Although 70-80% of newly diagnosed ovarian cancer patients respond to first-line therapy, almost all relapse and five-year survival remains below 50%. One strategy to increase five-year survival is prolonging time to relapse by improving first-line therapy response. However, no biomarker today can accurately predict individual response to therapy. In this study, we present analytical and prospective clinical validation of a new test that utilizes primary patient tissue in 3D cell culture to make patient-specific response predictions prior to initiation of treatment in the clinic. Test results were generated within seven days of tissue receipt from newly diagnosed ovarian cancer patients obtained at standard surgical debulking or laparoscopic biopsy. Patients were followed for clinical response to chemotherapy. In a study population of 44, the 32 test-predicted Responders had a clinical response rate of 100% across both adjuvant and neoadjuvant treated populations with an overall prediction accuracy of 89% (39 of 44, p < 0.0001). The test also functioned as a prognostic readout with test-predicted Responders having a significantly increased progression-free survival compared to test-predicted Non-Responders, p = 0.01. This correlative accuracy establishes the test's potential to benefit ovarian cancer patients through accurate prediction of patient-specific response before treatment.

As patient derived xenograft (PDX) models are increasingly used for preclinical drug development, strategies to account for the nonhuman component of PDX RNA expression data are critical to its interpretation. A bioinformatics pipeline to separate donor tumor and mouse stroma transcriptome profiles was devised and tested. To examine the molecular fidelity of PDX versus donor tumors, we compared mRNA differences between paired PDX-donor tumors from nine ovarian cancer patients. 1,935 differentially expressed genes were identified between PDX and donor tumors. Over 90% (n = 1767) of these genes were down-regulated in PDX models and enriched in stroma-specific functions. Several protein kinases were also differentially expressed in PDX tumors, e.g. PDGFRA, PDGFRB and CSF1R. Upon in silico removal of these PDX-donor tumor differentially expressed genes, a stronger transcriptional resemblance between PDX-donor tumor pairs was seen (average correlation coefficient increases from 0.91 to 0.95). We devised and validated an effective bioinformatics strategy to separate mouse stroma expression from human tumor expression for PDX RNAseq. In addition, we showed most of the PDX-donor differentially expressed genes were implicated in stromal components. The molecular similarities and differences between PDX and donor tumors have implications in future therapeutic trial designs and treatment response evaluations using PDX models.

Gene set analysis (GSA) has been used for analysis of microarray data to aid the interpretation and to increase statistical power. With the advent of next-generation sequencing, the use of GSA is even more relevant, as studies are often conducted on a small number of samples. We propose the use of soft truncation thresholding and the Gamma Method (GM) to determine significant gene set (GS), where a generalized linear model is used to assess per-gene significance. The approach was compared to other methods using an extensive simulation study and RNA-seq data from smallpox vaccine study. The GM was found to outperform other proposed methods. Application of the GM to the smallpox vaccine study found the GSs to be moderately associated with response, including focal adhesion (p = 0.04) and extracellular matrix receptor interaction (p = 0.05). The application of GSA to RNA-seq data will provide new insights into the genomic basis of complex traits.

Ovarian cancer (OC) is the second most common gynecological malignancy and the fifth leading cause of death due to cancer in women in the United States mainly due to the late-stage diagnosis of this cancer. It is, therefore, critical to identify potential indicators to aid in early detection and diagnosis of this disease. We investigated the microbiome associated with OC and its potential role in detection, progression as well as prognosis of the disease. We identified a distinct OC microbiome with general enrichment of several microbial taxa, including Dialister, Corynebacterium, Prevotella, and Peptoniphilus in the OC cohort in all body sites excluding stool and omentum which were not sampled from the benign cohort. These taxa were, however, depleted in the advanced-stage and high-grade OC patients compared to early-stage and low-grade OC patients suggestive of decrease accumulation in advanced disease and could serve as potential indicators for early detection of OC. Similarly, we also observed the accumulation of these mainly pathogenic taxa in OC patients with adverse treatment outcomes compared to those without events and could also serve as potential indicators for predicting patients' responses to treatment. These findings provide important insights into the potential use of the microbiome as indicators in (1) early detection of and screening for OC and (2) predicting patients' response to treatment. Given the limited number of patients enrolled in the study, these results would need to be further investigated and confirmed in a larger study.