|Robotic versus laparoscopic proctectomy for rectal cancer: a meta-analysis
|Memon S, Heriot AG, Murphy DG, Bressel M, Lynch AC
This review found that robot assistance reduced the rate of conversion to open surgery, compared with conventional laparoscopic proctectomy for rectal cancer, while the other clinical and oncological outcomes were comparable. Due to limitations in the review, particularly the high risk of selection bias in the primary studies, these conclusions may not be reliable.
To compare the safety and efficacy of robot-assisted versus conventional laparoscopic proctectomy for rectal cancer.
MEDLINE and EMBASE were searched for studies published in English, from 2002 to March 2011. Search terms were reported and the reference lists of the selected articles were checked.
Studies comparing robot-assisted versus conventional laparoscopic proctectomy, for rectal cancer, that reported clinical or oncological outcomes, were eligible for inclusion. Proctectomy included low anterior resection, total mesorectal excision, coloanal anastomosis, and abdominoperineal resection. Studies had to use robotic systems that were manufactured at the time of this review. Studies of robotic or voice-activated endoscopes or remote tele-robotics, and studies of high anterior resection or rectopexy, were excluded.
The participants in the included studies had a mean body mass index of 23 to 27 kilograms per square metre, and most studies included more men than women (where reported). The rates of previous surgery ranged, across study groups, from 2% to 62%. Studies were set in Korea, Italy or the USA, and different surgical techniques were used in each setting. The control group underwent restorative low or ultra-low anterior resection, in most cases; some participants had abdominoperineal resection. Most studies used hybrid techniques for robotic rectal dissection, and for colonic mobilisation and vessel ligation in the control group. The clinical outcomes were the rate of conversion to open surgery, operation time, hospital stay, and complication rate. The oncological outcomes were circumferential margin involvement, number of lymph nodes collected, and distal resection margin distances.
The authors did not state how many reviewers selected the studies.
Assessment of study quality
Study quality was assessed, using an adapted published checklist, considering study design, objectives, randomisation, completeness of follow-up, blinding, baseline similarity of groups, and the intervention (was this the only difference in treatment between groups). One study with unequal proportions of restorative versus non-restorative surgery in the two groups was excluded after quality assessment.
The authors did not state how many reviewers assessed quality.
For binary outcomes, risk ratios, and for continuous outcomes, mean differences were extracted or calculated, with 95% confidence intervals. If there were no events in one treatment group, risk differences were calculated.
The data were extracted independently by two reviewers. Additional information was requested from study authors if required.
Methods of synthesis
The data were combined to calculate pooled risk ratios, mean differences, or risk differences, with 95% confidence intervals; the authors stated that they used a random-effects model. Heterogeneity was assessed with Ι². Sensitivity analysis was conducted by excluding one outlying study.
Results of the review
Seven case-control studies were included, with 754 participants (range 50 to 200). Two studies matched their participants before surgery; the others did not. Study quality was deemed satisfactory overall. All studies reported prospectively collected data on consecutive patients. The study groups were well-matched at baseline for most characteristics, except in three studies where they differed in tumour characteristics and/or preoperative treatment. All but one study gave details of the operative management, and in all studies, both groups underwent treatment during the same period, and at the same institution. All studies used intention-to-treat analysis and followed-up an adequate number of participants.
The rates of conversion to open surgery were significantly lower with robot-assisted resection than with conventional laparoscopic surgery (RD -0.07, 95% CI -0.12 to -0.01; seven studies; Ι²=80%). There was no significant difference between the two groups in operating time (seven studies; Ι²=95%) hospital stay (six studies; Ι²=68%), complications (seven studies; Ι²=29%), circumferential margin involvement (seven studies; Ι²=0), number of lymph nodes collected (number of studies not stated, Ι²=57%), and distal resection margin distances (seven studies; Ι² not reported).
The rate of ileostomy ranged from 5% to 94% and the leak rate ranged from 2% to 14%, across the seven studies.
Robot assistance reduced the rate of conversion to open surgery, compared with conventional laparoscopic proctectomy for rectal cancer, while the other clinical and oncological outcomes were comparable.
The objectives and inclusion criteria were clear. Relevant sources were searched, but this included only two databases and the restriction to published studies in English meant that some studies may have been missed. The potential for publication bias was not discussed. Steps were taken to minimise the risk of reviewer bias and error in data extraction but it was unclear whether this applied to study selection and quality assessment. The criteria for quality assessment were not clearly described, especially for the intervention, and they were more suitable for randomised controlled trials, than for case-control studies. Study quality was described as satisfactory, but the case-control design meant that there was a high risk of selection bias, as the authors acknowledged. The study follow-up rates were not reported.
The authors chose acceptable methods to combine the studies and assess heterogeneity, but it may not have been appropriate to pool the data for all outcomes, as there were high levels of heterogeneity, for some analyses, and this was largely unexplained. It was unclear why a fixed-effect model was used for one analysis, as the authors stated that a random-effects model was used. In the main paper, the authors suggested that their conclusions should be interpreted with caution, but this was not included in the abstract conclusions.
Due to limitations in the review, particularly the high risk of selection bias in the primary studies, the authors' conclusions may not be reliable.
Implications of the review for practice and research
Practice: The authors did not state any implications for practice.
Research: The authors stated that large randomised controlled trials were required to assess the clinical, functional and oncologic outcomes of robot-assisted compared with conventional laparoscopic proctectomy, for the management of rectal cancer.
Memon S, Heriot AG, Murphy DG, Bressel M, Lynch AC. Robotic versus laparoscopic proctectomy for rectal cancer: a meta-analysis. Annals of Surgical Oncology 2012; 19(7): 2095-2101
Subject indexing assigned by NLM
Case-Control Studies; Humans; Laparoscopy; Length of Stay; Lymph Nodes /pathology; Postoperative Complications; Prognosis; Rectal Neoplasms /pathology /surgery; Robotics
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.