A zero passage ex-vivosingle cell mass response assay to predict tumor response: An observational analysis in gastrointestinal adenocarcinomas.

e16065 Background: Gastrointestinal cancer remains a significant challenge due to limited effectiveness of current therapies. Clinical and molecular tumor features provide some prognostic and predictive value but apply to a limited number of patients. Novel approaches to predict therapeutic response in a timely manner may improve treatment selection and clinical outcomes. Methods: We performed drug screening with mass response testing (MRT) analysis on fresh gastrointestinal adenocarcinoma samples. Samples were shipped to Travera Inc. (Medford, MA) and underwent CLIA-certified MRT with results returned within 48 hours. Drug response profiles were generated by comparing the cell mass distributions of vehicle-treated replicates (baseline heterogeneity) and drug-treated cells (drug effect). A bootstrapping procedure produced a P-value which was mapped to a 0–100 mass response score (MRS). An MRS ≥ 50 indicated a statistically significant effect for a given drug. This study was an exploratory analysis and did not include formal statistical testing. Results: From February to September 2024, 13 samples containing viable tumor cells of adenocarcinoma histological subtype were collected. Three samples were excluded from MRT analysis —one sample did not have sufficient tumor cell purity and two samples demonstrated unsuitable mass change observed in replicate controls. Median age was 68 years (range 54-89), and 9 samples (69%) were from males. Biopsy locations included the esophagus (N = 3), gastroesophageal junction (N = 6), stomach (N = 3), and liver (N = 1). Six patients (46%) were locally advanced upon collection, with the remaining subjects (54%) presenting with advanced disease. The majority of biopsies (N = 12) were collected during a routine esophagogastroduodenoscopy (EGD) prior to start of therapy. Collections yielded a median of 313,000 cells (range: 0.49-38 x10^6 ). Tumor cell enrichment was confirmed via cytometric EpCAM readout with Coulter Counter measurements where indicated. Median purity was 38% (range 6-91%) and median viability was 96% (range 80-100%) at time of analysis. Clinical responders were defined as having exhibited a radiographic response on subsequent clinical imaging. Response status was utilized to assess concordance of clinical response with MRT-based predictions. MRT analysis guided second line treatment-selection for one patient with advanced gastric cancer and reported clinically actionable results in a second patient with advanced rectal cancer. Conclusions: Rapid zero-passage ex-vivo drug predictions from routine clinical samples in a cohort of gastrointestinal adenocarcinomas were feasible across both core needle biopsy and endoscopic samples. If validated in a larger cohort, mass response testing (MRT) may represent a complement to current clinical tools for response prediction and further study is warranted.