A Machine Learning Approach for Postoperative Outcome Prediction: Surgical Data Science Application in a Thoracic Surgery Setting.

BACKGROUND: The use of innovative methodologies, such as Surgical Data Science (SDS), based on artificial intelligence (AI) could prove to be useful for extracting knowledge from clinical data overcoming limitations inherent in medical registries analysis. The aim of the study is to verify if the application of an AI analysis to our database could develop a model able to predict cardiopulmonary complications in patients submitted to lung resection. METHODS: We retrospectively analyzed data of patients submitted to lobectomy, bilobectomy, segmentectomy and pneumonectomy (January 2006-December 2018). Fifty preoperative characteristics were used for predicting the occurrence of cardiopulmonary complications. The prediction model was developed by training and testing a machine learning (ML) algorithm (XGBOOST) able to deal with registries characterized by missing data. We calculated the receiver operating characteristic curve, true positive rate (TPR), positive predictive value (PPV) and accuracy of the model. RESULTS: We analyzed 1360 patients (lobectomy: 80.7%, segmentectomy: 11.9%, bilobectomy 3.7%, pneumonectomy: 3.7%) and 23.3% of them experienced cardiopulmonary complications. XGBOOST algorithm generated a model able to predict complications with an area under the curve of 0.75, a TPR of 0.76, a PPV of 0.68. The model's accuracy was 0.70. The algorithm included all the variables in the model regardless of their completeness. CONCLUSIONS: Using SDS principles in thoracic surgery for the first time, we developed an ML model able to predict cardiopulmonary complications after lung resection based on 50 patient characteristics. The prediction was also possible even in the case of those patients for whom we had incomplete data. This model could improve the process of counseling and the perioperative management of lung resection candidates.

Redefining the Risk of Surgery for Clinical Stage IIIA (N2) Non-Small Cell Lung Cancer: A Pooled Analysis of the STS GTSD and ESTS Registry.

BACKGROUND: Management of clinical stage IIIA-N2 (cIIIA-N2) non-small cell lung cancer (NSCLC) remains controversial. We evaluated treatment strategies and outcomes in cIIIA-N2 NSCLC patients who underwent pulmonary resection in The Society of Thoracic Surgeons General Thoracic Surgery Database (STS GTSD) and the European Society of Thoracic Surgeons (ESTS) Registry. METHODS: The STS GTSD and ESTS Registry were queried for patients who underwent pulmonary resection for cIIIA-N2 NSCLC between 2012 and 2016. Demographic variables, treatment strategies, and outcome measures were collected and analyzed. Significance of differences was determined using the χ2 test for categorical variables and the Wilcoxon rank sum test for continuous variables. RESULTS: Pulmonary resection was performed in 4279 cIIIA-N2 NSCLC patients (2928 STS GTSD; 1351 ESTS). Induction therapy was administered to 49%. Lobectomy was performed in 67.1% and pneumonectomy in 13%. Lobectomy was associated with 19.2% major morbidity and 1.6% operative mortality, while pneumonectomy was associated with 34.1% and 5%, respectively. Induction therapy was associated with a higher rate of major morbidity or mortality than upfront surgery (23.2% vs 19.5%, p = 0.004), driven by pneumonectomy (40.7% vs 30.3%, p = 0.012) rather than lobectomy (20.3% vs 18.8%, p = 0.31). CONCLUSIONS: Pulmonary resection for cIIIA-N2 NSCLC is associated with low rates of operative morbidity and mortality, with lobectomy having lower morbidity and mortality than pneumonectomy. Induction therapy, particularly chemoradiotherapy, is associated with a higher rate of composite morbidity or mortality than upfront surgery in pneumonectomy patients but not lobectomy patients.

Computed tomography-guided microcoil placement for localizing small pulmonary nodules before uniportal video-assisted thoracoscopic resection.

PURPOSE: The increasing number of computed tomography (CT) performed allows the more frequent identification of small, solid pulmonary nodules or ground-glass opacities. Video-assisted thoracic surgery (VATS) represents the standard in most lung resections. However, since VATS limit is the digital palpation of the lung parenchyma, many techniques of nodule localization were developed. The aim of this study was to determine the feasibility and safety of CT-guided microcoil insertion followed by uniportal VATS wedge resection (WR). MATERIALS AND METHODS: Retrospective study in a single institution, including patients undergone CT-guided microcoil insertion prior to uniportal VATS resection between May 2015 and December 2018. The lesion was identified using fluoroscopy. RESULTS: Forty-six consecutive patients were enrolled (22 male and 24 female). On CT: 5 cases of GGO, 2 cases of semisolid nodules, 39 cases of solid nodules. The median pathologic tumor size was 1.21 cm. Neither conversion to thoracotomy nor microcoil dislodgement was recorded. All patients underwent uniportal VATS WR (9/46 underwent completion lobectomy after frozen section). WR median time was 105 min (range 50-150 min). No patients required intraoperative re-resection for positive margins. After radiological procedure, 1 case of hematoma and 2 cases of pneumothorax were recorded. Four complications occurred in the postoperative period. The mean duration of chest drain and length of stay were 2.9 and 4.6 days, respectively. CONCLUSIONS: CT-guided microcoil insertion followed by uniportal VATS resection was a safe and feasible procedure having a minimal associated complications rate and offering surgeons the ease of localization of small intrapulmonary nodules.

How the Pathological Microenvironment Affects the Behavior of Mesenchymal Stem Cells in the Idiopathic Pulmonary Fibrosis.

Idiopathic pulmonary fibrosis (IPF) is a chronic disease characterized by fibroblasts activation, ECM accumulation, and diffused alveolar inflammation. The role of inflammation in IPF is still controversial and its involvement may follow nontraditional mechanisms. It is seen that a pathological microenvironment may affect cells, in particular mesenchymal stem cells (MSCs) that may be able to sustain the inflamed microenvironment and influence the surrounding cells. Here MSCs have been isolated from fibrotic (IPF-MSCs) and control (C-MSCs) lung tissue; first cells were characterized and compared by the expression of molecules related to ECM, inflammation, and other interdependent pathways such as hypoxia and oxidative stress. Subsequently, MSCs were co-cultured between them and with NHLF to test the effects of the cellular crosstalk. Results showed that pathological microenvironment modified the features of MSCs: IPF-MSCs, compared to C-MSCs, express higher level of molecules related to ECM, inflammation, oxidative stress, and hypoxia; notably, when co-cultured with C-MSCs and NHLF, IPF-MSCs are able to induce a pathological phenotype on the surrounding cell types. In conclusion, in IPF the pathological microenvironment affects MSCs that in turn can modulate the behavior of other cell types favoring the progression of IPF.

Endovascular Treatment of Blunt Thoracic Aortic Injury by Fractured Rib.

We treated an 89-year-old patient affected by a descending thoracic aorta lesion due to a rib fracture with a penetrating costal stump. An urgent combined thoracic and endovascular surgical approach was performed, removing the rib fragment and positioning an aortic endoprosthesis simultaneously. Postoperative angio-computed tomography scan demonstrated the correct position of the endoprosthesis without any leakage or periaortic hemorrhage.

Uniportal Video-Assisted Thoracoscopic Anatomic Lung Resection after Neoadjuvant Chemotherapy for Lung Cancer: A Case-Matched Analysis.

BACKGROUND: The advantages of video-assisted thoracic surgery (VATS) are well-recognized in several studies. However, in the cases of advanced lung cancer after neoadjuvant chemotherapy (nCT), the role of VATS is still questionable, with concerns about safety, technical feasibility, and oncological completeness. The aim of this study was to assess the impact of nCT on patients who had undergone uniportal VATS (U-VATS) anatomic lung resections for lung cancer, by comparing the short-term outcomes of patients after nCT with case-matched counterparts (treated by surgery alone). METHODS: We performed a retrospective, comparative study enrolling 927 patients (nCT: 60; non-nCT:867) who underwent U-VATS anatomic lung resections from 2014 to 2020 in two centers. Data were collected in a shared database with standardized variables' definition. Propensity score matching using 15 baseline preoperative patients' characteristics was performed in order to minimize selection-confounding factors between the two groups, which then were directly compared in terms of perioperative outcomes. RESULTS: After propensity score matching, two groups of 60 patients had been defined. The nCT-group had a higher conversion rate compared to the control group (13.3% vs. 0%, p = 0.003) without an increase in operation time or cardiopulmonary complications. In addition, no differences between the two groups were recorded in terms of prolonged air leaks, length of stay, and readmission. CONCLUSIONS: U-VATS after nCT is a feasible approach, showing a similar rate of cardiopulmonary complications and length of stay when compared with the control group. However, it remains a challenging surgery due to its great technical complexity as well as the clinical status of the patients.

The ESTS-AIR database-initial results of a multi-institutional database on airway surgery.

OBJECTIVES: Compared to lung resections, airway procedures are relatively rare in thoracic surgery. Despite this, a growing number of dedicated airway centres have formed throughout Europe. These centres are characterized by a close interdisciplinary collaboration and they often act as supra-regional referring centres. To date, most evidence of airway surgery comes from retrospective, single-centre analysis as there is a lack of large-scale, multi-institutional databases. METHODS: In 2018, an initiative was formed, which aimed to create an airway database within the framework of the ESTS database (ESTS-AIR). Five dedicated airway centres were asked to test the database in a pilot phase. A 1st descriptive analysis of ESTS-AIR was performed. RESULTS: A total of 415 cases were included in the analysis. For adults, the most common indication for airway surgery was post-tracheostomy stenosis and idiopathic subglottic stenosis; in children, most resections/reconstructions had to be performed for post-intubation stenosis. Malignant indications required significantly longer resections [36.0 (21.4-50.6) mm] when compared to benign indications [26.6 (9.4-43.8) mm]. Length of hospital stay was 11.0 (4.1-17.3) days (adults) and 13.4 (7.6-19.6) days (children). Overall, the rates of complications were low with wound infections being reported as the most common morbidity. CONCLUSIONS: This evaluation of the 1st cases in the ESTS-AIR database allowed a large-scale analysis of the practice of airway surgery in dedicated European airway centres. It provides proof for the functionality of ESTS-AIR and sets the basis for rolling out the AIR subsection to all centres participating in the ESTS database.

Uniportal Video-Assisted Thoracoscopic Surgery Completion Lobectomy Long after Wedge Resection or Segmentectomy in the Same Lobe: A Bicenter Study.

BACKGROUND: Completion lobectomy (CL) following a prior resection in the same lobe may be complicated by severe pleural or hilar adhesions. The role of uniportal video-assisted thoracoscopic surgery (U-VATS) has never been evaluated in this setting. METHODS: Data were collected from two Italian centers. Between 2015 and 2022, 122 patients (60 men and 62 women, median age 67.7 ± 8.913) underwent U-VATS CL at least 4 weeks after previous lung surgery. RESULTS: Twenty-eight (22.9%) patients were affected by chronic obstructive pulmonary disease (COPD) and twenty-five (20.4%) were active smokers. Among the cohort, the initial surgery was performed using U-VATS in 103 (84.4%) patients, triportal-VATS in 8 (6.6%), and thoracotomy in 11 (9.0%). Anatomical segmentectomy was the initial surgery in 46 (37.7%) patients, while hilar lymphadenectomy was performed in 16 (13.1%) cases. CL was performed on 110 (90.2%) patients, segmentectomy on 10 (8.2%), and completion pneumonectomy on 2 (1.6%). Upon reoperation, moderate pleural adhesions were observed in 38 (31.1%) patients, with 2 (1.6%) exhibiting strong adhesions. Moderate hilar adhesions were found in 18 (14.8%) patients and strong adhesions in 11 (9.0%). The median operative time was 203.93 ± 74.4 min. In four (3.3%) patients, PA taping was performed. One patient experienced intraoperative bleeding that did not require conversion to thoracotomy. Conversion to thoracotomy was necessary in three (2.5%) patients. The median postoperative drainage stay and postoperative hospital stay were 5.67 ± 4.44 and 5.52 ± 2.66 days, respectively. Postoperative complications occurred in 34 (27.9%) patients. Thirty-day mortality was null. Histology was the only factor found to negatively influence intraoperative outcomes (p = 0.000). Factors identified as negatively impacting postoperative outcomes at univariate analyses were male sex (p = 0.003), age > 60 years (p = 0.003), COPD (p = 0.014), previous thoracotomy (p = 0.000), previous S2 segmentectomy (p = 0.001), previous S8 segmentectomy (p = 0.008), and interval between operations > 5 weeks (p= 0.005). In multivariate analysis, only COPD confirmed its role as an independent risk factor for postoperative complications (HR: 5.12, 95% CI (1.07-24.50), p = 0.04). CONCLUSIONS: U-VATS CL seems feasible and safe after wedge resection and anatomical segmentectomy.

Preoperative assessment of patients for lung cancer surgery.

PURPOSE OF REVIEW: The aim of this work was to present and analyze the latest published documents about the functional evaluation of patients undergoing lung resection and review articles from the past two years addressing the same topic. RECENT FINDINGS: In 2009 and 2010, two important task forces, appointed by international scientific societies, have published documents to guide the preoperative evaluation and risk stratification of lung resection candidates. In both documents, cardiac evaluation is prioritized. Detailed cardiologic guidelines have been proposed. After this first step, functional assessment should include a spirometric assessment, asystematic measurement of carbon monoxide diffusion capacity and a cardiopulmonary exercise test evaluation. Differences in the relative importance of these tests in the two guidelines were discussed. Most recent evidences focused on the role of cardiopulmonary exercise test and the use of several direct and indirect ergometric parameters that may refine risk assessment. SUMMARY: The use of evidence-based clinical guidelines on preoperative evaluation is recommendable. Nevertheless, scientific evidence is still suboptimal in this field. Aggregate analyses on larger series are needed to improve risk stratification.

Impact of enhanced pathway of care in uniportal video-assisted thoracoscopic surgery.

Enhanced Recovery After Surgery (E.R.A.S.) is a multimodal, evidence-based and patient-centered pathway designed to minimize surgical stress, enhancing recovery and improving perioperative outcomes. However, considering that the potential clinical implication of E.R.A.S. on patients undergoing video-assisted thoracic surgery (V.A.T.S.) has not properly defined, we proposed to implement our minimally invasive program with a specific clinical pathway able to enhance recovery after lung resection. Aim of this study was to assess the impact of this integrated program of Enhanced Pathway of Care (E.P.C.) in Uniportal V.A.T.S. patients undergoing lung resection, in terms of efficiency and safety. We conducted a retrospective, observational study enrolling patients undergoing uniportal V.A.T.S. resections from January 2015 to May 2020. Two groups were created: pre-E.P.C. and E.P.C. Propensity score matching analysis was performed to evaluate length of stay (LOS), postoperative cardiopulmonary complications (CPC) and readmission rate (READM). We analyzed 1167 patients (E.P.C. group: 182; pre-E.P.C. group: 985). E.P.C. group has a mean LOS shorter compared to pre-E.P.C. group (3.13 vs 4.19 days, p < 0.0001) without increasing on CPC (E.P.C. 12% vs pre-E.P.C. 11%, p = 0.74) and READM rate (E.P.C. 1.6% vs pre-E.P.C. 4.9%, p = 0.07). In particular, the LOS was shortened in the E.P.C. patients submitted to lobectomy, segmentectomy and wedge resection. Moreover, the three subgroups had similar CPC and READM rates for E.P.C. and control patients. In conclusion, this study demonstrated the benefits and safety of E.P.C. program showing a reduction of LOS for patients undergoing uniportal V.A.T.S. resection.