Phase I study of escalating doses of carfilzomib with HyperCVAD in patients with newly diagnosed acute lymphoblastic leukemia

Outcomes remain poor for many adults with acute lymphoblastic leukemia (ALL) with conventional chemotherapy. Hyperfractionated cyclophosphamide, vincristine, doxorubicin and dexamethasone (HyperCVAD) is a multi-agent and multi-course high-intensity chemotherapy regimen that results in high complete remission (CR) rates, long-term survival in 30–40%,1 and measurable/minimal residual disease (MRD) negativity in 50–60% adults with ALL.2, 3 As an established backbone regimen for ALL treatment, the HyperCVAD regimen has evolved to incorporate recent advances in anti-ALL therapy, and new agents continue to be tested in combination with the HyperCVAD backbone in efforts to improve outcomes.4, 5 Bortezomib, a first-generation proteasome inhibitor, has demonstrated synergism with chemotherapy, including cyclophosphamide, vincristine, cytarabine, rituximab,6, 7 and as an inhibitor of NF-κB, is hypothesized to act that results in accumulation of pro-apoptotic proteins that could sensitize ALL cells to dexamethasone and doxorubicin.8 Bortezomib has been tested clinically in combination with HyperCVAD for the treatment of multiple myeloma and mantle cell lymphoma and the combinations have been tolerable without significant added toxicity from bortezomib.7, 9 Carfilzomib is a next-generation irreversible and selective inhibitor of the chymotrypsin-like activity of the proteasome, which shows increased specificity, potency, and cellular apoptotic sensitivity compared to bortezomib in preclinical studies of ALL.10 Carfilzomib shows promising preclinical activity in ALL ex vivo models, and has synergistic activity with dexamethasone.10, 11 Therefore, we hypothesized that the addition of carfilzomib to HyperCVAD would be safe and feasible and could improve outcomes, including improved MRD responses, for adults with ALL. To test our hypothesis, we conducted a phase one study testing the safety, tolerability and preliminary efficacy of the addition of carfilzomib to the HyperCVAD regimen (NCT02293109). The study was approved by the UC Davis institutional review board, and all patients provided informed consent. The primary objectives of this study were to determine safety, tolerability and recommended phase two dose (R2PD) of carfilzomib in combination with HyperCVAD. Adverse events (AEs) were graded using NCI Common Terminology for Adverse Events version 4.03 and assigned an association to Hyper-CVAD or carfilzomib. The secondary objectives were to determine the rate of CR and rate of MRD negativity. Bone marrow biopsies to assess disease status were performed at screening, after cycle one, after cycle four and as clinically indicated. The CR was defined in accordance with established response criteria for ALL patients.12 The MRD was measured with multiparameter flow cytometry (MFC) and <0.01% was considered negative.13 Descriptive statistics were computed using Microsoft Excel 2020 (Microsoft, Seattle, WA). Patients were followed for safety, relapse and survival for up to 28 days after completion of study therapy. Key eligibility criteria included treatment naïve Philadelphia (Ph) chromosome negative ALL patients aged 18–64 years with adequate left ventricular, renal, and hepatic function, in addition to an Eastern Cooperative Oncology Group (ECOG) Performance Status of 0–2. Both B-cell and T-cell ALL subtypes were eligible. Eligible patients received up to four 28-day cycles of the standard HyperCVAD regimen (e.g., two cycles each of HyperCVAD courses A and B), including central nervous system prophylaxis.1, 5, 14 Rituximab was added for CD20 positive patients.4 For each of the four HyperCVAD cycles, four doses of carfilzomib were added on days 0, 1, 7 and 8 (Figure S1A). Dose escalation commenced using a 3 + 3 design with carfilzomib dose levels (DL) of 20 mg/m2 (DL1), 27 mg/m2 (DL2) and 36 mg/m2 (DL3). Dose-limiting toxicities (DLTs) were assessed through completion of cycle two and included protocol-defined non-hematologic events at least possibly related to study therapy (Figure S1B). Patients not evaluable for DLT were replaced. After the conclusion of planned study therapy, patients could continue to receive further chemotherapy, maintenance therapy and/or allogeneic hematopoietic cell transplant according to patient and physician preference. Ten patients were enrolled, three at DL1, three at DL2 and four at DL3. Patient demographics and disease characteristics are shown in Table S1. Median age was 38 [range 23–61]. Fifty percent (n = 5) were female, and 60% (n = 6) were Hispanic. Eighty percent (n = 8) had B-ALL, and 20% (n = 2) had T-ALL. Seventy percent (n = 7) had abnormal cytogenetics, including complex IgH gene rearrangements (n = 4). Fifty percent (n = 5) had CD20 positive disease and received rituximab. One patient was found to have CNS involvement during cycle one CSF assessments. No patients experienced a DLT. The median number of cycles completed on study was four (range 1–4). Twenty percent (n = 2) of patients had a dose modification related to AE (one in cycle one and the other in cycle three) and 30% (n = 3) had a dose delay (two related to AE). One of the patients with a dose modification was replaced after cycle one for AE (proteinuria) not meeting DLT criteria. Treatment-emergent AEs (TEAEs) regardless of attribution and occurring in at least 20% of patients are shown in Table S2. All patients (n = 10) had a TEAE, with blood and lymphatic system, metabolism and nutrition disorders and infections and infestations occurring most frequently at grade three or higher. The TEAEs at least possibly, probably or definitely related to study therapy occurring at ≥ grade 3 in ≥20% of participants are shown in Table 1. All patients (n = 10) experienced a grade 3–4 TEAE at least possibly related to study therapy, including anemia (n = 10), leukopenia (n = 10), neutropenia (n = 10), thrombocytopenia (n = 10), lymphopenia (n = 9), febrile neutropenia (n = 7), hypocalcemia (n = 2) and hyponatremia (n = 2). There were no grade five events and no cardiac AEs apart from related grade one sinus tachycardia in one patient. There were three events of dyspnea, all grade one or two, and four events of paresthesia/neuropathy, all grade one. Treatment-emergent serious adverse events (SAEs) at least possibly, probably or definitely related to study therapy occurring at events ≥ grade three in ≥20% of participants are shown in Table 1. Seventy percent (n = 7) of patients experienced a SAE, and the most frequent SAEs included febrile neutropenia (n = 6), leukopenia (n = 2) and thrombocytopenia (n = 2). The SAEs regardless of attribution are shown in Table S3 and include additional events occurring in one patient each, such as sepsis, lung infection, hyponatremia, and proteinuria. Preliminary efficacy results for the carfilzomib plus HyperCVAD regimen are shown in Figure 1. The CR rate was 90% (n = 9) after cycle one. One additional patient achieved CR after cycle three resulting in a 100% CR rate after four cycles. Using MFC with a cutoff for negativity at <0.01%, 80% of patients (n = 8) achieved MRD negativity. Seven achieved MRD negativity after cycle one and one became MRD negative after cycle four. One B-ALL patient had reappearance of MRD after cycle four, leading to a 70% MRD negative rate (n = 7) after four cycles. In terms of ALL subtypes, seven out of the eight B-ALL patients (88%) and one out of the two T-ALL patients (50%) achieved MRD negativity at any time on study. None of the patients experienced morphologic relapse or death during the study treatment and 28-day follow-up period. Overall, our phase one study demonstrated that the addition of carfilzomib to the HyperCVAD regimen was safe and tolerable for untreated ALL patients. There were no DLTs observed with carfilzomib doses up to 36 mg/m2, and nine out of ten patients were able to complete the entire course of planned study therapy. Treatment-related AEs and SAEs were consistent with HyperCVAD historical controls,1 including primarily hematologic toxicities. There were no treatment-related deaths or any deaths on study, and cardiac and neuropathy AEs were uncommon and low-grade. In addition, the combination shows promising preliminary efficacy with a 100% CR rate after four cycles and 80% of patients achieving an MRD negative CR. Studies have consistently shown that achieving MRD negativity is associated with improved outcomes in ALL.3, 15 The carfilzomib-HyperCVAD regimen appeared particularly effective for the B-ALL subpopulation with 88% of patients achieving MRD negativity, which compares favorably with HyperCVAD historical controls,2, 3 with an MRD negativity rate of 53% in Ph-negative B-ALL after 90 days in patients treated with HyperCVAD alone.2 Furthermore, 70% of the patients achieved MRD negativity after one cycle, compared to the published rate of 27% after one cycle with standard HyperCVAD therapy.2 Additional molecular or cellular studies may help elucidate the mechanisms underlying the more favorable responses seen here with B-cell compared to T-cell disease. Based on the current data, the RP2D level of carfilzomib in combination with HyperCVAD is 36 mg/m2. Notably, no maximum tolerated dose was achieved, suggesting that additional dose levels, such as the labeled dose of 56 mg/m2 twice weekly, and an expansion cohort could be explored to gather additional safety and preliminary efficacy data and test whether the response in T-ALL is potentially dose-related. Despite the small sample size and short follow-up of our study, the early and deep responses seen in the majority of patients is encouraging and warrants future phase two studies to confirm efficacy and explore survival outcomes. In summary, the addition of carfilzomib to HyperCVAD is safe and tolerable in patients with untreated Ph-negative ALL. Promising preliminary efficacy in CR rate and achievement of MRD negativity was observed, particularly in the B-ALL subpopulation. These results support further study of HyperCVAD plus carfilzomib in Ph-negative B-ALL. We thank the patients for participating in this study and Takeda for funding this study. We thank the University of California, Davis Comprehensive Cancer Center (UCDCCC) Office of Clinical Research for support in the conduct and analysis of the trial. The abstract form of preliminary data from this study was presented at ASH 2019: Abstract #3884. ACUTE LYMPHOBLASTIC LEUKEMIA: THERAPY, EXCLUDING TRANSPLANTATION. This study was supported by the UCDCCC Support Grant awarded by the National Cancer Institute (NCI P30CA093373). Conception and design: Brian A. Jonas, Joseph M. Tuscano and Mehrdad Abedi; acquisition of data: Brian A. Jonas, Aaron S. Rosenberg, Rasmus T. Hoeg, Joseph M. Tuscano and Mehrdad Abedi; analysis and interpretation of data: Brian A. Jonas, Samantha C. Fisch, Aaron S. Rosenberg, Rasmus T. Hoeg, Joseph M. Tuscano and Mehrdad Abedi; writing, review and/or revision of the manuscript: Brian A. Jonas, Samantha C. Fisch, Aaron S. Rosenberg, Rasmus T. Hoeg, Joseph M. Tuscano and Mehrdad Abedi. All authors approved of the final version. BA Jonas has served in a consulting/advisory role for AbbVie, Amgen, Celgene, Genentech, GlycoMimetics, Jazz, Takeda, Tolero, and Treadwell. He has received travel support from 47, AbbVie, Amgen, and Glycomimetics. He has received research funding to his institution from AbbVie, Accelerated Medical Diagnostics, AROG, Celgene, Daiichi Sankyo, F. Hoffmann-La Roche, Forma, Genentech/Roche, GlycoMimetics, Hanmi, Incyte, Jazz, LP Therapeutics, Pfizer, Pharmacyclics, and Sigma Tau. SC Fisch reports no conflicts of interest. AS Rosenberg has received research funding from Amgen. He has participated in Speakers Bureaus for Jannsen and Millenium-Takeda. He has served in a consulting/advisory role for Seattle Genetics and Karyopharm. RT Hoeg reports no conflicts of interest. JM Tuscano has received research funding from Celgene, Novartis, Achrotech, Pharmacyclics, Genentech and Takada. He has received honoraria from Celgene, Amgen and Seattle Genetics. M Abedi has participated in Speakers Bureaus for AbbVie, Celgene, BMS and Gilead. The data that support the findings of this study are available from the corresponding author upon reasonable request. Figure S1. Study treatment schema (A) and dose-limiting toxicity criteria (B) Table S1. Patient demographics and disease characteristics Table S2. Treatment-emergent adverse events (AEs) regardless of attribution occurring in ≥20% of participants Table S3. All treatment-emergent serious adverse events (SAEs) regardless of attribution Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.