Virologic effects of broadly neutralizing antibody VRC01 administration during chronic HIV-1 infection

Abstract

Passive immunization with HIV-1-neutralizing monoclonal antibodies (mAbs) is being considered for prevention and treatment of HIV-1 infection. As therapeutic agents, mAbs could be used to suppress active virus replication, maintain suppression induced by antiretroviral therapy (ART), and/or decrease the size of the persistent virus reservoir. We assessed the impact of VRC01, a potent human mAb targeting the HIV-1 CD4 binding site, on ART-treated and untreated HIV-1-infected subjects. Among six ART-treated individuals with undetectable plasma viremia, two infusions of VRC01 did not reduce the peripheral blood cell-associated virus reservoir measured 4 weeks after the second infusion. In contrast, six of eight ART-untreated, viremic subjects infused with a single dose of VRC01 experienced a 1.1 to 1.8 log10 reduction in plasma viremia. The two subjects with minimal responses to VRC01 were found to have predominantly VRC01-resistant virus before treatment. Notably, two subjects with plasma virus load <1000 copies/ml demonstrated virus suppression to undetectable levels for over 20 days until VRC01 levels declined. Among the remaining four subjects with baseline virus loads between 3000 and 30,000 copies, viremia was only partially suppressed by mAb infusion, and we observed strong selection pressure for the outgrowth of less neutralization-sensitive viruses. In summary, a single infusion of mAb VRC01 significantly decreased plasma viremia and preferentially suppressed neutralization-sensitive virus strains. These data demonstrate the virological effect of this neutralizing antibody and highlight the need for combination strategies to maintain virus suppression.

Trial registration: ClinicalTrials.gov NCT01950325.

Fig. 1.. Consolidated Standards of Reporting Trials flow diagram of VRC trial.

Consolidated Standards of Reporting Trials (CONSORT) diagram delineates the study enrollment of 27 subjects who were allocated to five groups. After VRC01 dose escalation in groups 1 to 5, enrollments to group 5 were restricted to viremic subjects on a one-infusion schedule. GM, gamma marker.

Fig. 2.. Detection of antibody VRC01 in sera.

(A to G) Serum VRC01 concentration in subjects after infusion with VRC01 is graphed over time for HIV-infected subjects enrolled in trial VRC 601 (Fig. 1 and table S1) as well as for HIV-1–uninfected subjects previously enrolled in trial VRC 602 (45). Day of VRC01 infusion IV or SC is indicated with black arrows and dose is labeled on the graphs. Limit of detection (LOD) of the assay is marked at 0.98 μg/ml. (H) Mean VRC01 serum concentration is graphed after a single infusion with 40 mg/kg IV in HIV-1–uninfected subjects (light brown; n = 5), aviremic subjects (maroon; n = 3), and viremic subjects (yellow; n = 8). Error bars are derived from SEM.

Fig. 3.. The effect of VRC01 infusion on virus load in plasma and blood CD4 T cells during ART.

(A) Plasma viremia was measured by SCA (n = 5). (B and C) The percentage of total peripheral CD4 T cells in each sample containing total HIV DNA (B) or integrated HIV DNA (C) was measured by quantitative polymerase chain reaction (qPCR) (n = 6). (D) The percentage of cells inducibly transcribing multiply spliced tat/rev RNA was quantified by TILDA (n = 6). (E) The percentage of cells producing replication-competent HIV after in vitro stimulation was measured by quantitative coculture (n = 4). (F to K) The percentages of sorted CM and EM CD4 T cell subsets containing total HIV DNA were measured by fluorescence-assisted clonal amplification (FCA) (F and G) (n = 6). Copies of unspliced gag RNA (H and I) and multiply spliced rev RNA (J and K) per HIV DNA copy in CM (H and J) and EM (I and K) CD4 T cell subsets were measured by qRT-PCR (n = 6). Vertical dashed lines in each plot indicate the days of the two VRC01 infusions. Participants who received VRC01 doses of 20 mg/kg (n = 3) are indicated with circular symbols and solid lines; those who received 40 mg/kg (n = 3) are indicated with triangular symbols and dashed lines. Where samples yielded values below the assay detection limit, this detection limit is plotted as an open symbol.

Fig. 4.. The effect of VRC01 on plasma virus load after infusion.

(A) Plasma virus load over time for eight viremic subjects before and after infusion with VRC01 (day 0, indicated by arrow). Limit of detection of the assay is indicated with a tick at 20 copies/ml. (B) Log₁₀ change in virus load from average individual baseline for each of the eight subjects longitudinally for 90 days after infusion. Black line indicates mean log₁₀ change in virus load for all viremic subjects. (C) Early changes in virus load are magnified by graphing all collected data points from day of infusion (day 0) to 28 days after infusion. Gray box highlights days 3 to 21, at which times there was a statistically significant decline in mean virus load from baseline by paired t test. Mean virus load nadir occurred on day 9 after infusion, indicated by a black vertical line. Asterisks denote subjects for whom virus became undetectable (<20 copies/ml) and thus have limited decline in virus load.

Fig. 5.. Kinetics of virus and antibody clearance.

(A) About 10 representative preinfusion env genes were cloned and expressed as pseudoviruses for each subject. Sensitivity to VRC01 is plotted as IC₈₀ values for each virus, and line indicates geometric mean. (B) Plasma virus load (left y axis) and serum VRC01 concentration (right y axis) are graphed over time for each viremic individual. Limit of detection of each assay (20 copies/ml and 0.098 μg/ml) is indicated on the respective axes with a tick. Solid gray boxes outline the range of preinfusion IC₈₀s as shown in (A), and geometric mean IC₈₀ values are labeled. Arrows indicate day of VRC01 infusion (day 0).

Fig. 6.. Selection for reduced sensitivity to VRC01 in postinfusion virus.

Envs cloned from baseline and 1-month post-infusion for each subject were tested for neutralization sensitivity to VRC01. (A) IC₈₀s of each Env clone are plotted for two time points (before infusion in circles and after infusion in squares) for subjects who had detectable virus (>20 copies/ml). Black line indicates geometric mean IC₈₀, and groups were compared by Mann-Whitney. Significant P values (<0.05) are shown. (B) V1-to-V5 sequences amplified by SGA and FCA from preinfusion and 1-month postinfusion were aligned and analyzed using a NEP algorithm (http://exon.niaid.nih.gov) to estimate the likelihood of residue changes being associated with VRC01 neutralization. Subject 23, whose majority of viruses at each time are genetically distinct, and subject 26, for whom no selection was detected, were excluded from this analysis. Residue changes for the top-ranked residue for each subject is shown by pie charts that indicate percent of preinfusion sequences (inner pie) or postinfusion sequences (outer pie) containing indicated residues at specified position (HXB2 numbering). (C) Length of V5 sequences amplified by SGA is plotted for preinfusion (circles) and postinfusion (squares) sequences. Lines indicate median and interquartile range, and significance was calculated by Mann-Whitney. Significant P values (<0.05) are shown. AA, amino acid.