Durable protection against repeated penile exposures to simian-human immunodeficiency virus by broadly neutralizing antibodies

Abstract

Penile acquisition of HIV accounts for most infections among men globally. Nevertheless, candidate HIV interventions for men advance to clinical trials without preclinical efficacy data, due primarily to a paucity of relevant animal models of penile HIV infection. Using our recently developed macaque model, we show that a single subcutaneous administration of broadly neutralizing antibody (bNAb) 10-1074 conferred durable protection against repeated penile exposures to simian-human immunodeficiency virus (SHIV_SF162P3). Macaques co-administered bNAbs 10-1074 and 3BNC117, or 3BNC117 alone, also exhibited significant protection against repeated vaginal SHIV_AD8-EO exposures. Regression modeling estimated that individual plasma bNAb concentrations of 5 μg ml^-1 correlated with ≥99.9% relative reduction in SHIV infection probability via penile (10-1074) or vaginal (10-1074 or 3BNC117) challenge routes. These results demonstrate that comparably large reductions in penile and vaginal SHIV infection risk among macaques were achieved at clinically relevant plasma bNAb concentrations and inform dose selection for the development of bNAbs as long-acting pre-exposure prophylaxis candidates for use by men and women.

Fig. 1. Passive immunization of macaques with bNAb 10–1074 delays SHIV acquisition following repeated penile SHIV challenges.

a Study design to assess the protective efficacy of 10–1074 against repeated penile SHIV challenges. Rhesus macaques (N = 6; Indian origin) were injected subcutaneously once with 10–1074 (10 mg kg⁻¹). Beginning 1 week later, macaques were challenged repeatedly, once per week, via the penis with SHIV_SF162P3 (200 TCID₅₀ into the prepuce pouch and 16 TCID₅₀ into the distal urethra) until systemic SHIV infection was confirmed via plasma viral load assay. Control macaques (N = 10) received no antibody, but were challenged identically. Plasma viral load (vRNA copies ml⁻¹) determined in 10–1074-treated macaques (b) or untreated controls (c). Symbols denote individual animals. d Percentages of macaques remaining uninfected at 7 days following administration of the indicated cumulative number of SHIV challenges. Statistical difference between groups was measured using a two-sided log-rank test (P = 0.0007). e Plasma levels of 10–1074 (μg ml⁻¹) were determined via TZM-bl neutralization assays using 10–1074-sensitive pseudovirus X2088.c9. f Plasma concentrations of 10–1074 in individual macaques (N = 6) at the time of SHIV breakthrough (values from e).The solid line denotes the group median; the dotted line denotes the limit of detection for 10–1074. Source data are provided in Supplementary Tables or a Source data file. SHIV infection data for animals in the control group have been reported previously in a paper describing the development of our penile SHIV infection model and are reproduced here to facilitate comparison between the treatment and control groups.

Fig. 2. 10–1074 antibody concentration predicts the probability of infection following penile SHIV_SF162P3 challenge.

a Probit regression analysis was used to estimate the per-challenge probability of penile SHIV acquisition as a function of imputed 10–1074 concentration in plasma. The per-challenge probability of infection among untreated control animals was 0.25 (95% CI 0.13–0.41) and is denoted in blue. Tick marks represent SHIV challenge events that either did (red) or did not (green) result in systemic infection. The solid black line is the model-based prediction; dotted black lines depict 90% pointwise confidence intervals. b Model-based estimates of SHIV infection probabilities following a single penile challenge and reduction of infection probability as compared with controls, at the indicated plasma concentrations of 10–1074.

Fig. 3. Passive immunization with bNAbs 10–1074 and 3BNC117 in combination, or 3BNC117 singly, delays SHIV acquisition following repeated vaginal challenges in DMPA-treated macaques.

a Rhesus macaques (Chinese origin) were injected subcutaneously once with a combination of bNAbs 10–1074 and 3BNC117 (10 mg each kg⁻¹; N = 5) or 3BNC117 singly (10 mg kg⁻¹; N = 6). Commencing 1 week later, macaques were challenged repeatedly, once per week, intravaginally with SHIV_AD8-EO (300 TCID₅₀) until systemic SHIV infection was confirmed via plasma viral load assay. Control macaques (N = 3) received no antibody but were challenged identically. Animals in both groups received DMPA (30 mg) intramuscularly at 2 weeks before the first SHIV challenge and every 4 weeks thereafter to normalize SHIV susceptibility. Plasma viral loads (vRNA copies ml⁻¹) for macaques in the combination bNAb group (b), single bNAb group (c) or untreated control group (d). Symbols denote individual animals. e Percentages of macaques remaining uninfected at 7 days following administration of the indicated cumulative number of SHIV challenges. Statistical differences between groups were analyzed using a two-sided log-rank test. Plasma levels of 10–1074 (μg ml⁻¹) or 3BNC117 (μg ml⁻¹) were determined via TZM-bl neutralization assays using 10–1074-sensitive pseudovirus X2088.c9 or 3BNC117-sensitive pseudovirus Q769.d22 for macaques that received both 10–1074 and 3BNC117 (f) or 3BNC117 alone (g). h Plasma concentrations of 10–1074 or 3BNC117 in individual macaques (N = 5 in Group-1, N = 6 in Group-2) at the time of SHIV breakthrough (values from f (Group-1) or g (Group-2)). Solid lines denote group medians; dotted lines indicate the limits of detection for 10–1074 (black) or 3BNC117 (gray). Source data are provided in Supplementary Tables or a Source data file.

Fig. 4. Antibody concentration predicts the probability of infection following vaginal SHIV_AD8-EO challenge.

Probit regression analysis was used to estimate the per-challenge probability of vaginal SHIV acquisition as a function of imputed 3BNC117 concentrations in plasma among macaques that were administered 3BNC117 singly (a) or 10–1074 concentrations in plasma among animals that received 10–1074 in combination with 3BNC117 (b). The per-challenge probability of infection among untreated control animals was 0.50 and is denoted in blue. Tick marks represent SHIV challenge events that either did (red) or did not (green) result in systemic infection. The solid black or gray line is the model-based prediction; dashed black lines depict 90% pointwise confidence intervals. Model-based estimates of SHIV infection probabilities following a single vaginal SHIV challenge at the indicated plasma concentrations of 3BNC117 (c) or 10–1074 (d).