High frequency of acquired aplastic anemia in Tanzania

Acquired aplastic anemia (AA) is a rare disease, with an estimated frequency—in Europe and in North America—of about 2 cases per million population per year.1 There are reports of AA from North Africa,2 West Africa,3 East Africa,4 and Southern Africa,5 but no data on the frequency of this condition. In an observational study approved by the Ethics Committee of Muhimbili National Hospital (MNH), we have enrolled 119 patients with pancytopenia over a period of 1 year (from September 1, 2016 to August 31, 2017). Systematic investigations included a bone marrow aspirate in all cases. In the majority of cases, pancytopenia could be attributed to acute leukemia, hypersplenism, or severe nutritional deficiency; other causes included lymphoma, myelo-proliferative disorders, myelodysplastic syndromes, and HIV infection. In 37 cases, the diagnosis was AA, but 3 were excluded from further analysis because they might have been familial. As for the remaining 34 cases, in 30 the diagnosis of AA was confirmed by bone marrow trephine biopsy, showing severely hypocellular marrow (cellularity estimated to be less than 20% of normal for age); in 4 cases the hematologic findings (see Table 1), including the bone marrow aspirate, were deemed sufficient for a diagnosis of AA. Each patient was then classified as having nonsevere AA, severe AA, or very severe AA (see Table 1), according to standard criteria.6 MNH is a tertiary health facility in Tanzania, and therefore, there is a high rate of referral from different parts of the country, as reflected by the geographic distribution of the place of residence of our 34 patients (see Table 1). Nineteen patients were from the Dar-es-Salaam region (DSR: population about 5 million). Therefore, a crude minimum figure for acquired AA in DSR is 19/5 million, or 3.8 cases per million per year (95% confidence limits 2.3 to 5.9). Several considerations indicate that this figure is an under-estimate. (a) Not all patients with AA report to MNH, since some may not have been diagnosed. (b) Since mortality from untreated AA is high,7 some patients may have died before referral. (c) We made the unexpected observation that at least 76% of patients with AA had higher education, as against an approximate estimate of 49% (54% in men and 44% in women) in the DSR population (https://dhsprogram.com/pubs/pdf/fr321/fr321.pdf). Since it seems a priori unlikely that higher education increases the risk of AA, perhaps the educational status played instead a role in making it more likely for a patient to come or being referred to MNH. Taking (c) into account, our minimum estimate becomes 3.8 × 76/49 = 5.9 cases per million per year (confidence limits 4.0 to 8.3). Unraveling possible causes of AA was not within the scope of this study. However, from the individual responses to a conventional questionnaire covering medical history, possible exposure to myelo-toxic agents and exposure to drugs, no significant findings have emerged to date. Unfortunately, the treatment of AA in our hospital has been sub-optimal until now, compared to international standards.8 In the course of our study, we have investigated, in addition to the 34 reported here, 18 patients who had been diagnosed previously. In the total of 52 patients, the overall mortality was 48%. Only one patient had allogeneic bone marrow transplantation (abroad) and was cured. Four patients received intensive immunosuppressive treatment (abroad) with anti-thymocyte globulin: of these one died, and three are in long-standing remission. All other patients received either supportive treatment plus cyclosporine A, or supportive treatment alone. In an unpublished series of 100 AA patients from South Africa (downloadable as Aplastic anemia at Chris Hani Baragwanath academic hospital), clinical outcomes were better, with a 5-year survival of 69%. Given the known association of AA with PNH,9 and given that small populations of glucosyl-phosphatidyl-inositol (GPI)-negative blood cells are a common finding in AA,10 we used flow cytometry in order to detect such cells (see Figure 1A). In 11 out of 26 patients tested (42%), we did indeed find a population of GPI(−) red cells significantly above background (see Figure 1B). The percentage of GPI(−) cells is known to be much higher among granulocytes (which we were unable to test) than among red cells:11 therefore, 42% is again a minimum estimate of all AA patients who have “PNH clones.” Interestingly, the proportion of patients with severe (or very-severe) AA was significantly lower, and the time since diagnosis was significantly longer in patients who did have “PNH clones” (see Figure 1, inset): in keeping with the notion that PNH clones may ameliorate the course of AA and may expand with time.12 The incidence of acquired AA, compared to Europe and North America, is estimated to be about twice as high in South-East Asia.13 We now report that in Tanzania the incidence is even higher, of the order of six cases per million per year. We do not believe this figure to be a vagary of the time during which the study was carried out, because in the course of a previous one-year period (2010) one of us has seen at MNH 35 cases of AA (P. Magesa, 2011, unpublished data). We do not know at the moment the reason for this relatively high incidence. Given the overwhelming evidence that acquired AA is a cell-mediated auto-immune disease,9, 14 it would appear that the risk of a derangement of the immune system leading to an immune attack against hematopoietic stem cells is greater in Tanzania than elsewhere. This increased risk might be related, in turn, to a different pattern of exposure to pathogenic organisms, including the malarial Plasmodia, from early in life. Whatever the reason for this increased incidence, in nearly one-half of our patients we found GPI(−) red cells at levels that indicate clonal advantage, suggesting that the target of the auto-immune attack is the GPI molecule.9 Thus, we have evidence that one and the same pathogenetic mechanism of acquired AA operates not only in Europe9 and in Japan15 but also in Africa. We are very grateful to all our medical colleagues as well as nursing and supporting staff in the Hematology Department and in other departments of MNH and of MUHAS for their role in patient care; and we thank colleagues in other hospitals for referring patients. We are immensely grateful to Dr Alex Magesa, Director of the Clinical Laboratories of MNH, for supervising laboratory work; to Sean Rooney, Hasan Kimweko, and Zadock Mbizi for their invaluable help with flow cytometry studies; to Mama Christina Tairo and her staff for laboratory tests. We thank Dr LoAnn Paterson and Dr Leah Mnango for reviewing some of the bone marrows. We thank Dr Luca Boni for help with statistics; Dr Rosario Notaro for critical comments on the manuscript, and the Istituto Toscano Tumori of Florence for donating monoclonal antibodies. We express deep appreciation to Julie Makani, Head of the Department of Hematology, for her constant and constructive encouragement. Finally, we thank our patients and we wish we can do more for them in the future. Nothing to report.