CNS Activity in Acute Leukemia No Barrier to CAR-T Remissions

CAR T-cell therapy was active and led to durable remissions in pediatric and young adult patients with acute lymphocytic leukemia or lymphocytic lymphoma and central nervous system (CNS) involvement, pooled data from five clinical trials found.

In the post hoc analysis involving 195 patients with relapsed or refractory disease, complete responses at 28 days post-infusion of CD19-directed CAR T cells were not significantly different for those with CNS activity and those without (97% vs 94%, respectively), reported Shannon Maude, MD, of Children’s Hospital of Philadelphia.

Similarly, at 2 years, the relapse-free survival rate was identical (at 60% in both groups), while overall survival (OS) rates were 83% in the group with CNS involvement and 71% in the group without CNS involvement (P=0.39), according to the findings in Lancet Haematology.

Importantly, there was no increased risk for severe neurotoxicity, as the higher overall rates seen in the CNS-positive group were driven by lower grade events.

Roughly 20% of acute leukemia relapses occur in the CNS, which often requires treatment with hematopoietic stem-cell transplantation and/or cranial irradiation, and these patients tend to have a poor prognosis, wrote Maude and colleagues.

They explained how efficacy and safety concerns resulted in the exclusion of patients with active CNS disease from most CD19-directed CAR T-cell therapy trials, leading to scant evidence to guide treatment in this population.

Writing in an accompanying editorial, Susanna Rives, MD, of Hospital Sant Joan de Deu Barcelona, said the study supports the use of CD19-directed CAR T-cell therapy for “patients with relapsed or refractory B-acute lymphocytic leukemia who have isolated or combined CNS disease that is adequately controlled,” but also highlighted some of the limitations with the findings.

“These results were obtained in patients without acute neurotoxicities greater than grade 1, in whom the disease was not under control or there were brain lesions deemed to increase the risk of neurotoxicity,” Rives noted. “Therefore, caution should be exercised when treating patients with advanced CNS involvement.”

“In patients with particular locations of CNS infiltration (e.g., orbital, brain parenchyma, or paraspinal regions), the potential trafficking and inflammation caused by CAR T cells (e.g., pseudotumor growth) could lead to severe toxicity,” she continued. “Bridging therapy with systemic and intrathecal chemotherapy targeted at the CNS should then be considered, as well as seizure prophylaxis.”

For their study, Maude’s group pooled data on 154 patients treated with tisagenlecleucel (Kymriah) across four trials (ELIANA, ENSIGN, Pedi CART19, and 16CT022) and 41 patients treated with an investigational CD19-directed product, huCART19, in the 13BT022 trial. All patients were treated from 2012 to 2016 and had relapsed or refractory acute lymphocytic leukemia or lymphocytic lymphoma. Median follow-up was 36 months for the CNS-negative group and 39 months for the CNS-positive group.

Patients included ranged from ages 1 to 29 years (41% under age 10) and 56% were male. Of the total population, 66% had no CNS involvement while 34% had CNS involvement. Most had CNS1 disease (91%) prior to infusion, while 7% had CNS2 and 2% had CNS3 disease.

The study also examined outcomes in the 22% of patients with isolated CNS involvement at relapse. Here, complete response rates 28 days following CAR T-cell infusion were identical (98%) when compared with patients with bone marrow involvement at relapse with or without CNS involvement. The 2-year OS rate was significantly higher among the group with isolated CNS involvement (91% vs 71%, respectively, P=0.046).

Rives noted that both tisagenlecleucel and huCART19 are 41BB constructs, so the findings in the study may not be generalizable to other CAR T-cell products.

For the total population, all-grade neurotoxicity occurred in 58% of the CNS-positive group and 41% of the CNS-negative group: grade 1 (30% vs 19%, respectively), grade 2 (15% vs 11%), grade 3 (9% each), grade 4 (3% vs 2%).

Rates of all-grade cytokine release syndrome were 80% for the CNS-positive group and 85% for the CNS-negative group: grade 1 (3% vs 9%, respectively), grade 2 (58% vs 47%), grade 3 (11% vs 14%), grade 4 (9% vs 15%).