Objective
The purpose of this study was to compare operative outcomes and complications for patients with endometrial cancer who underwent staging by laparoscopy vs laparotomy in a low-volume facility.
Study Design
Research was conducted with a retrospective cohort of surgical patients with clinical stage I endometrial cancer from 2004-2009.
Results
Eighty-six demographically similar patients (50 laparotomy and 36 laparoscopy) were identified. Laparoscopy had less estimated blood loss (339 vs 558 mL; P = .013) and lower rates of transfusion (5.6% vs 24%; P = .02). Laparoscopy was longer (281 vs 202 minutes; P < .0005) but required a shorter hospital stay (2.2 vs 5.5 days; P < .0005). Laparoscopy patients had fewer overall complications (16.7% vs 32%; P = .11). No differences in final surgical stage or lymph node yields between the groups were present.
Conclusion
Although a longer procedure, laparoscopy had fewer complications and shorter hospital stays. Prolonged operative time, compared with published experience, is potentially the result of unique factors in our center.
Although the role of comprehensive surgical staging continues to be debated since 1988 with the change in FIGO staging from clinical to surgical, the traditional approach for endometrial cancer surgical staging remains a total abdominal hysterectomy, bilateral salpingo-oophorectomy, pelvic washings, and pelvic/paraaortic lymph node dissection performed through a midline vertical incision. However, the evolution of gynecologic laparoscopy and robotic-assisted laparoscopy has allowed more difficult procedures (such as endometrial cancer staging) to be done in a minimally invasive manner. It is thought that a minimally invasive approach provides similar or better outcomes for the patient such as a shorter hospital stay and fewer postoperative complications (such as wound breakdown and infection) without an impact on her cancer diagnosis and outcome.
Although there are relatively few randomized controlled trials regarding this subject, a recent metaanalysis of the available data confirmed optimal patient outcomes that favor laparoscopy and observed no differences in disease-free and cancer-related survival. The largest randomized trial to compare laparoscopy and laparotomy for early endometrial cancer management, Gynecologic Oncology Group LAP2, reported reassuring perioperative complication and outcome data in late 2009 that was largely consistent with previous trials. More importantly, preliminary 3-year overall survival was similar at 89.8% for patients who were treated with laparoscopy compared with 89.9% for women who were treated with laparotomy. Recent studies have compared both actual and modeled costs and found that laparoscopy was significantly more cost-efficient than laparotomy in patients with early-stage endometrial cancer.
Although surgical volume has been evaluated extensively as a predictor for outcomes in patients with ovarian cancer, limited information is present regarding surgical volume and impact on endometrial cancer outcomes. As laparoscopic staging of endometrial cancer has gained widespread acceptance in the gynecologic oncology community in appropriate surgical patients, we increasingly have been incorporating laparoscopic treatment of early endometrial cancer over the past several years at our institution; however, difficulties in the implementation of this new treatment involve an absence of dedicated laparoscopy rooms, the lack of specially trained operative nurses/technicians for assistance with complex laparoscopic procedures, and an absence of laparoscopic teams. These difficulties can lengthen procedure times and, in some cases, possibly impact outcome and/or complications. The objective of our study was to compare operative outcomes and complications for endometrial cancer in patients who undergo staging with laparoscopy vs laparotomy in a low-volume facility to determine whether findings in other studies can be extrapolated to patients in a low-volume center where the operating circumstances may be different.
Materials and Methods
After institutional review board approval, a query into the inpatient electronic medical records system at Brooke Army Medical Center (San Antonio, TX) was performed to identify patients who underwent primary surgery for endometrial cancer between 2004 and 2009. An additional query was made into the tumor registry database at Brooke Army Medical Center for the same time period to ensure that all eligible patients were identified. Patients were excluded if there was evidence of previous pelvic radiation, uterine sarcoma on preoperative biopsy or final pathologic evaluation, a clearly delineated surgical plan for extrafascial hysterectomy only, or clinical stage II-IV endometrial cancer. Operative reports, discharge summaries, and readmissions were reviewed from the inpatient electronic medical record system. Applicable demographic and follow-up information was obtained from the outpatient medical record system. The information was entered into a database in compliance with the Health Insurance Portability and Accountability Act of 1996.
A retrospective cohort study was performed that compared (1) patients who were taken to the operating room for planned laparoscopic treatment (including both laparoscopic-assisted vaginal hysterectomy and total laparoscopic hysterectomy) of their clinical stage I endometrial cancer and (2) patients who had their endometrial cancer surgically treated by a laparotomy. All surgeries were performed by 1 of 4 Board Certified Gynecologic Oncologists (E.R.K., M.J.S., W.E.W., or C.A.L.). Patients were selected for the laparoscopic approach at the discretion of each surgeon. Surgical staging was performed per the Gynecologic Oncology Group surgical manual. None of the cases in this study involved the use of robotic-assisted laparoscopic surgery. Per local practice, surgery start time is recorded as the time of sterile placement of the Foley catheter by the operative team.
The primary study outcome was to compare the operative complication rates. Operative complications were a composite endpoint that totaled the incidence of (1) urologic, vascular, or gastrointestinal injury (unless planned), (2) readmission within 30 days from discharge for any cause with or without a repeat surgery unless it is clearly unrelated to initial surgery (for instance, a motor vehicle accident with an extremity fracture), and (3) unplanned postoperative intensive care unit (ICU) admission from the operating room. Secondary outcomes consisted of operative time, nodal counts, estimated blood loss (EBL), hospital length of stay, and a description of the conversion rate from laparoscopy to laparotomy. Furthermore, descriptive statistics were obtained with means for continuous data or described for nominal or categoric data for the 2 operative populations. Additional information that was collected included standard demographics, type of surgery, blood transfusion data, and pathologic data that included final surgical stage and grade. For this study, the data were analyzed on the basis of intention to treat; thus, patients who underwent conversion from laparoscopy to laparotomy are counted in the laparoscopy group.
SPSS software (version 16.0; SPSS Inc, Chicago, IL) was used to perform the statistical calculations. Probability values < .05 were considered to be statistically significant. Mean values are reported, unless otherwise noted. The χ 2 test or Fisher’s exact test, in instances in which the expected count was <5, was used to compare categoric data; parametric continuous variables were compared with the use of the 2-sample t test. The Mann-Whitney U test was used to compare data that were not normally distributed. Tests for equal variances were performed, and the appropriate probability value was chosen. All probability values were 2-tailed.
Results
Eighty-six patients met our inclusion criteria. The mean age for all patients was 61.6 years (range, 34–89 years). Overall, the 2 groups of patients were similar in age, race, American Society of Anesthesiologists classification and body mass index (BMI; Table 1 ). Laparotomy was performed for 50 patients; laparoscopy was initiated for 36 patients, but 4 were converted to laparotomy, for a conversion rate of 11.1%. The reasons for conversion were (1) common iliac injury during node dissection, (2) pelvic nodes suspicious for metastatic disease, (3) inability to visualize paraaortic nodes, and (4) distended bowel that limited visualization.
| Demographic | Laparoscopy (n = 36) | Laparotomy (n = 50) | P value |
|---|---|---|---|
| Age, y a | 61 (36–83) | 62 (34–89) | .75 |
| Body mass index, kg/m 2 a | 31.4 (19.6–50.9) | 33.3 (19–58.3) | .34 |
| Race, n | .053 | ||
| White | 25 | 33 | |
| Hispanic | 2 | 10 | |
| African American | 5 | 2 | |
| Asian | 4 | 2 | |
| Other | 0 | 3 | |
| American Society of Anesthesiologists Classification System, n | .194 | ||
| 1 | 2 | 0 | |
| 2 | 20 | 22 | |
| 3 | 13 | 24 | |
| 4 | 0 | 1 |
Operative characteristics and complications are given in Table 2 . Although laparoscopy was associated with a longer operative time (281 vs 202 minutes; P < .005), more EBL (339 vs 558 mL; P = .013), higher transfusion rate (5.6% vs 24%; P = .02), longer length of stay (2.2 vs 5.5 days; P < .0005), and more ICU admissions (0% vs 14%; P = .038), all favored laparoscopy. Reasons for ICU admissions included sepsis, fluid overload, acute myocardial infarction, atrial fibrillation, and hypotension. There was not a difference in hospital readmission rate with 4 laparoscopy patients (11%) vs 8 laparotomy patients (16%; P = .55). Two of the laparoscopy readmissions were for port site herniations, which were treated with repeat laparoscopy. The third laparoscopy readmission was for wound dehiscence in a patient who had been converted from laparoscopy to laparotomy, with the final laparoscopy readmission having a postoperative ileus because of a pelvic abscess that was drained by an interventional radiologic procedure. The reasons for readmission in the laparotomy patients were urosepsis (n = 1), pneumonia (n = 1), wound cellulitis or infection (n = 4; included 1 patient with a concomitant fascial dehiscence), ileus (n = 1), and isolated fascial dehiscence (n = 1). There was 1 perioperative death in the laparotomy group because of sepsis; the surgery had been complicated by an incidental enterotomy that had been repaired at the time of surgery. No perioperative deaths occurred in the laparoscopy group. Overall, there were no differences in the intraoperative injury rate (8% vs 5.6%; P = 1.0) or in the composite rate of all complications (16.7% vs 32%; P = .11), which nonetheless favored the laparoscopy group.
