The aim of the current study was to compare the safety of the laparoscopic vs abdominal approach to staging endometrial cancer. A search was conducted for randomized controlled trials that reported data from women with histologically confirmed endometrial cancer who underwent laparoscopic or abdominal surgery. An additional metaanalysis was performed. The primary endpoints were the rates of intraoperative and postoperative complications. A total of 8 original randomized controlled trials were included in the final analysis. No significant difference was observed in the relative risk (RR) for intraoperative complications between laparoscopy and laparotomy (RR, 1.25; 95% confidence interval, 0.99–1.56; P = .062). In contrast, a significant advantage of laparoscopy over laparotomy was obtained in terms of postoperative complications (RR, 0.71; 95% confidence interval, 0.63–0.79; P = .016). In comparison with abdominal surgery, the safety of the laparoscopic approach for surgical staging of endometrial cancer is similar in terms of intraoperative complications but results in fewer postoperative complications.
Endometrial cancer is the third most common neoplasia in women in developed countries.
In approximately 80% of patients, it is diagnosed at an early stage, and surgery is the treatment of choice. The standard treatment of endometrial cancer is peritoneal washing and total hysterectomy with bilateral salpingo-oophorectomy. Pelvic, with or without paraaortic, lymphadenectomy is also considered a standard procedure for staging endometrial cancer in several centers.
The advantages of minimally invasive surgical techniques, such as laparoscopy, for treating gynecologic diseases are unquestioned, and several authors reported the feasibility and safety of laparoscopy in treating endometrial cancer in obese and elderly patients and in populations that could receive the greatest benefit from this surgical approach.
Heterogeneous data obtained in small studies have been previously published on the role of laparoscopy for endometrial cancer staging. In a previous metaanalysis of randomized controlled trials (RCTs), we compared the laparoscopy with the traditional surgery in the treatment of endometrial cancer. Laparoscopy resulted in less blood loss, shorter hospital stay, and fewer complications than laparotomy. A similar overall survival was also demonstrated. Oncologic data alone were successively updated including Malzoni’s study because that RCT was published without previous protocol registration.
Notwithstanding the scientific effort to obtain type I clinical evidence on this emerging issue, available metaanalytic data remained of limited value because of the few and small clinical studies included, and to their flaws in the study design and in the outcomes reported. Specifically, only 4 RCTs were included in the final analysis for an overall population of 172 patients randomized to the endoscopic surgery. Three study protocols comparing laparoscopic and abdominal stadiation of endometrial carcinoma were in progress and registered on websites for clinical trials. To date, these large RCTs have been published and the safety data of the laparoscopic approach are available.
Based on these considerations, the aim of this study was to perform a systematic research of published and unpublished RCTs with metaanalysis to obtain conclusive and updated evidence on the safety of laparoscopic vs abdominal surgery for staging patients with endometrial cancer.
Materials and methods
The protocol design followed the Preferred Reporting Items for Systematic Reviews and Metaanalyses (PRISMA) guidelines for reporting systematic reviews and metaanalyses of RCTs.
To be included, trials had to be randomized, that is not confounded by additional therapeutic differences between the 2 groups and without a limit for the randomization time. Trials should have aimed to include patients who had a histologic diagnosis of endometrial cancer. Moreover, trials should have compared laparoscopic vs abdominal surgery.
The bibliographic search for RCTs (articles and/or abstracts) was conducted every month until August 2011, without limits for the English language. A combination of the following medical subject headings or keywords was included: “endometrial cancer,” “endometrial carcinoma,” “complication(s),” “uterine cancer,” “uterine carcinoma,” “laparoscopy,” “laparotomy,” “surgery,” “safety,” “efficacy,” “effectiveness,” “treatment,” and “therapy”.
We searched Medline (through PubMed), the Institute for Scientific Information Web of Science database, and websites for the registration of controlled trials ( http://clinicaltrials.gov/ct2/search ). The bibliographies of retrieved articles, books, expert opinion review articles, and reviewed bibliographies from subject experts were manually searched.
The titles and abstracts were initially screened, and potential relevant articles were identified and reviewed for inclusion/exclusion criteria. Successively, the protocols and results of the studies were examined according to specific inclusion criteria. Lastly, only studies that met the inclusion criteria were considered for the final analysis.
Two independent reviewers (A.F., S.P.) not blinded at any point to the authors or sources of publication reviewed, in parallel, the bibliographic sources, whereas disagreements between the reviewers were discussed and solved by consensus or arbitration (F.Z.).
The primary endpoints were the rates of intraoperative and postoperative complications. The secondary endpoints included the following: conversion to laparotomy, operative time, blood loss, and number of pelvic and paraaortic nodes removed.
The following data were recorded: demographic characteristics of the study population, histologic, morphologic and clinical cancer characteristics, surgical procedures performed, and surgical outcomes. To obtain data for each endpoint from all of the RCTs, whenever it was possible, the collaboration of all corresponding authors was requested to obtain data missing from the papers included in the study.
The overall effect of laparoscopy on the primary and secondary endpoints was shown. The analysis of the treatment effect was performed on an “intention-to-treat (ITT) basis,” considering dropouts and missing data as treatment failures and recalculating the results to adhere to ITT principles, and a “per-protocol basis,” considering the results from the evaluated patients alone.
For dichotomous data, we presented the results as a relative risk (RR) with a 95% CI and combined it for a metaanalysis to calculate a pooled estimate of the treatment effect for each outcome across the studies.
For continuous data, we used the weighted mean difference (WMD) if the outcomes were measured the same way between the trials, whereas we used the standardized mean difference to combine trials that measured the same outcome but used different methods.
To measure heterogeneity (the variation in the outcomes among the studies), we used Cochran’s Q test and a P value equal or higher than .05 to represent statistical homogeneity. For data with statistical homogeneity, the Mantel-Haenszel method was used to calculate the weighted summary RR or the WMD under the fixed effects model, whereas the random effects model of metaanalysis was used in case of statistical heterogeneity.
To test the effect of the publication year of the papers and the complexity of the surgical procedures, Spearman’s rank correlations were used. The complexity of the surgical procedures was defined as “low,” “medium,” and “high” if the surgery included hysterectomy alone, pelvic lymphadenectomy, and pelvic and paraaortic lymphadenectomy, respectively. A P value lower than .05 of 95% confidence interval (CI) that did not contain the unit was considered statistically significant. The StatsDirect software package (CamCode, Ashwell, UK) was used for the statistical analysis.
Results
Figure 1 shows the flow diagram of the study selection according to the PRISMA statement.
After checking for the inclusion and exclusion criteria, we selected a total of eight original RCTs to include in the final analysis.
The quality of the selected RCTs is detailed in Figure 2 . Of the 8 included studies, 3 trials were classified as having unclear sequence generation and allocation concealment. The blinding method was unclear in 2 studies, whereas 4 trials were unblinded. The outcome data were incomplete in 2 studies and unclear in 1 study. The selective outcome reporting was inadequate in 3 trials, and the risk of bias was unclear in the same studies.
The main characteristics of the populations studied are summarized in Table 1 . In all of the trials, patients with stage I endometrial cancer were studied, whereas stages IIa and II-III were included in only 2 of the trials. The surgical procedures performed differed among the studies. In 3 trials, the hysterectomy was laparoscopically assisted vaginal, whereas in the others, it was totally laparoscopic. In addition, Tozzi et al and Walker et al also included laparoscopic/abdominal radical hysterectomy.
| Study (ref) | Sample, n a | Inclusion criteria | Exclusion criteria | Lymphadenectomy, % a | |
|---|---|---|---|---|---|
| Pelvic | Paraaortic | ||||
| Fram et al | 29 vs 32 | Stage I | Not available | 55.2 vs 53.1 | — |
| Zorlu et al | 40 vs 38 | Stage I | Not available | 100 vs 100 | — |
| Tozzi et al | 63 vs 59 | Stage I-III | Uterine size exceeding transversal diameter of 8 cm | 84 vs 83.5 | 60.3 vs 61.2 |
| Zullo et al | 40 vs 38 | Stage I | Other premalignancies or malignancies Major medical conditions Psychiatric disorders Current or past history of acute or chronic physical illness Premenstrual syndrome Current or past use of drugs influencing cognition, vigilance, and/or mood | 100 vs 100 | 10 vs 7.9 |
| Malzoni et al | 81 vs 78 | Stage I | Evidence of more advanced clinical stages Prior pelvic radiotherapy and/or chemotherapy Ovarian lesions Metastasis beyond the uterus Contraindications for general anesthesia Systemic infections Abnormal Papanicolau smear Bulky uterus ≥12-wk size or where vaginal removal of the uterus may require morcellation Documented significant cardiopulmonary disease Severe hip disease precluding the use of the dorsolithotomy position Inadequate bone marrow, renal, and hepatic function BMI 40 kg/m 2 Age >80 y | 100 vs 100 | 24 vs 28 |
| Walker et al | 1696 vs 920 | Stage I-IIA° | Prior retroperitoneal surgery Prior pelvic or abdominal radiation therapy Pregnancy Contraindications to laparoscopy Metastasis on chest x-ray | 98 vs 99 | 94 vs 97 |
| Mourits et al | 185 vs 94 | Stage I, grade 1-2 EA or CH | Any non-EA Uterine size larger than that expected 12 wk of pregnancy Cardiopulmonary contraindication for laparoscopy or laparotomy | — | — |
| Janda et al | 190 vs 142 | Stage I, EC | Histologic cell-type other than EA on curettings Bulky lymph nodes on imaging Uterine size larger than 10 wk of gestation Estimated life expectancy less than 6 mo Medically unfit for surgery Noncompliant patients Geographic area not allowing an adequate follow-up Unfit to complete QoL assessments | 40.5 vs 67.6 | — |
a For laparoscopic and abdominal surgery, respectively. b Endometrial adenocarcinoma or sarcoma with no clinical evidence of metastasis beyond the uterine corpus or with macroscopic involvement of the endocervix; adequate bone marrow, renal, and hepatic function; no contraindication for surgery; English, Canadian, French, and Spanish languages; no current evidence of other malignancy.
A wide variability was found between pelvic and paraaortic lymphadenectomy. Pelvic node dissection was always performed in four trials, whereas in a variable proportion of patients (range, 40.5–84%), it was performed in 3 trials and never in 1 trial. Paraaortic node dissection was performed in a high percentage of subjects (range, 60.3–97%) in 2 studies, rarely (range, 7.9–28%) in 2 trials, and never in 4 studies.
The rate of intraoperative complications was reported in 7 studies. After the metaanalysis, the RR for intraoperative complications was not different between laparoscopy and laparotomy (RR, 1.25; 95% CI, 0.99–1.56; P = .062) without significant heterogeneity across the studies ( P = .827) ( Figure 3 , A) .
After combining the data from all of the RCTs included in the metaanalysis, a significant advantage of laparoscopy over laparotomy was obtained in terms of postoperative complications (RR, 0.71; 95% CI, 0.63–0.79; P = .016) with significant heterogeneity across the studies ( P = .007) ( Figure 3 , B).
The results for operative time and blood loss were based on 6 RCTs. After combining the data from the 6 RCTs, a significantly longer operative time was observed during the laparoscopic procedure (WMD = 51.46; 95% CI, 46.56–58.36; P < .0001) ( Figure 4 , A) . In addition, a significant ( P < .0001) heterogeneity across the studies was observed. In contrast, a significant reduction of blood loss in patients who received laparoscopy was reported (WMD = −17.82; 95% CI, −20.86 to −14.79; P < .0001) after a metaanalysis of the RCTs ( Figure 4 , B). In addition, a significant ( P < .0001) heterogeneity across the studies was observed.
