BackgroundTP53 is the most commonly mutated gene in cancer and codes for the best studied tumor suppressor, p53. MDM2 is involved in the negative regulation of p53 and itself serves as an oncogene, reported to be overexpressed in several cancer tumor types. In this retrospective study, we assessed the occurrence of MDM2 amplification among patients with various types of cancers and its association with clinical factors, other genetic aberrations, and response to targeted therapy in a phase I clinical trial setting.MethodsSamples from patients with advanced solid tumors who had been referred to the MD Anderson phase I clinical trials program between January 2011 and January 2016 were collected and analyzed for MDM2 amplification using FoundationOne’s genomic profiling assay. Patients whose tumors expressed MDM2 amplification were compared to those with tumors of the same histologic types without MDM2 amplification.ResultsWe tested tumors from 523 patients, of which 23 (4.4%) had MDM2 amplification. The highest prevalence of MDM2 amplification was in sarcoma (57%), breast cancer (13%) and bladder cancer (9%). Six patients with liposarcoma were treated on phase I protocol with an MDM2 inhibitor. The most common molecular aberrations co-occurring with MDM2 amplification was CDK4 amplification (70%). TP53 mutation was also detected in 7 patients (30%).ConclusionMDM2 amplification was most commonly associated with liposarcoma. Concomitant alterations in additional genes such as CDK4 amplification and TP53 mutations, along with variable responses to targeted therapies including MDM2 inhibitors, suggest that further combinational studies are needed to target this population.

3511 Background: Kirsten rat sarcoma viral oncogene homolog ( KRAS) p.G12C mutation occurs in approximately 13% of NSCLC and 1%–3% of CRC and other solid tumors. AMG 510 is a first-in-class small molecule that specifically and irreversibly inhibits KRASG12C. Previously, AMG 510 showed preliminary antitumor activity and favorable tolerability in pts with KRAS p.G12C mutant NSCLC or CRC in the phase 1, first-in-human trial. Here, we report results in pts with other tumor types from the same trial. Methods: This study evaluates AMG 510 in pts with locally-advanced or metastatic KRAS p.G12C mutant solid tumors. Key inclusion criteria: KRAS p.G12C mutation via local testing and prior systemic anticancer treatment (tx). Oral daily doses of 180, 360, 720, and 960 mg were tested in the dose escalation, and 960 mg was selected for expansion. Primary endpoint is safety; key secondary endpoints include pharmacokinetics and objective response rate as assessed per RECIST 1.1. Response is assessed every 6 weeks (wks) for 24 wks then every 12 wks thereafter. Results: As of January 8, 2020, 25 pts (9 female, median age 60 years [range: 40–75]) with tumor types categorized by investigators as histology other than NSCLC and CRC were enrolled and dosed (10 pancreatic cancer, 4 appendiceal cancer, 2 endometrial cancer, 2 unknown primary cancer, 1 bile duct cancer, 1 sinonasal cancer, 1 ampullary cancer, 1 small bowel cancer, 1 melanoma, 1 small cell lung cancer, and 1 esophageal cancer). 23 pts received 960 mg dose. 20 pts (80.0%) had ≥2 prior lines of tx. At data cutoff, 13 pts (52.0%) remained on tx; 9 (36.0%) and 3 (12.0%) pts remained on tx for ≥3 and ≥6 months, respectively. Median follow up was 4.3 months (range: 0.1–12.6). Tx-related adverse events (TRAEs) occurred in 9 pts (36.0%). 2 pts (8.0%) had grade 3 TRAEs, including diarrhea (1/25) and pneumonia (1/25, serious AE). No dose-limiting toxicities, grade ≥4, or fatal TRAEs were reported. No TRAEs led to tx discontinuation. 3 pts had not been followed up for ≥7 wks by the data cutoff. 22 pts were followed up for ≥7 wks, and their best overall responses were: 3 confirmed partial response (1 appendiceal, 1 melanoma, and 1 endometrial), 13 stable disease (6 pancreatic, 2 appendiceal, 1 ampullary, 1 bile duct, 1 endometrial, 1 sinonasal, and 1 unknown primary), and 6 progressive disease. Conclusions: AMG 510 was well tolerated and demonstrated clinical activity in pts with advanced KRAS p.G12C mutant solid tumors other than NSCLC and CRC. Clinical trial information: NCT03600883 .

Pazopanib is US FDA approved for the treatment of advanced soft tissue sarcomas. All patients with this disease ultimately develop resistance to therapy. Mechanisms of resistance include activation of the mTOR, histone deacetylase (HDAC), MAPK, and ERBB4 pathways. We hypothesized that combining pazopanib with other targeted agents inhibiting these pathways would increase response rates. We retrospectively evaluated the safety and efficacy of pazopanib plus vorinostat, everolimus, lapatinib or trastuzumab, and MEK inhibitor in patients with advanced sarcoma. The Cancer Geneome Atlas (TCGA) data was analyzed for HDAC, PI3K, HER2, and MAPK/RAS/RAF gene alterations from sarcoma TCGA. Of the 44 advanced sarcoma patients in these trials, 27 (61%) were male; 18 (41%) had bone sarcoma, and 26 (59%) had soft tissue sarcoma. Best response was partial response (PR) in four patients [(overall response rate (ORR) = 9%, 95% confidence interval [CI] 3% to 22%)]. The median progression-free survival (PFS) for all patients was 9.6 weeks (95% CI 8.0 to 15.7 weeks). Analysis of TCGA data revealed HDAC, PI3K, HER2, and MAPK/RAS/RAF gene alterations in 112/243 (46%) of patients predominantly HDAC1–11 (41%) alterations. Pazopanib combinations did demonstrate safety in combination with other agents. TCGA data suggests further evaluation of epigenetic pathway inhibitors in sarcoma.

3112 Background: Despite the broad activity of checkpoint inhibitors across tumor types, primary or secondary resistance after initial response represents a major challenge. Tomivosertib (T), a potent and highly selective inhibitor of the immunosuppressive kinases MNK-1 and 2, blocks expression of checkpoint proteins PD-1, PD-L1, and LAG-3 as well as immunosuppressive cytokines IL-6 and IL-8. In preclinical models, T was shown to trigger an anti-tumor immune response and enhance the activity of checkpoint inhibitors in a T-cell dependent manner. In prior clinical studies, T had an acceptable safety profile as a single agent and in combination with anti-PD-L1 agent avelumab. Methods: Patients experiencing insufficient response (progression or stable disease for 12 weeks or more) to any FDA-approved checkpoint inhibitor in any approved indication were eligible. T at 200 mg oral (PO) BID was added to the existing checkpoint inhibitor until disease progression or unacceptable toxicity was noted. Results: 39 pts (23 male, 16 female) were enrolled across seven cancer types. Median age was 68 (range 42-85). Median prior therapies were 2 (range 1-6). The most common cancers were lung (N = 17), urothelial (N = 6), renal (N = 5) and head and neck (N = 5). 36 pts continued on anti PD-1 antibody (Pembrolizumab and Nivolumab, 18 each) and 3 on anti PD-L-1 antibody (Durvalumab 2, Atezolizumab 1) . The most common grade 3/4 treatment related adverse events occurring in more than 1 pt were alanine aminotransferase increase (2), blood creatine phosphokinase increase (2) and maculo-papular rash (2). 7 patients discontinued treatment (18%) due to adverse events attributable to either drug. Three partial responses (PR) per RECIST 1.1 were observed in pts with previous progression on checkpoint inhibitor therapy, one each in NSCLC (1/17), gastric (1/1) and renal cancer (1/5). 7 NSCLC pts (41%) were progression free for ≥ 24 weeks. All NSCLC patients entered the study with progression by RECIST 1.1 on single agent checkpoint inhibitor prior to adding T. Conclusions: The addition of T to existing checkpoint therapy was well tolerated and manifested clinical activity including objective responses in pts with progression on existing checkpoint inhibitor. A Progression Free Survival rate at 24 weeks of 41% was noted in NSCLC patients. Additional studies evaluating the addition of T to checkpoint inhibitor therapy after progression on anti PD-1 or PD-L-1 therapy are planned. Clinical trial information: NCT03616834 .

No abstract