Abstract CT224: Phase II open-label multi-cohort study evaluating CPI-613 (devimistat) in combination with hydroxychloroquine and 5-fluorouracil or gemcitabine in patients with advanced chemo-refractory solid tumors

Abstract Background: Cancer cells use TCA cycle to generate metabolites as precursors for macromolecule synthesis. Preclinical studies show preventing TCA cycle generation of metabolites decrease tumor growth. CPI-613 (Devimistat) is a non-redox-active lipoate derivative that inhibits mitochondrial TCA cycle enzymes, pyruvate dehydrogenase (PDH) and α-ketoglutarate dehydrogenase (α-KGDH), to impair cells’ ability to utilize the TCA cycle to produce metabolites for growth. As an adaptive mechanism, cancer cells induce autophagy to generate metabolites. Pre-clinical studies report anti-tumor activity when CPI-613 is combined with autophagy inhibitor, hydroxychloroquine (HCQ). Methods: We designed an investigator initiated, open-label, phase II trial for patients (pts) with advanced chemo-refractory colorectal (Cohort 1), pancreatic (Cohort 2), and solid cancers (Cohort 3, exploratory). Cohorts 1+2 received combination of 2000 mg/m2 CPI-613+2400 mg/m2 5-fluorouracil (5FU) IV over 46hrs D1+D15 on 28D cycle. Cohort 3 received 1000 mg/m2 gemcitabine (GEM) + 2000 mg/m2 CPI-613 on D1+D15. For all cohorts, HCQ was given on C1D15 at 400mg PO BID for sequential metabolic analysis. Primary endpoint in cohorts 1+2 was objective response rate (ORR), and Simon’s 2-stage design was used, requiring ≥ 1/10 ORR’s in Stage 1 to proceed to Stage 2, and ≥4/29 ORR’s at trial end to claim the therapy promising. Secondary endpoints: safety, progression-free survival (PFS), overall survival (OS). Correlative analysis: evaluate metabolic signature changes with therapy from blood done at baseline, C1D15 and C2D1. (NCT05733000) Results: Between March 2023-Sept 2024, 26 evaluable patients enrolled, median age 60; 65% female; 54% ECOG=1; 84% White. Cohorts 1 + 2 each enrolled 10 pts in Stage 1. Cohort 3 included 6 pts (3 lung, 2 biliary,1 ovarian) to date. There were no ORR’s in Cohorts 1 + 2. 1 pt in cohort 2 had 38% decrease in target lesions with equivocal new lesion confirmed on next imaging. Best response stable disease in 3(30%) and 2(20%); median PFS 2.3 and 1.7 months; median OS 5.5 and 3.1 months, in cohort 1 and 2 respectively. Overall, 22/26 (85%) pts had at least one TRAE, including 8 (31%) highest grade 3-4 and 14 (54%) highest grade 1-2 AEs. Most frequent TRAEs across patients were vomiting (46%), fatigue (38%), lymphopenia (38%), anemia (35%). Metabolic signature analysis show alteration in PDH, α-KGDH and other TCA cycle functions with therapy, with differential expression based on clinical benefit. Conclusion: To date, in all cohorts, CPI-613 + HCQ + (5FU or GEM) did not add new safety signals compared with 5FU or GEM alone. Early efficacy data suggests clinical benefit for certain patients but no ORR’s were observed in Cohorts 1 and 2, and they will not continue to Stage 2. Further subgroup efficacy, correlative analysis and updates from ongoing cohort 3 will be presented. Citation Format: Devalingam Mahalingam, Carolyn Moloney, Masha Kocherginsky, Claudie Bosc, Katrina Dobinda, Aparna Kalyan, Sheetal M. Kircher, Therese Davis Brown, Ramya Myneni, Mary F. Mulcahy, Al B. Benson, Navdeep Chandel. Phase II open-label multi-cohort study evaluating CPI-613 (devimistat) in combination with hydroxychloroquine and 5-fluorouracil or gemcitabine in patients with advanced chemo-refractory solid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr CT224.