We thank Dr Ali and colleagues for their thoughtful comments and for their interest in our study examining cardiovascular events (CVEs) during T cell engager therapy.
We acknowledge that differential surveillance intensity is an important methodological consideration in observational studies of cardiotoxicity. In response to this concern, we examined the frequency of cardiac testing according to cytokine release syndrome (CRS) and/or immune effector cell–associated neurotoxicity syndrome (ICANS) severity in our data set. On-treatment echocardiography was performed in a total of 291 patients: 50 of 88 patients (56.8%) with grade ≥ 2 CRS or ICANS and 241 of 479 patients (50.3%) with grade < 2 CRS or ICANS (P = 0.279), suggesting no significant difference in echocardiographic surveillance between groups. Troponin measurements were obtained more frequently among patients with grade ≥ 2 CRS or ICANS than among those with less severe toxicity (13 of 88 [14.8%] vs 33/479 [6.9%]; P = 0.019). However, troponin testing during treatment was infrequent overall (46 total measurements across the cohort) and was not performed systematically, limiting interpretation and raising concern for ascertainment bias. For these reasons, these data were not included in the original paper.
To further evaluate the potential impact of surveillance bias, we performed an additional sensitivity analysis excluding left ventricular dysfunction (LVD) from the composite CVE endpoint, as this outcome may be more susceptible to detection bias related to differences in cardiac imaging intensity. In a time-dependent Cox model, grade ≥ 2 CRS or ICANS remained associated with subsequent CVEs in univariable (HR: 2.28; 95% CI: 1.11-4.67; P = 0.025) and multivariable (HR: 2.01; 95% CI: 0.97-4.18; P = 0.061) analyses. Although statistical significance was attenuated after adjustment, the effect estimate remained similar, likely reflecting reduced power and precision after exclusion of LVD events. These results suggest that the association between grade ≥ 2 CRS or ICANS and CVEs was not driven solely by surveillance-detected LVD.
We appreciate the opportunity to clarify the event adjudication process. All CVEs were independently reviewed by 2 investigators (O.I.B.Z. and A.N.) using prespecified endpoint definitions. Adjudication incorporated review of clinical documentation, laboratory data, electrocardiograms, imaging studies, and hospitalization records. Any discrepancies were resolved by consensus.
Once again, we thank Dr Ali and colleagues for highlighting these important methodological considerations. Although our findings should be interpreted within the limitations inherent to retrospective observational studies, the additional analyses support the robustness of the observed association between CRS or ICANS and cardiovascular risk. Future prospective studies with standardized cardiovascular surveillance protocols and systematic biomarker and imaging assessment are needed to more precisely define the relationship between immune-mediated toxicities and cardiovascular complications during T cell engager therapy.
Footnotes
Dr Nohria has received research support from Bristol Myers Squibb; has received consulting fees from AstraZeneca, Teledoc, and Takeda Oncology; and has received speaker fees from MLI Education. Dr Itzhaki Ben Zadok has received advisory board honoraria from Pfizer, Neopharm, and Medison.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.