Response to lymphodepleting regimen is associated with PFS in CAR-T infused NHL patients

On 19 February 2019, Alexandre Hirayama from the Fred Hutchinson Cancer Research Center, Seattle, WA, USA, and colleagues, published in Blood results from multivariate analyses on factors affecting outcomes and survival in aggressive non-Hodgkin lymphoma (NHL) patients receiving CAR T-cell therapy.

Chimeric antigen receptor T cells (CAR-T) have opened up a new therapeutic avenue for relapsed or refractory (R/R) NHL patients, thus identifying factors that could be associated with even longer remissions after CD19 CAR T-cell therapy is crucial. In this phase I-II trial (NCT01865617), the authors performed a series of multivariate analyses in order to investigate various factors impacting the response rates and progression-free survival (PFS) of aggresive NHL patients, who had received CD19 CAR-T cells.

Study design

• N = 48 aggressive NHL patients with primary refractory disease after chemoimmunotherapy, or R/R disease after at least two lines of therapy, or not eligible for hematopoietic stem cell transplantation (HSCT) or had relapsed after HSCT:
  • Diffuse large B-cell lymphoma (DLBCL; not otherwise specified [NOS] or transformed from indolent lymphoma): n = 28 patients
  • Mantle cell lymphoma (MCL): n = 6 patients
  • Follicular lymphoma (FL): n = 1 patient
  • Other aggressive NHL: n = 5 patients
  • High-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements (HGBL-DH/TH): n = 8 patients
• All patients included in this analysis received Cy/Flu lymphodepletion and a subsequent CAR-T infusion during the phase I trial
• CAR-T infusion (2−4 days after lymphodepletion):
  • 2 x 10⁶ CD19 CAR-T cells/kg at a 1:1 ratio of CD4:CD8 cells
• CAR-T construct:
  • FMC63-derived anti-CD19 single chain variable fragment (scFv) and 4-1BB costimulatory domain
  • Truncated human epidermal growth factor receptor (EGFRt) separated from the CAR construct by ribosomal skip sequence to allow for CAR-T cells quantification by flow cytometry
• Lymphodepletion regimens:
  • Cyclophosphamide and fludarabine (Cy/Flu):
    • High intensity Cy/Flu (n = 31):
      • Cy: 60 mg/kg (x1) + Flu 25 mg/m² (x3) [n = 30]
      • Cy: 60 mg/kg (x1) + Flu 25 mg/m² (x5) [n = 1]
    • Low intensity Cy/Flu (n = 17):
      • Cy 30 mg/kg total dose (x1) + Flu 25 mg/m² (x3) [n = 6]
      • Cy 300 mg/kg total dose (x3) + Flu 30 mg/m² (x3) [n = 10]
      • Cy 500 mg/kg total dose (x3) + Flu 30 mg/m² (x3) [n = 1]
    • Median patient age (range): 58.5 (52−63) years
    • Median number of prior lines (range): 4 (1−11)
    • Patients who had previously failed autologous SCT: 48% (n =23)
    • Patients who had previously failed allogeneic SCT: n =4
    • Patients who had previously failed both allogeneic and autologous SCT: n = 3
    • Bridging therapy was required by 25% (n = 12) of patients

Key results

• Patients evaluable for response: n = 47
• Best overall response rate (ORR) without additional therapy after CAR-T infusion: 51% (95% CI, 36−66%):
  • Complete response (CR): 40% of patients (95% CI, 27−56%)
  • Four patients with partial response (PR) and two with stable disease (SD) converted to CR after the first assessment without additional antitumor therapy
• Median time to best response (range): 1.0 (0.9−6.2) month
• For DLBCL patients:
  • ORR: 50% (95% CI, 33−67%)
  • CR rate: 43% (95% CI, 25−63%)
• For HGBL-DH/TH patients:
  • ORR: 38% (95% CI, 10−74%)
  • CR rate: 25% (95% CI, 4−64%)
• At a median follow-up of 26.9 months (range, 2.5−32.4 months) for all patients:
  • Median PFS: 2.9 months (95% CI, 1.7−6.6)
  • Median overall survival (OS): 11.4 months (95% CI, 7.0−28.9)
• At a median follow-up of 26.9 months (range, 2.5−32.4 months) for those who achieved CR:
  • Median PFS: 20.0 months (95% CI, 9.2−not reached)
  • Median OS: 28.9 months (95% CI, 24.2−not reached)
• Estimated 24-month PFS probability (among CR patients): 46% (95% CI, 28−76%)
• Estimated 24-month OS probability (among CR patients): 72% (95% CI, 54−96%)

Multivariate analysis

• The likelihood of CR after CAR-T infusion was higher in patients with:
  • Less aggressive disease before treatment (lower lactate dehydrogenase [LDH] levels; HR = 0.24; [95% CI, 0.08−0.53]; P = 0.003), or
  • Higher levels of the serum monocyte chemoattractant protein-1 (MCP-1) in response to lymphodepletion (HR = 1.36; [95% CI, 1.12−1.79]; P = 0.007)
• The likelihood of PFS after CAR-T infusion was higher in patients with:
  • Less aggressive disease before treatment (lower LDH levels)
  • Higher levels of MCP-1 on the day of the infusion
  • Higher levels of interleukin 7 (IL-7) after infusion (median day 4)
• Patients who received high-intensity Cy/Fly lymphodepletion had a more favourable cytokine profile than those receiving low-intensity Cy/Flu regimen
• In a subgroup analysis involving only patients achieving CR:
  • Patients who had peak IL-7 concentration after CAR-T cell infusion above the median for all patients (23.4 pg/ml) had better survival than those who had peak IL-7 levels below or equal to the median with 24-month probabilities of PFS and OS of 62% (95% CI, 39−100%) and 91% (95% CI, 75−100%), respectively

Conclusions

• LDH, MCP-1, and IL-7 seem to be independent factors that are associated with better PFS in patients with aggressive NHL after CD19 CAR-T infusion
• Cytokine levels were associated with CR and/or PFS increase in lymphodepleted patients and were higher in patients receiving high intensity Cy/Flu lymphodepletion compared to those receiving low intensity Cy/Flu