Developing and validating machine learning-based prediction models for frailty occurrence in those with chronic obstructive pulmonary disease.

Background: Frailty is a medical syndrome caused by multiple factors, characterized by decreased strength, endurance, and diminished physiological function, resulting in increased susceptibility to dependence and/or death. Patients with chronic obstructive pulmonary disease (COPD) tend to be more vulnerable to frailty due to their physical and psychological burdens. Therefore, the aim of this study was to develop a reliable and accurate vulnerability risk prediction model for frailty in patients with COPD in order to improve the identification and prediction of patient frailty. The specific objectives of this study were to determine the prevalence of frailty in patients with COPD and develop a prediction model and evaluate its predictive power. Methods: Clinical information was analyzed using data from the 2018 China Health and Retirement Longitudinal Study (CHARLS) database, and 34 indicators, including behavioral factors, health status, mental health parameters, and various sociodemographic variables, were examined in the study. The adaptive synthetic sampling technique was used for unbalanced data. Three methods, ridge regressor, extreme gradient boosting (XGBoost) classifier, and random forest (RF) regressor, were used to filter predictors. Seven machine learning (ML) techniques including logistic regression (LR), support vector machines (SVM), multilayer perceptron, light gradient-boosting machine, XGBoost, RF, and K-nearest neighbors were used to analyze and determine the optimal model. For customized risk assessment, an online predictive risk modeling website was created, along with Shapley additive explanation (SHAP) interpretations. Results: Depression, smoking, gender, social activities, dyslipidemia, asthma, and residence type (urban vs. rural) were predictors for the development of frailty in patients with COPD. In the test set, the XGBoost model had an area under the curve of 0.942 (95% confidence interval: 0.925-0.959), an accuracy of 0.915, a sensitivity of 0.873, and a specificity of 0.911, indicating that it was the best model. Conclusions: The ML predictive model developed in this study is a useful and easy-to-use instrument for assessing the vulnerability risk of patients with COPD and may aid clinical physicians in screening high-risk patients.

Chest wall resection with robotic-assisted thoracoscopic surgery for a Pancoast tumour: a case report.

We describe a rare procedure involving near-total robotic-assisted thoracoscopic surgery resection of a right posterior Pancoast tumour. Four ports and an assistant port were used. The DaVinci X system was used. The lobectomy was performed first to allow for adequate exposure to the apex and spine. The lateral aspect of ribs 1 to 4 was resected next, and the extrathoracic space was entered. Dissection proceeded through this space superiorly up to the level of the scapula and then posteriorly towards the spine. The second to the fifth ribs were dissected off the chest wall and resected medially off the spine at the rib heads. Further postero-superior exploration revealed the tumour to be invading the transverse process of the second rib, with ill-defined margins. Because of this development, and with the support of the spinal surgeons, a small high posterior thoracotomy was performed to complete the procedure and remove the specimen en bloc. The postoperative recovery was uneventful, and the patient was discharged on post-operative day 5. The final histological report confirmed a squamous non-small-cell lung cancer (pT3N0M0) with negative margins (R0). Asymptomatic recurrence was noted near the margin of the second rib resection posteriorly 1 year postoperatively and was successfully treated with radiotherapy.

Right robotic S4 segmentectomy.

Among the various segmentectomy procedures performed for lung diseases, the right S4 segmentectomy stands out as an exceptionally rare operation. Its infrequent occurrence can be attributed to several factors, including the relatively lower incidence of pathologies necessitating isolated resection of the S4 segment of the right lung. This video tutorial showcases the robotic approach to the rare right S4 segmentectomy, guiding surgeons through the step-by-step procedures for better understanding and proficiency.

Robotic resection of anterior mediastinal masses >10 cm: a case series.

Background: Robot-assisted thoracic surgery (RATS) for intrathoracic pathology and especially for mediastinal mass resection has been increasingly accepted as an alternative method to open sternotomy and video-assisted thoracic surgery (VATS). However, the utilization of this approach for complex and advanced in size cases needs more clinical evidence. We are presenting a series of 4 patients who had resection of >10 cm mediastinal masses via RATS. Cases Description: The mean age was 76.25±10.3 years and 3 were males (75%). All masses were positron emission tomography (PET) positive, and 1 patient had positive Acetyl-cholinesterase antibodies and myasthenia gravis (MG). All patients underwent RATS resection via DaVinci® X system. The dissections were conducted with spatula and/or Maryland bipolar forceps. In 2 cases, the resection was done with bilateral docking, and in 1 case, a drain was not inserted at the end. In 1 patient, pericardial resection was necessitated. All masses were thymomas with 1 dimension measured >10 cm on pathology. All patients were discharged on day 1 or 2 postoperatively with uneventful recoveries. There was no in-hospital, 30- or 90-day mortality. All patients were found to be without issues on follow-up. Conclusions: This report shows that RATS is safe and can be offered in the management of >10 cm anterior mediastinal masses. The previous size limit of the tumor for minimally invasive and especially RATS approach of 5 cm should be challenged.

Robotic reconstruction of a left hemi-diaphragm defect with double mesh in a patient with trisomy 21.

Robotic reconstructions of large diaphragmatic defects with mesh reconstructions are rare in the literature. We present a case of a complicated diaphragmatic defect, in an adult with trisomy 21, which was successfully repaired robotically with double mesh reinforcement. The meshes were sutured together via a separate suture in the middle to avoid fluid accumulation between them. The patient recovered quickly and uneventfully. On follow-up, he reported no pain, and his performance score improved dramatically. We present this complicated reconstruction in this specific patient, who we think benefitted from avoiding a thoraco-abdominal incision, demonstrating the merits of persevering with a robotic approach.

Trans-fissure and trans-lobar coil insertion technique for small lesions positioned inside the fissure.

The objective of this study is to present a technique of localization of difficult-to-locate lesions inside the fissure by deploying a coil with one of its ends left inside the fissure, and also, to evaluate the safety and the efficiency of this localization technique. Ten (10) patients with lesions in the fissure were identified during discussion at the multidisciplinary team meeting (MDT) and were recruited in the study. Attending surgeons verified that the lesions were unable to be located with minimally invasive approaches. These patients underwent localization of their lesion/s inside the fissure with a coil which was inserted through an adjacent lobe, through the fissure and inside the lesion with the other end protruding inside the fissure. All patients were next subjected to resection of their lesions with minimally invasive approaches. All patients had their lesions resected (with wedge resection or segmentectomies) with adequate margins (R0 resection). No conversion to thoracotomy was necessitated to attempt to palpate the lesion. Small, localized pneumothorax was noted in 7 patients; whereas, all patients had small laceration of their lung parenchyma along the course of the needle without, however, any action needed for these sequalae. The presented technique offers precise localization which leads to successful and safe resection of difficult-to-locate lesions inside the fissure via minimally invasive approaches. The clinical implications of this technique are numerous and its utilization can augment the successful performance of minimally invasive techniques for lesions situated inside the fissure.

Impact of the type of initial intervention on the outcomes of lung volume reduction for emphysema: a propensity matched study.

Background: The type of initial intervention i.e., endobronchial valve (EBV) implantation or lung volume reduction surgery (LVRS) to be offered as initial intervention remains vague in the treatment of emphysema-chronic obstructive pulmonary disease (COPD) patients. Aim of the present study was to compare the outcomes of EBV with that of LVRS in emphysema patients who could have both offered as an initial intervention. Methods: The outcomes of 44 EBV patients were retrospectively compared to the outcomes of 44 matched LVRS patients (matched for age, gender, performance status, body mass index (BMI), lung functions, comorbidities and exercise tolerance, matching tolerance 0.2) treated in a single institute within a 5-year period. The median follow-up was 32 months (maximum duration 84 months). Results: Mean age was 61.91±9.48 years and 55 (62.5%) were male. Postoperative morbidity was similar but length of stay (LOS) was longer in the LVRS group (median 10 vs. 6 days, P=0.006). Re-interventions were more frequent in the EBV versus LVRS group (52.3% vs. 20.5%, P=0.002) and so was the overall number of re-interventions (median 2 vs. 1, P<0.01). Breathing improved in more LVRS patients (86.4% vs. 70.5%, P<0.002). The decrease of the COPD Assessment Test (CAT) score was less significant in the EBV group (P=0.034). Survival was similar between 2 groups (P=0.350). Conclusions: EBV or LVRS as initial intervention are similar in terms of morbidity and mortality. EBV showed shorter LOS whilst LVRS necessitated less but more severe re-interventions and led to better overall quality of life.

Continuous endobronchial suctioning for refractory post lobectomy lung atelectasis: a case report.

The presentation of post lung resection atelectasis can vary between simple atelectasis and total lung collapse i.e., "white - out", making its treatment demanding in many occasions. We herein present the technique of continuous suctioning of the right upper lobe (RUL) by positioning a suction catheter inside the right upper lobe bronchus (RULB) through a tracheostomy in a sedated patient. This technique was used in the case of a 70-year-old patient who underwent a complicated redo thoracotomy and right lower lobectomy for lung cancer after a previous middle lobectomy via double thoracotomy for similar pathology. He had a significant ankylosis spondylitis past medical history with bamboo spine treated with long term high doses of steroids and methotrexate. Post redo surgery he developed respiratory failure with a radiologically significant RUL collapse, i.e., a "white-out", of the operated side which was refractory to usual conservative or bronchoscopic treatment. As a last resort, and in an effort to avoid high risk pneumonectomy, the patient was sedated, and a suction catheter was left inside the RULB under direct bronchoscopic guidance. This allowed the secretions inside the airways to be cleared, giving the remaining upper lobe infection time to subside, protected the stump from infective secretions and blind suctioning and led to avoidance of a high-risk pneumonectomy. The upper lobe cleared up from its collapse and patient's discharge from high dependency unit was achieved. This described maneuver can be useful in refractory cases of atelectasis when other measures have failed, in borderline patients or in patients where further surgery is technically cumbersome.

Procalcitonin Measurement in Pleural Fluid to Predict Infectious Complications of the Chest Post Lung Resection.

AIM: Procalcitonin (PCT) is variably used in clinical practice to identify infectious processes. This study investigated whether PCT level in pleural fluids could predict the infectious complications in the chests of patients undergoing lobectomy. PATIENTS AND METHODS: Thirty-four patients undergoing lobectomy for lung cancer were enrolled. PCT levels were measured in serum (S-PCT) and pleural effusion (PF-PCT) on consecutive postoperative days (PODs). The patients were grouped according to the development of chest infectious complications (atelectasis/pneumonia, postoperative infected pleural effusion/empyema/infected space, prolonged air leak >5 days with evidence of infection, lung torsion, and lung infarction). Multivariate analysis was performed to identify if S-PCT or PF-PCT and on which PODs were predictive of chest infectious complications. Receiver operating characteristic (ROC) analysis was further performed to identify cutoff values. RESULTS: Eleven patients experienced infectious complications within a median of 4 days (range 3-5 days) postoperatively. S-PCT and PF-PCT in non-complicated patients did not significantly increase postoperatively and followed a decreasing course. Only PF-PCT was significantly increased in complicated patients from POD1; the level peaked on POD4, as did that of S-PCT. PF-PCT on POD2 and 3 and S-PCT on POD3 independently predicted chest infectious complications. ROC analysis showed that PF-PCT > 0.88 ng/dL on POD2 was the most sensitive predictor of such complications (area under the ROC curve [AUC]: 0.979, sensitivity 85%/specificity 91%, p < .001) compared to S-PCT POD3 and PF-PCT POD3. CONCLUSION: Compared to PCT concentrations in serum, those in pleural fluids were more sensitive and predicted chest infectious complications earlier in patients undergoing lobectomy.

Tracheal laceration causing important post-intubation delayed subcutaneous emphysema and ventilatory deterioration in a COVID-19 patient with severe rheumatoid arthritis: a case report.

A 68-year-old man with a background of severe active rheumatoid arthritis (RA) was admitted to Intensive Care Unit (ICU) for respiratory support due to COVID-19 infection. Two days after an elective and uneventful intubation he developed severe and worsening surgical emphysema affecting his face, neck and both upper limbs. Ventilation was difficult to be achieved. Based on a negative chest X-ray, a CT scan of the chest was organized which showed extensive pneumomediastinum with no obvious cause. Therefore, urgent bronchoscopy was performed which showed a glassy lesion/laceration measuring 2 cm × 2 cm at the level of mid-trachea but no other signs of penetration through the airways were noted. Since events appeared 2 days after intubation, this was perceived as secondary to trauma during intubation on an inflammatory process background from RA and COVID-19 in the airways. The endotracheal tube was progressed beyond the site of laceration and bilateral pectoral fasciotomies were performed with negative suction vacuum dressings, which was successful in decreasing the surgical emphysema and achieving decreased ventilation requirements. Despite multi-organ support the patient continued to deteriorate and unfortunately passed away a week following admission. This scenario hightlighted that endotracheal sequalae should be suspected in patients with similar background and presentation.

Preoperative training education with incentive spirometry may reduce postoperative pulmonary complications.

OBJECTIVE: To assess whether preoperative incentive spirometer training would influence the development of postoperative pulmonary complications after lung resection. METHODS: Sixty-two lung resection patients were prospectively investigated; 17 were given an incentive spirometer preoperatively and 45 did not have an incentive spirometer preoperatively. Postoperatively, both arms exercised with an incentive spirometer. The number of repetitions per day, balls raised per repetition, correct technique of exercising, and postoperative pulmonary complications were compared between the 2 groups. Univariate binary logistic regression analysis of potential predictors of postoperative pulmonary complications led to multivariate analysis of independent predictors. Receiver operating characteristic analysis established the cutoff points of predictors. RESULTS: The group with no preoperative incentive spirometer developed more postoperative pulmonary complications than the preoperative incentive spirometer group (24.4% vs. 5.9%, respectively, p = 0.045). The preoperative incentive spirometer arm achieved more repetitions per day, balls per repetition, and correct incentive spirometer technique (p = 0.002, p < 0.001, p = 0.034, respectively). Balls raised per repetition and repetitions per day postoperatively were identified as independent predictors of postoperative pulmonary complications (p = 0.032 and p = 0.021, respectively). Less than 5 repetitions per day (sensitivity 93%, specificity 77%, p < 0.001) and less than 2 balls per repetition (sensitivity 93%, specificity 77%, p < 0.001) were predictive of postoperative pulmonary complications. CONCLUSION: Preoperative incentive spirometer exposure ensured better compliance with postoperative treatment and a more accurate technique (balls raised per repetition, repetitions per day). These variables correlated with a lower postoperative pulmonary complication rate.

Pleural Pressure Differences Before Removal Are Greater in Patients Who Develop Residual Pneumothorax Post Chest Drain Removal.

Aim of study: We aimed to investigate whether the difference in pleural pressures (ΔP) is wider among patients who develop a residual pneumothorax after chest tube removal following lung resection surgery. Materials and methods: Ninety-eight patients who underwent lung resection were included in the study over a period of 12 months. The ΔP prior to chest tube removal in patients who developed a residual pneumothorax after chest tube removal was compared with that of patients who did not develop this complication. The receiver operating characteristic (ROC) curve analysis was performed to identify cutoff values of ΔP for the prediction of residual pneumothorax. Logistic regression analysis was used to formulate a prediction model for the occurrence of residual pneumothorax based on ΔP. Results: Thirteen patients who developed a residual pneumothorax were compared with 85 patients without this complication. The ΔP in the residual pneumothorax group was significantly higher (10.8 versus 4.2 cm H2O, p < 0.01). The ΔP in patients who required intervention was also significantly higher (14.8 versus 4.2 cm H2O, p < 0.01). A ΔP cutoff value of 8 cm H2O was predictive of the occurrence of residual pneumothorax (sensitivity 85.6%, specificity 84.6%) and a value of 12 cm H2O was predictive of intervention (sensitivity 84%, specificity 85%). Increasing ΔP was an independent predictor of the occurrence of residual pneumothorax (p = 0.008) on the multivariate logistic regression model. Conclusion: Patients with wide ΔP before chest drain removal may be complicated with residual pneumothorax.

The Impact of a Newly Established Multidisciplinary Team on the Interventional Treatment of Patients With Emphysema.

BACKGROUND: The emphysema interventional treatment involves mainly lung volume reduction surgery (LVRS) and endobronchial valve (EBV) implantation. Few institutes discuss these cases at a dedicated emphysema multidisciplinary team (MDT) meeting. OBJECTIVES: To investigate the impact of a newly established dedicated emphysema MDT meeting on the interventional treatment of such patients. METHODS: During a study period of 4 years, the outcome of 44 patients who underwent intervention according to the proposal of the emphysema MDT (group A) was compared with the outcome of 44 propensity score matched patients (group B) treated without the emphysema MDT proposal. RESULTS: More LVRS and less EBV insertions were performed in group A (P = .009). In group B, the interventions were performed sooner than in group A (P = .003). Postoperative overall morbidity and length of in-hospital stay were similar in the 2 groups (P = .918 and .758, respectively). Improvement of breathing ability was reported in more patients from group A (P = .012). In group B, the total number of re-interventions was higher (P = .001) and the time to re-intervention had the tendency to be less (P = .069). Survival was similar between the 2 groups (P = .884). Intervention without discussion at the MDT and EBV as initial intervention was an independent predictor of re-intervention. CONCLUSIONS: Interventional treatment for patients with chronic obstructive pulmonary disease (COPD) after discussion at a dedicated MDT involved more LVRS performed, required fewer interventions for their disease, and had longer re-intervention-free intervals and better breathing improvement.

Ipsilateral pulmonary infarction after major lung resection.

Background Lung infarction is a rare complication of lung resection, developing mainly because of technical errors. In some cases, a specific reason cannot be identified. This study aimed to investigate the occurrence, characteristics, and outcome of this pathology in a series of patients. Methods The medical records of patients who underwent reoperation for lung infarction without an apparent cause (based on imaging, reoperation findings, and histopathology) after major lung resection at our institution from 2006 to 2015, were investigated. Results Seven patients were identified. The mean age was 62.2 years (range 51-75 years), and 5 were male. Copious dissection or adverse events during surgery were recorded in all but 2 cases. The main presenting symptom was unsettling frank hemoptysis (4 cases) with a variable time of onset of symptoms (4-164 h). All reoperations necessitated further lung resection (4 patients had a further lobectomy and 3 had a completion pneumonectomy). During reoperation, all vessels and bronchi were intact. No apparent cause of infarction could be identified according to the histopathology report. Morbidity after reoperation was atrial fibrillation in 3 cases and bronchopleural fistula in 2, one of which required a transsternal pneumonectomy and this was the only mortality. Length of stay ranged from 8 to 90 days. Conclusion Ipsilateral lung infarction after lobectomy is a rare complication and the reason may not be identifiable. Treatment usually requires reoperation. Extensive manipulation or adverse events during surgery could induce this rare complication.

Use of Bioglue™ to seal a difficult to treat air leak in a complicated empyema patient.

Surgical sealants have been used in thoracic surgery in an effort to reduce air leak duration, intercostal drain duration, length of stay and complications. They are instilled over a defect usually treated with other means. We herein present the technique of controlling a difficult to treat defect by directly instilling Bioglue™ alone into a crater, caused during empyema Video-assisted thoracic surgery in a 50-year-old patient with a trapped lung. This deep crater had caused a significant air leak rendering intraoperative ventilation challenging. After instillation, the dependent lung was kept blocked. With this technique, the Bioglue™ polymerized and the air leak was dramatically decreased making ventilation and eventually extubation of the patient feasible. Instillation of Bioglue™ directly into a large lung defect could be a choice of action to decrease complicated air leaks, otherwise impossible to treat with other means, in patients with trapped lung.

Do pleural adhesions influence the outcome of patients undergoing major lung resection?

OBJECTIVES: Our goal was to investigate whether pleural adhesions identified during an operation can induce adverse events. We investigated the outcome of major lung resection in patients with pleural adhesions encountered on entry into the pleural cavity. METHODS: We conducted a retrospective analysis of 144 patients undergoing major lung resection over a period of 9 months. Recorded data included demographics, comorbidities, surgical data, fluid volume drainage (on postoperative days [POD] 1 and 2 and in total), the overall and pleural space-associated morbidity (empyema, prolonged air leak or drainage, space issues), 30-day and late mortality rates. Patients were grouped according to the presence or not of adhesions observed when we entered the chest. RESULTS: Differences between patients without versus patients with adhesions were recorded for operative time (138 vs 169.3 min, P < 0.02), postoperative drainage on POD1 and POD2 (328.6 vs 478.5 ml, P < 0.01 and 214 vs 378 ml, P < 0.01 respectively), duration of air leak (1 vs 2 days, P = 0.03), duration of chest tube stay (2 vs 4 days, P < 0.01) and pleural morbidity (21.1% vs 38.8%, P = 0.02). There were no differences recorded in the 2 groups on conversion rates (2.5% vs 14.3%, P = 0.46), 30-day (1.1% vs 4.1%, P = .73) and late deaths (log-rank, P = 0.70). Pleural morbidity differed if the chest tube was removed on or earlier than POD2 (57.9% vs 36.9%, P = 0.02). We also calculated differences between those patients with adhesions involving the lower chest (55.1%) versus the rest of the group and specifically drainage on POD1 and POD2 (540.9 vs 372.1 ml, P < 0.01 and 392.5 vs 261 ml, P = 0.02, respectively) and pleural morbidity (46.4% vs 28.6%, P < 0.01). Logistic regression identified that firm, extensive adhesions, present in the lower third of the pleural cavity, are important predictors of pleural morbidity. CONCLUSIONS: Patients undergoing major lung resection who have pleural adhesions have an increased incidence of adverse surgical outcomes and higher pleural morbidity.

Performance of wider parenchymal lung resection than preoperatively planned in patients with low preoperative lung function performance undergoing video-assisted thoracic surgery major lung resection.

OBJECTIVES: Pulmonary assessment before major lung resections is used to determine patient's operability. In marginal cases, extensive pulmonary assessment is additionally important under the fear of a more radical parenchymal removal. This study investigates the outcome of wider lung parenchymal resections in patients with low lung functional status undergoing video-assisted thoracic surgery (VATS) major lung resection. METHODS: The medical records of patients who underwent VATS major lung resection for cancer, over a period of 5 years (August 2009-August 2014), were retrospectively reviewed. Patients with postoperative forced expiratory volume in first second (ppoFEV1) or postoperative diffusional capacity for carbon monoxide (ppoDLCO) <40% who underwent wider lung resection than preoperatively planned (Group A) were compared with patients with ppoFEV1 or ppoDLCO <40% who underwent the planned operation (Group B) and patients with ppoFEV1 and ppoDLCO >40% who underwent wider resection than preoperatively planned (Group C). Data analysed included demographics, past medical history, the surgery planned and performed, the reason for higher parenchymal resection, the clinical and pathological stage, the length of stay (LOS), the morbidity, the 30-day mortality and the survival. RESULTS: Overall, 73 patients were analysed (15 patients in Group A, 50 patients in Group B and 8 patients in Group C). The mean age was 68.5 years and 31.5% were males. The wider lung resection regarded 7 patients who underwent bilobectomy instead of lobectomy and 16 patients who underwent pneumonectomy instead of lobectomy. The main reason for higher resection was the wider invasion of the mass (21 patients). The age, gender and body mass index between three groups were similar, whereas ppoFEV1 and ppoDLCO were different (P < 0.001 and P < 0.001 respectively). Conversions, pulmonary morbidity and the 30-day mortality between groups were similar (P = 0.67, P = 0.88 and P = 0.33, respectively). LOS between groups was not different (P = 0.46). Survival rate between groups was also similar (log-rank, P = 0.79). CONCLUSIONS: Wider lung parenchymal resection than preoperatively anticipated may be performed, even in patients with low lung functional status, without increased adverse outcome when compared with patients with good lung function. This finding indicates that the preoperative risk stratification based on lung function tests is questionable.

Delayed presentation of aortoesophageal fistula after endovascular repair.

The incidence of aortoesophageal fistula after thoracic endovascular aortic repair has increased following an increase in thoracic endovascular aortic procedures. A 68-year-old man was diagnosed with aortoesophageal fistula 6 years after thoracic endovascular aortic repair of a descending aortic aneurysm. Due to massive hematemesis and instability, he underwent a left thoracotomy, graft repair of the aorta with stent removal, gastrostomy, and proximal esophagostomy. Unfortunately, he died 1 week later due to sepsis. Aortoesophageal fistula may present long after thoracic endovascular aortic repair, and clinicians treating such cases should always be aware of this complication.