INTRODUCTION: Burkitt lymphoma (BL) is the most common pediatric cancer in sub-Saharan Africa (SSA), and also occurs frequently in adolescents and adults, often associated with human immunodeficiency virus (HIV). BL is highly curable in high-income countries, but outcomes are worse for children in SSA due to advanced disease and inability to achieve cytotoxic intensity comparable to high-income countries. Moreover, unlike pediatric BL, there are few high-quality prospective studies describing non-pediatric BL in SSA.

METHODS: We prospectively enrolled patients with newly diagnosed BL from 2013 to 2018 in Malawi, a low-income country in SSA, and restricted analyses to patients aged ≥15 years. Diagnoses were confirmed using local performed immunohistochemistry, real-time telepathology, and subsequent secondary review in the United States. Participants received standardized treatment according to local pediatric or adult standards of care, with concurrent antiretroviral therapy (ART) if HIV+.

RESULTS: 36 patients ≥15 years with BL were enrolled, with median age 21 years (range 15-61), and 44% being HIV+. Five (14%) participants were over 40 years of age. At diagnosis, 23 (64%) participants presented with abdominal disease, 5 (14%) had CNS disease, and 5/23 (22%) had bone marrow involvement. 25 (69%) participants had B symptoms, 27 (75%) had stage 3/4 disease, and 19 (56%) had ECOG performance status >1. Among 16 HIV+ participants, 9 (56%) were aware of their HIV diagnosis for a median of 1.4 years (range 0.4-9.4) prior to BL diagnosis, and 8 (50%) were on ART for a median of 1.1 years (range 0.4-2.6). Among all HIV+ BL patients, median CD4 count was 134 (range 29-2235), and 8 (50%) had HIV VL <400 copies/mL. Four patients (11%) died before chemotherapy initiation. First-line chemotherapy consisted of bolus anthracycline-based chemotherapy (CHOP) in 25 (78%), infusional anthracycline-based chemotherapy (EPOCH) in four (12%), and high-dose methotrexate-based chemotherapy (COPADM) in two (6%). Only one participant received rituximab with CHOP as part of a clinical trial, since rituximab is not otherwise routinely available in Malawi. Participants received a median of 4 cycles of first-line chemotherapy (range 1-8). 14/29 (48%) evaluable participants achieved a complete response, 7 (24%) had a partial response and 8 (28%) were refractory. Median overall survival (OS) was 7 months and one-year OS was 36% (95% CI 20-52%). In a multivariate Cox proportional hazards model, death was associated with stage 3/4 disease (HR 6.3 [95% CI 1.4-27.2]; p=0.01) and ECOG performance status >1 (HR 7.4 [95% CI 1.5-35.1]; p=0.01) with a trend toward worse survival in HIV+ participants (HR 5.9 [95% CI 0.9-39.4]; p=0.07). There were no clear differences in outcome observed based on choice of first-line chemotherapy. Of 23 deaths during the follow-up period, six (26%) were classified as treatment-related.

CONCLUSIONS: This is one of the best characterized prospective cohorts of non-pediatric BL in SSA, which occurred primarily among adolescents and young adults. Outcomes were worse than pediatric cohorts from the region, with most deaths due to progressive BL. This study highlights the need to develop effective treatments for non-pediatric BL with and without HIV in resource-limited settings, where high-intensity strategies from high-income countries may have limited applicability.


Fedoriw:ALEXION PHARMACEUTICALS: Consultancy, Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.