AUTHORS
Issa F. Khouri, MD, Dan Li, Ph.D.2, Alison Gulbis, PharmD3, Denai Milton4, Bouthaina S. Dabaja, MD5, Celina Ledesma6, Qing Ma, Ph.D.7
AFFILIATIONS
1 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
2 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
3 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
4 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
5 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
6 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
7 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
ABSTRACT
Background:
Post-transplant cyclophosphamide (PTCy) is an effective platform for graft-versus-host disease (GVHD) prophylaxis after allogeneic hematopoietic cell transplantation (allo-HCT). Alternative approaches that maintain GVHD control while reducing cyclophosphamide exposure and favorably modulating immune reconstitution warrant investigation. Post-transplant bendamustine (PT-BEN) has demonstrated immunomodulatory activity in preclinical transplantation models. We conducted a phase I study evaluating stepwise substitution of PTCy with PT-BEN and incorporated preclinical and longitudinal immune studies to explore its biologic effects.
Methods:
Adults with hematologic malignancies undergoing reduced-intensity allo-HCT from matched sibling, matched or mismatched unrelated, or haploidentical donors were enrolled in a single-center phase I study. Conditioning consisted of fludarabine, melphalan, and 200-cGy total-body irradiation. Four day +3 dose levels (DLs) were evaluated: DL1, cyclophosphamide (CY) 50 mg/kg; DL2, CY 30 mg/kg plus bendamustine (BEN) 30 mg/m²; DL3, CY 20 mg/kg plus BEN 60 mg/m²; and DL4, BEN 100 mg/m² without CY. All patients received BEN 100 mg/m² on day +4. Tacrolimus and mycophenolate mofetil (MMF) were initially started on day +5; following a protocol amendment in July 2024, both were initiated on day 0. Dose finding used a Bayesian optimal interval design.
Preclinical studies assessed BEN-mediated inhibition of human alloreactive T-cell proliferation, preservation of regulatory T cells (Tregs), and survival in a human PBMC-induced xenogeneic GVHD model. Exploratory longitudinal immune profiling at the recommended phase II dose (RP2D) evaluated Treg/CD8 and Treg/CD8 effector T-cell ratios before HCT and at 1, 2, 3, and 6 months after HCT.
Results:
Twenty-six patients were enrolled: 4 at DL1, 14 at DL2 (3 during dose escalation and 11 in the expansion cohort), 5 at DL3, and 3 at DL4. Median age was 52 years, and 40% had an HCT-CI ≥3. Diagnoses included acute myeloid leukemia (n=10, 38%), myelodysplastic syndrome (n=5, 19%), acute lymphoblastic leukemia (n=4, 15%), lymphoma/CLL with Richter transformation (n=4, 15%), and chronic myeloid leukemia (n=3, 12%). Donors were haploidentical in 12 (46%), matched unrelated in 9 (35%), mismatched unrelated in 3 (12%), and matched sibling in 2 (8%). All patients achieved initial engraftment, with median neutrophil and platelet recovery at 13 and 15 days, respectively.
DL2 was selected as the RP2D based on the overall balance of safety, engraftment, and GVHD control, with increasing toxicity observed with greater substitution of CY at higher dose levels. No grade III-IV acute GVHD or moderate/severe chronic GVHD occurred at DL2. At the RP2D, day-100 non-relapse mortality (NRM) was 7%, GVHD-free, relapse-free survival (GRFS) was 71%, and 1-year overall survival (OS) was 80%. For the overall cohort, with a median follow-up of 29.7 months among survivors, OS was 81% at both 1 and 2 years; NRM was 8% at day 100 and 19% at both 1 and 2 years. Grade III-IV acute GVHD occurred exclusively at DL4.
In preclinical studies, BEN inhibited human CD4+ and CD8+ T-cell proliferation while relatively preserving CD4+FoxP3+ Tregs and improved survival in the xenogeneic GVHD model. In patients treated at DL2, the Treg/CD8 ratio increased significantly over time (overall P=0.0147), as did the Treg/CD8 effector T-cell ratio (overall P=0.0099). Both ratios were significantly increased at 3 months compared with pre-HCT (P=0.0047 and P=0.0062, respectively). A sensitivity analysis restricted to patients with complete serial sampling demonstrated concordant longitudinal changes.
Conclusions:
Partial substitution of PTCy with PT-BEN was feasible after allo-HCT, and DL2—CY 30 mg/kg plus BEN 30 mg/m² on day +3 followed by BEN 100 mg/m² on day +4—was identified as the RP2D. The absence of grade III-IV acute GVHD and moderate/severe chronic GVHD at DL2, together with preserved engraftment, encouraging GRFS, and durable overall survival, supports further prospective evaluation of this approach. Preclinical suppression of alloreactive T-cell responses with relative preservation of Tregs, together with concordant longitudinal increases in Treg/CD8 and Treg/CD8 effector ratios in patients, provides biologic support for the immunomodulatory activity of PT-BEN.
DISCLOSURES
Issa F. Khouri, MD: No relevant conflicts of interest to declare.