Feng, Yang

In this article, we present a flexible model for microbiome count data. We consider a quasi‐likelihood framework, in which we do not make any assumptions on the distribution of the microbiome count except that its variance is an unknown but smooth function of the mean. By comparing our model to the negative binomial generalized linear model (GLM) and Poisson GLM in simulation studies, we show that our flexible quasi‐likelihood method yields valid inferential results. Using a real microbiome study, we demonstrate the utility of our method by examining the relationship between adenomas and microbiota. We also provide an R package “fql” for the application of our method.

The microbiome plays an important role in human health by mediating the path from environmental exposures to health outcomes. The relative abundances of the high‐dimensional microbiome data have an unit‐sum restriction, rendering standard statistical methods in the Euclidean space invalid. To address this problem, we use the isometric log‐ratio transformations of the relative abundances as the mediator variables. To select significant mediators, we consider a closed testing‐based selection procedure with desirable confidence. Simulations are provided to verify the effectiveness of our method. As an illustrative example, we apply the proposed method to study the mediation effects of murine gut microbiome between subtherapeutic antibiotic treatment and body weight gain, and identify Coprobacillus and Adlercreutzia as two significant mediators.

PurposeSingle-agent gemcitabine (GEM) is standard treatment of metastatic pancreatic cancer. Fixed-dose rate (FDR) GEM and GEM plus oxaliplatin have shown promise in early clinical trials. E6201 was designed to compare overall survival (OS) of standard weekly GEM 1,000 mg/m2/30 minutes versus GEM FDR 1,500 mg/m2/150 minutes or GEM 1,000 mg/m2/100 minutes/day 1 plus oxaliplatin 100 mg/m2/day 2 every 14 days (GEMOX).MethodsThis trial included patients with metastatic or locally advanced pancreatic cancer, normal organ function, and performance status of 0 to 2. The study was designed to detect a 33% difference in median survival (hazard ratio [HR] ≤ 0.75 for either of the experimental arms) with 81% power while maintaining a significance level of 2.5% in a two-sided test for each of the two primary comparisons.ResultsEight hundred thirty-two patients were enrolled. The median survival and 1-year survival were 4.9 months (95% CI, 4.5 to 5.6) and 16% for GEM, 6.2 months (95% CI, 5.4 to 6.9), and 21% for GEM FDR (HR, 0.83; stratified log-rank P = .04), and 5.7 months (95% CI, 4.9 to 6.5) and 21% for GEMOX (HR, 0.88; stratified log-rank P = .22). Neither of these differences met the prespecified criteria for significance. Survival was 9.2 months for patients with locally advanced disease, and 5.4 months for those with metastatic disease. Grade 3/4 neutropenia and thrombocytopenia were greatest with GEM FDR. GEMOX caused higher rates of nausea, vomiting, and neuropathy.ConclusionNeither GEM FDR nor GEMOX resulted in substantially improved survival or symptom benefit over standard GEM in patients with advanced pancreatic cancer.