Advertisement

Comparative Effectiveness of Surgical Approaches for Lung Cancer

Published:December 09, 2020DOI:https://doi.org/10.1016/j.jss.2020.10.020

      Abstract

      Background

      The magnitude of association and quality of evidence comparing surgical approaches for lung cancer resection has not been analyzed. This has resulted in conflicting information regarding the relative superiority of the different approaches and disparate opinions on the optimal surgical treatment. We reviewed and systematically analyzed all published data comparing near- (30-d) and long-term mortality for minimally invasive to open surgical approaches for lung cancer.

      Methods

      Comprehensive search of EMBASE, MEDLINE, and the Cochrane Library, from January 2009 to August 2019, was performed to identify the studies and those that passed bias assessment were included in the analysis utilizing propensity score matching techniques. Meta-analysis was performed using random-effects and fixed-effects models. Risk of bias was assessed via the Newcastle–Ottawa Scale and the ROBINS-I tool. The study was registered in PROSPERO (CRD42020150923) prior to analysis.

      Results

      Overall, 1382 publications were identified but 19 studies were included encompassing 47,054 patients after matching. Minimally invasive techniques were found to be superior with respect to near-term mortality in early and advanced-stage lung cancer (risk ratio 0.45, 95% confidence interval [CI] 0.21-0.95, I2 = 0%) as well as for elderly patients (odds ratio 0.45, 95% CI 0.31-0.65, I2 = 30%), but did not demonstrate benefit for high-risk patients (odds ratio 0.74, 95% CI 0.06-8.73, I2 = 78%). However, no difference was found in long-term survival.

      Conclusions

      We performed the first systematic review and meta-analysis to compare surgical approaches for lung cancer which indicated that minimally invasive techniques may be superior to thoracotomy in near-term mortality, but there is no difference in long-term outcomes.

      Keywords

      Introduction

      Lung cancer is the leading cause of cancer-related death among men and women. Each year, more patients die from lung cancer than from colon, breast, and prostate cancers combined.
      American Cancer Society
      Cancer Facts & Figures 2019.
      For early-stage non-small cell lung cancer (NSCLC), including Stage I and II, surgical resection remains the primary treatment modality, and for more advanced stages, surgery is an essential aspect of management.
      • Lackey A.
      • Donington J.
      Surgical management of lung cancer.
      Surgical resection of lung cancer can be accomplished using various approaches. Three techniques are most commonly used today: thoracotomy (also known as the open approach), video-assisted thoracoscopic surgery (VATS), and robot-assisted thoracoscopic surgery (RATS), each with its own variations. Many studies have been conducted comparing these approaches, and while there is consensus that minimally invasive techniques may be preferred for resection of Stage I lung cancer,
      • Boffa D.J.
      • Kosinski A.S.
      • Furnary A.P.
      • et al.
      Minimally invasive lung cancer surgery performed by thoracic surgeons as effective as thoracotomy.
      • Yang C.J.
      • Kumar A.
      • Klapper J.A.
      • et al.
      A national analysis of long-term survival following thoracoscopic versus open lobectomy for stage I non-small-cell lung cancer.
      • Gossot D.
      Comparing open and closed chest surgery for early-stage lung cancer: still relevant?.
      • Bhagat R.
      • Bronsert M.R.
      • Henderson W.G.
      • et al.
      Analysis of discharge destination after open vs. minimally invasive surgery for lung cancer.
      • Detterbeck F.
      • Molins L.
      Video-assisted thoracic surgery and open chest surgery in lung cancer treatment: present and future.
      • Bendixen M.
      • Kronborg C.
      • Jorgensen O.D.
      • Andersen C.
      • Licht P.B.
      Cost-utility analysis of minimally invasive surgery for lung cancer: a randomized controlled trial.
      • Long H.
      • Tan Q.
      • Luo Q.
      • et al.
      Thoracoscopic surgery versus thoracotomy for lung cancer: short-term outcomes of a randomized trial.
      • Liang W.
      Comparison of lobectomy, segmentectomy and wedge resection for early stage NSCLC: a direct and network meta-analysis.
      • Yang H.X.
      • Woo K.M.
      • Sima C.S.
      • et al.
      Long-term survival based on the surgical approach to lobectomy for clinical stage I nonsmall cell lung cancer: comparison of robotic, video-assisted thoracic surgery, and thoracotomy lobectomy.
      limited evidence exists to guide the selection of surgical approach.
      • Rajaram R.
      • Mohanty S.
      • Bentrem D.J.
      • et al.
      Nationwide assessment of robotic lobectomy for non-small cell lung cancer.
      It is imprudent to extend the data regarding surgical approaches from early stage to more advanced lung cancers as the latter represents a different and heterogenous treatment group, frequently with greater technical challenges and risks of surgical complications. In addition, variations may exist in the efficacy of these techniques for different subgroups such as the elderly or high-risk patients. Furthermore, within each subgroup the different techniques might have different efficacy profiles, leading to complex decisions and tradeoffs.
      As a result of the nature and constant evolution of surgery
      • McCulloch P.
      • Feinberg J.
      • Philippou Y.
      • et al.
      Progress in clinical research in surgery and IDEAL.
      • Hirst A.
      • Philippou Y.
      • Blazeby J.
      • et al.
      No surgical innovation without evaluation: evolution and further development of the IDEAL framework and recommendations.
      • Vannucci F.
      • Gonzalez-Rivas D.
      Is VATS lobectomy standard of care for operable non-small cell lung cancer?.
      combined with the complexity of performing high-quality randomized clinical trials in surgical research, retrospective studies have dominated the literature on surgical technique. This leads to a high prevalence of bias and reduces the generalizability of findings. From the perspective of a systematic review, a very careful assessment of internal and external validity needs to be performed to offset these inherent biases.
      • Rothman K.J.
      • Greenland S.
      • Lash T.L.
      Modern Epidemiology.
      We conducted a systematic review and meta-analysis with three principal aims: (1) identify the current clinical evidence comparing mortality across different surgical techniques for resection of lung cancer; (2) determine the effects of surgical approach on specific subgroups of patients that are under-represented in the published studies; and (3) understand the prevalence of bias in this field and its effect on generalizability. Given the practical limitations precluding true randomized trials to assess surgical approach, we believe that a carefully conducted systematic review and meta-analysis offers valuable information regarding these important clinical questions.
      The focus of this paper is on survival. Quality of life is explored in a smaller set of papers
      • Bendixen M.
      • Jørgensen O.D.
      • Kronborg C.
      • Andersen C.
      • Licht P.B.
      Postoperative pain and quality of life after lobectomy via video-assisted thoracoscopic surgery or anterolateral thoracotomy for early stage lung cancer: a randomised controlled trial.
      • Zhao J.
      • Zhao Y.
      • Qiu T.
      • et al.
      Quality of life and survival after II stage nonsmall cell carcinoma surgery: video-assisted thoracic surgery versus thoracotomy lobectomy.
      • Rizk N.P.
      • Ghanie A.
      • Hsu M.
      • et al.
      A prospective trial comparing pain and quality of life measures after anatomic lung resection using thoracoscopy or thoracotomy.
      and with considerably increased heterogeneity. Costs, despite their obvious importance,
      • Bendixen M.
      • Kronborg C.
      • Jorgensen O.D.
      • Andersen C.
      • Licht P.B.
      Cost-utility analysis of minimally invasive surgery for lung cancer: a randomized controlled trial.
      ,
      • Rodgers-Fischl P.M.
      • Martin J.T.
      • Saha S.P.
      Video-assisted thoracoscopic versus open lobectomy: costs and outcomes.
      ,
      • Kneuertz P.J.
      • Singer E.
      • D'Souza D.M.
      • Abdel-Rasoul M.
      • Moffatt-Bruce S.D.
      • Merritt R.E.
      Hospital cost and clinical effectiveness of robotic-assisted versus video-assisted thoracoscopic and open lobectomy: a propensity score-weighted comparison.
      could not be evaluated given the heterogeneity in payment structures and contractual relationships, while perioperative outcomes have been extensively reviewed.
      • Detterbeck F.
      • Molins L.
      Video-assisted thoracic surgery and open chest surgery in lung cancer treatment: present and future.
      ,
      • Ng C.S.H.
      • MacDonald J.K.
      • Gilbert S.
      • et al.
      Optimal approach to lobectomy for non-small cell lung cancer: systemic review and meta-analysis.

      Material and methods

      Literature search

      The review was registered on the International Prospective Register of Systematic Reviews before the primary literature search (PROSPERO CRD42020150923). A comprehensive and systematic search of MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, and Scopus was performed for studies published after 2009. Databases of ongoing trials such as the WHO international trials registry, US National Institutes of Health Ongoing Trials Register, EU Clinical Trials Register, and the ISRCTN register of controlled trials were also searched for active clinical trials. The following keywords were used in several logical combinations: surgery, resection, lobectomy, segmentectomy, wedge resection, pneumonectomy, thoracic, lung, pulmonary, thoracoscopic, robotic surgery, robot assisted, da vinci, video-assisted thoracoscopic surgery, open thoracic, cancer. Bibliographies of the selected studies and articles citing selected studies were also reviewed to find additional relevant studies. Manual search of the issues where the selected papers were found and “suggestions” by editor algorithms were also considered in the search. The study was reported according to the PRISMA statement.
      • Liberati A.
      • Altman D.G.
      • Tetzlaff J.
      • et al.
      The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.

      Eligibility criteria

      For inclusion, studies had to discuss resection of patients’ lung cancer via thoracotomy, VATS, or RATS and compare mortality across at least two approaches. Only studies on human data published in or translated into English were included. Although no constraint on study design was set in the literature search, only evidence at least at the level of independent database or retrospective cohort study (both referred to as retrospective observational studies in the following) was considered for the review portion of this article. Retrospective observational studies were included only if sufficient care in adjusting for confounders was applied. In principle, randomized controlled trials (RCTs) would have been the best option, but none could be found that was relevant for the scope of this paper. Exclusion criteria included studies discussing only one surgical approach, studies published before 2009, and studies focusing on metastasectomies as opposed to primary lung cancer. Our rationale for setting 2009 as the cutoff was that technology, techniques, and indications for VATS and RATS have evolved significantly since their advent, and their utilization has increased from a small fraction to considerably more than half of all procedures since 2009.
      • Detterbeck F.
      • Molins L.
      Video-assisted thoracic surgery and open chest surgery in lung cancer treatment: present and future.
      We did not exclude single-center studies.

      Data extraction

      Three independent reviewers selected studies on the basis of the defined inclusion and exclusion criteria. Two reviewers initially reviewed all studies and were blinded to each other’s decisions via the use of the Rayyan tool.
      • Ouzzani M.
      • Hammady H.
      • Fedorowicz Z.
      • Elmagarmid A.
      Rayyan—a web and mobile app for systematic reviews.
      Once each reviewer had completed decisions, they were unblinded and discrepancies were reconciled through group discussion; a third reviewer served to arbitrate discrepancies. Studies were initially reviewed on the basis of title and abstract, and those deemed relevant were reviewed in full text to determine the final selection of studies to be used for bias assessment. In cases of likely study duplication, the more recent and complete study was selected.

      Risk of bias assessment

      Risk of bias was assessed for each article using the Newcastle–Ottawa Scale (NOS)
      • Wells BS G.
      • O’Connell D.
      • Peterson J.
      The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses 2000.
      and the ROBINS-I tool.
      • Sterne J.A.
      • Hernan M.A.
      • Reeves B.C.
      • et al.
      ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions.
      Studies deemed at low or moderate risk of bias were then included in the statistical analysis. The assessment of papers was then repeated in order to have a consistent evaluation. The authors collectively participated in meetings for discussing and arbitrating risk of bias assessments. Details are presented in the supplemental material.

      Statistical analysis

      Papers were stored in Endnote and data were manually extracted into Excel spreadsheets, except for when data had to be read off plots, where we utilized WebPlotDigitizer.
      WebPlotDigitizer [computer program]. Version 4.2.
      The statistical analysis was performed using R 3.6.0 (The R Foundation for Statistical Computing, http://www.R-project.org). The meta-analysis and its plots were done using the package “meta.”
      • Schwarzer G.
      meta: an R package for meta-analysis.
      The function “metagen” was utilized for hazard (HR) and odd ratios (OR) using the inverse variance method. We used HR for long-term survival and OR for near-term mortality. The function “metabin” was used for binary data such as near-term (30-d) mortality when reported as adjusted number of events (or the number of events could be reconstructed for all studies selected) and its outcome reported as risk ratios (RRs). For transparency and readability, we always included a random effect model estimate even when it had zero effect. The analysis was considered significant when P-values were smaller than 0.05; however, we attempted to qualify our statements depending on how much potential bias we believed remained in the selected studies. Lack of significance alone was not considered sufficient evidence for equivalence.

      Patient subgroups

      By its very nature, advanced lung cancer is a heterogeneous disease. Based on the literature, we defined advanced lung cancer as that requiring induction therapy, by stage (higher than IIA), or by size thresholds (>3 cm). However, studies may have been conducted in different countries, with slightly different regulations and protocols. Moreover, surgical procedures can be performed in a variety of ways and are constantly improved or modified likely even within each study. We chose to ignore most of these differences to see if general statements could be made. It is important to notice that the use of propensity scores essentially selects the patients from the open population that can be matched to the VATS or RATS populations, limiting the generalizability of the results to only those patients in whom a minimally invasive approach was selected and not the entire population of patients with advanced lung cancer. If procedure selection is at the discretion of the surgeon, the statements are also limited to surgeons who perform both approaches. High risk was defined as any patient with predicted postoperative forced expiratory volume in 1 s (FEV1%) <40%, diffusing capacity of the lung for carbon monoxide (DLCO%) <40%, or those meeting the requirement for high risk based on the American College for Surgery Oncology Group trial.
      • Fernando H.C.
      • Landreneau R.J.
      • Mandrekar S.J.
      • et al.
      The impact of adjuvant brachytherapy with sublobar resection on pulmonary function and dyspnea in high-risk patients with operable disease: preliminary results from the American College of Surgeons Oncology Group Z4032 Trial.

      Results

      Search results

      The literature search identified 1382 publications and meeting abstracts. After initial review of abstracts and titles, 282 were excluded because duplicates and 967 because they did not make the eligibility criteria (Supplemental Fig. 1). The remaining 132 studies were reviewed in full. This process was repeated again using the papers citing these studies and the bibliographies of these studies. Ultimately, a total of 38 studies met the inclusion criteria for areas including advanced lung cancer (14), elderly patients (6), high-risk patients (6), and robotic surgery (12). Of these, 5 studies on advanced lung cancer, 5 including elderly patients, 3 studies on high-risk patients, and 6 addressing early-stage lung cancer were sufficiently unbiased to be included in the quantitative analysis although not all of these 19 papers included every outcome of interest (near-term mortality, long-term mortality, and disease-free survival). Notably, only studies that conducted propensity score matching were deemed sufficiently unbiased for inclusion in the analysis.

      Study characteristics

      The 19 studies described 47,054 patients; open thoracotomy was the surgical approach in 22,958 patients, VATS in 16,448, and RATS in the remaining 7648. All included manuscripts were retrospective cohort studies. The number of patients in the included studies ranged from 56 to 12,339 demonstrating the wide variability in study approach and sample size (Supplemental Table 1).

      Study quality

      Scores from the NOS are listed in Supplemental Table 2, with risk assessments for each study. Following ROBINS-I, 15 of the 38 studies (39%) were deemed to have moderate risk of bias due to the presence of multivariable data or insufficient propensity score matching, 12 were considered at serious risk of bias (32%), and the other 11 studies (29%) were of critical risk of bias or not scored most often due to lack of adjustment for confounding variables or failure to use an intention-to-treat approach.

      Advanced stage cancer

      We did not find studies reporting outcomes of RATS in this disease population. The overall survival (OS) was not significantly different between patients who underwent VATS when compared with propensity score-matched patients who underwent open thoracotomy (HR of 1.00, 95% confidence interval [CI] 0.86-1.16, I2 = 0%). Disease-free survival was also not significantly different between patients who underwent VATS compared to propensity score-matched patients undergoing open thoracotomy (HR 0.82, 95% CI 0.63-1.08, I2 = 0%). However, the near-term (30 d) mortality rate appeared to be significantly lower for patients who underwent VATS compared to those undergoing thoracotomy (RR 0.45, 95% CI 0.21-0.95, I2 = 0%). These results are summarized in Figure 1. Median length of follow-up ranged from 24 mo to 75 mo.
      Figure thumbnail gr1
      Fig. 1Near- and long-term survival for VATS and Open surgical approach for advanced cancer. Forest plots of comparison show dark blue squares which represent single studies (area proportional to sample size). Light blue diamonds are 95% confidence intervals for the outcomes as estimated by the meta-analysis. Values smaller than 1 favor VATS. (Color version of figure is available online.)
      For OS, disease-free survival, and near-term mortality, a sensitivity analysis was conducted by including the studies with an NOS score of 5 (Supplemental Fig. 2). This score represented papers that were at a significant but not critical risk of bias. As expected, adding these more biased papers moved estimates closer to favoring VATS. This is likely because less advanced cancers tend to be resected using VATS and failing to use intention-to-treat removes the most challenging surgeries from VATS as well. However, inclusion of these studies did not affect conclusions for overall or near-term survival. Nevertheless, when examining disease-free survival with this more biased cohort of papers, a statistically significant improvement in survival for patients who underwent surgery via a VATS approach was found when compared to thoracotomy (HR 0.80, 95% CI 0.69-0.94, I2 = 0%). This illustrates how the bias in the surgical literature can alter conclusions inaccurately. These results can be seen in Supplemental Figure 2.

      Elderly

      Five studies were found comparing optimal surgical approach for cancer resection in elderly patients (see Supplemental Fig. 1). Only those studies directly comparing open versus minimally invasive surgical approaches in elderly patients were included. Of these, 4 were large database analyses, including the National Cancer Data Base, the Surveillance, Epidemiology, and End Results, the Dutch Lung Surgery Audit database, and the French National Administrative Database. Unfortunately, the selected papers contained a remarkable degree of heterogeneity. Detillon and Veen
      • Detillon D.D.
      • Veen E.J.
      Postoperative outcome after pulmonary surgery for non-small cell lung cancer in elderly patients.
      and Endoh et al.
      • Endoh H.
      • Yamamoto R.
      • Satoh Y.
      • Kuwano H.
      • Nishizawa N.
      Risk analysis of pulmonary resection for elderly patients with lung cancer.
      included patients ≥60 and ≥75 y old respectively, who underwent any type of lung resection for a variety of lung cancer stages. Ezer et al.
      • Ezer N.
      • Kale M.
      • Sigel K.
      • et al.
      Outcomes after video-assisted thoracoscopic lobectomy versus open lobectomy for early-stage lung cancer in older adults.
      and Smith et al.
      • Smith C.B.
      • Kale M.
      • Mhango G.
      • et al.
      Comparative outcomes of elderly stage I lung cancer patients treated with segmentectomy via video-assisted thoracoscopic surgery versus open resection.
      focused on patients ≥65 y old, with Stage I lung cancers treated with segmentectomies and Stage I and II with lobectomies. Finally, Pages et al.
      • Pages P.B.
      • Mariet A.S.
      • Madelaine L.
      • et al.
      Impact of video-assisted thoracic surgery approach on postoperative mortality after lobectomy in octogenarians.
      performed their study on patients ≥80 y old who underwent lobectomies, but no stage data were included. Despite this heterogeneity, which renders any outcome very tentative, we explored the possibility of finding overall trends or consistencies.
      Pooled OS and near-term survival were determined using two sets of studies (four studies for OS and three for near-term survival). Only ORs could be found consistently across papers because of the methodologies utilized to adjust for confounding and thus it was used in our analysis. For near-term survival, statistical significance was demonstrated in favor of VATS (random-effects OR 0.45, 95% CI 0.31-0.65, I2 = 30%), with a moderate and acceptable degree of heterogeneity even if this result remains uncertain because of the limitations described above. However, for OS, no significant difference among approaches was demonstrated (HR 0.94, 95% CI 0.83-1.05, I2 = 0%). For the OS studies that we included, it is important to notice that when we extracted data that did not adjust for surgeon and hospital variables, we obtained a significant result (HR 0.88, 95% CI 0.80-0.98, I2 = 0%). These results are summarized in Figure 2.
      Figure thumbnail gr2
      Fig. 2Near- and long-term survival for VATS and Open surgical approach for elderly patients. Forest plots of comparison show dark blue squares which represent single studies (area proportional to sample size). Light blue diamonds are 95% confidence intervals for the outcomes as estimated by the meta-analysis. Values smaller than 1 favor VATS. (Color version of figure is available online.)

      High risk

      After the initial search, six papers were found discussing cancer resection in high-risk patients as indicated in Supplemental Table 1. Of these, three papers were removed because of high risk of bias. Of the remaining studies, two involved large databases, the French Society of Thoracic and Cardiovascular Surgery Database and Society of Thoracic Surgeons - General Thoracic Database, and one a single institution consecutive cohort. Studies included any cancer stage where curative surgical resection could play a role in treatment. One study reported both overall and near-term (30-d) survival, while the other two studies reported only one outcome. Therefore the clinical evidence supporting one approach or the other for this group is of low quality and should be taken as uncertain at best. For near-term survival, highly heterogeneous and contradictory outcomes were reported (random-effects OR 0.74, 95% CI 0.06-8.73, I2 = 87%), similarly for OS (HR 0.55, 95% CI 0.23-1.30, I2 = 51%). It is not surprising to find some degree of heterogeneity as studies are from different populations and countries and included slightly different criteria; however, the inconsistency of these results is remarkable. We used the propensity-matched instead of the inverse probability of treatment weighting because it was used in all papers and the results for OS and DFS were very similar between the two PS methods and had no material effect on the conclusions. These results are summarized in Figure 3.
      Figure thumbnail gr3
      Fig. 3Near- and long-term survival for VATS and Open surgical approach for high-risk patients. Forest plots of comparison show dark blue squares which represent single studies (area proportional to sample size). Light blue diamonds are 95% confidence intervals for the outcomes as estimated by the meta-analysis. Values smaller than 1 favor VATS. (Color version of figure is available online.)

      Early-stage lung cancer

      Although it has been contended that minimally invasive resection may be superior to open approach for early-stage lung cancer, the efficacy of RATS and VATS compared to the open approach remains unclear. Accordingly, we evaluated the efficacy of RATS compared to both VATS and thoracotomy using studies in which RATS had been used and compared to these approaches for early-stage lung cancer. Since such a comparison has not previously been performed, we felt that this was an important clinical question to address in order to improve consensus on the utilization of these two divergent minimally invasive approaches in the treatment of lung cancer. In the literature, there are papers reporting simply RATS outcomes
      • Veronesi G.
      • Park B.
      • Cerfolio R.
      • et al.
      Robotic resection of Stage III lung cancer: an international retrospective study.
      ,
      • Cerfolio R.J.
      • Ghanim A.F.
      • Dylewski M.
      • Veronesi G.
      • Spaggiari L.
      • Park B.J.
      The long-term survival of robotic lobectomy for non-small cell lung cancer: a multi-institutional study.
      or comparing RATS with VATS.
      • Louie B.E.
      • Wilson J.L.
      • Kim S.
      • et al.
      Comparison of video-assisted thoracoscopic surgery and robotic approaches for clinical stage I and stage II non-small cell lung cancer using the Society of Thoracic Surgeons database.
      We focused on papers that compared all three approaches at the same time in order to reduce the risk of bias secondary to lack of matching with an open population. We also chose to follow the literature approach of using RATS as a reference. We did not perform a network meta-analysis because there was not sufficient evidence to rank the three approaches. Forest plots were made comparing RATS versus VATS and RATS versus open approaches for 30-d survival, disease-free survival, and OS. When comparing RATS to VATS, it appeared that disease-free survival was significantly greater when a robotic surgical approach was used (HR 1.47, 95% CI 1.07-2.02, I2 = 0%). However, there appeared to be no significant difference in OS (HR 1.38, 95% CI 0.94-2.00, I2 = 0%) or 30-d survival (RR 1.13, 95% CI 0.88-1.46, I2 = 0). These results are summarized in Figure 4.
      Figure thumbnail gr4
      Fig. 4Near- and long-term survival for RATS and VATS surgical approach for early-stage lung cancer. Forest plots of comparison show dark blue squares which represent single studies (area proportional to sample size). Light blue diamonds are 95% confidence intervals for the outcomes as estimated by the meta-analysis. Note that for these plots values larger than 1 favor RATS. (Color version of figure is available online.)
      Figure thumbnail gr5
      Fig. 5Near- and long-term survival for RATS and Open surgical approach for early-stage lung cancer. Forest plots of comparison show dark blue squares which represent single studies (area proportional to sample size). Dark blue squares represent single studies (area proportional to sample size). Light blue diamonds are 95% confidence intervals for the outcomes as estimated by the meta-analysis. Note that for these plots values larger than 1 favor RATS. (Color version of figure is available online.)
      When comparing RATS to open approaches, no statistically significant difference could be found regarding OS (RR 0.99, 95% CI 0.71-1.39, I2 = 0%) or disease-free survival (RR 1.09, 95% CI 0.82-1.45, I2 = 0%). For 30-d survival, statistical significance was demonstrated in favor of RATS, when using a fixed-effects model (RR 1.28, 95% CI 1.00-1.62, I2 = 41%). However, when a random-effects model was used, the results were no longer significant (RR 1.31, 95% CI 0.82-2.11). These results are summarized in Figure 5.

      Discussion

      Since the advent of VATS and, more recently, RATS approaches to the surgical treatment of lung cancer, a number of studies have attempted to examine outcomes of these procedures compared to thoracotomy. As experience and understanding of these techniques grows, variations upon these approaches continue to develop and their dissemination into more widespread use continues.
      • McCulloch P.
      • Altman D.G.
      • Campbell W.B.
      • et al.
      No surgical innovation without evaluation: the IDEAL recommendations.
      • Ergina P.L.
      • Cook J.A.
      • Blazeby J.M.
      • et al.
      Challenges in evaluating surgical innovation.
      • Barkun J.S.
      • Aronson J.K.
      • Feldman L.S.
      • et al.
      Evaluation and stages of surgical innovations.
      Although these procedures have widely been accepted for early-stage NSCLC,
      • Boffa D.J.
      • Kosinski A.S.
      • Furnary A.P.
      • et al.
      Minimally invasive lung cancer surgery performed by thoracic surgeons as effective as thoracotomy.
      • Yang C.J.
      • Kumar A.
      • Klapper J.A.
      • et al.
      A national analysis of long-term survival following thoracoscopic versus open lobectomy for stage I non-small-cell lung cancer.
      • Gossot D.
      Comparing open and closed chest surgery for early-stage lung cancer: still relevant?.
      • Bhagat R.
      • Bronsert M.R.
      • Henderson W.G.
      • et al.
      Analysis of discharge destination after open vs. minimally invasive surgery for lung cancer.
      • Detterbeck F.
      • Molins L.
      Video-assisted thoracic surgery and open chest surgery in lung cancer treatment: present and future.
      • Bendixen M.
      • Kronborg C.
      • Jorgensen O.D.
      • Andersen C.
      • Licht P.B.
      Cost-utility analysis of minimally invasive surgery for lung cancer: a randomized controlled trial.
      • Long H.
      • Tan Q.
      • Luo Q.
      • et al.
      Thoracoscopic surgery versus thoracotomy for lung cancer: short-term outcomes of a randomized trial.
      • Liang W.
      Comparison of lobectomy, segmentectomy and wedge resection for early stage NSCLC: a direct and network meta-analysis.
      • Yang H.X.
      • Woo K.M.
      • Sima C.S.
      • et al.
      Long-term survival based on the surgical approach to lobectomy for clinical stage I nonsmall cell lung cancer: comparison of robotic, video-assisted thoracic surgery, and thoracotomy lobectomy.
      there is still much debate regarding their utility. Ideally, well-designed, multi-institutional RCTs should be conducted to reach a consensus on the issue. Unfortunately, RCTs are difficult to conduct given issues including technical skills training, lack of clinical equipoise among surgeons,
      • Scott W.J.
      • Allen M.S.
      • Darling G.
      • et al.
      Video-assisted thoracic surgery versus open lobectomy for lung cancer: a secondary analysis of data from the American College of Surgeons Oncology Group Z0030 randomized clinical trial.
      • Yamamoto K.
      • Ohsumi A.
      • Kojima F.
      • et al.
      Long-term survival after video-assisted thoracic surgery lobectomy for primary lung cancer.
      • Rueth N.M.
      • Andrade R.S.
      Is VATS lobectomy better: perioperatively, biologically and oncologically?.
      and logistical issues involving blinding for surgery.
      In our analysis for advanced cancer, 1346 patients, propensity score matched 1:1 between VATS and open surgery were assessed for overall, disease-free, and near-term survival. Meta-analysis found VATS to be associated with decreased near-term mortality when compared to open thoracotomy. It seems more likely than not that the results for high-risk patients are aligned with the other categories and consistent with the largest study we found. For disease-free survival, the meta-analysis itself did not demonstrate statistically significant results. However, significance was found when conducting the sensitivity analysis (Supplemental Fig. 2). This discrepancy highlights the selection bias present in these studies which tend to favor VATS approaches, particularly when propensity matching is not used, but also when it is used. Furthermore, although the analysis demonstrated VATS to be superior for near-term mortality, there was no significant difference for long-term mortality. This may draw into question the utility of VATS for advanced stage cancer if the rates of mortality are similar after 30 d, particularly considering the learning curve, which has been shown to require a minimum of at least 20-30 cases to develop basic proficiency and potentially more than a hundred for true mastery.
      • Li X.
      • Wang J.
      • Ferguson M.K.
      Competence versus mastery: the time course for developing proficiency in video-assisted thoracoscopic lobectomy.
      Given such large numbers, it is very likely that comparisons involving VATS will involve surgeons at various points along this learning curve, making it difficult to draw generalizable conclusions about the technique. Furthermore, these large numbers imply that learning such techniques may be difficult for general thoracic surgeons not located in high-volume centers, and that their adoption in low-volume settings may lead to different outcomes. Moreover, the effect of enhanced recovery protocols mentioned previously
      • Krebs E.D.
      • Mehaffey J.H.
      • Sarosiek B.M.
      • Blank R.S.
      • Lau C.L.
      • Martin L.W.
      Is less really more? Reexamining video-assisted thoracoscopic versus open lobectomy in the setting of an enhanced recovery protocol.
      further highlights that surgical approach may have a more limited impact on patient outcomes than previously thought. In addition, advanced cancer management is evolving toward more complex multi-modality treatments including chemo-, immuno-, and/or radiotherapy, either before/after surgery or both, for which patients usually require an open approach.
      • Bryan D.S.
      • Donington J.S.
      The role of surgery in management of locally advanced non-small cell lung cancer.
      ,
      • Bott M.J.
      • Cools-Lartigue J.
      • Tan K.S.
      • et al.
      Safety and feasibility of lung resection after immunotherapy for metastatic or unresectable tumors.
      As mentioned, a significant number of these complex surgeries might be missing from this analysis because of lack of matching.
      In our analysis for robotic surgery in early-stage cancer resection, we compared RATS to VATS along with RATS to open approaches. When comparing RATS to VATS, it appeared that disease-free survival was significantly greater when a robotic surgical approach was used. However, there appeared to be no significant differences in OS or near-term survival. When comparing RATS to open approaches, no statistically significant difference could be found regarding OS or disease-free survival. The results regarding near-term survival are more ambiguous. A fixed-effects model demonstrated significance, while a random-effects model did not. This further highlights the difficulty associated with making generalizations from these studies.
      Regarding high-risk patients, our analysis demonstrated that there was no difference in OS between patients undergoing VATS or open procedures. However, VATS was associated with remarkable heterogeneity with respect to near-term mortality. Studies found differences in specific and important outcomes that were not the focus of this review, such as atelectasis, pneumonia, and length of stay.
      • Pages P.B.
      • Delpy J.P.
      • Orsini B.
      • et al.
      Propensity score analysis comparing videothoracoscopic lobectomy with thoracotomy: a French Nationwide study.
      Only two papers contained information about 30-d survival.
      • Pages P.B.
      • Delpy J.P.
      • Orsini B.
      • et al.
      Propensity score analysis comparing videothoracoscopic lobectomy with thoracotomy: a French Nationwide study.
      ,
      • Burt B.M.
      • Kosinski A.S.
      • Shrager J.B.
      • Onaitis M.W.
      • Weigel T.
      Thoracoscopic lobectomy is associated with acceptable morbidity and mortality in patients with predicted postoperative forced expiratory volume in 1 second or diffusing capacity for carbon monoxide less than 40% of normal.
      Despite their potential bias related to heterogeneity, particularly in terms of conversions, the results seem to be more consistent with the rest of the literature, making it moderately likely that VATS is favorable for near-term mortality. Only two studies contained information about OS. One study
      • Lau K.K.
      • Martin-Ucar A.E.
      • Nakas A.
      • Waller D.A.
      Lung cancer surgery in the breathless patient—the benefits of avoiding the gold standard.
      reported a much larger difference in OS between surgical approaches compared to the other.
      • Pages P.B.
      • Delpy J.P.
      • Orsini B.
      • et al.
      Propensity score analysis comparing videothoracoscopic lobectomy with thoracotomy: a French Nationwide study.
      This is reflected in an I2 >50% in our analysis. Importantly, these studies show that their conclusions are sensitive to hidden bias such as insufficient incorporation of conversions,
      • Pages P.B.
      • Delpy J.P.
      • Orsini B.
      • et al.
      Propensity score analysis comparing videothoracoscopic lobectomy with thoracotomy: a French Nationwide study.
      which is likely to apply to all or nearly all the studies in this field. Therefore, no conclusive statements can be made. Only one study
      • Lau K.K.
      • Martin-Ucar A.E.
      • Nakas A.
      • Waller D.A.
      Lung cancer surgery in the breathless patient—the benefits of avoiding the gold standard.
      contained information on disease-free survival, showing a generally similar performance for the minimally invasive and open approaches. Of the three studies we evaluated, one is at serious risk of bias (Burt et al. 2014
      • Burt B.M.
      • Kosinski A.S.
      • Shrager J.B.
      • Onaitis M.W.
      • Weigel T.
      Thoracoscopic lobectomy is associated with acceptable morbidity and mortality in patients with predicted postoperative forced expiratory volume in 1 second or diffusing capacity for carbon monoxide less than 40% of normal.
      ) and did not contain any survival data past the first 30 d. Moreover, all of these studies failed to address how conversions from minimally invasive to open approach were treated. Therefore, these conclusions should be interpreted cautiously, even if all the epidemiologically actionable information appears to be currently available. Additional high-quality studies are necessary to make better informed decisions as surgeons currently rely mostly on their intuition and experience.
      Regarding elderly patients, our results once again demonstrate that minimally invasive approaches are superior for near-term mortality, but have limited effect regarding OS. Only two studies were deemed to be at low risk of bias, and the results of the analysis were largely carried by a database study. One study
      • Smith C.B.
      • Wolf A.
      • Mhango G.
      • Wisnivesky J.P.
      Impact of surgeon volume on outcomes of older stage I lung cancer patients treated via video-assisted thoracoscopic surgery.
      seemed to reinforce our suspicion that the results favoring VATS were mostly due to the effect of the surgeons and the hospitals where VATS was performed, rather than the technique itself. This was demonstrated when outcomes were adjusted for surgeon volume, as VATS was not associated with improved survival.
      As with all meta-analyses, this study is not without its limitations. First, it is possible that relevant studies were not identified because of limitations in database literature searches. Second, the evidence is entirely acquired from non-randomized retrospective cohort studies, implying that our results must be interpreted cautiously. Although the predominant inclusion of propensity score-matched studies strengthened our results, these are not a replacement for RCTs. The wide range of institutions and hospitals in the included studies means that there will undoubtedly be variability in both patient characteristics and surgeon skills. Moreover, complex medical issues such as these bias assessments tend to be subjective, and other reviewers might have chosen their studies differently.

      Conclusion

      Although VATS use is increasing nationally, the findings of our analysis seem to suggest that when outcomes are compared to thoracotomy through the use of limited-bias propensity score-matched studies, the benefit of VATS in the setting of advanced lung cancer, high risk, and elderly patients appears to be marginal at most. Though the advantage of VATS seems more clear for early-stage NSCLC, interpretation of the role of VATS in other studied patient cohorts must be considered more cautiously, taking into account selection bias, cost, and learning curves. OS appeared similar between open approaches and VATS in all subgroup comparisons. The steep learning curve associated with mastering VATS may indicate that for surgeons who have not yet learned this technique, the benefit may be minimal or absent. These conclusions suggest that organization of care favoring referrals for minimally invasive approaches to high-volume centers may reduce the impact of learning curves on patient outcomes and lead to favorable mortality and length of stay compared to low-volume centers, as studies have suggested.
      • Tchouta L.N.
      • Park H.S.
      • Boffa D.J.
      • Blasberg J.D.
      • Detterbeck F.C.
      • Kim A.W.
      Hospital volume and outcomes of robot-assisted lobectomies.

      Acknowledgment

      This work was supported by grants from NIH grants HL145478, HL147290, and HL147575 (to AB). The authors are thankful to Ms. Elena Susan for formatting and submission of the manuscript to the journal. LLP would like to thank Thomas Ahern, Larner College of Medicine at the University of Vermont, for many fruitful conversations about epidemiological research related to this project.
      Author Contributions: Adwaiy Manerikar, Lorenzo L. Pesce, and Ankit Bharat contributed to data collection, study planning, critical analysis, and statistical analysis; Melissa Querrey and Emily Cerier contributed to data collection; and David D. Odell and Samuel Kim contributed to study planning and analysis. All authors contributed to manuscript writing.

      Disclosure

      The authors have no conflict of interest to declare. This work was supported by the National Institutes of Health, NIH HL145478, HL147290, and HL147575 (to AB).

      Supplementary data

      References

        • American Cancer Society
        Cancer Facts & Figures 2019.
        American Cancer Society, Atlanta2019
        • Lackey A.
        • Donington J.
        Surgical management of lung cancer.
        Semin Interv Radiol. 2013; 30: 133-140
        • Boffa D.J.
        • Kosinski A.S.
        • Furnary A.P.
        • et al.
        Minimally invasive lung cancer surgery performed by thoracic surgeons as effective as thoracotomy.
        J Clin Oncol. 2018; 36: 2378
        • Yang C.J.
        • Kumar A.
        • Klapper J.A.
        • et al.
        A national analysis of long-term survival following thoracoscopic versus open lobectomy for stage I non-small-cell lung cancer.
        Ann Surg. 2019; 269: 163-171
        • Gossot D.
        Comparing open and closed chest surgery for early-stage lung cancer: still relevant?.
        J Thorac Dis. 2019; 11: S1307-S1309
        • Bhagat R.
        • Bronsert M.R.
        • Henderson W.G.
        • et al.
        Analysis of discharge destination after open vs. minimally invasive surgery for lung cancer.
        Ann Thorac Surg. 2019; 109: 375-382
        • Detterbeck F.
        • Molins L.
        Video-assisted thoracic surgery and open chest surgery in lung cancer treatment: present and future.
        J Vis Surg. 2016; 2: 173
        • Bendixen M.
        • Kronborg C.
        • Jorgensen O.D.
        • Andersen C.
        • Licht P.B.
        Cost-utility analysis of minimally invasive surgery for lung cancer: a randomized controlled trial.
        Eur J Cardiothorac Surg. 2019; 56: 754-761
        • Long H.
        • Tan Q.
        • Luo Q.
        • et al.
        Thoracoscopic surgery versus thoracotomy for lung cancer: short-term outcomes of a randomized trial.
        Ann Thorac Surg. 2018; 105: 386-392
        • Liang W.
        Comparison of lobectomy, segmentectomy and wedge resection for early stage NSCLC: a direct and network meta-analysis.
        J Thorac Oncol. 2018; 13: S889-S890
        • Yang H.X.
        • Woo K.M.
        • Sima C.S.
        • et al.
        Long-term survival based on the surgical approach to lobectomy for clinical stage I nonsmall cell lung cancer: comparison of robotic, video-assisted thoracic surgery, and thoracotomy lobectomy.
        Ann Surg. 2017; 265: 431-437
        • Rajaram R.
        • Mohanty S.
        • Bentrem D.J.
        • et al.
        Nationwide assessment of robotic lobectomy for non-small cell lung cancer.
        Ann Thorac Surg. 2017; 103: 1092-1100
        • McCulloch P.
        • Feinberg J.
        • Philippou Y.
        • et al.
        Progress in clinical research in surgery and IDEAL.
        Lancet. 2018; 392: 88-94
        • Hirst A.
        • Philippou Y.
        • Blazeby J.
        • et al.
        No surgical innovation without evaluation: evolution and further development of the IDEAL framework and recommendations.
        Ann Surg. 2019; 269: 211-220
        • Vannucci F.
        • Gonzalez-Rivas D.
        Is VATS lobectomy standard of care for operable non-small cell lung cancer?.
        Lung Cancer. 2016; 100: 114-119
        • Rothman K.J.
        • Greenland S.
        • Lash T.L.
        Modern Epidemiology.
        3rd ed. Wolters Kluwer Health/Lippincott Williams & Wilkins, Philadelphia2008
        • Bendixen M.
        • Jørgensen O.D.
        • Kronborg C.
        • Andersen C.
        • Licht P.B.
        Postoperative pain and quality of life after lobectomy via video-assisted thoracoscopic surgery or anterolateral thoracotomy for early stage lung cancer: a randomised controlled trial.
        Lancet Oncol. 2016; 17: 836-844
        • Zhao J.
        • Zhao Y.
        • Qiu T.
        • et al.
        Quality of life and survival after II stage nonsmall cell carcinoma surgery: video-assisted thoracic surgery versus thoracotomy lobectomy.
        Indian J Cancer. 2015; 52: e130-e133
        • Rizk N.P.
        • Ghanie A.
        • Hsu M.
        • et al.
        A prospective trial comparing pain and quality of life measures after anatomic lung resection using thoracoscopy or thoracotomy.
        Ann Thorac Surg. 2014; 98: 1160-1166
        • Rodgers-Fischl P.M.
        • Martin J.T.
        • Saha S.P.
        Video-assisted thoracoscopic versus open lobectomy: costs and outcomes.
        South Med J. 2017; 110: 229-233
        • Kneuertz P.J.
        • Singer E.
        • D'Souza D.M.
        • Abdel-Rasoul M.
        • Moffatt-Bruce S.D.
        • Merritt R.E.
        Hospital cost and clinical effectiveness of robotic-assisted versus video-assisted thoracoscopic and open lobectomy: a propensity score-weighted comparison.
        J Thorac Cardiovasc Surg. 2019; 157 (e2012): 2018-2026
        • Ng C.S.H.
        • MacDonald J.K.
        • Gilbert S.
        • et al.
        Optimal approach to lobectomy for non-small cell lung cancer: systemic review and meta-analysis.
        Innovations (Phila). 2019; 14: 90-116
        • Liberati A.
        • Altman D.G.
        • Tetzlaff J.
        • et al.
        The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.
        Ann Intern Med. 2009; 151: W65-W94
        • Ouzzani M.
        • Hammady H.
        • Fedorowicz Z.
        • Elmagarmid A.
        Rayyan—a web and mobile app for systematic reviews.
        Syst Rev. 2016; 5: 210
        • Wells BS G.
        • O’Connell D.
        • Peterson J.
        The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses 2000.
        • Sterne J.A.
        • Hernan M.A.
        • Reeves B.C.
        • et al.
        ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions.
        BMJ. 2016; 355: i4919
      1. WebPlotDigitizer [computer program]. Version 4.2.
        • Schwarzer G.
        meta: an R package for meta-analysis.
        R News. 2007; 7: 40-45
        • Fernando H.C.
        • Landreneau R.J.
        • Mandrekar S.J.
        • et al.
        The impact of adjuvant brachytherapy with sublobar resection on pulmonary function and dyspnea in high-risk patients with operable disease: preliminary results from the American College of Surgeons Oncology Group Z4032 Trial.
        J Thorac Cardiovasc Surg. 2011; 142: 554-562
        • Detillon D.D.
        • Veen E.J.
        Postoperative outcome after pulmonary surgery for non-small cell lung cancer in elderly patients.
        Ann Thorac Surg. 2018; 105: 287-293
        • Endoh H.
        • Yamamoto R.
        • Satoh Y.
        • Kuwano H.
        • Nishizawa N.
        Risk analysis of pulmonary resection for elderly patients with lung cancer.
        Surg Today. 2013; 43: 514-520
        • Ezer N.
        • Kale M.
        • Sigel K.
        • et al.
        Outcomes after video-assisted thoracoscopic lobectomy versus open lobectomy for early-stage lung cancer in older adults.
        Ann Am Thorac Soc. 2018; 15: 76-82
        • Smith C.B.
        • Kale M.
        • Mhango G.
        • et al.
        Comparative outcomes of elderly stage I lung cancer patients treated with segmentectomy via video-assisted thoracoscopic surgery versus open resection.
        J Thorac Oncol. 2014; 9: 383-389
        • Pages P.B.
        • Mariet A.S.
        • Madelaine L.
        • et al.
        Impact of video-assisted thoracic surgery approach on postoperative mortality after lobectomy in octogenarians.
        J Thorac Cardiovasc Surg. 2019; 157: 1660-1667
        • Veronesi G.
        • Park B.
        • Cerfolio R.
        • et al.
        Robotic resection of Stage III lung cancer: an international retrospective study.
        Eur J Cardiothorac Surg. 2018; 54: 912-919
        • Cerfolio R.J.
        • Ghanim A.F.
        • Dylewski M.
        • Veronesi G.
        • Spaggiari L.
        • Park B.J.
        The long-term survival of robotic lobectomy for non-small cell lung cancer: a multi-institutional study.
        J Thorac Cardiovasc Surg. 2018; 155: 778-786
        • Louie B.E.
        • Wilson J.L.
        • Kim S.
        • et al.
        Comparison of video-assisted thoracoscopic surgery and robotic approaches for clinical stage I and stage II non-small cell lung cancer using the Society of Thoracic Surgeons database.
        Ann Thorac Surg. 2016; 102: 917-924
        • McCulloch P.
        • Altman D.G.
        • Campbell W.B.
        • et al.
        No surgical innovation without evaluation: the IDEAL recommendations.
        Lancet. 2009; 374: 1105-1112
        • Ergina P.L.
        • Cook J.A.
        • Blazeby J.M.
        • et al.
        Challenges in evaluating surgical innovation.
        Lancet. 2009; 374: 1097-1104
        • Barkun J.S.
        • Aronson J.K.
        • Feldman L.S.
        • et al.
        Evaluation and stages of surgical innovations.
        Lancet. 2009; 374: 1089-1096
        • Scott W.J.
        • Allen M.S.
        • Darling G.
        • et al.
        Video-assisted thoracic surgery versus open lobectomy for lung cancer: a secondary analysis of data from the American College of Surgeons Oncology Group Z0030 randomized clinical trial.
        The J Thorac Cardiovasc Surg. 2010; 139: 976-983
        • Yamamoto K.
        • Ohsumi A.
        • Kojima F.
        • et al.
        Long-term survival after video-assisted thoracic surgery lobectomy for primary lung cancer.
        Ann Thorac Surg. 2010; 89: 353-359
        • Rueth N.M.
        • Andrade R.S.
        Is VATS lobectomy better: perioperatively, biologically and oncologically?.
        Ann Thorac Surg. 2010; 89: S2107-S2111
        • Li X.
        • Wang J.
        • Ferguson M.K.
        Competence versus mastery: the time course for developing proficiency in video-assisted thoracoscopic lobectomy.
        J Thorac Cardiovasc Surg. 2014; 147: 1150-1154
        • Krebs E.D.
        • Mehaffey J.H.
        • Sarosiek B.M.
        • Blank R.S.
        • Lau C.L.
        • Martin L.W.
        Is less really more? Reexamining video-assisted thoracoscopic versus open lobectomy in the setting of an enhanced recovery protocol.
        J Thorac Cardiovasc Surg. 2019; S0022-5223: 31771-31774
        • Bryan D.S.
        • Donington J.S.
        The role of surgery in management of locally advanced non-small cell lung cancer.
        Curr Treat Options Oncol. 2019; 20: 27
        • Bott M.J.
        • Cools-Lartigue J.
        • Tan K.S.
        • et al.
        Safety and feasibility of lung resection after immunotherapy for metastatic or unresectable tumors.
        Ann Thorac Surg. 2018; 106: 178-183
        • Pages P.B.
        • Delpy J.P.
        • Orsini B.
        • et al.
        Propensity score analysis comparing videothoracoscopic lobectomy with thoracotomy: a French Nationwide study.
        Ann Thorac Surg. 2016; 101: 1370-1378
        • Burt B.M.
        • Kosinski A.S.
        • Shrager J.B.
        • Onaitis M.W.
        • Weigel T.
        Thoracoscopic lobectomy is associated with acceptable morbidity and mortality in patients with predicted postoperative forced expiratory volume in 1 second or diffusing capacity for carbon monoxide less than 40% of normal.
        J Thorac Cardiovasc Surg. 2014; 148 (dicussion 28-29 e11): 19-28
        • Lau K.K.
        • Martin-Ucar A.E.
        • Nakas A.
        • Waller D.A.
        Lung cancer surgery in the breathless patient—the benefits of avoiding the gold standard.
        Eur J Cardiothorac Surg. 2010; 38: 6-13
        • Smith C.B.
        • Wolf A.
        • Mhango G.
        • Wisnivesky J.P.
        Impact of surgeon volume on outcomes of older stage I lung cancer patients treated via video-assisted thoracoscopic surgery.
        Semin Thorac Cardiovasc Surg. 2017; 29: 223-230
        • Tchouta L.N.
        • Park H.S.
        • Boffa D.J.
        • Blasberg J.D.
        • Detterbeck F.C.
        • Kim A.W.
        Hospital volume and outcomes of robot-assisted lobectomies.
        Chest. 2017; 151: 329-339