Chimeric Antigen Receptor 4SCAR19-Modified T Cells in Acute Lymphoid Leukemia: a Phase II Multi-Center Clinical Trial in China

Background: Relapsed/refractory leukaemia is associated with poor prognosis. T cells genetically modified to express CD19-specific chimeric antigen receptor (CD19CAR) in patients (pts) with chronic lymphoblastic leukemia (CLL) and acute lymphoblastic leukemia (ALL) have shown a remarkable ant-cancer activity. However, many questions remain related to the predictive indicators of long-term response and the management of cytokine release syndrome (CRS) after CAR-T cell infusion. In this phase II multi-center clinical trial, we evaluated the safety and efficacy of a fourth generation, safety-improved CD19-CAR (4SCAR19) in B-ALL pts.

Patients and Methods: Fifty evaluable B-ALL patients (pts), with demonstrated persistent disease following salvage chemotherapy from 14 hospitals in China between July 2013 and June 2015 have been enrolled. The mean age is 14 (from 3 to 65) including 26 children and 24 adults. Leukemic genotypes include 16 Bcr-Abl (13 pT315I), 11 WT-1, 3 MLL-AF4, 3 E2A-PBX-1, 1 TEL/AML1, 1 IKZF1, 1 K-ras and the remaining 14 pts have undetectable genotype. Their disease characteristics include: 4 hypercritical acute leukemia, 10 extramedullary leukemia (7 CNSL, 3 multiple sites), with associated co-morbidity: 12 Aspergillus pneumonia, 4 cGVHD, 2 pleura/pericardial cavity effusion, 2 hepatitis B, 2 diabetes mellitus, 1 gastrointestinal hemorrhage, and 1 liver/spleen abscess. Twenty-one (42%) pts received allo-HSCT including 15 haplo-identical, 5 matched related donor (MRD) and 1 unrelated cord blood (URD-CB). These pts who relapsed after transplantation have received chemotherapy (chemo), combined with Tyrosine Kinase Inhibitors (TKIs, 10), donor leukocyte infusion (DLI, 19), or dendritic cells-cytokine induced killer cell (DC-CIK)/NK cell infusions (7). CAR-T cells were prepared from autologous (37), transplant donor (12) or non-transplant donor mother (1). T cells. Peripheral lymphocytes were collected from leukapheresis, and T cells were transduced with a 4th generation, safety-engineered, CD19scFv/CD28/CD137/CD27/CD3ζ-iCasp9 (4SCAR-19) lentivector. Pre-CAR-T lymphodepleting chemotherapy includes individualized chemo in 17 pts, and the others received Fludarabine (Flu) + Cyclophosphamide (Cy)(FC), or either Flu or Cy regimen: (1) FC: Cy 250mg/m2/d x3d and Flu 30mg/m2/d x3d ,or (2) either Flu x3 days, or (3) Cy x3 days, followed by CAR-T infusions at a dose of 2.13 (range from 0.42 - 5.9) x10⁶ CAR-T cells per kg body weight per infusion.

Results: For statistical analysis, 50 patients were divided into 2 cohorts: Cohort 1: B-ALL with morphological blast <50%, 35; and Cohort 2: Blast ≥50%, 15 pts before CAR-T treatment. The end points are: 1. early response (1-3 mon. after CAR-T infusion), and 2. leukemia-free survival (LFS) and overall survival (OS). LFS is defined as survival without recurrent malignancy and MRD (-). Early response evaluation is summarized below:

Toxicities: CRS occurred in most pts within the first 10 days of CAR-T cell infusion. 47/50 (94%) pts developed fever with elevated IL-2, IL-6, IL-10, and interferon gamma. Eight (16%) pts required either 12.5 mg Etanercept or 8 mg/kg tocilizumab, and three pts were treated with both drugs. Four pts developed hypotension and fully recovered after receiving dopamine. Four pts were treated once by methylprednisolone (1 mg/kg/day). The median follow-up was 4 months (range from 3~24 month), with 16 pts followed up for more than 6 months. Clinical outcomes: 1. LFS: The 120 days LFS for pts in cohort 1 and cohort 2 were 86% (CI,80%~93%) and 44.4% (CI, 31%~58%, P=0.0030), respectively. 2. OS (10 month) probabilities for patients in cohort 1 and cohort 2 were 82% (73%~91%) and 36% (19%~52%) (P= 0.0029), respectively.

Conclusion: Our results indicate the potential of rapid leukemia eradication kinetics of 4SCAR19 therapy in treating chemo-resistant B-ALL. This therapy dramatically improves the prognosis of B-ALL pts by either providing a bridging approach to allo-HSCT or a better remission induction with longer period of CR than the routine treatment. Moreover, pts with morphological blasts <50% have demonstrated significantly better outcomes than pts with blasts ≥50%.