Comparison of survival outcomes between robotic and laparoscopic radical hysterectomies for early-stage cervical cancer: a systemic review and meta-analysis

Abstract

Objective

Survival outcomes of robotic radical hysterectomy (RRH) remain controversial. Therefore, we performed a meta-analysis to evaluate survival outcomes between RRH) and laparoscopic radical hysterectomy (LRH) in patients with early-stage cervical cancer.

Methods

Studies comparing between RRH and LRH published up to November 2022 were systemically searched in the PubMed, Cochrane Library, Web of Science, ScienceDirect, and Google Scholar databases. Manual searches of related articles and relevant bibliographies of the published studies were also performed. Two researchers independently extracted data. Studies with information on recurrence and death after minimally invasive radical hysterectomy were also included. The extracted data were analyzed using the Stata MP software package version 17.0.

Results

Twenty eligible clinical trials were included in the meta-analysis. When all studies were pooled, the odds ratios of RRH for recurrence and death were 1.19 (95% confidence interval [CI]=0.91–1.55; p=0.613; I²=0.0%) and 0.96 (95% CI=0.65–1.42; p=0.558; I²=0.0%), respectively. In a subgroup analysis, the quality of study methodology, study size, country where the study was conducted, and publication year were not associated with survival outcomes between RRH and LRH.

Conclusion

This meta-analysis demonstrates that the survival outcomes are comparable between RRH and LRH.

Trial Registration

International Prospective Register of Systematic Reviews Identifier: CRD42023387916

Synopsis

Previous studies were underpowered to detect clinically meaningful difference in survival outcomes between robotic radical hysterectomy (RRH) and laparoscopic radical hysterectomy (LRH). We assessed the survival outcome between RRH and LRH in patients with early-stage cervical cancer. Survival outcomes are comparable between RRH and LRH.

INTRODUCTION

Cervical cancer is the third most commonly diagnosed cancer and the fourth leading cause of cancer-related death in women worldwide [1]. Radical hysterectomy and bilateral pelvic lymph node dissection are the standard procedures for patients with early-stage cervical cancer. A previous trial [2] demonstrated that minimally invasive radical hysterectomy (MIRH), including robotic radical hysterectomy (RRH) and laparoscopic radical hysterectomy (LRH), had lower rates of disease-free and overall survival than abdominal radical hysterectomy (ARH). Recently, MIRH is no longer the mainstream method because of its poor survival outcomes after the LACC trial. Chiva et al. reported that patients who underwent MIRH without a uterine manipulator had a similar disease-free survival rate to the ARH group [3]. Although a prospective study is warranted, MIRH without a uterine manipulator is currently considered an acceptable method. Currently, the RACC and ROCC trials, which compare the survival outcomes between RRH and ARH, are underway. Once their results are available, the status of minimally invasive surgery will likely be reevaluated. If there is no difference in survival outcomes between MIRH and ARH, MIRH is beneficial as it offers several advantages over ARH [4].

Laparoscopic surgery has limitations in the straight instrument, amplified tremor, and uncomfortable position of surgeons. The da Vinci robotic surgery system has gained the favor of surgeons due to its 3-dimensional definition vision, multi-degree-of-freedom rotatable wrist device, tremor filtration, and better ergonomics [5]. Previous studies [6,7,8,9,10] comparing RRH and LRH have reported no difference in survival outcomes between RRH and LRH. However, most of the studies [6,8,9,10] were retrospective, and the number of study subjects was relatively small to evaluate survival outcomes. Therefore, this study aimed to determine the survival outcomes between RRH and LRH in uterine cervical cancer through a meta-analysis.

MATERIALS AND METHODS

1. Literature search

The protocol was designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and registered in the International Prospective Register of Systematic Reviews (CRD42023387916). We conducted a comprehensive literature search using the PubMed, Cochrane Library, Web of Science, ScienceDirect, and Google Scholar databases for studies published from 2008 to November 2022 comparing RRH and LRH for the treatment of early-stage uterine cervical cancer. The following Medical Subject Heading terms and free-text terms were used for the search: “robotic radical hysterectomy,” “robotic surgery,” and “laparoscopic radical hysterectomy” in combination with “survival outcome,” “overall survival,” “disease-free survival,” “recurrence,” and “cervical cancer or carcinoma.” Additionally, we manually searched the references of the eligible studies to identify additional related studies.

2. Selection criteria

We included comparative studies on RRH and LRH in patients with early-stage cervical cancer. Studies assessing ARH and laparoscopic radical trachelectomy were excluded. Case series, editorials, letters to the editor, and conference abstracts were also excluded. Studies without available data on recurrence and death were excluded. We also excluded studies with fewer than 10 patients per arm. Non-English-language studies were excluded from the analysis.

3. Data extraction

Two researchers independently evaluated the eligibility of all the studies retrieved from the database based on the inclusion criteria. Any disagreements between authors were resolved through discussion. Among all articles found in the databases, we excluded duplicate articles and those that did not meet the inclusion criteria. We extracted the following data from the studies: first author, year of publication, published journal, year registered and enrolled, country where surgery was performed, study design, clinical stage (International Federation of Gynecology & Obstetrics 2008 staging system), study population, number of resected lymph nodes, overall follow-up period, and number of recurrences and deaths. When the follow-up period was different, the number of recurrences and deaths was corrected under the assumption that the hazard ratio was constant.

4. Main and subgroup analyses

We investigated the association between RRH and survival outcomes and compared the association between LRH and survival outcomes in the main meta-analysis. We also performed subgroup analyses for publication year, study size (small vs. large), quality of study methodology (high quality vs. low quality), and country to search for causes of heterogeneity. In this meta-analysis, matched retrospective and prospective cohort studies were classified as high quality, whereas retrospective studies were classified as low quality. When ≥40 patients were included in both the RRH and LRH groups, it was considered a large-scale study. The published articles were divided into the first and second halves of the time series. The countries were categorized as Asia, the USA, and Europe.

5. Statistical analyses

Data for dichotomous variables were analyzed using odds ratios (ORs); ORs with 95% confidence intervals (CIs) were calculated for recurrence and death after RRH and LRH. Forest plots were used to evaluate overall effects. We also performed a subgroup analysis to assess the effects of covariates, such as publication year, study size, quality of study methodology, and country where surgery was performed. Heterogeneity was assessed using the Higgins I² statistic and Cochran’s Q test [11]. The I² values ranged from 0% to 100%; we considered an I² value of >50% or a Q-test p-value of <0.1 as indicative of significant heterogeneity. We planned to estimate a pooled OR with a 95% CI using a fixed-effects model [12] in the absence of significant heterogeneity and a random-effects model [13] in the presence of significant heterogeneity. However, we only used the fixed-effects model because there was no heterogeneity in any of the analyses. Begg’s funnel plot and Egger’s test were used to assess for publication bias. All meta-analyses were performed using the Stata MP software package version 17.0 (StataCorp, College Station, TX, USA). The p-values of <0.05 were considered statistically significant.

RESULTS

After identifying potentially relevant studies from electronic databases and deleting duplicate articles, 95 articles were screened by the reviewers according to their titles, abstracts, and article types. Of these articles, 59 did not meet the selection criteria and were excluded. After reviewing the full texts of the remaining 36 articles, we found 20 relevant studies [14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33], published between 2008 and 2019, suitable for inclusion in the meta-analysis. Eleven out of the 16 studies excluded from the final review had insufficient data regarding survival outcomes. Other articles were excluded because they reported only abstraction (n=3), and 2 articles were excluded because the number of patients in one arm was less than 10. Two of the 20 relevant studies were excluded from the analysis of overall survival owing to unavailable data (Fig. 1).

Fig. 1. Flow diagram to identify relevant studies for the meta-analysis.

Table 1 presents the main characteristics of the 20 studies included in the final analysis. Their study design types were retrospective (n=12) [17,18,19,20,22,24,25,26,27,30,31,33], matched retrospective (n=2) [21,29], and prospective (n=6) [14,15,16,23,28,32]. The locations of these studies were Asia (n=8) [20,21,22,25,27,28,31,33], Europe (n=8) [17,18,19,23,26,29,30,32], and the USA (n=4) [14,15,16,24]. The countries with the largest representation in this meta-analysis were Italy (n=5) [18,23,26,29,30], South Korea (n=4) [21,22,27,31] and the USA (n=4) [14,15,16,24]. The enrollment period (year) of participants across the studies ranged from 2003 to 2017. We identified 1,050 patients who underwent RRH and 2,071 who underwent LRH. Out of the total 3,121 patients, 259 (8.3%) had recurrences (98 of the 1,050 patients underwent RRH, and 161 of the 2,071 patients underwent LRH), and 111 (3.6%) died of disease (38 of the 1,050 patients underwent RRH and 73 of the 2,071 patients underwent LRH). There were no statistically significant differences in the number of resected lymph nodes between RRH and LRH in all studies, except for 2 [16,24].

Table 1. Characteristics of studies included in the current meta-analysis (n=20).

Authors	Published year	Journal	Year enrolled	Country	Study design	Stage	Population		Follow-up (mo)
							RRH	LRH	RRH*	LRH*
Magrina et al. [14]	2008	Gynecol Oncol	2003–2006	USA	Prospective	IA2–IB2	27	31	31.1	31.1
Nezhat et al. [15]	2008	JSLS	2000–2008	USA	Prospective	IA2–IIA	13	30	12	29
Estape et al. [16]	2009	Gynecol Oncol	2006–2008	USA	Prospective	IA2–IB2	32	17	7.8	31.4
Lambaudie et al. [17]	2010	Eur J Surg Oncol	2007–Unknown	France	Retrospective	IA2–IVA	22	16	11.6	19.5
Tinelli et al. [18]	2011	Ann Surg Oncol	2003–2010	Italy	Retrospective	IA1–IIA	23	76	24.5	46.5
Gortchev et al. [19]	2012	Gynecol Surg	2006–2010	Bulgaria	Retrospective	IB1	73	46	10.5	51.1
Chen et al. [20]	2014	Int J Gynecol Cancer	Unknown	Taiwan	Retrospective	IA–IIB	24	32	13.9	34.6
Kim et al. [21]	2014	Int J Gynecol Cancer	2008–2013	ROK	Matched	IB1–IIA1	23	69	UK	UK
Yim et al. [22]	2014	Yonsei Med J	2009–2013	ROK	Retrospective	IA1–IIA2	60	42	25	19.5
Corrado et al. [23]	2016	Int J Gynecol Cancer	2010–2013	Italy	Prospective	IB2–IIB	41	41	30	60.1
Mendivil et al. [24]	2016	Surg Oncol	2009–2013	USA	Retrospective	IA2–IIB	58	49	39	39
Nie et al. [25]	2017	Int J Gynecol Cancer	2009–2016	China	Retrospective	IA2–IIA2	100	833	UK	UK
Pellegrino et al. [26]	2017	Acta Biomed	2010–2016	Italy	Retrospective	IA2–IIA1	34	18	59	30
Heo et al. [27]	2018	Obstet Gynecol Sci	2009–2013	ROK	Retrospective	IA1–IIB	19	22	UK	UK
Luo et al. [28]	2018	BMC Womens Health	2014–2015	China	Prospective	IA–IIB	30	30	UK	UK
Gallotta et al. [29]	2018	Eur J Surg Oncol	2010–2016	Italy	Matched	IA2–IIB	70	140	24	36
Corrado et al. [30]	2018	Int J Gynecol Cancer	2001–2016	Italy	Retrospective	IB1	88	152	46.6	41.7
Chong et al. [31]	2018	Int J Gynecol Cancer	2008–2013	ROK	Retrospective	IB1–IIA2	65	60	60	72
Ind et al. [32]	2019	Int J Med Robot	2010–2017	Unknown	Prospective	IB1	32	25	UK	UK
Chen et al. [33]	2019	World J Clin Cases	2014	China	Retrospective	IA1–IIB	216	342	29	30

LRH, laparoscopic radical hysterectomy; ROK, republic of Korea; RRH, robotic radical hysterectomy.

^*Median or mean.

Figs. 2 and 3 present the effect of RRH compared to LRH on the risk of recurrence and death in this meta-analysis of 20 and 18 studies, respectively. In a fixed-effects meta-analysis of 20 studies investigating recurrence, there was no statistically significant difference between RRH and LRH (OR=1.19; 95% CI=0.91–1.55; p=0.204). In a fixed-effects meta-analysis of 18 studies investigating death, there was no statistically significant difference between RRH and LRH (OR=0.96; 95% CI=0.65–1.42; p=0.204). There was no heterogeneity among studies of recurrences (I²=0.0%; p=0.827) or among studies of deaths (I²=0.0%; p=0.558). Funnel plot analysis showed a lack of publication bias in the incidence of recurrence and death (Fig. 4). Egger’s regression plot also demonstrated no publication bias. A cumulative meta-analysis of recurrence after RRH compared with LRH was performed. As shown in the cumulative meta-analysis, the summary estimate for recurrence tended to decrease over time (Fig. S1). Table 2 presents a summary of the subgroup analysis that assessed the potential effects of publication year, study size, quality of study methodology, and study location. There were no significant differences between RRH and LRH in the subgroup analysis categories.

Fig. 2. The forest plot shows the incidence of recurrences of RRH compared to LRH in a fixed-effect model of 20 studies. No significant between-study heterogeneity was detected (I²=0.0%; p=0.613) (OR=1.19; 95% CI=0.91–1.55; p=0.204).

CI, confidence interval; LRH, laparoscopic radical hysterectomy; OR, odd ratio; RRH, robotic radical hysterectomy.

Fig. 3. The forest plot shows the incidence of deaths of RRH compared to LRH in a fixed-effect model of 18 studies. No significant between-study heterogeneity was detected (I²=0.0%; p=0.558) (OR=0.96; 95% CI=0.64–1.42; p=0.827).

CI, confidence interval; LRH, laparoscopic radical hysterectomy; OR, odd ratio; RRH, robotic radical hysterectomy.

Fig. 4. Funnel plots of the incidence of recurrences (A) and deaths (B) of robotic radical hysterectomy compared to laparoscopic radical hysterectomy.

Table 2. Association between type of minimally invasive radical hysterectomy and recurrence in subgroup analysis (n=20).

Category		No. of studies	Summary OR (95% CI)	Heterogeneity, I2 (%)	p	Model used
All		20	1.19 (0.91–1.55)	0.0	0.613	Fixed effect
Published year
	Before 2015	9	1.30 (0.70–2.39)	0.0	0.874	Fixed effect
	After 2015	11	1.16 (0.87–1.56)	23.1	0.224	Fixed effect
Study size
	Large (both arm >40)	8	1.11 (0.81–1.52)	0.0	0.529	Fixed effect
	Small (at least 1 arm <40)	12	1.40 (0.86–2.29)	0.0	0.486	Fixed effect
Quality of study methodology
	High quality	9	1.33 (0.81–2.19)	0.0	0.479	Fixed effect
	Low quality	11	1.14 (0.83–1.55)	0.0	0.503	Fixed effect
Country
	Asia	8	1.23 (0.86–1.76)	0.0	0.544	Fixed effect
	USA	4	1.46 (0.52–4.10)	0.0	0.726	Fixed effect
	Europe	8	1.09 (0.71–1.67)	22.9	0.247	Fixed effect

CI, confidence interval; OR, odd ratio.

DISCUSSION

Radical hysterectomy is the standard treatment for early-stage cervical cancer. The LACC trial [2] and subsequent meta-analyses [34,35] demonstrated that the survival outcomes of MIRH were not as good as those of ARH. Subsequently, the number of ARH increased relative to MIRH [36]. The LACC trial has been criticized for various reasons, such as inadequate preoperative imaging and evaluation, the utilization of transcervical uterine manipulators, and insufficient tumor containment, which resulted in peritoneal contamination. Several studies [3,37] have shown no difference in survival between the 2 groups without the use of a cervical manipulator. Moreover, a prior study demonstrated no significant difference in survival rates between ARH and MIRH even when using a uterine manipulator in small cervical cancer <2 cm [34]. A vaginal cuff closure technique can be used to prevent tumor exposure and manipulation during MIRH [38]. Although current evidence favors ARH and the frequency of performing MIRH is relatively lower than before, MIRH remains a valid strategy in a limited range without compromising oncological outcomes. Several meta-analyses [34,35] have analyzed the survival outcomes of MRIH and ARH, but few have compared RRH and LRH.

RRH is superior to LRH in terms of estimated blood loss, but there is no difference in survival outcomes and perioperative complications between LRH and RRH. Radical hysterectomy for early-stage cervical cancer has a survival rate of more than 90%. As most of the studies were retrospective and the follow-up period was short, the number of patients who experienced recurrence may have been underestimated. Therefore, we attempted to confirm this finding through meta-analysis. We attempted a meta-analysis using the hazard ratio to compare the survival outcomes of RRH and LRH, but there were no studies presenting hazard ratios. Only 9 [18,19,21,23,24,26,27,30,31] of the 20 studies included in the meta-analysis showed survival curves. Therefore, we evaluated the OR for the risk of recurrence and death after RRH compared with that after LRH.

Guo et al. [39] performed a network meta-analysis to evaluate survival and surgical outcomes according to different surgical approaches for uterine cervical cancer. There were no significant differences in the recurrence or overall mortality between the RRH and LRH groups. However, they did not focus on survival outcomes. Zhang et al. [40] evaluated the efficacy of RRH for cervical cancer compared with LRH through a meta-analysis. They reported that there were no significant differences in recurrences between RRH and LRH using pooled data from 7 studies (OR=0.96; 95% CI=0.50–1.87; p=0.91). Meta-analysis for overall survival was not performed owing to limited data.

Our meta-analysis has some limitations. First, this meta-analysis failed to provide the highest level of evidence because of the retrospective nature of the included studies. Thus, selection bias and missing data may have reduced the quality of this study. Second, survival outcomes were calculated using ORs because most studies did not suggest hazard ratios or Kaplan-Meier curves. When the overall follow-up differed, the number of recurrences and deaths was revised under the assumption that the hazard ratio was the same. However, the overall follow-up period of each group was not specified in 30% of studies included in this meta-analysis. Third, the number of patients who underwent RRH in most studies was small, and the follow-up period was short because RRH was introduced later than LRH. The learning curve of early RRH might have influenced the results. Cumulative meta-analysis showed that the summary estimate of RRH compared to LRH for recurrence tended to decrease over time. A better survival outcome might be obtained in the RRH group if performed by surgeons who have mastered robotic surgery. Fourth, several techniques, such as routine use of a uterine manipulator [2,3] or an effect of the insufflation gas that might affect survival outcomes in MIRH, were not considered. Fifth, most patients had early-stage uterine cervical cancer from IA to IIA. However, among the 20 studies, 8 included stage IIB disease. Patients with stage IIB disease have been treated with radical hysterectomy in several countries. In the high-risk group with lymph node metastasis, parametrial involvement, positive resection margins, large tumors, and postoperative adjuvant therapies could have affected the results. Sixth, the mass size of cervical cancer, which could affect survival outcomes, was not considered. Finally, excluding languages other than English may have affected our outcomes.

To the best of our knowledge, this is the largest meta-analysis to focus on survival outcomes between RRH and LRH. The meta-analysis included 20 studies. Publication bias was minimal in this study. Although most of the included studies were retrospective and nonrandomized, this study included more than 3,121 patients and a comprehensive review of the available literature. The current meta-analysis demonstrated that oncological efficacy was comparable between RRH and LRH. Well-designed, prospective, randomized controlled trials are required to verify our findings.

SUPPLEMENTARY MATERIAL

Fig. S1

Cumulative meta-analysis of the incidence of recurrences of robotic radical hysterectomy compared to laparoscopic radical hysterectomy.