Erwin P Bottinger

Icahn School of Medicine at Mount Sinai, Manhattan, New York, United States

Are you Erwin P Bottinger?

Claim your profile

Publications (75)552.48 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Genome-wide association studies (GWASs) have identified many genetic variants underlying complex traits. Many detected genetic loci harbor variants that associate with multiple-even distinct-traits. Most current analysis approaches focus on single traits, even though the final results from multiple traits are evaluated together. Such approaches miss the opportunity to systemically integrate the phenome-wide data available for genetic association analysis. In this study, we propose a general approach that can integrate association evidence from summary statistics of multiple traits, either correlated, independent, continuous, or binary traits, which might come from the same or different studies. We allow for trait heterogeneity effects. Population structure and cryptic relatedness can also be controlled. Our simulations suggest that the proposed method has improved statistical power over single-trait analysis in most of the cases we studied. We applied our method to the Continental Origins and Genetic Epidemiology Network (COGENT) African ancestry samples for three blood pressure traits and identified four loci (CHIC2, HOXA-EVX1, IGFBP1/IGFBP3, and CDH17; p < 5.0 × 10(-8)) associated with hypertension-related traits that were missed by a single-trait analysis in the original report. Six additional loci with suggestive association evidence (p < 5.0 × 10(-7)) were also observed, including CACNA1D and WNT3. Our study strongly suggests that analyzing multiple phenotypes can improve statistical power and that such analysis can be executed with the summary statistics from GWASs. Our method also provides a way to study a cross phenotype (CP) association by using summary statistics from GWASs of multiple phenotypes. Copyright © 2015 The American Society of Human Genetics. Published by Elsevier Inc. All rights reserved.
    The American Journal of Human Genetics 12/2014; · 11.20 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Renal fibrosis is the histological manifestation of a progressive, usually irreversible process causing chronic and end-stage kidney disease. We performed genome-wide transcriptome studies of a large cohort (n = 95) of normal and fibrotic human kidney tubule samples followed by systems and network analyses and identified inflammation and metabolism as the top dysregulated pathways in the diseased kidneys. In particular, we found that humans and mouse models with tubulointerstitial fibrosis had lower expression of key enzymes and regulators of fatty acid oxidation (FAO) and higher intracellular lipid deposition compared to controls. In vitro experiments indicated that inhibition of FAO in tubule epithelial cells caused ATP depletion, cell death, dedifferentiation and intracellular lipid deposition, phenotypes observed in fibrosis. In contrast, restoring fatty acid metabolism by genetic or pharmacological methods protected mice from tubulointerstitial fibrosis. Our results raise the possibility that correcting the metabolic defect in FAO may be useful for preventing and treating chronic kidney disease.
    Nature medicine. 12/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Predictive models built using temporal data in electronic health records (EHRs) can potentially play a major role in improving management of chronic diseases. However, these data present a multitude of technical challenges, including irregular sampling of data and varying length of available patient history. In this paper, we describe and evaluate three different approaches that use machine learning to build predictive models using temporal EHR data of a patient.The first approach is a commonly used non-temporal approach that aggregates values of the predictors in the patient’s medical history. The other two approaches exploit the temporal dynamics of the data. The two temporal approaches vary in how they model temporal information and handle missing data. Using data from the EHR of Mount Sinai Medical Center, we learned and evaluated the models in the context of predicting loss of estimated glomerular filtration rate (eGFR), the most common assessment of kidney function.Our results show that incorporating temporal information in patient’s medical history can lead to better prediction of loss of kidney function. They also demonstrate that exactly how this information is incorporated is important. In particular, our results demonstrate that the relative importance of different predictors varies over time, and that using multi-task learning to account for this is an appropriate way to robustly capture the temporal dynamics in EHR data. Using a case study, we also demonstrate how the multi-task learning based model can yield predictive models with better performance for identifying patients at high risk of short-term loss of kidney function.
    Journal of Biomedical Informatics. 11/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Self-reported ancestry, genetically determined ancestry, and APOL1 polymorphisms are associated with variation in kidney function and related disease risk, but the relative importance of these factors remains unclear. We estimated the global proportion of African ancestry for 9048 individuals at Mount Sinai Medical Center in Manhattan (3189 African Americans, 1721 European Americans, and 4138 Hispanic/Latino Americans by self-report) using genome-wide genotype data. CKD-EPI eGFR and genotypes of three APOL1 coding variants were available. In admixed African Americans and Hispanic/Latino Americans, serum creatinine values increased as African ancestry increased (per 10% increase in African ancestry, creatinine values increased 1% in African Americans and 0.9% in Hispanic/Latino Americans; P≤1x10(-7)). eGFR was likewise significantly associated with African genetic ancestry in both populations. In contrast, APOL1 risk haplotypes were significantly associated with CKD, eGFR<45 ml/min per 1.73 m(2), and ESRD, with effects increasing with worsening disease states and the contribution of genetic African ancestry decreasing in parallel. Using genetic ancestry in the eGFR equation to reclassify patients as black on the basis of ≥50% African ancestry resulted in higher eGFR for 14.7% of Hispanic/Latino Americans and lower eGFR for 4.1% of African Americans, affecting CKD staging in 4.3% and 1% of participants, respectively. Reclassified individuals had electrolyte values consistent with their newly assigned CKD stage. In summary, proportion of African ancestry was significantly associated with normal-range creatinine and eGFR, whereas APOL1 risk haplotypes drove the associations with CKD. Recalculation of eGFR on the basis of genetic ancestry affected CKD staging and warrants additional investigation.
    Journal of the American Society of Nephrology : JASN. 10/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Herpes zoster, commonly referred to as shingles, is caused by the varicella zoster virus (VZV). VZV initially manifests as chicken pox, most commonly in childhood, can remain asymptomatically latent in nerve tissues for many years and often re-emerges as shingles. Although reactivation may be related to immune suppression, aging and female sex, most inter-individual variability in re-emergence risk has not been explained to date. We performed a genome-wide association analyses in 22 981 participants (2280 shingles cases) from the electronic Medical Records and Genomics Network. Using Cox survival and logistic regression, we identified a genomic region in the combined and European ancestry groups that has an age of onset effect reaching genome-wide significance (P>1.0 × 10(-8)). This region tags the non-coding gene HCP5 (HLA Complex P5) in the major histocompatibility complex. This gene is an endogenous retrovirus and likely influences viral activity through regulatory functions. Variants in this genetic region are known to be associated with delay in development of AIDS in people infected by HIV. Our study provides further suggestion that this region may have a critical role in viral suppression and could potentially harbor a clinically actionable variant for the shingles vaccine.Genes and Immunity advance online publication, 9 October 2014; doi:10.1038/gene.2014.51.
    Genes and immunity 10/2014; · 4.22 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Using genome-wide data from 253,288 individuals, we identified 697 variants at genome-wide significance that together explained one-fifth of the heritability for adult height. By testing different numbers of variants in independent studies, we show that the most strongly associated approximately 2,000, approximately 3,700 and approximately 9,500 SNPs explained approximately 21%, approximately 24% and approximately 29% of phenotypic variance. Furthermore, all common variants together captured 60% of heritability. The 697 variants clustered in 423 loci were enriched for genes, pathways and tissue types known to be involved in growth and together implicated genes and pathways not highlighted in earlier efforts, such as signaling by fibroblast growth factors, WNT/beta-catenin and chondroitin sulfate-related genes. We identified several genes and pathways not previously connected with human skeletal growth, including mTOR, osteoglycin and binding of hyaluronic acid. Our results indicate a genetic architecture for human height that is characterized by a very large but finite number (thousands) of causal variants.
    Nature Genetics 10/2014; · 35.21 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Phenome-wide association studies (PheWAS) have demonstrated utility in validating genetic associations derived from traditional genetic studies as well as identifying novel genetic associations. Here we used an electronic health record (EHR)-based PheWAS to explore pleiotropy of genetic variants in the fat mass and obesity associated gene (FTO), some of which have been previously associated with obesity and type 2 diabetes (T2D). We used a population of 10,487 individuals of European ancestry with genome-wide genotyping from the Electronic Medical Records and Genomics (eMERGE) Network and another population of 13,711 individuals of European ancestry from the BioVU DNA biobank at Vanderbilt genotyped using Illumina HumanExome BeadChip. A meta-analysis of the two study populations replicated the well-described associations between FTO variants and obesity (odds ratio [OR] = 1.25, 95% Confidence Interval = 1.11-1.24, p = 2.10 × 10(-9)) and FTO variants and T2D (OR = 1.14, 95% CI = 1.08-1.21, p = 2.34 × 10(-6)). The meta-analysis also demonstrated that FTO variant rs8050136 was significantly associated with sleep apnea (OR = 1.14, 95% CI = 1.07-1.22, p = 3.33 × 10(-5)); however, the association was attenuated after adjustment for body mass index (BMI). Novel phenotype associations with obesity-associated FTO variants included fibrocystic breast disease (rs9941349, OR = 0.81, 95% CI = 0.74-0.91, p = 5.41 × 10(-5)) and trends toward associations with non-alcoholic liver disease and gram-positive bacterial infections. FTO variants not associated with obesity demonstrated other potential disease associations including non-inflammatory disorders of the cervix and chronic periodontitis. These results suggest that genetic variants in FTO may have pleiotropic associations, some of which are not mediated by obesity.
    Frontiers in Genetics 08/2014; 5:250.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Type 2 diabetes (T2D) is more prevalent in African Americans than in Europeans. However, little is known about the genetic risk in African Americans despite the recent identification of more than 70 T2D loci primarily by genome-wide association studies (GWAS) in individuals of European ancestry. In order to investigate the genetic architecture of T2D in African Americans, the MEta-analysis of type 2 DIabetes in African Americans (MEDIA) Consortium examined 17 GWAS on T2D comprising 8,284 cases and 15,543 controls in African Americans in stage 1 analysis. Single nucleotide polymorphisms (SNPs) association analysis was conducted in each study under the additive model after adjustment for age, sex, study site, and principal components. Meta-analysis of approximately 2.6 million genotyped and imputed SNPs in all studies was conducted using an inverse variance-weighted fixed effect model. Replications were performed to follow up 21 loci in up to 6,061 cases and 5,483 controls in African Americans, and 8,130 cases and 38,987 controls of European ancestry. We identified three known loci (TCF7L2, HMGA2 and KCNQ1) and two novel loci (HLA-B and INS-IGF2) at genome-wide significance (4.15×10-94<P<5×10-8, odds ratio (OR) = 1.09 to 1.36). Fine-mapping revealed that 88 of 158 previously identified T2D or glucose homeostasis loci demonstrated nominal to highly significant association (2.2×10-23 < locus-wide P<0.05). These novel and previously identified loci yielded a sibling relative risk of 1.19, explaining 17.5% of the phenotypic variance of T2D on the liability scale in African Americans. Overall, this study identified two novel susceptibility loci for T2D in African Americans. A substantial number of previously reported loci are transferable to African Americans after accounting for linkage disequilibrium, enabling fine mapping of causal variants in trans-ethnic meta-analysis studies.
    PLoS Genetics 08/2014; 10(8):e1004517. · 8.52 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: We describe here the design and initial implementation of the eMERGE-PGx project. eMERGE-PGx, a partnership of the eMERGE and PGRN consortia, has three objectives : 1) Deploy PGRNseq, a next-generation sequencing platform assessing sequence variation in 84 proposed pharmacogenes, in nearly 9,000patients likely to be prescribed drugs of interest in a 1-3 year timeframe across several clinical sites; 2) Integrate well-established clinically-validated pharmacogenetic genotypes into the electronic health record with associated clinical decision support and assess process and clinical outcomes of implementation; and 3) Develop a repository of pharmacogenetic variants of unknown significance linked to a repository of an EHR-based clinical phenotype data for ongoing pharmacogenomics discovery. We describe site-specific project implementation and anticipated products, including genetic variant and phenotype data repositories, novel variant association studies, clinical decision support modules, clinical and process outcomes, approaches to manage incidental findings, and patient and clinician education methods.Clinical Pharmacology & Therapeutics (2014); Accepted article preview online 24 June 2014; doi:10.1038/clpt.2014.138.
    Clinical pharmacology and therapeutics. 06/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Mitochondrial dysfunction is increasingly recognized as contributing to glomerular diseases, including those secondary to mitochondrial DNA (mtDNA) mutations and deletions. Mitochondria maintain cellular redox and energy homeostasis and are a major source of intracellular reactive oxygen species (ROS) production. Mitochondrial ROS accumulation may contribute to stress-induced mitochondrial dysfunction and apoptosis and thereby to glomerulosclerosis. In mice, deletion of the gene encoding Mpv17 is associated with glomerulosclerosis, but the underlying mechanism remains poorly defined. Here we report that Mpv17 localizes to mitochondria of podocytes and its expression is reduced in several glomerular injury models and in human focal segmental glomerulosclerosis (FSGS), but not in minimal change disease (MCD). Using models of mild or severe nephrotoxic serum nephritis (NTSN) in Mpv17+/+ wildtype (WT) and Mpv17-/- knockout mice we found that Mpv17-deficiency resulted in increased proteinuria (mild NTSN) and renal insufficiency (severe NTSN) compared with WT. These lesions were associated with increased mitochondrial ROS generation and mitochondrial injury such as oxidative DNA damage. In vitro, podocytes with loss of Mpv17 function were characterized by increased susceptibility to apoptosis and ROS injury including decreased mitochondrial function, loss of mtDNA content and change in mitochondrial configuration. In summary, the inner mitochondrial membrane protein Mpv17 in podocytes is essential for the maintenance of mitochondrial homeostasis and protects podocytes against oxidative stress-induced injury both in vitro and in vivo.
    AJP Renal Physiology 03/2014; · 4.42 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Focal segmental glomerular sclerosis (FSGS) is a primary kidney disease that is commonly associated with proteinuria and progressive loss of glomerular function, leading to development of chronic kidney disease (CKD). FSGS is characterized by podocyte injury and depletion and collapse of glomerular capillary segments. Progression of FSGS is associated with TGF-β activation in podocytes; however, it is not clear how TGF-β signaling promotes disease. Here, we determined that podocyte-specific activation of TGF-β signaling in transgenic mice and BALB/c mice with Adriamycin-induced glomerulosclerosis is associated with endothelin-1 (EDN1) release by podocytes, which mediates mitochondrial oxidative stress and dysfunction in adjacent endothelial cells via paracrine EDN1 receptor type A (EDNRA) activation. Endothelial dysfunction promoted podocyte apoptosis, and inhibition of EDNRA or scavenging of mitochondrial-targeted ROS prevented podocyte loss, albuminuria, glomerulosclerosis, and renal failure. We confirmed reciprocal crosstalk between podocytes and endothelial cells in a coculture system. Biopsies from patients with FSGS exhibited increased mitochondrial DNA damage, consistent with EDNRA-mediated glomerular endothelial mitochondrial oxidative stress. Our studies indicate that segmental glomerulosclerosis develops as a result of podocyte-endothelial crosstalk mediated by EDN1/EDNRA-dependent mitochondrial dysfunction and suggest that targeting the reciprocal interaction between podocytes and endothelia may provide opportunities for therapeutic intervention in FSGS.
    The Journal of clinical investigation 03/2014; · 15.39 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Low-frequency coding DNA sequence variants in the proprotein convertase subtilisin/kexin type 9 gene (PCSK9) lower plasma low-density lipoprotein cholesterol (LDL-C), protect against risk of coronary heart disease (CHD), and have prompted the development of a new class of therapeutics. It is uncertain whether the PCSK9 example represents a paradigm or an isolated exception. We used the "Exome Array" to genotype >200,000 low-frequency and rare coding sequence variants across the genome in 56,538 individuals (42,208 European ancestry [EA] and 14,330 African ancestry [AA]) and tested these variants for association with LDL-C, high-density lipoprotein cholesterol (HDL-C), and triglycerides. Although we did not identify new genes associated with LDL-C, we did identify four low-frequency (frequencies between 0.1% and 2%) variants (ANGPTL8 rs145464906 [c.361C>T; p.Gln121(∗)], PAFAH1B2 rs186808413 [c.482C>T; p.Ser161Leu], COL18A1 rs114139997 [c.331G>A; p.Gly111Arg], and PCSK7 rs142953140 [c.1511G>A; p.Arg504His]) with large effects on HDL-C and/or triglycerides. None of these four variants was associated with risk for CHD, suggesting that examples of low-frequency coding variants with robust effects on both lipids and CHD will be limited.
    The American Journal of Human Genetics 02/2014; 94(2):223-32. · 11.20 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Elevated low-density lipoprotein cholesterol (LDL-C) is a treatable, heritable risk factor for cardiovascular disease. Genome-wide association studies (GWASs) have identified 157 variants associated with lipid levels but are not well suited to assess the impact of rare and low-frequency variants. To determine whether rare or low-frequency coding variants are associated with LDL-C, we exome sequenced 2,005 individuals, including 554 individuals selected for extreme LDL-C (>98(th) or <2(nd) percentile). Follow-up analyses included sequencing of 1,302 additional individuals and genotype-based analysis of 52,221 individuals. We observed significant evidence of association between LDL-C and the burden of rare or low-frequency variants in PNPLA5, encoding a phospholipase-domain-containing protein, and both known and previously unidentified variants in PCSK9, LDLR and APOB, three known lipid-related genes. The effect sizes for the burden of rare variants for each associated gene were substantially higher than those observed for individual SNPs identified from GWASs. We replicated the PNPLA5 signal in an independent large-scale sequencing study of 2,084 individuals. In conclusion, this large whole-exome-sequencing study for LDL-C identified a gene not known to be implicated in LDL-C and provides unique insight into the design and analysis of similar experiments.
    The American Journal of Human Genetics 02/2014; 94(2):233-45. · 11.20 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The electronic Medical Records and Genomics (eMERGE) (Phase I) network was established in 2007 to further genomic discovery using biorepositories linked to the electronic health record (EHR). In Phase II, which began in 2011, genomic discovery efforts continue and in addition the network is investigating best practices for implementing genomic medicine, in particular, the return of genomic results in the EHR for use by physicians at point-of-care. To develop strategies for addressing the challenges of implementing genomic medicine in the clinical setting, the eMERGE network is conducting studies that return clinically-relevant genomic results to research participants and their health care providers. These genomic medicine pilot studies include returning individual genetic variants associated with disease susceptibility or drug response, as well as genetic risk scores for common "complex" disorders. Additionally, as part of a network-wide pharmacogenomics-related project, targeted resequencing of 84 pharmacogenes is being performed and select genotypes of pharmacogenetic relevance are being placed in the EHR to guide individualized drug therapy. Individual sites within the eMERGE network are exploring mechanisms to address incidental findings generated by resequencing of the 84 pharmacogenes. In this paper, we describe studies being conducted within the eMERGE network to develop best practices for integrating genomic findings into the EHR, and the challenges associated with such work.
    Frontiers in Genetics 01/2014; 5:50.
  • Source
  • Genetics in medicine: official journal of the American College of Medical Genetics 09/2013; · 3.92 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To describe a collaborative approach for developing an electronic health record (EHR) phenotyping algorithm for drug-induced liver injury (DILI). We analyzed types and causes of differences in DILI case definitions provided by two institutions-Columbia University and Mayo Clinic; harmonized two EHR phenotyping algorithms; and assessed the performance, measured by sensitivity, specificity, positive predictive value, and negative predictive value, of the resulting algorithm at three institutions except that sensitivity was measured only at Columbia University. Although these sites had the same case definition, their phenotyping methods differed by selection of liver injury diagnoses, inclusion of drugs cited in DILI cases, laboratory tests assessed, laboratory thresholds for liver injury, exclusion criteria, and approaches to validating phenotypes. We reached consensus on a DILI phenotyping algorithm and implemented it at three institutions. The algorithm was adapted locally to account for differences in populations and data access. Implementations collectively yielded 117 algorithm-selected cases and 23 confirmed true positive cases. Phenotyping for rare conditions benefits significantly from pooling data across institutions. Despite the heterogeneity of EHRs and varied algorithm implementations, we demonstrated the portability of this algorithm across three institutions. The performance of this algorithm for identifying DILI was comparable with other computerized approaches to identify adverse drug events. Phenotyping algorithms developed for rare and complex conditions are likely to require adaptive implementation at multiple institutions. Better approaches are also needed to share algorithms. Early agreement on goals, data sources, and validation methods may improve the portability of the algorithms.
    Journal of the American Medical Informatics Association 07/2013; · 3.57 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Rationale: Vascular smooth muscle cell (VSMC) differentiation from neural crest cells (NCCs) is critical for cardiovascular development, but the mechanisms remain largely unknown. Objective: TGF-β function in VSMC differentiation from NCCs is controversial. We therefore determined the role and the mechanism of a TGF-β downstream signaling intermediate Smad2 in NCC differentiation to VSMCs. Methods and Results: By using Cre/loxP system, we generated NCC tissue-specific Smad2 knockout mouse model and found that Smad2 deletion resulted in defective NCC differentiation to VSMCs in aortic arch arteries during embryonic development and caused vessel wall abnormality in adult carotid arteries where the VSMCs are derived from NCCs. The abnormalities included missing one layer of VSMCs in the media of the arteries with distorted and thinner elastic lamina, leading to a thinner vessel wall as compared to the wild type vessel. Mechanistically, Smad2 interacted with MRTFB to regulate VSMC marker gene expression. Smad2 was required for TGF-β-induced MRTFB nuclear translocation whereas MRTFB enhanced Smad2 binding to VSMC marker promoter. Moreover, we found that Smad2, but not Smad3, was a progenitor-specific transcription factor mediating TGF-β-induced VSMC differentiation from NCCs. Smad2 appeared to be also involved in determining the physiological differences between NCC- and mesoderm-derived VSMCs. Conclusions: Smad2 is an important factor in regulating progenitor-specific VSMC development and physiological differences between NCC- and mesoderm-derived VSMCs.
    Circulation Research 07/2013; · 11.86 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Genome-wide association studies (GWAS) have identified 36 loci associated with body mass index (BMI), predominantly in populations of European ancestry. We conducted a meta-analysis to examine the association of >3.2 million SNPs with BMI in 39,144 men and women of African ancestry and followed up the most significant associations in an additional 32,268 individuals of African ancestry. We identified one new locus at 5q33 (GALNT10, rs7708584, P = 3.4 × 10-11) and another at 7p15 when we included data from the GIANT consortium (MIR148A-NFE2L3, rs10261878, P = 1.2 × 10-10). We also found suggestive evidence of an association at a third locus at 6q16 in the African-ancestry sample (KLHL32, rs974417, P = 6.9 × 10-8). Thirty-two of the 36 previously established BMI variants showed directionally consistent effect estimates in our GWAS (binomial P = 9.7 × 10-7), five of which reached genome-wide significance. These findings provide strong support for shared BMI loci across populations, as well as for the utility of studying ancestrally diverse populations.
    Nature Genetics 04/2013; · 35.21 Impact Factor
  • Clinical Pharmacology &#38 Therapeutics 04/2013; · 6.85 Impact Factor

Publication Stats

2k Citations
552.48 Total Impact Points

Institutions

  • 2006–2014
    • Icahn School of Medicine at Mount Sinai
      • Department of Medicine
      Manhattan, New York, United States
  • 2011
    • Loyola University Medical Center
      Maywood, Illinois, United States
  • 2007–2011
    • Mount Sinai Medical Center
      New York City, New York, United States
  • 1999–2009
    • Albert Einstein College of Medicine
      • • Department of Pediatrics
      • • Nephrology
      New York City, NY, United States
  • 2008
    • Albert Einstein Medical Center
      Philadelphia, Pennsylvania, United States
  • 2004
    • Thomas Jefferson University
      • Division of Hospital Medicine
      Philadelphia, PA, United States