Prognostic value of pre-infusion tumor growth rate for patients with lymphoma receiving chimeric antigen receptor T-cell therapy

Abstract

Background aimsChimeric antigen receptor T-cell therapy (CART) prolongs survival for patients with refractory or relapsed lymphoma, yet its efficacy is affected by the tumor burden. The relevance of tumor kinetics before infusion is unknown. We aimed to study the prognostic value of the pre-infusion tumor growth rate (TGRpre-BL) for progression-free (PFS) and overall survival (OS). MethodsConsecutive patients with available pre-baseline (pre-BL) and baseline (BL) computed tomography or positron emission tomography/computed tomography scan before CART were included. TGR was determined as change of Lugano criteria-based tumor burden between pre-BL, BL and follow-up examinations (FU) in relation to days between imaging exams. Overall response rate (ORR), depth or response (DoR) and PFS were determined based on Lugano criteria. Multivariate regression analysis studied association of TGR with ORR and DoR. Proportional Cox regression analysis studied association of TGR with PFS and OS. ResultsIn total, 62 patients met the inclusion criteria. The median TGRpre-BL was 7.5 mm2/d (interquartile range –14.6 mm2/d to 48.7 mm2/d); TGRpre-BL was positive (TGRpre-BL POS) in 58% of patients and negative (TGRpre-BL NEG, indicating tumor shrinkage) in 42% of patients. Patients who were TGRpre-BL POS had a 90-day (FU2) ORR of 62%, a DoR of –86% and a median PFS of 124 days. Patients who were TGRpre-BL NEG had a 90-day ORR of 44%, DoR of –47% and a median PFS of 105 days. ORR and DoR were not associated with slower TGR (P = 0.751, P = 0.198). Patients with an increase of TGR from pre-BL over BL to 30-day FU (FU1) ≥100% (TGRpre-BL-to-FU1≥100%) showed a significant association with shorter median PFS (31 days versus 343 days, P = 0.002) and shorter median OS after CART (93 days versus not reached, P < 0.001), compared with patients with TGRpre-BL-to-FU1<100%. ConclusionsIn the context of CART, differences in pre-infusion tumor kinetics showed minor differences in ORR, DoR, PFS and OS, whereas the change of the TGR from pre-BL to 30-day FU significantly stratified PFS and OS. In this patient population of refractory or relapsed lymphomas, TGR is readily available based on pre-BL imaging, and its change throughout CART should be explored as a potential novel imaging biomarker of early response.