CAR-T therapy with or without conditioning chemotherapy: Results from a phase I trial

On 2 April 2019, Mark Geyer and colleagues from the Memorial Sloan Kettering Cancer Center, New York, NY, USA, published in JCI Insight the results of a phase I trial. This study investigated the safety and long-term follow-up of chimeric antigen receptor T-cell (CAR-T) therapy with or without conditioning chemotherapy in patients with relapsed or refractory (R/R) chronic lymphocytic leukemia (CLL) or indolent B-cell non-Hodgkin lymphoma (B-NHL).

CAR-T therapy has shown great promise for the treatment of patients with R/R CLL or B-cell NHL who cannot be cured with standard of care regimens and who constitute a great unmet medical need. In this single-center, non-blinded, phase I trial (NCT00466531), the investigators sought to evaluate the efficacy and toxicity their CAR-T construct with or without conditioning chemotherapy.

Study design & baseline characteristics

• N = 20 patients included in this analysis:
  • R/R CLL: n = 16 patients
  • R/R B-NHL: n = 4 patients
    • Marginal zone lymphoma (MZL): n = 2
    • Follicular lymphoma (FL): n = 1
    • Mantle cell lymphoma (MCL): n = 1
  • Male patients: 70%
  • Median age (range): 63 (43−75) years
  • Median number of prior lines (range):
    • CLL cohort: 4 (1−11)
    • B-NHL cohort: 8 (6−10)
  • Molecular and cytogenetic abnormalities in the CLL cohort:

Unmutated immunoglobulin heavy-chain variable region gene (IgHV)	n = 9
11q deletion	n = 5
17p deletion or TP53 loss	n = 4
Complex karyotype	n = 3

• CAR-T construct & infusion:
  • CD28 co-stimulatory domain: 19−28z
  • Median duration of autologous T-cell collection (range) prior CAR-T infusion:
    • CLL cohort: 38 (20−225) days
    • B-NHL cohort: 109 (68−139) days
  • Median CD4⁺:CD8⁺ ratio in collected autologous T-cells (range):
    • CLL cohort: 1.9:1 (0.3:1−5:1)
    • B-NHL cohort: 1.9:1 (0.9:1−13.2:1)
  • Median CD4⁺:CD8⁺ ratio in infused CAR-T cells (range):
    • CLL cohort: 5.7:1 (0.3:1−0:1)
    • B-NHL cohort: 1.8:1 (0.8:1−3.1:1)
  • Median transduction efficiency (range): 30% (22−59%)
  • Median CAR-T cell product manufacturing (range) for entire cohort: 15 (11−19) days
• Study treatment stages for CLL cohort. The B-NHL cohort followed treatment Stage 3 only:
• Stage 1 (n = 3):
  • CAR-T infusion dose: 1.2−0 x 10⁷ cells/kg
  • Conditioning chemotherapy: none
• Stage 2A (n = 1 patient):
  • CAR-T infusion dose: 3.0 x 10⁷ cells/kg
  • Conditioning chemotherapy: 1.5 g/m² cyclophosphamide (Cy)
  • Outcome: this one patient died 48 hours after CAR-T infusion. This protocol was not used again
• Stage 2B (n = 3):
  • CAR-T infusion dose: 0.40−0 x 10⁷ cells/kg (split dose: 1/3 was administered on Day 0 and the rest on Day 1)
  • Conditioning chemotherapy: 1.5 g/m² Cy
• Stage 3 (n = 9):
  • CAR-T infusion total dose: 3.0 x 10⁷ cells/kg (split dose as above)
  • Conditioning chemotherapy: Investigator’s choice:
    • Cy (0.3, 3 or 1.5 g/m²)
    • Fludarabine (Flu; 25 mg/m²) and Cy (1.5 g/m²)
    • Bendamustine (70 or 90 mg/m²)
  • Five patients in the CLL cohort were allowed to receive continuous ibrutinib treatment up to CAR-T infusion (median: 7 months) and four patients up to leukapheresis (median: 4.8 months)

Key findings

	CLL cohort  (n = 16)	B-NHL cohort (n = 4)
Median follow-up (range)	40.6 (1.8−79.8) months	-
Median event-free survival	3.1 months	33.4 months
Median overall survival	17.1 months	Not reached
Objective response	38%	-
Complete response (CR) by IWCLL or Lugano criteria, respectively	n = 3/12 (two were MRD negative)	n = 2 (these patients were in CR at the time of CAR-T infusion and remained in CR thereafter)
Stable disease by IWCLL or Lugano criteria, respectively	n = 9	n = 2
Patients remaining in CR at median follow-up of 53 months or 24.7 months, respectively	100% (3/3)	50% (the other patient died 27 months post-infusion due to lymphoma unrelated causes)

• Amongst the patients with CLL, who did not achieve CR (n = 12):
  • Lost to follow-up: n = 1 patient
  • Died in active follow-up due to disease progression: n = 8 patients
  • Remained alive in active follow-up and received alternative treatment: n = 3 patients
    • Ibrutinib: n = 1
    • Allogeneic stem cell transplantation: n = 1
    • Other investigational therapy: n = 1
  • Amongst the patients with B-NHL who did not achieve CR (n = 2):
    • Patients achieving stable disease (SD): n = 2
  • Ex vivo expansion of T-cells (P = 0.040) and CD4⁺:CD8⁺ CAR-T cell ratios were significantly greater in patients receiving ibrutinib on leukapheresis

Safety

Cytokine release syndrome (CRS)

• CRS was observed in all patients
• The second fraction of CAR-T cells was withheld in 6 out of 11 patients with CLL (for whom split-dose infusion had been planned) due to early CRS development (including 4 out of 5 patients who were receiving concomitant ibrutinib)
• CRS severity (total cohort):
  • Grade 1: n = 8 patients
  • Grade 2: n = 10 patients
  • Grade 3: n = 1 patient
  • Grade 5: n = 1 patient died 48 hours after CAR-T infusion (suspected sepsis syndrome)
• Median CRS onset (total cohort): Day 1 post-infusion
• Latest CRS onset (total cohort): Day 3 post-infusion
• Median CRS duration (range): 2 (1−9) days
• Patients receiving tocilizumab for CRS management: n = 3

Neurological toxicity (NT)

• NT was observed in:
  • CLL cohort: n = 6
  • B-NHL cohort: n = 3
• Patients receiving corticosteroids for NT management: n = 2
• Median NT onset (total cohort): Day 2 post-infusion
• Latest NT onset (total cohort): Day 11 post-infusion
• Median NT duration (range): 1 (1−61) days
• NT severity:
  • Grade 3 encephalopathy: n = 2 patients (reversible)
  • Prolonged encephalopathy and dysphasia: n = 1 (gradual improvement to baseline)
  • All other neurological events were considered as Grade 1 or 2 and were reversible:
    • Encephalopathy: n = 5
    • Dysphasia: n = 3
    • Dysarthria: n = 1
    • Hallucinations: n = 1

Other adverse events (AEs)

• Common Grade 3−5 AEs observed in the total cohort:
  • Neutropenia: n = 8
  • Anemia: n = 6
  • Febrile neutropenia: n = 6
  • Thrombocytopenia: n = 5
  • Hypophosphatemia: n = 5
  • Lymphopenia: n = 4
  • Leukopenia: n = 4
  • Hypotension: n = 4
  • Hyperglycemia: n = 4
  • Hyponatremia: n = 4

Conclusions

• All conditioning chemotherapy regimens tested and the 19-28z CAR-T infusions were acceptably tolerated by patients with R/R CLL and indolent B-NHL, with a few patients achieving a durable CR (CLL cohort only)
• Since ex vivo T-cell expansion and CD4⁺:CD8⁺ CAR-T cell ratios were significantly greater in patients receiving ibrutinib on leukapheresis the authors suggested that ibrutinib may modulate autologous T-cell phenotype