Anushi Shah

Validating drug repurposing signals using electronic health records: a case study of metformin associated with reduced cancer mortality

Objectives Drug repurposing, which finds new indications for existing drugs, has received great attention recently. The goal of our work is to assess the feasibility of using electronic health records (EHRs) and automated informatics methods to efficiently validate a recent drug repurposing association of metformin with reduced cancer mortality. Methods By linking two large EHRs from Vanderbilt University Medical Center and Mayo Clinic to their tumor registries, we constructed a cohort including 32 415 adults with a cancer diagnosis at Vanderbilt and 79 258 cancer patients at Mayo from 1995 to 2010. Using automated informatics methods, we further identified type 2 diabetes patients within the cancer cohort and determined their drug exposure information, as well as other covariates such as smoking status. We then estimated HRs for all-cause mortality and their associated 95% CIs using stratified Cox proportional hazard models. HRs were estimated according to metformin exposure, adjusted for age at diagnosis, sex, race, body mass index, tobacco use, insulin use, cancer type, and non-cancer Charlson comorbidity index. Results Among all Vanderbilt cancer patients, metformin was associated with a 22% decrease in overall mortality compared to other oral hypoglycemic medications (HR 0.78; 95% CI 0.69 to 0.88) and with a 39% decrease compared to type 2 diabetes patients on insulin only (HR 0.61; 95% CI 0.50 to 0.73). Diabetic patients on metformin also had a 23% improved survival compared with non-diabetic patients (HR 0.77; 95% CI 0.71 to 0.85). These associations were replicated using the Mayo Clinic EHR data. Many site-specific cancers including breast, colorectal, lung, and prostate demonstrated reduced mortality with metformin use in at least one EHR. Conclusions EHR data suggested that the use of metformin was associated with decreased mortality after a cancer diagnosis compared with diabetic and non-diabetic cancer patients not on metformin, indicating its potential as a chemotherapeutic regimen. This study serves as a model for robust and inexpensive validation studies for drug repurposing signals using EHR data.

Applying active learning to high-throughput phenotyping algorithms for electronic health records data.

OBJECTIVES Generalizable, high-throughput phenotyping methods based on supervised machine learning (ML) algorithms could significantly accelerate the use of electronic health records data for clinical and translational research. However, they often require large numbers of annotated samples, which are costly and time-consuming to review. We investigated the use of active learning (AL) in ML-based phenotyping algorithms. METHODS We integrated an uncertainty sampling AL approach with support vector machines-based phenotyping algorithms and evaluated its performance using three annotated disease cohorts including rheumatoid arthritis (RA), colorectal cancer (CRC), and venous thromboembolism (VTE). We investigated performance using two types of feature sets: unrefined features, which contained at least all clinical concepts extracted from notes and billing codes; and a smaller set of refined features selected by domain experts. The performance of the AL was compared with a passive learning (PL) approach based on random sampling. RESULTS Our evaluation showed that AL outperformed PL on three phenotyping tasks. When unrefined features were used in the RA and CRC tasks, AL reduced the number of annotated samples required to achieve an area under the curve (AUC) score of 0.95 by 68% and 23%, respectively. AL also achieved a reduction of 68% for VTE with an optimal AUC of 0.70 using refined features. As expected, refined features improved the performance of phenotyping classifiers and required fewer annotated samples. CONCLUSIONS This study demonstrated that AL can be useful in ML-based phenotyping methods. Moreover, AL and feature engineering based on domain knowledge could be combined to develop efficient and generalizable phenotyping methods.

Opioid use after cardiac surgery in children with Down syndrome.

Objectives: To determine the cumulative opioid doses administered to patients with Down syndrome after cardiac surgery and compare them with patients without Down syndrome. Design: Retrospective observational comparative study. Setting: PICU in a university-affiliated freestanding pediatric teaching hospital. Patients: Infants and children who presented to our institution for heart surgery after July 1, 2008, and met the following criteria: 1) no opioid medications for 48 hours prior to surgery, 2) sternotomy approach with primary closure, and 3) no additional operative procedures in the 5 days after surgery. All patients with Down syndrome were included, and patients without Down syndrome with similar age, type of cardiac lesion, and length of surgical procedure were selected in a ~2:1 ratio, blinded to opioid exposure. Interventions: None. Measurements and Main Results: Clinical and demographic data were extracted from electronic medical record data. Univariate analyses and multivariate linear regression modeling were performed to determine the influence of Down syndrome, patient characteristics, and clinical covariates on weight-adjusted opioid dose. The differences in median cumulative opioid doses between those with Down syndrome (n = 44) and those without Down syndrome (n = 77) were not significant in the first 24 hours (+0.39 mg/kg [95% CI, –0.45 to +1.39 mg/kg]) or 96 hours (+0.54 mg/kg [95% CI, –0.59 to +2.07 mg/kg]) after surgery. Age, cardiac bypass time, benzodiazepines, and neuromuscular blocking agents were significantly correlated with opioid dose, but Down syndrome, gender, pain score, creatinine, acetaminophen, nonsteroidal anti-inflammatory drugs, and steroid medications were not. Patients with Down syndrome had longer hospital stays; in multivariate analysis, higher opioid exposures in the first 96 hours after surgery and higher peak serum creatinine values correlated with longer hospitalization. Conclusions: This cohort did not provide evidence for opioid resistance in patients with Down syndrome. Younger age, longer cardiac bypass time, exposure to benzodiazepines, and neuromuscular blockade did correlate with increased opioid doses after cardiac surgery.

A polymorphism in HLA-G modifies statin benefit in asthma

Several reports have shown that statin treatment benefits patients with asthma; however, inconsistent effects have been observed. The mir-152 family (148a, 148b and 152) has been implicated in asthma. These microRNAs suppress HLA-G expression, and rs1063320, a common SNP in the HLA-G 3′UTR that is associated with asthma risk, modulates miRNA binding. We report that statins upregulate mir-148b and 152, and affect HLA-G expression in an rs1063320-dependent fashion. In addition, we found that individuals who carried the G minor allele of rs1063320 had reduced asthma-related exacerbations (emergency department visits, hospitalizations or oral steroid use) compared with non-carriers (P=0.03) in statin users ascertained in the Personalized Medicine Research Project at the Marshfield Clinic (n=421). These findings support the hypothesis that rs1063320 modifies the effect of statin benefit in asthma, and thus may contribute to variation in statin efficacy for the management of this disease.

Alpha2A adrenergic receptor genetic variation contributes to hyperglycemia after myocardial infarction

BACKGROUND Acute myocardial infarction (AMI) is frequently associated with transient hyperglycemia even in patients without pre-existing diabetes. Acute stress can lead to increased blood glucose through the effect of catecholamines on alpha2A-adrenergic receptors (α2A-ARs) present in pancreatic islet β-cells. Variation in the gene (ADRA2A) that encodes the α2A-AR affects insulin release and glucose control and may play a particularly important role during times of stress. METHODS We performed a retrospective cohort study using de-identified electronic medical records linked to a DNA repository in 521 Caucasians and 55 African-American non-diabetic patients with AMI. We examined the association between admission blood glucose concentrations and ten selected ADRA2A SNPs in Caucasians. RESULTS Three ADRA2A SNPS were associated with stress-induced hyperglycemia in Caucasians. Individuals homozygous for the rs10885122 variant (n=9) had a 23% lower admission glucose (geometric mean [95% CI], 99 [83-118]mg/dl) compared with non-carriers (121 [118-125] mg/dl; n=401; P=0.001). Admission glucose was 14% higher in rs1800544 variant homozygotes (134 [119-150]mg/dl; n=36) compared to non-carriers (118 [115-121]mg/dl; n=290, P=0.046). Furthermore, homozygotes of the rs553668 variant (n=13) had a 13% higher glucose (133 [110-160]mg/dl) compared to non-carriers (118 [115-122]mg/dl; n=366; P=0.056). Haplotypes including these ADRA2A SNPs were associated with higher admission glucose levels. CONCLUSIONS Three ADRA2A genetic variants are associated with blood glucose and stress-induced hyperglycemia after AMI in Caucasians.

Maximizing Service Uptime of Smartphone-Based Distributed Real-Time and Embedded Systems

Smart phones are starting to find use in mission critical applications, such as search-and-rescue operations, wherein the mission capabilities are realized by deploying a collaborating set of services across a group of smart phones involved in the mission. Since these missions are deployed in environments where replenishing resources, such as smart phone batteries, is hard, it is necessary to maximize the lifespan of the mission while also maintaining its real-time quality of service (QoS) requirements. To address these requirements, this paper presents a deployment framework called Smart Deploy, which integrates bin packing heuristics with evolutionary algorithms to produce near-optimal deployment solutions that are computationally inexpensive to compute for maximizing the lifespan of smart phone-based mission critical applications. The paper evaluates the merits of deployments produced by Smart Deploy for a search-and-rescue mission comprising a heterogeneous mix of smart phones by integrating a worst-fit bin packing heuristic with particle swarm optimization and genetic algorithm. Results of our experiments indicate that the missions deployed using Smart Deploy have a lifespan that is 20% to 162% greater than those deployed using just the bin packing heuristic or evolutionary algorithms. Although Smart Deploy is slightly slower than the other algorithms, the slower speed is acceptable for offline computations of deployment.