Multicenter Phase II Study of Cabazitaxel in Advanced Gastroesophageal Cancer: Association of HER2 Expression and M2-Like Tumor-Associated Macrophages with Patient Outcome.

PURPOSE: We examined cabazitaxel, a novel next-generation taxoid, in patients with metastatic gastric cancer in a multicenter phase II study. PATIENTS AND METHODS: Patients who have progressed on one or more prior therapies for locally advanced, unresectable, or metastatic disease were eligible, and prior taxane therapy was allowed. Taxane-naïve and pretreated cohorts were analyzed independently for efficacy. The primary endpoint for both cohorts was progression-free survival (PFS) using RECIST 1.1, using a Simon's two-stage design (10% significance and 80% power) for both cohorts. Comprehensive molecular annotation included whole exome and bulk RNA sequencing. RESULTS: Fifty-three patients enrolled in the taxane-naïve cohort (Arm A) and 23 patients in the prior-taxane cohort (Arm B), from January 8, 2013, to April 8, 2015: median age 61.7 years (range, 35.5-91.8 years), 66% male, 66% Caucasian. The most common adverse events included neutropenia (17% Arm A and 39% Arm B), fatigue/muscle weakness (13%), and hematuria (12%). In Arm A, the 3-month PFS rate was 28% [95% confidence interval (CI), 17%-42%] and did not meet the prespecified efficacy target. The 3-month PFS rate in Arm B was 35% (95% CI, 16%-57%) and surpassed its efficacy target. HER2 amplification or overexpression was associated with improved disease control (P = 0.003), PFS (P = 0.04), and overall survival (P = 0.002). An M2 macrophage signature was also associated with improved survival (P = 0.031). CONCLUSIONS: Cabazitaxel has modest activity in advanced gastric cancer, including in patients previously treated with taxanes. Her2 amplification/overexpression and M2 high macrophage signature are potential biomarkers for taxane efficacy that warrant further evaluation.

Microtubule Engagement with Taxane Is Altered in Taxane-Resistant Gastric Cancer.

PURPOSE: Although taxane-based therapy is standard treatment for advanced gastric cancer, a majority of patients exhibit intrinsic resistance to taxanes. Here, we aim to identify the molecular basis of taxane resistance in gastric cancer. EXPERIMENTAL DESIGN: We performed a post hoc analysis of the TAX-325 clinical trial and molecular interrogation of gastric cancer cell lines to assess the benefit of docetaxel in diffuse (DIF-GC) versus intestinal (INT-GC) gastric cancer. We assessed drug-induced microtubule stabilization in gastric cancer cells and in biopsies of patients with gastric cancer treated with taxanes. We performed transcriptome analysis in taxane-treated gastric cancer cells and patients to identify molecular drivers of taxane resistance. RESULTS: Patients with DIF-GC did not derive a clinical benefit from taxane treatment suggesting intrinsic taxane resistance. DIF-GC cell lines displayed intrinsic resistance specific to taxanes because of impaired drug-induced microtubule stabilization, in the absence of tubulin mutations or decreased drug accumulation. Using taxane-treated gastric cancer patient biopsies, we demonstrated that absence of drug-target engagement was correlated with clinical taxane resistance. Taxane-sensitive cell lines displayed faster microtubule dynamics at baseline, implicating proteins that regulate cytoskeletal dynamics in intrinsic taxane resistance. Differential gene expression analysis of untreated and docetaxel-treated gastric cancer lines and patient samples identified kinesins to be associated with taxane sensitivity in vitro and in patient samples. CONCLUSIONS: Our data reveal that taxane resistance is more prevalent in patients with DIF-GC, support assessment of drug-target engagement as an early read-out of taxane clinical efficacy, and encourage the investigation of kinesins and other microtubule-associated proteins as potentially targetable mediators of taxane resistance in gastric cancer.

Identification of low abundance microbiome in clinical samples using whole genome sequencing.

Identifying the microbiome composition from primary tissues directly affords an opportunity to study the causative relationships between the host microbiome and disease. However, this is challenging due the low abundance of microbial DNA relative to the host. We present a systematic evaluation of microbiome profiling directly from endoscopic biopsies by whole genome sequencing. We compared our methods with other approaches on datasets with previously identified microbial composition. We applied this approach to identify the microbiome from 27 stomach biopsies, and validated the presence of Helicobacter pylori by quantitative PCR. Finally, we profiled the microbial composition in The Cancer Genome Atlas gastric adenocarcinoma cohort.