Co-infusion of haplo-identical CD19-chimeric antigen receptor T cells and stem cells achieved full donor engraftment in refractory acute lymphoblastic leukemia

A 71-year-old female was admitted to the hospital because of pancytopenia. Complete blood count with manual differentiation showed 68% blasts, and bone marrow smear showed 68.5% blasts. Flow cytometry analysis revealed 41.7% cCD79a+, 97.2% CD19+, 69.5% CD20+, 54.5% CD10+, 97.3% CD38+, and 91.4% HLA-DR. She had normal female karyotype, and no fusion genes were detected. The primary diagnosis was ALL. She was treated with vindesine, mitoxantrone, cyclophosphamide, and dexamethasone as induction and achieved complete remission. A consolidation was given with the same regimen as induction. The patient suffered from fever and bone pain 2 months after the last chemotherapy, and repeated bone marrow smear showed 73% lymphoblasts. Two cycles of re-induction chemotherapy were initiated consisting of vindesine, fludarabine, cyclophosphamide, and prednisone; however, she failed to achieve the second remission. The bone marrow examination was repeated and showed 98% lymphoblasts with the same immunophenotype as primary diagnosis. Her karyotype shifted to 42-46, X, -X[11], add(2)(P25)[7], ?del(6)(q22)[8], -7[7], ?add(8)(q24)[3], +mar1-2[11][cp15] / 46, XX[7]. She was diagnosed with relapsed and refractory B cell ALL and was the first patient enrolled in this phase I clinical trial (NCT02799550). A re-induction chemotherapy was administered (vindesine, idarubicin, pegaspargase, and dexamethasone) which was followed by co-infusion of haplo-identical donor-derived CD19-directed CAR-T cells and G-PBSCs. Both CAR-T cells and G-PBSCs were from her son, who had 6/10 human leukocyte antigen (HLA) loci matched with the patient. Peripheral blood for CAR-T cell preparation was collected ahead of mobilization, and the materials and methods to produce, detect, and quantify CD19-directed CAR-T cells were reported previously [7] and detailed in Additional file 1. Mobilization and apheresis of donor peripheral mononuclear cells were described previously [5]. CAR-T cells were infused from the second to fifth day with escalating doses, and G-PBSCs were infused at the fourth day post chemotherapy (Fig. 1a).

In this study, a 71-year-old relapsed and refractory ALL patient received co-infusion of haplo-identical donor-derived CD19-directed CAR-T cells and G-PBSCs and achieved full donor cell engraftment with mild toxicity.

It is speculated that allogeneic CAR-T cells may have potential benefit of reserving alloreactive-attacking capability to leukemic cells and help overcome limitation of autologous CAR-modified T cells [9]. However, it is unclear whether allogeneic CAR-T cells may persist and proliferate in vivo. In this patient, we persistently detected donor CAR-T cells within twofold of the baseline in host peripheral blood. These results suggested that allogeneic CAR-T cells can survive in vivo although their proliferative efficiency is much lower than that previously reported in which autologous CAR-T cell number amplifies thousands of times in vivo [1]. The anti-CD25 monoclonal antibody has been reported to eliminate donor-specific alloreactive T cells via blockade of the IL-2 binding site and then to prevent and treat acute GVHD [10, 11, 12]. However, no studies have investigated its influence to CAR-T cells. We treated this patient with anti-CD25 antibody on days 14 and 16, which was in accordance with the drop of the allogeneic CAR gene (Figs. 1b and 2e), implying the potential adverse impact of anti-CD25 antibody on CAR-T cell persistence and activation. It is also uncertain whether sustained pancytopenia in this patient suppressed proliferation of CD19-positive cells and consequently attenuated amplification activity of CAR-T cells. In fact, the persistence of CAR-T cells leads to prolonged B cell aplasia [1]. Allogeneic CAR-T cells may also cause donor/recipient-derived B cell clearance (Fig. 2b). Therefore, transient or low level expression of allogeneic CAR-T cells may facilitate donor/recipient-derived B cell reconstitution and simultaneously reserve anti-leukemic effect.

Another major concern is the safety of allogeneic CAR-T cells, particularly for severe GVHD. In this study, the patient developed obvious liver dysfunction and skin rash associated with donor cell engraftment, but no severe hypotension and polypnea were observed although mild elevated levels of cytokines including IL-6, IL-8, IL-10, and TNF-α were found. Therefore, we prefer to name it as a mild “GVHD-like” reaction instead of CRS in this patient. This reaction may correlate with low-intensity chemotherapy conditioning and CAR-T cell-mediated specific anti-leukemic action prior to the G-PBSC infusion, which suppressed patient’s normal immune function and consequently induced development of GVHD under the absence of GVHD prevention. This patient established full donor chimerism following a simple combined chemotherapy and infusions of allogeneic G-PBSCs and CD19-directed CAR-T cells, which may in part attribute to low proportion of recipient CD3⁺CD19⁻ T cells prior to cell infusions (Fig. 2b). Although a preclinical study reported CAR-T cells as a component of conditioning regimen of standard transplantation [13], whether allogeneic CAR-T cells improve engraftment is still undetermined. Further investigations will focus on a clinical study to evaluate the feasibility of this strategy in future.