Remissions of Acute Myeloid Leukemia and Blastic Plasmacytoid Dendritic Cell Neoplasm Following Treatment with CD123-Specific CAR T Cells: A First-in-Human Clinical Trial

The current treatment of relapsed or refractory AML is associated with low rates of complete response (CR) and considerable complications. As a result, only a minority of patients (pts) proceed to allogeneic hematopoietic stem cell transplantation (alloHSCT) with curative intent. Furthermore, outcomes for AML pts with disease relapse after alloHSCT are dismal. Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare and incurable blood cancer with a median survival of <18 months and no standard of care. Thus, there are clear unmet therapeutic needs in both these conditions. CD123 is overexpressed on AML blasts and leukemic stem cell (LSC)-enriched cell subpopulations compared to normal hematopoietic stem cells and myeloid progenitors. High levels of CD123 expression is also one of the diagnostic hallmarks for BPDCN. All these features make CD123 an attractive target for T cell based adoptive immunotherapy.

We have previously demonstrated preclinically the anti-AML activity of CD123-chimeric antigen receptor (CAR) T cell therapy (Mardiros, Blood 2013). The CD123CAR contains an anti-CD123 single-chain variable fragment, an optimized IgG4 CH2CH3 linker, a CD28 co-stimulatory domain, and a CD3 zeta signaling domain and is used to engineer T cells for patients enrolled on a single center, first-in-human phase I dose escalation clinical trial open at our Institution (NCT02159495). The primary objective is to test the safety and activity of escalating doses of CD123CAR T cells for patients with relapsed or refractory AML (cohort 1) and BPDCN (cohort 2). Donor-derived or autologous T cells from leukapheresed peripheral blood mononuclear cells were lentivirally transduced with the CD123CAR vector. Prior to T cell infusion, all patients undergo a lymphodepleting regimen including fludarabine 25-30 mg/m² daily for 3 days and cyclophosphamide 300 mg/m² daily for 3 days. Pts receive a single dose of CD123CAR T cells with an option for a second infusion if they continue to meet safety and eligibility criteria and still have CD123+ disease. To date, 14 patients have been enrolled and 7 treated (6 AML, 1 BPDCN).

All 6 patients in the AML cohort had refractory AML following alloHSCT, and a median of 4 (range: 4-7) prior lines of therapy. Of the 2 pts treated at dose level (DL) 1 (50M CAR+ T), 1 achieved a morphologic leukemic-free state, which lasted 2 months. She received a second infusion 3 months later with subsequent blast reduction from 77.9% to 0.9% (flow cytometry) after 35 days. Of the 4 pts on DL 2 (200M CAR+ T), 1 achieved CR and became transfusion independent. She proceeded to a second alloHSCT on day 70. Restaging on day +161 post-transplant showed she has remained in MRD-negative CR with good engraftment and 100% donor chimerism. Another pt with CR prior to treatment remained in CR at day 28 and has proceeded to a second alloHSCT. The remaining 2 patients had reductions in blast counts, but did not achieve remission. All toxicities were reversible and manageable: cytokine release syndrome (CRS; 4 grade 1, 1 grade 2); 1 adenoviral pneumonia requiring intubation; and 1 grade 3 rash due to drug hypersensitivity. There were no dose limiting toxicities and no treatment-related cytopenias. One pt with prior alloHSCT had mild recurrent cutaneous GVHD, which occurred after the completion of CAR T treatment. Correlative studies including T cell persistence and CD123 expression are underway and will be reported.

In the BPDCN cohort, 1 pt has been treated so far, a 74-year-old man with a persistent bulky subcutaneous mass after clinical trial treatment with a CD123 antibody-drug conjugate. Following lymphodepletion, he received a single dose of 100M autologous CD123CAR T cells and continues to be in CR at 60 days post-infusion. Skin biopsies at the tumor site on days 14 and 28 showed no evidence of disease. Restaging work-up at day 28 revealed disease-free bone marrow, no new masses by CT scan, normalized blood counts, and complete resolution of clinical symptoms. The pt tolerated the treatment well with no CRS or neurologic toxicity.

In this first-in-human clinical trial of CD123CAR T cell therapy, we have demonstrated the feasibility and safety of targeting CD123. We have also observed promising anti-leukemic activity in both AML and BPDCN. Importantly, no myeloablative effects have been observed, supporting further testing of this immunotherapeutic strategy in both transplant eligible and ineligible patients.