Higher CAR T-Cell Dosage Improves Efficacy Without Toxicity in Relapsed or Refractory CLL


Patients with relapsed or refractory chronic lymphocytic leukemia who received a higher dose of CAR T cells, achieved complete responses, regardless of cell dose, which was highly correlated with long-term survival.

Chimeric antigen receptor (CAR) modified T cells targeting CD19 (CAR T-19) may be more effective at a dose of 5 x 108 than 5 x 107 in patients with relapsed or refractory chronic lymphocytic leukemia (CLL). Toxicity was comparable in both dose groups.

Achieving a complete response (CR), regardless of cell dose, was highly correlated with long-term survival, according to a dose-optimization study published recently in The Journal of Clinical Oncology.

Among 32 evaluable patients, the CR rate was 28% (90% Confidence Interval [CI], 16%-44%). The overall response rate (ORR) was 44% (90% CI, 29%-60%). The median overall survival (OS) was 64 months for all patients, with no statistically significant between-group differences (64 vs. 68 months; P = .84). Patients who achieved a CR experienced prolonged survival compared to those who did not, regardless of dose group (P = .035). The median OS was not reached in patients with CR versus 64 months in those without CR. The median progression-free survival (PFS) was 40.2 months in patients who achieved CR compared to 1 month for patients who did not have a CR (P < .0001).

The authors, led by Noelle V. Frey, MD, MS, of the Perelman Center for Advanced Medicine at the University of Pennsylvania, believe that their study includes both the largest prospective cohort of CLL patients to receive CAR T cells and the longest follow-up. “We sought to determine an association of cell dose with response or toxicity in CLL. Although small numbers limit the statistical significance, the higher target dose (5 x 108 CAR T-19) using an adaptive split-dosing strategy was safe and possibly more effective than the lower dose,” they wrote. “Using this strategy, CRS [cytokine release syndrome] was generally moderate and manageable, and neurotoxicity was negligible.”

This split dosing strategy emerged during the stage 2 expansion phase. Stage 1, the dose-finding portion, had randomized patients to either the high-dose (5 x 108, minimum of 1 x 108) or the low-dose (5 x 107, minimum of 1 x 107) group. Upon completion of this stage, the authors chose the high-dose group for expansion due to its higher CR rate.

When starting stage 2, investigational new drug-compliant manufacturing modifications began. The first patient treated developed CRS within 12 hours of receiving T cells, and received tocilizumab and steroids within 24 hours of infusion. “On the basis of this experience and to improve patient safety, doses of CAR T-19 in all subsequent patients were administered via split dosing: 10% on day 1, 30% on day 2, and 60% on day 3,” the authors wrote. “A total of eight patients were infused with this fractionated dosing schedule. This design allowed withholding of subsequent doses after early signs of


The study’s primary end point was to identify an optimal dose of CAR T-19 as inducing CR in 30% or more of the patients at 3 months. Secondary end points included ORR, PFS, OS, toxicity including CRS, and manufacturing feasibility.

Eligible patients were adults with relapsed or refractory CLL who had completed at least 2 prior regimens. Patients with p53-related aberrations were eligible if they did not achieve a CR with initial therapy, or if they had disease progression within 2 years. Investigators recommended lymphodepleting chemotherapy but did not require it.

Among 32 evaluable patients, the median age was 61.6 years (range, 48.8-76.1 years). Patients had received a median of 3.5 prior therapies (range, 2-7). Nine patients had received BTK inhibitor and one patient had received venetoclax.

In stage 1, 28 patients were infused, but only 24 received the target dose and were evaluable for response. At the 90-day mark, 4 of 11 patients in the high-dose group achieved CR (36%; 90% CI, 14%-65%). There were 2 CRs among 13 patients in the low-dose group (15%; 90% CI, 3%-41%).

Ten patients were infused in stage 2; of these, 8 were evaluable. Thus, the high-dose patient population consisted of 19 patients (11 in stage 1, 8 in stage 2). ORR in this combined high-dose group was 53% (n = 10 of 19; 90% CI, 32%-73%) including 37% CR (n = 7). Three of these patients (16%) achieved a partial response (PR).

In the high-dose group, the overall response was 42% (n = 10). Notably, all 5 patients who were not able to receive the full dose did not respond to CAR T-cell therapy. In the low-dose group, the overall response among 14 intention-to-treat patients was 29% (n = 4). One patient who was unable to take the full dose was a non-responder. The ORR for all evaluable patients was 44% (n = 14 of 32; 90% CI, 29%-60%), with 28% CR (90% CI, 16%-44%) and 16% PR (90% CI, 6%-30%).

The authors found no significant association between response and patient age, number of prior therapies, stage at enrollment, p53-related aberrations, or IGHV mutation status. Follow-up lasted for a median of 31.5 months (range, 2-75 months).

Among patients who achieved a CR, the authors estimated the 36-month PFS to be 67% for patients with CR. “None of the patients without CR remained progression-free or alive beyond 36 months,” they wrote. “Disease in four of the nine patients achieving a CR relapsed 2, 20, 25, and 40 months after CAR T-19 infusion.”

At last follow-up, 18 patients remained alive (n = 7 of 13 in the low-dose group and n = 11-19 in the high-dose group). This includes 8 of 9 patients who achieved CR and 10 of 23 who did not.

During the first month of treatment, there were no treatment-related deaths. One patient in each group died within the first 3 months of progressive disease (one also had pneumonia). Among 13 patients in the low-dose group, 4 had grade 3 or 4 infections. Three had grade 1 or 2 CRS. Among 19 patients in the high-dose group, 5 had grade 3 or 4 infections and 5 had grade 3 or 4 CRS. One patient also developed a second malignancy.

Among the total trial population of 38 patients, 24 (63%) developed CRS. Of these, 7 patients (18%) were grade 3, and 2 patients (5%) were grade 4. Five patients required intervention with tocilizumab. In the high-dose group, 13 of 18 evaluable patients developed CRS. This includes 7 patients (36%) with grade 1 or 2 CRS and 6 patients (32%) with grade 3 or 4 CRS. The authors found that developing CRS did not affect a patient’s odds of achieving a CR (P = .44).

Frey et al concluded that efforts to further increase the CR rate to CAR T cells in CLL should be pursued, including with dose optimization as needed. Other important strategies could include prospective selection of patients who are most likely to benefit from CAR T cells, such as those exhibiting a CD271PD-12IL-6R1CD81 T-cell phenotype, or rationally combining CAR T cells with small-molecule inhibitors,” they wrote. “We conclude that a single course of treatment with CAR T-19 can induce durable remissions with an acceptable toxicity profile in some patients with advanced CLL.”


Frey NV, Gill S, Hexner EO, et al. Long-term outcomes from a randomized dose optimization study of chimeric antigen receptor modified t cells in relapsed chronic lymphocytic leukemia. J Clin Oncol. Published online April 16, 2020. doi https://doi.org/10.1200/JCO.19.03237

Related Videos
Related Content