While chimeric antigen receptor (CAR) T cell therapy is an essential therapeutic option for prolonging survival in patients with relapsed or refractory non-Hodgkin lymphomas (NHL), profound immune suppression and predicting the likelihood of treatment failure remain significant challenges.
Physician-scientists at Memorial Sloan Kettering Cancer Center (MSK) recently published two papers that advance the science of addressing these critical issues.
“MSK played a pioneering role in the development and is one of the leading centers nationwide authorized to administer CAR T cell therapy,” said bone marrow transplant specialist and cellular therapist Miguel-Angel Perales, MD, Chief of the Adult Bone Marrow Transplant Service at MSK. “We remain dedicated to conducting multidisciplinary research to improve patient outcomes.”
Immune Reconstitution for Patients with LBCL Varies by CAR T Product
A study by MSK researchers has revealed that immune reconstitution post CAR T cell therapy in this patient population varies by CAR T product, and that early natural killer cell recovery is a key predictor of survival. The paper was published in December 2024 online ahead of print in Blood Cancer Discovery. (1)
The findings are important since patients with relapsed or refractory large B-cell lymphoma (LCBL) receiving CAR T cell therapy experience profound immune vulnerability that can persist for an extended and undefined period after infusion. (2) (3) (4) (5) (6) Infections account for more than 50% of all non-relapse deaths and are the leading cause of morbidity and mortality after CAR T cell therapy, according to a recent systematic review and meta-analysis led by MSK investigators. (7)
“Our study is the first to provide comprehensive insights into immune reconstitution post-CAR T cell therapy in patients with LBCL in the largest cohort to date,” said bone marrow transplant specialist and cellular therapist Roni Shouval, MD, PhD, Director of the Precision Cellular Therapy Laboratory at MSK and co-senior author of the paper together with Dr. Perales. “It is also the first to highlight significant differences in immune cell recovery between the three available CAR T cell products for patients with lymphoma.”
For their study, Drs. Shouval, Perales, and colleagues profiled the dynamics of immune reconstitution and its influence on outcomes for 263 patients with relapsed or refractory LBCL treated with CAR T cell therapy. (1)
Previous studies have mainly focused on limited immune cell subsets, such as CD4+ and CD8+ T cells and CD3-19+ B cells, in patients treated with axicabtagene ciloleucel (axi-cel) (3) (4) or tisagenlecleucel (tisa-cel). (8) Patterns of immune reconstitution with lisocabtagene maraleucel (liso-cel), as well as the dynamics of more specialized immune cells, including memory and effector subcategories of CD4+ and CD8+ T cell subsets and natural killer cells, have not previously been reported.
In this study, 40% of patients received axi-cel, 30% received liso-cel, and 25% received tisa-cel.
Overall, patients remained persistently immunosuppressed after infusion, with 48.1% having CD4+ T cell counts of less than 200/µL and median CD3-19+ B cell counts of zero through one-year post-CAR T. (1)
Patterns of immune reconstitution varied by CAR T product, with the fastest CD4+ T cell recovery observed for tisa-cel, driven mainly by a more rapid recovery of CD4+CCR7-45RA-effector memory cells. Specifically, by 30 days after CAR T infusion, the median IQR value of CD3+ T cells among patients treated with tisa-cel was 710 (447-1070) compared to 309 (151-618) and 467 (273-688) for those who received axi-cel and liso-cel, respectively. (1)
Natural killer cell recovery was significantly associated with a more favorable progression-free survival (PFS) (hazard ratio = 0.647) and overall survival (OS) (hazard ratio = 0.637) and negatively correlated with inflammatory markers measured at the time of infusion. Higher natural killer cell counts at day 30 were associated with improved OS and PFS, and each log-fold increase in natural killer cell count was associated with a 36% reduced risk of death and a 35% lower risk of disease progression. A similar pattern was observed through landmark days 55 for OS and 65 for PFS: each log-fold increase in natural killer cell count was associated with a 48% reduced risk of death. (1)
The researchers’ analysis also revealed that higher levels of C-reactive protein, ferritin, tumor necrosis factor alpha, and interleukin 10 were significantly associated with delayed natural killer cell recovery. By contrast, the number of pre-apheresis treatment lines, pre-CAR T disease response, and age had no bearing on natural killer cell recovery. (1)
“Our results suggest the choice of CAR T product is a major driver of differences in immune recovery dynamics after infusion, with some indirect contribution from lymphodepletion intensity,” said Dr. Shouval. “They also highlight the need for tailored post-therapy management strategies and future research into optimizing immune recovery in this patient population.”
The study was supported by the following grants: National Institutes of Health (NIH) P01 CA23766 and NIH/National Cancer Institute (NCI) MSK Support Grant P30 CA008748; NIH/NCI K08-CA282987 and the Long Island Sound Chapter of Swim Across America (Dr. Shouval); FEHH/Gilead 2022 grant (Dr. Alejandro Luna de Abia); and the Alfonso Martin Escudero grant 2023 (Dr. Magdalena Corona de Lapuerta). Access disclosures for Dr. Shouval and Dr. Perales.
A Single Blood Test Can Identify Treatment Failure Risk Pre-CAR T in NHL
Dr. Shouval, Dr. Perales and MSK bone marrow transplant specialist Sandeep Raj, MD, co-led an international team of investigators that developed and validated a new model for identifying patients with a higher risk of disease relapse and mortality before receiving CAR T cell therapy. The model is based on an inflammatory signature found in a single blood test.
Called InflaMix (INFLAmation MIXture Model), the quantitative tool integrates 14 pre-CAR T infusion markers of inflammation and end-organ function into an inflammatory signature that strongly correlates with and predicts poor disease response and survival after CD19-directed CAR T therapy in patients with NHL. Dr. Raj was lead author of the paper describing the innovation and research findings published in Nature Medicine on April 1, 2025. (9)
The research team developed the model using a cohort of 149 NHL patients at MSK and validated it using three independent cohorts of 688 patients with NHL from diverse global treatment centers. (9)
InflaMix consistently and reproducibly identified patients with a higher risk of disease relapse and mortality. Assignment to the inflammatory cluster was associated with almost five-fold greater odds of not achieving a complete response by day 100 (odds ratio 4.76), a three-fold reduced PFS (hazard ratio 2.98), and a three-fold lower OS (hazard ratio 2.90) compared to assignment to the non-inflammatory cluster. Interestingly, cluster assignment was not associated with increased odds of cytokine release syndrome or immune-effector cell-associated neurotoxicity syndrome (ICANS). (9)
Notably, patients assigned to the inflammatory cluster at pre-apheresis who moved to the non-inflammatory cluster by pre-infusion showed significantly better survival and disease response rates than those who remained in the inflammatory cluster, even after adjusting for clinical risk factors such as bridging therapy. The investigators observed a similar survival improvement among patients who moved to a non-inflammatory profile between pre-lymphodepletion and pre-infusion. By contrast, patients with a non-inflammatory signature at earlier times who shifted to an inflammatory signature before infusion experienced worse outcomes. (9)
“InflaMix provided additional predictive value beyond established prognostic markers, such as tumor burden, and also performed robustly in cases with missing data, remaining accurate when considering only six readily available laboratory measures (albumin, Hgb, AST, ALP, CRP, and LDH), supporting its value as a prognostication tool in global settings with less robust blood testing capabilities,” said Dr. Shouval. “Our study findings suggest InflaMix will provide essential information for clinical decision making, especially as we found CAR T failure risk can change over time.”
Given that InflaMix was derived from a retrospective, single cohort at MSK, the study authors noted a need for more research in two areas: prospective and mechanistic studies to evaluate the tool’s real-time utility in risk stratification, and studies investigating the addition of clinical, tumor genomic, and radiomic features to the prognostic model.
“InflaMix may also have broader utility for evaluating the likelihood of treatment failure with other T cell immunotherapies, such as immune checkpoint blockade and stem cell transplant,” Dr. Shouval said.
The research team has posted an online calculator on GitHub for applying InflaMix at the bedside. Clinicians enter as many of the 14 laboratory results as possible, with priority to the six measurements from the limited panel.
Refer to the paper to review the range of funding support received by the study investigators. Access disclosures for Dr. Shouval, Dr. Perales, and Dr. Raj.