Leonardo O. Rodrigues

Non-obvious correlations to disease management unraveled by Bayesian artificial intelligence analyses of CMS data

OBJECTIVE Given the availability of extensive digitized healthcare data from medical records, claims and prescription information, it is now possible to use hypothesis-free, data-driven approaches to mine medical databases for novel insight. The goal of this analysis was to demonstrate the use of artificial intelligence based methods such as Bayesian networks to open up opportunities for creation of new knowledge in management of chronic conditions. MATERIALS AND METHODS Hospital level Medicare claims data containing discharge numbers for most common diagnoses were analyzed in a hypothesis-free manner using Bayesian networks learning methodology. RESULTS While many interactions identified between discharge rates of diagnoses using this data set are supported by current medical knowledge, a novel interaction linking asthma and renal failure was discovered. This interaction is non-obvious and had not been looked at by the research and clinical communities in epidemiological or clinical data. A plausible pharmacological explanation of this link is proposed together with a verification of the risk significance by conventional statistical analysis. CONCLUSION Potential clinical and molecular pathways defining the relationship between commonly used asthma medications and renal disease are discussed. The study underscores the need for further epidemiological research to validate this novel hypothesis. Validation will lead to advancement in clinical treatment of asthma & bronchitis, thereby, improving patient outcomes and leading to long term cost savings. In summary, this study demonstrates that application of advanced artificial intelligence methods in healthcare has the potential to enhance the quality of care by discovering non-obvious, clinically relevant relationships and enabling timely care intervention.

Identification of Filamin-A and -B as potential biomarkers for prostate cancer

Aim: A novel strategy for prostate cancer (PrCa) biomarker discovery is described. Materials & methods: In vitro perturbation biology, proteomics and Bayesian causal analysis identified biomarkers that were validated in in vitro models and clinical specimens. Results: Filamin-B (FLNB) and Keratin-19 were identified as biomarkers. Filamin-A (FLNA) was found to be causally linked to FLNB. Characterization of the biomarkers in a panel of cells revealed differential mRNA expression and regulation. Moreover, FLNA and FLNB were detected in the conditioned media of cells. Last, in patients without PrCa, FLNA and FLNB blood levels were positively correlated, while in patients with adenocarcinoma the relationship is dysregulated. Conclusion: These data support the strategy and the potential use of the biomarkers for PrCa.

Abstract 786: Project Survival: Interrogative Biology® platform mediated discovery of molecular markers for detection, stratification and outcomes in pancreatic cancer

Pancreatic adenocarcinoma generally presents with a poor prognosis and an extremely low response rate to first line therapies. There is a critical unmet need to discover and implement effective diagnostic panels to stratify outcomes and tailor treatment strategies to improve survival. The BERG Interrogative Biology® platform utilizes Artificial Intelligence to analyze and integrate multi-omic profiles with clinical annotation to define novel biomarkers and improve treatment interventions. Herein, we analyzed the serum proteome, signaling lipidome, structural lipidome, and metabolome of 163patients at multiple timepoints (pancreatic cancer: 115; pancreatitis: 15; and age-matched healthy controls: 33). Utilizing the power of the Bayesian Network learner, bAIcis™ (BERG Artificial Intelligence Clinical Information System), multi-omic profiles were aligned to the longitudinal clinical information and subjected to AI-algorithms to infer probabilistic cause-and-effect relationships among molecular and clinical variables explicitly explaining pancreatic cancer status, cancer progression, and survival, and defining the interconnectivity of molecular features with clinical phenotype. Network features linking into clinical endpoints and key network pressure points were identified as molecular drivers. The drivers of clinical endpoints were analyzed to rank potential biomarkers. Novel biomarkers were discovered that demonstrate diagnostic potential for pancreatic cancer detection, disease prognosis including metastatic disease progression, patient stratification associated with survival, and response to standard chemotherapy agents like Gemcitibine. A prospective study is underway to validate the biomarkers and discover additional diagnostic and therapeutic molecules to improve the outcomes for patients affected by adenocarcinoma of the pancreas. Citation Format: Niven R. Narain, A James Moser, Ramesh Ramanathan, John Crowley, Amy Stoll-D’Astice, Yezhou Sun, Leonardo O. Rodrigues, Eric M. Grund, Emily Chen, Vivek K. Vishnudas, Michael A. Kiebish, Viatcheslav R. Akmaev, Rangaprasad Sarangarajan. Project Survival: Interrogative Biology® platform mediated discovery of molecular markers for detection, stratification and outcomes in pancreatic cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 786.

Abstract 3376: Molecular drivers of endothelial hypoxic adaption during angiogenesis deciphered using the Berg Interrogative Biology® platform

Angiogenesis is a key feature of tumor progression providing oxygen and nutrients required for tumor cell growth, hypoxia being a major driver of this phenomenon. A systems biology approach using the Berg Interrogative Biology® platform was implemented to identify drivers of the endothelial cell (EC) angiogenic response to hypoxia. To determine the proteomic profile of proliferating ECs and non-proliferating confluent ECs exposed to normoxia or hypoxia, a functional proteomic approach employing activity-based probes and phosphorylation analysis was utilized. Kinases and ATPases were labeled with ATP-binding domain enrichment probes and titanium dioxide enrichment of phosphopeptides was employed for capture of protein phosphorylation events. Phenotypic assays including proliferation rates, apoptosis, mitochondrial superoxide and ROS/NO signaling were included as a measure of EC phenotype. Comparative proteomics, kinase activity, phosphoproteomics and assay data were integrated using an AI based Bayesian Network Inference approach to investigate causal signaling networks in order to elucidate the complexity and dynamics of angiogenesis and more specifically, the role of hypoxia in driving intracellular signaling in response to changes in oxygen tension. High confidence causal networks identified novel proteins that modulate the EC hypoxic response, validation of which are in process. Previously characterized proteins that are responsive to hypoxia were also identified in the hypoxic, but not normoxic signaling networks, namely Aldoc, Rac1, mTor, Cav-1 and Bax. Endothelial proliferation rate is closely related to both hydrogen peroxide and nitric oxide signaling, as has previously been reported. Novel regulatory networks that determine these interactions were identified. Using the Berg Interrogative Biology® platform, we are deciphering both the effects of the hypoxic microenvironment, and the unique characteristics of proliferating ECs by applying integrated functional proteomic assays and a systems approach to determine global changes in intracellular signaling in response to hypoxia. Citation Format: Tony E. Walshe, Justin Bourdelais, Viatcheslav R. Akmaev, Leonardo O. Rodrigues, Socheata Lao, Stephane Gesta, Michael A. Kiebish, Vivek K. Vishnudas, Rangaprasad Sarangarajan, Niven R. Narain. Molecular drivers of endothelial hypoxic adaption during angiogenesis deciphered using the Berg Interrogative Biology® platform. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3376.

Abstract 538: Identification and validation of novel prostate cancer biomarkers using the Berg Interrogative Biology™ platform

Prostate cancer is the most frequent cancer diagnosis among men and the second leading cause of cancer-related death. Despite the widespread use of digital rectal exam (DRE) and blood-based screening of prostate-specific antigen (PSA) for prostate cancer screening, there are significant limitations in their specificity and prognostic value. Biomarkers which distinguish i) PSA-low prostate cancer from benign prostatic hyperplasia (BPH), and ii) indolent versus aggressive disease course represent unmet clinical needs. Experimentally, a panel of prostate cancer cell lines and non-tumorigenic, human primary cells were exposed to in vitro conditions designed to simulate poor oxygenation, low pH, diminished nutrient microenvironments, and metabolic perturbations (24-48 h) followed by iTRAQ proteomic analysis of cell lysates. Using the Berg Interrogative Biology™ platform, proteomic data were then subjected to Bayesian network learning to map molecular interactions, with cytoskeletal and scaffolding proteins Filamin A (FLNA), Filamin B (FLNB), and Keratin 19 (KRT19) identified as candidate prostate cancer biomarkers. To validate biomarker expression, mRNA and protein was quantified in panel of primary human prostate epithelial cells (HPrEC) and androgen-sensitive (LnCAP) or refractory (DU145, PC-3) prostate cancer cells, and each was differentially detected in one or more prostate cancer cell lines compared to HPrEC. Using proteomic analysis, peptides from FLNA, FLNB, and KRT19 were also detected cell culture media conditioned by prostate cancer cells (24 h), indicating they can be secreted. Importantly, unlike PSA expression, global regulation of FLNA, FLNB, and KRT19 expression remained unaltered after treatment with multiple prostate-cancer relevant stimuli (e.g., hypoxia, androgens, and inflammatory stimuli). In vivo validation was next conducted in sera from men (N = 447) with confirmed prostate cancer, benign prostate tumors, or BPH using LDT ELISA assays in a CLIA-certified laboratory. To assess the sensitivity and specificity of FLNA, FLNB, and KRT19 compared to PSA, ROC curve analysis was performed. The individual predictive power of each biomarker alone was comparable to that of PSA. However, the combination of age, levels of FLNA, FLNB, and KRT19, and PSA out-performed PSA alone in identification of patients with prostate cancer stratified compared to benign status, gleason scores and incidence of BPH. Together, these data validate the use of the Berg Interrogative Biology™ platform for biomarker discovery and indicate that FLNA, FLNB, and KRT19 can be used in conjunction with PSA for more sensitive and specific prostate cancer screening, a critical unmet need in the field. Citation Format: Niven R. Narain, Anne Diers, Rakibou Ouro-Djobo, Joyce Chan, Leonardo O. Rodrigues, Vivek K. Vishnudas, Eleftherios P. Diamandis, Viatcheslav R. Akmaev, Rangaprasad Sarangarajan. Identification and validation of novel prostate cancer biomarkers using the Berg Interrogative Biology™ platform. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 538. doi:10.1158/1538-7445.AM2015-538

Abstract CT315: Phase I study of BPM 31510 (ubidecarenone) in patients with advanced solid tumors

Background: BPM 31510 is a novel small molecule that targets the metabolic machinery of the cancer microenvironment to create a hallmark shift from lactate dependency towards mitochondrial oxidative phosphorylation, reversing the Warburg effect. Preclinical data indicates Ubidecarenone causes this shift resulting in tumor regression and enhances the antitumor activity in combination with chemotherapy agents in a priming schedule. This is the first clinical study to evaluate the BPM 31510 at a 4-days continuous infusion in four arms; as a single agent, and in combination with Gemcitabine, 5-FU or Docetaxel. Methods: Eligible patients (pts) (aged ≥18 y) had previously treated relapsed/refractory solid tumors. Pts in the monotherapy arm received IV BPM 31510 for 4 days in continuous infusion in 28-d cycles. Patients in the combination arms were primed for 3 weeks with BPM 31510 and then started in a weekly dosing (either gemcitabine, 5-FU or docetaxel) after the BPM 31510 infusion in a 6-week cycle. Doses were escalated in a 3+3 schema. Phase I endpoints were safety, pharmacokinetics (PK) and Multi-Omics based pharmacodynamics (PD). Dose limiting toxicities (DLTs) are determined using Cycle 1 safety data. Tumor response is evaluated at week 2 and every 4 -6 weeks. Results: As of 01 Dec 2014, 56 pts with advanced solid tumors have been enrolled. Pts have been treated at 3 dose levels up to 137 mg/kg of BPM 31510. No DLTs or study treatment-related SAEs have been reported. The MTD has not yet been established. The most frequently reported related AEs in all 4 arms were grade 1-2 INR prolongation that was resolved after Vitamin K administration. Pre-load of pts with Vitamin K have resolved these events. No bleeding reported. Grade 1-2 thrombocytopenia has been seen in the Gemcitabine arm requiring dose modification. Preliminary PK data indicated linear distribution. Tmax and Cmax are associated with the end of the infusion. Twelve out of twenty five patients (48%) that are evaluable for efficacy after cycle 2 showed various responses including: tumor reductions, decrease FDG, arrested tumor progression, stable disease, decrease in tumor markers, clinical improvements reflected on QOL. Conclusions: Emerging data from this study suggest that BPM 31510 is well tolerated in monotherapy or in combination with chemotherapy agents. Early anti-tumor activity is seen. Dose-escalation on a 6-day infusion schedule is ongoing to determine the recommended phase II dose. Citation Format: Manish A. Shah, Peter Yu, Niven Narain, Rangaprasad Sarangarajan, Michael Kiebish, Vivek Vishnudas, Yezhou Sun, Leonardo Rodrigues, Viatcheslav R. Akmaev, Susan Brouwer, Janice Stevens, Ely Benaim, Ralph Zinner. Phase I study of BPM 31510 (ubidecarenone) in patients with advanced solid tumors. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr CT315. doi:10.1158/1538-7445.AM2015-CT315

Abstract 5230: Berg Interrogative Biology™ Informatics Suite: data driven integration of multi-omic technologies using Bayesian AI.

The Berg Interrogative Biology™ Informatics Suite is an automated data-processing instrument for generation of actionable hypotheses using data generated exclusively via the Berg Interrogative Biology™ approach. The Berg Interrogative Biology™ is a platform that systematically interrogates the biological environment in proprietary in-vitro and in-vivo biological model systems. The biologically relevant data output include molecular data from multi-omics technologies such as proteomics, lipidomics, metabolomics and genomics generated within the context of Berg Interrogative Biology™ is subsequently processed by a set of mathematical algorithms within Informatics Suite. The steps include filtering of datasets with methods that allows for missing data without compromising data quality, normalization of data using technology specific methods, imputation of missing data by rigorous statistical approaches, and generation of a molecular interactome model by integrating data across experiments and technologies. Consequently, the multi-omics data is subjected to analysis using a Bayesian Network inference approach and a multi-omic cause-and-effect relationships are inferred for each analyzed condition de novo. In addition to inferring cross-molecular species interaction networks, in-silico perturbation experiments may be performed to predict cascades of molecular and phenotypic responses to a gene or protein knock-down or over-expression model. Model response is analyzed by statistical techniques and submitted to a Rich Internet Application (RIA) that allows a dynamic and interactive meta-analysis of integrated molecular interaction networks. The Informatics Suite pipeline was applied to multi-omic data set generated via the use of the Berg Interrogative Biology™ process in an in-vitro model of angiogenesis. Comprehensive implementation of the platform technology with the informatics workflow not only identified new and physiologically relevant molecular interactions, but also confirmed previously known canonical interaction pathways described in the literature. Thus, the Informatics Suite within the Berg Interrogative Biology™ platform represents a novel computational component for integrative analysis of multi-omics molecular data that is fast, accurate and leads to a rank-ordered number of actionable hypotheses positioning Berg Interrogative Biology™ as one of the most innovative and efficient approaches in drug and biomarker discovery. Citation Format: Leonardo Rodrigues, Vijetha Vemulapalli, Anthony Walshe, Min Du, Michael Keibish, Joaquin J. Jimenez, Vivek K. Vishnudas, Rangaprasad Sarangarajan, Viatcheslav R. Akmaev, Niven R. Narain. Berg Interrogative Biology™ Informatics Suite: data driven integration of multi-omic technologies using Bayesian AI. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 5230. doi:10.1158/1538-7445.AM2013-5230