Maternal HIV Infection as a Risk Factor for Primary Epstein-Barr Virus Infection in Kenyan Infants

Abstract

Human immunodeficiency virus (HIV) infection is known to be associated with EBV shedding in saliva suggesting an increased risk of EBV transmission to infants born to mothers with HIV at an earlier age. In this study we investigated (i) whether maternal HIV status was a risk factor for EBV in blood at delivery or for shedding in saliva and breast milk of 6- and 10-weeks post-partum mothers, (ii) if there was a difference in EBV strains shed between HIV+ and HIV- mothers, and (iii) if maternal HIV status was a determinant of EBV viral load in their infants. Samples were collected as part of a prospective cohort study that followed HIV-positive (HIV+) and HIV-negative (HIV-) pregnant women in Western Kenya through delivery and post-partum period. EBV viral load in blood was found to be significantly higher in mothers with HIV (p-value = 0.04). Additionally, a statistically significant difference was observed between EBV viral load in saliva samples and HIV status where HIV+ mothers had a higher EBV viral load in saliva at 6-weeks post-partum compared to HIV- mothers (p-value < 0.01). The difference in EBV shedding in breast milk was not found to be statistically significant. Furthermore, no difference in frequency of EBV strain was attributable to HIV- or HIV+ mothers. Interestingly, we found that infants born to HIV+ mothers had a higher EBV viral load at the time of their first EBV detection in blood than infants born to HIV- mothers and this was independent of age at detection. Overall, our study suggests that HIV infected mothers shed more virus in saliva than HIV-negative mothers and infants born to HIV+ mothers were at risk for loss of control of primary EBV infection as evidenced by higher EBV viral load following primary infection.

Keywords: Burkitt lymphoma (BL); EBV transmission; Epstein - Barr virus; HIV - human immunodeficiency virus; HIV- unexposed uninfected (HUU) infants; HIV-exposed uninfected (HEU) infants; Kenya.

Figure 1

Diagram of study procedures. Flow chart summarizing the source of maternal samples analyzed. Wks, weeks; EBV, Epstein-Barr Virus; HIV, Human immunodeficiency virus.

Figure 2

EBV Viral load in HIV negative and positive mothers. (A) Maternal venous blood cell pellet at time of delivery. Analysis made on a zero-inflated model. DNA from venous blood from HIV- mothers (n=141) and HIV+ mothers (n=61) was tested for EBV viral load, only those who had detectable EBV are shown. HIV-mothers had a mean log EBV copies per μg of DNA of 3.49 and HIV+ mothers had a mean log EBV copies per μg of DNA of 3.73 (p-value = 0.04) (B) DNA from saliva from mother at 6-weeks post-partum that were HIV- (n=97) and HIV+ (n=52) (p-value < 0.01), as well as mothers at 10-weeks post-partum that were HIV- (n=103) and HIV+ (n=56) were tested for EBV shedding, only those that are positive shedders are shown. Comparison across time points account for the paired nature of the data and were analyzed with Wilcoxon signed rank exact test; differences among HIV+ (p-value < 0.01) and HIV- (p-value < 0.01) were observed. (C) Breast Milk DNA from mothers at 6- weeks post-partum both HIV- (n=25) and HIV+ (n=12) and breast milk of mothers 10-weeks post-partum HIV- (n=111) and HIV+ (n=34) were analyzed for shedding of EBV. Only mothers in which shedding was detected are shown. *P < 0.05, **P < 0.005, ***P < 0.0005, ****P < 0.0001, ns, not significant.

Figure 3

EBV viral load by EBV strain for HIV negative (black) and HIV positive (red) mothers. (A) Maternal venous blood cell pellet at time of delivery (n=108), 21 samples were excluded as they were not typable. We show 61 samples were EBV- 1, 44 HIV- and 17 HIV+ (3.62 EBV copies per μg of DNA), 32 were EBV- 2, 25 were HIV- and 7 HIV+ (3.50 EBV copies per μg of DNA) and 15 were coinfected, 7 HIV- and 8 HIV+ (3.77 EBV copies per μg of DNA). No significant difference was found. (B) Saliva 6-weeks post-partum (n=81), 4 samples were excluded as they were not typeable. EBV-1 (n=43, 37 HIV- and HIV+) had a mean viral load of 5.54 EBV copies per mL of saliva, EBV-2 (n=18, 12 HIV- and 6 HIV+) had a mean viral load of 5.90 EBV copies per mL of saliva, and coinfected (n=20, 13 HIV- and 7 HIV+) had a mean viral load of 6.26 EBV copies per mL of saliva. The difference was not found to be significantly different. (C) Saliva 10-weeks post-partum (n = 120), 5 samples were excluded as they were not typeable. EBV-1 (n=60, 38 HIV- and 22 HIV+) had a mean viral load of 4.78 EBV copies per mL of saliva, EBV-2 (n=23, 13 HIV- and HIV+) had a mean viral load of 5.03 EBV copies per mL of saliva, and coinfected (n=37, 21 HIV- and 16 HIV+) had a mean viral load of 5.30 EBV copies per mL of saliva. No statistical difference was found. (D) Breast Milk 10-weeks post-partum (n=91), 35 samples were excluded as they were not typeable. EBV-1 (n=52, 43 HIV- and 9 HIV+) had a mean viral load of 5.60 EBV copies per mL of breast milk, EBV-2 (n=22, 15 HIV- and 7 HIV+) had a mean viral load of 6.00 EBV copies per mL of breast milk, and coinfected (n=17, 14 HIV- and 3 HIV+) had a mean viral load of 5.46 EBV copies per mL of breast milk. No statistical difference was found.

Figure 4

Children EBV viral load in the first positive venous blood sample. Children are categorized on maternal HIV status as HUUs (Healthy Unexposed Uninfected Children) or HEUs (Healthy Exposed Uninfected Children). HUU children have a mean of 4.21 log EBV copies per μg of DNA and 4.63 log EBV copies per μg of DNA for HEU children (p-value = 0.01). *P < 0.05.