Primary graft dysfunction after lung transplantation

Advancing lung transplantation through machine learning and artificial intelligence

PURPOSE OF REVIEW

To explore the current applications of artificial intelligence and machine learning in lung transplantation, including outcome prediction, drug dosing, and the potential future uses and risks as the technology continues to evolve.

RECENT FINDINGS

While the use of artificial intelligence (AI) and machine learning (ML) in lung transplantation is relatively new, several groups have developed models to predict short-term outcomes, such as primary graft dysfunction and time-to-extubation, as well as long-term outcomes related to survival and chronic lung allograft dysfunction. Additionally, drug dosing models for Tacrolimus levels have been designed, demonstrating proof of concept for modelling treatment as a time-series problem.

SUMMARY

The integration of ML models with clinical decision-making has shown promise in improving post-transplant survival and optimizing donor lung utilization. As technology advances, the field will continue to evolve, with enhanced datasets supporting more sophisticated ML models, particularly through real-time monitoring of biological, biochemical, and physiological data.

Interventional Strategies for Children with Progressive Pulmonary Hypertension Despite Optimal Therapy: An Official American Thoracic Society Clinical Practice Guideline

Background: Pulmonary hypertension in children is progressive with wide variability in prognosis. This document provides an evidence-based clinical practice guideline for the management of children with progressive pulmonary hypertension despite optimal therapy. Methods: A multidisciplinary panel identified pertinent questions regarding the management of children with pulmonary hypertension that has progressed despite optimal therapy, conducted systematic reviews of the relevant literature, and applied the Grading of Recommendations, Assessment, Development and Evaluation approach to develop clinical recommendations. Results: After reviewing the research evidence, the panel considered the balance of desirable (benefits) and undesirable (harms and burdens) effects of the interventions in each proposed question. Valuation of our main outcomes was also considered, together with resources required, equity, acceptability, and feasibility. Recommendations were developed for or against interventional strategies specific to children with pulmonary hypertension that has progressed despite optimal therapy. Conclusions: Although there is a growing population of children with pulmonary hypertension, there is a striking lack of empirical evidence regarding management of those whose disease has progressed despite optimal pharmacotherapy. The panel formulated and provided the rationale for clinical recommendations for or against interventional strategies on the basis of this limited empirical evidence, coupled with expert opinion, to aid clinicians in the management of these complex pediatric patients. In addition, we identified important areas for future research.

Nrf2/HO-1 Pathway Mediated Protective Effects of Hydrogen in a Model of Lung Transplantation Simulated by Rat Pulmonary Microvascular Endothelial Cells

This study aimed to observe the mechanism of hydrogen (H2) in a lung transplantation model simulated by pulmonary microvascular endothelial cells (PMVECs), which were divided into 5 groups. The blank group was the normal PMVECs. During cold ischemia period, PMVECs in the control, O2, or H2 groups were aerated with no gas, O2, or 3% H2, and 3% H2 after transfected with a small interfering RNA targeting Nrf2 in the H2+si-Nrf2 group. Treatment with O2 and H2 decreased the oxidative stress injury, inflammation, cell apoptosis, and attenuated energy metabolism compared with the control group (P < 0.05). And the H2 group showed a better outcome with the increased protein expression of the Nrf2 and HO-1, which were conversed in the H2+si-Nrf2 group. In conclusion, H2 attenuated inflammation, oxidative stress injury, cell apoptosis, and maintained the balance between energy supply and demand in a rat PMVECs lung transplantation model via Nrf2/HO-1.

Machine Learning for Predicting Primary Graft Dysfunction After Lung Transplantation: An Interpretable Model Study

BACKGROUND

Primary graft dysfunction (PGD) develops within 72 h after lung transplantation (Lung Tx) and greatly influences patients' prognosis. This study aimed to establish an accurate machine learning (ML) model for predicting grade 3 PGD (PGD3) after Lung Tx.

METHODS

This retrospective study incorporated 802 patients receiving Lung Tx between July 2018 and October 2023 (640 in the derivation cohort and 162 in the external validation cohort), and 640 patients were randomly assigned to training and internal validation cohorts in a 7:3 ratio. Independent risk factors for PGD3 were determined by integrating the univariate logistic regression and least absolute shrinkage and selection operator regression analyses. Subsequently, 9 ML models were used to construct prediction models for PGD3 based on selected variables. Their prediction performances were further evaluated. Besides, model stratification performance was assessed with 3 posttransplant metrics. Finally, the SHapley Additive exPlanations algorithm was used to understand the predictive importance of selected variables.

RESULTS

We identified 9 independent clinical risk factors as selected variables. Among 9 ML models, the random forest (RF) model displayed optimal performance (area under the curve [AUC] = 0.9415, sensitivity [Se] = 0.8972, specificity [Sp] = 0.8795 in the training cohort; AUC = 0.7975, Se = 0.7520, Sp = 0.7313 in the internal validation cohort; and AUC = 0.8214, Se = 0.8235, Sp = 0.6667 in the external validation cohort). Further assessments on calibration and clinical usefulness indicated the promising applicability of the RF model in PGD3 prediction. Meanwhile, the RF model also performed best in terms of risk stratification for postoperative support (extracorporeal membrane oxygenation time: P < 0.001, mechanical ventilation time: P = 0.006, intensive care unit time: P < 0.001).

CONCLUSIONS

The RF model had the optimal performance in PGD3 prediction and postoperative risk stratification for patients after Lung Tx.

Enhancing lung transplantation with ECMO: a comprehensive review of mechanisms, outcomes, and future considerations

BACKGROUND

Lung transplantation (LTx) is a critical intervention for patients with end-stage lung disease. However, challenges such as donor organ scarcity and post-transplant complications significantly affect its success. Recent advancements in Extracorporeal Membrane Oxygenation (ECMO) have shown promise in improving the outcomes and expanding eligibility for LTx.

METHODS

A comprehensive literature review was conducted, focusing on studies that explore the use of ECMO in lung transplantation. A thorough search of relevant studies on ECMO and LTx was conducted using multiple scholarly databases and relevant keywords, resulting in 73 studies that met the inclusion criteria. Sources included peer-reviewed journals and clinical trial results, with emphasis on articles captured recent advancements in ECMO technology and techniques.

RESULTS

ECMO has been crucial in supporting patients before, during, and after LTx. It serves as a bridge to transplantation by maintaining pulmonary and circulatory stability in critically ill patients awaiting donor organs. ECMO also aids in the evaluation of marginal donor lungs and supports patients through acute post-transplant complications. Recent technological advancements have improved the safety and efficacy of ECMO, further solidifying its role in LTx.

CONCLUSION

In conclusion, this review underscores ECMO's critical role in enhancing outcomes across all stages of lung transplantation. Its various configurations and strategies have shown promise in stabilizing critically ill patients and improving transplant success rates. Looking ahead, it's important to gather more information about the long-term outcomes and potential complications associated with ECMO use. More research and data collection will help us understand the benefits and risks better, leading to improved decision-making and patient care in this field.

Transient heat stress protects from severe endothelial damage and dysfunction during prolonged experimental ex-vivo lung perfusion

Introduction

The pulmonary endothelium is the primary target of lung ischemia-reperfusion injury leading to primary graft dysfunction after lung transplantation. We hypothesized that treating damaged rat lungs by a transient heat stress during ex-vivo lung perfusion (EVLP) to elicit a pulmonary heat shock response could protect the endothelium from severe reperfusion injury.

Methods

Rat lungs damaged by 1h warm ischemia were reperfused on an EVLP platform for up to 6h at a constant temperature (T°) of 37°C (EVLP37°C group), or following a transient heat stress (HS) at 41.5°C from 1 to 1.5h of EVLP (EVLPHS group). A group of lungs exposed to 1h EVLP only (pre-heating conditions) was added as control (Baseline group). In a first protocol, we measured lung heat sock protein expression (HSP70, HSP27 and Hsc70) at selected time-points (n=5/group at each time). In a second protocol, we determined (n=5/group) lung weight gain (edema), pulmonary compliance, oxygenation capacity, pulmonary artery pressure (PAP) and vascular resistance (PVR), the expression of PECAM-1 (CD31) and phosphorylation status of Src-kinase and VE-cadherin in lung tissue, as well as the release in perfusate of cytokines (TNFα, IL-1β) and endothelial biomarkers (sPECAM, von Willebrand Factor -vWF-, sE-selectin and sICAM-1). Histological and immunofluorescent studies assessed perivascular edema and formation of 3-nitrotyrosine (a marker of peroxinitrite) in CD31 lung endothelium.

Results

HS induced an early (3h) and persisting expression of HSP70 and HSP27, without influencing Hsc70. Lungs from the EVLP37°C group developed massive edema, low compliance and oxygenation, elevated PAP and PVR, substantial release of TNFα, IL-1β, s-PECAM, vWF, E-selectin and s-ICAM, as well as significant Src-kinase activation, VE-cadherin phosphorylation, endothelial 3-NT formation and reduced CD31 expression. In marked contrast, all these alterations were either abrogated or significantly attenuated by HS treatment.

Conclusion

The therapeutic application of a transient heat stress during EVLP of damaged rat lungs reduces endothelial permeability, attenuates pulmonary vasoconstriction, prevents src-kinase activation and VE-cadherin phosphorylation, while reducing endothelial peroxinitrite generation and the release of cytokines and endothelial biomarkers. Collectively, these data demonstrate that therapeutic heat stress may represent a promising strategy to protect the lung endothelium from severe reperfusion injury.

The role of lung microbiota in primary graft dysfunction in lung transplant recipients

BACKGROUND

Recent studies have shown that the lung microbiota is altered in critically ill patients and predicts clinical outcomes. Primary graft dysfunction (PGD) is a common complication and a leading cause of death within 1 month of lung transplantation, but the clinical significance of changes in the lung bacterial community during PGD is unclear. The aim of this study was to determine the contribution of the lung microbiota to the development and course of severe PGD.

METHODS

We conducted a retrospective study to characterize the lung microbiota of 32 lung transplant patients with combined PGD using next-generation sequencing of bronchoalveolar lavage samples. The relationship between lung flora dysbiosis and lung immunity in PGD was assessed by quantification of alveolar cytokines. The contribution of microbiota characteristics to patient outcomes was assessed by estimating overall survival.

RESULTS

Patients diagnosed with PGD grade 3 showed a reduction in alpha diversity, driven by a significant increase in the abundance of the genera Modestobacter, Scardovia and Selenomonas, and a reduction in the proportion of the genera Klebsiella and Oribacterium. Alpha diversity of the lung microbiota in PGD3 patients was negatively correlated with BALF interleukin (IL)-2 (r = -.752, p < .05). In addition, bacterial diversity in the lung microbiota of non-survivors was lower than that of survivors (p = .041).

CONCLUSIONS

There is variation in the lung microbiota of PGD grade 3 patients and dysbiosis of the lung microbiota is associated with lung immunity. The lung microbiota has potential in the diagnosis and treatment of PGD grade 3.