Growing Understanding of the Clinical and Serologic Effects of COVID-19 Vaccines in Patients Undergoing Long-Term Dialysis

Individuals on maintenance dialysis are disproportionately affected by the coronavirus disease 2019 (COVID-19) pandemic. Infection rates are substantially higher among patients on dialysis compared with individuals not on dialysis. This was particularly true during the early stages of the pandemic when the incidence of infection among patients on dialysis was more than double that of the general population (1). Contributing factors to this heightened incidence include the high levels of contact with the health care system (including both providers and other patients) and a limited ability to physically distance inherent to the nature of how maintenance dialysis takes place. Additionally, many patients on dialysis reside in congregate settings, a risk factor linked with over 17-fold higher odds of contracting COVID-19 (2). Unfortunately, not only are infection rates higher among patients on dialysis, but the severity of infection is also higher, culminating in a large number of deaths. Prior to the vaccine rollout, mortality rates among patients on maintenance dialysis who contracted COVID-19 were reported at an astonishing 25% (2).

Given the higher occurrence and severity of COVID-19 disease, a number of containment measures were implemented early in the pandemic in many dialysis units around the world. These measures included universal masking for both patients and providers, individualized transportation of patients to and from the unit, social distancing in the waiting room, screening upon entrance to the unit, and early detection and isolation of positive cases (3).

Moreover, COVID-19 vaccines play a major role in both reducing the incidence and mitigating the severity of infection. Many nephrologists advocated for the prioritization of early vaccination for patients on dialysis given the sheer amount of morbidity and mortality that this group has sustained during the pandemic (4). Ultimately, there was a wide degree of variation in the vaccine priority level provided to patients on dialysis both between and within countries.

Notably, patients on dialysis were largely excluded from the vaccine development trials for safety reasons. Combine this with the fact that patients on dialysis are known to have attenuated responses to other vaccines, we were left to deal with the underlying possibility that patients on dialysis may not mount an adequate response to COVID-19 vaccination. Therefore, we now rely upon “real-world” and almost “real-time” data to understand how well these vaccines actually perform in the dialysis population. Among patients on maintenance dialysis, seropositivity has been reported at approximately 90% following two doses of an mRNA vaccine (5). However, when compared with the nondialysis populations, patients on dialysis demonstrated incomplete and delayed humoral response along with blunted cellular response to these vaccines (6). Moreover, reports of waning immunity leading to breakthrough infections are being reported in the general population (7). New evidence regarding whether this effect is magnified in the dialysis population is discussed below.

In this issue of CJASN, two additional population-level cohort studies further enhance our understanding of the clinical and serologic effects of COVID-19 vaccination on the maintenance dialysis population (8,9). The study by El Karoui et al. (8) utilizing Bayesian spatiotemporal modeling compared trends in COVID-19 hospitalizations between the dialysis population and the general population throughout the waves of the pandemic. In the process, the investigators assessed the association between preventive measures, including vaccination, and COVID-19 hospitalization rates. This French study was performed at the population level by utilizing a nationwide registry, which included 53,365 at-risk patients on dialysis. In a novel approach, three analytic models incorporating data on previous COVID-19 hospitalizations were used to predict subsequent COVID-19 hospitalizations at three distinct time points: between the first and second waves, at the beginning of the vaccination campaign for the dialysis population, and following full rollout of vaccination availability to the general population.

During the first wave of the pandemic, the incidence of COVID-19 hospitalizations among patients on dialysis was comparable with that of the general population. Notably, this differs compared with prior studies from other geographic settings, which reported higher incidence rates among the dialysis population (1). Interestingly, the exception to this was younger patients on dialysis (25–35 years old) who experienced an approximately four-fold higher hospitalization rate relative to the same age group in the general population (8). Model 1 (on the basis of first wave data) overestimated the hospitalization rate among patients on dialysis during the second wave, thereby suggesting that perhaps containment measures implemented between the first and second waves successfully reduced cases. However, on the basis of the inherent limitations to the study design, we do not know whether specific containment measures or declines in infection severity (i.e., intrinsic viral characteristics) were responsible. Meanwhile, model 2 (on the basis of data up to the beginning of the vaccination campaign for patients on dialysis) overestimated the hospitalization rate among patients on dialysis during the third wave, thereby suggesting that vaccination of patients on dialysis was associated with a further reduction in severe COVID-19 cases.

Model 3 (on the basis of data up through when vaccine coverage was expanded to the general population) found that not only was personal vaccination associated with a three-fold reduction in hospitalization risk among patients on dialysis, but also higher vaccination coverage of the general population was independently associated with a lower hospitalization risk among patients on dialysis. In fact, every 10% higher vaccination coverage for the same-age general population was associated with a reduction in hospitalization risk by approximately half for patients on dialysis. This suggests that herd immunity from vaccination of the general population as a whole provides further protection to patients on dialysis. Importantly, the beneficial effect of vaccination in preventing hospitalization was most pronounced among the youngest patients on dialysis. Although the results of this study must be interpreted within the inherent limitations of administrative data where granular details are lacking (e.g., individual vaccination data were not available), these results do speak to the protective effects of containment measures, immunity through personal vaccination, and herd immunity through vaccination of the general population.

The study by Hsu et al. (9), also published in this issue of CJASN, focuses specifically on the serologic response of patients on maintenance dialysis to COVID-19 vaccinations over time. The study cohort included 1870 patients treated at Dialysis Clinic, Inc. who received two doses of a COVID-19 vaccine and had their postvaccine antibody titers measured as part of routine monthly blood work. From the overall cohort, 16% were previously infected with COVID-19, whereas 84% were not. The breakdown of vaccine types administered was 50% Moderna (mRNA), 28% Pfizer (mRNA), and 22% Janssen (adenovirus vector). The primary analysis consisted of comparing antibody titers over time by vaccine type. The secondary analysis assessed the durability of seroresponse on the basis of the magnitude of the initial seroresponse via a time-to-event analysis.

Among patients on dialysis without a history of COVID-19 infection, the Moderna and Pfizer vaccines resulted in a strong initial rise in median antispike IgG antibody titers to above the upper limit of assay detection (≥20) at 1 month postvaccination. The titer levels waned to approximately eight and two, respectively, at 6 months postvaccination. Concerningly, patients without a history of COVID-19 infection who received the Janssen vaccine showed almost no serologic response, with median titers below the lower limit of assay detection (less than one) throughout the postvaccination period. Overall, this translated into undetectable titers in 10% of Moderna recipients, 33% of Pfizer recipients, and 52% of Janssen recipients at 6 months postvaccination. Patients with a previous COVID-19 infection had median titers that remained above the 75th percentile of the upper limit of assay detection throughout the postvaccination period, independent of vaccine type. Moreover, the magnitude of the initial seroresponse was highly correlated with more sustained seroresponse across all vaccine types. In fact, titers between one and 19 within 2 months postvaccination were associated with a 15-fold higher risk of loss of seroresponse (defined as either an undetectable antibody titer or development of COVID-19 infection) compared with an initial maximum titer of ≥20.

Similar results of rapid waning of antibody response to vaccination were recently reported from a prospective study of 4791 patients on dialysis from the US Renal Care network (10). In this study, the proportion of patients with an undetectable antibody titer increased from 7% at 14–30 days postvaccination to 20% at 5–6 months postvaccination. Moreover, individuals with lower antibody titers were at a higher risk for breakthrough COVID-19 infection (10).

As we continue to learn about the effects of COVID-19 on the dialysis population and the real-world effectiveness of different mitigation strategies, we must continue to adapt our approach to protect this vulnerable population as best as possible. Evidence suggests that containment strategies both within and outside of dialysis units reduce infection rates, and we must remain vigilant about maintaining these precautions. Further, maximizing vaccination rates among not only patients on dialysis (where mRNA vaccines are clearly preferred) but also the general population to provide herd immunity must remain a top priority. Given the attenuated initial immunologic response combined with rapidly waning immunity over several months postvaccination, booster doses of COVID-19 vaccines will likely be critical for this population to prevent breakthrough cases. More widespread testing for sustained serologic immunity may even need to be considered to determine the potential need for and frequency of ongoing booster doses within the dialysis population.

Disclosures

M.M. Sood reports speaker fees from AstraZeneca; serving on the editorial boards of American Journal of Kidney Disease, Canadian Journal of Cardiology, and CJASN; serving as a deputy editor of Canadian Journal of Kidney Disease and Health; and serving as a member of the American Society of Nephrology Highlights ESRD Team. The remaining author has nothing to disclose.

Funding

G.L. Hundemer is supported by Kidney Research Scientist Core Education and National Training (KRESCENT) Program New Investigator Award 2019KP-NIA626990. M.M. Sood is supported by the Jindal Research Chair for the Prevention of Kidney Disease.

Author Contributions

G.L. Hundemer and M.M. Sood conceptualized the study; G.L. Hundemer was responsible for data curation; M.M. Sood provided supervision; G.L. Hundemer wrote the original draft; and G.L. Hundemer and M.M. Sood reviewed and edited the manuscript.