Background: Myocardial Gal3 (galectin-3) expression is associated with cardiac inflammation and fibrosis. Increased Gal3 portends susceptibility to heart failure and death. There are no data reporting the causative role of Gal3 to mediate cardiac fibro-inflammatory response and heart failure. Methods: We developed a cardioselective Gal3 gain-of-function mouse ( Gal3+/+ ) using α-myosin heavy chain promotor. We confirmed Gal3-transgene expression with real-time polymerase chain reaction and quantified cardiac/circulating Gal3 with Western blot and immunoassays. We used echocardiogram and cardiac magnetic resonance imaging to measure cardiac volumes, function, and myocardial velocities. Ex vivo, we studied myocardial inflammation/fibrosis and downstream TGF (transforming growth factor) β1-mRNA expression. We examined the effects of acute myocardial ischemia in presence of excess Gal3 by inducing acute myocardial infarction in mice. Two subsets of mice including mice treated with N-acetyl-seryl-aspartyl-lysyl-proline (a Gal3-inhibitor) and mice with genetic Gal3 loss-of-function ( Gal3 −/−) were studied for comparative analysis of Gal3 function. Results: Gal3+/+ mice had increased cardiac/circulating Gal3. Gal3+/+ mice showed excess pericardial fat pad, dilated ventricles and cardiac dysfunction, which was partly normalized by N-acetyl-seryl-aspartyl-lysyl-proline. Cardiac magnetic resonance imaging showed reduced myocardial contractile velocities in Gal3+/+ . The majority of Gal3+/+ mice did not survive acute myocardial infarction, and the survivors had profound cardiac dysfunction. Myocardial histology of Gal3+/+ mice showed macrophage/mast-cell infiltration, fibrosis and higher TGFβ1-mRNA expression, which were mitigated by both Gal3 gene deletion and N-acetyl-seryl-aspartyl-lysyl-proline administration. Conclusions: Our study shows that cardioselective Gal3 overexpression leads to multiple cardiac phenotypic defects including ventricular dilation and cardiac dysfunction. Pharmacological Gal3 inhibition conferred protective effects with reduction of inflammation and fibrosis. Our study highlights the importance of translational studies to counteract Gal3 function and prevent cardiac dysfunction.

2545 Background: While Immune checkpoint inhibitors (ICI) have revolutionized the field of oncology, the benefits have come at the cost of serious side effects known as immune-related adverse events (irAEs). Approaches that can predict patients’ susceptibility to irAEs are key to their early detection and management. In the present study, we investigate the association between irAEs reported during ICI therapy across multiple cancer types and markers of tumor immune response. Our primary objective is to explore potential biomarkers for assessing patients’ risk of irAEs. Methods: 472 patients were evaluated who had tumor immune profiling performed paraffin embedded formalin fixed archival tumor biopsy samples using Omniseq Immune Report Card (IRC) and subsequently underwent ICI therapy. The IRC consisted of enumeration of tumor infiltrating lymphocytes (TILs) by immunohistochemistry (IHC) and TIL-associated genes by RNA-Seq, PD-L1 expression by IHC, and tumor mutational burden (TMB) by DNA-Seq. irAE type and grade were determined based on retrospective chart review. Fisher’s exact test was used to determined statistically significant associations between immune markers and irAE development. Results: Patients with lung (55%), ovarian (9%), and melanoma (5%) cancers constituted the majority of the cases. The median age of patients was 61, with 56% being female and 44% male. Most patients underwent treatment with (94%). irAEs developed in 36% of patients, with 2% of patients developing high-grade irAEs (Grade 3 or 4). Skin (11%), thyroid (10%), and GI (9%), were the most commonly affected organ systems. Increased TILs were associated with increased risk for any irAE (p = 0.04). A stronger association was noted in patients who underwent anti-PD-1/L1 monotherapy (p = 0.01) and/or in cases of lung cancer (p = 0.01). Interestingly, subanalyses by gender showed a statistically significant correlation between increased TILs and risk for any irAE in males (p = 0.006) but not in females (p = 0.63). High PD-L1 (defined as > 70% by IHC) was also significantly associated with increased risk for any irAE (p = 0.03). Subanalyses by gender and age again showed a similar association in females (p = 0.0002) and/or patients < 65 years (p = 0.04). high TMB and any irAE in female patients (p = 0.01) and in breast cancer cases (p = 0.03). On multivariate analysis, TILs by IHC appeared to be the strongest predictor of irAEs (p = 0.03). Conclusions: The tumor immune microenvironment (TME) has been shown to influence response to ICI, yet its association with irAEs has not been well studied. Our analysis sheds light on potential TME predictors for irAE in patients receiving ICI therapy. Further studies are needed to deepen our understanding of immune toxicity and to develop tools for identifying patients who are at risk.

6519 Background: HRQoL data in cancer clinical trials can inform tolerability of new drugs, facilitate informed decision-making, and influence health care and policy decisions, but are frequently underreported. We reviewed all registration trials that informed Food and Drug Administration (FDA) approval between 2015-2020 for latency and quality of HRQoL reporting. Methods: HRQoL data for each clinical trial associated with FDA drug approval between 7/2015-5/2020 was collected retrospectively from multiple sources including the FDA and clinicaltrials.gov website, conference abstracts, and journal manuscripts. The aim of the study was to analyze the proportion of trials reporting HRQoL, quality of HRQoL data, latency between FDA approval and first reporting of HRQoL data, and association between changes in HRQOL and overall survival (OS) and progression-free survival (PFS) outcomes. Results: Of the 259 trials involving 245 drug approvals, majority involved solid tumors (61.4%), were phase III (59.1%), and led to approval based on a non-PFS/OS endpoint (52.9%). HRQoL was a pre-specified endpoint in 55.2% and reported in 49.8% trials. HRQoL data was published by the time of FDA approval in only 41.8% cases, 24.8% reported HRQoL data > 12 mo after approval. Further, among trials reporting HRQoL (n = 129), HRQoL data was first reported in the primary paper in only 34.1%, and either in an ancillary paper in 41.9% or an ancillary abstract in 24% trials. Of the 129 trials with HRQoL data, an improvement in HRQoL was seen in 44.2%, no significant change in 41.9%, mixed results in 11.6%, and worsening in 2.3% of trials. Overall, by the time of FDA approval, OS and PFS data were reported in 59% (152/259) and 65% (168/259) trials respectively with an OS benefit seen in 23.9% (62/259) trials, and PFS benefit in 38.6% (100/259) trials. Of the 84 trials that led to FDA approvals based solely on response rate, HRQoL was reported in only 23.8% (n = 24) with a HRQoL benefit seen in only 9.5% (n = 8) trials. Trials reporting either no significant impact on HRQoL or a mixed impact on HRQoL reported median OS benefit of 4.6 months and 4.2 months respectively. In trials reporting HRQoL data > 6 mo from FDA approval, OS benefit of < 3 mo was seen in 17.8% (8/45) trials. No significant time trends were noted during the study period. Conclusions: There was significant underreporting of HRQoL outcomes in trials ( < 50%) associated with FDA drug approvals between 2015-2020 with majority of trials reporting HRQoL data in an ancillary paper/abstract, and at a much later time than the FDA approval. While it is widely accepted that timely dissemination of HRQoL data from cancer drug trials are vital for clinical and regulatory decisions, no improvement in reporting rates were noted over past 5 years. Only 10% of drugs approved on the basis of response rates showed improvement in HRQoL in the registration trials.

Summary Despite decreasing overall morbidity with minimally invasive esophagectomy (MIE), conduit functional outcomes related to delayed emptying remain challenging, especially in the immediate postoperative setting. Yet, this problem has not been described well in the literature. Utilizing a single institutional prospective database, 254 patients who underwent MIEs between 2012 and 2020 were identified. Gastric conduit dilation was defined as a conduit occupying &gt;40% of the hemithorax on the postoperative chest X-ray. Sixty-seven patients (26.4%) demonstrated acute conduit dilation. There was a higher incidence of conduit dilation in the patients who underwent Ivor Lewis esophagectomy compared to those with a neck anastomosis (67.2% vs. 47.1%; P = 0.03). Patients with dilated conduits required more esophagogastroduodenoscopies (EGD) (P &lt; 0.001), conduit-related reoperations within 180 days (P &lt; 0.001), and 90-day readmissions (P = 0.01). Furthermore, in 37 patients (25.5%) undergoing Ivor Lewis esophagectomy, we returned to the abdomen after intrathoracic anastomosis to reduce redundant conduit and pexy the conduit to the crura. While conduit dilation rates were similar, those who had intraabdominal gastropexy required EGD significantly less and trended toward a lower incidence of conduit-related reoperations (5.6% vs. 2.7%). Multivariable analysis also demonstrated that conduit dilation was an independent predictor for delayed gastric conduit emptying symptoms, EGD within 90 days, conduit-related reoperation within 180 days, and 30-day as well as 90-day readmission. Patients undergoing MIE with acute gastric conduit dilation require more endoscopic interventions and reoperations.

No abstract