- Email: [email protected]
Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients
Journal Pre-proof Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients Rajwanth R. Veluswamy, MD, MSCR, Stacey-Ann Whittaker Brown, MD, Grace Mhango, MPH, Keith Sigel, MD, PhD, Daniel G. Nicastri, MD, Cardinale B. Smith, MD, PhD, Marcelo Bonomi, MD, Matthew D. Galsky, MD, Emanuela Taioli, MD, PhD, Alfred I. Neugut, MD, PhD, Juan P. Wisnivesky, MD, DrPH PII:
S0012-3692(19)33946-7
DOI:
https://doi.org/10.1016/j.chest.2019.09.017
Reference:
CHEST 2646
To appear in:
CHEST
Received Date: 29 November 2018 Revised Date:
29 August 2019
Accepted Date: 14 September 2019
Please cite this article as: Veluswamy RR, Whittaker Brown SA, Mhango G, Sigel K, Nicastri DG, Smith CB, Bonomi M, Galsky MD, Taioli E, Neugut AI, Wisnivesky JP, Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients CHEST (2019), doi: https:// doi.org/10.1016/j.chest.2019.09.017. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Copyright © 2019 Published by Elsevier Inc under license from the American College of Chest Physicians.
Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients Rajwanth R. Veluswamy, MD, MSCR,1,2 Stacey-Ann Whittaker Brown, MD,3 Grace Mhango, MPH,4 Keith Sigel, MD, PhD,4 Daniel G. Nicastri, MD,5 Cardinale B. Smith, MD, PhD,1,6 Marcelo Bonomi, MD,7 Matthew D. Galsky, MD,1 Emanuela Taioli, MD, PhD,2,8 Alfred I. Neugut, MD, PhD,9,10 Juan P. Wisnivesky, MD, DrPH3,4 1
Division of Hematology/Oncology, Icahn School of Medicine at Mount Sinai, New York, NY; Institute of Translational Epidemiology, Icahn School of Medicine at Mount Sinai, New York, NY; 3Division of Pulmonary, Critical Care, and Sleep Medicine, Icahn School of Medicine at Mount Sinai, New York, NY; 4Division of General Internal Medicine, Icahn School of Medicine at Mount Sinai, New York, NY; 5Department of Thoracic Surgery, Icahn School of Medicine at Mounts Sinai, New York, NY; 6Hertzberg Palliative Care Institute of the Brookdale Department of Geriatrics, Icahn School of Medicine at Mount Sinai, New York, NY; 7Section of Hematology and Medical Oncology, Department of Internal Medicine, Wake Forest School of Medicine, Winston-Salem, NC; 8Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai, New York, NY; 9Division of Hematology/Oncology, Department of Medicine, Columbia University, New York, NY; and 10Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY 2
Corresponding Author:
Rajwanth Veluswamy, MD Division of Hematology and Medical Oncology Icahn School of Medicine at Mount Sinai One Gustave L. Levy Place, Box 1079 New York, NY 10029-6574 Phone: (212) 824-8580 Fax: (646) 537-9639 E-mail: [email protected]
Running Title: Robotic-Assisted Surgery for Older NSCLC Patients COI disclosure: RRV has received consulting honorarium from Onconova Therapeutics and Astrazeneca. JPW is a member of the research board of EHE International, has received consulting honorarium from Merck, Quintiles and Astrazeneca and research grants from Sanofi and Quorum. AIN is a member of the research board of EHE International and has received consulting honoraria from Otsuka, Hospira, and United Biosource Corporation. RRV, SAW, GM, KS, DGN, CBS, MB, MDG and ET have no conflict of interest to declare. Financial support: This study was supported by the Conquer Cancer Foundation of the American Society of Clinical Oncology Young Investigator Award and the ISMMS Clinical and Translational Science Award (CTSA). Word count: Abstract: 245; Manuscript text: 2500 (11 pages); Tables: 3; Figures: 1 1
Keywords: NSCLC, Robotic Surgery, Minimally Invasive, Early Stage, Treatment Abbreviations List: CI: confidence interval GLLM: generalized linear mixed model HR: hazard ratio ICU: intensive care unit LOS: length of stay NSCLC: non-small cell lung cancer OR: odds ratio PET: positron emission tomography RAS: robotic-assisted surgery SEER: Surveillance, Epidemiology, and End Results US: United States VATS: video-assisted thoracic surgery
2
Abstract
Background: Robotic-assisted surgery (RAS) is a novel surgical approach increasingly used for patients with non-small cell lung cancer (NSCLC). However, data comparing the effectiveness and costs of RAS vs. open thoracotomy and video-assisted thoracic surgery (VATS) for NSCLC is limited. Methods: Patients >65 years old with stage I-IIIA NSCLC treated with RAS, VATS or open thoracotomy were identified from the Surveillance, Epidemiology, and End Results-Medicare database and matched by age, sex, stage, and extent of resection. We used propensity score methods to compare adjusted rates of postoperative complications, adequate lymph node staging, survival, and treatment-related costs. Results: In our matched study cohort of 2, 766 resected NSCLC patients, RAS was associated with lower complication rates (odds ratio [OR]: 0.57; 95% confidence interval [CI]: 0.42-0.79) compared to open thoracotomy, and similar complication rates (OR: 1.02; 95% CI: 0.76-1.37) compared to VATS. RAS patients were as likely to have adequate lymph node sampling as those undergoing open thoracotomy (OR: 1.28; 95% CI: 0.94-1.74) or VATS (OR: 0.88; 95% CI: 0.66-1.18). There was no significant difference in overall survival after RAS vs. open thoracotomy (hazard ratio [HR]: 0.81, 95% CI: 0.63-1.04) or VATS (HR: 0.91; 95% CI: 0.701.18). Costs were similar for RAS ($54,702) vs. open thoracotomy ($57,104; p=0.08), and higher compared to VATS ($48,729; p=0.02). Conclusions: RAS leads to improved operative outcomes compared to open thoracotomy but may not offer an advantage over VATS. The comparative effectiveness of RAS should be further evaluated prior to widespread adoption.
3
Introduction
Optimal surgical management of non-small cell lung cancer (NSCLC) is critical for maximizing chances of cure and achieving good long-term outcomes. These treatment decisions are becoming even more important given the expected increase in the proportion of tumors diagnosed at earlier stages following full implementation of lung cancer screening recommendations.1 However, with a mean age at diagnosis of 70 years, most NSCLC patients are older and frequently have smoking-related comorbidities, creating challenges for resection.2 Thus, the identification of effective and well-tolerated surgical approaches has been a major area of clinical interest.
The standard surgical approach for resectable stage I-IIIA NSCLC has been lobectomy via open thoracotomy.3 However, open lobectomy is associated with considerable postoperative complications and mortality, especially in the elderly.4,5 As a consequence, minimally invasive techniques, such as video-assisted thoracoscopic surgery (VATS), have been increasingly adopted.3 While there have been no randomized controlled trials, several retrospective studies have shown that VATS lobectomy is associated with fewer postoperative complications compared to open lobectomy.6-8 Additionally, the total costs for lobectomy have been shown to be lower when using a VATS approach.9
More recently, surgical approaches using robotic systems have been designed to enhance instrument maneuverability and visualization, enabling surgeons to perform intricate maneuvers with greater precision. Prior case series have demonstrated acceptable perioperative results, low thoracotomy conversion rates, and comparable survival among NSCLC patients treated with 4
robotic-assisted surgery (RAS).10-12 However, there is limited data regarding the comparative effectiveness of RAS vs. VATS or open resection for older NSCLC patients, a group at high risk for poor outcomes. Furthermore, the financial cost of RAS has not been adequately described. As a consequence, RAS is being adopted into routine care without adequate scrutiny.
In this study, we used population-based data from the Surveillance, Epidemiology, and End Results (SEER)-Medicare database to assess postoperative, oncologic and cost outcomes of older patients with stage I-IIIA NSCLC treated with RAS vs. VATS or open thoracotomy.
Methods
Study Variables
We included all patients >65 years from SEER-Medicare with stage I-IIIA NSCLC who were diagnosed between 2008 (first year of RAS in registry) and 2013, and treated with surgical resection via RAS, open thoracotomy or VATS.13 In order to create a well-balanced cohort, we matched RAS-treated patients with up to four patients of the same age (within 5 years), sex, stage, and extent of surgery (wedge, segmentectomy or lobectomy) who underwent open thoracotomy or VATS. We excluded patients treated with neoadjuvant treatment (may represent higher stage), in health maintenance organizations or without outpatient Medicare coverage due to lack of claims data.
We obtained pretreatment information, including sociodemographics, comorbidities, tumor characteristics and stage, and diagnostic, staging, and preoperative evaluation, from both SEER 5
and Medicare data.14-17 Using procedure codes from Medicare claims, we classified study subjects as having undergone RAS, VATS or open thoracotomy. We ranked surgeons according to the volume of total resections they performed in the year of the patient’s surgery.18,19 See online appendix for more details.
Postoperative morbidity and mortality (primary outcomes) were defined as the presence of complications and death, respectively, within 30 days of surgery or during the hospitalization in which surgery was performed. Complications were determined from Medicare claims and included: 1) extrapulmonary infections; 2) cardiovascular complications; 3) thromboembolic events; 4) respiratory complications; 5) reoperations; 6) readmissions; 7) prolonged length of stay (>14 days; LOS); and 8) intensive care unit (ICU) management. Postoperative mortality was determined using the dates of surgery and death, obtained from Medicare files.
From SEER data, we compared the extent of lymph node dissection (≤10 vs. >10 nodes; cutoff recommended for adequate staging) with each surgical approach as a secondary outcome.20 Survival was calculated from the date of surgery to the date of death or last follow-up (December 31, 2015 for censored observations) according to Medicare data. For analyses involving lungcancer specific survival, we used SEER to ascertain the cause of death from state death certificates.
Treatment-related costs were estimated from the payer perspective using Medicare reimbursement data from inpatient, outpatient, physician, durable medical equipment, and home health agency files.21 This costs data does not include capital and maintenance costs for surgical
6
equipment incurred by the institution. We categorized costs as: 1) pre-operative (within thirty days prior to surgery); 2) operative (within 30 days of surgery); and 3) post-operative (from postoperation day 31 through 365).22 Total cost was calculated as the sum of the costs of each phase of treatment and standardized to 2012 United States (US) dollars using the Consumer Price Index.23
Statistical Analysis
We compared the distribution of pretreatment characteristics of patients in each study group using analysis of variance or the chi-square test. Given the lack of randomization, we applied propensity score methods to minimize the effects of measured confounders on treatment allocation. We used nominal regression to calculate propensity scores indicating the probability of patients undergoing RAS, open thoracotomy, or VATS based on pretreatment information.
We compared adjusted rates of postoperative complications and mortality, as well as the number of lymph nodes assessed, among patients treated with open thoracotomy, VATS or RAS with a generalized linear mixed model (GLMM; adjusting confidence intervals for clustering due to matching). We adjusted for covariates using inverse probability weighting; weights were calculated as the inverse of the estimated propensity score for undergoing the type of surgical treatment each patient received. All models were also adjusted for surgeon’s case volume. We conducted secondary analyses: 1) limiting the cohort to patients with stage I-II NSCLCs as stage IIIA tumors may not be amenable to primary surgical resection; and 2) stratifying patients who underwent RAS during the early (2008-2011) and later (2012 and 2013) years of uptake to
7
determine if operator experience improved outcomes. We conducted adjusted survival comparisons using Cox regression analysis with inverse probability weighting by propensity scores and robust standard errors to incorporate clustering. Finally, we compared costs using a GLMM with an identity link and γ-distributed variance function.
Results
Of the 2,766 stage I-IIIA NSCLC patients included in the study, 338 (12%) underwent RAS, 1230 (45%) received VATS, and 1198 (43%) were managed with open thoracotomy. There were significant differences in the distribution of age (p<0.001), race (p<0.001), household income (p<0.001), comorbidities (p<0.001), tumor size (p<0.001), histology (p=0.006) and preoperative assessment with positron emission tomography (PET; p<0.001) or mediastinoscopy (p<0.001) across the three treatment groups. After adjusting for propensity scores, all measured variables were well balanced (Table 1).
Propensity score adjusted analysis showed that overall surgical complication rates were lower for patients undergoing RAS compared to those treated with open thoracotomy (odds ratio [OR]: 0.57; 95% confidence interval [CI]: 0.42-0.79). Specifically, rates of blood transfusions (OR: 0.31; 95% CI: 0.17-0.60), ICU stays (OR: 0.58; 95% CI: 0.44-0.77) and prolonged LOS (OR: 0.59; 95% CI: 0.35-0.98) were significantly lower in patients undergoing RAS (Table 2). There was no statistical difference in adequate lymph node staging (OR: 1.28; 95% CI: 0.94-1.74) and exploratory survival analyses (hazard ratio [HR]: 0.81, 95% CI: 0.63-1.04; HR: 0.75, 95% CI: 0.51-1.12 for overall and lung cancer-specific survival, respectively) when comparing RAS to
8
open thoracotomy (Table 3). Finally, while total costs for patients who underwent RAS were similar than for those who underwent open thoracotomy ($54,702 vs. $57,104; p=0.08), costs during the operative period were significantly lower for RAS ($28,732 vs. $32,746; p<0.0001). Pre-operative ($3,668 vs. $3,199; p=0.12) and post-operative ($22,302 vs. $21,158; p=0.95) costs were similar between RAS and open thoracotomy (Figure 1).
When compared to patients who underwent VATS, those treated with RAS had similar overall complication rates related to surgery (OR: 1.02; 95% CI: 0.76-1.37; p>0.05 for individual complications; Table 2). Both approaches had similar odds of undergoing adequate lymph node evaluation (OR: 0.88; 95% CI: 0.66-1.18) and did not have statistically different survival (HR: 0.91; 95% CI: 0.70-1.18; HR: 0.87; 95% CI: 0.56-1.34 for overall and lung cancer-specific survival, respectively; Table 3). RAS patients incurred higher total costs ($54,702 vs. $48,729; p=0.02) and pre-operative costs ($3,668 vs. $2,803; p<0.0001) compared to patients undergoing VATS. Costs during the operative ($28,732 vs. $27,209; p=0.078) and post-operative ($22,302 vs. $18,718; p=0.15) periods were similar between the two minimally invasive procedures (Figure 1).
Secondary analyses limiting the cohort to stage I-II NSCLC patients (N=2529) showed similar postoperative and oncological outcomes as those observed in the analysis of the entire cohort, with the exception that prolonged LOS rates were now similar in all three surgical groups (Tables 2 and 3). When limited to the initial years of uptake (2008-2011; N=123), RAS patients had lower rates of blood transfusions (OR: 0.28, 95% CI: 0.10-0.83) compared to open thoracotomy; otherwise, they experienced similar rates of complications as patients who
9
underwent open thoracotomy (OR: 0.78; 95% CI: 0.45-1.35) and VATS (OR: 1.17; 95% CI: 0.70-1.96). Complication rates in RAS patients, however, improved in the more recent years of uptake (2012-2013; N=215) when compared to those who received open thoracotomy, with lower rates of: total complications (OR: 0.49; 95% CI: 0.33-0.73), blood transfusions (OR: 0.34; 95% CI: 0.15-0.75) and ICU stays (OR: 0.49; 95% CI: 0.35-0.68). RAS continued to be associated with similar rates of complications compared to VATS (OR: 0.95; 95% CI: 0.661.36).
Discussion
RAS has rapidly expanded into the routine practice of various surgical specialties, leading to almost a 200% increase in utilization between 1999 and 2014.28 Currently, it is estimated that one in four US hospitals have a robotic system.24 However, high quality evidence is still needed to determine the comparative effectiveness and costs of RAS in NSCLC resections.25 In this population-based study of older patients, we found RAS was associated with reduced rate of complications when compared to open thoracotomy, and similar complication rates when compared to VATS. We further found that the improvement of postoperative outcomes of RAS vs. open thoracotomy occurred in the later years of uptake. While RAS was associated with similar total costs than open thoracotomy, we found significantly higher costs when compared to VATS. Therefore, it remains unclear if RAS offers any benefit beyond VATS.
Data evaluating postoperative outcomes of RAS for NSCLC mostly comes from smaller case series of single institutions.10-12 More recently, comparative effectiveness studies using various databases have reported conflicting results. Data from the Premier Hospital Network found no 10
difference between RAS and VATS in overall complication rates, LOS, and surgical mortality, but the former was associated with higher costs and longer operating times.26 Conversely, analysis of the National Inpatient Sample found that patients undergoing RAS compared to VATS were associated with higher rates of intraoperative bleeding complications and were more likely to be discharged to facilities rather than home.27 However, these studies did not adjust for important imbalances in tumor characteristics, nor account for surgeon case volume.28,29 Our study used nationally representative data and rigorously adjusted for all measured patient, tumor, and surgery characteristics. We additionally focused on older patients, the age group most commonly affected by NSCLC and more vulnerable to experiencing surgical complications.
Interestingly, we found that RAS was associated with lower postoperative complications when compared to open thoracotomy only in the later study years (2012-2013), suggesting that postoperative outcomes improved with increasing experience and/or better patient selection with this technique. Improved operative outcomes for RAS were also found in an analysis of the State Inpatient Database, using a study cohort where the majority of resections (>75%) were performed by high-volume surgeons.30 There is a steep learning curve for surgeons to become proficient with RAS.31 New generations of robotic systems (i.e. Xi platform) are being developed which may improve user experience and clinical outcomes. In addition, widespread, standardized training of RAS using virtual training modules through the robotic console may allow surgeons to gain proficiency in robotic skills and reduce complication rates.
The oncologic outcomes of RAS for NSCLC have also not yet been adequately described. Prior case series have demonstrated intraoperative mediastinal lymph node staging to be similar with
11
both VATS and RAS.20,32 Additionally, the 5-year survival of NSCLC patients undergoing robotic lobectomy was similar to those historically reported for VATS or open thoracotomy. 33 However, these studies lack a comparison arm, have potential biases in patient selection, and have limited generalizability. Our findings, accounting for these limitations, provide stronger level of evidence confirming these conclusions. Of note, our results are an overall average treatment effect after adjusting for important clinical confounders. We are unable to determine potential differing treatment effects in individual patient subgroups, as this would significantly reduce the power of our study. Future studies can be undertaken to explore the role of each surgical approach in specific subgroups.
The implementation of new technology must consider the relative costs compared to existing interventions. We found RAS to be associated with similar total costs compared with open thoracotomy, but lower costs during the operative period. These results are consistent with prior studies evaluating prostatectomy and hysterectomy.34,35 When compared to VATS, RAS was associated with significantly higher total costs, specifically those incurred in the pre-operative period. Of note, our findings do not take into account expenses associated with robotic surgical systems that are incurred by institutions, such as costs for the robot itself ($1-$2.5 million), maintenance ($80-$170 thousand annually) and additional consumables ($700-$3,500 per robotic case).24,36,37 Medicare claims data instead provides reimbursements paid (i.e., from the payer perspective) for all services rendered to beneficiaries. Over 55 million Americans (17% of US population) rely on Medicare as their primary insurance coverage; 80% of beneficiaries are age 65 or older. Therefore, Medicare data provides highly representative information for evaluating costs of medical interventions in the US.38
12
There are some limitations of the present study. Lack of random treatment allocation can result in systematic differences in the distribution of pretreatment characteristics between treatment groups, which could affect both the type of treatment a patient receives and their clinical outcomes. To mitigate this potential bias, we matched patients for key factors and applied propensity score methods to balance groups for all other measured confounders. However, we were unable to account for unmeasured confounders, a particular limitation inherent to all observational studies. Additionally, we did not have data on completeness of resection (i.e., R0 vs. R1 vs. R2 resection) or conversion rates for minimally invasive procedures to open thoracotomy. Conversions may unfavorably bias both operative and costs outcomes for open thoracotomy. However, a large, single institution case series previously reported rates of incomplete resections and conversions are low and similar between RAS and VATS.39 Finally, we were not able enumerate all costs related to each surgical approach, and therefore future research should explore insights into mechanisms of costs difference between each procedure.
Conclusions
Our population-based study of older NSCLC patients found that RAS offers improved postoperative outcomes compared to open thoracotomy, but no such benefit was observed vs. VATS. Furthermore, while we found total costs were similar between RAS and open thorocatomy, RAS was associated with higher costs compared to VATS. Lymph node evaluation and exploratory survival rates were similar between all three approaches. While RAS is innovative, we found no distinct advantage over VATS to support its adoption into routine care
13
for the surgical management of patients with NSCLC. As technology and operator experience with RAS improve with time, continued reassessment of comparative effectiveness should be undertaken to ensure patients receive the most effective and safe surgical option.
14
Acknowledgments
The authors would like to thank the Conquer Cancer Foundation of the American Society of Clinical Oncology; the Applied Research Branch, Division of Cancer Prevention and Population Science, National Cancer Institute; the Office of Information Services and the Office of Strategic Planning, Health Care Finance Administration; Information Management Services (IMS), Inc.; the SEER Program tumor registries for their efforts in the creation of the SEER-Medicare Database; and the ISMMS Clinical and Translational Science Award (CTSA). The collection, interpretation and reporting of these data are the sole responsibilities of the authors. All significant contributors of this manuscript have been listed.
Author Contributions: Conception and design, Manuscript Writing, Final approval: All authors Collection, assembly, analysis and interpretation of data: RRV, SAW, GM, JPW
15
References 1. 2. 3. 4.
5.
6.
7.
8.
9.
10. 11. 12. 13.
14. 15. 16.
17. 18.
Moyer VA. Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014;160(5):330-338. Howlader N NA, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2011. 2013; http://seer.cancer.gov/csr/1975_2011/. Accessed August, 2015. National Comprehensive Cancer Network. Non-Small Cell Lung Cancer (Version 4.2016). http://www.nccn.org/professionals/physician_gls/pdf/nscl.pdf. Accessed March 11. Scott WJ, Matteotti RS, Egleston BL, Oseni S, Flaherty JF. A comparison of perioperative outcomes of video-assisted thoracic surgical (VATS) lobectomy with open thoracotomy and lobectomy: results of an analysis using propensity score based weighting. Ann Surg Innov Res. 2010;4(1):1. Whitson BA, Groth SS, Duval SJ, Swanson SJ, Maddaus MA. Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy. Ann Thorac Surg. 2008;86(6):2008-2016; discussion 2016-2008. Yan TD, Black D, Bannon PG, McCaughan BC. Systematic review and meta-analysis of randomized and nonrandomized trials on safety and efficacy of video-assisted thoracic surgery lobectomy for early-stage non-small-cell lung cancer. J Clin Oncol. 2009;27(15):2553-2562. Swanson SJ, Herndon JE, 2nd, D'Amico TA, et al. Video-assisted thoracic surgery lobectomy: report of CALGB 39802--a prospective, multi-institution feasibility study. J Clin Oncol. 2007;25(31):4993-4997. Cajipe MD, Chu D, Bakaeen FG, et al. Video-assisted thoracoscopic lobectomy is associated with better perioperative outcomes than open lobectomy in a veteran population. Am J Surg. 2012;204(5):607-612. Swanson SJ, Meyers BF, Gunnarsson CL, et al. Video-assisted thoracoscopic lobectomy is less costly and morbid than open lobectomy: a retrospective multiinstitutional database analysis. Ann Thorac Surg. 2012;93(4):1027-1032. Toker A. Robotic thoracic surgery: from the perspectives of European chest surgeons. J Thorac Dis. 2014;6 Suppl 2:S211-216. Augustin F, Bodner J, Wykypiel H, Schwinghammer C, Schmid T. Initial experience with robotic lung lobectomy: report of two different approaches. Surg Endosc. 2011;25(1):108-113. Gharagozloo F, Margolis M, Tempesta B, Strother E, Najam F. Robot-assisted lobectomy for early-stage lung cancer: report of 100 consecutive cases. Ann Thorac Surg. 2009;88(2):380-384. Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40(8 Suppl):Iv-3-18. Wisnivesky JP, Henschke CI, Swanson S, et al. Limited resection for the treatment of patients with stage IA lung cancer. Annals of surgery. 2010;251(3):550-554. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 2000;53(12):1258-1267. Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996-1998 practice patterns. Radiology. 2002;222(1):144-148. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-smallcell lung cancer. J Clin Oncol. 2006;24(3):413-418. Earle CC, Schrag D, Neville BA, et al. Effect of surgeon specialty on processes of care and outcomes for ovarian cancer patients. J Natl Cancer Inst. 2006;98(3):172-180. 16
19. 20.
21. 22.
23. 24.
25. 26.
27. 28. 29. 30.
31. 32. 33. 34.
35. 36. 37. 38.
Schrag D, Earle C, Xu F, et al. Associations between hospital and surgeon procedure volumes and patient outcomes after ovarian cancer resection. J Natl Cancer Inst. 2006;98(3):163-171. Shapiro M, Mhango G, Kates M, et al. Extent of lymph node resection does not increase perioperative morbidity and mortality after surgery for stage I lung cancer in the elderly. Eur J Surg Oncol. 2012;38(6):516-522. Brown ML, Riley GF, Schussler N, Etzioni R. Estimating health care costs related to cancer treatment from SEER-Medicare data. Med Care. 2002;40(8 Suppl):Iv-104-117. Smith BD, Jiang J, Chang JY, et al. Cost-effectiveness of stereotactic radiation, sublobar resection, and lobectomy for early non-small cell lung cancers in older adults. J Geriatr Oncol. 2015;6(4):324-331. http://www2.census.gov/geo/docs/maps-data/maps/reg_div.txt CbradwsFcaFAf. Intuitive Surgical Investor Presentation Q4 2014. Available from: https://www.rts.ch/emissions/tempspresent/6617343.html/BINARY/La%20pr%C3%A9sentation%20d%E2%80%99Intuitive%20Surgic al. Swanson SJ. Robotic pulmonary lobectomy--the future and probably should remain so. J Thorac Cardiovasc Surg. 2010;140(5):954. Swanson SJ, Miller DL, McKenna RJ, Jr., et al. Comparing robot-assisted thoracic surgical lobectomy with conventional video-assisted thoracic surgical lobectomy and wedge resection: results from a multihospital database (Premier). J Thorac Cardiovasc Surg. 2014;147(3):929-937. Paul S, Jalbert J, Isaacs AJ, Altorki NK, Isom OW, Sedrakyan A. Comparative effectiveness of robotic-assisted vs thoracoscopic lobectomy. Chest. 2014;146(6):1505-1512. Irie M, Nakanishi R, Yasuda M, Fujino Y, Hamada K, Hyodo M. Risk factors for short-term outcomes after thoracoscopic lobectomy for lung cancer. Eur Respir J. 2016;48(2):495-503. Husain ZA, Kim AW, Yu JB, Decker RH, Corso CD. Defining the High-Risk Population for Mortality After Resection of Early Stage NSCLC. Clin Lung Cancer. 2015;16(6):e183-187. Kent M, Wang T, Whyte R, Curran T, Flores R, Gangadharan S. Open, video-assisted thoracic surgery, and robotic lobectomy: review of a national database. Ann Thorac Surg. 2014;97(1):236-242; discussion 242-234. Suda T. Transition from video-assisted thoracic surgery to robotic pulmonary surgery. The Journal of Visualized Surgery. 2017;3(4). Lee HS, Jang HJ. Thoracoscopic mediastinal lymph node dissection for lung cancer. Semin Thorac Cardiovasc Surg. 2012;24(2):131-141. Park BJ. Robotic lobectomy for non-small cell lung cancer: long-term oncologic results. Thorac Surg Clin. 2014;24(2):157-162, vi. Mouraviev V, Nosnik I, Sun L, et al. Financial comparative analysis of minimally invasive surgery to open surgery for localized prostate cancer: a single-institution experience. Urology. 2007;69(2):311-314. Halliday D, Lau S, Vaknin Z, et al. Robotic radical hysterectomy: comparison of outcomes and cost. Journal of robotic surgery. 2010;4(4):211-216. Park BJ, Flores RM. Cost comparison of robotic, video-assisted thoracic surgery and thoracotomy approaches to pulmonary lobectomy. Thorac Surg Clin. 2008;18(3):297-300, vii. Barbash GI, Glied SA. New technology and health care costs--the case of robot-assisted surgery. The New England journal of medicine. 2010;363(8):701-704. Mues KE, Liede A, Liu J, et al. Use of the Medicare database in epidemiologic and health services research: a valuable source of real-world evidence on the older and disabled populations in the US. Clin Epidemiol. 2017;9:267-277.
17
39.
Yang HX, Woo KM, Sima CS, 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. Annals of surgery. 2017;265(2):431-437.
18
Table 1. Characteristics of Stage I-IIIA Non-Small Cell Lung Cancer Patients Treated with Robotic-Assisted Surgery versus VATS and Open Thoracotomy Robotic-Assisted Surgery (N=338) 73.0 (8.0) 190 (56.2) 190 (56.2)
Characteristics Age3, median (IQR4) Female3, N (%) Married, N (%) Race/Ethnicity, N (%) White 275 (81.4) Black ≥11 (≥5) Hispanic ≤11 (≤5) Other 36 (10.7) Median Household Annual Income Quartile, N (%) First (Lowest) 84 (25.1) Second 80 (23.9) Third 80 (23.9) Fourth (Highest) 91 (27.2) Comorbidity Score, N (%) 0-1 134 (39.6) 1-2 95 (28.1) >2 109 (32.3) Year of Diagnosis, N (%) 2008 or 2009 14 (4.1) 2010 28 (8.3) 2011 81 (24.0) 2012 112 (33.1) 2013 103 (30.5) Cancer Stage3 , 7th Edition AJCC, N (%) Stage I 241 (71.3) Stage II 62 (18.3) Stage IIIA 35 (10.4) Tumor Size, in mm, N (%) ≤ 20 134 (39.6) 21 to 30 86 (25.4) 31 to 50 93 (27.5) 51 to 70 ≥11 (≥10) ≥ 71 ≤11 (≤10) Histology, N (%) Adenocarcinoma 215 (63.6) Squamous Cell Carcinoma 93 (27.5) Large Cell Carcinoma ≤11 (≤10) Other Histology/NOS5 ≥11 (≥10)
VATS1 (N=1,230) 72.0 (7.0) 688 (55.9) 705 (57.3)
Open Thoracotomy (N=1,198) 73.0 (7.0) 672 (56.1) 672 (56.1)
1,080 (87.8) 65 (5.3) 29 (2.4) 56 (4.6)
1,037 (86.6) 80 (6.7) 37 (3.1) 44 (3.7)
274 (22.7) 287 (23.8) 262 (21.7) 384 (31.8)
413 (35.0) 290 (24.6) 240 (20.3) 238 (20.2)
463 (37.6) 323 (26.3) 444 (36.1)
365 (30.5) 341 (28.5) 492 (41.1)
49 (4.0) 113 (9.2) 291 (23.7) 384 (31.2) 393 (32.0)
56 (4.7) 107 (8.9) 276 (23.0) 409 (34.1) 350 (29.2)
918 (74.6) 214 (17.4) 98 (8.0)
P-value Adjusted2 Unadjusted <0.001 0.39 0.99 0.24 0.82 0.48
<0.001
0.75
<0.001
>0.99
<0.001
0.95
0.84
0.58
879 (73.4) 215 (18.0) 104 (8.7)
0.66
0.98
520 (42.3) 377 (30.7) 238 (19.4) 62 (5.0) 33 (2.7)
370 (30.9) 342 (28.6) 351 (29.3) 85 (7.1) 50 (4.2)
<0.001
0.75
812 (66.0) 295 (24.0) 25 (2.0) 98 (8)
706 (58.9) 383 (32.0) 23 (1.9) 86 (6.9)
0.006
0.97
19
Tumor Site, N (%) Upper Lobe 193 (54.1) 676 (55.0) 675 (56.3) Middle Lobe ≥11 (≥5) 106 (8.6) 67 (5.6) 0.064 >0.99 Lower Lobe 122 (36.1) 432 (35.1) 431 (36.0) Other Site ≤11 (≤10) 16 (1.3) 25 (2.1) PET6 Scan, N (%) 259 (76.6) 847 (68.9) 945 (78.9) <0.001 0.64 Chest CT7, N (%) 222 (65.7) 793 (65.5) 813 (67.9) 0.21 0.94 Mediastinoscopy, N (%) 60 (17.8) 129 (10.5) 116 (9.7) <0.001 0.73 Surgery Type3, N (%) Limited Resection 26 (7.7) 103 (8.4) 76 (6.3) 0.16 0.08 Lobectomy 312 (92.3) 1,127 (91.6) 1,122 (93.7) 1 Video-Assisted Thoracic Surgery; 2Adjusted for propensity scores; 3Matching characteristics used to form cohorts; 4 Interquartile Range; 5 Not Otherwise Specified; 6 Positron Emission Tomography; 7Computer Tomography; Cells with <11 patients were masked to maintain patient confidentiality.
20
Table 2: Postoperative Complication Rates of Robotic-Assisted Surgery Compared to Video-Assisted Thoracic Surgery and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients Entire Cohort
Limited to Stage I-II
Limited to Years Limited to Years 2008 – 2011 2012 - 2013 RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS1 VATS2 Open3 VATS VATS VATS VATS Open Open Open Open Outcome P-value N (%) N (%) N (%) OR4 OR OR OR OR OR OR OR 5 (95% CI ) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) Any 251 888 1000 <0.001 1.02 1.00 0.78 1.17 0.95 0.57 0.60 0.49 Complication (74.3) (72.2) (83.5) (0.76 - 1.37) (0.74 - 1.36) (0.45 - 1.35) (0.70 - 1.96) (0.66 - 1.36) (0.42 - 0.79) (0.43 - 0.83) (0.33 - 0.73) Respiratory 113 362 464 <0.001 0.89 1.18 0.87 1.13 1.01 1.30 0.82 1.11 Complication (33.4) (29.4) (38.7) (0.68 - 1.18) (0.89 - 1.56) (0.65 - 1.17) (0.84 - 1.51) (0.63 - 1.64) (0.84 - 2.00) (0.58 - 1.16) (0.77 - 1.61) Extrapulmonary 17 54 104 <0.001 0.72 1.31 0.81 1.35 0.63 1.02 0.77 1.53 Infection (5.0) (4.4) (8.7) (0.40 - 1.29) (0.71 - 2.40) (0.45 - 1.46) (0.73 - 2.49) (0.23 - 1.70) (0.40 - 2.57) (0.37 - 1.60) (0.69 - 3.42) Cardiovascular ≤11 29 28 0.79 1.21 1.05 1.12 0.84 1.46 0.78 1.10 1.30 Complication (≤3) (2.4) (2.3) (0.53 - 2.75) (0.47 - 2.33) (0.43 - 2.89) (0.34 - 2.08) (0.31 - 6.86) (0.18 - 3.44) (0.42 - 2.89) (0.52 - 3.25) Thromboembolic 17 63 60 0.99 1.00 1.03 1.00 0.97 0.86 1.11 1.14 0.97 Complication (5.0) (5.1) (5.0) (0.55 - 1.82) (0.56 - 1.88) (0.53 - 1.90) (0.51 - 1.84) (0.34 - 2.21) (0.41 - 3.00) (0.53 - 2.49) (0.46 - 2.05) Blood 12 63 125 <0.001 0.59 0.60 0.57 0.60 0.31 0.33 0.28 0.34 Transfusion (3.6) (5.1) (10.4) (0.30 - 1.14) (0.30 - 1.18) (0.20 - 1.66) (0.26 - 1.42) (0.17 - 0.60) (0.17 - 0.64) (0.10 - 0.83) (0.15 - 0.75) Reoperation ≤11 14 27 0.09 0.67 1.55 0.73 1.55 1.02 1.87 0.56 1.44 (≤3) (1.1) (2.3) (0.24 - 1.85) (0.52 - 4.63) (0.26 - 2.05) (0.52 - 4.62) (0.20 - 5.30) (0.33 - 10.7) (0.15 - 2.06) (0.35 - 5.95) Prolonged LOS6 21 72 147 <0.001 1.13 0.64 1.10 0.65 0.82 0.57 1.40 0.59 (6.3) (6.0) (12.4) (0.66 - 1.92) (0.37 - 1.10) (0.63 - 1.91) (0.26 - 1.58) (0.36 - 1.90) (0.30 - 1.07) (0.70 - 2.80) (0.35 - 0.98) 7 ICU Stay 208 740 871 <0.001 0.96 0.95 0.81 1.29 1.23 0.58 0.60 0.49 (0.74 - 1.26) (0.71 - 1.26) (0.49 - 1.33) (0.82 - 2.00) (0.58 - 1.14) (61.9) (61.2) (73.6) (0.44 - 0.77) (0.45 - 0.81) (0.35 - 0.68) Readmission 26 100 112 0.47 0.76 0.87 0.76 0.87 0.91 0.82 0.67 0.92 (7.7) (8.3) (9.5) (0.48 - 1.21) (0.54 - 1.43) (0.47 - 1.24) (0.52 - 1.46) (0.39 - 2.11) (0.36 - 1.87) (0.39 - 1.16) (0.51 - 1.69) Postoperative ≤11 17 35 0.01 0.46 0.76 0.51 0.78 0.18 0.35 0.62 0.96 Mortality (≤3) (1.4) (2.9) (0.16 - 1.34) (0.24 - 2.38) (0.17 - 1.51) (0.25 - 2.48) (0.03 - 1.35) (0.04 - 3.23) (0.18 - 2.11) (0.26 - 3.59) 1 Robotic-Assisted Surgery; 2Video-Assisted Thoracic Surgery; 3Open Thoracotomy; 4Odds Ratio (adjusted); 5Confidence Interval; 6Length of Stay; 7Intensive Care Unit; Cells with <11 patients were masked to maintain patient confidentiality.
21
Table 3: Oncologic Outcomes of Robotic-Assisted Surgery Compared to VATS and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients Outcome Lymph Node Staging (>10 LNs Examined) OR3 (95% CI4) Survival HR5 (95% CI) Overall survival
Cohort Entire Cohort: Stage I-IIIA Limited to Stage I-II
Robotic-Assisted Surgery vs. Open Thoracotomy Adjusted2 Unadjusted 1.36 (1.03 - 1.80) 1.28 (0.94 - 1.74) 1.35 (0.98 - 1.86) 1.45 (1.08 - 1.94)
Robotic-Assisted Surgery vs. VATS1 Unadjusted 0.97 (0.75 - 1.26) 1.02 (0.78 - 1.33)
Adjusted2 0.88 (0.66 - 1.18) 0.93 (0.69 - 1.26)
Entire Cohort: Stage I-IIIA 0.79 (0.64 - 0.97) 0.81 (0.63 - 1.04) 1.05 (0.83 - 1.32) 0.91 (0.70 - 1.18) Limited to Stage I-II 0.79 (0.59 – 1.05) 0.95 (0.73 – 1.23) 0.87 (0.65 – 1.17) 0.72 (0.56 – 0.92) Lung cancer-specific survival Entire Cohort: Stage I-IIIA 0.79 (0.56 - 1.13) 0.75 (0.51 - 1.12) 1.13 (0.77 - 1.66) 0.87 (0.56 - 1.34) Limited to Stage I-II 0.67 (0.41 – 1.10) 0.91 (0.56 – 1.50) 0.78 (0.45 – 1.32) 0.61 (0.39 – 0.96) 1 Video-Assisted Thoracic Surgery; 2Adjusted for propensity scores and surgical volume; 3Odds Ratio; 4Confidence Interval: adjusted for clustering by surgeons due to matching; 5Hazard Ratio
22
Figure 1: Propensity Score Adjusted Costs of Robotic-Assisted Surgery Compared to VATS and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients According to Phase of Care B: p=0.08
$60,000.00
A: p=0.02
A: RAS vs VATS B: RAS vs Open
$50,000.00
B: p<0.001
$40,000.00
A: p=0.08
B: p=0.95
$30,000.00
Robotic VATS
A: p=0.15
OPEN $20,000.00 B: p=0.12
$10,000.00
A: p<0.001
$0.00 Total Costs
Pre-operative
Operative
Post-operative
e-Appendix 1. Methods As discussed in the manuscript, study patients were identified using the SEER-Medicare database, a large population-based source of longitudinal data that includes demographic and clinical information for Medicare beneficiaries with cancer in various United States (US) regions.(E1) The SEER program has been expanded to cover 28% of the US population, with 94% of eligible patients linked to Medicare. Using SEER data, we obtained each patient’s sociodemographic information, including age, gender, marital status, race, and ethnicity. Medicare data allowed us to group patients into quartiles based on socioeconomic status according to census tract level information. Comorbidities were assessed using Medicare data by applying a modified version of the Charlson index and lung cancer-specific weights.(E2,E3) Claims indicating home services (limited to homebound patients) were used as a proxy for poor performance status.(E4) Each patient’s cancer was staged according to the American Joint Commission on Cancer (seventh edition) using SEER data providing tumor size, extension, location and lymph node status. Cancer histology and grade were determined according to the International Classification of Diseases for Oncology, Third Edition, and morphology codes available in SEER. Information about tests performed for cancer diagnosis (e.g., fine needle aspiration, bronchoscopy), preoperative evaluation (e.g., ventilation/perfusion scan, cardiac/cardiopulmonary stress testing), and staging work-up (e.g., positron emission tomography [PET], mediastinoscopy) was ascertained using International Classification of Diseases, 9 th revision diagnostic and procedure codes (ICD-9) and Current Procedural Terminology-4 (CPT-4) codes from Medicare claims.(E5, E6) The SEER registry routinely collects treatment data for 4 months after diagnosis and, therefore, this time period was used to assess the initial course of therapy. ICD-9 procedure and CPT-4 codes submitted as part of Medicare claims were used to classify study subjects as having undergone robotic-assisted surgery (ICD-9: 17.45 and CPT-4: S2900), VATS (ICD-9: 34.21 and CPT-4: 32657), or open thoracotomy (ICD-9: 34.01 and CPT-4: 32100 ). We focused on patients who underwent lobectomy (SEER codes 30, 31, and 33 and ICD-9 code 32.4) or limited resection (SEER codes 20 to 22 and ICD-9 code 32.29 and 32.3), as these procedures can be performed via these surgical approaches. Physician-specific surgical case volume, defined as the volume of resections performed by a patient’s surgeon in the year of the patient’s surgery, was ascertained from Medicare billing claims. Using this data, surgeons were ranked according to their average volume of claims for lung cancer resections.(E7,E8) Patients who received adjuvant treatments (i.e., post-operative RT and chemotherapy) were identified using both SEER and Medicare information using validated algorithms.(E9,E10) Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.
Consistent with prior studies, postoperative morbidity and mortality (primary outcomes) were defined as the presence of complications and death, respectively, within 30 days of surgery or during the hospitalization in which the primary surgical procedure was performed. We also used prolonged length of stay (>14 days), obtained from MedPAR files, as an indicator of postoperative outcomes. We identified patients requiring intensive care unit (ICU) admission following surgery based on information provided as part of Medicare inpatient files. Sample size calculations determined that the study included the necessary number of patients to achieve 80% power to identify statistically significant differences (α level 0.05) in the rate of postoperative complications among patients treated with robotic-assisted surgery vs. VATS and open surgery (OR of 1.3 to 2.0). Survival analyses were exploratory in nature given that robotic-assisted surgery is a relatively new procedure for lung cancer resection without long-term follow up data. All analyses were conducted with SAS statistical software (SAS, version 9.4; Cary, NC). The Icahn School of Medicine at Mount Sinai Institutional Review Board considered the study exempt.
Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.
References E1. Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Medical care 2002; 40: Iv-3-18. E2. Wisnivesky JP, Henschke CI, Swanson S, Yankelevitz DF, Zulueta J, Marcus S, Halm EA. Limited resection for the treatment of patients with stage IA lung cancer. Annals of surgery 2010; 251: 550-554. E3. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. Journal of clinical epidemiology 2000; 53: 1258-1267. E4. Wisnivesky JP, Smith CB, Packer S, Strauss GM, Lurslurchachai L, Federman A, Halm EA. Survival and risk of adverse events in older patients receiving postoperative adjuvant chemotherapy for resected stages II-IIIA lung cancer: observational cohort study. BMJ (Clinical research ed) 2011; 343: d4013. E5. Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996-1998 practice patterns. Radiology 2002; 222: 144-148. E6. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-small-cell lung cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2006; 24: 413-418. E7. Earle CC, Schrag D, Neville BA, Yabroff KR, Topor M, Fahey A, Trimble EL, Bodurka DC, Bristow RE, Carney M, Warren JL. Effect of surgeon specialty on processes of care and outcomes for ovarian cancer patients. Journal of the National Cancer Institute 2006; 98: 172-180. E8. Schrag D, Earle C, Xu F, Panageas KS, Yabroff KR, Bristow RE, Trimble EL, Warren JL. Associations between hospital and surgeon procedure volumes and patient outcomes after ovarian cancer resection. Journal of the National Cancer Institute 2006; 98: 163-171. E9. Warren JL, Harlan LC, Fahey A, Virnig BA, Freeman JL, Klabunde CN, Cooper GS, Knopf KB. Utility of the SEER-Medicare data to identify chemotherapy use. Medical care 2002; 40: Iv-55-61. E10. Virnig BA, Warren JL, Cooper GS, Klabunde CN, Schussler N, Freeman J. Studying radiation therapy using SEER-Medicare-linked data. Medical care 2002; 40: Iv-49-54.
182739
Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.
S0012-3692(19)33946-7
DOI:
https://doi.org/10.1016/j.chest.2019.09.017
Reference:
CHEST 2646
To appear in:
CHEST
Received Date: 29 November 2018 Revised Date:
29 August 2019
Accepted Date: 14 September 2019
Please cite this article as: Veluswamy RR, Whittaker Brown SA, Mhango G, Sigel K, Nicastri DG, Smith CB, Bonomi M, Galsky MD, Taioli E, Neugut AI, Wisnivesky JP, Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients CHEST (2019), doi: https:// doi.org/10.1016/j.chest.2019.09.017. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Copyright © 2019 Published by Elsevier Inc under license from the American College of Chest Physicians.
Comparative Effectiveness of Robotic-Assisted Surgery for Resectable Lung Cancer in Older Patients Rajwanth R. Veluswamy, MD, MSCR,1,2 Stacey-Ann Whittaker Brown, MD,3 Grace Mhango, MPH,4 Keith Sigel, MD, PhD,4 Daniel G. Nicastri, MD,5 Cardinale B. Smith, MD, PhD,1,6 Marcelo Bonomi, MD,7 Matthew D. Galsky, MD,1 Emanuela Taioli, MD, PhD,2,8 Alfred I. Neugut, MD, PhD,9,10 Juan P. Wisnivesky, MD, DrPH3,4 1
Division of Hematology/Oncology, Icahn School of Medicine at Mount Sinai, New York, NY; Institute of Translational Epidemiology, Icahn School of Medicine at Mount Sinai, New York, NY; 3Division of Pulmonary, Critical Care, and Sleep Medicine, Icahn School of Medicine at Mount Sinai, New York, NY; 4Division of General Internal Medicine, Icahn School of Medicine at Mount Sinai, New York, NY; 5Department of Thoracic Surgery, Icahn School of Medicine at Mounts Sinai, New York, NY; 6Hertzberg Palliative Care Institute of the Brookdale Department of Geriatrics, Icahn School of Medicine at Mount Sinai, New York, NY; 7Section of Hematology and Medical Oncology, Department of Internal Medicine, Wake Forest School of Medicine, Winston-Salem, NC; 8Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai, New York, NY; 9Division of Hematology/Oncology, Department of Medicine, Columbia University, New York, NY; and 10Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY 2
Corresponding Author:
Rajwanth Veluswamy, MD Division of Hematology and Medical Oncology Icahn School of Medicine at Mount Sinai One Gustave L. Levy Place, Box 1079 New York, NY 10029-6574 Phone: (212) 824-8580 Fax: (646) 537-9639 E-mail: [email protected]
Running Title: Robotic-Assisted Surgery for Older NSCLC Patients COI disclosure: RRV has received consulting honorarium from Onconova Therapeutics and Astrazeneca. JPW is a member of the research board of EHE International, has received consulting honorarium from Merck, Quintiles and Astrazeneca and research grants from Sanofi and Quorum. AIN is a member of the research board of EHE International and has received consulting honoraria from Otsuka, Hospira, and United Biosource Corporation. RRV, SAW, GM, KS, DGN, CBS, MB, MDG and ET have no conflict of interest to declare. Financial support: This study was supported by the Conquer Cancer Foundation of the American Society of Clinical Oncology Young Investigator Award and the ISMMS Clinical and Translational Science Award (CTSA). Word count: Abstract: 245; Manuscript text: 2500 (11 pages); Tables: 3; Figures: 1 1
Keywords: NSCLC, Robotic Surgery, Minimally Invasive, Early Stage, Treatment Abbreviations List: CI: confidence interval GLLM: generalized linear mixed model HR: hazard ratio ICU: intensive care unit LOS: length of stay NSCLC: non-small cell lung cancer OR: odds ratio PET: positron emission tomography RAS: robotic-assisted surgery SEER: Surveillance, Epidemiology, and End Results US: United States VATS: video-assisted thoracic surgery
2
Abstract
Background: Robotic-assisted surgery (RAS) is a novel surgical approach increasingly used for patients with non-small cell lung cancer (NSCLC). However, data comparing the effectiveness and costs of RAS vs. open thoracotomy and video-assisted thoracic surgery (VATS) for NSCLC is limited. Methods: Patients >65 years old with stage I-IIIA NSCLC treated with RAS, VATS or open thoracotomy were identified from the Surveillance, Epidemiology, and End Results-Medicare database and matched by age, sex, stage, and extent of resection. We used propensity score methods to compare adjusted rates of postoperative complications, adequate lymph node staging, survival, and treatment-related costs. Results: In our matched study cohort of 2, 766 resected NSCLC patients, RAS was associated with lower complication rates (odds ratio [OR]: 0.57; 95% confidence interval [CI]: 0.42-0.79) compared to open thoracotomy, and similar complication rates (OR: 1.02; 95% CI: 0.76-1.37) compared to VATS. RAS patients were as likely to have adequate lymph node sampling as those undergoing open thoracotomy (OR: 1.28; 95% CI: 0.94-1.74) or VATS (OR: 0.88; 95% CI: 0.66-1.18). There was no significant difference in overall survival after RAS vs. open thoracotomy (hazard ratio [HR]: 0.81, 95% CI: 0.63-1.04) or VATS (HR: 0.91; 95% CI: 0.701.18). Costs were similar for RAS ($54,702) vs. open thoracotomy ($57,104; p=0.08), and higher compared to VATS ($48,729; p=0.02). Conclusions: RAS leads to improved operative outcomes compared to open thoracotomy but may not offer an advantage over VATS. The comparative effectiveness of RAS should be further evaluated prior to widespread adoption.
3
Introduction
Optimal surgical management of non-small cell lung cancer (NSCLC) is critical for maximizing chances of cure and achieving good long-term outcomes. These treatment decisions are becoming even more important given the expected increase in the proportion of tumors diagnosed at earlier stages following full implementation of lung cancer screening recommendations.1 However, with a mean age at diagnosis of 70 years, most NSCLC patients are older and frequently have smoking-related comorbidities, creating challenges for resection.2 Thus, the identification of effective and well-tolerated surgical approaches has been a major area of clinical interest.
The standard surgical approach for resectable stage I-IIIA NSCLC has been lobectomy via open thoracotomy.3 However, open lobectomy is associated with considerable postoperative complications and mortality, especially in the elderly.4,5 As a consequence, minimally invasive techniques, such as video-assisted thoracoscopic surgery (VATS), have been increasingly adopted.3 While there have been no randomized controlled trials, several retrospective studies have shown that VATS lobectomy is associated with fewer postoperative complications compared to open lobectomy.6-8 Additionally, the total costs for lobectomy have been shown to be lower when using a VATS approach.9
More recently, surgical approaches using robotic systems have been designed to enhance instrument maneuverability and visualization, enabling surgeons to perform intricate maneuvers with greater precision. Prior case series have demonstrated acceptable perioperative results, low thoracotomy conversion rates, and comparable survival among NSCLC patients treated with 4
robotic-assisted surgery (RAS).10-12 However, there is limited data regarding the comparative effectiveness of RAS vs. VATS or open resection for older NSCLC patients, a group at high risk for poor outcomes. Furthermore, the financial cost of RAS has not been adequately described. As a consequence, RAS is being adopted into routine care without adequate scrutiny.
In this study, we used population-based data from the Surveillance, Epidemiology, and End Results (SEER)-Medicare database to assess postoperative, oncologic and cost outcomes of older patients with stage I-IIIA NSCLC treated with RAS vs. VATS or open thoracotomy.
Methods
Study Variables
We included all patients >65 years from SEER-Medicare with stage I-IIIA NSCLC who were diagnosed between 2008 (first year of RAS in registry) and 2013, and treated with surgical resection via RAS, open thoracotomy or VATS.13 In order to create a well-balanced cohort, we matched RAS-treated patients with up to four patients of the same age (within 5 years), sex, stage, and extent of surgery (wedge, segmentectomy or lobectomy) who underwent open thoracotomy or VATS. We excluded patients treated with neoadjuvant treatment (may represent higher stage), in health maintenance organizations or without outpatient Medicare coverage due to lack of claims data.
We obtained pretreatment information, including sociodemographics, comorbidities, tumor characteristics and stage, and diagnostic, staging, and preoperative evaluation, from both SEER 5
and Medicare data.14-17 Using procedure codes from Medicare claims, we classified study subjects as having undergone RAS, VATS or open thoracotomy. We ranked surgeons according to the volume of total resections they performed in the year of the patient’s surgery.18,19 See online appendix for more details.
Postoperative morbidity and mortality (primary outcomes) were defined as the presence of complications and death, respectively, within 30 days of surgery or during the hospitalization in which surgery was performed. Complications were determined from Medicare claims and included: 1) extrapulmonary infections; 2) cardiovascular complications; 3) thromboembolic events; 4) respiratory complications; 5) reoperations; 6) readmissions; 7) prolonged length of stay (>14 days; LOS); and 8) intensive care unit (ICU) management. Postoperative mortality was determined using the dates of surgery and death, obtained from Medicare files.
From SEER data, we compared the extent of lymph node dissection (≤10 vs. >10 nodes; cutoff recommended for adequate staging) with each surgical approach as a secondary outcome.20 Survival was calculated from the date of surgery to the date of death or last follow-up (December 31, 2015 for censored observations) according to Medicare data. For analyses involving lungcancer specific survival, we used SEER to ascertain the cause of death from state death certificates.
Treatment-related costs were estimated from the payer perspective using Medicare reimbursement data from inpatient, outpatient, physician, durable medical equipment, and home health agency files.21 This costs data does not include capital and maintenance costs for surgical
6
equipment incurred by the institution. We categorized costs as: 1) pre-operative (within thirty days prior to surgery); 2) operative (within 30 days of surgery); and 3) post-operative (from postoperation day 31 through 365).22 Total cost was calculated as the sum of the costs of each phase of treatment and standardized to 2012 United States (US) dollars using the Consumer Price Index.23
Statistical Analysis
We compared the distribution of pretreatment characteristics of patients in each study group using analysis of variance or the chi-square test. Given the lack of randomization, we applied propensity score methods to minimize the effects of measured confounders on treatment allocation. We used nominal regression to calculate propensity scores indicating the probability of patients undergoing RAS, open thoracotomy, or VATS based on pretreatment information.
We compared adjusted rates of postoperative complications and mortality, as well as the number of lymph nodes assessed, among patients treated with open thoracotomy, VATS or RAS with a generalized linear mixed model (GLMM; adjusting confidence intervals for clustering due to matching). We adjusted for covariates using inverse probability weighting; weights were calculated as the inverse of the estimated propensity score for undergoing the type of surgical treatment each patient received. All models were also adjusted for surgeon’s case volume. We conducted secondary analyses: 1) limiting the cohort to patients with stage I-II NSCLCs as stage IIIA tumors may not be amenable to primary surgical resection; and 2) stratifying patients who underwent RAS during the early (2008-2011) and later (2012 and 2013) years of uptake to
7
determine if operator experience improved outcomes. We conducted adjusted survival comparisons using Cox regression analysis with inverse probability weighting by propensity scores and robust standard errors to incorporate clustering. Finally, we compared costs using a GLMM with an identity link and γ-distributed variance function.
Results
Of the 2,766 stage I-IIIA NSCLC patients included in the study, 338 (12%) underwent RAS, 1230 (45%) received VATS, and 1198 (43%) were managed with open thoracotomy. There were significant differences in the distribution of age (p<0.001), race (p<0.001), household income (p<0.001), comorbidities (p<0.001), tumor size (p<0.001), histology (p=0.006) and preoperative assessment with positron emission tomography (PET; p<0.001) or mediastinoscopy (p<0.001) across the three treatment groups. After adjusting for propensity scores, all measured variables were well balanced (Table 1).
Propensity score adjusted analysis showed that overall surgical complication rates were lower for patients undergoing RAS compared to those treated with open thoracotomy (odds ratio [OR]: 0.57; 95% confidence interval [CI]: 0.42-0.79). Specifically, rates of blood transfusions (OR: 0.31; 95% CI: 0.17-0.60), ICU stays (OR: 0.58; 95% CI: 0.44-0.77) and prolonged LOS (OR: 0.59; 95% CI: 0.35-0.98) were significantly lower in patients undergoing RAS (Table 2). There was no statistical difference in adequate lymph node staging (OR: 1.28; 95% CI: 0.94-1.74) and exploratory survival analyses (hazard ratio [HR]: 0.81, 95% CI: 0.63-1.04; HR: 0.75, 95% CI: 0.51-1.12 for overall and lung cancer-specific survival, respectively) when comparing RAS to
8
open thoracotomy (Table 3). Finally, while total costs for patients who underwent RAS were similar than for those who underwent open thoracotomy ($54,702 vs. $57,104; p=0.08), costs during the operative period were significantly lower for RAS ($28,732 vs. $32,746; p<0.0001). Pre-operative ($3,668 vs. $3,199; p=0.12) and post-operative ($22,302 vs. $21,158; p=0.95) costs were similar between RAS and open thoracotomy (Figure 1).
When compared to patients who underwent VATS, those treated with RAS had similar overall complication rates related to surgery (OR: 1.02; 95% CI: 0.76-1.37; p>0.05 for individual complications; Table 2). Both approaches had similar odds of undergoing adequate lymph node evaluation (OR: 0.88; 95% CI: 0.66-1.18) and did not have statistically different survival (HR: 0.91; 95% CI: 0.70-1.18; HR: 0.87; 95% CI: 0.56-1.34 for overall and lung cancer-specific survival, respectively; Table 3). RAS patients incurred higher total costs ($54,702 vs. $48,729; p=0.02) and pre-operative costs ($3,668 vs. $2,803; p<0.0001) compared to patients undergoing VATS. Costs during the operative ($28,732 vs. $27,209; p=0.078) and post-operative ($22,302 vs. $18,718; p=0.15) periods were similar between the two minimally invasive procedures (Figure 1).
Secondary analyses limiting the cohort to stage I-II NSCLC patients (N=2529) showed similar postoperative and oncological outcomes as those observed in the analysis of the entire cohort, with the exception that prolonged LOS rates were now similar in all three surgical groups (Tables 2 and 3). When limited to the initial years of uptake (2008-2011; N=123), RAS patients had lower rates of blood transfusions (OR: 0.28, 95% CI: 0.10-0.83) compared to open thoracotomy; otherwise, they experienced similar rates of complications as patients who
9
underwent open thoracotomy (OR: 0.78; 95% CI: 0.45-1.35) and VATS (OR: 1.17; 95% CI: 0.70-1.96). Complication rates in RAS patients, however, improved in the more recent years of uptake (2012-2013; N=215) when compared to those who received open thoracotomy, with lower rates of: total complications (OR: 0.49; 95% CI: 0.33-0.73), blood transfusions (OR: 0.34; 95% CI: 0.15-0.75) and ICU stays (OR: 0.49; 95% CI: 0.35-0.68). RAS continued to be associated with similar rates of complications compared to VATS (OR: 0.95; 95% CI: 0.661.36).
Discussion
RAS has rapidly expanded into the routine practice of various surgical specialties, leading to almost a 200% increase in utilization between 1999 and 2014.28 Currently, it is estimated that one in four US hospitals have a robotic system.24 However, high quality evidence is still needed to determine the comparative effectiveness and costs of RAS in NSCLC resections.25 In this population-based study of older patients, we found RAS was associated with reduced rate of complications when compared to open thoracotomy, and similar complication rates when compared to VATS. We further found that the improvement of postoperative outcomes of RAS vs. open thoracotomy occurred in the later years of uptake. While RAS was associated with similar total costs than open thoracotomy, we found significantly higher costs when compared to VATS. Therefore, it remains unclear if RAS offers any benefit beyond VATS.
Data evaluating postoperative outcomes of RAS for NSCLC mostly comes from smaller case series of single institutions.10-12 More recently, comparative effectiveness studies using various databases have reported conflicting results. Data from the Premier Hospital Network found no 10
difference between RAS and VATS in overall complication rates, LOS, and surgical mortality, but the former was associated with higher costs and longer operating times.26 Conversely, analysis of the National Inpatient Sample found that patients undergoing RAS compared to VATS were associated with higher rates of intraoperative bleeding complications and were more likely to be discharged to facilities rather than home.27 However, these studies did not adjust for important imbalances in tumor characteristics, nor account for surgeon case volume.28,29 Our study used nationally representative data and rigorously adjusted for all measured patient, tumor, and surgery characteristics. We additionally focused on older patients, the age group most commonly affected by NSCLC and more vulnerable to experiencing surgical complications.
Interestingly, we found that RAS was associated with lower postoperative complications when compared to open thoracotomy only in the later study years (2012-2013), suggesting that postoperative outcomes improved with increasing experience and/or better patient selection with this technique. Improved operative outcomes for RAS were also found in an analysis of the State Inpatient Database, using a study cohort where the majority of resections (>75%) were performed by high-volume surgeons.30 There is a steep learning curve for surgeons to become proficient with RAS.31 New generations of robotic systems (i.e. Xi platform) are being developed which may improve user experience and clinical outcomes. In addition, widespread, standardized training of RAS using virtual training modules through the robotic console may allow surgeons to gain proficiency in robotic skills and reduce complication rates.
The oncologic outcomes of RAS for NSCLC have also not yet been adequately described. Prior case series have demonstrated intraoperative mediastinal lymph node staging to be similar with
11
both VATS and RAS.20,32 Additionally, the 5-year survival of NSCLC patients undergoing robotic lobectomy was similar to those historically reported for VATS or open thoracotomy. 33 However, these studies lack a comparison arm, have potential biases in patient selection, and have limited generalizability. Our findings, accounting for these limitations, provide stronger level of evidence confirming these conclusions. Of note, our results are an overall average treatment effect after adjusting for important clinical confounders. We are unable to determine potential differing treatment effects in individual patient subgroups, as this would significantly reduce the power of our study. Future studies can be undertaken to explore the role of each surgical approach in specific subgroups.
The implementation of new technology must consider the relative costs compared to existing interventions. We found RAS to be associated with similar total costs compared with open thoracotomy, but lower costs during the operative period. These results are consistent with prior studies evaluating prostatectomy and hysterectomy.34,35 When compared to VATS, RAS was associated with significantly higher total costs, specifically those incurred in the pre-operative period. Of note, our findings do not take into account expenses associated with robotic surgical systems that are incurred by institutions, such as costs for the robot itself ($1-$2.5 million), maintenance ($80-$170 thousand annually) and additional consumables ($700-$3,500 per robotic case).24,36,37 Medicare claims data instead provides reimbursements paid (i.e., from the payer perspective) for all services rendered to beneficiaries. Over 55 million Americans (17% of US population) rely on Medicare as their primary insurance coverage; 80% of beneficiaries are age 65 or older. Therefore, Medicare data provides highly representative information for evaluating costs of medical interventions in the US.38
12
There are some limitations of the present study. Lack of random treatment allocation can result in systematic differences in the distribution of pretreatment characteristics between treatment groups, which could affect both the type of treatment a patient receives and their clinical outcomes. To mitigate this potential bias, we matched patients for key factors and applied propensity score methods to balance groups for all other measured confounders. However, we were unable to account for unmeasured confounders, a particular limitation inherent to all observational studies. Additionally, we did not have data on completeness of resection (i.e., R0 vs. R1 vs. R2 resection) or conversion rates for minimally invasive procedures to open thoracotomy. Conversions may unfavorably bias both operative and costs outcomes for open thoracotomy. However, a large, single institution case series previously reported rates of incomplete resections and conversions are low and similar between RAS and VATS.39 Finally, we were not able enumerate all costs related to each surgical approach, and therefore future research should explore insights into mechanisms of costs difference between each procedure.
Conclusions
Our population-based study of older NSCLC patients found that RAS offers improved postoperative outcomes compared to open thoracotomy, but no such benefit was observed vs. VATS. Furthermore, while we found total costs were similar between RAS and open thorocatomy, RAS was associated with higher costs compared to VATS. Lymph node evaluation and exploratory survival rates were similar between all three approaches. While RAS is innovative, we found no distinct advantage over VATS to support its adoption into routine care
13
for the surgical management of patients with NSCLC. As technology and operator experience with RAS improve with time, continued reassessment of comparative effectiveness should be undertaken to ensure patients receive the most effective and safe surgical option.
14
Acknowledgments
The authors would like to thank the Conquer Cancer Foundation of the American Society of Clinical Oncology; the Applied Research Branch, Division of Cancer Prevention and Population Science, National Cancer Institute; the Office of Information Services and the Office of Strategic Planning, Health Care Finance Administration; Information Management Services (IMS), Inc.; the SEER Program tumor registries for their efforts in the creation of the SEER-Medicare Database; and the ISMMS Clinical and Translational Science Award (CTSA). The collection, interpretation and reporting of these data are the sole responsibilities of the authors. All significant contributors of this manuscript have been listed.
Author Contributions: Conception and design, Manuscript Writing, Final approval: All authors Collection, assembly, analysis and interpretation of data: RRV, SAW, GM, JPW
15
References 1. 2. 3. 4.
5.
6.
7.
8.
9.
10. 11. 12. 13.
14. 15. 16.
17. 18.
Moyer VA. Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2014;160(5):330-338. Howlader N NA, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2011. 2013; http://seer.cancer.gov/csr/1975_2011/. Accessed August, 2015. National Comprehensive Cancer Network. Non-Small Cell Lung Cancer (Version 4.2016). http://www.nccn.org/professionals/physician_gls/pdf/nscl.pdf. Accessed March 11. Scott WJ, Matteotti RS, Egleston BL, Oseni S, Flaherty JF. A comparison of perioperative outcomes of video-assisted thoracic surgical (VATS) lobectomy with open thoracotomy and lobectomy: results of an analysis using propensity score based weighting. Ann Surg Innov Res. 2010;4(1):1. Whitson BA, Groth SS, Duval SJ, Swanson SJ, Maddaus MA. Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy. Ann Thorac Surg. 2008;86(6):2008-2016; discussion 2016-2008. Yan TD, Black D, Bannon PG, McCaughan BC. Systematic review and meta-analysis of randomized and nonrandomized trials on safety and efficacy of video-assisted thoracic surgery lobectomy for early-stage non-small-cell lung cancer. J Clin Oncol. 2009;27(15):2553-2562. Swanson SJ, Herndon JE, 2nd, D'Amico TA, et al. Video-assisted thoracic surgery lobectomy: report of CALGB 39802--a prospective, multi-institution feasibility study. J Clin Oncol. 2007;25(31):4993-4997. Cajipe MD, Chu D, Bakaeen FG, et al. Video-assisted thoracoscopic lobectomy is associated with better perioperative outcomes than open lobectomy in a veteran population. Am J Surg. 2012;204(5):607-612. Swanson SJ, Meyers BF, Gunnarsson CL, et al. Video-assisted thoracoscopic lobectomy is less costly and morbid than open lobectomy: a retrospective multiinstitutional database analysis. Ann Thorac Surg. 2012;93(4):1027-1032. Toker A. Robotic thoracic surgery: from the perspectives of European chest surgeons. J Thorac Dis. 2014;6 Suppl 2:S211-216. Augustin F, Bodner J, Wykypiel H, Schwinghammer C, Schmid T. Initial experience with robotic lung lobectomy: report of two different approaches. Surg Endosc. 2011;25(1):108-113. Gharagozloo F, Margolis M, Tempesta B, Strother E, Najam F. Robot-assisted lobectomy for early-stage lung cancer: report of 100 consecutive cases. Ann Thorac Surg. 2009;88(2):380-384. Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40(8 Suppl):Iv-3-18. Wisnivesky JP, Henschke CI, Swanson S, et al. Limited resection for the treatment of patients with stage IA lung cancer. Annals of surgery. 2010;251(3):550-554. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 2000;53(12):1258-1267. Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996-1998 practice patterns. Radiology. 2002;222(1):144-148. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-smallcell lung cancer. J Clin Oncol. 2006;24(3):413-418. Earle CC, Schrag D, Neville BA, et al. Effect of surgeon specialty on processes of care and outcomes for ovarian cancer patients. J Natl Cancer Inst. 2006;98(3):172-180. 16
19. 20.
21. 22.
23. 24.
25. 26.
27. 28. 29. 30.
31. 32. 33. 34.
35. 36. 37. 38.
Schrag D, Earle C, Xu F, et al. Associations between hospital and surgeon procedure volumes and patient outcomes after ovarian cancer resection. J Natl Cancer Inst. 2006;98(3):163-171. Shapiro M, Mhango G, Kates M, et al. Extent of lymph node resection does not increase perioperative morbidity and mortality after surgery for stage I lung cancer in the elderly. Eur J Surg Oncol. 2012;38(6):516-522. Brown ML, Riley GF, Schussler N, Etzioni R. Estimating health care costs related to cancer treatment from SEER-Medicare data. Med Care. 2002;40(8 Suppl):Iv-104-117. Smith BD, Jiang J, Chang JY, et al. Cost-effectiveness of stereotactic radiation, sublobar resection, and lobectomy for early non-small cell lung cancers in older adults. J Geriatr Oncol. 2015;6(4):324-331. http://www2.census.gov/geo/docs/maps-data/maps/reg_div.txt CbradwsFcaFAf. Intuitive Surgical Investor Presentation Q4 2014. Available from: https://www.rts.ch/emissions/tempspresent/6617343.html/BINARY/La%20pr%C3%A9sentation%20d%E2%80%99Intuitive%20Surgic al. Swanson SJ. Robotic pulmonary lobectomy--the future and probably should remain so. J Thorac Cardiovasc Surg. 2010;140(5):954. Swanson SJ, Miller DL, McKenna RJ, Jr., et al. Comparing robot-assisted thoracic surgical lobectomy with conventional video-assisted thoracic surgical lobectomy and wedge resection: results from a multihospital database (Premier). J Thorac Cardiovasc Surg. 2014;147(3):929-937. Paul S, Jalbert J, Isaacs AJ, Altorki NK, Isom OW, Sedrakyan A. Comparative effectiveness of robotic-assisted vs thoracoscopic lobectomy. Chest. 2014;146(6):1505-1512. Irie M, Nakanishi R, Yasuda M, Fujino Y, Hamada K, Hyodo M. Risk factors for short-term outcomes after thoracoscopic lobectomy for lung cancer. Eur Respir J. 2016;48(2):495-503. Husain ZA, Kim AW, Yu JB, Decker RH, Corso CD. Defining the High-Risk Population for Mortality After Resection of Early Stage NSCLC. Clin Lung Cancer. 2015;16(6):e183-187. Kent M, Wang T, Whyte R, Curran T, Flores R, Gangadharan S. Open, video-assisted thoracic surgery, and robotic lobectomy: review of a national database. Ann Thorac Surg. 2014;97(1):236-242; discussion 242-234. Suda T. Transition from video-assisted thoracic surgery to robotic pulmonary surgery. The Journal of Visualized Surgery. 2017;3(4). Lee HS, Jang HJ. Thoracoscopic mediastinal lymph node dissection for lung cancer. Semin Thorac Cardiovasc Surg. 2012;24(2):131-141. Park BJ. Robotic lobectomy for non-small cell lung cancer: long-term oncologic results. Thorac Surg Clin. 2014;24(2):157-162, vi. Mouraviev V, Nosnik I, Sun L, et al. Financial comparative analysis of minimally invasive surgery to open surgery for localized prostate cancer: a single-institution experience. Urology. 2007;69(2):311-314. Halliday D, Lau S, Vaknin Z, et al. Robotic radical hysterectomy: comparison of outcomes and cost. Journal of robotic surgery. 2010;4(4):211-216. Park BJ, Flores RM. Cost comparison of robotic, video-assisted thoracic surgery and thoracotomy approaches to pulmonary lobectomy. Thorac Surg Clin. 2008;18(3):297-300, vii. Barbash GI, Glied SA. New technology and health care costs--the case of robot-assisted surgery. The New England journal of medicine. 2010;363(8):701-704. Mues KE, Liede A, Liu J, et al. Use of the Medicare database in epidemiologic and health services research: a valuable source of real-world evidence on the older and disabled populations in the US. Clin Epidemiol. 2017;9:267-277.
17
39.
Yang HX, Woo KM, Sima CS, 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. Annals of surgery. 2017;265(2):431-437.
18
Table 1. Characteristics of Stage I-IIIA Non-Small Cell Lung Cancer Patients Treated with Robotic-Assisted Surgery versus VATS and Open Thoracotomy Robotic-Assisted Surgery (N=338) 73.0 (8.0) 190 (56.2) 190 (56.2)
Characteristics Age3, median (IQR4) Female3, N (%) Married, N (%) Race/Ethnicity, N (%) White 275 (81.4) Black ≥11 (≥5) Hispanic ≤11 (≤5) Other 36 (10.7) Median Household Annual Income Quartile, N (%) First (Lowest) 84 (25.1) Second 80 (23.9) Third 80 (23.9) Fourth (Highest) 91 (27.2) Comorbidity Score, N (%) 0-1 134 (39.6) 1-2 95 (28.1) >2 109 (32.3) Year of Diagnosis, N (%) 2008 or 2009 14 (4.1) 2010 28 (8.3) 2011 81 (24.0) 2012 112 (33.1) 2013 103 (30.5) Cancer Stage3 , 7th Edition AJCC, N (%) Stage I 241 (71.3) Stage II 62 (18.3) Stage IIIA 35 (10.4) Tumor Size, in mm, N (%) ≤ 20 134 (39.6) 21 to 30 86 (25.4) 31 to 50 93 (27.5) 51 to 70 ≥11 (≥10) ≥ 71 ≤11 (≤10) Histology, N (%) Adenocarcinoma 215 (63.6) Squamous Cell Carcinoma 93 (27.5) Large Cell Carcinoma ≤11 (≤10) Other Histology/NOS5 ≥11 (≥10)
VATS1 (N=1,230) 72.0 (7.0) 688 (55.9) 705 (57.3)
Open Thoracotomy (N=1,198) 73.0 (7.0) 672 (56.1) 672 (56.1)
1,080 (87.8) 65 (5.3) 29 (2.4) 56 (4.6)
1,037 (86.6) 80 (6.7) 37 (3.1) 44 (3.7)
274 (22.7) 287 (23.8) 262 (21.7) 384 (31.8)
413 (35.0) 290 (24.6) 240 (20.3) 238 (20.2)
463 (37.6) 323 (26.3) 444 (36.1)
365 (30.5) 341 (28.5) 492 (41.1)
49 (4.0) 113 (9.2) 291 (23.7) 384 (31.2) 393 (32.0)
56 (4.7) 107 (8.9) 276 (23.0) 409 (34.1) 350 (29.2)
918 (74.6) 214 (17.4) 98 (8.0)
P-value Adjusted2 Unadjusted <0.001 0.39 0.99 0.24 0.82 0.48
<0.001
0.75
<0.001
>0.99
<0.001
0.95
0.84
0.58
879 (73.4) 215 (18.0) 104 (8.7)
0.66
0.98
520 (42.3) 377 (30.7) 238 (19.4) 62 (5.0) 33 (2.7)
370 (30.9) 342 (28.6) 351 (29.3) 85 (7.1) 50 (4.2)
<0.001
0.75
812 (66.0) 295 (24.0) 25 (2.0) 98 (8)
706 (58.9) 383 (32.0) 23 (1.9) 86 (6.9)
0.006
0.97
19
Tumor Site, N (%) Upper Lobe 193 (54.1) 676 (55.0) 675 (56.3) Middle Lobe ≥11 (≥5) 106 (8.6) 67 (5.6) 0.064 >0.99 Lower Lobe 122 (36.1) 432 (35.1) 431 (36.0) Other Site ≤11 (≤10) 16 (1.3) 25 (2.1) PET6 Scan, N (%) 259 (76.6) 847 (68.9) 945 (78.9) <0.001 0.64 Chest CT7, N (%) 222 (65.7) 793 (65.5) 813 (67.9) 0.21 0.94 Mediastinoscopy, N (%) 60 (17.8) 129 (10.5) 116 (9.7) <0.001 0.73 Surgery Type3, N (%) Limited Resection 26 (7.7) 103 (8.4) 76 (6.3) 0.16 0.08 Lobectomy 312 (92.3) 1,127 (91.6) 1,122 (93.7) 1 Video-Assisted Thoracic Surgery; 2Adjusted for propensity scores; 3Matching characteristics used to form cohorts; 4 Interquartile Range; 5 Not Otherwise Specified; 6 Positron Emission Tomography; 7Computer Tomography; Cells with <11 patients were masked to maintain patient confidentiality.
20
Table 2: Postoperative Complication Rates of Robotic-Assisted Surgery Compared to Video-Assisted Thoracic Surgery and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients Entire Cohort
Limited to Stage I-II
Limited to Years Limited to Years 2008 – 2011 2012 - 2013 RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS vs. RAS1 VATS2 Open3 VATS VATS VATS VATS Open Open Open Open Outcome P-value N (%) N (%) N (%) OR4 OR OR OR OR OR OR OR 5 (95% CI ) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) Any 251 888 1000 <0.001 1.02 1.00 0.78 1.17 0.95 0.57 0.60 0.49 Complication (74.3) (72.2) (83.5) (0.76 - 1.37) (0.74 - 1.36) (0.45 - 1.35) (0.70 - 1.96) (0.66 - 1.36) (0.42 - 0.79) (0.43 - 0.83) (0.33 - 0.73) Respiratory 113 362 464 <0.001 0.89 1.18 0.87 1.13 1.01 1.30 0.82 1.11 Complication (33.4) (29.4) (38.7) (0.68 - 1.18) (0.89 - 1.56) (0.65 - 1.17) (0.84 - 1.51) (0.63 - 1.64) (0.84 - 2.00) (0.58 - 1.16) (0.77 - 1.61) Extrapulmonary 17 54 104 <0.001 0.72 1.31 0.81 1.35 0.63 1.02 0.77 1.53 Infection (5.0) (4.4) (8.7) (0.40 - 1.29) (0.71 - 2.40) (0.45 - 1.46) (0.73 - 2.49) (0.23 - 1.70) (0.40 - 2.57) (0.37 - 1.60) (0.69 - 3.42) Cardiovascular ≤11 29 28 0.79 1.21 1.05 1.12 0.84 1.46 0.78 1.10 1.30 Complication (≤3) (2.4) (2.3) (0.53 - 2.75) (0.47 - 2.33) (0.43 - 2.89) (0.34 - 2.08) (0.31 - 6.86) (0.18 - 3.44) (0.42 - 2.89) (0.52 - 3.25) Thromboembolic 17 63 60 0.99 1.00 1.03 1.00 0.97 0.86 1.11 1.14 0.97 Complication (5.0) (5.1) (5.0) (0.55 - 1.82) (0.56 - 1.88) (0.53 - 1.90) (0.51 - 1.84) (0.34 - 2.21) (0.41 - 3.00) (0.53 - 2.49) (0.46 - 2.05) Blood 12 63 125 <0.001 0.59 0.60 0.57 0.60 0.31 0.33 0.28 0.34 Transfusion (3.6) (5.1) (10.4) (0.30 - 1.14) (0.30 - 1.18) (0.20 - 1.66) (0.26 - 1.42) (0.17 - 0.60) (0.17 - 0.64) (0.10 - 0.83) (0.15 - 0.75) Reoperation ≤11 14 27 0.09 0.67 1.55 0.73 1.55 1.02 1.87 0.56 1.44 (≤3) (1.1) (2.3) (0.24 - 1.85) (0.52 - 4.63) (0.26 - 2.05) (0.52 - 4.62) (0.20 - 5.30) (0.33 - 10.7) (0.15 - 2.06) (0.35 - 5.95) Prolonged LOS6 21 72 147 <0.001 1.13 0.64 1.10 0.65 0.82 0.57 1.40 0.59 (6.3) (6.0) (12.4) (0.66 - 1.92) (0.37 - 1.10) (0.63 - 1.91) (0.26 - 1.58) (0.36 - 1.90) (0.30 - 1.07) (0.70 - 2.80) (0.35 - 0.98) 7 ICU Stay 208 740 871 <0.001 0.96 0.95 0.81 1.29 1.23 0.58 0.60 0.49 (0.74 - 1.26) (0.71 - 1.26) (0.49 - 1.33) (0.82 - 2.00) (0.58 - 1.14) (61.9) (61.2) (73.6) (0.44 - 0.77) (0.45 - 0.81) (0.35 - 0.68) Readmission 26 100 112 0.47 0.76 0.87 0.76 0.87 0.91 0.82 0.67 0.92 (7.7) (8.3) (9.5) (0.48 - 1.21) (0.54 - 1.43) (0.47 - 1.24) (0.52 - 1.46) (0.39 - 2.11) (0.36 - 1.87) (0.39 - 1.16) (0.51 - 1.69) Postoperative ≤11 17 35 0.01 0.46 0.76 0.51 0.78 0.18 0.35 0.62 0.96 Mortality (≤3) (1.4) (2.9) (0.16 - 1.34) (0.24 - 2.38) (0.17 - 1.51) (0.25 - 2.48) (0.03 - 1.35) (0.04 - 3.23) (0.18 - 2.11) (0.26 - 3.59) 1 Robotic-Assisted Surgery; 2Video-Assisted Thoracic Surgery; 3Open Thoracotomy; 4Odds Ratio (adjusted); 5Confidence Interval; 6Length of Stay; 7Intensive Care Unit; Cells with <11 patients were masked to maintain patient confidentiality.
21
Table 3: Oncologic Outcomes of Robotic-Assisted Surgery Compared to VATS and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients Outcome Lymph Node Staging (>10 LNs Examined) OR3 (95% CI4) Survival HR5 (95% CI) Overall survival
Cohort Entire Cohort: Stage I-IIIA Limited to Stage I-II
Robotic-Assisted Surgery vs. Open Thoracotomy Adjusted2 Unadjusted 1.36 (1.03 - 1.80) 1.28 (0.94 - 1.74) 1.35 (0.98 - 1.86) 1.45 (1.08 - 1.94)
Robotic-Assisted Surgery vs. VATS1 Unadjusted 0.97 (0.75 - 1.26) 1.02 (0.78 - 1.33)
Adjusted2 0.88 (0.66 - 1.18) 0.93 (0.69 - 1.26)
Entire Cohort: Stage I-IIIA 0.79 (0.64 - 0.97) 0.81 (0.63 - 1.04) 1.05 (0.83 - 1.32) 0.91 (0.70 - 1.18) Limited to Stage I-II 0.79 (0.59 – 1.05) 0.95 (0.73 – 1.23) 0.87 (0.65 – 1.17) 0.72 (0.56 – 0.92) Lung cancer-specific survival Entire Cohort: Stage I-IIIA 0.79 (0.56 - 1.13) 0.75 (0.51 - 1.12) 1.13 (0.77 - 1.66) 0.87 (0.56 - 1.34) Limited to Stage I-II 0.67 (0.41 – 1.10) 0.91 (0.56 – 1.50) 0.78 (0.45 – 1.32) 0.61 (0.39 – 0.96) 1 Video-Assisted Thoracic Surgery; 2Adjusted for propensity scores and surgical volume; 3Odds Ratio; 4Confidence Interval: adjusted for clustering by surgeons due to matching; 5Hazard Ratio
22
Figure 1: Propensity Score Adjusted Costs of Robotic-Assisted Surgery Compared to VATS and Open Thoracotomy in Stage I-IIIA Non-Small Cell Lung Cancer Patients According to Phase of Care B: p=0.08
$60,000.00
A: p=0.02
A: RAS vs VATS B: RAS vs Open
$50,000.00
B: p<0.001
$40,000.00
A: p=0.08
B: p=0.95
$30,000.00
Robotic VATS
A: p=0.15
OPEN $20,000.00 B: p=0.12
$10,000.00
A: p<0.001
$0.00 Total Costs
Pre-operative
Operative
Post-operative
e-Appendix 1. Methods As discussed in the manuscript, study patients were identified using the SEER-Medicare database, a large population-based source of longitudinal data that includes demographic and clinical information for Medicare beneficiaries with cancer in various United States (US) regions.(E1) The SEER program has been expanded to cover 28% of the US population, with 94% of eligible patients linked to Medicare. Using SEER data, we obtained each patient’s sociodemographic information, including age, gender, marital status, race, and ethnicity. Medicare data allowed us to group patients into quartiles based on socioeconomic status according to census tract level information. Comorbidities were assessed using Medicare data by applying a modified version of the Charlson index and lung cancer-specific weights.(E2,E3) Claims indicating home services (limited to homebound patients) were used as a proxy for poor performance status.(E4) Each patient’s cancer was staged according to the American Joint Commission on Cancer (seventh edition) using SEER data providing tumor size, extension, location and lymph node status. Cancer histology and grade were determined according to the International Classification of Diseases for Oncology, Third Edition, and morphology codes available in SEER. Information about tests performed for cancer diagnosis (e.g., fine needle aspiration, bronchoscopy), preoperative evaluation (e.g., ventilation/perfusion scan, cardiac/cardiopulmonary stress testing), and staging work-up (e.g., positron emission tomography [PET], mediastinoscopy) was ascertained using International Classification of Diseases, 9 th revision diagnostic and procedure codes (ICD-9) and Current Procedural Terminology-4 (CPT-4) codes from Medicare claims.(E5, E6) The SEER registry routinely collects treatment data for 4 months after diagnosis and, therefore, this time period was used to assess the initial course of therapy. ICD-9 procedure and CPT-4 codes submitted as part of Medicare claims were used to classify study subjects as having undergone robotic-assisted surgery (ICD-9: 17.45 and CPT-4: S2900), VATS (ICD-9: 34.21 and CPT-4: 32657), or open thoracotomy (ICD-9: 34.01 and CPT-4: 32100 ). We focused on patients who underwent lobectomy (SEER codes 30, 31, and 33 and ICD-9 code 32.4) or limited resection (SEER codes 20 to 22 and ICD-9 code 32.29 and 32.3), as these procedures can be performed via these surgical approaches. Physician-specific surgical case volume, defined as the volume of resections performed by a patient’s surgeon in the year of the patient’s surgery, was ascertained from Medicare billing claims. Using this data, surgeons were ranked according to their average volume of claims for lung cancer resections.(E7,E8) Patients who received adjuvant treatments (i.e., post-operative RT and chemotherapy) were identified using both SEER and Medicare information using validated algorithms.(E9,E10) Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.
Consistent with prior studies, postoperative morbidity and mortality (primary outcomes) were defined as the presence of complications and death, respectively, within 30 days of surgery or during the hospitalization in which the primary surgical procedure was performed. We also used prolonged length of stay (>14 days), obtained from MedPAR files, as an indicator of postoperative outcomes. We identified patients requiring intensive care unit (ICU) admission following surgery based on information provided as part of Medicare inpatient files. Sample size calculations determined that the study included the necessary number of patients to achieve 80% power to identify statistically significant differences (α level 0.05) in the rate of postoperative complications among patients treated with robotic-assisted surgery vs. VATS and open surgery (OR of 1.3 to 2.0). Survival analyses were exploratory in nature given that robotic-assisted surgery is a relatively new procedure for lung cancer resection without long-term follow up data. All analyses were conducted with SAS statistical software (SAS, version 9.4; Cary, NC). The Icahn School of Medicine at Mount Sinai Institutional Review Board considered the study exempt.
Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.
References E1. Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Medical care 2002; 40: Iv-3-18. E2. Wisnivesky JP, Henschke CI, Swanson S, Yankelevitz DF, Zulueta J, Marcus S, Halm EA. Limited resection for the treatment of patients with stage IA lung cancer. Annals of surgery 2010; 251: 550-554. E3. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. Journal of clinical epidemiology 2000; 53: 1258-1267. E4. Wisnivesky JP, Smith CB, Packer S, Strauss GM, Lurslurchachai L, Federman A, Halm EA. Survival and risk of adverse events in older patients receiving postoperative adjuvant chemotherapy for resected stages II-IIIA lung cancer: observational cohort study. BMJ (Clinical research ed) 2011; 343: d4013. E5. Levin DC, Parker L, Intenzo CM, Sunshine JH. Recent rapid increase in utilization of radionuclide myocardial perfusion imaging and related procedures: 1996-1998 practice patterns. Radiology 2002; 222: 144-148. E6. Lathan CS, Neville BA, Earle CC. The effect of race on invasive staging and surgery in non-small-cell lung cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2006; 24: 413-418. E7. Earle CC, Schrag D, Neville BA, Yabroff KR, Topor M, Fahey A, Trimble EL, Bodurka DC, Bristow RE, Carney M, Warren JL. Effect of surgeon specialty on processes of care and outcomes for ovarian cancer patients. Journal of the National Cancer Institute 2006; 98: 172-180. E8. Schrag D, Earle C, Xu F, Panageas KS, Yabroff KR, Bristow RE, Trimble EL, Warren JL. Associations between hospital and surgeon procedure volumes and patient outcomes after ovarian cancer resection. Journal of the National Cancer Institute 2006; 98: 163-171. E9. Warren JL, Harlan LC, Fahey A, Virnig BA, Freeman JL, Klabunde CN, Cooper GS, Knopf KB. Utility of the SEER-Medicare data to identify chemotherapy use. Medical care 2002; 40: Iv-55-61. E10. Virnig BA, Warren JL, Cooper GS, Klabunde CN, Schussler N, Freeman J. Studying radiation therapy using SEER-Medicare-linked data. Medical care 2002; 40: Iv-49-54.
182739
Online supplements are not copyedited prior to posting and the author(s) take full responsibility for the accuracy of all data.