Durability of immunity and clinical protection in nursing home residents following bivalent SARS-CoV-2 vaccination

Abstract

Background: Vaccines have substantially mitigated the disproportional impact of SARS-CoV-2 on the high morbidity and mortality experienced by nursing home residents. However, variation in vaccine efficacy, immune senescence and waning immunity all undermine vaccine effectiveness over time. The introduction of the bivalent vaccine in September 2022 aimed to counter this increasing susceptibility and consequences of breakthrough infection, however data on the durability and protection of the vaccine are limited. We evaluated the durability of immunity and protection after the first bivalent vaccination to SARS-CoV-2 in nursing home residents.

Methods: For the immunologic evaluation, community nursing home volunteers agreed to serial blood sampling before, at two weeks, three and six months after each vaccination for antibodies to spike protein and pseudovirus neutralization activity over time. Concurrent clinical outcomes were evaluated by reviewing electronic health record data from residents living in Veterans Administration managed nursing home units. Residents without recent infection but prior vaccination to SARS-CoV-2 were followed over time beginning with administration of the newly available bivalent vaccine using a target trial emulation (TTE) approach; TTE compared time to breakthrough infection, hospitalization and death between those who did and did not receive the bivalent vaccine.

Results: We evaluated antibodies in 650 nursing home residents; 452 had data available following a first monovalent booster, 257 following a second monovalent booster and 321 following a bivalent vaccine. We found a rise in BA.5 neutralization activity from the first and second monovalent boosters through the bivalent vaccination regardless of prior SARS-CoV-2 history. Titers declined at three and six months after the bivalent vaccination but generally exceeded those at three months compared to either prior boost. BA.5 neutralization titers six months after the bivalent vaccination were diminished but had detectable levels in 80% of infection-naive and 100% of prior infected individuals. TTE evaluated 5903 unique subjects, of whom 2235 received the bivalent boost. TTE demonstrated 39% or greater reduction in risk of infection, hospitalization or death at four months following the bivalent boost.

Conclusion: Immunologic results mirrored those of the TTE and suggest bivalent vaccination added substantial protection for up to six months after bivalent vaccination with notable exceptions. However, the level of protection declined over this period, and by six months may open a window of added vulnerability to infection before the next updated vaccine becomes available. We strongly agree with the CDC recommendation that those who have not received a bivalent vaccination receive that now and these results support a second bivalent booster for those at greatest risk which includes many nursing home residents.

Figure 1.. Neutralization assays after boosting

This figure illustrates pseudovirus neutralization assay results for Omicron BA.5 (top panels), and Wuhan strains (bottom panels) in nursing home residents before and after first and second monovalent booster and the bivalent booster. The blue panels are infection-naive and the red panels had prior infection. Not all subjects had all timepoints drawn. N is the subjects in that group. The solid line is the median response. pNT50 = Pseudovirus neutralization. X axis timepoints: Preboost is 0-14 days prior the the first monovalent booster, Postboost 1st, 2nd, BV are a median 17 days, M3 Post is median 105 days, and M6 Post is median 183 days.

Figure 2.. Decline of neutralization titers after boosting

pNT50 neutralization titer of BA.5 (top) and Wuhan (bottom) of days from each booster dose in infection naive (blue) and prior infected (red). The lines represent the estimated fixed effects from a mixed-effects model predicting log-transformed titer response as a function of days since vaccine dose with a 2nd degree polynomial function and random intercepts at the subject level.

Figure 3.. Anti-spike titers after boosting

This figure illustrates anti-spike for BA.5 (top) in AU/ml and Wuhan (bottom) in BAU/ml in nursing home residents before and after first and second monovalent booster and the bivalent booster. The blue panels are infection-naive and the red panels had prior infection. Not all subjects had all timepoints drawn. N is the subjects in that group. The solid line is the median response. X axis timepoints: Preboost is 0-14 days prior the the first monovalent booster, Postboost 1st, 2nd, BV are a median 17 days, M3 Post is median 105 days, and M6 Post is median 183 days.

Figure 4.

Km plots for infection following bivalent vaccination in comparison to controls in a TTE design.

Figure 5.

Risk differences for SARS-CoV-2 clinical outcomes of infection, hospitalization, death or a combination of hospitalization or death up to 16 weeks after bivalent vaccination.