Overcoming disease refractoriness through CD7-targeting CAR T cells in acute myeloid leukemia Authors: Marco Becilli, Franco Locatelli Published: 2026 Main text Over the last decade, CD19-targeting chimeric antigen receptor (CAR) T cell therapy has emerged as a major breakthrough in the treatment of B-cell malignancies, leading to significantly improved outcomes in both children and adults with relapsed/refractory (R/R) acute lymphoblastic leukemia.1 However, replicating these results in the setting of acute myeloid leukemia (AML), refractory or relapsing after conventional or even innovative approaches, has proven challenging. For patients with R/R AML, outcomes remain dismal, and allogeneic hematopoietic stem cell transplantation (HSCT) represents the only potentially curative option, provided that a new complete remission is achieved, ideally with measurable residual disease (MRD) negativity. The major challenge for CAR T cell therapy in AML lies in identifying targetable antigens that are selectively expressed on leukemia blasts, but are absent on the surface of healthy hematopoietic precursors, progenitors, and stem cells, thereby sparing patients from CAR T cell-induced profound myeloablation or myelosuppression. In collaboration with MediSix Therapeutics Pte. Ltd., we are testing an innovative approach for T cell acute lymphoblastic leukemia (T-ALL) that is based on a bicistronic lentiviral vector encoding a second-generation anti-CD7 41BB/CD3ζ CAR construct and an anti-CD7 protein expression blocker (PEBL), which retains the CD7 antigen in the endoplasmic reticulum and Golgi apparatus, thus preventing its expression on the T cell surface (PCART7). Indeed, the CD7 molecule, a transmembrane glycoprotein member of the immunoglobulin superfamily, is expressed on most healthy T cells and natural killer (NK) cells.2 Consequently, T cells transduced with an anti-CD7 CAR would undergo a reciprocal self-killing effect, known as “fratricide.” This approach, conceived and initially investigated by Campana and colleagues at the National University of Singapore, has been shown to efficiently overcome the fratricide issue, promoting a potent antileukemic effect in vitro and in vivo in T-ALL.3,4 Starting from these preliminary results, we have activated a phase I/II clinical trial currently enrolling children, adolescents, and young adults with R/R CD7-positive T-ALL and T cell lymphoblastic lymphoma (T-LBL) (clinicaltrials.gov: NCT06064903). We hypothesized that patients with CD7-positive AML (approximately 15%–20% of total cases5), who have exhausted all available effective options, might benefit from the same approach (Figure 1). Therefore, we began screening all patients with either newly diagnosed or relapsed AML for CD7 expression on leukemic blasts. Patients were considered eligible to receive PCART7 if >98% of blasts expressed CD7, the same threshold applied in our trial for T-ALL/T-LBL. Three consecutive patients, aged 9–19 years, affected by CD7-positive AML, had failed all available therapies.6 Two of them were resistant to or had relapsed after a treatment combining the CD33-directed antibody-drug conjugate gemtuzumab ozogamicin and the BCL-2 inhibitor venetoclax. One of them had relapsed after HSCT. We considered them for treatment with PCART7 under the Hospital Exemption pathway, a European regulatory tool that allows swift access to a novel advanced therapy medicinal product (ATMP) prepared on a non-routine, individual patient basis, for urgent clinical conditions with an unmet medical need that are potentially life-threatening or seriously detrimental to health.7Figure 1Schematic representation of the cytotoxic activity of a T cell engineered with a lentiviral vector to express a CD7-directed CAR and an anti-CD7 PEBL (PCART7), targeting both CD7-expressing T-cell acute lymphoblastic leukemia (T-ALL) and acute myeloid leukemia (AML) blasts Schematic representation of the cytotoxic activity of a T cell engineered with a lentiviral vector to express a CD7-directed CAR and an anti-CD7 PEBL (PCART7), targeting both CD7-expressing T-cell acute lymphoblastic leukemia (T-ALL) and acute myeloid leukemia (AML) blasts Manufacturing was conducted starting from a cryopreserved leukapheresis product, and T cell transduction was performed after CD4+/CD8+ cell selection using the lentiviral vector supplied by MediSix Therapeutics Pte. Ltd. Manufacturing was feasible for all patients, and all requirements for formal Good Manufacturing Practice (GMP) release of the cell products were successfully met. Bridging therapy failed to control disease progression. After obtaining approval from the Institutional Ethical Committee and the Italian regulatory agency (Agenzia Italiana del Farmaco, AIFA), we infused 1 × 106 transduced T cells/kg of recipient body weight following lymphodepletion with fludarabine and cyclophosphamide. After infusion, we observed robust CAR T cell expansion in peripheral blood, with a peak achieved during the second week. PCART7 cells were also detectable in bone marrow aspirates up to the last evaluation before HSCT. In parallel, circulating CD7+ cells (including leukemic blasts and healthy T and NK cells) declined and were cleared during the second week. We performed bone marrow aspirates for disease assessment before lymphodepletion, showing 85%, 8.5%, and 62% leukemic blasts in the three patients, respectively, and on days +14 and +28 following PCART7 infusion. At post-treatment time points, a morphologic leukemia-free state (MLFS) was achieved in all patients, with leukemic blasts no longer detectable by multiparameter flow cytometry (MFC), demonstrating remarkable anti-leukemia activity of PCART7 against AML blasts. To consolidate these results, we performed HSCT 2 months following PCART7 infusion after confirming MFC MRD negativity in the bone marrow, which has persisted over time. The treatment showed a favorable toxicity profile, with grade 1 and grade 3 cytokine release syndrome (CRS) occurring in two patients and one patient, respectively. Grade 3 CRS resolved after administration of tocilizumab without the need of corticosteroids and occurred in a particularly frail patient who had previously experienced relevant treatment-related toxicities during chemo-immunotherapy. Neither immune effector cell-associated neurotoxicity syndrome (ICANS) nor hemophagocytosis syndrome (IEC-HS) was observed. The limited CAR-related toxicity, regardless of leukemia burden prior to lymphodepletion, may be explained by the reduced cytokine-secreting potential of CAR T cells, without affecting antitumor activity, and is potentially driven by pre-activation mediated by CD7-expressing T lymphocytes during the manufacturing process.4 As expected from lymphodepletion and CAR-related inflammation, we recorded substantial early hematotoxicity (from PCART7 infusion up to day 30), with grade 4 Neutropenia-Immune Effector Cell-Associated Hematotoxicity (N-ICAHT)8 and grade 4 Thtombocytopenia-Immune Effector Cell-Associated Hematotoxicity (T-ICAHT)9 in all patients. Infections (probable invasive pulmonary aspergillosis, Human Herpesvirus-6 (HHV-6) reactivation, and BK virus-associated hemorrhagic cystitis) were all manageable, and both CAR+ T cells and reconstituting CD7- un-transduced T cells contributed to pathogen clearance. Altogether, these preliminary data show that targeting CD7 with an anti-CD7 CAR construct and adopting an anti-CD7 PEBL to prevent reciprocal killing of the transduced T cells without additional genetic manipulation may be a promising strategy to induce deep remissions in a subset of heavily pre-treated, patients with highly refractory AML, including pediatric population. Of note, it has been reported that CD7 is more frequently expressed on leukemia blasts in AML presenting with high-risk features according to the European LeukemiaNet 2022 criteria.5 Formal clinical trials and longer follow-up will be essential to better understand the key clinical and biological features that may help select patients who would particularly benefit from this approach, one of the most important being the minimum threshold of CD7-expressing blasts and the role of CD7 antigen density on the leukemia cell surface in achieving bone marrow MRD negativity. The duration of responses and the risk of CD7-negative relapses, with or without subsequent consolidative HSCT, will contribute to clarify the role of HSCT in the setting of CAR T cell therapy for R/R CD7+ AML. Evaluation of the incidence of prolonged severe hematotoxicity in a larger cohort of patients will help define the contribution of factors other than target antigen expression on healthy hematopoietic progenitors (including prior lines of therapy, pre-lymphodepletion disease burden, bone marrow microenvironment, or specific serum biomarkers) to the development of myelosuppression requiring HSCT as a salvage strategy. Long-term T- and NK-cell lymphopenia represents another theoretical concern with anti-CD7 CAR T cells, potentially rendering HSCT necessary even in patients with prolonged persistence of circulating CAR+ T cells, for whom CAR T cell therapy might otherwise serve as a curative strategy. Notably, reconstitution of CD7-negative T and NK cells has been documented in the first months after infusion of PCART7 in patients with T-ALL,4 and expansion of virus-specific CD7-negative T cells has also been observed in patients with AML6; these findings suggest that HSCT as a rescue strategy for CAR-induced immunodeficiency might be avoided in selected individuals. There have been tremendous efforts to develop AML-targeting immunotherapies, with nearly all approaches directed against shared antigens found on both AML blasts and normal myeloid cells; however, results from phase I studies of CD33- and CD123-targeting CAR T cells have not produced the anticipated outcomes, with response rates of only 25% or less and limited durability.10 Thus, although limited to a minority of patients with AML, the results obtained with PCART7 infusion pave the way for translating the unprecedented success of CAR T cell therapy also to patients with AML. Acknowledgments This manuscript was partly supported by the PALM (Pediatric Acute Leukemia Myeloid) grant from Umberto Veronesi Foundation. Declaration of interests F.L. serves on the scientific advisory boards of Amgen, Novartis, Sanofi, and Vertex and on the speaker’s bureau for Miltenyi Biomedicine, Amgen, Novartis, Bristol Myers Squibb, Gilead, MEDAC, and SOBI.