Effectiveness of recombinant herpes zoster vaccine by immune status and prior receipt of live zoster vaccine in Medicare enrollees, 2018-2019

Background: The two-dose recombinant zoster vaccine (RZV) is highly efficacious against herpes zoster (HZ) in adults ages 50+ and replaced the live zoster vaccine (ZVL), which was contraindicated for adults with immunocompromising conditions. Immunocompromised adults were excluded from RZV trials, thus limiting recommendations for use in adults at greatest risk for HZ.

Objectives: We used the target trial emulation framework to evaluate the effectiveness of RZV in older adults in the US with and without immunocompromising conditions.

Methods: We emulated 12 year-long pragmatic randomized trials beginning January 2018. Medicare enrollees were included in as many emulated trials for which they were eligible (no prior history of HZ or prior RZV). Vaccination assignment for each emulated trial was based on participants’ RZV receipt within a one-month enrollment period. RZV was defined using medication and procedure codes. HZ (the primary outcome of interest) was defined using a diagnosis code-based algorithm. Nine immunocompromising conditions were defined using codes reported by the Food and Drug Administration and measured as baseline covariates during the enrollment period for each emulated trial. Participants in each trial were followed for up to 12 months until a diagnosis of HZ, receipt of ZVL, admission to a skilled nursing facility or hospice, death or December 31, 2019. Risk ratios and risk differences were estimated from survival analysis with robust variance estimation. Next, we used the clone-censor-weight method to estimate the effectiveness of receiving two versus one RZV dose. Each RZV recipient was cloned and assigned to each treatment strategy. Those who did not comply with their assigned RZV strategy were censored. Participants were weighted on the inverse probability of being censored and the hazard ratio for HZ was estimated comparing the two versus one dose strategies.

Results: From 12 emulated trials including ~3 million participants each, the vaccine effectiveness (VE) of RZV was 53% (95% CI: 50%, 56%). VE was similar between immunocompromised adults (51%, 95% CI: 41%, 59%) and immunocompetent adults (54%, 95% CI: 50%, 57%), and was lower among adults who previously received ZVL (47%, 95% CI: 40%, 53%) versus those who did not (55%, 95% CI: 51%, 58%). Among RZV recipients, two doses conferred a 30% reduction in HZ incidence compared to a single dose .

Conclusions: RZV reduces HZ incidence in healthy Medicare enrollees, and two doses protect better than one dose. Prior recipients of the discontinued ZVL might benefit from revaccination with RZV.