The Cancer Journal - Volume 7, Number 4
Cancer and race
In reports of epidemiological studies of cancer carried out in America, racial or ethnic origin stands out as a frequently cited parameter, regardless of whether the natural history of the disease, prognosis, morbidity or mortality is being considered. This deserves some attention. What do these terms really mean? Are the racial and ethnic origins discussed based on genetic information or on social criteria? In either case, are they relevant? If they are, we should continue to use them, whatever we think of such classifications; if not, we should stop talking about them and applying them to cancerology.
Today it is clear, if not certain, that many cancers have a genetic element; that transformation at the cellular level is the result of mutation or gene malfunction. It is no less evident that certain types of cancer have a strong or very strong familial component, whether this be a direct effect on neoplastic transformation or an increased risk of developing a lesion or a precancerous disease.
The question which we are asking here is in what measure are the risks associated with "race" hereditary? If this question is asked with respect to dogs, the answer would be unambiguous. The different races of dogs have been and continue to be maintained by breeders on morphological criteria. Vetinary surgeons are well acquainted with the pathologies associated with different races: the frequency of osteosarcoma in large pedigree dogs, the frequency of mastocytomas in boxers, the very high frequency of cancers in strongly endogamic races. As a general rule the risk factors are higher in pedigree dogs than in mongrels. The situation is even more accentuated in the mouse, where isogenic inbred lines have been created. This is both one of the most artificial genetic models and one of the most instructive. Today it has been largely replaced by transgenic mice, which are even further from the human case. In dogs and in mice, we can be almost completely certain of the genetic background.
This is not true for humans. Apart from skin colour, a Hispanic, or rather Latin-American surname or accent, slanting eyes, prominent cheekbones and a few other physical traits, what do we know of the people whom we put into the usual categories: white or Caucasian, black or African (Afro-American in the United States), Oriental or Asiatic, Hispanic (never in Spain), Amerindian, etc.? In no publication, at least in no publication which I have read, is the basis of this racial classification described. So what is it based on? Is it the subject who gives his racial or ethnic origin in reply to a question from the epidemiologist, or does the epidemiologist himself decide on the classification? Furthermore, when a genetic approach is being used in a mouse model, we carefully distinguish isogenic mice from first and second generation hybrids and back-crosses. In man we do nothing of the sort, as if all subjects were "pure race". However, we can be certain, looking at ancient and recent history, that migrations, mixing and inter-breeding have made the human race one of the most heterogeneous in nature. There is abundant proof of this, ranging from the adaptability of man to all climates to the growing knowledge of genetic markers. Heterogeneity between the main groups shows itself in morphological terms, but this heterogeneity is much larger within each of these groups, with the possible exception of some isolated communities. This is evidence, if any were needed, of a history based on diversity.
So, at the scientific level, what is left of the association between race and cancer? Can one pass from studies of the familial basis of cancer, which is being better and better described (1) to a study of racial or ethnic origins? In their paper entitled "Differences in stage at presentation of breast and gynecologic cancers among whites, blacks and hispanics" (2), are the authors studying anything other than social, cultural and economic differences between these categories? The fact that blacks and Hispanics in Florida consult a doctor later is more related to their income level and their degree of integration into the American way of life than to the colour of their skin or their mother tongue. When we take a closer look we see that the real independent variables in all these studies in which "race" is mentioned are social parameters, which can be analysed, as long as they have not been overlooked. Furthermore, truly scientific genetic factors, such as HLA haplotypes, are the only acceptable risk factors of real value, even if masked behind "race", in an epidemiological study of cancer (3).
For the sake of scientific rigour, let us stop publishing work which, while not taking up a clear position on the question of race and cancer, leaves room for doubt. Let us say once and for all that this concept does nothing to advance our understanding of the disease, that it is either false genetics or misleading sociology and that it encourages intellectual laziness or the latent racism of both authors and readers.
CNRS, BP8, 94801 Villejuif, France
1. Cannon-Albright L.A., Thomas A., Goldgar D.E. et al. Familiality of cancer in Utah. Cancer Research 54, 2378-2385, 1994.
2. Chen F., Trapido E., Davis K. Differences in stage at presentation of breast and gynecologic cancers among whites, blacks, and hispanics. Cancer 73, 2838-2842, 1994.
3 - Gregoire L., Lawrence D.W., Kukuruga D. et al. Association between HLA-DQBI alleles and risk for cervical cancer in African-American women. Int. J. Cancer 57, 504-507, 1994.