Endometrial cancer is the most common gynecologic malignancy in developed countries, with an estimated 60,000 new cases diagnosed in 2016 in the United States.1Surgical staging for endometrial cancer can be performed through open surgery through laparotomy or minimally invasive surgery (laparoscopy and robotic surgery) with no differences between them in terms of oncologic outcomes.2,3 Minimally invasive surgery has been shown to have several perioperative benefits compared with open surgery in both benign and malignant disease: less blood loss and need for transfusion, decreased postoperative pain, shorter hospital stay, and better quality of life.4–7 Vaginal surgery is generally accepted as an alternative for selected patients at high risk for surgical morbidity.8,9
Laparoscopic surgery has been used for the treatment of uterine cancer since the early 1990s,10 but after more than 20 years of use of laparoscopic surgery and vaginal surgery for endometrial cancer, still less than one fourth of patients with endometrial cancer in the United States were treated with minimally invasive surgery.11 The introduction of robotic surgery for hysterectomy (U.S. Food and Drug Administration approval in 2005), however, was crucial for the implementation of minimally invasive surgery for benign and malignant disease at the national level.12,13
In the current analysis, we focused on the association between the adoption of minimally invasive surgery and 30-day morbidity and mortality in endometrial cancer staging in the United States. Also, we report a detailed comparison of 30-day surgical outcomes between minimally invasive surgery and open surgery through laparotomy adjusted for potential confounders.
MATERIALS AND METHODS
For our retrospective cohort study, the American College of Surgeons’ (ACS) National Surgical Quality Improvement Project database was reviewed for patients who had undergone surgical staging for endometrial cancer. The ACS National Surgical Quality Improvement Project is the first nationally validated, risk-adjusted, outcomes-based mechanism able to accurately collect and analyze surgical-related outcomes data in the United States. In 2016, more than 750 hospitals had joined the project. Data are collected according to strictly standardized definitions. Patients enrolled are followed up during their hospital course and after discharge until 30 days postoperatively. A total of 323 variables (including preoperative risk factors, intraoperative characteristics, and 30-day postoperative mortality and morbidity outcomes) are collected in the ACS National Surgical Quality Improvement Project database, many of them available beginning in 2005. This study was exempt from institutional review board approval because the National Surgical Quality Improvement Project database we used in the analysis is population-based and deidentified.
The ACS National Surgical Quality Improvement Project database was reviewed for 2008–2014 for patients undergoing hysterectomy for endometrial cancer. Endometrial cancer was identified using the patient's primary postoperative International Classification of Diseases, 9th Revision or 10th Revision diagnosis code. Current Procedural Terminology (CPT) codes were used to identify hysterectomy, which was stratified by approach as follows: open surgery through laparotomy, minimally invasive surgery, and vaginal surgery (Box 1). To avoid including patients with stage IV disease, women with disseminated cancer and CPT codes for associated surgical procedures involving nongynecologic anatomic regions were excluded.
First, trends in the use of each approach over the study period were analyzed, comparing patient characteristics and 30-day outcomes between the different surgical approaches: minimally invasive surgery, vaginal surgery, and open surgery through laparotomy. Univariate analysis was performed comparing the three surgical approaches for patient baseline characteristics (age, body mass index [BMI, calculated as weight (kg)/[height (m)]2], functional status, comorbid conditions, American Society of Anesthesiologists score, preoperative anemia), perioperative characteristics (operative time, intraoperative complications, length of stay), and 30-day morbidity (minor and major complications, readmissions, reoperations, deaths). Operative time was considered a continuous and dichotomous variable (arbitrarily assuming a cutoff at 2 hours). Complications were categorized according to the ACS National Surgical Quality Improvement Project classification. Major complications were defined as unplanned intubation, wound disruption, ventilator support for longer than 48 hours, sepsis, septic shock or systemic inflammatory respiratory syndrome, pneumonia, deep incisional surgical site infection, acute renal failure, organ space surgical site infection, progressive renal insufficiency, pulmonary embolism, myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, stroke or cerebrovascular accident with neurologic deficit, deep vein thrombosis, or thrombophlebitis. Minor complications were defined as any urinary tract infection or superficial surgical site infection. In addition, multivariable logistic regression models were used to investigate the independent effect of minimally invasive surgery (compared with open surgery through laparotomy) on 30-day outcomes (major complications, intraoperative and postoperative transfusions, superficial surgical site infections, readmissions, and reoperations).
Statistical analysis was performed using SAS 9.3. The Cochran-Armitage test for trend was used to assess trends over time. In the univariate analyses, comparisons were made among the three surgical groups using the χ2 test or Fisher exact test for categorical variables and the Kruskal-Wallis test for continuous variables. In open surgery compared with minimally invasive surgery comparisons, t tests and Wilcoxon rank-sum tests were also used for continuous variables. Multivariable logistic regression models were used to assess the association between minimally invasive surgery and 30-day outcomes while controlling for clinically significant comorbid conditions, associated procedures (pelvic lymphadenectomy), and patient characteristics.
In addition, we used inverse probability of treatment weighting adjustment for the propensity score as an alternative method for estimating the effects of open surgery and minimally invasive surgery and to eliminate potential residual imbalance in prognostically significant covariates in multivariable logistic regression models.14 Inverse of the propensity score was obtained by using a multivariable logistic regression model with outcome of minimally invasive surgery. The model adjusted for age, BMI, presence of pelvic lymphadenectomy, any functional dependency, hypertension requiring medication, diabetes mellitus, current smoker (within 1 year), American Society of Anesthesiologists classification, inpatient status, preoperative albumin level, and year of operation. The predicted values were then converted to the inverse propensity score of treatment to produce the weighted propensity score using w1=P(E)/P(Ei) (exposed) and w0=1–P(E)/1–P(Ei) (unexposed) through transformed logit. All calculated P values were two-sided, and P<.05 was considered statistically significant.
The ACS National Surgical Quality Improvement Project and the hospitals participating in the ACS National Surgical Quality Improvement Project are the source of the data used here; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
During the study period, 12,752 patients with primary International Classification of Diseases, 9th Revision or 10th Revision diagnosis codes for endometrial cancer who underwent hysterectomy were identified. After exclusion of women with disseminated cancer (n=307) and nongynecologic surgical procedures (n=162), 12,283 patients met the inclusion criteria: 4,246 patients (34.6%) underwent open surgery through laparotomy, 7,737 (63.0%) underwent minimally invasive surgery, and 300 (2.4%) underwent vaginal surgery. A significant implementation of minimally invasive surgery (24.2–71.4% of procedures) and a concomitant decrease in the use of open surgery (74.3–26.4% of procedures) was observed from 2008 to 2014 (P<.001 for trend), whereas the rate of vaginal surgery (1.5–2.2% of procedures) did not change (P=.06 for trend) (Fig. 1).
Baseline characteristics of patients who underwent minimally invasive surgery compared with open surgery through laparotomy are reported in Table 1. No significant differences in BMI were found between the groups (mean [SD]: open surgery, 35.0 [10.1] versus minimally invasive surgery, 34.8 [9.8]); more than 25% of women in each group were extremely obese (BMI 40 or greater, obesity class III according to the World Health Organization classification). Patients who underwent minimally invasive surgery were slightly younger (mean [SD]: open surgery, 62.9 years versus minimally invasive surgery, 62.2 years; P<.001). A more favorable preoperative American Society of Anesthesiologists score (American Society of Anesthesiologists 1 or 2) was associated with minimally invasive surgery (open surgery, 43.7% versus minimally invasive surgery, 52.3%; P<.001); moreover, the minimally invasive surgery group had a significantly higher percentage of women with an independent functional status (open surgery, 97.5% versus minimally invasive surgery, 98.8%; P<.001). After propensity score balancing, lower albumin level and current smoking remained clinically associated with performance of open surgery compared with minimally invasive surgery (P<.001 for both).
Operative time was significantly longer in the minimally invasive surgery cohort (P<.001); 38.5% with open surgery through laparotomy and 27.2% with minimally invasive surgery had a duration of surgery less than 2 hours (P<.001) (Table 2). The open surgery through laparotomy group had significantly more minor complications (open surgery, 7.9% versus minimally invasive surgery, 2.6%) and major complications (open surgery, 9.1% versus minimally invasive surgery, 3.2%) 30 days after surgery (both P<.001). A higher rate of transfusion (13.7% versus 1.7%) and longer median hospital stay (3 days versus 1 day) were also observed in the open surgery group compared with minimally invasive surgery (both P<.001). Readmission rate within 30 days after surgery was 9.1% for open surgery compared with 3.4% for minimally invasive surgery with reoperation rates of 2.1% and 1.1%, respectively (both P<.001). Mortality rate associated with surgery (within 30 days) was also significantly higher for open surgery (open surgery, 0.8% versus minimally invasive surgery, 0.2%; P<.001).
On multivariable logistic regression performed with inverse probability of treatment weighting and adjusting for possible confounders, open surgery through laparotomy (compared with minimally invasive surgery) was independently associated with increased odds of major complications (adjusted odds ratio [OR] 2.36, 95% CI 2.00–2.79), readmission (adjusted OR 2.17, 95% CI 1.81–2.60), and return to the operating room (adjusted OR 1.52, 95% CI 1.12–2.06) (Table 3). Open surgery through laparotomy also was strongly associated with superficial surgical site infection (adjusted OR 6.75, 95% CI 4.97–9.17), perioperative blood transfusion (adjusted OR 5.85, 95% CI 4.82–7.11), and death within 30 days after surgery (adjusted OR 3.79, 95% CI 2.19–6.56).
From 2008 to 2014, the proportions of any complication (12.8–7.5%) (Fig. 2), superficial surgical site infection (4.5–1.9%) (Fig. 2), deep incisional surgical site infection (0.6–0.5%), sepsis or septic shock (1.2–0.9%), and deep vein thrombosis (1.2–0.5%) within 30 days after surgery decreased significantly (all P<.05 for trend). Accompanying the increase in the use of minimally invasive surgery for the surgical treatment of endometrial cancer, a comprehensive and significant decrease was observed in 30-day morbidity (major complications, P<.001 for trend) along with a slight decrease in 30-day mortality (P=.24 for trend) during the study period (Fig. 3).
Our results suggest that the adoption of minimally invasive surgery is associated with an overall decrease in 30-day related morbidity and mortality for endometrial cancer treatment in the United States. Also, we observed a steady and fast increase in the utilization of minimally invasive surgery in endometrial cancer at a national level during the last decade. It is reasonable to assume that robotic surgery contributed considerably to this change.
In our study, we demonstrated on a large scale that minimally invasive surgery for endometrial cancer staging is associated with fewer surgically related complications and better surgical outcomes than open surgery through laparotomy. Our findings strengthen the results of prior randomized trials comparing laparotomy with minimally invasive techniques for endometrial cancer treatment15 and allow us to generalize those observations to the broad national practice in the United States.
Of interest, we also observed advantages of minimally invasive surgery regarding perioperative death with a 3.8-fold lower odds of 30-day mortality. Assuming 60,000 new cases of endometrial cancer treated yearly in the United States, this might translate to 360 lives saved per year. However, considering the low 30-day mortality rate for both approaches (open surgery, 0.8%; minimally invasive surgery, 0.2%), the clinical significance of this difference is unclear.
During the study period, we observed a steady increase in the use of minimally invasive surgery, up to 71.4% of all procedures, with a contemporary inverse trend of open surgery utilization (from 74.3% to 26.4% of procedures). Our results agree with the findings of a recently published study performed using the Surveillance, Epidemiology, and End Results–Medicare database.13 A previous study reported that only 16.7% of hysterectomies for endometrial cancer staging in the United States were accomplished through a minimally invasive approach in 2006, more than a decade after the acceptance of laparoscopy for the treatment of endometrial cancer.16 In light of this, even if the current analysis could not distinguish between different minimally invasive approaches (laparoscopy versus robot-assisted), we hypothesize that the steady increase in minimally invasive surgery might be the result of the recent broad diffusion of robotically assisted surgery into gynecologic practice. In fact, the da Vinci Surgical System achieved U.S. Food and Drug Administration clearance for gynecologic procedures in April 2005. This hypothesis has been previously suggested also by other investigators.11
In the current study, we also were able to separate vaginal surgery from minimally invasive surgery (laparoscopy or robot-assisted). Although not considered the standard of care for endometrial cancer, vaginal hysterectomy has been shown to be appropriate in selected patients at high risk for surgical morbidity.8,17,18 Because vaginal surgery is likely performed in a different subset of patients, and because of its marginal effect on the study population (2.4% overall rate without significant change through the years), we excluded this approach from the surgery-related analysis. As a consequence, we focused only on the comparison between open surgery through laparotomy and minimally invasive surgery.
Although our analysis included a large sample of patients undergoing primary surgery for endometrial cancer, with detailed analysis of 30-day complications, some limitations must be addressed. First, the ACS National Surgical Quality Improvement Project database does not include information on tumor characteristics and adjuvant treatment received by the patients enrolled. For this reason, oncologic outcomes were not considered in our analysis. However, the equivalence in long-term survival between open surgery through laparotomy and minimally invasive surgery has already been demonstrated elsewhere.19,20 Second, because the data collected in the database come from different types of institutions (community hospitals, teaching hospitals, public hospitals), referral bias could potentially affect our results. Moreover, because the database does not release surgeon and hospital identifiers, we were unable to assess surgeon volume and number of hospitals that treated endometrial cancer over the study period, respectively. Third, the percentage of lymphadenectomy performed registered in the present investigation is much lower than that reported in other registry-based national studies.11,13 This discrepancy might be related to a miscoding of the associated procedures within the National Surgical Quality Improvement Project database. Nonetheless, considering that lymph node assessment does not influence surgical approach (minimally invasive surgery versus open surgery), we expected both groups of our analysis would be equally affected by this underestimation. Fourth, we were not able to distinguish between laparoscopy and robot-assisted surgery. Such limits related to an inherent lack of CPT modifiers did not allow for precise identification of which technique contributed more to minimally invasive surgery implementation.
Our results suggest that the increase in the use of minimally invasive surgery was associated with a detectable decrease in 30-day morbidity and mortality among patients treated for early-stage endometrial cancer in the United States between 2008 and 2014. The increased knowledge of the benefits of laparoscopy and the advancement of laparoscopic techniques certainly contributed to this implementation. However, the introduction of robotic surgery undoubtedly was instrumental in this change. Because a direct correlation has been demonstrated among decreases in 30-day morbidity, shortening of hospital stay, and reduction of hospital and societal costs,19,20 we can suppose that the trend in favor of minimally invasive surgery could bring tangible savings to the U.S. health care system. Nevertheless, the potential benefits to the health care system should be further explored in future dedicated analyses to support our hypothesis.
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin 2016;66:7–30.
2. Walker JL, Piedmonte MR, Spirtos NM, Eisenkop SM, Schlaerth JB, Mannel RS, et al. Laparoscopy compared with laparotomy for comprehensive surgical staging of uterine cancer: Gynecologic Oncology Group Study LAP2. J Clin Oncol 2009;27:5331–6.
3. Janda M, Gebski V, Forder P, Jackson D, Williams G, Obermair A, et al. Total laparoscopic versus open surgery for stage 1 endometrial cancer: the LACE randomized controlled trial. Contemp Clin Trials 2006;27:353–63.
4. Janda M, Gebski V, Brand A, Hogg R, Jobling TW, Land R, et al. Quality of life after total laparoscopic hysterectomy versus total abdominal hysterectomy for stage I endometrial cancer (LACE): a randomised trial. Lancet Oncol 2010;11:772–80.
5. Mourits MJ, Bijen CB, Arts HJ, ter Brugge HG, van der Sijde R, Paulsen L, et al. Safety of laparoscopy versus laparotomy in early-stage endometrial cancer: a randomised trial. Lancet Oncol 2010;11:763–71.
6. Kornblith AB, Huang HQ, Walker JL, Spirtos NM, Rotmensch J, Cella D. Quality of life of patients with endometrial cancer undergoing laparoscopic international federation of gynecology and obstetrics staging compared with laparotomy: a Gynecologic Oncology Group study. J Clin Oncol 2009;27:5337–42.
7. Corrado G, Cutillo G, Pomati G, Mancini E, Sperduti I, Patrizi L, et al. Surgical and oncological outcome of robotic surgery compared to laparoscopic and abdominal surgery in the management of endometrial cancer. Eur J Surg Oncol 2015;41:1074–81.
8. Beck TL, Morse CB, Gray HJ, Goff BA, Urban RR, Liao JB. Route of hysterectomy and surgical outcomes from a statewide gynecologic oncology population: is there a role for vaginal hysterectomy? Am J Obstet Gynecol 2016;214:348.e1–9.
10. Childers JM, Brzechffa PR, Hatch KD, Surwit EA. Laparoscopically assisted surgical staging (LASS) of endometrial cancer. Gynecol Oncol 1993;51:33–8.
11. Fader AN, Weise RM, Sinno AK, Tanner E III, Borah BJ, Moriarty JP, et al. Utilization of minimally invasive surgery in endometrial cancer care: a quality and cost disparity. Obstet Gynecol 2016;127:91–100.
12. Wright JD, Ananth CV, Lewin SN, Burke WM, Lu YS, Neugut AI, et al. Robotically assisted vs laparoscopic hysterectomy among women with benign gynecologic disease. JAMA 2013;309:689–98.
13. Wright JD, Burke WM, Tergas AI, Hou JY, Huang Y, Hu JC, et al. Comparative effectiveness of minimally invasive hysterectomy for endometrial cancer. J Clin Oncol 2016;34:1087–96.
14. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 2011;46:399–424.
15. Obermair A, Janda M, Baker J, Kondalsamy-Chennakesavan S, Brand A, Hogg R, et al. Improved surgical safety after laparoscopic compared to open surgery for apparent early stage endometrial cancer: results from a randomised controlled trial. Eur J Cancer 2012;48:1147–53.
16. Scalici J, Laughlin BB, Finan MA, Wang B, Rocconi RP. The trend towards minimally invasive surgery (MIS) for endometrial cancer: an ACS-NSQIP evaluation of surgical outcomes. Gynecol Oncol 2015;136:512–5.
17. Massi G, Savino L, Susini T. Vaginal hysterectomy versus abdominal hysterectomy for the treatment of stage I endometrial adenocarcinoma. Am J Obstet Gynecol 1996;174:1320–6.
18. Berretta R, Merisio C, Melpignano M, Rolla M, Ceccaroni M, DE Ioris A, et al. Vaginal versus abdominal hysterectomy in endometrial cancer: a retrospective study in a selective population. Int J Gynecol Cancer 2008;18:797–802.
19. Bogani G, Multinu F, Dowdy SC, Cliby WA, Wilson TO, Gostout BS, et al. Incorporating robotic-assisted surgery for endometrial cancer staging: analysis of morbidity and costs. Gynecol Oncol 2016;141:218–24.
© 2018 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
20. Lau S, Vaknin Z, Ramana-Kumar AV, Halliday D, Franco EL, Gotlieb WH. Outcomes and cost comparisons after introducing a robotics program for endometrial cancer surgery. Obstet Gynecol 2012;119:717–24.