Clinical Impact of a Germline CD19 Variant on Treatment Outcome After CAR-T Cell Therapy in Relapsed/Refractory Mantle Cell Lymphoma

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study looked at how a patient's genetics might affect their outcome after a specific type of CAR-T cell therapy for mantle cell lymphoma. It's a single-center retrospective review, so the data is exploratory. But the thinking behind it makes sense. The authors are honest about the study's limits, and they don't overstate their conclusions. The main issues? The sample size is small, which means the statistics aren't very powerful. The survival test results don't agree with each other. And there's no adjustment for other factors that could be at play. The figures could also be cleaner. Fixing these things would make the paper a lot stronger.

1.  Survival Analysis Presentation

    The survival curves for progression-free and overall survival need some work. Standard practice is to show the number of patients still at risk at regular time points. You also want to see confidence intervals for each genetic group. Without these, it's hard to tell how reliable the differences really are.

2.  Statistical Methods

    The authors used two different statistical tests—the log-rank test and the Gehan–Breslow–Wilcoxon test. They need to explain why they used both. Was this planned from the start? Or was it a way to explore the data after the fact? Since the two tests gave different results, being upfront about this would help readers trust the analysis more.

3.  Confounders and Multivariable Analysis

If the sample size allows, the authors should try a multivariable analysis. This means looking at other factors that could influence the results—things like how much disease the patient had, how many treatments they'd tried before, what type of BTK inhibitor they used, and how severe their CAR-T side effects were. If that's not possible, they should at least talk about it as a limitation. Otherwise, it's hard to rule out that something else is driving the findings.

4. Making Sense of the Inverse Association

The Discussion needs to dig deeper into why the CD19 rs2904880 genotype links to survival in opposite ways for MCL versus DLBCL. Figure 2 gives us a helpful visual, but we need more explanation. How might differences in CAR-T construct design—like the FMC63-based brexu-cel compared to other products—play a role? What about target antigen density or the unique biology of MCL tumors? These are the kinds of details that could help explain this puzzling discrepancy.

5. What This Means for Clinical Practice

The authors suggest CD19 rs2904880 might one day help with risk stratification in MCL. To their credit, they're careful to call this a conceptual idea, not something ready for the clinic. Still, this section could be stronger with a brief outline of what evidence we'd actually need. Think about specific sample size thresholds, ways to integrate this with other biomarkers, and the types of prospective studies that would be necessary before this marker could move toward real-world use.

6. Figure 2: Labels and Captions Need Clarity

The idea of showing CD19 variants and how they bind differently to FMC63 is a smart one. But the figure legend and caption need revision. Readers need to know: are these binding affinity differences based on actual experiments, or are they hypothetical? Without that clarity, it's easy to overstate what the evidence actually shows.

7. Broadening the Conclusions

The conclusion is nicely concise and scientifically measured. But it could be even better by directly mentioning the inverse direction of this association compared to what we've seen in DLBCL. That's a clinically important detail, and it will shape how readers interpret the whole study.

Author Response

Cancers-4386733 reviewer 1

 

Comments and Suggestions for Authors

This study looked at how a patient's genetics might affect their outcome after a specific type of CAR-T cell therapy for mantle cell lymphoma. It's a single-center retrospective review, so the data is exploratory. But the thinking behind it makes sense. The authors are honest about the study's limits, and they don't overstate their conclusions. The main issues? The sample size is small, which means the statistics aren't very powerful. The survival test results don't agree with each other. And there's no adjustment for other factors that could be at play. The figures could also be cleaner. Fixing these things would make the paper a lot stronger.

Response: We thank the reviewer for the valuable comments. We have integrated all points in the revised manuscript and trust that the manuscript has been improved.

1.  Survival Analysis Presentation

    The survival curves for progression-free and overall survival need some work. Standard practice is to show the number of patients still at risk at regular time points. You also want to see confidence intervals for each genetic group. Without these, it's hard to tell how reliable the differences really are.

Response: The revised survival curves include confidence intervals and number of patients at risk.

1.  Statistical Methods

    The authors used two different statistical tests—the log-rank test and the Gehan–Breslow–Wilcoxon test. They need to explain why they used both. Was this planned from the start? Or was it a way to explore the data after the fact? Since the two tests gave different results, being upfront about this would help readers trust the analysis more.

Response: PFS and OS curves were estimated using the Kaplan-Meier method and curves were com-pared using the log rank (Mantel-Cox) test as the primary method, as it is the standard approach and equally weights events over the entire follow-up. In addition, we calculated the log rank test for trend and the Gehan-Breslow-Wilcoxon test as pre-specified secondary, exploratory analyses because they apply different weighting schemes to events over time. The explanation has been added to chapter 2.4. Statistical analysis.

1.  Confounders and Multivariable Analysis

If the sample size allows, the authors should try a multivariable analysis. This means looking at other factors that could influence the results—things like how much disease the patient had, how many treatments they'd tried before, what type of BTK inhibitor they used, and how severe their CAR-T side effects were. If that's not possible, they should at least talk about it as a limitation. Otherwise, it's hard to rule out that something else is driving the findings.

Response: Treatment outcomes were evaluated in multivariate analysis and included parameters of possible impact on clinical outcome including LDH levels, prior therapy lines and CRS. The multivariate analysis supported the trend observed in the univariate analysis for the CD19 major allele V174 to be a prognostic indicator for overall survival with a HR 0.08 at a p-value of 0.04 (Table 5). High LDH levels were also predictive for outcomes with HR 63 at p-value of 0.01, higher number of prior therapy lines with HR 5.8 at p-value 0.15, and CRS with HR 2.45 at p-value 0.48). This paragraph has been added to the results section.

 

1. Making Sense of the Inverse Association

The Discussion needs to dig deeper into why the CD19 rs2904880 genotype links to survival in opposite ways for MCL versus DLBCL. Figure 2 gives us a helpful visual, but we need more explanation. How might differences in CAR-T construct design—like the FMC63-based brexu-cel compared to other products—play a role? What about target antigen density or the unique biology of MCL tumors? These are the kinds of details that could help explain this puzzling discrepancy.

Response:

Comparative analyses of outcomes with the same CD19 CAR products across multiple indications consistently show that efficacy, primary resistance, CD19-negative relapse and acute toxicities vary according to the underlying disease, supporting the view that entity-specific biology significantly shapes CAR-T cell performance [29]. Available clinical data further suggest that CD19-directed therapies can retain substantial activity even in cases with dim or undetectable CD19 expression and baseline CD19 status has non consistently predicted response to axi-cel, tisa-cel or brexu-cel [32]. Thus, it seems unlikely that the opposite genotype-outcome associations observed in DLBCL and MCL are driven solely by gross differences in CD19 antigen density. Since both our cohort and the PPM1D analysis [33] focus on patients with MCL, it is plausible that unique MCL-specific tumor biology – including its characteristic CCND1-driven cell-cycle deregulation, frequent DNA damage response pathway alterations and distinctive patterns of immune evasion involving checkpoint upregulation and microenvironmental reprogramming [34]– may influence how rs2904880-related changes in the CD19 epitope and product-specific CAR architectures translate into clinical benefit, potentially contributing to the observed opposite survival effects of the germline variant in MCL versus DLBCL.

Structural analyses of CD19-antibody complexes have shown that clinically used binders such as FMC63 and alternative CD19 binders can differ markedly in their monovalent affinity for CD19. These differences translate into distinct functional behavior of the corresponding CD19 CAR constructs. Specific CD19 variants such as Arg163Leu and Leu174Val, as well as deletions affecting exons 1-3, have been mapped to key interaction sites and linked to impaired binding, loss of cytotoxicity and clinical immune escape with relapse under FMC63-based CD19 CAR-T cell therapy, demonstrating that even subtle antigen variants can substantially alter CAR engagement [38–40]. These observations support the concept that both CAR design and CD19 sequence variation can generate variant-specific affinity differences, supporting the plausibility of our working model that the germline polymorphism rs2904880 may modulate the interaction of individual CD19 CAR products.

These paragraphs have been added to the discussion.

1. What This Means for Clinical Practice

The authors suggest CD19 rs2904880 might one day help with risk stratification in MCL. To their credit, they're careful to call this a conceptual idea, not something ready for the clinic. Still, this section could be stronger with a brief outline of what evidence we'd actually need. Think about specific sample size thresholds, ways to integrate this with other biomarkers, and the types of prospective studies that would be necessary before this marker could move toward real-world use.

Response: Future implementation would require validation in large, prospectively collected multi-center cohorts, ideally including several hundred brexu-cel treated MCL patients and integration of CD19 rs2904880 into multivariable models alongside established clinical and molecular risk factors. Only if the polymorphism provides reproducible, independent prognostic information and improves the performance of such composite risk scores would its use in routine risk stratification be justified. This paragraph has been added to the discussion.

1. Figure 2: Labels and Captions Need Clarity

The idea of showing CD19 variants and how they bind differently to FMC63 is a smart one. But the figure legend and caption need revision. Readers need to know: are these binding affinity differences based on actual experiments, or are they hypothetical? Without that clarity, it's easy to overstate what the evidence actually shows.

Response: The differential binding affinities are hypothetical. This sentence has been added to the figure legend.

1. Broadening the Conclusions

The conclusion is nicely concise and scientifically measured. But it could be even better by directly mentioning the inverse direction of this association compared to what we've seen in DLBCL. That's a clinically important detail, and it will shape how readers interpret the whole study.

Response: Notably, the direction of the association between CD19 rs2904880 and outcome in our MCL cohort appears inverse to prior reports in DLBCL and also varies across different CD19 CAR-T cell products, indicating that the impact of this polymorphism is highly con-text dependent. This sentence has been added to the conclusion.

 

Submission Date 03 June 2026

Date of this review 13 Jun 2026 01:56:11

 

Date of revision 25 Jun 2026.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

In this article Ruckstuhl et al present a    retrospective, single-center observational study on the effect of germline CD19 mutations on the outcomes of anti-CD19 CAR  T cell therapy in h R/R mantle cell lymphoma. The description of this germline single nucleotide polymorphism (SNP) in CD19, rs2904880,  influencing anti-CD19 CAR T cell activity has been previously reported in DLBCL, which somewhat reduces the novelty of this study. However, the results are interesting and the manuscript is well written and clear.  The discussion is good at stating the limited power of the study due to the size and is realistic about the meaning of the increase in overall survival and the different tests used. 

The introduction could have more background on what CAR T cell therapy is as it is not described. 

 

in section 2.4. Statistical analysis its says "Statistical significance was defined as a p value > 0.05" this should be changed to <

 

in figure 1 the statistical test used needs to be added to the figure legend. 

 

The idea that the different CAR constructs have have different affinities for the different variants is interesting, but more data to support this would be good.

 

Overall interesting and well written paper. 

Author Response

Cancers-4386733 reviewer 2

 

Comments and Suggestions for Authors

In this article Ruckstuhl et al present a   retrospective, single-center observational study on the effect of germline CD19 mutations on the outcomes of anti-CD19 CAR T cell therapy in h R/R mantle cell lymphoma. The description of this germline single nucleotide poly-morphism (SNP) in CD19, rs2904880, influencing anti-CD19 CAR T cell activity has been previously reported in DLBCL, which somewhat reduces the novelty of this study. However, the results are interesting and the manuscript is well written and clear.  The discussion is good at stating the limited power of the study due to the size and is realistic about the meaning of the increase in overall survival and the different tests used. 

Response: We thank the reviewer for the valuable comments. We have integrated all points in the revised manuscript and trust that the manuscript has been improved.

The introduction could have more background on what CAR T cell therapy is as it is not described. 

Response: Chimeric antigen receptor (CAR)-T cell therapy is an adoptive cellular immuno-therapy in which autologous T cells are collected, genetically modified to express a synthetic chimeric antigen receptor (CAR) targeting a defined surface antigen such as CD19, thereby enabling antigen-specific recognition and killing of malignant B cells. This sentence has been added to the introduction.  

in section 2.4. Statistical analysis its says "Statistical significance was defined as a p value > 0.05" this should be changed to <

Response: The definition of the p value has been changed to p value < 0.05.

in figure 1 the statistical test used needs to be added to the figure legend. 

Response: The statistical test has been added to the figure legend of Figure 1.

The idea that the different CAR constructs have have different affinities for the different variants is interesting, but more data to support this would be good.

Response: Structural analyses of CD19-antibody complexes have shown that clinically used binders such as FMC63 and alternative CD19 binders can differ markedly in their mono-valent affinity for CD19. These differences translate into distinct functional behavior of the corresponding CD19 CAR constructs. Specific CD19 variants such as Arg163Leu and Leu174Val, as well as deletions affecting exons 1-3, have been mapped to key interaction sites and linked to impaired binding, loss of cytotoxicity and clinical immune escape with relapse under FMC63-based CD19 CAR-T cell therapy, demonstrating that even subtle antigen variants can substantially alter CAR engagement [36–38]. These observations support the concept that both CAR design and CD19 sequence variation can generate variant-specific affinity differences, supporting the plausibility of our working model that the germline polymorphism rs2904880 may modulate the interaction of individual CD19 CAR products.

 Overall interesting and well written paper. 

Submission Date 03 June 2026

Date of this review 17 Jun 2026 18:35:10

Date of revision 25 June 2026

Author Response File: Author Response.docx