Early clinical and radiological predictors of relapse following partial response to chimeric antigen receptor T-cell therapy in patients with aggressive B-cell Non-Hodgkins lymphoma
OA Version
Citation
Abstract
INTRODUCTION: Chimeric antigen receptor (CAR) T-cell therapy is a relatively new cancer immunotherapy that is offering remarkable benefits for patients in the third line setting for relapsed/refractory aggressive large B-cell non-Hodgkin’s lymphoma (NHL). At one-month restaging post CAR T-cell therapy, approximately 30% of patients are found to be in a partial response (PR) or to have stable disease (SD) to therapy, and these patients diverge into 30% who deepen into a complete response (CR) and the remaining 70% who ultimately progress (PD). To date there has been no way to predict who will enter remission and who will progress. Our study seeks to better understand these diverging patient groups and identify early predictive markers of progression. Patients projected to progress could then be treated with adjuvant therapies that could both kill residual tumor cells and improve residual CAR T-cell activation in order to increase the probability of a deepening of response.
METHODS: In this retrospective analysis, we evaluated 259 patients who were treated with anti-CD19 directed CAR T-cell therapy for r/r B-NHL at the Dana-Farber Cancer Institute between December 2017 and June 2021. Patients with a PR or SD at 1 month were further classified into two cohorts according to their subsequent response conversion. Those that are known to have progressed from their PR/SD were included in the PD cohort, and the rest of the PR/SD patients were included in the non-PD cohort. Early clinical and radiological parameters up to the 1 month restaging scan were included in our analysis.
RESULTS: The overall response rate for our patient sample was 79% with a best response of CR in 64% of patients. At 1 month restaging 52% of patients were in CR, 28% in PR, 2% SD, and 18% PD. Median follow-up was 9 months. 41% of patients deepened their response from PR/SD into a CR, 90% of whom remain in a CR to last follow-up, and 42% were known to have progressed. The PD cohort had poorer progression-free survival (3 months vs. 11 months) and overall survival (11 months vs. not yet reached) than the non- PD cohort. Parameters significantly correlated to progression included poor performance status; high CRP and low Hgb at Day -30; high CRP at lymphodepletion; need for early admission; low Hgb, high LDH, and high ferritin on Day 0; CRS Lee grade 2+ and low albumin at CRS onset; high ferritin at time of CRS treatment; and Deauville score of 5 on restaging scan. High MTV and TLG on pre- and post-infusion PET scans also trended with those who progressed.
CONCLUSIONS: The findings of our analysis that were correlated to progression are primarily related to poor lymphoma biology pre-therapy, residual lymphoma, and CRS onset kinetics. Notably, inflammatory markers and maximal CRS post-therapy were not significant. A primary difference between the two PR patient groups may thus relate to the tumor microenvironments pre-infusion with greater tumor bulk in the PD cohort possibly corresponding to tumor microenvironments that are more inflammatory and hostile to CAR T-cells. We therefore define a new prognostic scoring system, to be validated in future studies, that utilizes five early clinical parameters and establishes three risk groups with significantly different prognoses for PR/SD patients. Evaluation of post-infusion minimal residual disease, CD19 status, PD-L1 status, and other molecular markers could be useful in better understanding and differentiating the pathophysiologies characterizing these patient groups.