Diversity in Research

Medical & Graduate Students

Yoshiko Toyoda

"Effect of Affordable Care Act Medicaid Expansion on Mastectomy and Breast Reconstruction Rates"

States had the option to expand Medicaid under the Patient Protection and Affordable Care Act (passed in 2010). Thirty-two states (including DC) have expanded Medicaid since 2014, while 19 have not. This state specific discrepancy allows study of the effect of Medicaid expansion on health outcomes over time. Breast cancer is the most common non-skin cancer in women. While oncologic surgery is life-saving, the patient is left disfigured. Breast reconstruction has been demonstrated to improve quality of life through substantial psychological benefit and is readily available to insured women under the Women’s Health and Cancer Rights Act of 1998, as well as through state-specific mandates on breast reconstruction education and referrals. However, barriers to breast reconstruction remain, of which insurance coverage is a major factor. We hypothesize that Medicaid expansion states will have had an increase in insurance coverage, screening mammography prevalence, and mastectomy and breast reconstruction rates compared to states which did not expand Medicaid.

Rachel Umans

"Community Perspectives in Medicine: Evaluating Course Impact on First Year Medical Students"

The aim in this prospective cohort study with paired analysis is to test the hypothesis that students who participate in the student-run elective course “Community Perspectives in Medicine (CPIM)” will have significant changes in their attitudes towards under-served populations compared to their (control) classmates who could not register because of limited class size. CPIM provides a unique forum for first year medical students to interact and have open discussions with members of communities that are most impacted by social and health inequities. Thus, the goal of this study is to evaluate the impact of the course on medical students' attitudes towards underserved populations. To do this, students enrolled in the course will be invited to complete pre- and post-course surveys, using the previously-validated “Medical Students' Attitudes Towards the Underserved” (MSATU) to assess their attitudes towards marginalized patient populations. Paired by anonymized pre- and post-course responses will be collected using REDCap (as described below). A control group will be recruited from peer students in the same medical school class who attempted to enroll in CPIM but were unable to only due to limited class size. Students enrolled in the study (and time-matched controls) may be contacted for up to a maximum of four surveys: 1) before the start of the course, 2) immediately following the course, 3) one year after course completion, and 4) two years after course completion (during their clinical rotations).

Resident

Lauren Kelly, M.D.

"Disparities in end-of-life (EOL) outcomes in Latino immigrant vs. US-born Latino patients: Results from two sequential multi-institutional, longitudinal cohort studies"

Cancer is the leading cause of death among Latino persons who reside in the US, greater than one-third of whom are immigrants. Little is known about the influence of Latino immigrant status on end-of-life (EOL) cancer care and there is limited research guiding providers on how to deliver optimal EOL care in this population. Many studies have focused on racial, ethnic, and religious differences in care preferences. However, there is a lack of focus on the socioeconomic forces that impact Latino patients and families as they face serious illness, and the effect of immigrant status on EOL care outcomes has not yet been well characterized. The purpose of this study is to determine Latino immigrant/US-born differences in EOL cancer outcomes (e.g., ACP, palliative care, intensive care, value-consistent care, quality of life and death), and to determine the factors that significantly explain Latino immigrant/US-born disparities in EOL outcomes, including lack of available family support, low income, educational level, language proficiency, health insurance status, and access to hospice services. Data will be derived from two sequential multi-institutional, longitudinal cohort studies of patients with advanced cancer recruited from 2002-2008 and 2010-2015, both based in academic and community-based medical centers. Participants will include self-reported U.S.-born non-Latino whites (N=253), U.S.-born Latinos (N=34), and Latino immigrants (N=65) who had a diagnosis of a poor-prognosis advanced cancer. The results of this study will address a gap in the literature by further defining socioeconomic disparities in the Latino immigrant population, thus highlighting areas to target potential policy reforms, such as improved access to charitable hospice care among the undocumented population.

Robert White, M.D.

"Antiemetic prophylaxis as an anesthesia quality marker and its association with race in the Multicenter Perioperative Outcomes Group"

Health care disparities continue to persist, despite countermeasures, and they have been linked to wealth and wealth distribution, race, language, health insurance primary payer status, and geography. Differences in outcomes according to these social determinants of health have been shown to exist in both the medical and surgical populations. However, limited research has been done describing the role of anesthesiologists, in particular intraoperative anesthesiologists, in contributing to this epidemic. We previously used the National Anesthesia Clinical Outcomes Registry (NACOR) electronic anesthesia database and showed that patients with lower socioeconomic status (as indicated by Medicaid insurance or lower median income) may receive inferior anesthesia care provided by individual anesthesiologists, as indicated by statistically significant less administration of antiemetic prophylaxis medications (dexamethasone and ondansetron). The findings were robust to sensitivity analyses, challenging the notion that anesthesia providers do not contribute to health care disparities. Our findings were not without important limitations including generalizability, validity, and missing data. Hence, we seek to repeat our analysis in another large national electronic anesthesia database: the Multicenter Perioperative Outcomes Group (MPOG). Our hypothesis is that patient race is associated with reduced quality of intraoperative anesthesia care as evidenced by a reduction in receiving antiemetic prophylaxis medications and that this disparity is present after controlling for patient-, surgical-, and outcome-related variables.

Peter Kennel, M.D.

"Does Polypharmacy Contribute to Socioeconomic Disparities in Heart Failure?”

This project directly responds to the RFA by proposing to carry out outcomes research that addresses the well-known socioeconomic status-related health disparity observed in heart failure. Epidemiologic studies have revealed staggering health disparities in outcomes among individuals with heart failure (HF). In particular, individuals with HF who have low income experience higher rates of mortality and readmission. Although factors such as education, social support, lifestyle, health literacy, and access to care have been implicated as potentially contributing to this finding, prior literature suggests that these factors do not fully explain the observation that individuals with low socioeconomic status (SES) experience worse outcomes compared to those with higher SES. Consequently, there remains a need to identify underappreciated modifiable factors that are contributing to this disparity. This proposal seeks to improve the health of individuals with heart failure and low SES by exploring whether polypharmacy, a potentially modifiable factor, mediates the association between low SES and adverse outcomes.

Fellows/Postdoctoral Students

Mavee Witherspoon, Ph.D.
Research Associate in Medicine

"Delineating Genetic Predisposition for Multiple Myeloma in African Americans"

Multiple myeloma (MM) is a plasma cell (PC) malignancy that is largely incurable. Race is an important predisposing factor. MM risk is elevated 2-4 fold for persons of European ancestry (EA) and 5.5-fold for persons of African ancestry (AA) with a first-degree relative with MM. Despite intensive study, the etiology of MM is poorly understood. We have identified the first autosomal dominant MM predisposition gene, KDM1A. KDM1A is a histone 3 lysine-4 demethylase that transcriptionally represses gene targets critical for MM pathogenesis. KDM1A germline mutations are carried by ~1.2% of MM patients unselected for family history (increasing MM risk 5.4-fold). Additionally, in genetic high-risk MM patients we identified a high rate of ClinVar annotated pathogenic mutations in established cancer predisposition DNA repair genes (DRG) (~12%). However, our studies to date have been almost exclusively in MM patients of European ancestry (EA). Here, we will analyze AA MM patients. We hypothesize that germline KDM1A and DRG mutation rates are higher in African American (AA) MM patients than controls subjects, and also higher than in EA MM patients. We will analyze these genes in exomes from 1,200 AA MM patients and 1,200 AA controls. Overall, these studies will provide evidence for clinically relevant multiple myeloma susceptibility genes in individuals of African ancestry, and demonstrate the power of multiethnic patient cohorts to elucidate genetic mechanisms of myelomagenesis.

Elizabeth Luth, Ph.D.
Postdoctoral Associate in Medicine

"Racial Disparities in End-of-Life Care for Alzheimer's Disease and Related Dementia Patients"

Black individuals are at increased risk of ADRD diagnosis and more aggressive EOL care, relative to white individuals. They are also less likely to access EOL care associated with quality of life and complete ACP, which can reduce burdensome transitions and facilitate receipt of value-consistent EOL care. As such, identifying as black carries a potentially multiplicative burden at EOL in the ADRD population. The proposed study aims to identify black-white disparities in ACP and EOL care in ADRD patients who die in hospital using abstracted medical records (n=740). We expect that black patients will receive aggressive treatment more often, access palliative and hospice services less often, and complete ACP at lower rates than white patients. Additionally, key informant interviews with caregivers and healthcare providers of ADRD patients near EOL (n=20) will determine the factors influencing black-white disparities in the quality of EOL care patients receive, caregiver support, and interest in ACP. Interviews will probe for concepts related to EOL care not captured in medical charts (e.g. illness understanding, access to/acceptability of ACP, expectations for care, attitudes towards hospice and palliative care). We expect black caregivers will describe greater barriers to obtaining desired care and support than white caregivers. The results of this study will identify areas ripe for interventions to improve EOL care for ADRD patients and their caregivers without exacerbating existing racial disparities that can be developed in future research studies, including the PI’s K award application.

Fontasha Powell, Ph.D.

"A machine learning approach to prediction of disease conversion and clinical impairment in Multiple Sclerosis in African Americans compared to Caucasians"

Multiple sclerosis (MS) is a chronic inflammatory disease characterized by progressive degeneration of nerve cells in the central nervous system (CNS), disabling 400,000 individuals in the US and 2.5 million worldwide. Though analysis of metrics derived from brain MRI has provided insight into mechanisms of disease damage and recovery, accurate prognoses of cognitive and physical disability still remain challenging. This project proposes a cross-institutional mentorship collaboration between a neurologist (Dr. Susan Gauthier of NYP), neuroscientist (Dr. Amy Kuceyeski of WCM and Cornell BSCB) and computer engineer (Dr. Mert Subuncu of Cornell Engineering) to apply machine learning methods to multimodal human neuroimaging in order to build predictive, translational models of clinical outcomes in MS. The proposed study is innovative in that it takes advantage of an unprecedented, longitudinal multimodal MR imaging dataset, uniquely housed at Cornell, with over 1500 MS patients, capturing disease life course over a period of approximately 10 years. Specifically, we seek to achieve the following aims: 1) Predict future conversion and timing of conversion to different MS-subtypes over the life course of the disease and 2) Predict clinical impairment as measured by physical disability scores. Taken together, the overall goal of these aims is to create a tool for clinicians that can aid in creating more accurate prognoses for multiple sclerosis patients.

Faculty

Jessica Ancker, Ph.D.
Associate Professor of Healthcare Policy and Research

"Combining Social Determinants of Health Data with Electronic Health Record Data for Predictive Modeling"

Background: Socioeconomic status and other social determinants of health (SDH) are well established drivers of health inequality. Applying SDH data in real time appears increasingly feasible with ever-larger sets of clinical data from electronic health records (EHRs), more powerful machine learning capabilities, and novel sources of nonclinical data. However, models that have combined clinical data with SDH data have had mixed results due to limitations in the way community-level data has been used to estimate individual-level predictors. Objective: Our objectives are to estimate the independent contributions of community-level and individual-level SDH, and to apply improved techniques for estimating individual-level SDH predictors. Methods: We will use a large research data set combining clinical with SDH data currently being constructed with funding from the US Department of Transportation. To model likelihood of emergency department admissions among patients with diabetes or congestive heart failure, we will apply established machine learning techniques, elastic net regression and classification/regression trees with resampling, and compare the predictive performance of models to assess the contribution of different types of SDH. Relevance: Medical organizations seeking to improve outcomes and healthcare quality for disadvantaged patient groups need to use SDH data to identify and intervene with high-risk patients. However, it may be prohibitively labor-intensive for medical organizations to conduct primary data collection to collect and maintain SDH data on all patients. Furthermore, such data collection may be perceived as intrusive and inappropriate by patients. Improved methods for integrating publicly available community-specific data with clinical data would assist medical organizations with this important task.

Laura C. Pinheiro, Ph.D.
Assistant Professor of Health Services Research in Medicine

"Evaluating a Model to Improve Comorbidity Management for Minority Cancer Patient"

There are 16 million cancer survivors in the U.S, 70% of whom have co-occurring chronic conditions. Comorbidity management during cancer treatments is often sidelined due to competing priorities, lack of non-cancer expertise, and communication gaps with primary care physicians (PCPs). Evidence suggests that comorbidities may impact cancer treatment, both through delays and possibly direct effects through mechanisms like insulin resistance. As minorities have greater comorbidity burden, improved management may reduce disparities in cancer outcomes. However, care coordination between PCPs and oncologists is a known barrier to effective management. Therefore, “oncogeneralists”, PCPs with cancer-specific training, can help manage comorbidities alongside oncologists during active cancer treatment. Within the context of national efforts to reduce disparities, the oncogeneralist model could be a population health management innovation. Yet, this model has not been widely implemented, and its effects on outcomes are unknown. The objective of the proposed study is to assess the effectiveness of the oncogeneralist model to manage diabetes and hypertension in cancer patients at the only U.S hospital to implement this model. Using electronic medical records, surveys and semi-structured interviews, we will evaluate 1) diabetes and hypertension control, 2) emergency department visits and hospitalizations, and 3) patient and provider satisfaction and patient perceived care coordination. We will also investigate if outcomes vary by race/ethnicity. Findings from this highly innovative work will generate evidence to improve comorbidity management in minority cancer patients, and explore if Weill Cornell’s Cancer Center should consider implementing and studying the oncogeneralist model as a strategy to reduce disparities.

Erica Phillips, M.D., M.S.
Associate Professor of Clinical Medicine

"Reducing Breast Cancer Recurrence Among Black Women: The Breast cancer Weight Loss for Life Study (BWELL)"

Black women with breast cancer have higher cancer-specific and overall mortality rates, partially due to higher rates of obesity, a known contributor to breast cancer progression and mortality. Weight loss interventions among breast cancer survivors positively affect weight, behaviors, biomarkers, and psychosocial outcomes, however, few have targeted Black women. Most standard behavioral weight loss treatments have had limited success in achieving clinically meaningful weight loss (5% or more) in Black women. This application will test an innovative behavioral approach which has shown promising results in closing the gap in disparate weight loss outcomes among Black adults. Using a two phase design approach, we will conduct focus group interviews with patient and provider stakeholders to assess the needs and priorities of early stage (I-III) breast cancer survivors as it relates to weight control. Concepts generated from the qualitative interviews will be used to tailor and adapt a novel acceptance-based therapy (ABT) intervention for weight loss. The feasibility, acceptability and early efficacy of the intervention will be examined in 30 breast cancer survivors with a BMI > 30 kg/m2 over 16 weeks. The results from this study will be used to develop a fully powered RCT of the adapted ABT intervention. If successful this study will deepen our understanding of achieving clinically meaningful weight loss in cancer survivors, a modifiable factor in controlling obesity-related cancers. Our project builds on the research infrastructure of the Cornell Center for Health Equity and expands Weill Cornell’s footprint into communities with the greatest health inequities.

Jennifer Downs, M.D., Ph.D.
Assistant Professor of Medicine in Microbiology and Immunology
Friedman Family Research Scholar in Pediatric Infectious Diseases

"Reducing Disparities in Family Planning Access and Uptake in Tanzania"

Background: Over one-third of sexually-active women in northwest Tanzania who do not desire pregnancy report using no effective family planning method to prevent pregnancy. Our recent focus group interviews demonstrated that major barriers to family planning use include poor education, refusal of male partners, and uncertainty about its compatibility with religious beliefs. We hypothesize that the uptake of family planning can be increased if it is understood and endorsed by local religious leaders. These influential community leaders will then educate their congregations. We build on our methodology that was highly effective in promoting healthseeking behavior in a trial of >145,000 Tanzanian men (Downs, Lancet 2017).

Specific Aims: 1) To determine whether the uptake of family planning will increase to 46% in 10 intervention villages whose religious leaders receive the educational intervention, compared with 18% in 10 control villages that do not. 2) To compare the reported reasons for seeking family planning between women in intervention and control villages.

Medical & Graduate Students

Samuel Taylor, M.D.

"Characterizing the Effects of Fructose on Colorectal Cancer Growth and Metabolism"

Aim 1: To determine whether fructose metabolites regulate pyruvate kinase isoform M2 (PKM2) Hypothesis: Fructose 1-phosphate (F1P) inhibits the activity and tetramerization of PKM2 Aim 1a: To characterize PKM2 activity in the presence of F1P in vitro, in human CRC cell lines, and in mouse intestinal tumors Aim 1b: To characterize PKM2 tetramer formation in the presence of F1P in vitro, in human CRC cell lines, and in mouse intestinal tumors

Aim 2: To determine fructose’s effect on the growth and metabolism of CRC cell lines and mouse intestinal tissues Hypothesis: Fructose enhances growth in CRC cell lines and induces hypertrophy and hyperplasia in mouse intestinal tissues which can be reversed with PKM2 activation Aim 2a: To characterize the growth of human CRC cell lines with and without fructose in standard culture conditions as well as in conditions replicating the tumor microenvironment Aim 2b: To characterize the changes induced in wild-type mouse intestinal tissue following chronic exposure to high-fructose corn syrup (HFCS) Aim 2c: To establish whether fructose or HFCS-induced changes characterized in the above aims can be abrogated by using commercially available small molecule PKM2 activators

Daniel Wang, M.D.

"Social determinants of heart failure readmissions and mortality: A systematic review and meta-analysis"

The primary objective of this study is to conduct a systematic review and meta-analysis examining the social determinants of health that prognosticate 30-day readmission, 1-year mortality, and 1-year ICD placement following an index admission for heart failure (HF). Social determinants include but are not limited to race, age, sex, geography by region within the USA, geography by environment (urban, suburban, or rural), social support, smoking status, alcohol use, discharge facility post-index admission, socioeconomic status, and insurance status. In particular, this study aims to use increased power to identify weak and/or poorly understood elements within the literature, such as readmission, morbidity, and mortality risks among Asian-American HF patients, and why non-white HF patients have lower mortality rates but higher readmission rates than white patients. The current literature on social determinants of HF readmission is dominated by single-center or single-health-system experiences over a period of several years, but few unifying systematic reviews and metaanalyses exist that summarize the data and provide sufficient power to identify granular associations among determinants, readmissions, and mortality. Further, data on cardiovascular risk factors in Asian-Americans have traditionally been limited, and it may be that associations and risk factors previously undetected will come to light with larger sample sizes in a meta-analysis.

The secondary objective of this study is to examine medical prognostic factors for the same endpoints – 30day readmission, 1-year mortality, and 1-year ICD placement – as above. Medical prognosticators include but are not limited to co-morbidity severity by the Charlson Comorbidity Index (CCI), length of stay (LOS) of 2 index admission, and left ventricular ejection fraction (LVEF). The inclusion of medical predictors is in line with the literature, in which all papers examining social determinants also include examinations of medical determinants.

We hypothesize that increased age, male sex, non-white racial identification, non-Northeast or Northwest geography, non-urban living environment, lower social support, non-smoking status, low alcohol use, home discharge, low socioeconomic status, and non-insured status will be correlated with increased 30-day readmission rates after the index heart failure admission. We further hypothesize that increased comorbidity severity per the CCI, increased index LOS, and decreased LVEF are correlated with increased 30-day readmission and 1-year mortality. We anticipate higher 1-year mortality rates in white patients, as previously reported in the literature, though the reasoning for such a finding remains varied.

Resident

Jaime Berstein, M.D.

"Engineering Permanent Nipple Projection"

Nipple reconstruction is critical for successful breast reconstruction in women undergoing mastectomy and in gender affirmation surgery in transgender patients. A highquality nipple reconstruction has the potential to increase happiness and decrease rates of gender dysphoria/suicide rates in transgender patients. Unfortunately, current nipple reconstruction techniques are limited by scar contracture with significant loss of projection. This results in a flattened appearance of the nipple with a poor aesthetic outcome, leaving patients feeling as though their recovery/reconstruction is incomplete. In addition, no data exists regarding the biomechanical properties of the native nipple, making it nearly impossible to evaluate the success of reconstructive options.

Previously, we demonstrated the success of 3D-printed external scaffolds to mitigate volume loss of engineered auricular cartilage. We therefore propose to address the above issues by using minced autologous costal cartilage, harvested from the patient at the time of their surgery, to be placed immediately within a biocompatible, cylindrical, 3D-printed scaffold as an autologous tissue “rebar” for nipple reconstruction. After 6 months, the constructs will be explanted for histological, topographical, and volume analysis. Using autologous costal cartilage allows for a potentially life-long lasting implant. We will also perform the first known assessment of the biomechanical properties of native human nipples.

After accomplishing these aims, we will have created an autologous nipple that maintains permanent contour and projection. In addition, for the first time, the biomechanical properties of the native nipple will have been measured, serving as an important guideline in the fabrication of tissue engineered nipples moving forward.

Daniel Vanderbilt, M.D.

"Feasibility of pulmonary nodules screening at standard Computerized Tomography simulation (CTsim) for breast cancer radiotherapy"

Lung cancer is the leading contributor to cancer mortality in women worldwide. Although there is a clear survival benefit for early detection, screening is not uniformly applied and many patients present with unresectable disease. Women with breast cancer who chose breast conservation undergo CT imaging as part of breast radiotherapy after lumpectomy. We propose to investigate the feasibility of using these stored chest CT images, obtained during planning of breast cancer radiotherapy, for the early detection of suspicious lung nodules and potentially, lung cancer.

Our preliminary analyses suggest the quality of the planning CT scans are comparable to standard low-dose CT scans performed for lung cancer screening. We propose to use archived imaging to generate a database of breast cancer CT simulations, which we will then analyze by formal Lung-RADSTM criteria. The yield of suspicious nodules will then inform an intervention to offer appropriate additional follow up, allowing us to estimate the detection of early lung cancer in this population. This novel strategy of using existing images obtained for therapy of breast cancer for screening of another malignancy offers the potential to improve lung cancer early detection and outcomes in women, without additional exposure to radiation or imaging expense.

Matt Wingo, M.D.

"Transplanted Endothelial Cells for the Restoration of Myocardial Function in Heart Failure with Preserved Ejection Fraction"

Introduction: Women and men often develop heart failure differently. Men are more likely to develop plaques that block the blood flow in their arteries, but women's hearts can become too stiff to pump blood effectively. This is called heart failure with preserved ejection fraction (HFpEF). As of now, there are no proven therapies to treat HFpEF and it is rapidly becoming more widespread. This disease is thought to be caused by the small vessels or "microvasculature" in the heart and endothelial cells (cells that line the veins and arteries).

Hypothesis: Since we know endothelial cells can regenerate and alter their behavior based on their environment, we hypothesize that introducing engineered cardiac endothelial cells to heart tissue will help the prevent HFpEF from worsening or improve it by correcting the microvascular dysfunction.

Specific Aim 1: The goal of this experiment is to evaluate if transplanted EC cells can prevent progression or even rescue the phenotype of HFpEF in a mouse model. Short Term objective: We will then test to see if our cells can change the function and composition of the damaged heart.

Long Term Objective: This work will eventually result in a novel, scalable and economically feasible treatment for HFpEF, the most common type of heart failure in women.

Methods: Mice will receive medication to make them hypertensive. After 3 weeks, we will perform ultrasound examination of their hearts and other tests to verify that they have developed HFpEF. The experimental group treated with mouse cardiac endothelial cells injected into their hearts. The control group will also undergo injection, but they will receive no cells. We will measure the differences in heart function that develop in both groups.

Fellows/Postdoctoral Students

Kiel Telesford, Ph.D.
Postdoctoral Associate in Neuroscience

"Determining ethnicity-based differential B cell inflammation in multiple sclerosis"

Keith Chadwick, M.D.
Instructor in Otolaryngology

"Novel delivery method of transgender voice therapy using a mobile application"

Transgender patients’ voices are closely related to their gender identity. Social isolation and emotional distress can result when patients’ voices are inconsistent with their self-identified gender. Transgender patients’ voices can be rehabilitated to improve congruence with their gender identity through behavioral interventions such as voice therapy. Voice therapy is an effective means of improving satisfaction with voice in the transgender population and leads to an enriched quality of life; however access to voice therapy is a significant struggle due to insurance, financial, social and cultural concerns. The development of easily accessible alternatives for behavioral voice modification in transgender patients is crucial to improving their voice-related quality of life. This study addresses this deficit in access to these health services through the development of a mobile application to deliver transgender voice therapy. A prospective cohort study will be undertaken at Weill Cornell Medical College in collaboration with Ithaca College, during which patients will undergo either current standard of care in-person voice therapy or voice therapy utilizing a novel mobile application. Voice outcomes, including measures of self-reported voice-related quality of life and objective acoustic measurements of the voice, will be compared between groups. Through the development of this application, voice therapy services will be accessible to a greater proportion of transgender patients by reducing the barriers that prevent access to voice care. Additionally, through the efforts of community outreach and engagement during the study, awareness and knowledge of voice problems in the transgender community will increase.

Faculty

Eloise Chapman-Davis, M.D.
Donna Redel Clinical Scholar
Associate Professor of Obstetrics and Gynecology

"Utilization of a web-based platform (Patient Activated Learning System- PALS) to improve knowledge and follow-up among women with abnormal cervical cancer screening"

Cervical cancer incidence and mortality have decreased in the US over time due to effective screening programs. Despite being preventable, cervical cancer diagnosis and mortality are more common among certain ethnic groups, individuals with lower socioeconomic status, and those with limited education. Influences on these outcomes are multifactorial but include low understanding of the screening and diagnosis process. The overall goal of this research is to develop and test an intervention designed to improve knowledge and follow up among underserved women with abnormal pap smears. In this study, we will utilize information gained from patient interviews about their knowledge of the cervical cancer screening and treatment process, and perceived barriers surrounding appropriate follow up, to create content for the intervention. We will use a unique web-based platform, the Patient Activated Learning System (PALS), which provides engaging and informative video and text designed to improve patient knowledge specifically among underserved populations. Twenty diverse patients with abnormal pap smears will be recruited from the Women’s Health colposcopy clinic for initial qualitative interviews. Forty patients from the same population will then be engaged in a pre-post pilot to examine feasibility, acceptability, and impact of modules on improving knowledge. The secondary outcome will be guideline-compliant attendance at the follow up of visit, including a PALS-linked text message sent one week prior to the scheduled follow up appointment. This pilot study aims to develop a scalable, low burden intervention to improve knowledge and guideline compliant follow up in a racially diverse urban tertiary center.

Melissa Davis, Ph.D.
Assistant Professor of Cell and Developmental Biology Research in Surgery (Interim)

"Identification of Ancestry-Specific Drug targets using CRISPR Screening of Patient-Derived African Breast Cancer 3-D Organoid models"

Precision medicine approaches to cancer treatment exploit genetic differences to tailor therapy options that are specific to individual patients/groups. We hypothesize that shared genetic susceptibilities to cancer, which are harbored within populations of common ancestry, may translate to options for targeted therapies. Specifically, we will investigate whether genetic manipulation of kinase genes in tumors derived from African patients may uncover novel drug targets. Our team has uncovered patterns of tumor phenotypes across the African continent, showing similarities among West African women from Nigeria and Ghana that contrast with subtype prevalence in East African women from Ethiopia. We have previously described several gene signatures that are robustly associated with self-identified race in breast cancer - a key asset in designing clinical studies to identify novel treatment options. We will utilize a CRISPR-Cas9 mutagenesis screening method, in patientderived 3D organoid models, to determine which kinases directly impact the tumor viability and response to current therapy drugs in a genetic background that has not been investigated in this context to date. As part of our new WCM initiative, to attain NCI Cancer Center designation, we are charged to have better representation of translational research that develops into actionable discoveries and we are on track to achieve this goal. Our study, if only10% successful, can increase breast cancer IITs by 40% from our current status. This work will expand our capacity to potentially develop precision treatments in our underserved catchment areas giving WCM an unprecedented advantage toward ameliorating cancer disparities, worldwide.

Jialin Mao, M.S.
Assistant Professor of Healthcare Policy and Research

"Racial Disparity in Device-based Surgical Treatment for Prostate Cancer: Examining Surgeons’ Role"

Minority patients are less likely to receive robot-assisted radical prostatectomy (RARP) for prostate cancer treatment, which has been shown to reduce length of stay and readmission, when compared with white patients. It is unknown to what extent the difference in the use of RARP leads to the difference in outcomes by race and ethnicity. In addition, the role of surgeon in a patient-surgeon-facility multilevel structure contributing to disparities related to RARP has never been investigated. We propose to leverage the New York State Cancer Registry and statewide discharge records to create a linked database (2004-2016) of individual patient level prostate cancer data to examine the disparities related to RARP. We propose to assess to what extent the difference in postoperative outcomes across racial groups among prostate cancer patients and ethnicity is mediated by the differential use of RARP and determine surgeon’s influence on the differential use of RARP. Our central hypothesis is that surgeon is an important part in the multilevel framework, which can be broken down into two pathways—segregated and differential treatment. The segregated treatment hypothesis is that minority patients are disproportionately treated by surgeons who are infrequent RARP performers. The differential treatment hypothesis is that when treating minority and white patients with similar conditions, surgeons are less likely to perform RARP for minority patients than for white patients. The proposed study will fill knowledge gaps in understanding the racial disparities related to technology adoption in cancer care and pave the road for future patient engagement research.

Joseph Osborne, M.D., Ph.D.
Professor of Radiology

"WCM catchment Prostate Cancer Health Impact Program (pCHIP)"

Prostate cancer is the most common cancer in the world among men, with significant racial/ethnic disparities in treatment and survival observed. Our project focuses on AA men (a racial group that has the lowest PCa survival rates in New York city, state, and the US) and H/L men (an ethnic group that is less likely to receive therapy for PCa compared to Whites, even when they have more aggressive disease) in the NYPH Brooklyn/Queens catchment. Evidence from randomized clinical trials support the efficacy of decision navigation, with navigated patients showing greater confidence in their decisions about cancer treatment and less regret, both after initial consultation and follow up, when compared to patients in the control group [1, 2]. Despite this evidence supporting decision navigation as a feasible, acceptable, and effective strategy among newly diagnosed men with PCa, no studies to our knowledge have specifically focused on DNI among AA and H/L men. Therefore, our team has an opportunity to not only improve access for this group but also pioneer a proven methodology among this vulnerable population. Our study will explicitly identify the influences of shared decision makers along the cancer continuum, which is crucial for elucidating potential social and contextual factors that may contribute to disparities in PCa treatment and prognosis among AA and H/L men. If successful, our proposed intervention has widespread applicability to the treatment of other cancers (e.g. colorectal cancer) that are prevalent among racial/ethnic minorities.

Robert Peck, M.D., M.S.
Associate Professor of Medicine

"Reducing Post-hospitalization Mortality in Impoverished Adults with Hypertensive Emergency: a Pilot Study in Tanzania"

Hypertensive emergency – severe hypertension with end organ damage – is a leading cause of hospital admissions worldwide. Post-hospitalization outcomes for hypertensive emergency are poor, with high rates of re-hospitalization and death among low socioeconomic status populations, likely due to poor outpatient follow-up and medication nonadherence. At Weill-Bugando Hospital in Mwanza, Tanzania, we documented that 26% of patients with hypertensive emergency died and 20% were re-hospitalized within 6 months after hospital discharge. These outcomes were associated with 1) not attending clinic within one month of discharge; and 2) lower socioeconomic status. A major gap in our current therapeutic approach to hypertensive emergency is that we lack tools to address patients’ social barriers to good health outcomes. Lessons learned from HIV – another chronic disease associated with low socioeconomic status – and from our research in Africa, may help fill this gap. The objective for this application is to adapt and pilot test an evidence-based, social worker intervention used in hospitalized HIV-infected Tanzanian adults, for patients with hypertensive emergency. We hypothesize that this 5-visit intervention will reduce rates of death within 6 months to <5%, through enabling patients with hypertensive emergency to overcome social barriers and attend outpatient care. Our rationale for this research is to generate formative data for an NIH R01 application to conduct a randomized trial of this novel intervention. Our study has 2 specific aims. The expected outcome is to develop an innovative social medicine tool that can improve post-hospitalization outcomes for hypertensive emergency, thereby advancing the fight against healthcare disparities.

Medical & Graduate Students

Ishani Premaratne

"Impact of an Academic Relative Value Unit System on Gender-Based Differences in Surgical Faculty Compensation and Academic Productivity"

African-Americans (AA) are more likely to develop aggressive cancers and experience poorer cancer prognoses than non-Hispanic Caucasians. Examples include prostate, breast, ovarian, and non-small lung carcinomas. Despite such racial/ethnic disparities in cancer mortality that continue to widen, genomic studies rarely interrogate cancer in diverse populations. Recently, genome-wide somatic analysis have revealed more aggressive genomic characteristics of AA tumors including higher genomic instability (GI), homologous recombination-deficiency (HRD) levels, and more aggressive molecular features such as chromothripsis across many cancer types. GI and HRD levels are strongly correlated across AA tumors, indicating that HRD plays an important role in GI in these patients. Current practice requires clinically relevant biomarkers to tailor treatment of cancer, including testing for HRD in patients with ovarian, prostate, breast and pancreatic cancers (current National Comprehensive Cancer Network — NCCN guidelines). Using a low-pass whole-genome sequencing-based cell free DNA assay and novel detection computational algorithms to detect HRD mutational signatures, we will identify AA cancer patients who may benefit from PARP inhibitors and/or platinum-based chemotherapy. Given our robusl sequencing and bioinformatics platforms combined with our large volume clinical practice and clinical trials portfolio, we plan to leverage the results of the proposed study to reduce racial/ethnic disparities in cancer mortality.

Resident

Nabeel Wahid

"Impact of Affordable Care Act Medicaid Expansion on Disparities in Liver Transplant Listings"

The Affordable Care Act (ACA) was enacted in 2010 to improve health equity in part by expanding Medicaid in states that chose to do so. Since the expansion of Medicaid, some states have seen dramatic increases in the number of individuals insured through Medicaid which has in turn resulted in decreased sociodemographic disparities in a variety of healthcare domains.

However, in the field of liver transplantation, there is limited literature on the effects of ACA Medicaid expansion. Some limited studies have concluded that Medicaid expansion increased the number of liver patients on the transplant list who had Medicaid, but there is no conclusive evidence that Medicaid expansion has resulted in decreased longstanding disparities in patients on the transplant waiting list.

Our study will utilize the United Network for Organ Sharing (UNOS) / Organ Procurement and Transplantation Network (OPTN) Scientific Registry of Transplant Recipients (SRTR) database. We aim to describe how Medicaid expansion has affected the sociodemographics (namely age, gender, race/ethnicity, rural/urban residence, and insurance type) of patients on the liver transplant list using two complementary methods: 1) by comparing states that expanded Medicaid with states that did not expand Medicaid and 2) by assessing how the sociodemographics changed over time in states that expanded Medicaid before and after expansion. Additionally, we will assess whether or not Medicaid expansion impacted sociodemographic disparities in time-to-transplantation and time-to-death for individuals on the transplant list.

Fellow/Postdoctoral Students

John Keefe, Ph.D.
Fellow of Psychology in Psychiatry

"Trauma-focused psychodynamic therapy for underserved LGBT patients with post- traumatic stress disorder"

LGB individuals have twice the prevalence of post-traumatic stress disorder (PTSD) relative to heterosexuals. Childhood trauma exposures are also relatively more common among both LGB and transgender individuals. Yet, no data exist about whether or which PTSD therapies are effective for these groups. “Gold-standard” exposure therapy for PTSD may be less appropriate for this population, as exposure therapies have high dropout rates among patients with childhood trauma. An alternative evidence-based route to treating PTSD is through the attachment system, using affect-focused psychotherapies.

Attachment insecurity predisposes individuals to develop PTSD after experiencing trauma, and PTSD further disrupts the attachment system. This feedback loop results in increased anxiety, decreased ability to make sense of one’s experiences, and more fragile connections with others. We will conduct a pilot open trial of trauma-focused psychodynamic psychotherapy (TFPP), an attachment-focused treatment developed at Weill Medical College with funding from the WMC CTSC, in conjunction with VA NY Harbor to treat complex PTSD. Fifteen LGBT patients with Clinician Administered PTSD Scale (CAPS) diagnosed PTSD will be given 24 sessions of TFPP delivered by supervised psychotherapists at WCM/NYP-Weill. After a thorough baseline evaluation, patients will be CAPS assessed at mid-treatment (Week 5), termination (Week 12), and at 3 months post-treatment. Acceptability of the therapy will be assessed through treatment completion (>70% goal). Usefulness of the therapy will be assessed through the CAPS treatment response rate defined as a 30% CAPS improvement. If promising, these pilot data will justify a randomized controlled trial testing TFPP’s efficacy in this population.

John Vaughn

"Understanding Racial and Ethnic Differences in Utilization of Hematopoietic Cell Transplantation for Hematologic Malignancies"

Hematopoietic cell transplantation (HCT) is a potentially life-saving therapy for many diseases including hematologic malignancies, bone marrow failure syndromes, autoimmune disorders, and other hereditary disorders. However, past work has shown that Non-Hispanic Black (NHB) and Hispanic patients are significantly less likely to receive HCT than Non-Hispanic White (NHW) patients for leukemia, lymphoma, and multiple myeloma. The reasons for these patterns are not well-understood. We therefore propose a mixed methods study to understand the reasons for these differences in utilization and to identify barriers and facilitators to transplantation among racial/ethnic minorities. The conceptual framework for our proposal is the Andersen model of healthcare utilization, a well-established framework for understanding both individual and societal determinants of healthcare utilization. Specific Aim 1 will identify contemporary racial differences in utilization of autologous HCT in the linked Surveillance, Epidemiology, End Results (SEER)- Medicare database, controlling for determinants of health services utilization. The SEER- Medicare database is a valuable tool for following a large subset of cancer patients in the United States through diagnosis and treatment of their cancer. Specific Aim 2 will identify barriers and facilitators to HCT from the patient’s perspective using semi-structured interviews. We will conduct interviews at NewYork-Presbyterian Weill Cornell Medicine (a high-volume transplant center) and Brooklyn Methodist Hospital (a community referral source). By identifying predictors of HCT utilization and barriers to transplantation among NHB and Hispanic patients, we will provide critical information needed to develop interventions to improve utilization of HCT for these patients.

Faculty

Arnab Kumar Ghosh, M.B.B.S.
Assistant Professor of Clinical Medicine

"Effect of value-based payment models on racial and socioeconomic disparities in hospitalized medicine patients"

Today, the United States spends about 18% of GDP on healthcare, yet outcomes at every stage of a patient’s lifespan fall behind those of developed countries. Furthermore, marked variations in outcomes by race and wealth remain, driven by social factors influencing health. To address costs and variation in care, healthcare financing has moved from fee-for-service to paying for delivering higher quality outcomes at reduced cost (i.e., value-based payment models

[VBPMs]), leading to large-scale changes to hospital organizational processes. While studies have evaluated the effectiveness of VBPMs overall and their effect on outcomes such as readmission and mortality rates, there is little research on how the organization of inpatient care and its subsequent effects (e.g., length of stay [LOS] and discharge destination) within hospitals explains racial and socioeconomic disparities in hospitalized patients in the VBPM era.

Using quantitative methods designed to control for unobserved confounding, this project will use large administrative datasets to investigate how hospital-oriented VBPMs impact racial/ethnic and socio-economic variations in a) LOS and b) discharge destination in adult medicine inpatients.

This work is significant because hospitals, healthcare leaders, and policymakers need to understand whether and how institutional structures in the era of new payment models influence and potentially worsen existing healthcare disparities. The findings will be innovative because they examine the unexplored domain of how organizational attributes of healthcare delivery in hospitals influence healthcare disparities, and direct future research towards the organizational behavior of healthcare providers in response to external economic pressures such as VBPMs.

Caitlin Hoffman, M.D.
Assistant Professor of Neurological Surgery

"PrIMES: A Targeted, Data-Driven, Longitudinal Approach to Closing the Healthcare Education Diversity Gap"

Increasing ethnic and racial diversity of medical school matriculants is paramount for rectifying health care disparities. Despite national initiatives to increase participation of underrepresented groups in medicine (URiM), students from disenfranchised demographics have comprised -10% of medical school graduates since 1975. To help address this issue, in 2018 we founded PrlMES: a longitudinal, targeted, data-driven mentorship program pairing URiM pre-medical student Mentees with current medical student or resident Mentors. The PrIMES curriculum was designed via quantitative analyses of focus groups and interviews of URiM students and families regarding perceived barriers and challenges to meeting the Association of American Medical College (AAMC) application benchmarks. In the PrIMES Pilot Cycle, Mentors facilitated 11 local freshman and sophomore college students in completing the PrIMES curriculum, during which their preparedness for successful application and matriculation into medical school increased as much as 23% on average. Today, after pilot results analysis, programmatic modification, and expansion to multiple academic institutions, Cycle 2 of PrIMES supports 41 URiM college students in realizing their goals of becoming physicians. We also engineered tracking tools and data frameworks to facilitate longitudinal analysis of outcomes and metrics, individualization of the curriculum, and application to peer institutions. To support our continued expansion, we are developing sustainable technologies to automate our data pipeline, as well as other aspects of PrIMES. Additional initiatives include creation of a PrIMES High School curriculum, support of the PrIMES Wardrobe Fund and Mentee Hardship Fund, as well as expansion throughout all academic institutions in New York City by Fall 2020.

Bella Mehta, M.B.B.S.
Assistant Professor of Medicine

"Pregnancy-related Outcomes in Patients with Systemic Lupus Erythematosus: Impact  of Social Determinants of Health"

Systemic Lupus Erythematosus (SLE) disproportionally affects women and minorities of childbearing age. Pregnancy in women with SLE was once considered so high risk, physicians often counseled termination. Although outcomes have improved,  pregnancy  in  women  with SLE still carries high maternal and fetal risk.

Race and SES disparities in pregnancy outcomes are well documented and extend to women with SLE. Among pregnant women with SLE, African Americans (AAs) have 2-3 fold higher mortality risk compared to whites. Race has also been associated with other adverse pregnancy outcomes in SLE, and SES was a contributor to these disparities. While maternal and fetal outcomes of pregnancy in SLE have improved over the past two decades, it is not known whether the improved outcomes are shared equally among AA and low SES women. It is also unclear whether rates of fetal morbidity or severe maternal morbidity have improved differentially for SLE and non-SLE patients, and for AA and low SES women.

Our proposed retrospective, longitudinal, nationwide analysis of the largest publicly available inpatient database (National Inpatient Sample) will evaluate the effects of race and SES on pregnancy outcomes in SLE, including the rates of fetal and maternal morbidity, over time.

Findings from this pilot study will provide critical evidence to help guide clinical care and health policy for women of child-bearing potential with SLE and provide baseline data for future investigations focused on improving pregnancy outcomes in SLE.

Ayana Morales, M.D.
Instructor in Medicine

"Wilms’ Tumor 1 as an Immunotherapeutic Target in Kaposi sarcoma"

Kaposi Sarcoma (KS), a vascular neoplasm caused by human herpesvirus-8 or Kaposi sarcoma herpesvirus is the most common HIV associated malignancy globally. HIV associated KS is more aggressive than other epidemiologic forms of Mediterranean, classic, or transplant associated cases which can cause cutaneous skin lesions that can progress to lymphatics, lungs and other visceral organs. Persistent disparities exist in KS incidence and survival in the US by race and geography, predominantly affecting African American men.

Antiretroviral Advanced KS require combined antiretroviral therapy with chemotherapies, which may be toxic and are rarely curative. An improved understanding of the virus and host interactions are important in the development of improved targeted treatments for Kaposi sarcoma. We have shown that the oncoprotein WT1 is overexpressed in KS, predominantly studied in HIV associated KS. WT1 has been used as a prognostic marker in acute myeloid leukemia, myelodysplastic syndromes and in various solid tumors. Immunotherapy targeting WT1 has been used in clinical trials against several hematological malignancies and solid cancers and shown to be safe and to produce immunological and clinical responses. In this proposal, we plan to address the impact of KSHV and HIV co-infection on WT1 expression and the tumor microenvironment.

Investigating the oncogene WT1 through patient specimens, cell culture and animal model experiments by targeting tumor cells with overexpressed WT1 with WT1-CTLs and a therapeutic human monoclonal Ab-ESK-1 specific for WT1, will allow us to identify molecular mechanisms and provide preclinical evidence for a potential therapeutic approach specific for WT1 in KS.

Juan Miguel Mosquera, M.D.
Associate Professor of Pathology and Laboratory Medicine

"Detection of Homologous Recombination Deficiency in Cancers of African American Patients through a Whole- Genome Sequencing-based Cell Free DNA Assay"

African-Americans (AA) are more likely to develop aggressive cancers and experience poorer cancer prognoses than non-Hispanic Caucasians. Examples include prostate, breast, ovarian, and non-small lung carcinomas. Despite such racial/ethnic disparities in cancer mortality that continue to widen, genomic studies rarely interrogate cancer in diverse populations. Recently, genome-wide somatic analysis have revealed more aggressive genomic characteristics of AA tumors including higher genomic instability (GI), homologous recombination-deficiency (HRD) levels, and more aggressive molecular features such as chromothripsis across many cancer types. GI and HRD levels are strongly correlated across AA tumors, indicating that HRD plays an important role in GI in these patients. Current practice requires clinically relevant biomarkers to tailor treatment of cancer, including testing for HRD in patients with ovarian, prostate, breast and pancreatic cancers (current National Comprehensive Cancer Network — NCCN guidelines). Using a low-pass whole-genome sequencing-based cell free DNA assay and novel detection computational algorithms to detect HRD mutational signatures, we will identify AA cancer patients who may benefit from PARP inhibitors and/or platinum-based chemotherapy. Given our robusl sequencing and bioinformatics platforms combined with our large volume clinical practice and clinical trials portfolio, we plan to leverage the results of the proposed study to reduce racial/ethnic disparities in cancer mortality.