Background:  Risk of recurrent/severe herpes zoster (HZ) is increased in HIV⁺ patients. ZOSTAVAX^® (ZV; zoster vaccine live) is generally safe and effective in reducing HZ incidence/severity in HIV^– adults ≥50 years old, but has not been evaluated in HIV⁺ adults.

Methods: A randomized (3:1 ZV:placebo), double-blind, placebo-controlled trial assessed safety and immunogenicity of 2 doses of ZV in HIV⁺ adults ≥18 years old (CD4 >200 copies/µL; HIV RNA <75 copies/mL for ≥6 months on stable ART; varicella-zoster virus (VZV) seropositive, history of VZV or HZ >1 year prior to entry). Patients were stratified by screening CD4 (>350 copies/µL [H-CD4] vs ≥200 to 349 copies/µL [L-CD4]), received ZV or placebo on day 0 and week 6; and were evaluated at weeks 2, 6, 8, 12, and 24. Primary safety endpoint was International Conference on Harmonization (ICH)-defined serious adverse events or NIAID Division of AIDS (DAIDS) grade 3/4 signs/symptoms (excluding trauma-related) during 6-week post-vaccination periods. By predefined rule, ZV was considered generally safe if ≤18 patients in the ZV arm had primary safety endpoints. Immunogenicity endpoint was the mean natural log-transformed VZV ELISA titer at 6 weeks post each vaccination.

Results:  We enrolled 395 patients:  203 H-CD4 patients (152 ZV/51 placebo) and 192 L-CD4 patients (144 ZV/48 placebo); 3 (1 ZV, 2 placebo) received no vaccine and were excluded. Patients were 84% male; 66% white, 31% black, 22% Hispanic; median age 49 years; median H-CD4 602/µL, L-CD4 283/µL. Of 295 ZV patients, 15 experienced primary safety endpoints, none vaccine related. In the first 48 patients, median (IQR) baseline natural log VZV antibody titer was 5.60 (4.75 to 6.72) and was higher at week 12 for ZV (6.40, 5.57 to 7.01) vs placebo (5.15, 4.27 to 6.13); p = 0.017. Geometric mean fold-rise was 1.75 ZV vs 1.09 placebo. Week 12 VZV antibody titer (after 2 ZV doses) was similar to week 6 (1 dose). H-CD4 patients had higher antibody titer than L-CD4 patients over time (p = 0.024, linear mixed effects adjusted for baseline titer).

 *Number of patients with each adverse events; primary safety endpoint CI-based on exact permutation, adjusted for 2-stage design, others exact binomial; **p values Fisher’s exact test. ^†Rash consistent with HZ in 2 ZV and 2 placebo patients; 1 placebo patient PCR+ VZV wild type.

Conclusions:  Administration of 2 doses of ZV in HIV⁺ adults (CD4 ≥200 copies/µL) virologically suppressed on ART was generally safe, and preliminary data suggest it was immunogenic. Clinical HZ was observed in 2 ZV and 2 placebo patients; the only PCR confirmed and was VZV wild type.