A global systematic review and meta-analysis on laparoscopic vs open right hemicolectomy with complete mesocolic excision

Purpose The aim of this study was to compare the outcomes of right hemicolectomy with CME performed with laparoscopic and open surgery. Methods PubMed, Scopus, Web of Science, China National Knowledge Infrastructure, Wanfang Data, Google Scholar and the ClinicalTrials.gov register were searched. Primary outcome was the overall number of harvested lymph nodes. Secondary outcomes were short and long-term course variables. A meta-analysis was performed to calculate risk ratios. Results Twenty-one studies were identified with 5038 patients enrolled. The difference in number of harvested lymph nodes was not statistically significant (MD 0.68, − 0.41–1.76, P = 0.22). The only RCT shows a significant advantage in favour of laparoscopy (MD 3.30, 95% CI − 0.20–6.40, P = 0.04). The analysis of CCTs showed an advantage in favour of the laparoscopic group, but the result was not statically significantly (MD − 0.55, 95% CI − 0.57–1.67, P = 0.33). The overall incidence of local recurrence was not different between the groups, while systemic recurrence at 5 years was lower in laparoscopic group. Laparoscopy showed better short-term outcomes including overall complications, lower estimated blood loss, lower wound infections and shorter hospital stay, despite a longer operative time. The rate of anastomotic and chyle leak was similar in the two groups. Conclusions Despite the several limitations of this study, we found that the median number of lymph node harvested in the laparoscopic group is not different compared to open surgery. Laparoscopy was associated with a lower incidence of systemic recurrence. Supplementary Information The online version contains supplementary material available at 10.1007/s00384-021-03891-0.


Introduction
Colorectal cancer (CRC) is one of the most common healththreatening diseases around the globe. It is the third most frequent cancer worldwide (1.85 million new cases/year, representing 10.2% of total malignancies) and, according to predictions, is expected to increase in incidence by a further 20% before 2030 [1]. The number of annual worldwide deaths from CRC was approximately 880,000 in 2018, with an increasing trend year on year [2].
The prognosis of this disease is strongly related to the stage at the time of diagnosis, with a 5-year survival rate of around 90% when the cancer is diagnosed at an early stage, compared with 13% when the presentation is delayed and metastatic disease is present [2]. In particular, lymph node involvement determines important variations in outcome, with overall 5year CRC survival at 59% [3]. Patients with stage II and stage III CRC treated with potentially curative surgery will still sadly die of the disease in up to 30% of cases at 5 years. At least part of this may be determined by understaging of the disease due to an insufficient lymph node yield [4], as this factor is important in determining subsequent oncologic adjuvant treatment [5].
Surgery is the mainstay of potentially curative treatment, also playing a central role in staging. Standard segmental colectomy with D2 lymphadenectomy is based on the oncological principle that local control of disease determines survival, with lymphadenectomy mainly meant for prognosis rather than cure [6]. In fact, the indication for adjuvant chemotherapy is based on several factors, including nodal status, and may provide a reduction in mortality by up to 30%. According to guidelines [7], a minimum of 12 lymph nodes should be analysed for an accurate staging of the disease, whilst an understaging may result in patients not receiving adjuvant therapy.
It is in this context that in 2009 Hohenberger proposed to extend the lymphadenectomy, describing complete mesocolic excision (CME) for the treatment of cancer of the caecum and ascending colon [8]. In subsequent years, the technique spread, and it was later adapted to laparoscopy which had become the accepted standard of care in the surgical treatment of colon cancer. The aim of this systematic review and metaanalysis is to compare short-term and long-term outcomes of individuals with right colon cancer undergoing treatment by open or laparoscopic right hemicolectomy with CME.

Materials and methods
We performed a systematic review of the literature, which was searched up to 20 March 2020, according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [9], including the following databases: Medline/PubMed, Scopus, Web of Science (WOS), China National Knowledge Infrastructure (CNKI, 中国知网), Wanfang Data (万方)) and Google Scholar.
The following search statement was used in Medline/ PubMed:

Eligibility criteria
Studies that compared participants enrolled for either laparoscopic or open right hemicolectomy with CME were eligible for inclusion. Randomized controlled trials (RCTs) and clinical control studies (CCTs) (prospective and retrospective cohort studies) were included. Case report studies were excluded.

Study selection and data extraction
Two reviewers (RC and SZ) performed the search independently. A third author (AA) arbitrated any disagreements on inclusion or exclusion of studies. The reference lists of the included studies were searched manually. Only the data for patients who had undergone open or laparoscopic CME were included.

Outcomes of interest
Primary outcome was the overall number of harvested lymph nodes. Secondary outcomes were local and systemic recurrence at 3 and 5 years, operative time, post-operative mortality at 30 days, overall post-operative complications at 30 days, estimated blood loss, surgical intraoperative complications (vascular injuries, iatrogenic small bowel perforation), anastomotic leak rate, chyle leak rate, post-operative hospital stay, post-operative ileus, wound infections and pulmonary infections.

Quality assessment
All studies fulfilling the selection criteria for this systematic review and meta-analysis were assessed for methodological quality and risk of bias by two authors (RC and GA). The individual scores of quality assessment items per study were assessed using the Cochrane risk tool for Randomised Control Trials [10,11] and the methodological index for nonrandomised studies (MINORS) [12].

Statistical analyses
All analyses were performed according to original treatment allocation (intention-to-treat analysis). The categorical variables were described as absolute/relative frequencies and the continuous ones as median and interquartile range (IQR). Data were analysed for risk ratios (RR) in the case of dichotomous variables, and weighted mean differences (WMD) for continuous variables. The randomised Mantel-Haenszel method was used for the meta-analysis. All results were displayed in Forest plots. The I 2 and the Cochrane's Q were reported as statistical measures of heterogeneity. For outcomes with more than ten studies, funnel plots are shown. Other statistical measures of bias are not reported given the high study heterogeneity. The data analysis was performed using the meta-analysis software Review Manager (RevMan) v 5.3.5 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2018).
In this meta-analysis, a subgroup analysis according to the types of study design was performed. We also identified the studies which had both the largest variance (wide intervals) and the extreme outlier weight in each clinical outcome group.

Study selection
The PRISMA flow diagram is presented in Fig. 1. The initial search produced 2055 studies. After removal of duplicates, 721 citations remained. After screening of titles and abstracts, 36 studies were analysed in full text, with 12 studies excluded as reported in SDC1, leaving 23 studies matching the inclusion criteria for meta-analysis , one PhD thesis by El Nakeeb (not published) [13]. Of these, two references [15,16] report the same data as well as the PhD thesis [13]. In addition, we found one ongoing study (registered in ClinicalTrials.gov as NCT03826446) [48].
Among the examined studies, patients with stage IV disease were excluded, except for one [18]. The other studies include stages I, II and III, but two [14,25] only included stage III cancers and two [22,26] only included stages II and III; in addition, five studies [15,[28][29][30]32] do not indicate the TNM stage.
Concerning tumour localisation, all of the studies considered cancers of cecum, ascending colon and hepatic flexure, while only five studies [17,20,22,31,32] also included proximal transverse colonic neoplasms.
For what it concerns conversion from laparoscopic to open resection, most of the studies excluded the patients which required this procedure for any reason. From those other studies that considered in their data also converted surgeries, we excluded the converted procedures in our analysis.

Quality assessment
We assessed the risk of bias for each trial and summarised them using the criteria and the 'Risk of bias' Table (SDC 3). No data were reported on random sequence generation or allocation concealment in the RCT, while blinding of participants and personnel was not reported. A 'low risk of bias' was reported in the analysis of the attrition bias and an 'unclear risk of bias' for selective reporting. Risks of bias assessed with the

Discussion
The CME technique as part of a right hemicolectomy is technically difficult because of the large variability in the right colon vasculature, which may be the cause of challenging intraoperative bleeding complications. Moreover, CME should be performed by surgeons with a significant experience in colorectal surgery. It is important to recall that the localisation of the disease influences the extent of lymphadenectomy in the peripancreatic area: for experts of the CME technique, cancers located in the hepatic flexure and in proximal transverse colon require extension to the Henle trunk stations.
Several authors have published their data and results in the execution of CME in a laparoscopic setting, which clearly requires further experience in advanced laparoscopic surgery. The choice between laparoscopy or open surgery for CME remains a point of discussion and interest. Moreover, some argue that CME may be  [49,50], and further data are required to assess the utility of robotics in this setting.
We compared laparoscopic and open CME in terms of the number of lymph nodes harvested. Moreover, we evaluated the differences in long-term prognosis considering the 3-year and 5year post-operative overall survival and disease-free survival. Local control of the disease is improved by an increased number of harvested lymph nodes [7]. The samples of patients were comparable for age, gender, BMI, ASA and TNM stage. In addition, despite the majority of the studies indicated the TNM stage, the exact localisation of primary tumour was often not reported. This clearly represents a limit in the literature available regarding this topic. Finally, raw data or hazard ratios of survival curves, both overall and disease-free, would be more useful to understand the true equivalence of laparoscopic and open techniques in terms of oncologic outcome. During the entire analysis, a significant heterogeneity was often observed, so that we preferred to use, as standard, random effects. For these reasons, we would need an Individual Patient Database (IPD) or, at least, the hazard ratios of the published studies.
Laparoscopic CME may allow more lymph nodes to be harvested than in open surgery. Nevertheless, laparoscopy appeared superior to open surgery in terms of overall recurrence at 3 and 5 years, the only measurable parameter related to oncologic adequacy. However, there needs to be caution in interpretation of this result due to the possible influence of selection bias.
Improved staging of the disease does allow patients to be considered for the most appropriate therapeutic approach, in particular the receipt of adjuvant chemotherapy. Han et al. [33] also showed how laparoscopy resulted in an increased harvested number of so-called principal lymph nodes (along the course of the superior mesenteric artery), this, being more radical, and possibly aided by the improved magnification of the surgical field offered by laparoscopy.
Besides extended lymphadenectomy, CME requires the mesocolic fascia to be excised intact, in order to guarantee advantages in term of oncological radicality and survival [8,[51][52][53]. In our systematic review and meta-analysis, we did not report data about the quality of the surgical specimen, since only five studies [20,23,26,28,30], among those included, explicitly reported this feature. In fact, most of the authors, although describing in detail the technique used and highlighting the importance of the integrity of the mesocolon, focused their attention on the number of harvested lymph nodes in order to evaluate the local control of the disease. This bias could derive from low initial experience of pathologists in the examination of the mesocolic fascia, in a similar way that happened at the beginning of the application of total mesorectal excision (TME) for rectal cancer. However, it remains technically imperative to maintain the integrity of the mesocolic fascia during CME in order to maximise appropriate oncological radicality.
It is important to remember that laparoscopic approach may fail and require conversion to open surgery because of various reasons: the main factor that implicates conversion is a vascular injury that leads to uncontrollable bleeding or that cannot be detected in its source; other conditions that could require the change of procedure could be the excess of adhesions or an organ injury hard to repair laparoscopically. In case of conversion to the open approach, there are consequences for the patient that must be considered and discussed before choosing this method: in particular, the outcome of the patient worsens in term of longer and complicated post-operative stay, the possible need for post-operative ICU and longer operative time. These complications are a direct consequence of the conversion and are not a consequence of the open technique, that is why it is very important to select the patients that will undergo laparoscopic surgery [54,55].
The presented data show laparoscopy to be at least non inferior to open surgery when performing CME for cancer of the right colon, with other benefits of a minimally invasive approach also confirmed. Moreover, the laparoscopy group resulted in a lower 5-year systemic recurrence rate, even if this outcome was reported by only a few studies: the results might be considered promising but will need further confirmation by the enlargement of the cohort of studies. Laparoscopy offered better short-term outcomes including overall complications, lower estimated blood loss and lower wound infection rates, altogether translating into a shorter hospital stay. This appears at the expense of a longer operative time, likely due to the higher technical difficulty associated with the surgery, in particular when a laparoscopic extended right hemicolectomy is performed for tumours located at the hepatic flexure or proximal transverse colon. These endpoints are similar to other meta-analyses [37,38] regarding this topic, and a possible explanation is that the laparoscopic technique is harder to perform and requires a longer learning curve. No difference was observed in terms of post-operative mortality, anastomotic leak, chylous leak or pulmonary infection, which is reassuring.
The current study has several limitations. Although this systematic review included the highest number of studies reported in literature, the major limitation is represented by the geographical distribution of the studies. In fact, all but one studies which was performed in Germany [18] were published by non-western groups: 16 studies were from Asia (13 from China, 2 from South Korea and 1 from Bangladesh), 1 from Africa (Egypt) and 2 East Europe (Russia). The importance of this limitation is related to the possibility of differences not only in surgical standards but also in standards of adjuvant therapy and health care systems, difficult to assess. A second limitation is that among these studies only one is an RCT. Therefore, the meta-analysis is based on non randomised controlled trials. It would have been interesting to perform a subgroup analysis for TNM/UICC stage associated with the rate of laparoscopic and open surgery, in order to assess possible advantages of laparoscopy in terms of recurrence for a specific tumour stage which requires adjuvant chemotherapy. A third limitation is the impossibility to assess and meta-analyse the data regarding the number of cases in which the minimum of 12 lymph nodes were harvested, corresponding to current guidelines [7]. In order to achieve both these analyses, probably an Individual Participant Database analysis should be planned. Furthermore, it is likely that the control group (open surgery) is often not treated by a well-trained surgeon, being the operation time quite long (median 173 min). Finally, despite Hohenberger [8] recommended a minimum of 25 lymph nodes harvested, the overall median in the open group is only 21.8, with 5 papers reporting even a median of 15 or less.
Another developing field in the group of minimally invasive techniques is robotic surgery: some studies [56,57] analysed the feasibility, safety and effectiveness of robotic CME, which could be considered as another possible technique to assist CME and having potential for improving the outcomes and quality of life of patients, such as many studies also suggest for laparoscopic procedures as compared to traditional open surgery [58,59]. However, data are still missing, and the number of studies needs to be increased in order to evaluate this alternative minimally invasive approach.
In conclusion, the present systematic review and metaanalysis fails to show that a laparoscopic approach appears non-inferior to open surgery in terms of lymph nodes harvested but showed a benefit in terms of incidence of systemic recurrence, when performing right hemicolectomy with CME for colon cancer, while maintaining the usual benefits of a minimally invasive surgical approach.
Nevertheless, further prospective and randomised studies are awaited to increase the quantity of data and quality of the evidence. Funding Open access funding provided by Università degli Studi di Ferrara within the CRUI-CARE Agreement.

Supplementary Information
Data availability All data and materials as well as software application or custom code support the published claims and comply with field standards.

Declarations
Ethics approval and consent to participate (Not applicable)

Conflict of interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.