Background: AZD0156 is an oral inhibitor of ATM, a serine threonine kinase that plays a key role in DNA damage response (DDR) associated with double-strand breaks. Topoisomerase-I inhibitor irinotecan is used clinically to treat colorectal cancer (CRC), often in combination with 5-fluorouracil (5FU). AZD0156 in combination with irinotecan and 5FU was evaluated in preclinical models of CRC to determine whether low doses of AZD0156 enhance the cytotoxicity of irinotecan in chemotherapy regimens used in the clinic. Methods: Anti-proliferative effects of single-agent AZD0156, the active metabolite of irinotecan (SN38), and combination therapy were evaluated in 12 CRC cell lines. Additional assessment with clonogenic assay, cell cycle analysis, and immunoblotting were performed in 4 selected cell lines. Four colorectal cancer patient derived xenograft (PDX) models were treated with AZD0156, irinotecan, or 5FU alone and in combination for assessment of tumor growth inhibition (TGI). Immunofluorescence was performed on tumor tissues. The DDR mutation profile was compared across in vitro and in vivo models. Results: Enhanced effects on cellular proliferation and regrowth were observed with the combination of AZD0156 and SN38 in select models. In cell cycle analysis of these models, increased G2/M arrest was observed with combination treatment over either single agent. Immunoblotting results suggest an increase in DDR associated with irinotecan therapy, with a reduced effect noted when combined with AZD0156, which is more pronounced in some models. Increased TGI was observed with the combination of AZD0156 and irinotecan as compared to single-agent therapy in some PDX models. The DDR mutation profile was variable across models. Conclusions: AZD0156 and irinotecan provide a rational and active combination in preclinical colorectal cancer models. Variability across in vivo and in vitro results may be related to the variable DDR mutation profiles of the models evaluated. Further understanding of the implications of individual DDR mutation profiles may help better identify patients more likely to benefit from treatment with the combination of AZD0156 and irinotecan in the clinical setting.

Purpose: In this single-institution phase II investigator-initiated study, we assessed the ability of MAPK and VEGF pathway blockade to overcome resistance to immunotherapy in microsatellite-stable metastatic colorectal cancer (MSS mCRC). Patients and Methods: Patients with MSS, BRAF wild-type mCRC who progressed on ≥2 prior lines of therapy received pembrolizumab, binimetinib, and bevacizumab until disease progression or unacceptable toxicity. After a safety run-in, patients were randomized to a 7-day run-in of binimetinib or simultaneous initiation of all study drugs, to explore whether MEK inhibition may increase tumor immunogenicity. The primary endpoint was objective response rate (ORR) in all patients combined (by Response Evaluation Criteria in Solid Tumors v1.1). Results: Fifty patients received study drug treatment; 54% were male with a median age of 55 years (range, 31–79). The primary endpoint, ORR, was 12.0% [95% confidence interval (CI) 4.5%–24.3%], which was not statistically different than the historical control data of 5% (P = 0.038, exceeding prespecified threshold of 0.025). The disease control rate was 70.0% (95% CI, 55.4%–82.1%), the median progression-free survival 5.9 months (95% CI, 4.2–8.7 months), and the median overall survival 9.3 months (95% CI, 6.7–12.2 months). No difference in efficacy was observed between the randomized cohorts. Grade 3 and 4 adverse events were observed in 56% and 8% of patients, respectively; the most common were rash (12%) and increased aspartate aminotransferase (12%). Conclusions: Pembrolizumab, binimetinib, and bevacizumab failed to meet its primary endpoint of higher ORR compared with historical control data, demonstrated a high disease control rate, and demonstrated acceptable tolerability in refractory MSS mCRC.

(1) Background: Histone deacetylases (HDACs) play a critical role in epigenetic signaling in cancer; however, available HDAC inhibitors have limited therapeutic windows and suboptimal pharmacokinetics (PK). This first-in-human phase I dose escalation study evaluated the safety, PK, pharmacodynamics (PDx), and efficacy of the oral Class I-targeting HDAC inhibitor bocodepsin (OKI-179). (2) Patients and Methods: Patients (n = 34) with advanced solid tumors were treated with OKI-179 orally once daily in three schedules: 4 days on 3 days off (4:3), 5 days on 2 days off (5:2), or continuous in 21-day cycles until disease progression or unacceptable toxicity. Single-patient escalation cohorts followed a standard 3 + 3 design. (3) Results: The mean duration of treatment was 81.2 (range 11–447) days. The most frequent adverse events in all patients were nausea (70.6%), fatigue (47.1%), and thrombocytopenia (41.2%). The maximum tolerated dose (MTD) of OKI-179 was 450 mg with 4:3 and 200 mg with continuous dosing. Dose-limiting toxicities included decreased platelet count and nausea. Prolonged disease control was observed, including two patients with platinum-resistant ovarian cancer. Systemic exposure to the active metabolite exceeded the preclinical efficacy threshold at doses lower than the MTD and was temporally associated with increased histone acetylation in circulating T cells. (4) Conclusions: OKI-179 has a manageable safety profile at the recommended phase 2 dose (RP2D) of 300 mg daily on a 4:3 schedule with prophylactic oral antiemetics. OKI-179 is currently being investigated with the MEK inhibitor binimetinib in patients with NRAS-mutated melanoma in the phase 2 Nautilus trial.

Purpose: In this preclinical human immune system patient-derived xenograft (HIS-PDX) model and phase II clinical trial, we assessed evorpacept (anti-CD47 engineered fusion protein with inactive Fc), cetuximab, and pembrolizumab (triple therapy) in microsatellite-stable (MSS) colorectal cancer. Patients and Methods: HIS BALB/c-Rag2nullIl2rγnullSirpαNOD mice with PDXs were treated with triple therapy or its components. Patients with refractory MSS colorectal cancer were treated with triple therapy in a safety run-in (stage 1) followed by expansion (stage 2, planned N = 42). The co-primary objectives were to determine the recommended dose of evorpacept and objective response rate (vs. historic control). Results: In HIS-PDX mice, triple therapy decreased the growth of MSS colorectal cancer tumors and increased tumor-infiltrating CD8+ T cells. Sixteen patients were treated on the clinical trial across two evorpacept dose levels: N = 12 in stage 1 and N = 4 in stage 2. Trial enrollment was terminated early because of safety concerns (one treatment-related grade 5 event each of hemophagocytic lymphohistiocytosis and cytokine release syndrome). Otherwise, the adverse event profile was as expected. Among all patients, the objective response rate was 6.3%; formal hypothesis testing was not performed. The disease control rate was 12.5%, the median progression-free survival was 2.3 months, and the median overall survival was 10.9 months. Blood- and tumor-based clinical trial correlative analyses identified innate and adaptive immune system activation. Conclusions: Whereas triple therapy demonstrated evidence of efficacy in refractory MSS colorectal cancer, safety concerns halted enrollment. Further investigation is necessary to determine the optimal use of CD47-targeted therapies in MSS colorectal cancer. Significance: Evorpacept, cetuximab, and pembrolizumab demonstrated antitumor activity in a preclinical HIS-PDX model and clinical trial in refractory MSS colorectal cancer; however, immune-related adverse events prompted early termination of study enrollment. Evidence of innate and adaptive antitumor immune activation was identified. The role of SIRPα/CD47 blockade in the treatment of MSS colorectal cancer needs to be further elucidated in future trials.

249 Background: Vantictumab is a fully human monoclonal antibody that inhibits Wnt pathway signaling through binding FZD1, 2, 5, 7, and 8 receptors. A phase Ib study of VAN in combination with Nab-P and G was performed in patients with untreated stage IV pancreatic adenocarcinoma. Methods: Patients received VAN at escalating doses (3-7 mg/kg) in combination with standard dosing of Nab-P and G according to a 3+3 design. Due to fragility fractures occurring in this and other related clinical trials, dosing on an every 2 week schedule in cohorts 1 and 2 was transitioned to every 4 week dosing for cohorts 3 through 5. In these later cohorts, a minimum of six patients were treated at each dose level and additional criteria for maximum tolerated dose (MTD) integrating bone safety parameters were added. The bone safety plan was also revised for these cohorts. Sequential dosing of VAN followed by Nab-P and G was explored in cohort 5. Results: Thirty-one patients (52% male, 48% female) were enrolled and treated in 5 dosing cohorts. Median age was 66. Most common study-treatment related adverse events were nausea (68%) and fatigue (52%). One dose limiting toxicity (DLT) event occurred in the study population—grade 3 dehydration in 1 of 9 patients in cohort 4 (5 mg/kg q4w). Fragility fractures attributed to VAN occurred in two patients in cohort 2 (7 mg/kg q2w). Once the dosing schedule was revised to every 4 weeks, the maximum administered VAN dose was 5 mg/kg. No fragility fractures attributed to VAN occurred in these cohorts; pathologic fracture not attributed to VAN was documented in 2 patients. The study was terminated due to lack of an acceptable therapeutic index. Partial response was documented in 13 patients (42%) and stable disease in 11 (36%). Conclusions: The MTD of VAN plus Nab-P and G was not determined, but the maximum administered dose (MAD) of VAN, 7 mg/kg every 2 weeks, was considered unsafe related to bone toxicity, a known effect of WNT inhibition. After the study was revised, the MAD was 5 mg/kg every 4 weeks, with no protocol-specified bone toxicity observed (n = 16). Clinical trial information: NCT02005315.

Background: This study further evaluated the safety and efficacy of the combination of alisertib and sapanisertib in an expansion cohort of patients, including a subset of patients with refractory pancreatic adenocarcinoma, with further evaluation of the pharmacodynamic characteristics of combination therapy. Methods: Twenty patients with refractory solid tumors and 11 patients with pancreatic adenocarcinoma were treated at the recommended phase 2 dose of alisertib and sapanisertib. Adverse events and disease response were assessed. Patients in the expansion cohort were treated with a 7-day lead-in of either alisertib or sapanisertib prior to combination therapy, with tumor tissue biopsy and serial functional imaging performed for correlative analysis. Results: Toxicity across treatment groups was overall similar to prior studies. One partial response to treatment was observed in a patient with ER positive breast cancer, and a patient with pancreatic cancer experienced prolonged stable disease. In an additional cohort of pancreatic cancer patients, treatment response was modest. Correlative analysis revealed variability in markers of apoptosis and immune cell infiltrate according to lead-in therapy and response. Conclusions: Dual targeting of Aurora A kinase and mTOR resulted in marginal clinical benefit in a population of patients with refractory solid tumors, including pancreatic adenocarcinoma, though individual patients experienced significant response to therapy. Correlatives indicate apoptotic response and tumor immune cell infiltrate may affect clinical outcomes.