CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Methods: The study population was a subset of mother-infant pairs enrolled in NICHD HPTN 040 (perinatal trial with women identified with HIV during labor with no antiretroviral use in pregnancy) for whommaternal and infant urines were available. One mL aliquots of maternal urines were tested with qualitative Real-Time PCR for CMV DNA; those with positive results were tested by quantitative CMV PCR. Infant urines were similarly tested with mother-infant pair results correlated. Results: Urine specimens were available for 260 women, the majority (85.4%) from the Americas (Brazil, Argentina, U.S.) as opposed to South Africa (14.6%). Twenty-four women (9.2%) had detectable CMV viruria by qualitative PCR. Most women (91.6%) had viruria with low level detectable CMV (<200 copies/mL) while two had 236 and 53,524 CMV copies/mL urine respectively. Ten (3.8%) infants had CMV detected in their urine at the time of birth (range of 367-592,274 copies/mL). Among mothers with detectable urinary CMV at the time of delivery, 20.8% (5/24) had infants with congenital CMV. In contrast, only five of 236 infants (2.1%) born to mothers with undetectable urinary CMV had congenital CMV (p =0.0008). The two women with higher levels of CMV viruria had infants with congenital CMV. Mean CD4 counts were similar between women with CMV viruria versus those without (441.8 vs. 481.1 cells/mm 3 , p = 0.92). Women with CMV viruria had a higher mean HIV viral load in magnitude, but of no statistical significance (94,664 vs 70,906.8 copies/mL, p=0.91). These women were also more likely to transmit HIV to their infants (29.2% vs. 8.1%, p =0.005). Infant death was more frequent in children of women with CMV viruria as opposed to no CMV shedding (4.2% vs. 0.85%, p=0.25). Women with CMV viruria were 5 times more likely to transmit HIV (OR=4.68, 95%CI: 1.73-12.69) and 12 times more likely to transmit CMV (OR=12.16, 95%CI:3.23-45.73). Conclusions: Urinary CMV shedding is more frequent than expected in HIV-infected pregnant women not on antiretrovirals. Maternal CMV viruria at the time of birth is a significant risk factor for CMV and/or HIV transmission to infants and a potential marker for adverse infant outcomes.

780 TLR9 Variant Is AssociatedWith Earlier HIV, EBV, and CMV Acquisition in Infants Kristin Beima-Sofie 1 ; Grace C. John-Stewart 1 ; DaltonWamalwa 2 ; Elizabeth Maleche-Obimbo 2 ; Jennifer Slyker 1 1 Univ of Washington, Seattle, WA, USA; 2 Univ of Nairobi, Nairobi, Kenya

Background: Cytomegalovirus (CMV) and Epstein-Barr (EBV) virus are acquired by the majority of African HIV-infected infants during the first year of life. Both viruses establish persistent infection with prolonged systemic viremia. CMV is associated with an increased risk of HIV progression, morbidity, mortality and developmental impairment. EBV is an important contributor to lymphoma in the setting of HIV. We recently reported that a single-nucleotide polymorphism in Toll-like Receptor 9 (TLR9) is associated with an increased risk of HIV acquisition in Kenyan infants. Because TLR9 recognizes unmethylated CpG sequences in viral DNA, we hypothesized that this variant may also affect risk of CMV and EBV acquisition and viral replication. Methods: HIV, CMV and EBV outcomes were assessed longitudinally from birth to 1 year of age among HIV-infected and exposed infants from a Kenyan perinatal cohort. HIV infection was diagnosed using PCR for HIV DNA or the GenProbe Assay for HIV RNA. CMV and EBV DNA were measured from plasma specimens using real-time PCR, with serologic confirmation of DNA-negatives. Infants were genotyped for the TLR9 1635A/G (rs352140) polymorphism. Pearson’s Chi squared tests, Cox proportional hazards, and linear regression were performed to assess the association of the 1635A/G variant with acquisition, peak viral levels, and viral suppression. Results: CMV and TLR genotype data were available in 37 HIV-infected infants. More than 80% of CMV infections occurred before 3 months of age. At 1 month of age, 42% of infants with the 1635 AA or AG genotypes, and 11% of infants with 1635 GG genotype had acquired CMV (p=0.03). EBV and TLR genotype data were available in 104 HIV-exposed and infected infants. The probability of EBV infection at 12 months was 61% among HIV-infected and 19% among exposed infants. Infants with one or more copies of the 1635A allele showed an increased risk of EBV acquisition during the first year of life, adjusting for HIV status (HR=1.98, 95%CI=1.11, 3.54). No associations were found between TLR9 polymorphisms and CMV or EBV peak viral loads, or with time to CMV or EBV suppression (p>0.05 for all). Conclusions: The TLR9 1635A/G variant was associated with earlier acquisition of CMV and EBV, and in a prior analysis with earlier acquisition of HIV, suggesting a key role for this locus in defense against viral infection. Determining the functional phenotype of this TLR9 variant may inform novel therapeutic or vaccination strategies.

Poster Abstracts

325

CROI 2016