Human immune globulin treatment controls Zika viremia in pregnant rhesus macaques

Abstract

There are currently no approved drugs to treat Zika virus (ZIKV) infection during pregnancy. Hyperimmune globulin products such as VARIZIG and WinRho are FDA-approved to treat conditions during pregnancy such as Varicella Zoster virus infection and Rh-incompatibility. We administered ZIKV-specific human immune globulin as a treatment in pregnant rhesus macaques one day after subcutaneous ZIKV infection. All animals controlled ZIKV viremia following the treatment and generated robust levels of anti-Zika virus antibodies in their blood. No adverse fetal or infant outcomes were identified in the treated animals, yet the placebo control treated animals also did not have signs related to congenital Zika syndrome (CZS). Human immune globulin may be a viable prophylaxis and treatment option for ZIKV infection during pregnancy, however, more studies are required to fully assess the impact of this treatment to prevent CZS.

Fig 1. Animal demographics and study timeline.

(A) Animal ID, animal age at the time of ZIKV or mock inoculation, gestational day (GD) at the time of infection, and gestational day at the time of cesarean section (C-section). Each study arm is represented in subsequent graphs by the color shown in the monkey for each group. (B) Timeline representing the day of inoculation, treatment, blood sampling (ticks on the line), and C-section for animals in this study relative to either days post-inoculation (dpi) or gestational day (GD). Blood sampling occurred as shown through 27 dpi and then was twice weekly until two consecutive time points were negative for ZIKV RNA and then sampling was once weekly.

Fig 2. Dam viral load analysis between ZIKV-IG, placebo-IG, and untreated animals.

(A) Longitudinal plasma viral load for animals across each group. Red = ZIKV-IG-treated, blue = placebo-IG-treated, gray = untreated. The x-axis is set to the limit of quanti of the assay (100 copies/ml). (B) Peak viral load was statistically different for each group shown with the mean and standard deviation (Kruskall-Wallis chi-squared = 7.1263, df = 2, p = 0.03). ZIKV-IG-treated dams had lower viremia levels than untreated controls (p = 0.02; *p<0.05), but placebo-IG-treated animals did not (p = 0.89) and ZIKV-IG- and placebo-IG were also not statistically different (p = 0.130). (C) Area under the curve was statistically different between the ZIKV-IG group and untreated animals (p = 0.01; **p<0.01) but the placebo-IG group was not (p = 0.60). ZIKV-IG- and placebo-IG-treated groups were not statistically different from each other (p = 0.1859). The graph shows the mean and standard deviation. (D) Duration of viremia, as shown with the mean and standard deviation, was not different between groups (Kruskal-Wallis chi-squared = 5.05, df = 2, p = 0.08). (E) The total number of days with a positive viral load was different between the ZIKV-IG and untreated groups (p = 0.01) but not between placebo-IG and untreated animals (p = 1.00). ZIKV-IG- and placebo-IG-treated groups were not statistically different from each other (p = 0.085). Data shown as the mean with standard deviation.

Fig 3. Measurement of human or macaque-specific binding antibody titers by whole virion binding ELISA.

(A) Concentration of human ZIKV-specific IgG antibodies measured longitudinally in ZIKV-IG-treated animals. (B) EC50 of macaque ZIKV-specific IgG antibodies in ZIKV-IG-treated animals. (C) EC50 of macaque ZIKV-specific IgG antibodies in placebo-IG and untreated animals. Dotted lines represent the limit of detection for each assay. Vertical dotted lines indicate the time points of IgG infusion.

Fig 4. Plaque reduction neutralizing antibody titers (90%) of ZIKV-IG and placebo-IG-treated animals.

Longitudinal neutralizing antibody titers of ZIKV-IG-treated animals (red). Placebo-IG PRNT90 titers were measured at 0 and 27 dpi only (blue). Dotted vertical lines represent the time point 1 hour post-ZIKV-IG infusion for reference. Mean titers with 95% CI are plotted.

Fig 5. Ultrasound growth z-scores of ZIKV-IG, placebo-IG, untreated, and mock infected animals relative to normative data from CNPRC.

The legend shows individual symbols for each ZIKV-IG and placebo-IG-treated animal while untreated and mock infected groups are represented by gray and turquoise open circles, respectively. Lines represent the mean z-score at each time point for each group starting at 3–4 weeks post-infection when data could be collected from all pregnancies. The dotted line represents no change from normative data set to 0. (A) Biparietal diameter z-scores. (B) Head circumference z-scores. (C) Abdominal circumference z-scores. (D) Femur length z-scores.