Human Immunodeficiency Virus (HIV)-Antibody Repertoire Estimates Reservoir Size and Time of Antiretroviral Therapy Initiation in Virally Suppressed Perinatally HIV-Infected Children

Abstract

Background: Assays to estimate human immunodeficiency virus (HIV) reservoir size require large amounts of blood, which represents a drawback especially in pediatric settings. We investigated whether HIV-antibody repertoire could estimate the viral reservoir size. Moreover, we assessed the magnitude of HIV-antibody response as a predictor of time of antiretroviral therapy (ART) initiation.

Methods: Human immunodeficiency virus-antibody responses to 10 different viral proteins were evaluated by HIV Western blot (WB) kit and a WB score was assigned to each patient. Patients were classified in 2 subgroups based on the timing of ART initiation (early treated [ET], 0-24 weeks and late treated [LT], >24 weeks). Human immunodeficiency virus-deoxyribonucleic acid (DNA) was quantified using real-time quantitative polymerase chain reaction on total peripheral blood mononuclear cells. Logistic regression and principal component analysis were built on these data to test the ability of WB score to predict the expected value of HIV-DNA and the timing of ART initiation.

Results: Sixty-nine perinatally HIV-infected children were evaluated. Reduced HIV-specific antibody responses and lower size of HIV-DNA were observed in ET compared with LT patients (P < .001 and P = .02, respectively). We found that WB score correlates with HIV-DNA (P = .032) and timing of ART initiation (P < .001). Based on the logistic regression analysis, we found that WB score can predict the HIV-DNA size and the timing of ART initiation with an Akaike information criterion of -118.13 and -151.51, respectively.

Conclusions: Western blot score can estimate HIV-DNA size and timing of ART initiation in long-term virally suppressed children. This rapid, inexpensive, and easily reproducible tool can provide useful information to identify potential candidates for HIV remission studies.

Keywords: HIV reservoir; HIV-antibody repertoire; children; early antiretroviral therapy; serology.

Figure 1.

Human immunodeficiency virus (HIV) serostatus and viral reservoir size in early treated (ET) and late treated (LT): (A) representative Western blot (WB) patterns for positive control, strong positivity to gp160 and p24, weak positivity to gp160 and p24, and negative response are shown. (B) Proportion of seropositive (SP) children in ET (antiretroviral therapy [ART] start: 0–24 weeks) and in LT group (ART start >24 weeks). Patients were considered SP when WB score was ≥0.5 and positive to CMIA test. (C) Scatter dot plot depicts WB score analysis in ET and LT. (D) Contingency plot shows percentage of SP children in ET and LT groups who were positive for Envelope antigens (gp160, gp120, gp41), reverse transcriptase (p66 and p51), integrase protein (p31), core antigen (p55, p24, and p17), and p39 antigen. Only significant P values (P < .05) have been reported in the figure. (E) Comparison between HIV-deoxyribonucleic acid (DNA) measured in ET and LT group and (F) in seronegative (SN) and SP patients among ET group.

Figure 2.

Human immunodeficiency virus (HIV) antibody repertoire predicts size of viral reservoir and time of antiretroviral therapy (ART) initiation: correlation analysis (Spearman test) between (A) HIV-deoxyribonucleic acid (DNA) and Western blot (WB) score; (B) HIV-DNA and timing of ART initiation expressed in weeks, and (C) WB score and ART initiation expressed in weeks. Early treated and late treated individuals were analyzed together. (D) Principal component analysis (PCA) shows segregation among patients with different WB score according to HIV-DNA and ART initiation; (E) representative cartoon shows the relation between WB score, HIV-DNA, and timing of ART initiation based on logistic regression model and PCA. From left to right: time of ART initiation (weeks), range of WB score, and size of HIV-DNA. Different range of HIV-DNA and WB scores have been identified according to logistic regression model and PCA: Low (value from 0 to 200 copies/106 peripheral blood mononuclear cells [PBMCs]) associated with a WB score ranging from 0 to 2.5; Intermediate (200 to 800 copies/106 PBMC) associated with a WB score ranging from 3 to 6; High (800 to 1500 copies/106 PBMC) associated with a WB score over 6.