Ambulatory Extracorporeal Membrane Oxygenation Use in Patients Listed for Lung Transplant.

There has been an increase in the use of extracorporeal membrane oxygenation (ECMO) to bridge critically ill patients to lung transplant (LTX). This study evaluates how ambulatory status on ECMO affected waitlist and post-LTX outcomes. The United Network of Organ Sharing (UNOS) database was queried for patients aged of greater than or equal to 18 years and between 2016 and 2021 to identify pre-LTX patients supported by ECMO. The patients were classified in venous-arterial (VA) ECMO and veno-venous (VV) ECMO cohorts and further classified as ambulatory (AMB) and non-AMB (nAMB). Each cohort was controlled against the non-ECMO patients. Univariate statistical tests, as well as Kaplan-Meier survival curves, were used for analysis. The 90 day waitlist survival was the highest among the non-ECMO group (96%), but both AMB VV and VA groups had superior survival compared to the nAMB group (85% vs. 75%, 78% vs. 65%, p < 0.01). After adjusting for the median lung allocation score (LAS) (88) in the VV ECMO group, the waitlist survival was superior in the AMB VV ECMO compared to those not on ECMO (86% vs. 78%, p > 0.01). The 1 year post-LTX survival between non-ECMO and AMB VV ECMO was comparable (88% vs. 88%, p = 0.66). Ambulating patients or use of physical therapy while on ECMO can help improve lung transplant outcomes.

Disease diagnosis and severity classification in pulmonary fibrosis using carbonyl volatile organic compounds in exhaled breath.

BACKGROUND: Pathophysiological conditions underlying pulmonary fibrosis remain poorly understood. Exhaled breath volatile organic compounds (VOCs) have shown promise for lung disease diagnosis and classification. In particular, carbonyls are a byproduct of oxidative stress, associated with fibrosis in the lungs. To explore the potential of exhaled carbonyl VOCs to reflect underlying pathophysiological conditions in pulmonary fibrosis, this proof-of-concept study tested the hypothesis that volatile and low abundance carbonyl compounds could be linked to diagnosis and associated disease severity. METHODS: Exhaled breath samples were collected from outpatients with a diagnosis of Idiopathic Pulmonary Fibrosis (IPF) or Connective Tissue related Interstitial Lung Disease (CTD-ILD) with stable lung function for 3 months before enrollment, as measured by pulmonary function testing (PFT) DLCO (%), FVC (%) and FEV1 (%). A novel microreactor was used to capture carbonyl compounds in the breath as direct output products. A machine learning workflow was implemented with the captured carbonyl compounds as input features for classification of diagnosis and disease severity based on PFT (DLCO and FVC normal/mild vs. moderate/severe; FEV1 normal/mild/moderate vs. moderately severe/severe). RESULTS: The proposed approach classified diagnosis with AUROC=0.877 ± 0.047 in the validation subsets. The AUROC was 0.820 ± 0.064, 0.898 ± 0.040, and 0.873 ± 0.051 for disease severity based on DLCO, FEV1, and FVC measurements, respectively. Eleven key carbonyl VOCs were identified with the potential to differentiate diagnosis and to classify severity. CONCLUSIONS: Exhaled breath carbonyl compounds can be linked to pulmonary function and fibrotic ILD diagnosis, moving towards improved pathophysiological understanding of pulmonary fibrosis.

Evaluation of Lung Cancer Patient Response to First-Line Chemotherapy by Integration of Tumor Core Biopsy Metabolomics with Multiscale Modeling.

The standard of care for intermediate (Stage II) and advanced (Stages III and IV) non-small cell lung cancer (NSCLC) involves chemotherapy with taxane/platinum derivatives, with or without radiation. Ideally, patients would be screened a priori to allow non-responders to be initially treated with second-line therapies. This evaluation is non-trivial, however, since tumors behave as complex multiscale systems. To address this need, this study employs a multiscale modeling approach to evaluate first-line chemotherapy response of individual patient tumors based on metabolomic analysis of tumor core biopsies obtained during routine clinical evaluation. Model parameters were calculated for a patient cohort as a function of these metabolomic profiles, previously obtained from high-resolution 2DLC-MS/MS analysis. Evaluation metrics were defined to classify patients as Disease-Control (DC) [encompassing complete-response (CR), partial-response (PR), and stable-disease (SD)] and Progressive-Disease (PD) following first-line chemotherapy. Response was simulated for each patient and compared to actual response. The results show that patient classifications were significantly separated from each other, and also when grouped as DC vs. PD and as CR/PR vs. SD/PD, by fraction of initial tumor radius metric at 6 days post simulated bolus drug injection. This study shows that patient first-line chemotherapy response can in principle be evaluated from multiscale modeling integrated with tumor tissue metabolomic data, offering a first step towards individualized lung cancer treatment prognosis.

Lung cancer survival prediction and biomarker identification with an ensemble machine learning analysis of tumor core biopsy metabolomic data.

INTRODUCTION: While prediction of short versus long term survival from lung cancer is clinically relevant in the context of patient management and therapy selection, it has proven difficult to identify reliable biomarkers of survival. Metabolomic markers from tumor core biopsies have been shown to reflect cancer metabolic dysregulation and hold prognostic value. OBJECTIVES: Implement and validate a novel ensemble machine learning approach to evaluate survival based on metabolomic biomarkers from tumor core biopsies. METHODS: Data were obtained from tumor core biopsies evaluated with high-resolution 2DLC-MS/MS. Unlike biofluid samples, analysis of tumor tissue is expected to accurately reflect the cancer metabolism and its impact on patient survival. A comprehensive suite of machine learning algorithms were trained as base learners and then combined into a stacked-ensemble meta-learner for predicting "short" versus "long" survival on an external validation cohort. An ensemble method of feature selection was employed to find a reliable set of biomarkers with potential clinical utility. RESULTS: Overall survival (OS) is predicted in external validation cohort with AUROCTEST of 0.881 with support vector machine meta learner model, while progression-free survival (PFS) is predicted with AUROCTEST of 0.833 with boosted logistic regression meta learner model, outperforming a nomogram using covariate data (staging, age, sex, treatment vs. non-treatment) as predictors. Increased relative abundance of guanine, choline, and creatine corresponded with shorter OS, while increased leucine and tryptophan corresponded with shorter PFS. In patients that expired, N6,N6,N6-Trimethyl-L-lysine, L-pyrogluatmic acid, and benzoic acid were increased while cystine, methionine sulfoxide and histamine were decreased. In patients with progression, itaconic acid, pyruvate, and malonic acid were increased. CONCLUSION: This study demonstrates the feasibility of an ensemble machine learning approach to accurately predict patient survival from tumor core biopsy metabolomic data.

Lung cancer metabolomic data from tumor core biopsies enables risk-score calculation for progression-free and overall survival.

INTRODUCTION: Metabolomics has emerged as a powerful method to provide insight into cancer progression, including separating patients into low- and high-risk groups for overall (OS) and progression-free survival (PFS). However, survival prediction based mainly on metabolites obtained from biofluids remains elusive. OBJECTIVES: This proof-of-concept study evaluates metabolites as biomarkers obtained directly from tumor core biopsies along with covariates age, sex, pathological stage at diagnosis (I/II vs. III/VI), histological subtype, and treatment vs. no treatment to risk stratify lung cancer patients in terms of OS and PFS. METHODS: Tumor core biopsy samples obtained during routine lung cancer patient care at the University of Louisville Hospital and Norton Hospital were evaluated with high-resolution 2DLC-MS/MS, and the data were analyzed by Kaplan-Meier survival analysis and Cox proportional hazards regression. A linear equation was developed to stratify patients into low and high risk groups based on log-transformed intensities of key metabolites. Sparse partial least squares discriminant analysis (SPLS-DA) was performed to predict OS and PFS events. RESULTS: Univariable Cox proportional hazards regression model coefficients divided by the standard errors were used as weight coefficients multiplied by log-transformed metabolite intensity, then summed to generate a risk score for each patient. Risk scores based on 10 metabolites for OS and 5 metabolites for PFS were significant predictors of survival. Risk scores were validated with SPLS-DA classification model (AUROC 0.868 for OS and AUROC 0.755 for PFS, when combined with covariates). CONCLUSION: Metabolomic analysis of lung tumor core biopsies has the potential to differentiate patients into low- and high-risk groups based on OS and PFS events and probability.

Evaluation of disease staging and chemotherapeutic response in non-small cell lung cancer from patient tumor-derived metabolomic data.

OBJECTIVES: Despite extensive effort, the search for clinically-relevant metabolite biomarkers for early detection, disease monitoring, and outcome prediction in lung cancer remains unfulfilled. Although biofluid evaluation has been explored, the complexity inherent in metabolite data and the dynamic discrepancy between metabolites in biofluids vs. tumor tissue have prevented conclusive results. This proof-of-concept study explored models predictive of staging and chemotherapy response based on metabolomic analysis of fresh, patient-derived non-small cell lung cancer (NSCLC) core biopsies. MATERIALS AND METHODS: Samples (n = 36) were evaluated with high-resolution 2DLC-MS/MS and 13C-glucose enrichment, and the data were comprehensively analyzed with machine learning techniques. Patients were categorized as Disease-Control (DC) [encompassing complete-response (CR), partial-response (PR), and stable-disease (SD)] and Progressive-Disease (PD) in terms of first-line chemotherapy. Four major types of learning methods (partial least squares discriminant analysis (PLS-DA), support vector machines (SVM), artificial neural networks, and random forests (RF)) were applied to differentiate between positive (DC and CR/PR) and poor (PD and SD/PD) responses, and between stage I/II/III and stage IV disease. Models were trained with forward feature selection based on variable importance and tested on validation subsets. RESULTS: The models predicted patient classifications in the validation subsets with AUC (95 % CI): DC vs. PD (SVM), 0.970(0.961-0.979); CR/PR vs. SD/PD (PLS-DA), 0.880(0.865-0.895); stage I/II/III vs. IV (SVM), 0.902(0.880-0.924). Highest performing model was SVM for DC vs. PD (balanced accuracy = 0.92; kappa = 0.74). CONCLUSION: This study illustrates a comprehensive evaluation of patient tumor-specific metabolic profiles, with the potential to identify disease stage and predict response to first-line chemotherapy.

Intraoperative Esophagogastroduodenoscopy During Heller Myotomy: Evaluating Guidelines.

BACKGROUND: National guidelines suggest routine intraoperative esophagogastroduodenoscopy (EGD) during laparoscopic Heller myotomy (LHM) to assess for mucosal perforation and myotomy adequacy, but the utility of this is unknown. This study aimed to evaluate the effect of intraoperative EGD on outcomes after LHM. METHODS: Patients who underwent LHM in a single center were retrospectively identified. Outcomes were compared between patients who did and did not undergo intraoperative EGD. RESULTS: Sixty-one patients were reviewed: 46 (75%) underwent intraoperative EGD and 15 (25%) did not. Mucosal perforations occurred in 2 (4%) of the EGD group and 3 (20%) of the non-EGD group (p = 0.06). All perforations, regardless of EGD use, were recognized laparoscopically. There were no postoperative leaks. Failed myotomy occurred in 5 (11%) who underwent EGD and 1 (7%) who did not (p = 0.64). CONCLUSIONS: Because EGD does not appear to improve outcomes after LHM, we emphasize its selective, rather than routine, use.

Optimal Surgical Timing After Neoadjuvant Therapy for Stage IIIa Non-Small Cell Lung Cancer.

BACKGROUND: Patients with clinically/pathologically diagnosed stage IIIa non-small cell lung cancer (NSCLC) considered for surgery are recommended to undergo neoadjuvant chemotherapy with or without radiation. The timing of an operation after therapy is not standardized; therefore, we investigated the timing of intervention after neoadjuvant therapy and the impact on outcomes in this demographic. METHODS: The National Cancer Database was queried between 2010 and 2015 for patients with clinical/pathologic stage IIIa NSCLC. Patients were then divided into short (<77 days), mid (77-114 days), and long delay (>114 days) groups based on interquartile values. These groups were then compared for age, race, gender, insurance type, Charlson-Deyo score, length of stay, readmission rate, and overall survival based on timing of operation. RESULTS: There were 31,357 patients with clinical/pathologic stage IIIa NSCLC, and 5946 patients underwent surgical intervention. Preoperatively 3593 patients underwent chemoradiotherapy, 2185 underwent chemotherapy only, and 168 received radiation alone. The short, mid, and long delay groups were clinically and statistically similar in age, gender, insurance type, comorbidity index, treating facility type, and distance from home. Long delay groups had larger tumor size compared with other groups. Postoperative length of stay, rates of 30-day readmission, and 30- and 90-day mortality were similar across all groups. Cox modeling demonstrated a significant difference in survival when patients underwent earlier operative intervention compared with late operative intervention and when patients received chemoradiation compared with chemotherapy alone. Short, mid, and long delay group 1-year survivals were 82%, 83%, and 80% and 3-year survival 59%, 58%, and 52%, respectively (P = .0003). CONCLUSIONS: The delay in surgical resection of stage IIIa NSCLC is not associated increased early mortality; however, it is associated with worse 3-year postresection survival.

Objective Donor Scoring System for Lung Transplantation.

BACKGROUND: There is no objective method to estimate post-lung transplant survival solely on the basis of cumulative donor risk factors. METHODS: The United Network Organ Sharing thoracic transplant database was queried to identify patients who underwent lung transplantation between 2005 and 2015. A Cox proportional hazard model was generated using a training set to identify donor risk factors significantly associated with posttransplant survival. Significant donor risk factors were assigned a score on the basis of their hazard ratio. Donor risk score was calculated for each patient by adding the individual donor risk factor scores. Donors in the validation set were then categorized into low-risk (score = 0), intermediate-risk (score = 1), and high-risk (score >1) categories on the basis of the cumulative risk score. The Lung Allocation Score was used as a surrogate for recipient risk. Survival for each risk group was calculated using Kaplan-Meier curves. RESULTS: The donor risk groups' respective survival at 1 year was 85%, 81%, and 77%, and at 5 years it was 53%, 50%, and 42% (p < 0.001). The combination of low-risk recipients and low-risk donors had 1- and 5-year survival of 89% and 59%, respectively. The combination of high-risk recipients and high-risk donors had 1- and 5-year survival of 70% and 30%, respectively. CONCLUSIONS: The proposed lung donor scoring system is a simple, easy to use method that can aid transplant surgeons in the selection of a potential lung transplant donor. Using the lung donor score in conjunction with the Lung Allocation Score can allow for matching of recipients and donors, to optimize posttransplant outcomes.

Pharmacokinetic/pharmacodynamic modeling of combination-chemotherapy for lung cancer.

Chemotherapy for non-small cell lung cancer (NSCLC) typically involves a doublet regimen for a number of cycles. For any particular patient, a course of treatment is usually chosen from a large number of combinational protocols with drugs in concomitant or sequential administration. In spite of newer drugs and protocols, half of patients with early disease will live less than five years and 95% of those with advanced disease survive for less than one year. Here, we apply mathematical modeling to simulate tumor response to multiple drug regimens, with the capability to assess maximum tolerated dose (MTD) as well as metronomic drug administration. We couple pharmacokinetic-pharmacodynamic intracellular multi-compartment models with a model of vascularized tumor growth, setting input parameters from in vitro data, and using the models to project potential response in vivo. This represents an initial step towards the development of a comprehensive virtual system to evaluate tumor response to combinatorial drug regimens, with the goal to more efficiently identify optimal course of treatment with patient tumor-specific data. We evaluate cisplatin and gemcitabine with clinically-relevant dosages, and simulate four treatment NSCLC scenarios combining MTD and metronomic therapy. This work thus establishes a framework for systematic evaluation of tumor response to combination chemotherapy. The results with the chosen parameter set indicate that although a metronomic regimen may provide advantage over MTD, the combination of these regimens may not necessarily offer improved response. Future model evaluation of chemotherapy possibilities may help to assess their potential value to obtain sustained NSCLC regression for particular patients, with the ultimate goal of optimizing multiple-drug chemotherapy regimens in clinical practice.

Solid-Organ Transplantation and the Affordable Care Act: Accessibility and Outcomes.

The objective of the study is to evaluate the impact of the Affordable Care Act (ACA) on accessibility to solid organ transplant and outcomes. Data source registry: United Network of Organ Sharing database. Patients aged ≥18 years listed for kidney, liver, heart, and lung transplant between years 2010 and 2016 were classified by insurance and status of Medicaid adoption under ACA to evaluate insurance distribution. Between 2010 and 2016, states that adopted Medicaid had 2 to 4 per cent point increase in the proportion of patients listed with Medicaid across all organs. One-year waiting list survival of Medicaid patients was better in the ACA era. States that expanded Medicaid under the ACA had a significant increase in the proportion of patients listed with Medicaid and better one-year waiting list survival.

Does Donor Age and Double Versus Single Lung Transplant Affect Survival of Young Recipients?

BACKGROUND: In an effort to expand the donor pool for lung transplants, numerous studies have examined the use of advanced age donors with mixed results, including decreased survival among younger recipients. We evaluated the impact of the use of advanced age donors and single versus double lung transplantation on posttransplant survival. METHODS: The United Network for Organ Sharing database was retrospectively queried between January 2005 and June 2014 to identify lung transplant patients aged at least 18 years. Patients were stratified by recipient age 50 years or less, donor age 60 years or more, and single versus double lung transplantation. Overall survival was assessed using the Kaplan-Meier method. Multivariable survival analysis was performed using a Cox proportional hazards model. RESULTS: In all, 14,222 lung transplants were performed during the study period. With univariate analysis, donor lungs aged 60 years or more were associated with slightly worse 5-year survival (44% versus 52%; p < 0.001). Among recipients aged more than 50 years, this trend was not present in the multivariate model (hazard ratio 1.23, p = 0.055). Among recipients aged 50 years or more, receiving older donor lungs showed worse survival with the use of single lung transplant (5-year survival 15% versus 50%, p = 0.01). No significant difference in survival between young and old donors was seen when double lung transplant was performed (p = 0.491). Cox proportional hazards model showed a trend toward interaction between single lung transplantation and older donors (hazard ratio 2.36, p = 0.057). CONCLUSIONS: Reasonable posttransplant outcomes can be achieved with use of advanced age donors in all recipient groups. Double lung transplantation should be performed when older donors (age more than 60) are used in young recipients (age 50 or less).

Prediction of lung cancer patient survival via supervised machine learning classification techniques.

Outcomes for cancer patients have been previously estimated by applying various machine learning techniques to large datasets such as the Surveillance, Epidemiology, and End Results (SEER) program database. In particular for lung cancer, it is not well understood which types of techniques would yield more predictive information, and which data attributes should be used in order to determine this information. In this study, a number of supervised learning techniques is applied to the SEER database to classify lung cancer patients in terms of survival, including linear regression, Decision Trees, Gradient Boosting Machines (GBM), Support Vector Machines (SVM), and a custom ensemble. Key data attributes in applying these methods include tumor grade, tumor size, gender, age, stage, and number of primaries, with the goal to enable comparison of predictive power between the various methods The prediction is treated like a continuous target, rather than a classification into categories, as a first step towards improving survival prediction. The results show that the predicted values agree with actual values for low to moderate survival times, which constitute the majority of the data. The best performing technique was the custom ensemble with a Root Mean Square Error (RMSE) value of 15.05. The most influential model within the custom ensemble was GBM, while Decision Trees may be inapplicable as it had too few discrete outputs. The results further show that among the five individual models generated, the most accurate was GBM with an RMSE value of 15.32. Although SVM underperformed with an RMSE value of 15.82, statistical analysis singles the SVM as the only model that generated a distinctive output. The results of the models are consistent with a classical Cox proportional hazards model used as a reference technique. We conclude that application of these supervised learning techniques to lung cancer data in the SEER database may be of use to estimate patient survival time with the ultimate goal to inform patient care decisions, and that the performance of these techniques with this particular dataset may be on par with that of classical methods.

Application of unsupervised analysis techniques to lung cancer patient data.

This study applies unsupervised machine learning techniques for classification and clustering to a collection of descriptive variables from 10,442 lung cancer patient records in the Surveillance, Epidemiology, and End Results (SEER) program database. The goal is to automatically classify lung cancer patients into groups based on clinically measurable disease-specific variables in order to estimate survival. Variables selected as inputs for machine learning include Number of Primaries, Age, Grade, Tumor Size, Stage, and TNM, which are numeric or can readily be converted to numeric type. Minimal up-front processing of the data enables exploring the out-of-the-box capabilities of established unsupervised learning techniques, with little human intervention through the entire process. The output of the techniques is used to predict survival time, with the efficacy of the prediction representing a proxy for the usefulness of the classification. A basic single variable linear regression against each unsupervised output is applied, and the associated Root Mean Squared Error (RMSE) value is calculated as a metric to compare between the outputs. The results show that self-ordering maps exhibit the best performance, while k-Means performs the best of the simpler classification techniques. Predicting against the full data set, it is found that their respective RMSE values (15.591 for self-ordering maps and 16.193 for k-Means) are comparable to supervised regression techniques, such as Gradient Boosting Machine (RMSE of 15.048). We conclude that unsupervised data analysis techniques may be of use to classify patients by defining the classes as effective proxies for survival prediction.