Overall Survival (OS)

Axicabtagene Ciloleucel

Participants received cyclophosphamide 500 mg/m\^2/day IV infusion and fludarabine 30 mg/m\^2/day IV infusion conditioning chemotherapy for 3 days followed by axicabtagene ciloleucel administered as a single IV infusion at a target dose of 2 x 10\^6 anti-CD19 CAR transduced autologous T cells/kg on Day 0.

Standard of Care Therapy

Participants received 2 or 3, 21-day cycles of second-line chemotherapy regimen: * R-ICE: rituximab 375mg/m\^2 before chemotherapy, ifosfamide 5g/m\^2 24 hours CI on Day 2 + mesna, carboplatin AUC 5 Day 2, maximum dose 800mg, etoposide 100mg/m\^2/day on Days 1-3; * R-ESHAP: rituximab 375mg/m\^2 Day 1, etoposide 40mg/m\^2/day IV on Days 1-4, methylprednisolone 500mg/day IV on Days 1-4 or 5, cisplatin at 25mg/m\^2/day CI Days 1-4, cytarabine 2g/m\^2 on Day 5; * R-GDP: rituximab 375mg/m\^2 Day 1 (or Day 8), gemcitabine 1g/m\^2 on Days 1 and 8, dexamethasone 40mg on Days 1-4, cisplatin 75mg/m\^2 Day 1 or carboplatin AUC=5; or * R-DHAP: rituximab 375mg/m\^2 before chemotherapy, dexamethasone 40mg/day on Days 1-4, high dose cytarabine 2g/m\^2 every 12 hours for 2 doses on Day 2 following platinum, cisplatin 100mg/m\^2 24 hours CI on Day 1 or oxaliplatin 100mg/m\^2. Participants who responded to second-line chemotherapy got high dose therapy and autologous stem cell transplant.

Analysis denominator · Participants

1. ci Lower Limit

   0.530

   ci Num Sides

   TWO_SIDED

   ci Pct Value

   95

   ci Upper Limit

   1.007

   estimate Comment

   Stratified Cox regression models were used to estimate hazard ratio and 2-sided CIs for axicabtagene ciloleucel relative to standard of care therapy. The Breslow method was used to handle the ties for the Cox regression models.

   group Ids

   1. OG000
   2. OG001

   non Inferiority Comment

   One-sided p-value based on log-rank test stratified by response to first-line therapy and second-line age-adjusted IPI as data collected on case report forms.

   non Inferiority Type

   SUPERIORITY

   other Analysis Description

   The Rho family spending function with parameter (Rho = 6) was used to allocate alpha between the interim and primary OS analysis. Given that fewer than the anticipated 210 events were observed at the data cutoff date for the primary OS analysis (25 January 2023), the efficacy boundary for the primary OS analysis was recalculated using the Rho family spending function based on the actual observed event numbers (177 deaths) resulting in an efficacy boundary at 1-sided significance level of 0.0249.

   p Value

   0.027

   p Value Comment

   Stratified log-rank (randomization factor) p-value.

   param Type

   Hazard Ratio (HR)

   param Value

   0.730

   statistical Comment

   Stratified (randomization factor) log-rank test.

   statistical Method

   Log Rank