Faculty Bibliography

BACKGROUND:Fewer than 20% of donated lungs are transplanted. We examined the impacts of donor management and recovery at an in-hospital donor care unit (DCU) established in 2021 versus the donor hospital (DH). METHODS:(P/F ratio) were examined as a hypothesized effect modifier. We projected the national impact of improved utilization on lung transplant volume. RESULTS:After adjusting for donor differences, lung utilization was 88% higher (aRR: 1.88; 95% CI: 1.40, 2.53) in the DCU versus DH setting. Median P/F ratio increased from 275 to 348 mmHg (p<0.0001) in the DCU and explained approximately 38% of the improvement in lung utilization. Non-lung organ utilization was either improved or non-inferior in the DCU. Lung graft survival and pulmonary function were similar for recipients of DCU versus DH-managed donors. Nationally, an 88% improvement in lung utilization among donors not currently transferred to a DCU could theoretically result in ≥300 more lung transplants per year. CONCLUSIONS:Lung transplantation can be nearly doubled through lung-centric donor management in a DCU, without sacrificing recipient outcomes nor the utilization of other organs. Donor management practices to optimize lung utilization should be proliferated by establishing more DCUs and, where feasible, applying lung-centric protocols in donor hospitals.

The da Vinci 5 (DV5) system has purported advantages over existing platforms, including forced feedback. We compare perioperative safety, operative efficiency and early oncologic quality between DV5 and Xi robotic platforms for anatomic lung resection. A total of 419 robotic lung resections were performed between January 2025 and August 2026. Propensity-score overlap weighting balanced and matched 294 patients (47 DV5, 247 Xi) for age, sex, ASA class, BMI, forced expiratory volume in 1 s, tumor size, clinical T and N category, resection type and laterality. Effects were reported with robust and bootstrap 95% confidence intervals. Sensitivity analyses included 1:1 and 1:3 propensity matching comparisons and E-value assessment. From January 2025 to August 2026, 294 patients (47 DV5, 247 Xi) had balanced baseline characteristics (SMD < 0.01). Operative time was shorter with DV5 (96.4 vs. 118.7 min; 95% CI, - 36.2 to - 8.5). Lymph node yield was comparable (DV5 29.9 vs. Xi 31.6; [- 7.2 to + 3.7]). Blood loss (p < 0.01), length of stay (1.05 vs. 1.28 days; [- 0.53, + 0.06]) and stapler loads (8.5 vs. 9.6; [- 2.81 to + 0.55]) were numerically lower with DV5. Major complications were numerically lower with DV5 (1/47 [2.0%] vs. 20/247 [8.2%]; -16.0 to + 3.6). Both had 100% R0 resections and no conversion. There was no 30- or 90-day mortality. In this first clinical comparison of da Vinci 5 and Xi for robotic anatomic lung resection, DV5 preserved the safety and oncologic quality of Xi while being associated with shorter operative times. Prospective multi-institutional studies are required to confirm these findings.

BACKGROUND/UNASSIGNED:Radiation pneumonitis (RP) remains a significant treatment-related toxicity in patients with unresectable, locally advanced non-small cell lung cancer (NSCLC) undergoing chemoradiotherapy (CRT). Most existing predictive models rely on static baseline demographic or dosimetry variables and lack real-time clinical applicability. We developed a novel predictive framework that integrates longitudinal symptom data extracted from clinical notes using natural language processing (NLP) with clinical and dosimetry features to improve early RP prediction. METHODS/UNASSIGNED:We retrospectively identified 227 patients with locally advanced NSCLC treated with definitive CRT at a high-volume cancer center in the United States. We included all patients older than 18 years who were diagnosed between Jan 1, 2006, and Dec 31, 2022 with histologically or cytologically confirmed unresectable Stage 2 or 3 NSCLC and treated with conformal radiotherapy to a minimum dose of ≥45 Gy with or without chemotherapy. Of these, 31 RP events were identified through manual adjudication using radiologic criteria and chart review. NLP was used to extract the temporal relationship of 16 pre-specified symptoms with treatment from over 100,000 clinical notes spanning pre- and during-treatment intervals. We trained and validated machine learning models on combinations of baseline clinical data, radiation dosimetry, and NLP-derived symptom features. Model performance was evaluated using a nested cross-validation framework, with an outer cross-validation loop reserved for performance assessment and an inner cross-validation loop used for model training and integration, and summarized using area under the receiver operating characteristic curve (AUC) and partial AUC (pAUC) at high specificity thresholds. Clinical utility was evaluated using decision curve analysis (DCA). FINDINGS/UNASSIGNED:The best-performing model incorporated longitudinal NLP features and achieved a median AUC of 0.759 (90% confidence interval 0.753-0.766), significantly outperforming baseline models using only dosimetry (AUC 0.613) or clinical variables (AUC 0.635). NLP-based features such as cough trajectory, shortness of breath, and wheezing were among the most important predictors. Inclusion of NLP-derived symptom data improved early identification of high-risk patients, particularly in the clinically relevant high-specificity range (pAUC 0.021 vs. 0.010 for dosimetry alone). DCA showed that the calibrated MLP model provided greater net benefit than default strategies of treating all or no patients across clinically relevant threshold possibilities. INTERPRETATION/UNASSIGNED:In this early work, NLP-based extraction of longitudinal symptoms from routine clinical documentation meaningfully enhances RP prediction in patients undergoing CRT for NSCLC. This approach leverages existing electronic health record infrastructure to deliver real-time, scalable, and interpretable risk estimates, offering a pathway toward potential early intervention and personalized toxicity management. The model and DCA requires external and prospective validation before clinical deployment; as such, future work should focus on this validation and integration into clinical decision support systems. FUNDING/UNASSIGNED:AstraZeneca.

BACKGROUND/UNASSIGNED:Efforts to reduce waitlist mortality in lung transplantation have been hindered by a high donor lung non-use rate, with approximately one-fifth of deceased organ donors ultimately providing lungs for transplantation. Rationale for organ decline varies across individual providers and transplant centers. Greater donor lung utilization could significantly reduce waitlist mortality and improve clinical outcomes for adult lung transplant candidates. OBJECTIVE/UNASSIGNED:This study aimed to evaluate whether machine learning models can successfully predict donor lung non-use and characterize donor factors associated with current utilization patterns, with the long-term goal of informing more consistent and data-driven donor lung evaluation. METHODS/UNASSIGNED:Using U.S. national registry data collected between 2012 and 2022, we generated five machine learning models - logistic regression, decision tree, random forest, XGBoost, and Naive Bayes - all of which were trained with 19 donor variables to predict donor lung non-use. Models were tested with both full donor feature sets and restricted donor feature sets comprised of the top 15, 10, and 5 predictors. Performance metrics, including area under the receiver operating characteristic curve (AUROC), accuracy, precision, recall, and F1 score, were assessed using cross-validation. RESULTS/UNASSIGNED:ratio, donor age, body mass index, and serum aspartate aminotransferase levels. More restricted random forest models (i.e., those that only included 5 or 10 predictors) and any of the restricted logistic regression and XGBoost models had inferior but still acceptable performance (AUROC 0.867-0.958). CONCLUSION/UNASSIGNED:Machine learning models, particularly random forest, XGBoost, and logistic regression, accurately predicted donor lung non-use. Restricted models using fewer predictors also maintained strong performance, demonstrating technical feasibility for future evaluation in time-sensitive workflows. Because the models were trained on historical utilization decisions rather than recipient outcomes or independent measures of donor suitability, prospective validation is needed to determine whether they can improve appropriate donor lung utilization or support acceptance decisions without reinforcing practice-pattern bias.

BACKGROUND/UNASSIGNED:Minimally invasive pneumonectomy via robotic or video-assisted platforms is not yet widely adopted. Rates of conversion-to-thoracotomy and 90-day mortality are reported at 30%-50% and 12%, respectively. We describe our experience with the largest single-institution consecutive series of robotic pneumonectomy. METHODS/UNASSIGNED:This is a retrospective review of prospective database of a consecutive (non-selected) series of patients who underwent robotic pneumonectomy from January 2019 to December 2024 at our institution. RESULTS/UNASSIGNED:There were 21 patients (67% male) with median age 70 (range 15-79), DLCO 76% (range 48%-106%), and FEV1 76% (range 44%-126%). Indication for pneumonectomy was cancer in 17 patients (81%), of which 15 (88%) had a primary lung cancer. Four patients (19%) had destroyed lung from infection, and six (29%) underwent completion pneumonectomy. Thirteen (76%) of the 17 cancer patients had neoadjuvant therapy, including chemotherapy (11 patients), immunotherapy (10 patients), and radiation therapy (6 patients). A robotic approach was selected for all patients, with planned conversion-to-open in two patients (9%). There were no intraoperative transfusions. Median chest tube duration was one day (range 1-4). Median length of stay was five days (range 1-25). Thirty-day readmission was 19% (4 patients). Thirty-day morbidity was 19%, including atrial fibrillation in two patients, hemothorax in one patient, and chylothorax in one patient. Thirty-day mortality was 0% and 90-day mortality was 5% (one patient). CONCLUSION/UNASSIGNED:Pneumonectomy can be performed safely and effectively robotically even in high-risk patients with previous ipsilateral thoracotomy and/or neoadjuvant chemo-, immuno-, or radio-therapy. We demonstrate that a robotic approach enables minimal unplanned conversion-to-thoracotomy, no blood transfusions and excellent short-term outcomes. Our pre-, intra- and post-operative strategies are shared.

Primary graft dysfunction (PGD) is a proinflammatory syndrome occurring within the first days following lung transplantation. It is initiated by ischemia-reperfusion injury and perpetuated by donor and recipient immunologic factors, resulting in alveolar damage and progressive hypoxemic respiratory failure.¹ PGD is a known risk factor for both early allograft failure and chronic lung allograft dysfunction (CLAD).² Incidence of severe PGD remains high at 10-25%, though is variable; risk factors for PGD include center experience, underlying recipient disease type, size matching, donor lung storage conditions, operative time, and post-operative management.² Strategies to prevent PGD or mitigate the long -term consequences after it develops, are sorely needed. However, lack of specificity of the current PGD definition may hamper further progress in the field, especially as it pertains to the development of robust and relevant clinical trials. We propose that future modifications of the PGD definition incorporate more objective surrogates of allograft injury and subsequent diffuse alveolar damage, which may improve our ability to accurately study disease pathogenesis and improve outcomes.

In the face of a growing mismatch between candidates awaiting transplantation and the supply of conventional donor organs, attention has shifted toward novel methods to increase the donor pool, including the use of donation after circulatory death (DCD) and the refinement of procurement techniques that safeguard graft quality. Thoracoabdominal normothermic regional perfusion (TA-NRP) has emerged as a new strategy, leveraging extracorporeal support to curtail warm-ischemic injury while permitting in situ functional assessment. This review covers the rationale behind the use of TA-NRP, while outlining its use during procurement and the current body of evidence gathered on it implementation in lung transplantation specifically.

BACKGROUND:Donor to recipient size matching is an essential part of lung transplantation, with significant mismatch leading to worse patient outcomes. The current practice is based on limited data along with broadly accepted themes that have not been articulated in the form of objective and definitive guidelines. The objective of the American Association for Thoracic Surgery (AATS) Clinical Practice Standards Committee (CPSC) expert panel was to develop evidence- and expert-based recommendations for optimal donor to recipient lung allograft size matching based on review of the existing literature. METHODS:The AATS CPSC assembled an expert panel of 18 lung transplant surgeons from 15 centers who developed a consensus document of recommendations. The panel was divided into subgroups covering size-matching in (1) bilateral lung transplantation, (2) single lung transplantation, (3) lobar transplantation, (4) unique situations, and (5) management of complications following severe size mismatch. Following a focused literature review, each subgroup formulated recommendation statements for each subtopic, which were reviewed and further refined using a Delphi process until consensus was achieved on each final statement by the voting group. RESULTS:The expert panel achieved consensus on 20 recommendations for current best practices in donor and recipient size matching. These recommendations include utilization of a ratio of donor-to-recipient predicted total lung capacity between 0.8 to 1.2, with special considerations based on recipient pathology, single lung or lobar transplantation, and anatomic variations such as chest wall abnormalities or significant mediastinal shift. Furthermore, oversized allografts can be reduced in size via non-anatomic or anatomic resection in select cases when required. CONCLUSIONS:Consistent practice guidelines regarding donor to recipient size matching will be helpful and important to achieve optimal outcomes in lung transplantation. The recommendations described here provide guidance for professionals involved in the care of patients with end-stage lung disease considered for transplantation.

OBJECTIVE/UNASSIGNED:Mid-to-distal tracheal surgery for cancer can be safely performed minimally invasively with a one-day length of stay, avoiding a guardian chin suture, and ensuring a R0 resection in select patients. METHODS/UNASSIGNED:This is a retrospective technical review of the largest series to date of patients with mid-to-distal tracheal cancers. All were offered a right robotic approach using veno-venous extracorporeal membrane oxygenation (VV ECMO) support via percutaneous right internal jugular vein and right common femoral vein access. RESULTS/UNASSIGNED:From May 2019 to April 2024, five consecutive patients (3 men, 2 women; aged 11, 29, 37, 40, and 74 years) presented with a mid-to-distal tracheal cancer. All underwent right robotic mid-distal tracheal resections on VV ECMO for primary tracheal cancers. All patients had an end-to-end tracheal anastomosis and R0 resection and all avoided: systemic heparinization, suprahyoid release maneuvers and a postoperative guardian chin stitch. Median operative time was 258 min (range 227-292). All patients tolerated the operations well and were discharged home on the morning of postoperative day 1. There was no minor or major morbidity, no 30 or 90-day mortality, and no re-admissions. Two patients complained of cough. All had R0 resections and to date none have evidence of recurrent disease or stricture. CONCLUSION/UNASSIGNED:Resection of mid-to-distal primary tracheal cancers can be performed safely and efficiently via a right robotic approach while on VV ECMO with little to no morbidity or mortality and require only an overnight hospital stay. The techniques used to perform the operation and achieve these results are described.