The impact of CMV reactivation on mortality after chimeric antigen receptor T-cell therapy Open Access

TO THE EDITOR:

Cytomegalovirus (CMV) is increasingly recognized as an important pathogen in recipients of chimeric antigen receptor (CAR) T-cell therapy (CARTx), yet its clinical impact remains largely unexplored. CMV reactivation has been reported in 30% to 60% (clinically significant CMV [csCMV] requiring therapy in 10%-20%) of CMV-seropositive CARTx recipients, typically occurring within 2 to 6 weeks after infusion.^1-5 We previously reported results from a prospective study of 72 CMV-seropositive CD19- and B-cell maturation antigen (BCMA)–CARTx recipients with weekly testing for up to 12 weeks. We found a cumulative incidence of CMV reactivation of 27% and identified corticosteroid use for >3 days and BCMA-CARTx as key risk factors.⁵ 

Although CMV viremia is common, CMV end-organ disease is relatively infrequent, perhaps owing to the use of preemptive treatment.^2,6-9 Nonetheless, CMV has been implicated in “indirect effects,” including increased mortality^10-12 and a higher overall infection risk¹³ in hematopoietic cell transplant (HCT) recipients. The impact of CMV on mortality in HCT is mitigated by prophylaxis with letermovir, drawing a mechanistic link between CMV reactivation and increased risk of death.^11,12 In CARTx recipients, 2 studies reported an association between CMV reactivation and increased mortality within 1 year of CARTx.^4,9 An improved understanding of the clinical impact of CMV is needed to guide preventive strategies in the CARTx setting. In this study, we expand upon our previously reported cohort of CARTx recipients with weekly CMV testing to assess the impact of CMV reactivation on overall mortality (OM) by 1 year after CARTx in 84 prospectively monitored participants.⁵ 

We prospectively enrolled sequential CMV-seropositive adults planning to receive commercial or investigational CARTx targeting CD19, CD20, or BCMA for non-Hodgkin lymphoma, B-cell acute lymphoblastic leukemia, chronic lymphocytic leukemia, or multiple myeloma from August 2021 to March 2023 at Fred Hutchinson Cancer Center. We obtained plasma once before lymphodepleting chemotherapy and weekly after CARTx for up to 12 weeks. We tested plasma for CMV using a laboratory-developed quantitative polymerase chain reaction (PCR) assay with a lower limit of quantitation of 50 IU/mL.¹⁴ Independent of the current study, weekly CMV monitoring was recommended according to institutional guidelines in high-risk patients, defined as receiving >3 days of corticosteroids or >1 dose of tocilizumab and/or other anticytokine therapy (supplemental Methods). If CMV testing was performed for clinical purposes, the Abbott RealTime PCR was used with a limit of quantitation of 50 IU/mL. When results were available from both research and clinical samples within 24 hours, the higher value was included in the analysis. Results from research testing were not available to clinical teams and did not influence management. Preemptive treatment was recommended for CMV viral loads ≥150 IU/mL per institutional guidelines, and csCMV was defined as viremia above this treatment threshold of ≥150 IU/mL and requiring treatment.¹⁵ Cytokine release syndrome and immune effector cell–associated neurotoxicity syndrome were graded according to the American Society for Transplantation and Cellular Therapy Consensus Guidelines.¹⁶ The study was approved by the Fred Hutchinson Cancer Center Institutional Review Board. All participants provided a written informed consent.

We conducted landmark analyses to estimate the probability of 1-year overall survival (OS) with the Kaplan-Meier method starting 6 weeks after CARTx. The cadence of CMV reactivation informed our selection of 6 weeks for landmark analysis given that all CMV reactivation episodes occur within this time window.⁵ We used adjusted Cox proportional hazards regression to evaluate risk factors for OM. Patients were censored at the time of an HCT or second CARTx. Tests were 2 sided with a significance level of 0.05. Analyses were performed using SAS version (9.4 TS1M3) (SAS Institute, Cary, NC).

We followed 84 patients who received 89 CAR T-cell infusions including 67 CD19- or CD20- (78%) and 19 BCMA-targeted CAR T-cell products (22%) (supplemental Table 1). CMV reactivation occurred in 22 of 89 infusions (25%) with viral loads ≥150 IU/mL in 8 (9%), resulting in a cumulative incidence of 24.7% (95% confidence interval [CI], 16.3-34.1) and 9.0% (95% CI, 4.2-16.1) for detection at any level and at ≥150 IU/mL, respectively, within 12 weeks after CARTx (supplemental Figure 1). The median peak viral load was 124 IU/mL (interquartile range, 61-276), and the median time to first detection was 23 days (range, 11-47) (supplemental Table 2). No participants were diagnosed as having CMV end-organ disease. Preemptive treatment was administered in 5 participants with viral loads ≥150 IU/mL (ie, csCMV) (supplemental Figure 2).

One participant was lost to follow-up and 2 participants died before week 6 and were not included in the analysis (1 of them died within 6 weeks from their second infusion). Between 6 weeks and 1 year after CARTx, 21 participants died; 16 deaths were caused by relapse, whereas nonrelapse deaths occurred in 5 patients (infection [n = 2], organ failure [n = 2], brain herniation [n = 1]). OS was lower among those with previous CMV reactivation, particularly among those with CMV reactivation ≥150 IU/mL, although this did not reach statistical significance (Figure 1). Results of a univariate Cox model are presented in supplemental Table 3. In models adjusted for key confounders of CAR T-cell target and cytokine release syndrome and/or immune effector cell–associated neurotoxicity syndrome grade ≥2, CMV showed a trend toward increased OM, particularly in patients with CMV reactivation ≥150 IU/mL (adjusted hazard ratio, 2.42; 95% CI, 0.69-8.47; P = .17; Figure 2), but this did not reach statistical significance. Similar results were obtained in models adjusting for corticosteroid use for >3 days (adjusted hazard ratio, 1.95; 95% CI, 0.67-5.72; P = .22). Differences did not reach statistical significance, likely related to the relatively small sample size and number of events. Analyses were not conducted for nonrelapse mortality (NRM) owing to the low number of events precluding informative conclusions.

In this prospective study of weekly CMV PCR testing for up to 12 weeks after CARTx in CMV-seropositive adults, we found a trend between CMV reactivation ≥150 IU/mL and increased OM, but this did not reach statistical significance. These results are concordant with other findings to date, but the limited sample size precluded definitive conclusions and analyses for NRM, highlighting the need for larger studies to explore the impact of CMV on long-term outcomes.

In CARTx recipients, infection is the leading cause of NRM, accounting for more than half of all nonrelapse deaths through 5 years after treatment.^17,18 Although CMV reactivation is common after CARTx, data on its impact on survival are only now emerging. Our findings align with 2 recent studies. Khawaja et al⁹ retrospectively analyzed 51 patients with CMV detection in a cohort of 230 CD19 CARTx recipients. csCMV infection, defined as a viral load ≥500 IU/mL requiring antiviral therapy or CMV end-organ disease, was observed in 22 patients and independently associated with increased NRM at 1 year after CARTx. OM was also higher in patients with csCMV in unadjusted analysis. In another retrospective study of 95 CMV-seropositive CARTx recipients without systematic CMV testing, Lin et al⁴ found that, among patients with CMV reactivation by day 28 after infusion, 1-year OM was 57% compared with 23% (P = .001) in those without CMV reactivation. Although most deaths in our cohort were caused by relapse, our study lacked the power to explore mechanisms of CMV reactivation potentially driving mortality. Indirect effects of CMV may contribute to poor immune reconstitution and increased infection risk, which should be studied in larger cohorts. Finally, viral load thresholds at which to begin preemptive therapy in the CARTx setting are unclear⁶ and should be further evaluated in larger studies. In conclusion, we observed a potential association between CMV reactivation and decreased OS. However, these findings did not reach statistical significance, possibly owing to the relatively small sample size and number of events. Our results are consistent with the established effect of CMV reactivation on increasing the risk of death after allogeneic HCT and highlight the need to better understand the role of CMV in CARTx outcomes.^11,12 Such findings may support the need for screening and preemptive or prophylactic treatment in this population. Larger studies are needed to confirm the impact of CMV on outcomes in this patient population.

Acknowledgments: The authors thank Sarah S. Ibrahimi, Jessica B. Hecht, Alythia Vo, and Winnie L. Liu (Fred Hutchinson Cancer Center) for data collection; Haiying Zhu and Tracy Santo (University of Washington) for performing laboratory testing of samples; and Ryan S. Basom and Chris Davis (Fred Hutchinson Cancer Center) for support with data extraction.

This work was supported by grants from the Swiss National Science Foundation (P500PM_202961; E.K.), SICPA Foundation (E.K.), Merck (J.A.H.), and a Cancer Center Support grant from the National Institutes of Health to Fred Hutchinson Cancer Center (P30 CA015704-47).

Contribution: E.K. and J.A.H. designed the study; M.J.B. contributed to the study design; E.K., P.F., J.A.H., and H.X. analyzed and interpreted the data and created the figures; E.K., P.F., M.K.S., C.C., and J.A.H. enrolled participants and collected samples and data; T.L.S.-A., A.C.P.-O., and K.R.J. supervised the laboratory work; E.K. and J.A.H. prepared the first draft of the manuscript; and all authors contributed to the writing and revision of the manuscript and approved the final version.

Conflict-of-interest disclosure: E.K. reports advisory board participation with and received support for conferences from Merck. J.G. has served as ad hoc consultant and received honoraria from Sobi, Legend Biotech, Janssen, Kite Pharma, and MorphoSys; received research funding from Sobi, Juno Therapeutics (a Bristol Myers Squibb company), Celgene (a Bristol Myers Squibb company), and Angiocrine Bioscience; and participated in the independent data review committee for Century Therapeutics. M.S. has served as a consultant and participated in advisory boards, steering committees, or data safety monitoring committees for AbbVie, Genentech, AstraZeneca, Genmab, Janssen, BeiGene, Bristol Myers Squibb, MorphoSys/Incyte, Kite Pharma, Eli Lilly, Fate Therapeutics, Nurix, and Merck; received research funding from Mustang Bio, Genentech, AbbVie, BeiGene, AstraZeneca, Genmab, MorphoSys/Incyte, and Vincerx; has stock options in Koi Biotherapeutics; and reports employment with Bristol Myers Squibb (spouse). D.J.G. has served as an advisor and received research funding and royalties from Juno Therapeutics, a Bristol Myers Squibb company; served as an advisor and received research funding from Seattle Genetics, Janssen Biotech, and Bristol Myers Squibb; served as an advisor for GlaxoSmithKline, Celgene, Ensoma, and Legend Biotech; and received research funding from SpringWorks Therapeutics, Sanofi, and Cellectar Biosciences. M.J.B. has served as a consultant for Merck, Takeda, Symbio, Assembly Bio, AlloVir, and Moderna; received research funding from Merck and Moderna; and served as a consultant and had the option to acquire stock for Evrys Bio. J.A.H. has served as a consultant for Moderna, AlloVir, Gilead, Takeda, CSL Behring, and Karius, and received research funding from AlloVir, Gilead, Merck, and Takeda. The remaining authors declare no competing financial interests.

Correspondence: Eleftheria Kampouri, Infectious Diseases Service, Lausanne University Hospital and University of Lausanne, BH09-788, Bugnon 46, CH-1011 Lausanne, Switzerland; email: eleftheria-evdokia.kampouri@chuv.ch.