Elevated serum levels of interleukin-6 in endemic Burkitt lymphoma in Ghana

To the Editor

Endemic Burkitt lymphoma (eBL) is an aggressive non-Hodgkin lymphoma with a high incidence among children in equatorial Africa [1]. Infection with Epstein-Barr virus (EBV) and Plasmodium falciparum (Pf) have been implicated as etiological factors [2]. A cohort study conducted in the West Nile district in Uganda demonstrated elevated risk of eBL in children infected with EBV during infancy [3]. Several case-control studies have confirmed that children with high exposure to malaria, based on questionnaire and antibody measurements, have an elevated risk of eBL [4,5]. However, how these infections cause eBL is not well understood. Pf infection is thought to influence eBL risk by impairing EBV-specific T cell immunity, which leads to EBV reactivation and an increase in EBV load [6]. Moreover, EBV secretes functional cytokine homologs, for example BCRF-1, an IL-10-related peptide, which could be involved in eBL pathogenesis [7]. Interestingly, patients with acute malaria also often have elevated serum levels of cytokines, including tumor necrosis factor (TNF) α, interferon (IFN) γ, interleukin (IL)-6, and IL-10 [8]. Although there is some evidence that IL-6 serum levels in serum may be elevated before the development of AIDS-associated BL [9], the relationship between eBL and cytokines has not been evaluated before.

To obtain preliminary data, we measured levels of four cytokines, TNF-α, IFN-γ, IL-6, and IL-10, in 65 children with eBL (0<16 years) enrolled at the Korle-Bu Teaching Hospital, Accra, Ghana, and 41 apparently healthy children of a similar age enrolled contemporaneously from homes near the cases from 1965 to 1994. Cytokines were measured in serum by using commercial immunoassays kits according to the manufacturer’s instructions (R&D Systems Inc., Minneapolis, MN, USA) and compared using the Student’s t-test and chi-squared test. Odds ratios (ORs) and 95% confidence intervals (CI) of association were estimated using unconditional logistic regression with adjustment for sex, age, calendar year of enrollment, and malaria antibodies. Two-sided p-values <0.05 were considered statistically significant. Ethical approval for the study was obtained from the Office of Human Subject Research at the National Institutes of Health (OHSR #5430) and the George Washington University Institutional Review Board (Protocol #111040).

The cases and controls were similar with respect to gender, age group, and enrollment year (Data not shown). The levels of the different cytokines exhibited generally small but statistically significant correlations (0.25 for IL-6 and IL-10, 0.26 for IL-6 and TNFα, and 0.46 for TNF-α and IFNγ). Positivity of cytokine levels among the controls was highest for IL-6 (71%), intermediate for IFNγ and TNF-α (42% and 34%, respectively), and lowest for IL-10 (7%). The mean IL-6 level in eBL cases was 1.1 log higher than among controls (p = 0.001, Figure 1), but the levels for IFNγ, TNF-α, and IL-10 were similar. When the cytokine levels were categorized as positive or negative, IL-6 positivity was positively associated with eBL [OR= 3.5 (95% CI = 1.1–11.1), p = 0.03] after adjusting for sex, age, calendar-year period, and malaria antibodies. TNF-α positivity was positively associated with eBL (OR= 2.0, 95% CI = 0.7–5.5), but the result was not statistically significant. Positivity for IFNγ (p=0.60) and IL-10 (p = 0.70) was unrelated to eBL. In analyses for a dose response with IL-6 tertiles, the ORs for association with eBL rose from 1.7 (95% CI = 0.51–5.7) in children in the second tertile to 6.0 (95% CI = 1.9–18) among children in the third tertile as compared with children in the first tertile (P_trend = 0.001; Figure 2). No trend was observed in the risk of eBL and tertiles of TNF-α.

Figure 1.

Dot plot of mean cytokine levels for Burkitt lymphoma cases and controls for interleukin-6 (panel A), interleukin-10 (Panel B), tumor necrosis factor-α (Panel C), and interferonγ (Panel D). Each dot represents a single subject; the horizontal bar in the middle represents the mean value of the measurements

Figure 2.

Forest plot showing adjusted odds ratios (ORs) and 95% confidence intervals (CI) of association between endemic Burkitt lymphoma and tertile categories of interleukin-6 (first panel) and tumor necrosis factor (second panel). Associations adjusted for age group, sex, enrollment year, and anti-malaria antibodies

We present the first evaluation of the potential associations between eBL and serum cytokines measured in an individual-level study. We observed a significant association between IL-6 and eBL, which was associated with a significant statistical trend in analyses adjusting for age, sex, and malaria antibodies as confounders. Interestingly, eBL was not associated with TNFα or IL-10 levels in serum, although both cytokines are increased during acute malaria infection [10], suggesting that the IL-6 associations with eBL might be specific. IL-10 is jointly elevated with IL-6 in acute malaria infection [7], but our null findings with respect to IL-10 raise the question whether eBL occurs in the context of a blunted IL-10 response. Our study design does not allow us to determine whether the elevated IL-6 levels precede or are a consequence of eBL.

Assuming our results are not due to chance, IL-6 could contribute to lymphomagenesis by thwarting the apoptosis feedback loops in B cells that are triggered in B cells that have developed c-MYC chromosomal translocations to prevent the propagation of cells with chromosomal instability [11]. IL-6 is a pleiotropic cytokine that may exert local effects via the fixed IL-6 receptor (IL-6Rα) on initiated B cells (classical signaling pathway) or systemic effects by engaging its soluble receptor (sIL-6Rα) on activated B cells (trans-signaling pathway) [12]. Our results also could be due to secretion of IL-6 by eBL tumor cells [11]. If so, then levels of IL-6 might be correlated with disease burden. Irrespective of the cellular origin of IL-6 in eBL patients, our findings raise an important question about IL-6 as a risk or tumor biomarker in eBL and may justify conducting hypothesis-driven studies to investigate the association between serum IL-6 and eBL by using pre-eBL sera.

Our findings should be interpreted with caution because they are from a small study that used a case-control design. Although we adjusted for malaria antibody results, we lacked data on current clinical malaria status of the participants, which could explain the cytokine patterns. Our findings, although limited, were robust in adjusted analyses and are consistent with a previous study that reported elevated IL-6 serum levels before the development of AIDS-associated BL [9]. Interestingly, a recent study reported a similar molecular signature of eBL and AIDS-associated BL [13], in which elevated IL-6 levels have been found in pre-BL sera.

To summarize, we found significantly elevated levels of circulating IL-6 in the serum of children with eBL patients compared to children without eBL of a similar age living nearby in a Pf endemic area. Future studies are needed to investigate the association between cytokine levels in serum, especially IL-6, and eBL.