Multicenter, Open-Label, Phase I Study of Azacitidine-CHOP for Patients with Nodal T-Cell Lymphoma with T-Follicular Helper Phenotype

Background

The standard initial treatment for nodal T-cell lymphoma involves combination chemotherapy, commonly using the CHOP regimen (cyclophosphamide, doxorubicin, vincristine, and prednisone). In clinical practices worldwide, this chemotherapy is often followed by an autologous hematopoietic stem cell transplant during the first remission, despite the lack of randomized clinical trials supporting this approach. Complete response (CR) rates with CHOP-based chemotherapy range from 30-40%, with limited evidence suggesting significant improvements from modifying the regimen. Previous phase II or III studies have not significantly improved outcomes, potentially due to including various T-cell lymphoma subtypes. This study focuses on follicular helper T-cell (TFH) origin T-cell lymphomas, which are characterized by specific cytogenetic abnormalities affecting epigenetic regulation, DNA methylation, and T-cell receptor signaling. Key mutations include TET2, DNMT3A, IDH2, and RHOA, suggesting hypomethylating agents like 5-azacytidine as potential treatments.

Methods

This single-arm, multicenter, phase I trial investigates azacitidine combined with CHOP chemotherapy in newly diagnosed patients with three TFH origin T-cell lymphoma subtypes: Angioimmunoblastic T-cell lymphoma (AITL), Follicular T-cell lymphoma (FTCL), and Peripheral T-cell lymphomas, not otherwise specified with a TFH phenotype (PTCL-TFH). Patients receive six cycles of CHOP (cyclophosphamide 750 mg/m², doxorubicin 50 mg/m², vincristine 1.4 mg/m² [max 2 mg], prednisolone 100 mg PO) every three weeks. Azacitidine is administered subcutaneously on days -2, -1, and 1 of each cycle. Phase I evaluates four azacitidine dosage levels (50 mg/m², 75 mg/m², 100 mg/m², and 125 mg/m²) using the Bayesian Optimal Interval (BOIN) design to determine the maximum tolerated dose (MTD), with a target dose-limiting toxicity (DLT) rate of 25%. At each dosage level, the cohort size starts with three individuals. To minimize inappropriate dosage decisions, if the observed Dose-Limiting Toxicity (DLT) at that level is 0.197 or less (≤ 0.197), the dosage will be increased to the next higher level. If the observed DLT is greater than 0.298 (> 0.298), the dosage will be reduced to the lower level. If the DLT is between 0.197 and 0.298, the current dosage level will be maintained. Patients in phase I receive maintenance azacitidine (75 mg/m²) subcutaneously on days 1-5 every four weeks for 12 cycles.

Results

As of the analysis, 16 patients (nine male, seven female; median age 67, range 46-82) were enrolled, with diagnoses including AITL (n = 8), PTCL-TFH (n = 7), and FTCL (n = 1). Most patients had stage III/IV disease (n = 13) and ECOG performance grade 0 (n = 8). Dose level 1 completed treatment without DLT, while dose levels 2 and 3 involved seven and eight patients, respectively. Two patients at dose level 3 experienced DLT, leading to dose reductions in azacitidine and other CHOP components. Six patients experienced grade 4 neutropenia, while grade 3 toxicities included neutropenia, febrile neutropenia, thrombocytopenia, anemia, anorexia, deep vein thrombosis, AST/ALT increase, and sepsis. There were no treatment-related mortalities or discontinuations due to severe toxicities. Non-hematologic grade 1/2 toxicities were common, including constipation, nausea, weight loss, diarrhea, injection site reactions, and peripheral neuropathy. Of the 13 patients evaluated for response, 11 showed a response including complete response (n = 9) and partial response (n = 2). All responders continued maintenance therapy, with the phase 2 dose for azacitidine set at level 3.

Conclusions

This study is the first to establish the dosage of subcutaneous azacitidine combined with CHOP for newly diagnosed nodal TFH-origin T-cell lymphomas. The addition of azacitidine at 75 and 100 mg/m² on days -2, -1, and 1 to CHOP was feasible and effective, with manageable toxicity profiles. A phase II study will be conducted using the recommended phase II dose (RP2D) from phase I, with the primary endpoint being the complete response (CR) rate. Using Simon's Minimax two-stage design, the phase II study will test a null hypothesis CR rate of 40% against a target CR rate of 60%, with a significance level of 0.05 and 80% power. Including a 5% dropout rate, the total target enrollment is 41 patients, incorporating those from phase I receiving the RP2D.