COVID-19 vaccination during therapy in relation to COVID-19 death in CLL

TO THE EDITOR:

Patients with chronic lymphocytic leukemia (CLL) are immunocompromised and recommended to receive COVID-19 vaccination, regardless of treatment status or specific therapy; however, factors including active CLL treatment are associated with poor serological vaccination response.¹ This analysis investigated the relationship between COVID-19 vaccination status and COVID-19 death using blinded data from the ongoing AMPLIFY trial (ClinicalTrials.gov identifier #NCT03836261), including patients in the global safety population on study after 15 March 2021 (patients who died or exited the study before this date were not included).

AMPLIFY is a global open-label phase 3 trial evaluating acalabrutinib + venetoclax with or without obinutuzumab vs chemoimmunotherapy (fludarabine/cyclophosphamide/rituximab or bendamustine/rituximab) in patients with previously untreated CLL. Although the study was blinded at the time of this analysis, patients in the acalabrutinib-containing arms received oral acalabrutinib 100 mg twice daily for 14 28-day cycles. Oral venetoclax dosing included a ramp-up over a 5-week period starting on cycle 3 to a maximum dose of 400 mg daily in cycle 4, continuing until the end of cycle 14. Obinutuzumab was administered IV at 1000 mg once on days 1 (or 100 mg on day 1 and 900 mg on day 2 if the first dose was split), 8, and 15 of cycle 2, and day 1 of cycles 3 through 7. Patients receiving chemoimmunotherapy received up to 6 cycles of either fludarabine + cyclophosphamide + rituximab or bendamustine + rituximab, dosed according to standard institutional practice. Patients were categorized by COVID-19 vaccination status (vaccinated or unvaccinated); those who were vaccinated before the observation period or at any time during the observation period were included in the vaccinated group. The primary outcome of this ad hoc analysis was time to COVID-19 death, which was defined as time from analysis start date (the latter of 15 March 2021 or the randomization date) to date of death or censoring, and analyzed using a Cox model with a time-dependent covariate for COVID-19 vaccination and adjusted for geographic region, cumulative illness rating scale–geriatric total score, and any B symptoms at baseline. All analyses were intended to be descriptive, with no multiple testing control. The protocol was approved by an independent review and ethics committee at each site and all patients provided written informed consent.

As of data cutoff (31 July 2023), 799 patients were enrolled (median age, 61 years; 64.7% male); 372 (46.6%) had received ≥1 COVID-19 vaccination, and 427 (53.4%) were unvaccinated. The most common COVID-19 vaccination administered was tozinameran (65.9%), followed by unspecified COVID-19 vaccination (19.4%), elasomeran (17.5%), nonreplicating viral vector adenovirus (ChAdOx1 nCoV-19; 11.6%), inactivated (VeroCell) CZ02 (3.2%), nonreplicating viral vector adenovirus-26 (Gam-COVID-Vac)/nonreplicating viral vector adenovirus-5 (Gam-COVID-Vac; 1.3%), nonreplicating viral vector adenovirus-26 (JNJ 78436735; 1.3%), and unspecified COVID-19 messenger RNA vaccination (1.1%). In total, 279 of the vaccinated patients (75%) received their first COVID-19 vaccination dose before or within 12 months after the first dose of study treatment (6.2% before study treatment, 7.8% within 0 to <3 months, 19.4% within 3 to <6 months, and 41.7% within 6 to <12 months). Baseline characteristics were relatively well balanced between vaccinated and unvaccinated patients, although regional differences in vaccination status were seen, particularly for Eastern and Western Europe (Table 1). Baseline medical histories relevant to COVID-19 (vascular, metabolic, respiratory, cardiac, renal, and liver disorders; neoplasms; and infections) were similar between vaccination groups, although slightly more unvaccinated patients (6 [1.4%]) had pulmonary fibrosis than vaccinated patients (0%; Table 2).

Overall, 32 COVID-19 deaths were reported, with a numerically smaller proportion of COVID-19 deaths seen in vaccinated (n = 8/372 [2.2%]) vs unvaccinated (n = 24/427 [5.6%]) patients. Adjusting for switching from unvaccinated to vaccinated status during the observation period, the Cox model with the time-dependent covariate for vaccination status demonstrated consistent results, with a 37% longer time to COVID-19 death seen in vaccinated vs unvaccinated patients (hazard ratio: 0.63; 95% confidence interval, 0.26-1.38). Bulky nodal disease (≥5 or ≥10 cm), higher body mass index (>30 kg/m²), and B symptoms at baseline were more common in patients who died from COVID-19, regardless of vaccination status (Table 1), suggesting baseline weight and disease burden also have a potential role in COVID-19 deaths.

Multiple risk factors for poor COVID-19 vaccine response have been identified in patients treated for CLL, including therapy with Bruton tyrosine kinase inhibitors, B-cell lymphoma 2 protein inhibitors, anti-CD20 agents, or any active therapy in general.¹ Despite patients being treated with these agents in this study, our analyses, although limited and not statistically significant, suggest that COVID-19 vaccination may still confer some protection based on the trend of longer survival time seen in vaccinated patients who had died because of COVID-19 infection. An analysis of case fatality rate was not conducted in the current analysis because the true incidence of infection could not be assessed based on symptomatic cases alone. A previously published analysis of the ongoing phase 3b ASSURE trial examined COVID-19–related outcomes in a CLL population treated with acalabrutinib monotherapy; however, data regarding COVID-19 vaccination were limited.² That analysis reported a COVID-19 fatality rate of ∼30%, which was comparable with that reported in retrospective studies of COVID-19 infection in patients with CLL published during the same time frame.^2-5 A subsequent, more comprehensive COVID-19−focused analysis after the unblinding of the AMPLIFY study will further inform the impact of vaccination status on COVID-19 outcomes in patients receiving specific treatment regimens.

In conclusion, although the study was not prospectively designed to examine differences in COVID-19 deaths between vaccinated and unvaccinated groups, and the study remains blinded to the CLL treatment received, our limited analysis suggests prior vaccination may still provide a protective effect in patients with CLL on therapy who become infected with COVID-19. However, the death rates in the vaccinated and unvaccinated are low, which may reflect that the management of COVID-19 has improved over time. Overall, these results suggest that all patients with CLL should receive COVID-19 vaccination even while receiving therapies believed to inhibit vaccine response. Geographic location may result in disparity in vaccine access but prioritization of an elderly and immune-compromised population (such as those with CLL) should continue to be enforced even as the pandemic has waned.

Acknowledgment: Medical writing support was provided by Robert J. Schoen of Peloton Advantage, LLC, an OPEN Health company, and funded by AstraZeneca. This study was sponsored by AstraZeneca.

Contribution: K.M. was responsible for project/data management and statistical analyses; and all authors were responsible for data analysis/interpretation, critical revision and review of the manuscript, and approval of the final draft for submission.

Conflict-of-interest disclosure: J.R.B. reports consultancy for Pfizer, Numab Therapeutics, Acerta/AstraZeneca, Grifols Worldwide Operations, Loxo/Lilly, Kite, iOnctura, Hutchmed, Pharmacyclics, Merck, AbbVie, Genentech/Roche, BeiGene, and Alloplex Biotherapeutics, and research funding from MEI Pharma, SecuraBio, Loxo/Lilly, iOnctura, Gilead, BeiGene, and TG Therapeutics. S.R. and K.M. report employment with AstraZeneca. J.F.S. reports research funding from AbbVie, Celgene, Janssen, and Roche; consultant or advisory roles with AbbVie, AstraZeneca, Celgene, Genentech, Genor Bio, Gilead, Janssen, MorphoSys, Roche, Sunesis, and TG Therapeutics; and other roles with AbbVie, Celgene, Roche, and TG Therapeutics.

Correspondence: Jennifer R. Brown, Chronic Lymphocytic Leukemia Center, Dana-Farber Cancer Institute, 450 Brookline Ave, Boston, MA 02215; email: jennifer_brown@dfci.harvard.edu.