About

Dr. Scott Atay is an Associate Professor of Clinical Surgery at the Keck School of Medicine of USC. He specializes in thoracic surgery with a focus on malignant diseases of the chest, including esophageal cancer. His surgical training was completed at Beth Israel Deaconess Medical Center, Harvard Medical School, and he further specialized in thoracic surgery through a fellowship at MD Anderson Cancer Center. His training included advanced surgical techniques such as minimally invasive procedures (VATS) and robotic surgery. Dr. Atay also completed a clinical research fellowship at the National Cancer Institute (NCI) within the NIH, where his research concentrated on genetic alterations in malignant pleural mesothelioma. His current clinical and research efforts are centered on multimodality treatment approaches—combining chemotherapy, immunotherapy, radiation, and surgery—for patients with both early and advanced stage thoracic cancers. He is committed to delivering compassionate, high-quality care driven by the most current evidence, and he is actively involved in research that advances the understanding and treatment of thoracic malignancies.

Research topics

• Medicine
• Internal medicine
• Surgery
• Pathology
• Advertising
• Business
• Radiology
• Family medicine
• Psychology
• Medical education
• Intensive care medicine
• Oncology

Selected publications

• Association of Foregut Testing and Intervention on Lung Transplant Outcomes

  Journal of Clinical Gastroenterology · 2026-03-18

  article

  GOALS: To evaluate the association between multimodal foregut testing, surgical intervention, and outcomes after lung transplantation. BACKGROUND: Gastroesophageal reflux and other foregut disorders have been linked to adverse lung transplant outcomes. However, the impact of comprehensive pre- and posttransplant foregut evaluation and targeted intervention remains unclear. STUDY: We reviewed lung transplant recipients at a single center from February 2013 to April 2023. Data included demographics, pre- and posttransplant endoscopy, esophageal manometry, pH monitoring, esophagram, gastric emptying studies, and fundoplication status. The primary outcomes were the incidence, timing, and severity of acute cellular rejection, as well as mortality. RESULTS: Among 197 patients (median age 57 years old; 52% female), foregut testing patterns shifted post-transplant, with esophagram most frequent pre-transplant (44%) and gastric emptying studies most frequent post-transplant (62%). Manometry results were unchanged in 75% of patients, and gastric emptying studies in 27%. Over half developed new delayed gastric emptying posttransplant, including patients after fundoplication. Barrett's esophagus was associated with increased mortality risk. Fundoplication, typically performed in patients with abnormal pH studies, was linked to higher acute rejection incidence but lower severity and reduced mortality. CONCLUSIONS: Comprehensive foregut evaluation before and after lung transplantation reveals high rates of abnormal motility and reflux, dynamic posttransplant physiological changes, and an association between Barrett's esophagus and mortality. Structured foregut testing may help identify high-risk patients and guide timely intervention to improve outcomes.

  PublisherDOI

• Abstract 3407: Ensuring diverse <i>in vitro</i> lung cancer models: Development of alveolar epithelial cell lines, patient-derived xenografts, and lung adenocarcinoma cell lines from individuals of African ancestry

  Cancer Research · 2026-04-03

  article

  Abstract Lung cancer disproportionately affects individuals of African ancestry (AA), particularly men, who experience 12% higher incidence and 15% higher death rates than European ancestry men. The causes of this disparity remain unclear; while smoking is the most substantial risk factor for lung cancer, AA men do not smoke more than other groups. Despite the lack of understanding of the increased lung cancer risk and death rates of the AA population, the tools to study lung cancer in this group lag far behind; there are very limited in vitro models to study lung cancer in AA individuals. Lung adenocarcinoma (LUAD), which originates in the alveolar epithelium, is the most common histological subtype of lung cancer in all population groups. Immortalized human alveolar epithelial cells (ihAEC) are therefore key tools for studying the etiology and development of LUAD, as well as effects of environmental exposures. In addition to ihAECs, patient-derived xenografts (PDXs), in which human tumors are implanted and grown in immunodeficient mice, and in vitro cultured LUAD cell lines, are valuable tools to study lung adenocarcinoma. To address the shortage of these models developed from AA patients, we are taking a 3-pronged approach. We are collecting non-tumor and LUAD tissues from AA lung cancer patients with full consent. When sufficient non-tumor tissue is available, we isolate alveolar epithelial cells and use a CRISPR/Cas9-based gene-delivery approach to generate ihAEC lines. We direct genomic integration of simian virus 40 large-tumor antigen (SV40 LgT) and human telomerase reverse transcriptase (TERT) genes into the adeno-associated virus integration site 1 (AAVS1), a safe harbor region that prevents inserted genes from being silenced. When sufficient LUAD tumor tissue is available, we implant tumor sections subcutaneously in immunodeficient mice. Tumors that grow are explanted for a new round of propagation in mice as well as in vitro culture. To date one PDX has been established, and cells are in culture from non-tumor and tumor tissue; further characterization is in progress. The development of new ihAEC lines, PDXs, and LUAD cell lines will provide valuable tools for studying responses to environmental exposures, carcinogen detoxification processes, oncogenic transformation, and for testing novel therapies in ancestry-appropriate models. This work is an important step towards minimizing lung cancer health disparities caused by the absence of diverse model systems. Supported by grants U54CA233396, U54CA233444, and U54CA233465 from the National Institutes of Health (NIH)/National Cancer Institute (NCI), and the Norris Comprehensive Cancer Center core grant, award number P30CA014089 from the NIH/NCI. Citation Format: Karla E. Gonzalez, Bianca Dal Bo, Chunli Yan, Donna Loza, Matthew A. Gladstone, Anthony W. Kim, Scott M. Atay, Takashi Harano, W Dean Wallace, Ben Y. Tew, Bodour Salhia, Beiyun Zhou, Kweku Ofosu-Asante, Kyle R. Philips, Benjamin J. Ryder, Desmond Kwakye, Chase A. Lilly, Yong Huang, Nazarius S. Lamango, Ite A. Offringa. Ensuring diverse in vitro lung cancer models: Development of alveolar epithelial cell lines, patient-derived xenografts, and lung adenocarcinoma cell lines from individuals of African ancestry [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3407.

  PublisherDOI

• Lung transplant outcomes in recipients waitlisted for both single and double lung transplantation

  JHLT Open · 2026-04-02

  articleOpen access

  Background: Double lung transplantation (DLTx) generally provides better survival than single lung transplantation (SLTx).However, SLTx recipients may often be more deconditioned, making direct comparisons challenging.The objective of this study is to investigate the transplant outcomes in patients waitlisted for both SLTx and DLTx to minimize differences in recipient background. Methods:The United Network for Organ Sharing Database was retrospectively analyzed.The patients who were waitlisted from December 2017 to February 2023 then removed from the waiting list in the same time periods were included in the analysis.Pediatric lung transplantation, multiorgan transplantation, and lung retransplantation were excluded.Results: Of 4,056 patients who were waitlisted for both SLTx and DLTx, 2,127 patients (52.4%) received DLTx, 1,351 patients (33.3%) received SLTx, and 578 patients (14.3 %) were removed from waiting list without receiving lung transplantation.Median days on the waitlist were 36 days (IQR 12-93) for DLTx and 43 days (IQR 15-120) for SLTx.The lung allocation score at the time of transplantation was higher in DLTx group (50.58 19.02 vs 44.67 14.80, p < 0.001).DLTx recipients had significantly higher rates of prolonged intubation at 72 hours after transplantation (33.9% vs. 19.9%,p < 0.001) and reintubation (19.3% vs. 13.0%,p < 0.001).Median posttransplant length of hospital stay was longer in DLTx group (19 [IQR 14-33] vs 15 days [IQR 11-24], p < 0.001).DLTx group had significantly lower mortality than SLTx group (n= 3478, log-rank test p < 0.001).Conclusions: Among patients waitlisted for both DLTx and SLTx, DLTx recipients had lower post-transplant mortality, despite a more complicated immediate postoperative course compared to SLTx.

  PublisherOA PDFDOI

• Reconciling Survival Differences in Lobar and Sublobar Resection for Non–Small Cell Lung Cancer

  Journal of Surgical Research · 2026-03-31

  article
  PublisherDOI

• Impact of Ex Vivo Lung Perfusion on Inpatient Cost: A Propensity Score‐Matched Analysis of the US Nationwide Healthcare Cost and Utilization Project Database

  Clinical Transplantation · 2025-01-28 · 2 citations

  articleOpen access

  BACKGROUND: The goal of this study was to investigate the association between ex vivo lung perfusion (EVLP) use and inpatient hospitalization cost for lung transplantation in a nationwide sample. METHODS: Lung transplantation patients in 2018-2020 Nationwide Readmissions Database (NRD) were grouped based on use of EVLP. The primary outcome was total inpatient hospitalization cost. 1:2 propensity score matching by EVLP status was performed followed by multivariable linear regression to determine the association between inpatient cost and EVLP while adjusting for pre-transplant hospital days, high volume EVLP center status, and propensity score. RESULTS: There were 3902 lung transplants and 118 (3%) were recipients of EVLP lungs. Among EVLP patients, the median cost was $871 468 (IQR: $608 671-1 274 392), compared to $846 516 (IQR: $531 462-1 439 267, p = 0.871) among the total non-EVLP cohort. After 1:2 propensity score-matched cohort, recipients of EVLP lungs had longer median hospital length of stay (p = 0.046). In the multivariable model using the matched sample, increased cost was not associated with EVLP use (p = 0.783); however, high volume EVLP centers were associated with decreased cost (p = 0.018). CONCLUSIONS: EVLP use was not associated with greater inpatient costs and may be favorable at high volume centers.

  PublisherOA PDFDOI

• Social contributions as risk factors for readmissions after lung transplantation: Clinical and financial implications

  JHLT Open · 2025-05-26 · 1 citations

  articleOpen access

  <h2>Abstract</h2><h3>Introduction</h3> Lung transplant is associated with a 60-80% one-year post-transplant readmission rate. Social contributors represent potentially modifiable risk factors for readmission. We compared the clinical and financial of implications of readmissions associated with and without social factors. <h3>Methods</h3> Retrospective single-center study of lung transplant patients surviving to discharge between 2/2/2013 and 4/11/2023. Two reviewers categorized one-year readmissions into two groups: social (e.g., housing instability or rejection due to medication non-compliance) and non-social (e.g., pancreatitis). Sociodemographics, transplant indications, Stanford Integrated Psychosocial Assessment for Transplant (SIPAT) scores, lung allocation score (LAS), pre-operative hospitalization status, in-hospital post-operative course, and readmission costs were compared between patients with and without a social readmission. <h3>Results</h3> Among 192 transplants (109 double, 83 single) there were 436 one-year readmissions including 33 social readmissions. Reviewer inter-rater reliability was >95% and Kappa was 0.91. A social readmission occurred in 21 (11%) patients, and 9 of these patients had multiple social readmissions. A social readmission was either the first or second readmission for 81% of these patients. Patients with a social readmission had a greater median number of readmissions (4 vs. 2; p<0.001) and were associated with longer length of stay (8 vs 5 days; p<0.004), increased hospital costs ($23,813 vs. $14,245; p=0.04), and decreased margin (-$6,145 vs. $2,287; p<0.001). <h3>Conclusions</h3> Social readmissions represent a burden on patients and health systems. There is a strong association between social readmissions and increased costs, length of stay, and number of readmissions. Outpatient investment in patients with first-time social readmissions may improve outcomes and decrease healthcare costs.

  PublisherOA PDFDOI

• The Impact of Kidney Transplantation on Survival Outcomes for Lung Transplantation

  The Annals of Thoracic Surgery · 2025-10-10

  article
  PublisherDOI

• Left Upper Lobectomy After Coronary Artery Bypass Grafting: A National Analysis of Postoperative Outcomes

  Annals of Thoracic Surgery Short Reports · 2025-10-21

  articleOpen access

  <h2>Abstract</h2><h3>Background</h3> To determine whether previous coronary artery bypass grafting (CABG) is a risk factor for postoperative mortality, morbidity, or worse quality metrics in patients undergoing left upper lobectomy. <h3>Methods</h3> Using International Classification of Diseases 10th revision codes, the Healthcare Cost and Utilization Project Nationwide Readmissions Database was queried for patients with pulmonary neoplasms undergoing left upper lobectomy from 2016 to 2018 and categorized by history of CABG. Sociodemographic factors, comorbidities, and hospital characteristics were analyzed using univariable and multivariable regressions. <h3>Results</h3> A total of 11,118 patients met inclusion criteria, of whom 465 (4.2%) had a history of CABG. On bivariate analysis, postoperative myocardial infarction and atrial fibrillation rates were higher (<i>P</i> < .004) in patients with prior CABG. However, multivariable modeling revealed no association between history of CABG and worse outcomes across all metrics: in-hospital mortality (odds ratio [OR], 0.806; <i>P</i> = .563), cardiovascular complications (OR, 0.826; <i>P</i> = .0985), pulmonary complications (OR, 0.849; <i>P</i> = .154), length of stay (relative risk, 0.982; <i>P</i> = .619), 30-day readmission (OR, 1.057; <i>P</i> = .747), and 90-day readmission (OR, 1.137; <i>P</i> = .396). Thoracotomy patients experienced worse outcomes across all metrics (<i>P</i> < .05) compared with thoracoscopy. Prior CABG was not associated with worse outcomes in either thoracoscopy or thoracotomy subgroup analyses. Although not statistically significant (<i>P</i> = .255), patients with previous CABG had approximately $6500 in additional charges. <h3>Conclusions</h3> Prior CABG in patients undergoing left upper lobectomy is not associated with increased mortality, morbidity, length of stay, readmissions, or increased hospital charges. Thoracoscopy may be preferred in this population and is associated with improved outcomes compared with thoracotomy.

  PublisherOA PDFDOI

• Performance of the Charlson and Elixhauser Comorbidity Index Varies With the Type of Minimally Invasive Thoracic Surgery

  World Journal of Surgery · 2025-04-22

  article

  BACKGROUND: The Elixhauser (ECI) and Charlson-Deyo (CCI) comorbidity indices are two well-established measures used for assessing clinical prognosis and adjusting comorbidities in research. However, the optimal index is unclear within thoracic surgery. This study comparatively evaluates their effectiveness in predicting short-term outcomes (in-hospital mortality, complications, nonroutine discharge, and 30-/90-day readmissions) in minimally invasive pulmonary lobectomy (MIL) and minimally invasive Ivor Lewis esophagectomy (MIE). METHODS: Using the Healthcare Cost and Utilization Project National Readmission Database (2016-2018), MIL and MIE were identified using International Classification of Diseases, 10th Edition codes. Multivariable logistic regression models were constructed. The discriminative ability was quantified using the area under the receiver operating characteristic curve (AUC). The acceptable discriminative ability was defined as AUC > 0.70. RESULTS: CCI better predicted mortality (AUC 0.7866; 95% CI, 0.7549-0.8182) compared to ECI (AUC 0.7561; 95% CI, 0.7214-0.7908, p = 0.003) for MIL. The CCI marginally predicted nonroutine discharge (AUC 0.6427; 95% CI, 0.6362-0.6492 vs. ECI AUC 0.6399; 95% CI, 0.6333-0.6464, p = 0.01). In the MIE cohort, both the indices predicted mortality well (ECI 0.8038; 95% CI 0.7458-0.8618 vs. CCI 0.7969; 95% CI 0.7393-0.8546, p = 0.67). Neither index had acceptable discriminative ability for other outcomes. CONCLUSIONS: Based upon two commonly performed index thoracic procedures, the outcomes may differ by comorbidity measure employed and by surgery type, suggesting the need for careful selection of index, especially once patients are deemed fit for surgery. The CCI is superior in predicting mortality in patients with MIL. Both CCI and ECI are suitable for MIE. Furthermore, with the recent implementation of an updated ECI incorporating ICD-10 coding, these findings support the durability and robustness of the new ECI. Future research investigating their performances in predicting long-term outcomes in thoracic surgery may be warranted.

  PublisherDOI

• The surgical management of mediastinal paragangliomas: A modern series reflecting true collaboration between cardiac and thoracic surgeons

  JTCVS Open · 2025-11-20

  articleOpen accessSenior author

  Objective: The study objective was to evaluate the perioperative management and outcomes of operation for patients undergoing mediastinal paraganglioma resection at Keck Medical Center of University of Southern California. Methods: Six patients underwent surgical management for mediastinal paraganglioma from 2018 to 2024. Perioperative data were collected through retrospective review of the electronic medical record. Results: The median age was 61 years (range, 27-69). Five patients (83%) demonstrated elevated levels of urinary or plasma metanephrines. Germline genetic testing demonstrated a pathogenic mutation consistent with paraganglioma-pheochromocytoma syndrome in 3 patients (50%). A multidisciplinary approach was used in all cases, with cardiac and thoracic surgical staff attending to each patient. Operative approach was via median sternotomy in 4 patients (67%), clamshell thoracotomy in 1 patient (17%), and right posterolateral thoracotomy in 1 patient (17%). Cardiopulmonary bypass was used in 3 patients (50%); 2 patients required great vessel transection for exposure, and 1 patient required en bloc resection of the right main coronary due to tumor encasement. R0 resection was achieved in all patients. Median intensive care unit and hospital length of stay were 2.5 days (range, 1-4) and 5 days (range, 4-9), respectively. The predominant method of postoperative surveillance was biannual chest computed tomography, with no patients demonstrating radiographic evidence of recurrence during a median follow-up of 22.5 months (range, 6.5-85). Conclusions: Surgical resection of mediastinal paraganglioma is safe and feasible with a multidisciplinary approach. The use of cardiopulmonary bypass, although occasionally necessary, did not result in adverse outcomes. All patients achieved an R0 resection with minimal postoperative complications, and no evidence of recurrence has been observed during the follow-up period.

  PublisherOA PDFDOI

Frequent coauthors

• Anthony W. Kim

  University of Southern California

  42 shared

• Sean C. Wightman

  University of Southern California

  40 shared

• Julie A. Hong

  37 shared

• Mary Zhang

  National Cancer Institute

  37 shared

• David S. Schrump

  Center for Cancer Research

  37 shared

• Mahadev Rao

  36 shared

• Trevor Upham

  34 shared

• Xinmin Li

  31 shared