Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Trends in cancer incidence in Kyadondo County, Uganda, 1960–1997

Trends in cancer incidence in Kyadondo County, Uganda, 1960–1997 British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign DOI: 10.1054/ bjoc.1999.1071, available online at http://www.idealibrary.com on Trends in cancer incidence in Kyadondo County, Uganda, 1960Ð1997 1 2 3 1 HR Wabinga , DM Parkin , F Wabwire-Mangen and S Nambooze 1 2 Department of Pathology, Makerere University, PO Box 7072, Kampala, Uganda; The International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France; Institute of Public Health, Makerere University, PO Box 7072, Kampala, Uganda Summary Incidence rates of different cancers have been calculated for the population of Kyadondo County (Kampala, Uganda) for four time periods (1960–1966; 1967–1971; 1991–1994; 1995–1997), spanning 38 years in total. The period coincides with marked social and lifestyle changes and with the emergence of the AIDS epidemic. Most cancers have increased in incidence over time, the only exceptions being cancers of the bladder and penis. Apart from these, the most common cancers in the early years were cervix, oesophagus and liver; all three have remained common, with the first two showing quite marked increases in incidence, as have cancers of the breast and prostate. These changes have been overshadowed by the dramatic effects of the AIDS epidemic, with Kaposi’s sarcoma emerging as the most common cancer in both sexes in the 1990s, and a large increase in incidence of squamous cell cancers of the conjunctiva. In the most recent period, there also seems to have been an increase in the incidence of non-Hodgkin lymphomas. So far, lung cancer remains rare. Cancer control in Uganda, as elsewhere in sub-Saharan Africa, faces a threefold challenge. With little improvement in the incidence of cancers associated with infection and poverty (liver, cervix, oesophagus), it must face the burden of AIDS-associated cancers, while coping with the emergence of cancers associated with Westernization of lifestyles (large bowel, breast and prostate). © 2000 Cancer Research Campaign Keywords: cancer registry; time trends; Africa; AIDS Kampala Cancer Registry was established in 1954 with the aim of Previous incidence rates for Kyadondo County have been obtaining information on cancer occurrence in the population of published for the periods 1954–1960 (Davies et al, 1962, 1965) Kyadondo County in which the capital city of Kampala is situated and 1968–1970 (Templeton et al, 1972). In addition, a comprehen- (Davies et al, 1965). The registry is located in the Department of sive descriptive analysis for the period 1964–1968, based on all Pathology in Makerere University Medical School. It functioned registrations (not restricted to Kyadondo residents, and without continuously both pre- and post-independence (1962), until the calculation of incidence rates), was published as a monograph coup d’état of General Amin Dada in 1971. Thereafter, full popu- (Templeton, 1973). The first results from the revitalized registry lation coverage was not possible, although a register was main- for a 28-month period in 1989–1991 were published in 1993 tained within the Pathology Department until 1980, when all (Wabinga et al, 1993). Here we review incidence rates for registration ceased. With the return of political stability, the Kyadondo County for two extended periods, 1960–1971 and registry was restarted (in 1989) and has functioned continuously 1991–1997, with a focus on the trends in risk of different cancers since. This long period of operation provides a unique opportunity both during and between these periods. to study temporal changes in cancer patterns in an African setting. Within Kyadondo, there have been quite marked social changes MATERIALS AND METHODS over the last 40 years. Progressive urbanization of the population has meant that, although the total population has increased more Cancer records than fourfold, Kampala city has grown almost tenfold, from a population of 92 000 in 1959 to 870 000 in 1995, and now repre- In 1953 the request/result forms of the Department of Pathology sents three-quarters of the county population. There has also been were redesigned specifically to permit registration of cancers. an increase in educational standards: the population of 10 years Thus they contain demographic information on the patient such as and above who have ever been to school increased among males name, age, sex, tribe and place of residence, as well as the source from 4.7% in 1959 to 71.8% in 1991 and among females from of the specimen and results of the examination. In addition to data 1.8% to 61.9%. In addition, Uganda is one of the countries most collected in this way, tumour registrars have been employed to affected by the HIV epidemic, and therefore temporal changes search for cancer cases admitted to, or treated in, the four main reveal its effect on cancer occurrence in Africans. hospitals in Kampala (and more recently in the Uganda Hospice) and, for individuals resident in Kyadondo County, to extract some- what more extensive information onto special notification forms. Between 1954 and 1980, registration was manual apart from the Received 18 June 1999 period 1964–1968 when the data were transferred to punchcards Revised 20 October 1999 (Templeton, 1973), which are no longer available, with details of Accepted 4 November 1999 all cancer cases identified being entered in a large ‘register’. Since Correspondence to: DM Parkin 1989 the registration process has been computerized, using the 1585 1586 HR Wabinga et al Table 1 Average annual population at risk, Kyadondo County, Uganda, and percentage by age group and sex 1960–1966 1967–1971 1991–1994 1995–1997 M F M F MFM F 0–14 30.8 39.0 34.3 43.5 40.6 43.5 40.6 43.5 15–44 57.9 50.9 55.8 47.6 52.4 50.0 52.4 50.0 45–55 5.9 4.8 5.1 4.2 4.0 3.2 4.0 3.2 55–64 2.9 2.8 2.5 2.2 1.6 1.7 1.6 1.7 65+ 2.3 2.4 2.2 2.4 1.3 1.7 1.3 1.7 Total 152 000 116 500 225 400 185 400 526 900 546 600 577 200 592 400 CANREG system (Cooke, 1998) which, at the stage of data entry, Statistical methods prevents the use of non-existent codes and performs checks for Incidence rates for each of the four periods were calculated for internal consistency between variables. It also permits a search for 5-year age groups. Age-standardized rates (world standard popula- potential duplicate registrations. tion) and 95% confidence limits were calculated as described in The records from the registers of 1960–1980 have also been Boyle and Parkin (1991). Differences between rates were tested by entered into the CANREG system (the registers for 1954–1959 were means of the Mantel–Haenzel test, stratifying for age in 5-year no longer available, and had probably been removed from the groups (Estève et al, 1994). country). Several duplicate entries were identified, as well as a considerable number of errors and missing information. Doubtful RESULTS records were completed, corrected or deleted, after tracing the original archives in the records of the hospitals and the pathology During the 19 years of registration considered (1960–1971; department. 1991–1997) a total of 7312 cases (3576 male and 3736 female) Tumour site and morphology were coded according to the were registered. Tables 2 and 3 show, for males and females second edition of the International Classification of Diseases for respectively, the number of cases in each of the four time periods, Oncology (Percy et al, 1990). For tabulation of results, these were by major cancer sites, together with the age-standardized inci- converted to the 10th revision of the ICD (WHO, 1992) and to the dence rates, and their standard errors. The statistical significance International Classification for Childhood Cancer (Kramárová of the change in the incidence rate from the preceding period is et al, 1997). shown. Since registration ceased to be population-based between 1972 In both sexes, the total incidence of cancer (all sites combined) and 1980, the data for this period were not used in the current increased across the four periods. The most dramatic increase analyses. With the availability of 8 full years’ registration since the between the decades of the 1960s and the 1990s is in the incidence restart of the registry in late 1989, it was clear that the case finding of Kaposi’s sarcoma (KS), especially for females, in whom KS for 1990 was inadequate and so the data for this year, included was previously a rather rare disease (sex ratio almost 20:1) but in in a previous publication (Wabinga et al, 1993) were excluded. whom the incidence rate is now around half that of men. Figure 1 Otherwise, inspection of the annual number of registrations (and shows age-specific incidence rates for KS, for 1960–1971 and proportions of registrations based on histological examinations) 1991–1997 in men and, for the latter period only, in women. In showed a regular progression with time. We therefore chose to males, as well as the increase in incidence, the shape of the curve consider four time periods: 1960–1966 (7 years); 1967–1971 has changed: in the 1960s there was a progressive increase in risk (5 years), 1991–1994 (4 years); 1995–1997 (3 years). with age, while in the 1990s there is a small peak in incidence in childhood, and a more marked one in the late 30s. In females, inci- dence in childhood is lower, and the peak incidence in adults is at ages 25–29. The mean age at diagnosis in the 1991–1997 period is Population data significantly later (P < 0.001) in men than in women both for all cases (32.0 vs 27.1), and for cases in adults aged 15 or more Population censuses were performed in 1959, 1969 and 1991. For (34.9 vs 29.8). these years, the population of Kyadondo County was available by Other than KS, the cancers showing the most marked increases sex and 5-year age group. An estimate for 1995 was provided in incidence in males are cancers of the prostate and oesophagus. by the Department of Statistics of the Ministry of Finance and There have also been significant increases in rates for cancers of Economic Planning; this was a simple extrapolation of the 1991 the large bowel and eye and, in the last 3 years, of non-Hodgkin result, assuming an annual growth in each age group of 2.61% for lymphoma. Although the changes in incidence between the indi- males and 2.30% for females. An interpolation between the census vidual time periods are not statistically significant, the trends in of 1959 and 1969 was prepared, assuming a constant rate of incidence rates between 1960–1966 and 1995–1997 are highly increase within age/sex groups, and projections for 1996 and 1997 significant (P < 0.01) for both stomach and lung cancers. were made, assuming the same growth rate as that provided for Conversely, there have been significant decreases in the incidence 1991–1995. rates of cancers of the bladder and penis, and of Hodgkin’s disease For the calculation of incidence rates in the four periods consid- and leukaemia. Liver cancer, which was the most common cancer ered, we used: the estimated population for 1963, the census of men in the 1960s, appears to have declined in incidence during population of 1969 and the estimated average populations for the 1990s. 1991–1994 and 1995–1997. These are shown in Table 1. British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1587 Table 2 Incidence of the major cancers in Kyadondo County, Uganda, in four time periods: males 1960–1966 1967–1971 1991–1994 1995–1997 No. ASR(standard No. ASR (standard No. ASR (standard No. ASR (standard error) error) error) error) Nasopharynx 3 0.3 (0.2) 9 1.8 (0.7) 11 0.7 (0.2) 26 2.3 (0.6) c c Oesophagus 8 1.7 (0.6) 25 5.1 (1.1) 83 15.8 (1.9) 68 13.0 (1.8) Stomach 14 2.7 (0.8) 20 4.7 (1.1) 31 4.7 (1.0) 37 7.6 (1.4) Colon & rectum 14 3.0 (0.8) 23 4.8 (1.1) 52 8.3 (1.3) 38 6.8 (1.4) c d d Liver 44 6.0 (1.1) 75 11.7 (1.6) 73 9.8 (1.4) 41 5.9 (1.2) Lung 5 0.8 (0.4) 10 2.1 (0.8) 25 4.1 (1.0) 19 3.2 (1.0) Melanoma 1 0.1 (0.1) 8 1.3 (0.6) 7 1.5 (0.7) 7 1.1 (0.5) Skin 26 3.7 (0.8) 27 4.3 (0.9) 18 2.5 (0.7) 20 4.1 (1.0) Kaposi’s sarcoma 28 3.2 (0.7) 29 3.7 (0.8) 670 39.3 (2.1) 513 39.3 (2.3) a c a Prostate 13 3.1 (0.9) 27 6.8 (1.4) 113 26.3 (2.6) 139 39.2 (3.7) Penis 29 5.5 (1.1) 30 6.3 (1.2) 18 2.9 (0.8) 23 4.4 (1.1) Bladder 24 5.2 (1.1) 27 5.9 (1.2) 13 2.5 (0.8) 10 2.9 (0.9) Eye 5 0.4 (0.2) 12 1.1 (0.4) 43 2.3 (0.4) 47 3.0 (0.6) Hodgkin’s disease 16 1.9 (0.6) 17 1.7 (0.5) 8 0.8 (0.4) 11 1.3 (0.6) NHL 32 3.9 (0.8) 32 3.6 (0.7) 76 3.6 (0.5) 95 7.4 (1.1) Leukaemia 22 2.2 (0.6) 24 3.2 (0.9) 13 0.7 (0.2) 16 1.1 (0.3) ALL 352 54.2 (3.3) 478 81.2 (4.3) 1456 149.1 (5.2) 1290 166.6 (6.2) ALL (except KS) 324 51.0 449 77.5 780 109.8 777 127.3 a b c d e f Significant increase since preceding period: P < 0.05; P < 0.01; P < 0.001.Significant decrease since preceding period: P < 0.05; P < 0.05; P < 0.001. Male (1960–1971) 80 Male (1991–1997) Female (1991–1997) 5-9 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ 0-4 10-14 Age group Figure 1 Age-specific incidence rates of Kaposi’s sarcoma In females, the increasing trends for cancers of the oesophagus, nomas (SCCs). In 1960–1971, 28/48 (58%) of histologically stomach, large bowel, eye and non-Hodgkin lymphomas are confirmed bladder cancer cases were SCCs, and just 10/48 (21%) similar to those in men, and there have also been increases in transitional cell/adenocarcinomas. In 1991–1997 the percentages incidence of the two major cancers of women – breast cancer and were 19% (3/16) SCC and 56% (9/16) transitional cell carcinomas. cervix cancer. Figure 2 shows age-specific incidence rates of Incidence rates of oesophageal cancer are similar in the two cervix cancer in the four different periods. The progressive sexes, which is unusual for this tumour, even in the African conti- increase of incidence with age in the earlier periods is replaced in nent. Forty-one per cent of the oesophageal cancers had informa- the 1990s by a pattern of peak incidence around menopause, tion on histological subtypes. All were carcinomas, with 87.1% followed by a plateau. Nevertheless, there has been no significant squamous, 7.1% adenocarcinomas and 5.7% unspecified subtype; alteration in the mean age at diagnosis between the four time there was no evidence of a change in the proportions over time. periods (43.8, 42.8, 43.2 and 43.7). The large increase in incidence of eye cancers is the conse- The decline in bladder cancer incidence between the 1960s and quence of increasing incidence of SCCs of conjunctiva. These 1990s has concerned almost exclusively squamous cell carci- increased from 4/17 eye cancers (23.5%) in men in 1960–1971 © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 Incidence rates per 100 000 1588 HR Wabinga et al Table 3 Incidence of the major cancers in Kyadondo County, Uganda, in four time periods: females 1960–1966 1967–1971 1991–1994 1995–1997 No. ASR (standard No. ASR (standard No. ASR (standard No. ASR (standard error) error) error) error) Nasopharynx 3 0.7 (0.4) 2 0.3 (0.2) 13 0.9 (0.3) 21 1.6 (0.4) c a Oesophagus 9 2.6 (0.9) 31 7.9 (1.5) 55 9.4 (1.3) 63 14.2 (1.9) Stomach 4 0.8 (0.5) 14 3.4 (1.0) 22 3.2 (0.7) 28 5.6 (1.1) Colon & rectum 11 2.7 (0.9) 22 6.3 (1.5) 36 5.7 (1.1) 34 6.6 (1.2) Liver 9 1.8 (0.6) 21 5.0 (1.3) 42 5.1 (1.0) 35 6.3 (1.3) Lung 3 0.6 (0.4) 6 1.4 (0.6) 7 0.7 (0.3) 18 3.2 (0.9) Melanoma 7 1.8 (0.7) 9 2.5 (0.8) 8 1.3 (0.5) 8 2.2 (0.8) Skin 16 4.0 (1.1) 15 3.1 (0.9) 12 1.4 (0.5) 8 1.0 (0.4) Kaposi’s sarcoma 1 0.1 (0.1) 2 0.2 (0.1) 360 17.9 (1.2) 335 21.8 (1.5) Breast 52 11.7 (1.8) 45 9.8 (1.6) 161 19.1 (1.8) 146 22.0 (2.1) Cervix uteri 84 17.7 (2.2) 109 22.5 (2.5) 341 39.7 (2.5) 296 44.1 (3.0) Corpus uteri 14 3.1 (0.9) 18 4.6 (1.3) 30 4.1 (0.9) 21 4.0 (0.9) Ovary 26 5.7 (1.3) 19 3.2 (0.8) 62 6.9 (1.1) 41 5.3 (1.0) Vulva/vagina 8 1.8 (0.7) 10 2.1 (0.8) 7 0.6 (0.3) 11 1.6 (0.6) c a Eye 4 0.3 (0.2) 3 0.2 (0.1) 37 1.7 (0.4) 45 3.4 (0.7) Thyroid 5 1.3 (0.6) 12 3.0 (1.0) 22 2.6 (0.7) 34 5.6 (1.1) Hodgkin’s disease 2 0.6 (0.5) 7 0.7 (0.3) 7 0.2 (0.1) 13 0.9 (0.3) NHL 14 2.2 (0.7) 15 2.2 (0.7) 48 2.1 (0.4) 82 5.7 (0.9) Leukaemia 13 2.3 (0.7) 20 2.8 (0.7) 17 1.2 (0.4) 17 1.9 (0.6) ALL 338 73.0 (4.4) 469 98.9 (5.3) 1508 146.8 (4.8) 1421 179.7 (6.0) ALL (except KS) 337 72.9 467 98.7 1148 128.9 1086 157.9 a b c d e f Significant increase since preceding period: P < 0.05; P < 0.001; P < 0.001. Significant decrease since preceding period: P < 0.05; P < 0.05; P < 0.001. Table 4 Cancer in children (age 0–14), Kyadondo County, 1960–1971 and 1991–1997 1960–1971 1991–1997 6 6 n (%) M:F Age standard rate (per 10 ) n (%) M:F Age standard rate (per 10 ) Leukaemia 25 (18.4) 1.1 18.7 27 (4.9) 0.7 8.6 Hodgkin’s disease 11 (8.1) 2.7 8.7 9 (1.6) 1.3 2.8 Burkitt’s lymphoma 13 (9.6) 1.6 9.5 109 (19.7) 1.5 34.3 Other NHL 18 (13.4) 2.6 13.1 61 (11.1) 1.3 19.1 Brain & CNS 4 (2.9) 0.3 3.0 7 (1.3) 2.5 2.3 Neuroblastoma 5 (3.7) 1.5 2.9 1 (0.2) – 0.3 Retinoblastoma 16 (11.8) 1.7 9.4 33 (6.0) 1.4 9.3 Wilm’s tumour 10 (7.4) 0.7 6.1 29 (5.3) 1.6 8.6 Osteosarcoma 5 (3.7) 1.5 4.2 6 (1.1) 4.0 1.9 Kaposi’s sarcoma 3 (2.2) 2.0 2.5 183 (33.2) 1.5 55.8 Other STS 10 (7.4) 4.0 7.6 20 (3.6) 0.8 6.0 Carcinomas 9 (6.7) 2.0 7.4 18 (3.3) 1.3 5.7 Total 136 (100) 1.6 97.8 552 (100) 1.4 169.7 Includes lymphoma (unspecified). (none of the seven eye cancers in women), to 32/45 (71% in men) BL was significantly higher (P < 0.001) in the 1990s than in the and 34/40 (85%) in women in the most recent 3-year period (there 1960s. This is unlikely to be due to changing classification or were a few cases of unknown histological type). coding, since the incidence of other and unspecified non-Hodgkin Table 4 shows incidence rates for certain cancers of childhood lymphomas was also higher in the second period (Table 4). In (ages 0–14) in two periods, 1960–1971 and 1991–1997. The most contrast, the incidence of childhood leukaemia is low, and there striking change is the increased incidence of KS, from 2.5 per was a significant (P < 0.05) fall in incidence between the two million in 1960–1971, to 55.1 per million in 1991–1997. The male periods. predominance has persisted (sex ratio 1.5:1), and KS has become the most common tumour of childhood, ahead of the lymphomas. DISCUSSION The mean age of these cases was 5.0, with no difference between the sexes, incidence was slightly higher in the age group 5–9 than In order to study time trends, it is important that the degree of at 0–4 (Figure 4). Lymphomas remain relatively frequent, with completeness of registration of incident cancer cases should be Burkitt’s lymphoma (BL) by far the most frequently diagnosed in similar throughout the period under consideration. In recent years 1991–1997; the peak age is at 5–9 (Figure 3) with a mean age of (since 1991) we believe that registration has been relatively good; 6.6, and boys predominate with a ratio of 1.4:1. The incidence of an exercise in independent case ascertainment (Parkin et al, 1994), British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1589 Table 5 Percentage of cases registered with diagnosis based on cytology Cervix cancer or histology (MV%) Period Cancer site 1960–1966 1967–1971 1991–1994 1995–1997 Oral cavity and pharynx 91 97 59 84 Oesophagus 65 68 28 41 Stomach 83 74 30 57 Colon and rectum 86 87 58 69 Liver 85 79 33 36 Lung 75 81 41 76 Kaposi’s sarcoma 76 94 78 86 Breast 82 88 56 68 Cervix 74 70 58 64 Prostate 62 67 66 80 Bladder 89 77 45 50 Lymphoma 96 93 72 80 Leukaemia 54 71 43 39 All cancers 84 84 64 73 1960 1966 1967 1971 missed. Finally, the wider availability of diagnostic modalities, 1991 1994 such as endoscopy, in the 1990s may have improved case finding 1995 1997 for certain cancers. The results should be interpreted, therefore, in the light of increases which could be due to better identification of cases diagnosed without histology. Templeton et al (1972) note that decreasing incidence rates are likely to be more meaningful 25-29 35-39 40-44 50-54 15-19 20-24 30-34 45-49 55-59 60-64 65-69 70-74 75+ than small increases in apparent incidence which might be related Age group to improvements in diagnostic facilities. In addition to these considerations, comparison of incidence Figure 2 Age-specific incidence rates of cancer of the cervix uteri rates requires that accurate population denominators are available. For Kyadondo, we had available estimates by age group and sex for three points during the long period studied, and made use of interpolations and projections for the other years. There must be Burkitt’s lymphoma Kaposi’s sarcoma some question as to the accuracy of the estimates of person-years 78.0 at risk, therefore, particularly with respect to the age distributions, 66.6 65.7 although no better data are, of course, available. 47.8 Davies et al (1964) reviewed hospital records from Mengo 41.4 Hospital (in Kyadondo County) from 1897 to 1956. They noted 28.2 that the general pattern of malignant diseases admitted to the hospital changed little over this period. The first estimation of inci- dence rates from the cancer registry of Kyadondo County was for 0—4 5—9 10—14 0—4 5—9 10—14 1954–1960 (Davies et al, 1962, 1965). The overall incidence reported for 1954–1960 was a little higher than our result for Figure 3 Age-specific incidence rates for Burkitt’s lymphoma and Kaposi’s sarcoma of childhood in 1991–1997 (both sexes) 1960–1966 (crude rates of 39.5 vs 33.1 for men and 48.4 vs 41.5 for women). It is possible that registration was rather more complete during the earlier period, when the proportion of cases comparing data collected for epidemiological studies in Kampala registered without histology was 27.5%, compared with 16.4% in with the registry database, suggested that the registry had identi- 1960–1966. On the other hand, experience with computer checks fied 90% of incident cancers. For the earlier periods, we can be on the manual register for 1960–1971 showed that several dupli- less certain. cate registrations and non-malignant cases had been included, and Some indication of completeness of registration is also provided these may also have inflated the numbers in the earlier time period. by the level of histological (or cytological) verification of regis- Nevertheless, for the individual cancer sites, the age-standardized tered cancer cases (Parkin et al, 1994). Table 5 shows that histo- rates are very similar in the two periods. logical verification was considerably more frequent in the earlier Incidence rates for the period 1967–1971 were very similar to time periods. This may be only a partial reflection of reality; those reported for 1968–1970 by Templeton et al (1972) with certainly, before the country was plunged into political, social and crude all-sites rates of 42.5 and 42.4 per 100 000 in men, and 50.2 economic chaos in the 1970s the standard of diagnostic (and thera- and 50.6 per 100 000 in women. peutic) services was higher than in the period of reconstruction in Over the 38-year period, the cancers showing the most dramatic the early 1990s. But it is also possible that the case-finding mech- increases in incidence are those related to infection with HIV, anisms particularly in the first period studied here were less active particularly KS, and SCC of the conjunctiva (Beral and Newton, than in the 1990s and that some cases without biopsies were 1998). KS has always been observed in the Ugandan population, © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 Incidence rates per 100 000 1590 HR Wabinga et al although in the 1950s and 1960s it was of the typical ‘endemic’ in 1985, has since been increasing again (FAO, 1998). On the pattern, involving the skin, particularly the legs, and affecting other hand, the importance of cereal-based diets, particularly those principally males, with the risk rising progressively with age relying upon maize, has been stressed with respect to the risk in (Taylor et al, 1971; Templeton, 1981). The enormous increase in Africa, and in particular in explaining the marked differences in incidence of KS since the earlier periods, together with the risk within quite small geographic areas (Cook, 1971; van narrowing of the sex-ratio (from 18:1 in 1960–1971 to 1.7:1 in the Rensburg, 1981). The mechanism could be via nutritional defi- 1990s) was noted in a previous report (Wabinga et al, 1993), and ciencies associated with such diets (van Rensburg et al, 1983; Ziegler and Katangole-Mbidde (1996) have drawn attention to the Jaskiewicz, 1989) or contamination of maize with mycotoxins dramatic increase in the incidence of KS in children. Uganda, in (Sydenham et al, 1990). However, maize is not the staple food in common with other central African countries, has a very high Kyadondo area of Uganda, where consumption of a variety of prevalence of infection with HIV, and the age-specific incidence of plantains is the major source of carbohydrate. KS corresponds closely to the age-specific reporting rates for The rather higher incidence rates of gastric cancer in the more AIDS, which are highest at ages 30–39 for men and 20–29 for recent periods possibly reflects better diagnosis because of the women (UNAIDS, 1998). The prevalence of HIV infection among availability of gastroscopy; however, the incidence of this cancer pregnant women had reached 30% in the urban population of remains low. This is certainly not due to a low prevalence of Kampala in 1990–1992 (UNAIDS, 1998), but since that time it has Helicobacter pylori – recognized by the IARC (1994) as an impor- been reported to be in decline (to around 15% in 1996–1997). This tant cause – since the infection has been shown to be common, at may explain the relative stability in the incidence of KS in least in the western part of Uganda (Wabinga, 1996). Tumours of 1995–1997 relative to 1991–1994 observed in this study. the large bowel are also rare – an observation made by Burkitt 30 Currently, the evidence suggests that human herpes virus 8 years ago (Burkitt, 1971), although there seems to have been (HHV8) is the aetiological agent responsible for KS (IARC, something of an increase in incidence in both sexes since the 1998). HHV8 has been identified in over 85% of KS tissue speci- 1950s and early 1960s, but little evidence of any recent change. mens in Uganda (Chang et al, 1996). Seroprevalence studies The reported rates of liver cancer in Uganda have remained low suggest a relatively high prevalence of infection by HHV8 in the in comparison with other parts of sub-Saharan Africa although of general population of Uganda – considerably higher than in the course much higher than in Western populations. The reasons for USA and Europe, which would be consistent with the elevated this are not apparent. The prevalence of chronic carriage of frequency of ‘endemic’ KS which preceded the AIDS epidemic hepatitis B in Uganda is similar to that in other countries, and afla- (Gao et al, 1996; Simpson et al, 1996). toxin contamination of foodstuffs appears to be common (Sebunya Unlike KS, the incidence of non-Hodgkin lymphomas, which and Yourtee, 1990), though less perhaps in Kampala than in other are found to be particularly common among immunosuppressed parts of the country (Alpert et al, 1971). individuals in Europe and USA, had remained relatively stable in The incidence rates of both breast and cervix cancer have Kyadondo County in the early 1990s (Wabinga et al, 1993). approximately doubled between the 1950s and 1960s, and the However, in 1995–1997 there appears to have been a significant 1990s. The reasons for these changes are not immediately clear. increase in incidence both in males and females. This may relate to They are unlikely to be related to the epidemic of AIDS for, improved survival of patients with HIV infection as other oppor- although cervix cancer is considered to be an AIDS-defining tunistic infections are controlled, permitting a more prolonged condition in the US, the evidence of a link between infection with duration of immunosuppression, and the development of more HIV and the risk of invasive cervical cancer is not consistent. The clinically-evident lymphomas. role of oncogenic human papillomaviruses (HPV) in the aetiology Squamous cell carcinomas of the conjunctiva have been recog- of cervix cancer is clearly established (IARC, 1995). HPV was nized for many years as more common in Africans than in found to be present in the great majority of cervix cancers in Europeans, but there is also a considerable (tenfold or more) Uganda, with HPV type 16 in 53% (Bosch et al, 1995). It seems increase in risk in the presence of HIV infection (Newton, 1996). quite plausible that the social disruption of the Amin dictatorship Ateenyi-Agaba (1995) has reported a large increase in the and following civil wars (1972–1986) had favoured the spread of numbers of cases presenting clinically in Kampala since the onset HPV, like other sexually transmitted diseases. As far as breast of the AIDS epidemic. cancer is concerned, it is possible that some of the increase is The increase in the incidence of oesophageal cancer between related to declines in fertility. Census data suggest that urban the 1960s and 1990s, particularly among males, is difficult to dwellers, and those with higher educational levels, have lower explain. Studies in southern Africa, as elsewhere, have pointed to than average fertility, but it is unlikely that changes in these para- the importance of tobacco smoking and, less certainly, alcohol as meters could explain a doubling of incidence rates. Screening being important in aetiology (van Rensburg et al, 1985; Segal et al, programmes for these cancers are virtually non-existent and can 1988; Vizcaino et al, 1995). The sex ratio of oesophageal cancer in hardly have influenced trends. There has been virtually no change Uganda has always been quite close to one to one, which suggests in the incidence of other female cancers – corpus uteri, ovary and that tobacco and alcohol are not of major importance in the rela- vulva/vagina. tively high, and increasing, rates, since smoking (in particular) is In men, the incidence of cancer of the prostate has increased largely confined to males. Furthermore, tobacco smoking is not remarkably, from an age-standardized rate of 3–6 per 10 in the common (average consumption per adult is 300 manufactured 1950s/60s, to 40 per 10 in the late 1990s. This is one of the cigarettes per year), there has been little change between 1970 and highest incidence rates recorded in Africa (Ferlay et al, 1998). 1990 (WHO, 1997), and the incidence of lung cancer has remained Most of this increase is in elderly men – aged 65 or over – low in both sexes. Consumption of alcoholic beverages, on the although the actual rates are not exceptionally high, compared other hand, is quite high (IARC, 1988) and after a decline through with those reported in Europe and North America. The increase in the 1970s to a low point of around 140 calories per caput per day Uganda is certainly not due to screening, although it is quite likely British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1591 Ateenyi-Agaba C (1995) Conjunctival squamous-cell carcinoma associated with that increased awareness, a greater readiness to perform prostatec- HIV infection in Kampala, Uganda. Lancet 345: 695–696 tomy for urinary symptoms in elderly men, and histological exam- Beral V and Newton R (1998) Overview of the epidemiology of immunodeficiency- ination of operative biopsies have played a role. The level of associated cancers. J Natl Cancer Inst Monogr, pp. 1–6 histological confirmation of diagnosis has actually increased over Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, Schiffman MH, Moreno V, Kurman R and Shah KV (1995) Prevalence of human time (Table 3) in contrast to virtually all other sites. papillomavirus in cervical cancer: a worldwide perspective. International In contrast to prostate cancer, the incidence of penile cancer in biological study on cervical cancer (IBSCC) Study Group. J Natl Cancer Inst men is lower in the 1990s than in the 1960s (although the rates 87: 796–802 remain high – the age-standardized rate of 3–4 in Kampala, can be Boyle P, Parkin DM (1991) Statistical methods for registries In: Cancer compared with 0.8 in the black population in the US SEER Registration: Principles and Methods, Jensen OM, Parkin DM, MacLennan R, Muir CS and Skeet R (eds), pp. 126–158. IARC Scientific Publication No. 95. registries (Parkin et al, 1997)). Penile cancer was clearly very The International Agency for Research on Cancer: Lyon frequent in the early case series from Kampala (Davies et al, Bradley DJ, Sturrock RF and Williams PN (1967) The circumstantial epidemiology 1964), and Dodge et al (1973) found it ‘the commonest tumour of Schistosoma haematobium in Lango district, Uganda. East Afr Med J 44: registered in males’ in 1964–1968. This decline in incidence is 193–204 Burkitt DP (1971) Epidemiology of cancer of the colon and rectum. Cancer 28: 3–13 probably real, since penile cancer is easily diagnosed and probably Chang Y, Ziegler J, Wabinga H, Katangole-Mbidde E, Bashoff C, Schultz T, Whitby always brought for medical attention. Penile cancer has been D, Maddelena D, Jaffe HW, Weiss RA, the Uganda Kaposi’s Sarcoma Study related to genital hygiene (Kyalwazi, 1966) and the decline in inci- Group and Moore PS (1996) Kaposi’s sarcoma-associated herpesvirus and dence may be related to improved hygiene as a consequence of Kaposi’s sarcoma in Africa. Arch Int Med 156: 202–204 urbanization and greater availability of piped water supplies. Cook P (1971) Cancer of the oesophagus in Africa. Br J Cancer 25: 853–880 Cooke A (1998) CANREG3 Manual. Internal Report No 98/03. International Agency Bladder cancer is also significantly lower in the 1990s than in for Research on Cancer: Lyon the 1960s. In Uganda, bladder cancer has been linked to the pres- Davies JNP, Wilson BA and Knowelden J (1962) Cancer incidence of the African ence of urethral stricture. Dodge (1964) found 30% of bladder population of Kyadondo (Uganda). Lancet 2: 328–330 cancer causes had such strictures, and Owor (1975) found 4% of Davies JNP, Elmes S, Hutt MSR, Mtimavalye LAR, Owor R and Shaper L (1964) Cancer in an African community, 1897–1956. Br Med J 1: 259–264 patients with strictures developed bladder cancer. Since strictures Davies JNP, Knowelden J and Wilson BA (1965) Incidence rates of cancer in are a sequel to gonococcal infection, it is possible that better treat- Kyondondo county, Uganda 1954–1960. J Natl Cancer Inst 35: 789–821 ment for STDs may have reduced their prevalence. Bladder cancer Dodge OG (1964) Tumours of the bladder and urethra associated with urinary is also a consequence of infection with Schistosoma haematobium retention in Ugandan Africans. Cancer 17: 1433–1436 (IARC, 1994). However, in Uganda Schistosoma haematobium Dodge OG, Owor R and Templeton AC (1973) Tumours of the male genitalia. In: Tumours in a Tropical Country. Recent Results in Cancer Control 41, AC has a rather restricted range and Kyadondo County is not among Templeton (ed). Springer: Berlin the endemic areas (Bradley et al, 1967). Estève J, Benhamou E and Raymond L (eds) (1994) Statistical Methods in Cancer Research. Vol IV, Descriptive Epidemiology. IARC Scientific Publication No. CONCLUSIONS 128. The International Agency for Research on Cancer: Lyon FAO (1998) FAOSTAT Statistics Database: Food balance sheets This long time series shows clearly the development of the (http://apps.fao.org/) problem of cancer in modern sub-Saharan Africa. New problems Ferlay J, Parkin DM and Pisani P (1998) GLOBOCAN: Cancer Incidence and are emerging in the shape of AIDS-related cancers (KS, NHL, Mortality Worldwide. IARC CancerBase No. 3. IARC: Lyon Gao SJ, Kingsley L, Li M, Zheng W, Parravicini C, Ziegler J, Newton R, Rinaldo conjunctival neoplasms), and cancers associated with CR, Saah A, Phair J, Detels R, Chang Y and Moore PS (1996) KSHV Westernization of lifestyles (breast and prostate cancers in partic- antibodies among Americans, Italians and Ugandans with and without Kaposi’s ular, large bowel cancers less certainly). At the same time, there sarcoma. Nat Med 2: 925–928 has been little, if any, decline in the major cancers of this region – IARC (1988) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, Vol 44, Alcohol Drinking. International Agency for Research: Lyon cervix, liver and oesophagus. So far, Uganda, as most of the IARC (1994) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, region, has been spared the epidemic of tobacco-related cancer Volume 61, Schistosomes, Liver Flukes and Helicobacter pylori. International that has been such an important feature of the cancer profile in Agency for Research: Lyon economically developed countries. This reprieve is probably only IARC (1995) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, temporary, however, and, while economic hardship may limit Volume 64, Human Papillomaviruses. International Agency for Research: Lyon IARC (1998) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, tobacco consumption and its health consequences, it will also Vol 70, Epstein-Barr Virus and KaposiÕs Sarcoma Herpesvirus/Human restrict the capacity to mount effective programmes of cancer Herpesvirus 8. International Agency for Research: Lyon control. Jaskiewicz K (1989) Oesophageal carcinoma: cytopathology and nutritional aspects in aetiology. Anticancer Res 9: 1847–1852 Kramárová E, Stiller CA, Ferlay J, Parkin DM, Draper GJ, Michaelis J, Neglia J ACKNOWLEDGEMENTS and Quershi S (1997) The International Classification of Childhood Cancer. We acknowledge the cooperation and assistance of the medical IARC Technical Report No. 29. International Agency for Research on Cancer: nursing and records staff of Mulago, Rubaga, St Francis Lyon Kyalwazi SK (1966) Carcinoma of the penis. A review of 153 patients admitted to (Nsambya), and Mengo hospitals, and the Uganda Hospice, for Mulago Hospital, Kampala, Uganda. East Afr Med J 43: 415–425 their assistance in tracing and recording cancer patients, as well as Newton R (1996) A review of the aetiology of squamous cell carcioma of the the pathologists and haematologists for access to their laboratory conjunctiva. Br J Cancer 74: 1511–1513 records. We thank Ms Zam Kyalisiima for assistance in coding and Owor R (1975) Carcinoma of bladder and urethra in patients with urethral strictures. East Afr Med J 52: 12–18 data entry, and Mr Andy Cooke for help with the data analysis. Parkin DM, Chen VW, Ferlay J, Galceran J, Storm HH and Whelan SL (1994) Comparability and Quality Control in Cancer Registration. IARC Technical REFERENCES Report No. 19. International Agency for Research on Cancer: Lyon Parkin DM, Whelan SL, Ferlay J, Raymond L and Young J (Eds) (1997) Cancer Alpert ME, Hutt MSR, Wogan GN and Davidson CS (1971) Association between Incidence in Five Continents, Vol VII. IARC Scientific Publication No 143. aflatoxin content of food and hepatoma frequency in Uganda. Cancer 28: The International Agency for Research on Cancer: Lyon 253–260 © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 1592 HR Wabinga et al Percy C, Van Holten V and Muir C (eds) (1990) International Classification of UNAIDS (1998) Epidemiological fact sheet on HIV/AIDS and sexually transmitted Diseases for Oncology, 2nd edn. World Health Organization; Geneva, diseases Uganda. http://www.who.int/emc-hiu/fact_sheets/ Switzerland van Rensburg SJ (1981) Epidemiologic and dietary evidence for a specific Sebunya TK and Yourtee DM (1990) Aflatoxigenic Aspergilli in foods and feeds in nutritional predisposition to esophageal cancer. J Natl Cancer Inst 67: 243–251 Uganda. J Food Quality 13: 97–197 van Rensburg SJ, Ambrose SB, Rose EF and Plessis JP (1983) Nutritional status of Segal I, Reinach SG and de Beer M (1988) Factors associated with oesophageal African populations predisposed to esophageal cancer. Nutr Cancer 4: 206–214 cancer in Soweto, South Africa. Br J Cancer 58: 681–686 van Rensburg SJ, Bradshaw ES, Bradshaw D and Rose EF (1985) Oesophageal Simpson GR, Schulz TF, Whitby D, Cook PM, Boshoff C, Rainbow L, Howard MR, cancer in Zulu men. South Africa: a case–control study. Br J Cancer 51: Gao SJ, Bohenzky RA, Simmonds P, Lee C, de Ruiter A, Hatzakis A, Tedder 399–405 RS, Weller IV, Weiss RA and Moore PS (1996) Prevalence of Kaposi’s Vizcaino AP, Parkin DM and Skinner MEG (1995) Risk factors associated with sarcoma associated herpesvirus infection measured by antibodies to oesophageal cancer in Bulawayo, Zimbabwe. Brit J Cancer 72: 769–773 recombinant capsid protein and latent immunofluorescence antigen. Lancet Wabinga HR (1996) Frequency of Helicobacter pylori in gastroscopic biopsy of 348: 1133–1138 Ugandan Africans. East Afr Med J 73: 691–693 Sydenham EW, Thiel PG, Marasas FO, Shephard GS, van Schalkwyk DJ and Koch Wabinga HR, Parkin DM, Wabwire-Mangen F and Mugerwa JW (1993) Cancer in KR (1990) Natural occurrence of some fusarium mycotoxins in corn from low Kampala, Uganda, in 1989–91: Changes in incidence in the era of AIDS. Int J and high esophageal cancer prevalence areas of the Transkei, Southern Africa. Cancer 54: 26–36 J Agric Food Chem 38: 1900–1903 WHO (1992) International Statistical Classification of Diseases and Related Health Taylor JF, Templeton AC, Vogel CL, Ziegler JL and Kyalwazi SK (1971) Kaposi’s Problems, 10th Revision. World Health Organization: Geneva, Switzerland sarcoma in Uganda: a clinico-pathological study. Int J Cancer 8: 122–135 WHO (1997) Tobacco or Health: A Global Status Report. Country Profiles by Templeton AC (ed) (1973) Tumours in a Tropical Country. Recent Results in Region. World Health Organization: Geneva, Switzerland Cancer Research, No. 41. Springer Verlag: Berlin Ziegler JL and Katangole-Mbidde E (1996) Kaposi’s sarcoma in childhood: an Templeton AC (1981) Kaposi’s sarcoma. Pathol Ann 16: 315–336 analysis of 100 cases from Uganda and relationship to HIV infection. Int J Templeton AC, Buxton E and Bianchi A (1972) Cancer in Kyadondo County, Cancer 65: 200–203 Uganda, 1968–1970. J Natl Cancer Inst 48: 865–874 British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png British Journal of Cancer Springer Journals

Trends in cancer incidence in Kyadondo County, Uganda, 1960–1997

Loading next page...
 
/lp/springer-journals/trends-in-cancer-incidence-in-kyadondo-county-uganda-1960-1997-bKR4B8aw0u

References (53)

Publisher
Springer Journals
Copyright
Copyright © 2000 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Epidemiology; Molecular Medicine; Oncology; Drug Resistance
ISSN
0007-0920
eISSN
1532-1827
DOI
10.1054/bjoc.1999.1071
Publisher site
See Article on Publisher Site

Abstract

British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign DOI: 10.1054/ bjoc.1999.1071, available online at http://www.idealibrary.com on Trends in cancer incidence in Kyadondo County, Uganda, 1960Ð1997 1 2 3 1 HR Wabinga , DM Parkin , F Wabwire-Mangen and S Nambooze 1 2 Department of Pathology, Makerere University, PO Box 7072, Kampala, Uganda; The International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France; Institute of Public Health, Makerere University, PO Box 7072, Kampala, Uganda Summary Incidence rates of different cancers have been calculated for the population of Kyadondo County (Kampala, Uganda) for four time periods (1960–1966; 1967–1971; 1991–1994; 1995–1997), spanning 38 years in total. The period coincides with marked social and lifestyle changes and with the emergence of the AIDS epidemic. Most cancers have increased in incidence over time, the only exceptions being cancers of the bladder and penis. Apart from these, the most common cancers in the early years were cervix, oesophagus and liver; all three have remained common, with the first two showing quite marked increases in incidence, as have cancers of the breast and prostate. These changes have been overshadowed by the dramatic effects of the AIDS epidemic, with Kaposi’s sarcoma emerging as the most common cancer in both sexes in the 1990s, and a large increase in incidence of squamous cell cancers of the conjunctiva. In the most recent period, there also seems to have been an increase in the incidence of non-Hodgkin lymphomas. So far, lung cancer remains rare. Cancer control in Uganda, as elsewhere in sub-Saharan Africa, faces a threefold challenge. With little improvement in the incidence of cancers associated with infection and poverty (liver, cervix, oesophagus), it must face the burden of AIDS-associated cancers, while coping with the emergence of cancers associated with Westernization of lifestyles (large bowel, breast and prostate). © 2000 Cancer Research Campaign Keywords: cancer registry; time trends; Africa; AIDS Kampala Cancer Registry was established in 1954 with the aim of Previous incidence rates for Kyadondo County have been obtaining information on cancer occurrence in the population of published for the periods 1954–1960 (Davies et al, 1962, 1965) Kyadondo County in which the capital city of Kampala is situated and 1968–1970 (Templeton et al, 1972). In addition, a comprehen- (Davies et al, 1965). The registry is located in the Department of sive descriptive analysis for the period 1964–1968, based on all Pathology in Makerere University Medical School. It functioned registrations (not restricted to Kyadondo residents, and without continuously both pre- and post-independence (1962), until the calculation of incidence rates), was published as a monograph coup d’état of General Amin Dada in 1971. Thereafter, full popu- (Templeton, 1973). The first results from the revitalized registry lation coverage was not possible, although a register was main- for a 28-month period in 1989–1991 were published in 1993 tained within the Pathology Department until 1980, when all (Wabinga et al, 1993). Here we review incidence rates for registration ceased. With the return of political stability, the Kyadondo County for two extended periods, 1960–1971 and registry was restarted (in 1989) and has functioned continuously 1991–1997, with a focus on the trends in risk of different cancers since. This long period of operation provides a unique opportunity both during and between these periods. to study temporal changes in cancer patterns in an African setting. Within Kyadondo, there have been quite marked social changes MATERIALS AND METHODS over the last 40 years. Progressive urbanization of the population has meant that, although the total population has increased more Cancer records than fourfold, Kampala city has grown almost tenfold, from a population of 92 000 in 1959 to 870 000 in 1995, and now repre- In 1953 the request/result forms of the Department of Pathology sents three-quarters of the county population. There has also been were redesigned specifically to permit registration of cancers. an increase in educational standards: the population of 10 years Thus they contain demographic information on the patient such as and above who have ever been to school increased among males name, age, sex, tribe and place of residence, as well as the source from 4.7% in 1959 to 71.8% in 1991 and among females from of the specimen and results of the examination. In addition to data 1.8% to 61.9%. In addition, Uganda is one of the countries most collected in this way, tumour registrars have been employed to affected by the HIV epidemic, and therefore temporal changes search for cancer cases admitted to, or treated in, the four main reveal its effect on cancer occurrence in Africans. hospitals in Kampala (and more recently in the Uganda Hospice) and, for individuals resident in Kyadondo County, to extract some- what more extensive information onto special notification forms. Between 1954 and 1980, registration was manual apart from the Received 18 June 1999 period 1964–1968 when the data were transferred to punchcards Revised 20 October 1999 (Templeton, 1973), which are no longer available, with details of Accepted 4 November 1999 all cancer cases identified being entered in a large ‘register’. Since Correspondence to: DM Parkin 1989 the registration process has been computerized, using the 1585 1586 HR Wabinga et al Table 1 Average annual population at risk, Kyadondo County, Uganda, and percentage by age group and sex 1960–1966 1967–1971 1991–1994 1995–1997 M F M F MFM F 0–14 30.8 39.0 34.3 43.5 40.6 43.5 40.6 43.5 15–44 57.9 50.9 55.8 47.6 52.4 50.0 52.4 50.0 45–55 5.9 4.8 5.1 4.2 4.0 3.2 4.0 3.2 55–64 2.9 2.8 2.5 2.2 1.6 1.7 1.6 1.7 65+ 2.3 2.4 2.2 2.4 1.3 1.7 1.3 1.7 Total 152 000 116 500 225 400 185 400 526 900 546 600 577 200 592 400 CANREG system (Cooke, 1998) which, at the stage of data entry, Statistical methods prevents the use of non-existent codes and performs checks for Incidence rates for each of the four periods were calculated for internal consistency between variables. It also permits a search for 5-year age groups. Age-standardized rates (world standard popula- potential duplicate registrations. tion) and 95% confidence limits were calculated as described in The records from the registers of 1960–1980 have also been Boyle and Parkin (1991). Differences between rates were tested by entered into the CANREG system (the registers for 1954–1959 were means of the Mantel–Haenzel test, stratifying for age in 5-year no longer available, and had probably been removed from the groups (Estève et al, 1994). country). Several duplicate entries were identified, as well as a considerable number of errors and missing information. Doubtful RESULTS records were completed, corrected or deleted, after tracing the original archives in the records of the hospitals and the pathology During the 19 years of registration considered (1960–1971; department. 1991–1997) a total of 7312 cases (3576 male and 3736 female) Tumour site and morphology were coded according to the were registered. Tables 2 and 3 show, for males and females second edition of the International Classification of Diseases for respectively, the number of cases in each of the four time periods, Oncology (Percy et al, 1990). For tabulation of results, these were by major cancer sites, together with the age-standardized inci- converted to the 10th revision of the ICD (WHO, 1992) and to the dence rates, and their standard errors. The statistical significance International Classification for Childhood Cancer (Kramárová of the change in the incidence rate from the preceding period is et al, 1997). shown. Since registration ceased to be population-based between 1972 In both sexes, the total incidence of cancer (all sites combined) and 1980, the data for this period were not used in the current increased across the four periods. The most dramatic increase analyses. With the availability of 8 full years’ registration since the between the decades of the 1960s and the 1990s is in the incidence restart of the registry in late 1989, it was clear that the case finding of Kaposi’s sarcoma (KS), especially for females, in whom KS for 1990 was inadequate and so the data for this year, included was previously a rather rare disease (sex ratio almost 20:1) but in in a previous publication (Wabinga et al, 1993) were excluded. whom the incidence rate is now around half that of men. Figure 1 Otherwise, inspection of the annual number of registrations (and shows age-specific incidence rates for KS, for 1960–1971 and proportions of registrations based on histological examinations) 1991–1997 in men and, for the latter period only, in women. In showed a regular progression with time. We therefore chose to males, as well as the increase in incidence, the shape of the curve consider four time periods: 1960–1966 (7 years); 1967–1971 has changed: in the 1960s there was a progressive increase in risk (5 years), 1991–1994 (4 years); 1995–1997 (3 years). with age, while in the 1990s there is a small peak in incidence in childhood, and a more marked one in the late 30s. In females, inci- dence in childhood is lower, and the peak incidence in adults is at ages 25–29. The mean age at diagnosis in the 1991–1997 period is Population data significantly later (P < 0.001) in men than in women both for all cases (32.0 vs 27.1), and for cases in adults aged 15 or more Population censuses were performed in 1959, 1969 and 1991. For (34.9 vs 29.8). these years, the population of Kyadondo County was available by Other than KS, the cancers showing the most marked increases sex and 5-year age group. An estimate for 1995 was provided in incidence in males are cancers of the prostate and oesophagus. by the Department of Statistics of the Ministry of Finance and There have also been significant increases in rates for cancers of Economic Planning; this was a simple extrapolation of the 1991 the large bowel and eye and, in the last 3 years, of non-Hodgkin result, assuming an annual growth in each age group of 2.61% for lymphoma. Although the changes in incidence between the indi- males and 2.30% for females. An interpolation between the census vidual time periods are not statistically significant, the trends in of 1959 and 1969 was prepared, assuming a constant rate of incidence rates between 1960–1966 and 1995–1997 are highly increase within age/sex groups, and projections for 1996 and 1997 significant (P < 0.01) for both stomach and lung cancers. were made, assuming the same growth rate as that provided for Conversely, there have been significant decreases in the incidence 1991–1995. rates of cancers of the bladder and penis, and of Hodgkin’s disease For the calculation of incidence rates in the four periods consid- and leukaemia. Liver cancer, which was the most common cancer ered, we used: the estimated population for 1963, the census of men in the 1960s, appears to have declined in incidence during population of 1969 and the estimated average populations for the 1990s. 1991–1994 and 1995–1997. These are shown in Table 1. British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1587 Table 2 Incidence of the major cancers in Kyadondo County, Uganda, in four time periods: males 1960–1966 1967–1971 1991–1994 1995–1997 No. ASR(standard No. ASR (standard No. ASR (standard No. ASR (standard error) error) error) error) Nasopharynx 3 0.3 (0.2) 9 1.8 (0.7) 11 0.7 (0.2) 26 2.3 (0.6) c c Oesophagus 8 1.7 (0.6) 25 5.1 (1.1) 83 15.8 (1.9) 68 13.0 (1.8) Stomach 14 2.7 (0.8) 20 4.7 (1.1) 31 4.7 (1.0) 37 7.6 (1.4) Colon & rectum 14 3.0 (0.8) 23 4.8 (1.1) 52 8.3 (1.3) 38 6.8 (1.4) c d d Liver 44 6.0 (1.1) 75 11.7 (1.6) 73 9.8 (1.4) 41 5.9 (1.2) Lung 5 0.8 (0.4) 10 2.1 (0.8) 25 4.1 (1.0) 19 3.2 (1.0) Melanoma 1 0.1 (0.1) 8 1.3 (0.6) 7 1.5 (0.7) 7 1.1 (0.5) Skin 26 3.7 (0.8) 27 4.3 (0.9) 18 2.5 (0.7) 20 4.1 (1.0) Kaposi’s sarcoma 28 3.2 (0.7) 29 3.7 (0.8) 670 39.3 (2.1) 513 39.3 (2.3) a c a Prostate 13 3.1 (0.9) 27 6.8 (1.4) 113 26.3 (2.6) 139 39.2 (3.7) Penis 29 5.5 (1.1) 30 6.3 (1.2) 18 2.9 (0.8) 23 4.4 (1.1) Bladder 24 5.2 (1.1) 27 5.9 (1.2) 13 2.5 (0.8) 10 2.9 (0.9) Eye 5 0.4 (0.2) 12 1.1 (0.4) 43 2.3 (0.4) 47 3.0 (0.6) Hodgkin’s disease 16 1.9 (0.6) 17 1.7 (0.5) 8 0.8 (0.4) 11 1.3 (0.6) NHL 32 3.9 (0.8) 32 3.6 (0.7) 76 3.6 (0.5) 95 7.4 (1.1) Leukaemia 22 2.2 (0.6) 24 3.2 (0.9) 13 0.7 (0.2) 16 1.1 (0.3) ALL 352 54.2 (3.3) 478 81.2 (4.3) 1456 149.1 (5.2) 1290 166.6 (6.2) ALL (except KS) 324 51.0 449 77.5 780 109.8 777 127.3 a b c d e f Significant increase since preceding period: P < 0.05; P < 0.01; P < 0.001.Significant decrease since preceding period: P < 0.05; P < 0.05; P < 0.001. Male (1960–1971) 80 Male (1991–1997) Female (1991–1997) 5-9 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ 0-4 10-14 Age group Figure 1 Age-specific incidence rates of Kaposi’s sarcoma In females, the increasing trends for cancers of the oesophagus, nomas (SCCs). In 1960–1971, 28/48 (58%) of histologically stomach, large bowel, eye and non-Hodgkin lymphomas are confirmed bladder cancer cases were SCCs, and just 10/48 (21%) similar to those in men, and there have also been increases in transitional cell/adenocarcinomas. In 1991–1997 the percentages incidence of the two major cancers of women – breast cancer and were 19% (3/16) SCC and 56% (9/16) transitional cell carcinomas. cervix cancer. Figure 2 shows age-specific incidence rates of Incidence rates of oesophageal cancer are similar in the two cervix cancer in the four different periods. The progressive sexes, which is unusual for this tumour, even in the African conti- increase of incidence with age in the earlier periods is replaced in nent. Forty-one per cent of the oesophageal cancers had informa- the 1990s by a pattern of peak incidence around menopause, tion on histological subtypes. All were carcinomas, with 87.1% followed by a plateau. Nevertheless, there has been no significant squamous, 7.1% adenocarcinomas and 5.7% unspecified subtype; alteration in the mean age at diagnosis between the four time there was no evidence of a change in the proportions over time. periods (43.8, 42.8, 43.2 and 43.7). The large increase in incidence of eye cancers is the conse- The decline in bladder cancer incidence between the 1960s and quence of increasing incidence of SCCs of conjunctiva. These 1990s has concerned almost exclusively squamous cell carci- increased from 4/17 eye cancers (23.5%) in men in 1960–1971 © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 Incidence rates per 100 000 1588 HR Wabinga et al Table 3 Incidence of the major cancers in Kyadondo County, Uganda, in four time periods: females 1960–1966 1967–1971 1991–1994 1995–1997 No. ASR (standard No. ASR (standard No. ASR (standard No. ASR (standard error) error) error) error) Nasopharynx 3 0.7 (0.4) 2 0.3 (0.2) 13 0.9 (0.3) 21 1.6 (0.4) c a Oesophagus 9 2.6 (0.9) 31 7.9 (1.5) 55 9.4 (1.3) 63 14.2 (1.9) Stomach 4 0.8 (0.5) 14 3.4 (1.0) 22 3.2 (0.7) 28 5.6 (1.1) Colon & rectum 11 2.7 (0.9) 22 6.3 (1.5) 36 5.7 (1.1) 34 6.6 (1.2) Liver 9 1.8 (0.6) 21 5.0 (1.3) 42 5.1 (1.0) 35 6.3 (1.3) Lung 3 0.6 (0.4) 6 1.4 (0.6) 7 0.7 (0.3) 18 3.2 (0.9) Melanoma 7 1.8 (0.7) 9 2.5 (0.8) 8 1.3 (0.5) 8 2.2 (0.8) Skin 16 4.0 (1.1) 15 3.1 (0.9) 12 1.4 (0.5) 8 1.0 (0.4) Kaposi’s sarcoma 1 0.1 (0.1) 2 0.2 (0.1) 360 17.9 (1.2) 335 21.8 (1.5) Breast 52 11.7 (1.8) 45 9.8 (1.6) 161 19.1 (1.8) 146 22.0 (2.1) Cervix uteri 84 17.7 (2.2) 109 22.5 (2.5) 341 39.7 (2.5) 296 44.1 (3.0) Corpus uteri 14 3.1 (0.9) 18 4.6 (1.3) 30 4.1 (0.9) 21 4.0 (0.9) Ovary 26 5.7 (1.3) 19 3.2 (0.8) 62 6.9 (1.1) 41 5.3 (1.0) Vulva/vagina 8 1.8 (0.7) 10 2.1 (0.8) 7 0.6 (0.3) 11 1.6 (0.6) c a Eye 4 0.3 (0.2) 3 0.2 (0.1) 37 1.7 (0.4) 45 3.4 (0.7) Thyroid 5 1.3 (0.6) 12 3.0 (1.0) 22 2.6 (0.7) 34 5.6 (1.1) Hodgkin’s disease 2 0.6 (0.5) 7 0.7 (0.3) 7 0.2 (0.1) 13 0.9 (0.3) NHL 14 2.2 (0.7) 15 2.2 (0.7) 48 2.1 (0.4) 82 5.7 (0.9) Leukaemia 13 2.3 (0.7) 20 2.8 (0.7) 17 1.2 (0.4) 17 1.9 (0.6) ALL 338 73.0 (4.4) 469 98.9 (5.3) 1508 146.8 (4.8) 1421 179.7 (6.0) ALL (except KS) 337 72.9 467 98.7 1148 128.9 1086 157.9 a b c d e f Significant increase since preceding period: P < 0.05; P < 0.001; P < 0.001. Significant decrease since preceding period: P < 0.05; P < 0.05; P < 0.001. Table 4 Cancer in children (age 0–14), Kyadondo County, 1960–1971 and 1991–1997 1960–1971 1991–1997 6 6 n (%) M:F Age standard rate (per 10 ) n (%) M:F Age standard rate (per 10 ) Leukaemia 25 (18.4) 1.1 18.7 27 (4.9) 0.7 8.6 Hodgkin’s disease 11 (8.1) 2.7 8.7 9 (1.6) 1.3 2.8 Burkitt’s lymphoma 13 (9.6) 1.6 9.5 109 (19.7) 1.5 34.3 Other NHL 18 (13.4) 2.6 13.1 61 (11.1) 1.3 19.1 Brain & CNS 4 (2.9) 0.3 3.0 7 (1.3) 2.5 2.3 Neuroblastoma 5 (3.7) 1.5 2.9 1 (0.2) – 0.3 Retinoblastoma 16 (11.8) 1.7 9.4 33 (6.0) 1.4 9.3 Wilm’s tumour 10 (7.4) 0.7 6.1 29 (5.3) 1.6 8.6 Osteosarcoma 5 (3.7) 1.5 4.2 6 (1.1) 4.0 1.9 Kaposi’s sarcoma 3 (2.2) 2.0 2.5 183 (33.2) 1.5 55.8 Other STS 10 (7.4) 4.0 7.6 20 (3.6) 0.8 6.0 Carcinomas 9 (6.7) 2.0 7.4 18 (3.3) 1.3 5.7 Total 136 (100) 1.6 97.8 552 (100) 1.4 169.7 Includes lymphoma (unspecified). (none of the seven eye cancers in women), to 32/45 (71% in men) BL was significantly higher (P < 0.001) in the 1990s than in the and 34/40 (85%) in women in the most recent 3-year period (there 1960s. This is unlikely to be due to changing classification or were a few cases of unknown histological type). coding, since the incidence of other and unspecified non-Hodgkin Table 4 shows incidence rates for certain cancers of childhood lymphomas was also higher in the second period (Table 4). In (ages 0–14) in two periods, 1960–1971 and 1991–1997. The most contrast, the incidence of childhood leukaemia is low, and there striking change is the increased incidence of KS, from 2.5 per was a significant (P < 0.05) fall in incidence between the two million in 1960–1971, to 55.1 per million in 1991–1997. The male periods. predominance has persisted (sex ratio 1.5:1), and KS has become the most common tumour of childhood, ahead of the lymphomas. DISCUSSION The mean age of these cases was 5.0, with no difference between the sexes, incidence was slightly higher in the age group 5–9 than In order to study time trends, it is important that the degree of at 0–4 (Figure 4). Lymphomas remain relatively frequent, with completeness of registration of incident cancer cases should be Burkitt’s lymphoma (BL) by far the most frequently diagnosed in similar throughout the period under consideration. In recent years 1991–1997; the peak age is at 5–9 (Figure 3) with a mean age of (since 1991) we believe that registration has been relatively good; 6.6, and boys predominate with a ratio of 1.4:1. The incidence of an exercise in independent case ascertainment (Parkin et al, 1994), British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1589 Table 5 Percentage of cases registered with diagnosis based on cytology Cervix cancer or histology (MV%) Period Cancer site 1960–1966 1967–1971 1991–1994 1995–1997 Oral cavity and pharynx 91 97 59 84 Oesophagus 65 68 28 41 Stomach 83 74 30 57 Colon and rectum 86 87 58 69 Liver 85 79 33 36 Lung 75 81 41 76 Kaposi’s sarcoma 76 94 78 86 Breast 82 88 56 68 Cervix 74 70 58 64 Prostate 62 67 66 80 Bladder 89 77 45 50 Lymphoma 96 93 72 80 Leukaemia 54 71 43 39 All cancers 84 84 64 73 1960 1966 1967 1971 missed. Finally, the wider availability of diagnostic modalities, 1991 1994 such as endoscopy, in the 1990s may have improved case finding 1995 1997 for certain cancers. The results should be interpreted, therefore, in the light of increases which could be due to better identification of cases diagnosed without histology. Templeton et al (1972) note that decreasing incidence rates are likely to be more meaningful 25-29 35-39 40-44 50-54 15-19 20-24 30-34 45-49 55-59 60-64 65-69 70-74 75+ than small increases in apparent incidence which might be related Age group to improvements in diagnostic facilities. In addition to these considerations, comparison of incidence Figure 2 Age-specific incidence rates of cancer of the cervix uteri rates requires that accurate population denominators are available. For Kyadondo, we had available estimates by age group and sex for three points during the long period studied, and made use of interpolations and projections for the other years. There must be Burkitt’s lymphoma Kaposi’s sarcoma some question as to the accuracy of the estimates of person-years 78.0 at risk, therefore, particularly with respect to the age distributions, 66.6 65.7 although no better data are, of course, available. 47.8 Davies et al (1964) reviewed hospital records from Mengo 41.4 Hospital (in Kyadondo County) from 1897 to 1956. They noted 28.2 that the general pattern of malignant diseases admitted to the hospital changed little over this period. The first estimation of inci- dence rates from the cancer registry of Kyadondo County was for 0—4 5—9 10—14 0—4 5—9 10—14 1954–1960 (Davies et al, 1962, 1965). The overall incidence reported for 1954–1960 was a little higher than our result for Figure 3 Age-specific incidence rates for Burkitt’s lymphoma and Kaposi’s sarcoma of childhood in 1991–1997 (both sexes) 1960–1966 (crude rates of 39.5 vs 33.1 for men and 48.4 vs 41.5 for women). It is possible that registration was rather more complete during the earlier period, when the proportion of cases comparing data collected for epidemiological studies in Kampala registered without histology was 27.5%, compared with 16.4% in with the registry database, suggested that the registry had identi- 1960–1966. On the other hand, experience with computer checks fied 90% of incident cancers. For the earlier periods, we can be on the manual register for 1960–1971 showed that several dupli- less certain. cate registrations and non-malignant cases had been included, and Some indication of completeness of registration is also provided these may also have inflated the numbers in the earlier time period. by the level of histological (or cytological) verification of regis- Nevertheless, for the individual cancer sites, the age-standardized tered cancer cases (Parkin et al, 1994). Table 5 shows that histo- rates are very similar in the two periods. logical verification was considerably more frequent in the earlier Incidence rates for the period 1967–1971 were very similar to time periods. This may be only a partial reflection of reality; those reported for 1968–1970 by Templeton et al (1972) with certainly, before the country was plunged into political, social and crude all-sites rates of 42.5 and 42.4 per 100 000 in men, and 50.2 economic chaos in the 1970s the standard of diagnostic (and thera- and 50.6 per 100 000 in women. peutic) services was higher than in the period of reconstruction in Over the 38-year period, the cancers showing the most dramatic the early 1990s. But it is also possible that the case-finding mech- increases in incidence are those related to infection with HIV, anisms particularly in the first period studied here were less active particularly KS, and SCC of the conjunctiva (Beral and Newton, than in the 1990s and that some cases without biopsies were 1998). KS has always been observed in the Ugandan population, © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 Incidence rates per 100 000 1590 HR Wabinga et al although in the 1950s and 1960s it was of the typical ‘endemic’ in 1985, has since been increasing again (FAO, 1998). On the pattern, involving the skin, particularly the legs, and affecting other hand, the importance of cereal-based diets, particularly those principally males, with the risk rising progressively with age relying upon maize, has been stressed with respect to the risk in (Taylor et al, 1971; Templeton, 1981). The enormous increase in Africa, and in particular in explaining the marked differences in incidence of KS since the earlier periods, together with the risk within quite small geographic areas (Cook, 1971; van narrowing of the sex-ratio (from 18:1 in 1960–1971 to 1.7:1 in the Rensburg, 1981). The mechanism could be via nutritional defi- 1990s) was noted in a previous report (Wabinga et al, 1993), and ciencies associated with such diets (van Rensburg et al, 1983; Ziegler and Katangole-Mbidde (1996) have drawn attention to the Jaskiewicz, 1989) or contamination of maize with mycotoxins dramatic increase in the incidence of KS in children. Uganda, in (Sydenham et al, 1990). However, maize is not the staple food in common with other central African countries, has a very high Kyadondo area of Uganda, where consumption of a variety of prevalence of infection with HIV, and the age-specific incidence of plantains is the major source of carbohydrate. KS corresponds closely to the age-specific reporting rates for The rather higher incidence rates of gastric cancer in the more AIDS, which are highest at ages 30–39 for men and 20–29 for recent periods possibly reflects better diagnosis because of the women (UNAIDS, 1998). The prevalence of HIV infection among availability of gastroscopy; however, the incidence of this cancer pregnant women had reached 30% in the urban population of remains low. This is certainly not due to a low prevalence of Kampala in 1990–1992 (UNAIDS, 1998), but since that time it has Helicobacter pylori – recognized by the IARC (1994) as an impor- been reported to be in decline (to around 15% in 1996–1997). This tant cause – since the infection has been shown to be common, at may explain the relative stability in the incidence of KS in least in the western part of Uganda (Wabinga, 1996). Tumours of 1995–1997 relative to 1991–1994 observed in this study. the large bowel are also rare – an observation made by Burkitt 30 Currently, the evidence suggests that human herpes virus 8 years ago (Burkitt, 1971), although there seems to have been (HHV8) is the aetiological agent responsible for KS (IARC, something of an increase in incidence in both sexes since the 1998). HHV8 has been identified in over 85% of KS tissue speci- 1950s and early 1960s, but little evidence of any recent change. mens in Uganda (Chang et al, 1996). Seroprevalence studies The reported rates of liver cancer in Uganda have remained low suggest a relatively high prevalence of infection by HHV8 in the in comparison with other parts of sub-Saharan Africa although of general population of Uganda – considerably higher than in the course much higher than in Western populations. The reasons for USA and Europe, which would be consistent with the elevated this are not apparent. The prevalence of chronic carriage of frequency of ‘endemic’ KS which preceded the AIDS epidemic hepatitis B in Uganda is similar to that in other countries, and afla- (Gao et al, 1996; Simpson et al, 1996). toxin contamination of foodstuffs appears to be common (Sebunya Unlike KS, the incidence of non-Hodgkin lymphomas, which and Yourtee, 1990), though less perhaps in Kampala than in other are found to be particularly common among immunosuppressed parts of the country (Alpert et al, 1971). individuals in Europe and USA, had remained relatively stable in The incidence rates of both breast and cervix cancer have Kyadondo County in the early 1990s (Wabinga et al, 1993). approximately doubled between the 1950s and 1960s, and the However, in 1995–1997 there appears to have been a significant 1990s. The reasons for these changes are not immediately clear. increase in incidence both in males and females. This may relate to They are unlikely to be related to the epidemic of AIDS for, improved survival of patients with HIV infection as other oppor- although cervix cancer is considered to be an AIDS-defining tunistic infections are controlled, permitting a more prolonged condition in the US, the evidence of a link between infection with duration of immunosuppression, and the development of more HIV and the risk of invasive cervical cancer is not consistent. The clinically-evident lymphomas. role of oncogenic human papillomaviruses (HPV) in the aetiology Squamous cell carcinomas of the conjunctiva have been recog- of cervix cancer is clearly established (IARC, 1995). HPV was nized for many years as more common in Africans than in found to be present in the great majority of cervix cancers in Europeans, but there is also a considerable (tenfold or more) Uganda, with HPV type 16 in 53% (Bosch et al, 1995). It seems increase in risk in the presence of HIV infection (Newton, 1996). quite plausible that the social disruption of the Amin dictatorship Ateenyi-Agaba (1995) has reported a large increase in the and following civil wars (1972–1986) had favoured the spread of numbers of cases presenting clinically in Kampala since the onset HPV, like other sexually transmitted diseases. As far as breast of the AIDS epidemic. cancer is concerned, it is possible that some of the increase is The increase in the incidence of oesophageal cancer between related to declines in fertility. Census data suggest that urban the 1960s and 1990s, particularly among males, is difficult to dwellers, and those with higher educational levels, have lower explain. Studies in southern Africa, as elsewhere, have pointed to than average fertility, but it is unlikely that changes in these para- the importance of tobacco smoking and, less certainly, alcohol as meters could explain a doubling of incidence rates. Screening being important in aetiology (van Rensburg et al, 1985; Segal et al, programmes for these cancers are virtually non-existent and can 1988; Vizcaino et al, 1995). The sex ratio of oesophageal cancer in hardly have influenced trends. There has been virtually no change Uganda has always been quite close to one to one, which suggests in the incidence of other female cancers – corpus uteri, ovary and that tobacco and alcohol are not of major importance in the rela- vulva/vagina. tively high, and increasing, rates, since smoking (in particular) is In men, the incidence of cancer of the prostate has increased largely confined to males. Furthermore, tobacco smoking is not remarkably, from an age-standardized rate of 3–6 per 10 in the common (average consumption per adult is 300 manufactured 1950s/60s, to 40 per 10 in the late 1990s. This is one of the cigarettes per year), there has been little change between 1970 and highest incidence rates recorded in Africa (Ferlay et al, 1998). 1990 (WHO, 1997), and the incidence of lung cancer has remained Most of this increase is in elderly men – aged 65 or over – low in both sexes. Consumption of alcoholic beverages, on the although the actual rates are not exceptionally high, compared other hand, is quite high (IARC, 1988) and after a decline through with those reported in Europe and North America. The increase in the 1970s to a low point of around 140 calories per caput per day Uganda is certainly not due to screening, although it is quite likely British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign Trends in cancer incidence in Kyadondo, Uganda 1591 Ateenyi-Agaba C (1995) Conjunctival squamous-cell carcinoma associated with that increased awareness, a greater readiness to perform prostatec- HIV infection in Kampala, Uganda. Lancet 345: 695–696 tomy for urinary symptoms in elderly men, and histological exam- Beral V and Newton R (1998) Overview of the epidemiology of immunodeficiency- ination of operative biopsies have played a role. The level of associated cancers. J Natl Cancer Inst Monogr, pp. 1–6 histological confirmation of diagnosis has actually increased over Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, Schiffman MH, Moreno V, Kurman R and Shah KV (1995) Prevalence of human time (Table 3) in contrast to virtually all other sites. papillomavirus in cervical cancer: a worldwide perspective. International In contrast to prostate cancer, the incidence of penile cancer in biological study on cervical cancer (IBSCC) Study Group. J Natl Cancer Inst men is lower in the 1990s than in the 1960s (although the rates 87: 796–802 remain high – the age-standardized rate of 3–4 in Kampala, can be Boyle P, Parkin DM (1991) Statistical methods for registries In: Cancer compared with 0.8 in the black population in the US SEER Registration: Principles and Methods, Jensen OM, Parkin DM, MacLennan R, Muir CS and Skeet R (eds), pp. 126–158. IARC Scientific Publication No. 95. registries (Parkin et al, 1997)). Penile cancer was clearly very The International Agency for Research on Cancer: Lyon frequent in the early case series from Kampala (Davies et al, Bradley DJ, Sturrock RF and Williams PN (1967) The circumstantial epidemiology 1964), and Dodge et al (1973) found it ‘the commonest tumour of Schistosoma haematobium in Lango district, Uganda. East Afr Med J 44: registered in males’ in 1964–1968. This decline in incidence is 193–204 Burkitt DP (1971) Epidemiology of cancer of the colon and rectum. Cancer 28: 3–13 probably real, since penile cancer is easily diagnosed and probably Chang Y, Ziegler J, Wabinga H, Katangole-Mbidde E, Bashoff C, Schultz T, Whitby always brought for medical attention. Penile cancer has been D, Maddelena D, Jaffe HW, Weiss RA, the Uganda Kaposi’s Sarcoma Study related to genital hygiene (Kyalwazi, 1966) and the decline in inci- Group and Moore PS (1996) Kaposi’s sarcoma-associated herpesvirus and dence may be related to improved hygiene as a consequence of Kaposi’s sarcoma in Africa. Arch Int Med 156: 202–204 urbanization and greater availability of piped water supplies. Cook P (1971) Cancer of the oesophagus in Africa. Br J Cancer 25: 853–880 Cooke A (1998) CANREG3 Manual. Internal Report No 98/03. International Agency Bladder cancer is also significantly lower in the 1990s than in for Research on Cancer: Lyon the 1960s. In Uganda, bladder cancer has been linked to the pres- Davies JNP, Wilson BA and Knowelden J (1962) Cancer incidence of the African ence of urethral stricture. Dodge (1964) found 30% of bladder population of Kyadondo (Uganda). Lancet 2: 328–330 cancer causes had such strictures, and Owor (1975) found 4% of Davies JNP, Elmes S, Hutt MSR, Mtimavalye LAR, Owor R and Shaper L (1964) Cancer in an African community, 1897–1956. Br Med J 1: 259–264 patients with strictures developed bladder cancer. Since strictures Davies JNP, Knowelden J and Wilson BA (1965) Incidence rates of cancer in are a sequel to gonococcal infection, it is possible that better treat- Kyondondo county, Uganda 1954–1960. J Natl Cancer Inst 35: 789–821 ment for STDs may have reduced their prevalence. Bladder cancer Dodge OG (1964) Tumours of the bladder and urethra associated with urinary is also a consequence of infection with Schistosoma haematobium retention in Ugandan Africans. Cancer 17: 1433–1436 (IARC, 1994). However, in Uganda Schistosoma haematobium Dodge OG, Owor R and Templeton AC (1973) Tumours of the male genitalia. In: Tumours in a Tropical Country. Recent Results in Cancer Control 41, AC has a rather restricted range and Kyadondo County is not among Templeton (ed). Springer: Berlin the endemic areas (Bradley et al, 1967). Estève J, Benhamou E and Raymond L (eds) (1994) Statistical Methods in Cancer Research. Vol IV, Descriptive Epidemiology. IARC Scientific Publication No. CONCLUSIONS 128. The International Agency for Research on Cancer: Lyon FAO (1998) FAOSTAT Statistics Database: Food balance sheets This long time series shows clearly the development of the (http://apps.fao.org/) problem of cancer in modern sub-Saharan Africa. New problems Ferlay J, Parkin DM and Pisani P (1998) GLOBOCAN: Cancer Incidence and are emerging in the shape of AIDS-related cancers (KS, NHL, Mortality Worldwide. IARC CancerBase No. 3. IARC: Lyon Gao SJ, Kingsley L, Li M, Zheng W, Parravicini C, Ziegler J, Newton R, Rinaldo conjunctival neoplasms), and cancers associated with CR, Saah A, Phair J, Detels R, Chang Y and Moore PS (1996) KSHV Westernization of lifestyles (breast and prostate cancers in partic- antibodies among Americans, Italians and Ugandans with and without Kaposi’s ular, large bowel cancers less certainly). At the same time, there sarcoma. Nat Med 2: 925–928 has been little, if any, decline in the major cancers of this region – IARC (1988) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, Vol 44, Alcohol Drinking. International Agency for Research: Lyon cervix, liver and oesophagus. So far, Uganda, as most of the IARC (1994) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, region, has been spared the epidemic of tobacco-related cancer Volume 61, Schistosomes, Liver Flukes and Helicobacter pylori. International that has been such an important feature of the cancer profile in Agency for Research: Lyon economically developed countries. This reprieve is probably only IARC (1995) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, temporary, however, and, while economic hardship may limit Volume 64, Human Papillomaviruses. International Agency for Research: Lyon IARC (1998) IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, tobacco consumption and its health consequences, it will also Vol 70, Epstein-Barr Virus and KaposiÕs Sarcoma Herpesvirus/Human restrict the capacity to mount effective programmes of cancer Herpesvirus 8. International Agency for Research: Lyon control. Jaskiewicz K (1989) Oesophageal carcinoma: cytopathology and nutritional aspects in aetiology. Anticancer Res 9: 1847–1852 Kramárová E, Stiller CA, Ferlay J, Parkin DM, Draper GJ, Michaelis J, Neglia J ACKNOWLEDGEMENTS and Quershi S (1997) The International Classification of Childhood Cancer. We acknowledge the cooperation and assistance of the medical IARC Technical Report No. 29. International Agency for Research on Cancer: nursing and records staff of Mulago, Rubaga, St Francis Lyon Kyalwazi SK (1966) Carcinoma of the penis. A review of 153 patients admitted to (Nsambya), and Mengo hospitals, and the Uganda Hospice, for Mulago Hospital, Kampala, Uganda. East Afr Med J 43: 415–425 their assistance in tracing and recording cancer patients, as well as Newton R (1996) A review of the aetiology of squamous cell carcioma of the the pathologists and haematologists for access to their laboratory conjunctiva. Br J Cancer 74: 1511–1513 records. We thank Ms Zam Kyalisiima for assistance in coding and Owor R (1975) Carcinoma of bladder and urethra in patients with urethral strictures. East Afr Med J 52: 12–18 data entry, and Mr Andy Cooke for help with the data analysis. Parkin DM, Chen VW, Ferlay J, Galceran J, Storm HH and Whelan SL (1994) Comparability and Quality Control in Cancer Registration. IARC Technical REFERENCES Report No. 19. International Agency for Research on Cancer: Lyon Parkin DM, Whelan SL, Ferlay J, Raymond L and Young J (Eds) (1997) Cancer Alpert ME, Hutt MSR, Wogan GN and Davidson CS (1971) Association between Incidence in Five Continents, Vol VII. IARC Scientific Publication No 143. aflatoxin content of food and hepatoma frequency in Uganda. Cancer 28: The International Agency for Research on Cancer: Lyon 253–260 © 2000 Cancer Research Campaign British Journal of Cancer (2000) 82(9), 1585–1592 1592 HR Wabinga et al Percy C, Van Holten V and Muir C (eds) (1990) International Classification of UNAIDS (1998) Epidemiological fact sheet on HIV/AIDS and sexually transmitted Diseases for Oncology, 2nd edn. World Health Organization; Geneva, diseases Uganda. http://www.who.int/emc-hiu/fact_sheets/ Switzerland van Rensburg SJ (1981) Epidemiologic and dietary evidence for a specific Sebunya TK and Yourtee DM (1990) Aflatoxigenic Aspergilli in foods and feeds in nutritional predisposition to esophageal cancer. J Natl Cancer Inst 67: 243–251 Uganda. J Food Quality 13: 97–197 van Rensburg SJ, Ambrose SB, Rose EF and Plessis JP (1983) Nutritional status of Segal I, Reinach SG and de Beer M (1988) Factors associated with oesophageal African populations predisposed to esophageal cancer. Nutr Cancer 4: 206–214 cancer in Soweto, South Africa. Br J Cancer 58: 681–686 van Rensburg SJ, Bradshaw ES, Bradshaw D and Rose EF (1985) Oesophageal Simpson GR, Schulz TF, Whitby D, Cook PM, Boshoff C, Rainbow L, Howard MR, cancer in Zulu men. South Africa: a case–control study. Br J Cancer 51: Gao SJ, Bohenzky RA, Simmonds P, Lee C, de Ruiter A, Hatzakis A, Tedder 399–405 RS, Weller IV, Weiss RA and Moore PS (1996) Prevalence of Kaposi’s Vizcaino AP, Parkin DM and Skinner MEG (1995) Risk factors associated with sarcoma associated herpesvirus infection measured by antibodies to oesophageal cancer in Bulawayo, Zimbabwe. Brit J Cancer 72: 769–773 recombinant capsid protein and latent immunofluorescence antigen. Lancet Wabinga HR (1996) Frequency of Helicobacter pylori in gastroscopic biopsy of 348: 1133–1138 Ugandan Africans. East Afr Med J 73: 691–693 Sydenham EW, Thiel PG, Marasas FO, Shephard GS, van Schalkwyk DJ and Koch Wabinga HR, Parkin DM, Wabwire-Mangen F and Mugerwa JW (1993) Cancer in KR (1990) Natural occurrence of some fusarium mycotoxins in corn from low Kampala, Uganda, in 1989–91: Changes in incidence in the era of AIDS. Int J and high esophageal cancer prevalence areas of the Transkei, Southern Africa. Cancer 54: 26–36 J Agric Food Chem 38: 1900–1903 WHO (1992) International Statistical Classification of Diseases and Related Health Taylor JF, Templeton AC, Vogel CL, Ziegler JL and Kyalwazi SK (1971) Kaposi’s Problems, 10th Revision. World Health Organization: Geneva, Switzerland sarcoma in Uganda: a clinico-pathological study. Int J Cancer 8: 122–135 WHO (1997) Tobacco or Health: A Global Status Report. Country Profiles by Templeton AC (ed) (1973) Tumours in a Tropical Country. Recent Results in Region. World Health Organization: Geneva, Switzerland Cancer Research, No. 41. Springer Verlag: Berlin Ziegler JL and Katangole-Mbidde E (1996) Kaposi’s sarcoma in childhood: an Templeton AC (1981) Kaposi’s sarcoma. Pathol Ann 16: 315–336 analysis of 100 cases from Uganda and relationship to HIV infection. Int J Templeton AC, Buxton E and Bianchi A (1972) Cancer in Kyadondo County, Cancer 65: 200–203 Uganda, 1968–1970. J Natl Cancer Inst 48: 865–874 British Journal of Cancer (2000) 82(9), 1585–1592 © 2000 Cancer Research Campaign

Journal

British Journal of CancerSpringer Journals

Published: Apr 4, 2000

There are no references for this article.