Phase 1 Alexander Study of AUTO3, the First CD19/22 Dual Targeting CAR T Cell Therapy, with Pembrolizumab in Patients with Relapsed/Refractory (r/r) DLBCL

Background:

CD19 directed CAR T cells are effective in patients with r/r DLBCL. However, relapses due to CD19 loss or PDL1 upregulation are common. In this study, we evaluate the safety and efficacy of AUTO3, a CAR T targeting CD19/22 with limited duration of PD-1 blockade. Preliminary safety presented at ASCO 2020 demonstrated lack of dose limiting toxicities (DLT) during the dose escalation phase. Here we present updated efficacy and safety data with longer follow up that included additional patients treated at the RP2D dose range and in an outpatient setting.

Methods:

We generated a bicistronic retroviral vector encoding both an anti-CD19 (OX40 co-stim) and an anti-CD22 (41BB co-stim) CARs with humanized binders. The cell products were manufactured in a semi-automated, closed process using CliniMACS Prodigy.

Patients (≥ 18 years) with r/r DLBCL (NOS) or transformed (tDLBCL); ECOG < 2, adequate organ function were eligible. Lymphodepletion was with Flu/Cy prior to AUTO3 and bridging therapy was allowed. The three dose levels explored are 50, 150, and 450 x 10^6 CAR T-cells. Patients received AUTO3 alone, or with 3 doses of pembrolizumab (pem) 200 mg q 3 wks starting on D14 (regimen A), or with a single dose of pem 200 mg on D-1 (regimen B). After RP2D is determined, 12 patients were to be dosed at the RP2D dose range (150-450 x 10^6 CAR T-cells with D-1 pem). An expansion cohort of twenty patients was added to examine safety and tolerability in an outpatient setting. The primary endpoint is frequency of DLTs and grade (G) 3-5 adverse events (AE) and secondary endpoints included ORR, CRR, and biomarkers.

Results:

As of 13 July, 2020, 33 patients treated with AUTO3. The median age was 59 (28 - 83) and median number of prior therapies was 3 (1 - 10). 79% had refractory disease, 76% were DLBCL NOS, 18% were tDLBCL, and 6% were high grade B Cell Lymphoma. Among 15 patients with molecular data, 9 patients had double HIT and 3 patients had triple HIT. No DLT was observed during dose escalation. Treatment emergent AEs that occurred > 20% were neutropenia (73%), thrombocytopenia (64%), anemia (61%), cytokine release syndrome (33%), pyrexia (30%), constipation (27%), and fatigue (24%). G > 3 treatment emergent AEs that occurred > 20% were neutropenia (73%), thrombocytopenia (48%), and anemia (48%). Majority of Serious AE were hematological or infectious and largely reversible. Across all dose levels, there were 0% sCRS with primary infusion and 9% neurotoxicity/ICANS (3 cases). All 3 cases of NT/ICANS occurred in the setting of disease progression with very minimal to no CART cells in the peripheral blood, and associated with confounding factors such as sepsis, metabolic acidosis, hyponatremia, and narcotic use.

29 patients were evaluable for efficacy (PET + disease prior to pre-conditioning). The ORR and CRR were 69% and 52%. Fourteen out of 15 CRs did not experience relapse with a median f/u of 3 months (1-24 mth). Among the 15 evaluable patients treated at a dose > 50 x 10⁶ with D-1 pem, the ORR was 73% and CRR was 60%, and no relapse observed among CR patients (1-6 mth). In the outpatient cohort, twelve patients have been leukapheresed and three patients have been dosed thus far. Additional patients treated in an out-patient setting and longer follow up, as well as relevant biomarkers will be presented.

Conclusions:

AUTO3 at RP2D dose range of > 50 x 10⁶ CAR T cells with D-1 pembrolizumab induces durable complete remissions. None of the patients in CR experienced severe CRS or neurotoxicities of any grade. Outpatient cohort is currently enrolling.