CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Results: Comparison of CMV shedders and non shedders (table 1) of similar age, CD4 count, plasma HIV RNA, CMV IgG titers, and estimated duration of HIV infection showed that individuals shedding seminal CMV had significantly higher HIV RNA levels in semen (p<0.001). The presence of seminal CMV DNA was associated with a decreased CMV-specific response (IFN-γ production) in CD4 + T cells (p=0.088) in peripheral blood but no difference in CD8 + T-cell IFN-γ production. Interestingly, asymptomatic CMV shedding was also associated with a significantly poorer HIV-specific CD8+ T-cell response (p=0.013). CMV shedders also had significantly higher PD-1 expression on CD8 + T cells (p=0.011) but no difference in PD-1 expression on CD4 + T cells. Conclusions: A decreased CMV-specific CD4 + T-cell response during primary HIV infection is associated with increased frequency of seminal CMV shedding, which in turn may drive CD8 + T-cell exhaustion and lead to an impaired HIV-specific immune response. These observations could explain the connections between CMV replication and higher levels of seminal HIV shedding, as well as increased cell-associated HIV RNA and proviral HIV DNA levels in blood.

Table 1: Patient Characteristics and PBMC Profiles

p-value 1

Total (n=28)

CMV Non-Shedders (n=13)

CMV Shedders (n=15) 34.93 [30.44 – 39.42] 618.2 [428.4 – 808]

Age (years), mean [CI]

38.54 [33.50 – 43.58]

0.30 0.88

CD4 Count (cells/µL), mean [CI] Blood HIV RNA level (copies/mL) CMV IgG Titer (IU/mL), mean [CI] Seminal CMV level (copies/mL), mean [CI] Seminal HIV level (copies/mL), mean [CI] CMV-specific (IFNγ) CD4 + T cells 2 HIV-specific (IFNγ) CD4 + T cells 2 CMV-specific (IFNγ) CD8 + T cells 2 HIV-specific (IFNγ) CD8 + T cells 2 EDI (days), mean [CI]

593 [423 – 763]

2.32x10 6 [-2.53x10 6 – 7.67 x10 6 ] 3.43 x10 5 [-1.95 x10 5 – 8.81 x10 5 ] 0.61

23.81 [17.82 – 29.79]

24.51 [16.16 – 32.86]

0.72

148.9 [53.42 – 244.4]

205.9 [-3.34 – 415.2]

0.85

4.05x10 6 [-1.19 x10 6 – 9.29 x10 6 ]

<0.0001

0

0.0009

168.4 [-10.58– 347.3]

57,591 [-51,264 – 166,445]

0.088 0.300

1.007 [0.28 – 1.73] 0.084 [-0.014 – 0.18] 0.15 [-0.011 – 0.31] 0.14 [0.018 – 0.27] 0.29 [0.18 – 0.40] 0.21 [-0.12 – 0.54]

0.47 [0.12 – 0.81]

0.02 [-0.0077 – 0.050]

0.33 [0.032 – 0.63]

0.36

0.013

0.010 [-0.0096 – 0.030]

%PD-1 on CD4 + T cells 3 %PD-1 on CD8 + T cells 3

0.41 [0.25 – 0.58] 0.19 [0.12 – 0.26]

0.25

0.011

Confidence Interval (CI); International units (IU); Estimated Date of Infection (EDI), Interferon gamma (IFNγ), Programmed Death Receptor (PD-1); 1 p-value determined by the Mann-Whitney U test; 2 IFNγ production normalized to total CD4 + or CD8 + T cells and background subtracted; 3 PD-1 expression normalized to total CD4 + or CD8 + T cells

731

High Anti-CMV IgG Levels Predict HIV Disease Progression Eshan U. Patel 1 ; Sara Gianella 2 ; Kevin Newell 3 ; Aaron A.Tobian 4 ; Allison R. Kirkpatrick 5 ; Mary K. Grabowski 6 ; David Serwadda 7 ;Thomas C. Quinn 4 ; Andrew D. Redd 8 ; Steven J. Reynolds 1 1 NIH, Bethesda, MD, USA; 2 Univ of California San Diego, San Diego, CA, USA; 3 Rsr Data and Communication Technologies, Inc, Garrett Park, MD, USA; 4 Johns Hopkins Univ Sch of Med, Baltimore, MD, USA; 5 NIAID, NIH, Baltimore, MD, USA; 6 Johns Hopkins Univ, Baltimore, MD, USA; 7 Makerere Univ Schoool of PH, Kampala, Uganda; 8 NIAID, NIH, Bethesda, MD, USA Background: The relationship between the humoral immune response to cytomegalovirus (CMV) and HIV pathogenesis is not well characterized in sub-Saharan Africa. In this secondary analysis of a randomized trial in Rakai, Uganda, we examined correlates of anti-CMV IgG levels and assessed their prognostic value on HIV disease progression and inflammation. Methods: HIV-1/HSV-2 co-infected Ugandan women (N=300) with CD4 counts between 300-400 cells/μL were enrolled in a double-blind, randomized, placebo-controlled trial to examine the effect of acyclovir use on HIV disease progression. The primary outcome was ART eligibility (CD4 count ≤250 cells/μL, a WHO Stage IV condition, non-traumatic death, or ART initiation for any reason). Anti-CMV IgG, C-reactive protein (CRP), and soluble CD14 (sCD14) levels were measured on serum samples at baseline, year 1 and year 2 (n=812 visits). Anti-CMV IgG levels were categorized into tertiles, and then dichotomized to compare the 3 rd tertile to the 1 st and 2 nd tertiles in combination. Cox proportional hazard regression models were used to evaluate the predictive value of baseline anti-CMV IgG levels on ART eligibility. Poisson regression with GEE was used to calculate prevalence risk ratios (PRR) to indicate associations with high anti-CMV IgG levels. Correlations were assessed using Spearman’s rank-order test. Results: Compared to lower levels of anti-CMV IgG at baseline, women in the 3 rd tertile had a higher probability of ART eligibility (29.7/100py vs. 53.4/100py, P <0.001). The hazard of reaching ART eligibility was greater for the 3 rd tertile at baseline independent of study arm, age, body mass index, CRP, sCD14, CD4 count, and HIV viral load (aHR=1.51, 95%CI=1.08-2.10; P <0.016). In a multivariate analysis censored for ART initiation, high anti-CMV IgG levels were associated with high CRP, high HIV viral load, and low CD4 counts ( P<0.05 ). Among 72 women who initiated ART during the study period, anti-CMV IgG levels were significantly higher post-ART initiation (PRR=1.82, 95%CI=1.33-2.49, P <0.001; n=113 visits) than pre-ART (n=95 visits). Log 10 anti-CMV IgG levels correlated with log 10 CRP levels (rho=0.21, P =0.05) and with log 10 sCD14 levels (rho=0.21, P =0.045) in fully suppressed women (n=88 visits, N=65). Conclusions: High anti-CMV IgG levels were predictive of HIV disease progression and were associated with increased systemic inflammation. Determinants of elevated anti-CMV IgG levels post-ART initiation require further investigation.

Poster Abstracts

304

CROI 2016