Age of Infection and Risk of Virally Associated Cancers: New Clues to an Old Puzzle

(See the article by Plantinga et al, on pages 934–43 and see the article by Piriou et al, on pages 906–13.)

Endemic Burkitt lymphoma is a puzzling disease caused by Epstein–Barr virus (EBV) that occurs in children in sub-Saharan Africa, where it is the most common childhood malignancy. One aspect of the puzzle is that EBV is a ubiquitous infection in humans that is harmless to the vast majority of infected individuals. Other aspects are the young age of onset and the restricted geographical distribution of the disease. The geographical correlation of the incidence of endemic Burkitt lymphoma with the prevalence of malaria is suggestive that malaria is an important cofactor for endemic Burkitt lymphoma [1, 2]. Until recently, however, direct evidence for an association has been lacking. Such evidence has recently begun to accumulate [3, 4]. Evidence for the mechanisms of interactions between EBV and malaria that result in endemic Burkitt lymphoma are, however, very much needed.

In the 1970s, de-Thé [5] proposed that early age of infection of infants with EBV may play a role in the etiology of endemic Burkitt lymphoma. To address this hypothesis, as well as whether malaria may have a role in early infection by EBV, in this issue of The Journal of Infectious Diseases Piriou et al [6] studied the age of EBV infection in 2 cohorts of infants in Kenya, in areas that had year-round vs sporadic malaria exposure. An impressive aspect of this study is the adaptation of established methodologies to use with very small amounts of blood appropriate to collect in a study of such young children (<3 years). The comprehensive analysis of markers of EBV and malaria infection from such small samples enabled a clear picture of the infection dynamics in these cohorts to emerge. The findings are striking: Infants in an area where malaria exposure is high and year-round (Kisumu) became infected at significantly younger ages than their counterparts in an area with lower and intermittent malaria exposure (Nandi). In addition, a higher proportion of infants had evidence of EBV infection before 6 months of age in the high malaria exposure cohort. The detection of EBV infection before 6 months of age is in itself remarkable and in contrast with earlier reports [7]. The authors comment that the earlier studies were conducted in urban areas and that malaria exposures were not documented. It is also probable that the newer techniques employed in the current study and the use of both molecular and serological approaches improved sensitivity to detect EBV infection.

The current study does not, of course, provide direct evidence that early EBV infection is a risk factor for endemic Burkitt lymphoma, but such a risk is suggested by the correlation between endemic Burkitt lymphoma and malaria prevalence and the observations reported here. The authors also provide evidence that infants infected with EBV before 6 months of age, especially those living in a high malaria area, had more frequently detected and higher EBV loads. Children with endemic Burkitt lymphoma have higher EBV loads than healthy controls. The authors suggest a plausible scenario that early infection with EBV facilitated by high malaria exposure results in poor immune control of the infection.

It is also not apparent from the current study why children in Kisumu are infected by EBV at such an early age. This does not appear to be a defect in protection by maternal antibodies as detection of maternal antibodies was similar in both cohorts, and decline of maternal antibodies correlated with age of primary infection. The answer may be related to the higher EBV loads observed in the infants in Kisumu. Presumably, mothers and older siblings will also have elevated EBV viral loads. Previous work from this group has demonstrated more frequent EBV reactivation in children in Kisumu vs Nandi, and consequently, infants in Kisumu may be exposed more frequently and to higher levels of EBV than those in Nandi. Studies of EBV in saliva in comparable cohorts would be informative in this respect.

The study may have implications for cancers related to other infections. Risk factors for Kaposi sarcoma (KS) in subjects uninfected with human immunodeficiency virus are poorly understood, especially in sub-Saharan Africa where KS can occur in children, similar to endemic Burkitt lymphoma. KS is caused by the gammaherpesvirus Kaposi sarcoma–associated herpesvirus (KSHV), which is related to EBV. Malaria has recently been reported to be a risk factor for KSHV infection in Uganda [8]. Such observations give rise to speculations that age of infection may also play a role in endemic KS risk. Much work will need to be done to provide evidence to substantiate such speculations.

There are some important implications of the study relating to prevention. The observation that early age of infection with EBV predicts elevated viral load has implications for other EBV-related diseases and suggests that EBV vaccination could potentially prevent disease. These data also suggest that a reduction in malaria transmission, in addition to reducing the appalling toll of morbidity and mortality caused by malaria, could impact the incidence of endemic Burkitt lymphoma. Interventions such as use of bed nets and other mosquito control devices are increasingly proving effective in lowering malaria transmission [9]. With the recent hopeful reports on a putative malaria vaccine, we can hope for sustained reductions in malaria transmission rates, even in holoendemic regions [10]. A reduction in the incidence of endemic Burkitt lymphoma, the most common childhood malignancy in equatorial Africa, would be an unforeseen but most welcome additional outcome.