The Paul P. Carbone Award in International Oncology Recipient: Donald Maxwell Parkin, Bsc. MD, MRCP (UK), Dip.Soc.Med (Edin), FFCM

CANCER IN AFRICA

The provision of services for prevention and treatment of cancer has had a low priority with African governments and development agencies. In a way, it is reasonable to have focused on health problems that have been largely solved in the developed world (infant, child and maternal mortality, infectious diseases), but, unfortunately, these ‘old’ diseases co-exist in Africa with new ones, most evidently AIDS, but also non-communicable diseases, including cancer. Cancer is not rare in Africa. Even ignoring the huge load of AIDS-related Kaposi sarcoma, the probability of a woman living in present-day Kampala or Harare developing a cancer by age 65 is only about 20% lower than that of her sisters in Western Europe. Yet the facilities for cancer treatment in most of Africa are minimal, and there are almost no prevention programs worthy of the name.

WHO encourages all countries to establish a national cancer control program within a comprehensive, systematic framework, comprising prevention, early diagnosis, screening, curative therapy, and palliative care. Appropriate decision-making requires that epidemiological and programmatic data are available. Alas, knowledge of cancer patterns in Africa is woefully inadequate; until quite recently it was based primarily on the work of pioneering clinicians and pathologists who described the composition of series of cancer patients encountered in their professional lives in terms of age, sex, cancer site and histology. Unfortunately, comparisons based upon relative frequency of different cancers in case series can be very misleading. Pathology series under-represent cancers that are difficult to biopsy (such as liver, pancreas or brain). Hospital series are biased by the clinical facilities available (radiotherapy series always include, for example, many cancers of the head and neck and of the cervix, but few gastro-intestinal cancers). What is more, the use of proportions (or percentages) of different cancers as the statistic for comparison introduces a further problem: since the total must always equal 1 (or 100), if one cancer is ‘common’, all others in the series will appear to be rare.

The appropriate statistics for making comparisons of risk between populations are incidence rates; these derive from population-based cancer registries, which aim to record information on all new cases of cancer that occur in a defined population. Cancer registration has been slow to develop in Africa, for a variety of reasons. Certainly, defining the residents and cancer patients in a particular area, and obtaining accurate diagnostic information, is more difficult than in western countries. But it is not impossible. The monograph “Cancer in Africa: Epidemiology and Prevention,” just published by IARC, brings together information from all of the cancer registries in Africa (they cover about 8% of the population), and reviews the epidemiology of the major cancers.

The principal cancer today is cervix cancer, responsible for 12% of all new cases. There is not much evidence of a decrease in incidence (as observed elsewhere in the world), but the risk has not been increased by the AIDS epidemic. The other major cancers are breast (10%), liver (8%), Kaposi sarcoma (6%), non-Hodgkin’s lymphoma (5%), and prostate cancer (5%). Kaposi sarcoma has increased dramatically in parts of East and Central African where HIV prevalence is high, and in several countries it is the most common cancer of men (and second in women). The frequency of breast cancer in Africa appears surprising, as does the appearance of prostate cancer in the top 6; this may reflect the urban bias of existing data sources, where changing lifestyles mean that these malignancies are becoming increasingly common. Burkitt’s lymphoma was first described in Africa, and although now recognized to be a common form of lymphoma, specific features of the epidemiology in Africa may provide clues to understanding its aetiology. Liver cancer and bladder cancer also have singular epidemiological features that have been the focus of several studies. In most of Africa, two of the major cancers in more developed countries–lung cancer and colon cancer–remain rather rare.