Erratum to: Is chimaeric antigen receptor T-cell therapy really superior to standard of care as second-line therapy for large B-cell lymphoma? (British Journal of Haematology, (2023), 200, 1, 10.1111/bjh.18505)

To the Editor, We read with interest the letter by Luis Carlos Saiz and colleagues and thank you for the opportunity to discuss some of their comments. Saiz et al. raised concerns about analysing data, which is still immature, especially data from an ongoing trial designed for a 3-year follow-up, which reported interim results for both primary and secondary outcomes, with a median follow-up of only 6.2 months published as an abstract,¹ while the median follow up of the other studies was 24 months. We agree with the authors regarding the short follow up, and we pointed out and elaborated on this obstacle throughout the manuscript and especially in the limitations section.² We understand that the authors are concerned by the discrepancy in the reported interim efficacy analysis percentages between the various publications and the protocol of the TRANSFORM study, especially in view of the recent publication.³ Inclusion of a study which was published as an abstract, presented in a leading international meeting and meets the pre-specified eligibility criteria of our systematic review is in accordance with the methodology of a systematic review and meta-analysis as specified by the Cochrane handbook, which enables inclusion of both published and unpublished data, while acknowledging the limitations of unpublished data.⁴ Furthermore, there are no major differences in the results between the abstract included in our meta-analysis and the full manuscript which appeared at a later stage in the Lancet.³ We agree with the authors regarding their comment on Figure 3 and herein we attach the accurate forest plot of EFS. However, the numbers cited in Results section are correct, and the EFS including all 3 trials was significantly improved in the CAR-T group compared to standard of care (SOC) (HR 0.57, 95% CI 0.49–0.68). 3 FIGURE (Figure presented.) Event free survival. As the authors mentioned, there is variability in the design of the trials, in particular relating to EFS definition. Stable disease as event was defined at different time points; ZUMA-7: day 150, Belinda: week 12, and Transform: week 9, and unlike ZUMA-7 and TRANSFORM, Belinda did not include initiation of a second salvage regimen within the first 12 weeks as an event.^5,6 This variability translated to high heterogeneity in the EFS analysis of I² = 94% (HR 0.57 in favour of CAR-T, 95% CI 0.49–0.68, I² = 94%, 865 patients, 3 trials). Therefore, in order to explore this high heterogeneity of EFS, we conducted a sensitivity analysis excluding the BELINDA trial resulting in I² = 0% (HR 0.39 in favour of CAR-T, 95% CI 0.31–0.48, I² = 0%, 2 trials). This sensitivity analysis confirms that the heterogeneity probably stems from the differences between Belinda and the other 2 trials. In summary, the role of CAR-T as second-line treatment of DLBCL, is one of the hottest topics in haemato-oncology today. Recent countless experts' debates and reviews have tried to address this very issue, emphasizing on the dissimilarities between the studies and implementation of their results outside clinical trials.⁷ For example, a recent Expert Panel Opinion from the American Society for Transplantation and Cellular Therapy (ASTCT) provided a comprehensive review of the 3 studies, emphasizing potential differences, and shared perspectives on what these results mean to clinical practice.⁸ The panel elaborated on several controversies such as: should a difference in PFS/EFS only without influencing OS change practice? do these trials justify a one-size-fits-all approach? what is the place of auto-HCT in view of these data? What about cost-effectiveness and access to care? At the same time, they pointed out that by now the FDA has approved Axi-cel and Liso-cel in second-line therapy based on the inclusion criteria of ZUMA-7 and TRANSFORM, respectively. Shortly after, the National Comprehensive Cancer Network (NCCN) guidelines were updated accordingly.⁹ So, in our opinion, the key question now is not whether CAR-T should be applied in the second line settings, but who are the patients that are most likely to benefit, since country-specific resource limitations like affordability, cost-effectiveness and manufacturing barriers should be taken into consideration. Our study which includes a pooled analysis of all 3 trials showing that CAR-T significantly improves both EFS and OS compared to SOC, supports a paradigm shift and hopefully will contribute to our understanding of the role and place of CAR-T in the algorithm of treating the patient with refractory/relapsed lymphoma.