A Novel Composite Endpoint of Toxicity-Free, Progression-Free Survival (TPFS) after CD19-Directed Chimeric Antigen Receptor T-Cell Therapy for Relapsed/Refractory Large B-Cell Lymphoma

Background: CD19-directed chimeric antigen receptor T-cell (CAR T) therapy with axicabtagene ciloleucel (axi-cel), tisagenlecleucel (tisa-cel), or lisocabtagene maraleucel (liso-cel) are FDA-approved for relapsed and/or refractory (r/r) large B-cell lymphoma (LBCL) after ≥2 prior lines of systemic therapy (all 3) or earlier after 1 line (axi-cel and liso-cel) based on durable remissions of 30-40%. Pivotal trials and real-world data demonstrated notable heterogeneity in toxicity and efficacy between CAR T products. To account for such differences, we have established a novel composite endpoint of toxicity-free, progression-free survival within 6 months (TPFS6) after CAR T-cell therapy. We also studied factors associated with TPFS6 across CAR T-cell products and its impact on long-term success of CAR T-cell therapy.

Methods: This retrospective cohort analysis included 282 consecutive adult patients with r/r LBCL and its variants who received axi-cel (n=228), tisa-cel (n=47), or liso-cel (n=7) at Moffitt Cancer Center (05/2015-01/2022). TPFS6 was defined as absence of severe (≥Grade 3) cytokine release syndrome (CRS), immune effector cell associated neurotoxicity syndrome (ICANS), progressive disease (PD)/stable disease (SD) and non-relapse mortality (NRM) within 6 months of CAR T-cell infusion. CRS and ICANS were graded according to CTCAE or ASTCT Consensus grading. Factors associated with TPFS6 were assessed via logistic regression multivariate analysis (MVA). Landmark analysis at the 6-month point examined the impact of TPFS on overall survival (OS). All analyses were performed utilizing SAS and R.

Results: For the entire cohort (median age=64 years [range, 19-86]; 60.3% males), the TPFS6 was 37% (95% confidence interval [CI] 31-43, Figure 1A) with no difference between the recipients of axi-cel, tisa-cel, or liso-cel (p=0.97). Individual components of TPFS6 such as severe CRS/ICANS, NRM, and PD/SD accounted for 6.8%/30.3%, 1.7%, and 61.2%, respectively. Patients with TPFS6 differed from their counterparts by gender (males, 51% vs 65.7%, p =0.014), presence of bulky (8.7% vs 17.4%, p =0.045) or extra nodal (≥2 sites, 30.8% vs 46.6%, p=0.009) disease, ECOG performance (≥2, 7.7% vs 20.8%, p=0.004), prior autologous transplant (25% vs 13.5%, p=0.015), C-reactive protein (CRP) (>4 mg/dL, 10.6% vs 35.4%, p<0.001), ferritin (≥400 ng/mL, 37.5% vs 62.4%, p<0.001), lactate dehydrogenase (LDH) (≥225 U/L, 51% vs 76.4%, p<0.001) at lymphodepletion (LD). TPFS6 patients were comparable to those experiencing at least one of the TPFS6 endpoints according to age, race, number of prior lines or use of bridging therapy, LBCL variants, Ann Arbor staging, double/triple hit status, prior CNS involvement, and vein-to-vein time (all p>0.1).

In logistic regression MVA, lower levels of baseline LDH (odds ratio [OR]=0.44, p=0.004), CRP (OR=0.33, p=0.004), ECOG<2 (OR=0.43; p=0.05) and female gender (OR=0.5; p=0.01) were associated with TPFS6. In the landmark Cox MVA of OS, TPFS6 was strongly associated with 5.2-fold survival difference (p<0.001, Figure 1B) after adjustment for age (hazard ratio [HR]=1.24 per 10-year increment, p=0.077) male gender (HR=1.92, p=0.013), ECOG≥2 (HR=2.34, p=0.004), high LDH at LD (HR=2.63, p<0.001), and CAR T product (HR=1.67 for tisa-cel vs axi-cel, p=0.079). At 6 months, overall response and complete response rates were 48% and 45%, respectively. With median follow up of 15.4 months for surviving patients, the estimated 2-year OS and PFS were 54.5% and 39.3%, respectively.

Conclusions: Only 37% of patients were alive at 6 months without experiencing a TPFS-defining event. TPFS6 was associated with known clinical risk factors such as CRP, LDH, ECOG and gender but not the type of CAR T product. TPFS6 encompasses clinically meaningful components associated with OS and it therefore represents an ideal recovery outcome by measuring initial treatment success without progression, major morbidity, and mortality. Thus, TPFS6 needs to be further explored as a uniform endpoint for evaluating CAR T-cell safety and efficacy in future clinical trials of r/r LBCL.