Epidemiology of Burkitt Lymphoma in East Africa Children or Minors (EMBLEM)

Burkitt lymphoma (BL) is an aggressive monoclonal B-cell malignancy that is rare (sporadic) worldwide, but is 100-fold more common (endemic) in equatorial Africa, particularly among children. Epstein-Barr virus (EBV) and malaria are epidemiologically linked to endemic BL in epidemiologic studies, but questions remain about role of EBV variants and the evidence for association with malaria is weak. EBV is ubiquitous, yet only few children develop BL, possibly because only a few EBV variants are pathogenitically relevant. The association of BL with malaria is based on ecologic and non-comparative clinical studies. Two case-control studies have reported significant association of high anti-malarial antibodies with BL (OR=5_ among children in Uganda and in Malawi, but selection bias (cases and controls came from dissimilar geographical areas) and reverse causality bias were limitations. Three studies were conducted in the 1960s and 70s to test association of carriage of malaria-resistance gene with BL, two of which reported a significant or marginal inverse association. These pioneering studies were small (240 cases all together) and looked at one polymorphism in one gene (sickle cell gene). Improvements in technologies to characterize genetic variation allow the EBV and malaria hypotheses to be examined with greater power by looking at genetic variation across multiple genes.

The Epidemiology of Burkitt lymphoma in East African children and minors (EMBLEM) is a case-control study of 1500 BL cases and 3000 age-, sex- and residence-frequency matched controls we are proposing to conduct in East Africa. The study will enroll cases at four hospitals in four regions in East Africa, where malaria transmission is holoendemic and year round. The controls will be enrolled from general population attendees at Health Center II (HC-II) units where the cases originated. The primary study objectives are: 1) to test the hypothesis that genetic resistance to malaria is associated with a lower risk of BL, and 2) to use genome-wide association methods to discover genetic variation that may be associated with decreased or increased risk of BL. Because genetic variation conveys no information on actual exposure to malaria or EBV, in secondary analyses, we will use empiric epidemiological questionnaire and laboratory methods: a) to measure exposure to malaria and its association with BL, and b) to measure EBV variants and their association with BL. To examine issues related to bias and to obtain data to correct for deviations, we will also enroll 2250 population controls from 5% of the villages to obtain population distribution of key exposures variables. This data will be used to reweight the distribution in HC-II controls back to the general population.

• INCLUSION CRITERIA:

For case subjects:

• Newly diagnosed child with BL. New newly diagnosed means not more than 1 month since diagnosis, to minimize bias from mortality after diagnosis.
• Not initiated BL specific treatment.
• Age 0 through 15 years at diagnosis.
• Residing in a pre-defined geographic area for at least 4 months prior to onset of BL-related symptoms. The catchment geographic area will be defined in the Study Manual as districts for each region.
• Diagnosis based on local histology or cytology report.

For control subjects:

• Age 0-15 years.
• Residing in a defined geographic area for at least 4 months.

Exclusion criteria

For case subjects:

• Not residing within the pre-defined geographic area for at least 4 months before onset of BL-related symptoms.
• Clinically unstable condition; they will be stabilized first.
• Initiated BL treatment.
• Wrong diagnosis.
• Refusal or are inability to consent.

For control subjects:

• Mild clinical malaria (fever 37.5 degrees Celcius and a thick blood smear positive for malaria).
• Any severe illness requiring immediate admission to hospital, e.g. acute respiratory infection, diarrhea with dehydration, snake bites or fractures.
• Any cancer.
• Not a usual resident of an eligible geographic area.
• Non-consent.