Contributions
Abstract: EP497
Type: E-Poster Presentation
Session title: Aggressive Non-Hodgkin lymphoma - Clinical
Background
Anti-CD19 CAR T-cells, Tisagenleuclecel (Tisa-cel) and Axicabtagene ciloleucel (Axi-cel), have been approved for pts with relapsed or refractory aggressive large B-cell lymphomas (R/R LBCL) after 2 lines of therapy, with durable remissions observed in 30% to 40% of the pts, and median progression-free survival (PFS) at 2.9 and 5.9 months (M), respectively. Predictive value of early response assessments at M1 and M3 on patient outcome has not been described.
Aims
To evaluate the predictive value of early response at M1 and M3 after CAR T-cells infusion in R/R LBCL, using clinical evaluation and visual and semi-quantitative PET approach analysis.
Methods
Clinical, biological and metabolic characteristics, including, SUVmax, SUVpeak and total metabolic tumor volume (TMTV) before CART infusion (baseline), at M1 and at M3 were retrospectively collected. Complete metabolic response (CMR) was defined by a Deauville Score (DS) 1,2,3. Conversion rate from partial response (PR)/stable disease (SD) at M1 to complete response (CR) at M3 were assessed. Optimal ΔSUVmax at M1 or M3 able to predict PFS events were identified using time-ROC. Survivals were estimated by the Kaplan-Meier method. Cox models were used to select prognostic covariates based on landmark at M1 or M3.
Results
151 patients with R/R DLBCL (n= 122), transformed follicular lymphoma (tFL) (n= 19) or PMBL (n=10) were treated with Tisa-cel (n= 61) or Axi-cel (n=90). Median age was 63 (21-77). At time of decision, 131 (87%) patients were PS 0-1, 86 (57%) were IPI low or intermediate risks, 84 (56%) had abnormal LDH levels. Median number of prior lines was 3 (2-7). PFS at 6 and 12 months for the whole cohort were 48% and 42% respectively, and overall survival (OS) were 80% and 64%, respectively. Median time from infusion to progression was estimated at 3.7 months (95%CI, 3.0 to 16.7). At baseline: median SUVmax was 21.5 and median TMTV was 54ml. 66 (44%) patients presented a high TMTV (>80ml). At M1: 61/151 (40%) patients are in CR, 37/151 (24.5%) in PR and 10/151 (7%) in SD. Among 128 patients evaluable with DS, 52 (41%) are in CMR with DS 1-3, 32 (25%) have DS-4 and 44 (34%) have DS-5, with 6-month PFS and OS rates (from M1) at 78% and 95% for DS1-3, 76% and 97% for DS-4, 10% and 63% for DS-5, respectively. At M3: 64/108 (59%) were in CR and 14/108 (13%) were in PR or SD. 13 PR patients at M1 converted to CR at M3, while 13 stayed in PR/SD, and 11 progressed. 9/10 SD progressed. Among 103 patients evaluable with DS, 53 (51%) are in CMR with DS 1-3, 17 (17%) have DS-4 and 33 (32%) have DS-5, with 6-month PFS and OS rates (from M3) at 87% and 92% for DS1-3, 97% and 100% for DS-4, 9% and 71% for DS-5, respectively. ΔSUVmax cutoffs identifying good responder patients were a decrease of at least 73% at M1 and 85% at M3 with 51.6% and 51.6% of the patients who reached these cutoffs, respectively. Patients below the 73% threshold at M1 had decreased PFS (HR 3.96, p<0.001) and OS (HR=4.45, p<0.001). By contrast, patients below the 85% threshold at M3 had a decreased PFS (HR=4.18, p<0.001) but no different OS (HR=1.34, p=0.55), probably because early progressor with the poorest outcome have been excluded from this analysis.
Conclusion
Early metabolic response at M1 is strongly associated with better outcome. These data argue for an early monitoring of patients using functional imaging, and innovative strategies for very high-risk patients.
Keyword(s): CAR-T, Diffuse large B cell lymphoma, Positron emission tomography (PET)
Abstract: EP497
Type: E-Poster Presentation
Session title: Aggressive Non-Hodgkin lymphoma - Clinical
Background
Anti-CD19 CAR T-cells, Tisagenleuclecel (Tisa-cel) and Axicabtagene ciloleucel (Axi-cel), have been approved for pts with relapsed or refractory aggressive large B-cell lymphomas (R/R LBCL) after 2 lines of therapy, with durable remissions observed in 30% to 40% of the pts, and median progression-free survival (PFS) at 2.9 and 5.9 months (M), respectively. Predictive value of early response assessments at M1 and M3 on patient outcome has not been described.
Aims
To evaluate the predictive value of early response at M1 and M3 after CAR T-cells infusion in R/R LBCL, using clinical evaluation and visual and semi-quantitative PET approach analysis.
Methods
Clinical, biological and metabolic characteristics, including, SUVmax, SUVpeak and total metabolic tumor volume (TMTV) before CART infusion (baseline), at M1 and at M3 were retrospectively collected. Complete metabolic response (CMR) was defined by a Deauville Score (DS) 1,2,3. Conversion rate from partial response (PR)/stable disease (SD) at M1 to complete response (CR) at M3 were assessed. Optimal ΔSUVmax at M1 or M3 able to predict PFS events were identified using time-ROC. Survivals were estimated by the Kaplan-Meier method. Cox models were used to select prognostic covariates based on landmark at M1 or M3.
Results
151 patients with R/R DLBCL (n= 122), transformed follicular lymphoma (tFL) (n= 19) or PMBL (n=10) were treated with Tisa-cel (n= 61) or Axi-cel (n=90). Median age was 63 (21-77). At time of decision, 131 (87%) patients were PS 0-1, 86 (57%) were IPI low or intermediate risks, 84 (56%) had abnormal LDH levels. Median number of prior lines was 3 (2-7). PFS at 6 and 12 months for the whole cohort were 48% and 42% respectively, and overall survival (OS) were 80% and 64%, respectively. Median time from infusion to progression was estimated at 3.7 months (95%CI, 3.0 to 16.7). At baseline: median SUVmax was 21.5 and median TMTV was 54ml. 66 (44%) patients presented a high TMTV (>80ml). At M1: 61/151 (40%) patients are in CR, 37/151 (24.5%) in PR and 10/151 (7%) in SD. Among 128 patients evaluable with DS, 52 (41%) are in CMR with DS 1-3, 32 (25%) have DS-4 and 44 (34%) have DS-5, with 6-month PFS and OS rates (from M1) at 78% and 95% for DS1-3, 76% and 97% for DS-4, 10% and 63% for DS-5, respectively. At M3: 64/108 (59%) were in CR and 14/108 (13%) were in PR or SD. 13 PR patients at M1 converted to CR at M3, while 13 stayed in PR/SD, and 11 progressed. 9/10 SD progressed. Among 103 patients evaluable with DS, 53 (51%) are in CMR with DS 1-3, 17 (17%) have DS-4 and 33 (32%) have DS-5, with 6-month PFS and OS rates (from M3) at 87% and 92% for DS1-3, 97% and 100% for DS-4, 9% and 71% for DS-5, respectively. ΔSUVmax cutoffs identifying good responder patients were a decrease of at least 73% at M1 and 85% at M3 with 51.6% and 51.6% of the patients who reached these cutoffs, respectively. Patients below the 73% threshold at M1 had decreased PFS (HR 3.96, p<0.001) and OS (HR=4.45, p<0.001). By contrast, patients below the 85% threshold at M3 had a decreased PFS (HR=4.18, p<0.001) but no different OS (HR=1.34, p=0.55), probably because early progressor with the poorest outcome have been excluded from this analysis.
Conclusion
Early metabolic response at M1 is strongly associated with better outcome. These data argue for an early monitoring of patients using functional imaging, and innovative strategies for very high-risk patients.
Keyword(s): CAR-T, Diffuse large B cell lymphoma, Positron emission tomography (PET)