Walsh, Dana M.

Abstract Objectives Evaluation of differences in the intestinal microbiome and resistome among high-risk patients (i.e. intensive care, oncology, transplant recipients) who are and are not colonized with carbapenem-resistant Enterobacterales (CRE). Methods One hundred and twelve rectal swabs were obtained from 85 patients with known CRE colonization status and cohorted. Long-read metagenomic next-generation sequencing (mNGS) was performed on rectal swabs. Microbiome and resistome analysis were performed by assessing α-diversity, β-diversity, relative abundance assessment and linear discriminant analysis effect size (LEfSe), comparing patients colonized (CRE positive) and not colonized (CRE negative) with CRE. Longitudinal analysis of sequential swabs collected over multiple hospital encounters on a subset of patients was performed at the patient level. Results The microbiomes of cohorts were similar when comparing α- and β-diversity measures and relative abundance. LEfSe analysis identified Gram-negative pathobionts enriched among CRE-positive samples and Gram-positive taxa enriched among CRE-negative samples. α-Diversity of the resistome differed at class, gene and allele levels. Relative abundance and LEfSe analysis demonstrated enrichment of genes conferring β-lactam resistance among CRE-positive patients; LEfSe also demonstrated enrichment of antimicrobial resistance genes to multiple antimicrobial classes. At the patient level, fluctuations in the microbiome and resistome among longitudinally collected swabs were associated with antibiotic exposure. Conclusions Differences between the microbiomes of CRE-positive- and CRE-negative-colonized patients at the cohort level were relatively muted, whereas statistically significant differences were observed among their resistomes. In patients followed longitudinally, shifts in microbiome and resistome composition were dramatic in between encounters and antibiotic exposures.

Abstract Incidence rates for endometrial cancer (EC) are rising, particularly in postmenopausal and obese women. Our previous work has showed that the uterine and vaginal microbiome distinguishes patients with EC from those without. We now examine the impact of patient factors (such as menopause status, body mass index, and vaginal pH) in the microbiome in the absence of EC and how these might contribute to the microbiome signature in EC. We find that each factor independently alters the microbiome and identified postmenopausal status as the main driver of a polymicrobial network associated with EC (ECbiome). Of the 17 taxa we found enriched in EC patients, 8 were also enriched by postmenopause. Because postmenopausal status is a main risk factor for endometrial cancer, this system can be thought of as an ecological succession towards a disease state. Organisms within the ECbiome have been previously identified as polymicrobial associations in chronic soft tissue infections, ulcers, morbid obesity, and bacterial vaginosis. This suggests that this microbiome network is likely involved in mechanisms of host disease. Within the ECbiome, Porphyromonas somerae stands out because of its very significant and specific association with EC (p = 5.73E-6; Q = 0.009). We had previously identified this species as significantly associated with EC, and hereby validate that finding. Because the association of P. somerae with EC was robust and specific enough to potentially be considered as a biomarker for the disease (AUC 76.7%, CI: 67.9-85.5%), we developed a P. somerae-specific qPCR assay to detect this species in vaginal swabs. This becomes of particular interest because the lower tract and uterine microbiome are significantly correlated in these patients (p = 0.01), indicating that the vaginal microbiome can be used as a proxy for the uterine environment in EC. In agreement with the 16S rRNA gene sequencing data, P. somerae was detected significantly more frequently in vaginal swabs of patients with EC than those without EC (p = 3.02E-7). Overall, the test achieved a sensitivity of 74% and specificity of 63%, with a positive predictive value of 0.86 for obese and postmenopausal patients. In summary, we identify Porphyromas somerae presence as the most predictive microbial marker of EC, and we confirm this using targeted qPCR, which could be of use in detecting EC in high-risk, asymptomatic women. Given the established pathogenic behavior of P.somerae, we have also explored mechanistic routes through which this microbe and associated consortia may be involved in the mechanisms of disease. Citation Format: Marina Walther-Antonio, Dana Walsh, Alexis Hokenstad, Jun Chen, Jayeun Sung, Gregory Jenkins, Nicholas Chia, Heidi Nelson, Andrea Mariani. Postmenopause as a key factor in the composition of the Endometrial Cancer Microbiome (ECbiome): Putative role of Porphyromonas somerae in the disease [abstract]. In: Proceedings of the AACR Special Conference on the Microbiome, Viruses, and Cancer; 2020 Feb 21-24; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2020;80(8 Suppl):Abstract nr PR09.

Abstract Background Recurrent Clostridioides difficile infection (rCDI) is an urgent public health threat associated with significant mortality and medical cost. Microbiota therapy is gaining acceptance as a strategy to reduce rCDI recurrence. We present the final 24-month analysis of clinical safety, efficacy, and microbiome restoration from a Phase 2 open-label trial of RBX2660 for prevention of CDI recurrence. Methods Participants with multi-recurrent CDI received <2 doses of RBX2660 delivered via enema 7 days apart in this multicenter, open-label Phase 2 study. Efficacy was defined as the absence of CDI recurrence through 56 days after the last dose and was compared with 8-week recurrence-free rates for a historical control cohort that received standard-of-care antibiotic therapy. Fisher exact test compared the proportion of treatment participants who were CDI-free by age and sex. Durability was defined as continued absence of CDI episodes beyond 8 weeks. Safety and durability assessments occurred at 3, 6, 12, and 24 months. Participant stool samples were collected prior to and for up to 720 days after treatment, and microbiome changes were assessed by shallow shotgun sequencing. Results The efficacy of RBX2660 to prevent rCDI at 8 weeks (78.9%; 112/142) was higher than the CDI-free rate in the historical control group (30.7%, 23/75; P < 0.0001). Age and sex did not impact efficacy. Among participants who achieved treatment success at 8 weeks and were evaluable for long-term durability (n = 95), 8 experienced a new CDI episode by the 24-month follow-up for an overall durability of 91.6%. The safety profile was consistent with previous reports for RBX2660. In total, 503 stool samples from 110 treatment responders were analyzed. Within 7 days of treatment, the relative abundance of Bacteroidia and Clostridia remained shifted higher than pre-treatment levels while Gammaproteobacteria and Bacilli declined sharply after treatment, and these changes persisted to at least 24 months. Conclusion RBX2660, a microbiota-based drug, was safe and efficacious for preventing rCDI with clinical durability to 24 months after treatment, independent of age or sex, and RBX2660 durability associated with durable microbiome shifts from pre-treatment to a healthier composition. Disclosures Robert Orenstein, DO, Rebiotix Inc. (Advisor or Review Panel member), Sarah Mische, PhD, Rebiotix Inc. (Employee), Ken Blount, PhD, Rebiotix Inc. (Employee), Lindy Bancke, PharmD, Rebiotix Inc. (Employee), Xin Su, MD, MSci, Rebiotix Inc. (Employee), Dana Walsh, PhD, Rebiotix Inc. (Employee), Adam Harvey, PhD, Rebiotix Inc. (Employee), Carlos Gonzalez, MS, Rebiotix Inc. (Consultant), Dale N. Gerding, MD, Rebiotix Inc. (Board Member).

Abstract Ovarian cancer (OC) is the most lethal gynecologic malignancy and the fifth leading cause of cancer death in the United States. Metformin, an antidiabetic drug, has been shown to decrease the risk of mortality in cancer patients, particularly those with OC. Though the drug acts through inhibition of the PI3K/mTOR/AKT pathway, which is commonly altered in OC, knockout of key proteins in the pathway does not attenuate metformin’s antidiabetic effects. Further, intravenous metformin administration is less effective than oral, implying that some crucial interaction takes place in the gut. The gut is home to some 1013 bacterial cells, collectively known as the microbiota. These bacteria play important roles in fermentation of fiber, producing short-chain fatty acids (SCFA) that have positive health benefits, including prevention of metabolic syndrome, bowel disorders, and certain cancers. Mechanistically, SCFAs inhibit the PI3K/mTOR/AKT pathway, similar to metformin. Conditions such as obesity, metabolic syndrome, and diabetes significantly alter gut microbiota composition, possibly leading to a reduction in beneficial SCFAs. However, in type II diabetes, treatment with metformin leads to an increase in SCFA-producing bacteria, suggesting that metformin may positively influence the microbiota to improve patient metabolic profiles. In OC, metformin may play a similar role in promoting the presence of SCFA-producing bacteria. We hypothesized that metformin increases the abundance of SCFA-producing bacteria in the gut, with which it acts synergistically to decrease growth of OC cells in combination with OC chemotherapy. To investigate our hypothesis, we have employed the OC patient-derived xenograft (PDX) mouse model established at Mayo Clinic. Immunocompromised SCID mice received OC cells from Mayo Clinic patients and the tumors were allowed to grow to an appropriate size before treatment was started. Cage mates were split into one of four treatment arms: a control (no treatment) group, metformin only, carboplatin/paclitaxel (CP) chemotherapy only, or CP with metformin. Mice received 2mg/ml metformin in their water continuously and/or 51mg/kg carboplatin and 15mg/kg paclitaxel intravenously weekly. Stool was collected weekly prior to tumor injection, during tumor growth, and through the four weeks of treatment. DNA was extracted from stool and bacterial 16S rRNA amplicon sequencing was used to determine microbiota profiles. Mice with chemosensitive tumors treated with only CP chemotherapy had significant reductions in tumor growth along with significantly less microbial gut diversity than pretumor and pretreatment mice. Chemosensitive mice treated with metformin alone did not lose significant microbial diversity but did not have significant reductions in tumor growth as compared to the control mice. CP in combination with metformin both reduced tumor growth and protected against loss of microbiota diversity. Further, mice treated with only metformin had significant increases in abundance in the Firmicutes genera Ruminococcus and Oscillospira, SCFA producers that are associated with leanness and health. Mice treated with both metformin and CP had significant increases in another SCFA-producing Firmicute, Moryella. Finally, mice with chemoresistant tumors showed improved tumor growth inhibition with combination CP/metformin therapy, suggesting that metformin-associated changes in the gut microbiota may increase sensitivity to CP. These data support the hypothesis that metformin protects against detrimental loss of gut microbiota diversity induced by chemotherapy treatment, enriching beneficial SCFA producers. This work implies a role for the gut microbiota in OC treatment and provides support for metformin’s potential use as a beneficial addition to chemotherapy treatment. Citation Format: Dana M. Walsh, Ismail Mert, Chen Jun, Xiaonan Hou, Saravut John Weroha, Marina Walther-Antonio. Metformin alters the gut microbiota of ovarian cancer patients treated with carboplatin/paclitaxel chemotherapy and enhances sensitivity in resistant tumors. [abstract]. In: Proceedings of the AACR Conference: Addressing Critical Questions in Ovarian Cancer Research and Treatment; Oct 1-4, 2017; Pittsburgh, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(15_Suppl):Abstract nr A70.