcinogenic activity (Hecker, 1972; Chouand of the cocarcinogens involved in cocarcino- roulinkov e t a / . , 1974). Indeed, if merely the end result of pluricarcinogenesis versus cocarcinogenesis is congenesis remains to be investigated further. sidered4.e. generation of neoplasia-a qualitative The term “ cocarcinogenesis ” was sometimes also difference in the biological activities of solitary carcinoused to include conditioning exposure(s) prior to the start gens and cocarcinogens might appear not to exist. of the process of carcinogenesis (e.g. Berenblum, 1969, Clearly, such a point of view does not differentiate the 1974) as defined above. Such usage is at variance with diverse biological qualities of action of the circinogens the more restricted meaning attributed to it at first by involved. Such a differentiation, however, will affect and Shear, 1938, and also with the definition of syncarcino- improve benefit/risk evaluations preceding preventive genesis (Bauer, 1949,1963). It also has the disadvantage of legislation. mixing what-because of heuristic reasons-might be That cocarcinogens may play a role in the etiology kept separate : i.e. carcinogens and carcinogenic processes, from host or target tissue predisposition (see above of human tumors is suggested by clinical observations (for reviews see Bauer 1949, 1963). Also, epidemiological evidence as to the etiological role of For 12-0-tetradecanoylphorbol-13-acetate (TPA, croton oil factor A,) some authors prefer the name cocarcinogens in certain occupational cancers has been reported (e.g. Bingham and Horton, 1966) phorbol-myristate-acetate (PMA), perhaps because of historical associations. Boutwell (1974) rightly pro- and verified experimentally by a number of investiposed to give up this confusing duality. The reasons gators (Salaman, 1958; Shear, 1938). Within the last for using the systematic name 12-0-tetradecanoylphorbol13-acetate (TPA) have been detailed elsewhere (Hecker, decade, the existence of a wide variety of cocarcinogens in the plant kingdom has become apparent. As 1975). I’

128

HECKER

compared to solitary carcinogens, they may be considered as second order carcinogenic risk factors (Hecker, 1972, 1975). The still-increasing number of cocarcinogens calls for careful investigations of their contribution to the etiology of human cancer and for adoption of adequate preventive measures in environmental hygiene. Such measures must seek to prevent exposure not only to solitary (or incomplete) carcinogens but also to cocarcinogens (Roe, 1968; van Duurenetal., 1971, 1973; Hecker, 1972; Horton and Christian, 1974; Narisawa et al., 1974). Studies of cocarcinogenesis make it possible to resolve the complexity of cellular and molecular events comprising carcinogenesis in defined stages. Such stages may be investigated separately to analyse the mechanism of carcinogenesis at the organism, the cellular and the molecular levels (Hecker, 1968, 1971 ; Boutwell, 1974; Berenblum, 1974; Hecker and Schmidt, 1974). In mechanistic investigations of this kind for the (subcarcinogenic) exposure to the solitary carcinogen (or exposure to the incomplete carcinogen) the term “ initiation” and for the cocarcinogen the term “promotion” may be used, respectively. Carcinogenic factors and unspecified carcinogenesis

For causative agents of which the origin, nature and identity is not clarified unequivocally, the term carcinogenic factor may be used. If in any specific case of carcinogenesis the number and/or sequence of exposure to the carcinogens or carcinogenic factors involved have not been clarified unequivocally, the carcinogenic process may be classified preliminarily as “ unspecified carcinogenesis ”. If, for example, it were established that “ hormoneinduced cancers ” are being generated by an unifactorial process, the process would be considered as solitary carcinogenesis and the hormone, consequently, as a solitary carcinogen. If, however, such cancers were the result of the synergistic action of e.g. a virus (perhaps already present in the target tissue) and of a hormone, such process would be considered as syncarcinogenesis still leaving for clarification the possibilities of either pluri- or cocarcinogenesis and hence classifying the hormone alternatively either as a solitary carcinogen or as a cocarcinogen (Hecker 1972). CONCLUSIONS

In cancer etiology generation of neoplasia in the broadest possible sense may be described by the term “ carcinogenesis ”. The term ‘‘ carcinogen ” may be used to cover the causative agents involvsd in carcinogenesis if these are identified unequivocally. The multiplicity of possible exposures of the host or target tissue according to kind(s), number@) and sequence(s) of carcinogens may be called “ carcino-

genic processes ”. Together with “ host or target tissue predisposition ” carcinogens and carcinogenic processes are considered fundamental variables of carc i nogenesis,

A carcinogenic process evoked by unifactorial exposure of the host or target tissue is specified as “ solitary carcinogenesis ”. The carcinogen involved is a “ solitary carcinogen ”. Carcinogenic processes evoked by multifactorial exposure of the host or target tissue are collectively called syncarcinogenesis ”. In cancer etiology exposure to even a subcarcinogenic dose of a solitary carcinogen (or to an incomplete carcinogen), may be causative of cancer, if the biological effects caused by the initial exposure, no matter how much later in life, are augmented by exposure to another solitary carcinogen (carcinogenic process : “ pluricarcinogenesis ’’1 or to a cocarcinogen (carcinogenic process : ‘‘ cocarcinogenesis ”). Per se, cocarcinogens are non-carcinogenic and non-initiating. In the etiology of cancer in man solitary carcinogens and cocarcinogens may be considered first- and second-order carcinogenic risk factors, respectively. Their distinction may contribute considerably to more differentiated evaluations of benefit versus risk. Detailed knowledge in this field constitutes the scientific foundation of cancerprevertfion by environmental hygiene. In experimental cancer research certain carcinogenic processes verified in appropriate hosts or target tissues may be used as experimental model systems to elucidate the biology and biochemistry of the cancer cell and of carcinogenesis. Such elucidations are a prerequisite for the development of causal therapies of cancer. ACKNOWLEDGEMENTS

The present analysis and proposals were stimulated by informal discussions between the author and members

of the UICC Committee of Quantitative Carcinogenesis (Chairman: P. Shubik), the Feasibilty Committee of the IARC (Chairman: J. Higginson) in November 1968 (Kingston, Jamaica), and the Committee of International Collaborative Activities (CICA) of the UICC, and reached its present format by many discussions with individual cancer research workers, e.g. during the 2nd Meeting of the European Association for Cancer Research in October, 1973 (Heidelberg, Germany), the XIth International Cancer Congress in September, 1974 (Florence, Italy) and the Third International Symposium on Detection and Prevention of Cancer in April 1976 (New York, USA). In seeking for a wide distribution of this paper within the scientific community of Cancer Research to encourage international discussion, the cooperation of the editors of International Journal of Cancer, Zeitschrift fur Krebsforschung und Klinische Onkologie, GANN, WHO-Bulletin and Journal of the National Cancer Institute (USA) is gratefully acknowledged.

DEFINITIONS AND TERMINOLOGY IN CANCER

129

REFERENCES

BAUER,K. H., Uber Syn- und Anticarcinogenese. Klin. Wschr., 27, 118-121 (1949). BAUER,K. H., Das Krebsproblem, 2, Auflage. SpringerVerlag, Berlin-Gottingen-Heidelberg ( I 963). BERENBLUM, I., The cocarcinogenic action of croton resin. Cancer Res., 1, 44-48 (1941~). BERENBLUM, I., The mechanism of carcinogenesis. A study of the significance of cocarcinogenic action and related phenomena. Cancer Res., I , 807-814 (19416). BERENBLUM, I., The two-stage mechanism of carcinogenesis as an analytical tool. In: P. Emmelot and 0. Muhlbock (ed.), Cellular control mechanism and cancer, pp. 259-267, Elsevier Publishing Co. Amsterdam, London and New York (1964). BERENBLUM, l., A re-evaluation of the concept of cocarcinogenesis. frog,. exp. Tumor Res., 11, 21-30 (1969). BERENBLUM, I., Carcinogenesis as a biological problem. In: A. Neuberger and E. L. Tatum (ed.), Frontiers of biology, Vol. 34, North-Holland Publishing Co., Amsterdam and Oxford; American Elsevier Publishing Co., Inc., New York (1 974).

BERENBLUM, I., and SHUBIK, P., A new quantitative approach to the study of the stages of chemical carcinogenesis in the mouse’s skin. Brit. J . Cancer, 1, 383-391 (1947). BINGHAM, E., and HORTON,A. W., Environmental carcinogenesis: experimental observations related to occupational cancer. In: W. Montagna and R. L. Dobson (ed.), Advances in biology o f t h e skin, Vol. 7 , pp. 183-193, Pergamon Press, Oxford (1966). BOUTWELL, R. K., Some biological aspects of skin carcinogenesis. Progr. exp. Tumor Res., 4, 207-250 (1964). BOUTWELL, R. K., The function and mechanism of promoters of carcinogenesis. Crir. Rev. Toxicol., 2, 419-443 (1974). CAIRNS,J., The cancer problem. Scient. Amer., 233, 64-78

Vol. VI, Geschwilste, Tumors, II, pp. 651-676, SpringerVerlag, Berlin and Heidelberg (1975). HECKER,E., and SCHMIDT,R., Phorbolesters-the irritants and cocarcinogens of croton tiglium L. Progr. Chem. Organ. Natur Prod., 31, 377-467 (1974). HORTON,A. W., and CHRISTIAN,F. M., Cocarcinogenic versus incomplete carcinogenic activity among aromatic hydrocarbons: contrast between chrysene and benzo(b)triphenylene. J . nat. Cancer Inst., 53, 1017-1020 (1974). HUEBNER, R. J., and TODARO,G. J., Oncogens of RNA tumor viruses as determinants of cancer. Proc. nut. Acad. Sci. (Wash.), 64, 1087-1094 (1969). LIKHACHEV, A. Y., Combined effect of the carcinogenic substances. Vopr. Onkol., 14, 114-124 (1968). In Russian. J., and RAULOT-LAPOINTE, G., ContribuMARIE,P., CLUNET, tion B I’ktude du developpement des tumeurs malignes sur les ulckres de Roentgen. Tumeur maligne d6veloppee sur une radiodermite experimentale chez le rat blanc. Bull. Ass. franc. Et. Cancer, 3, 404-426 (1910). MOHR,U.,Pluricarcinogenesis. Abstracts 2nd Meeting of the European Association for Cancer Research, 2-5 October, 1973, Heidelberg, p. 4, Deutsches Krebsforschungszentrum, Heidelberg (1973). MOTTRAM,J. C., A developing factor in experimental blastogenesis. J. Path. Bart., 56, 181-187 (1944). NARISAWA, T., MAGADIA, N. E., WEISBURGER, J. H., and WYNDER,E. L., Promoting effect of bile acids on colon carcinogenesis after jntrarectal instillation of N-methy1-N’nitro-N-nitrosoguanidine in rats. J . nut. Cancer Inst., 53, 1093-1097 (1974).

ROE,F. J. C., Carcinogenesis and sanity. Fd. Cosmet. Toxicol., 6, 485-498 (1968).

ROE,F. J. C. and ROWSON,K. E. K., The induction of cancer by combinations of viruses and other agents. I n / . Rev. exp. Pathology, Vol. 6, 181-227 (1968). CHOUROULINKOV, I., and LAZAR,P., Actions canctrogkne et Rous, P., A transmissible avian neoplasm (sarcoma of the cocancerogkne du 12-0-tttradtcanoylphorbol-13-acttate common fowl). J. exp. Med., 12, 696-708 (1910). (TPA) sur la peau de souris. C . R . Acad. Sci. (Paris), 278, Serie D, 3027-3030 (1974). SALAMAN, M. H., Cocarcinogenesis. Brit. med. Bull., 14, 116-120 (1958). DRUCKREY, H., and KUPFMULLER, K., Dosis und Wirkung. Editio Cantor, Aulendorf/Wurtt (1949). SALAMAN, M. H., and ROE, F. J. C., Cocarcinogenesis. Brit. med. Bull., 20, 139-144 (1964). EMMETT,E. A,, Ultraviolet radiation as a cause of skin tumors. Crit. Rev. Toxicol., 2, 21 1-255 (1973). SCHMAHI., D., Entstehung, Wachstum und Chemotherapie maligner Tumoren, 2. Aufl. Editio Cantor, Aulendorf/Wiirtt EPSTEIN,S., Environmental determinants of human cancer. ( I 970). Cancer Res., 34, 2425-2435 (1974). SHEAR, M. J., Studies in carcinogenesis. V. Methyl derivatives FALK,H. L., Anticarcinogenesis-an alternative. Progr. exp. of 1,2-benzanthracene. Amer. J . Cancer, 33, 499-537, 532 Tumor Res., 14, 105-137 (1971). (1938). GRAFFI,A., Experimente und Betrachtungen zur Natur und TEMIN, H. M., The protovirus hypothesis: speculations on Ursache des Krebses. Sitrungsberichte der Deutschrn the significance of RNA-directed DNA-synthesis for normal Akademie der Wissenshaften zu Berlin. Klasse fur Medizin, development and for carcinogenesis. J. nut. Cancer Inst., 46, Nr.2, S . 1-44, Akademie Verlag, Berlin (1964). 111-VII (1971). HECKER,E., Die cocarcinogene Wirkung der Phorbolester. VAN DUUREN. B. L., Tumor-promoting agents in two-stage In: H. Holzer and A. W. Holldorf (ed.), Molekdare Biologie carcinogenesis. Progr. exp. Tumor Res., 11, 31-68 (1969). des malignen Wachstums, pp. 105-116, Springer Verlag, Berlin, Heidelberg and New York (1966). VAN DUUREN,B. L., BLAZEJ,T., GOLDSCHMIDT, B. M., S., and SIVAK,A., Cocarcinogenesis KATZ,C., MELCHIONNE, HECKER,E., Cocarcinogenic principles from the seed oil of studies on mouse skin and inhibition of tumor induction. croton tiglium and from other euphorbiaceae. Cancer Res., J. nut. Cancer Inst., 46, 1039-1044 (1971). 28, 2338-2349 (1968). See also PIanta Medica Suppl., 24, 45 (1968). VAN DUUREN, B. L., KATZ, C., and GOLDSCHMIDT, B. M., Cocarcinogenic agents in tobacco carcinogenesis. J . nat. HECKER,E., Isolation and characterization of the cocarciCancer Inst., 51, 703-705 (1973). nogenic principles from croton oil. In: H. Busch (ed.), Methods in cancer reseurch, Vol. 6, pp. 439-484, Academic WHO, Occupational Cancer. Report of a W H O Consultation Press, New York and London (1971). 1971, document number WHO/OH/72.1I , World Health Organization, Geneva. HECKER,E., Aktuelle Probleme der Krebsentstehung. KrebsJiwschung, 78, 99-122 (1972). YAMAGIWA, K., and ISCHIKAWA, K., Uber die atypische Epithelwucherung. Gann, 8, I I (1914). See: K. Yamagiwa, HECKER,E., Cocarcinogens and cocarcinogenesis (with a Collected papers on artificial production of cancer, pp. 1-9, note o n synergistic processes in carcinogenesis). In: Maruzen Co. Ltd., Tokyo (1965). E. Grundrnann (ed.), Handbuch der Ailgemeinen Pathologie, (1975).