Robotic versus laparoscopic Gastrectomy for gastric cancer: a systematic review and updated meta-analysis

Laparoscopic gastrectomy (LG) has been widely used for the treatment of gastric cancer and a number of other different minimally invasive procedures have been developed to date [1, 2]. There are several benefits for patients; including better cosmesis, reduced pain, early recovery of intestinal function, and shorter hospital stay, while maintaining comparable oncologic safety [1‐4].

Robotic surgery was first put into practice in 2000, after being approved by the US Food and Drug Administration (FDA). It plays an essential role in ergonomics and offers advantages such as motion scaling, less fatigue, tremor filtering, seven degrees of wrist-like motion, and three-dimensional vision [5, 6]. Surgeons hoped that such innovative technology could overcome some limitations innate to traditional laparoscopic surgery. Thus, experienced laparoscopic surgeons are increasingly trying to develop new procedures that best exploit the capabilities of robotic surgery in the treatment of gastric cancer [7].

Nonetheless, the present status of robotic gastrectomy (RG) is, as of the writing of this paper, still restricted and this is in part to due to the lack of randomized controlled trials (RCTs). Several previous studies including meta-analyses have argued that RG can be a more effective and safer operation in comparison with conventional LG. In spite of these studies, many questions still need to be answered, most notably, RG’s efficacy with regard to oncologic, long-term survival outcomes and its cost-effectiveness. Moreover, a series of studies on RG for the treatments of gastric cancer have been recently published. These studies are meaningful in highlighting the status of RG in the treatment of gastric cancer. Therefore, this paper’s current research is intended to conduct a comprehensive systematic review of all the currently available literature and a meta-analysis of RG in comparison to LG in order to assess the feasibility, security and efficacy of RG.

The cost-effectiveness and definite advantages of RG have not been well documented, which is different when compared to the evolution of LG versus conventional open surgery [20]. However, the number of publications on RG has gradually increased in recent years. The oncologic outcomes, postoperative outcome, intraoperative effects and costs of a total of 1830 patients who underwent RG for gastric cancer treatment in 19 studies were reviewed as we believe such research would contribute to a more objective and comprehensive assessment of the current RG surgical status.

In spite of the considerable heterogeneity, the prolonged operating time in RG was shown in almost all the included research studies. The prolonged exposure time to pneumoperitoneum and the associated increased time of anesthesia is a major concern. Few publications describe the effect of longer operation times during RG. However, previous research of LG in senior patients has shown that longer operation time did not result in detrimental effects with regard to surgical results [41]. Therefore, a prolonged operating time should not affect surgeons directly on conducting research on RG’s new utility. Inevitably, the docking time was considered as an essential factor, which enhanced the operating time. The docking time was between 20 min to 60 min as reported in our study [7, 13, 15, 31], We found RG had longer operation times than LG by 49 min, which suggested the ‘true’ time spent on operations was similar or even shorter than LG. Furthermore, with the increased utilization of the new robotic surgical system, operation times are expected to shorten. Several studies have reported that the da Vinci Xi robotic platform is more user-friendly and is easier to install in rectal and nephritic surgery [42, 43]. As a result, we believed that RG is technically feasible with regard to operation time.

Surgeons have to go through a learning curve to master a technique. The surgical results, such as operation time, oncological outcomes and postoperative complications can be affected by surgeon’s familiarity with the instrument, experience and assistant compliance. In general, before stabilization, LG should be conducted on around 40 to 60 cases [44]. The learning curve for RG was shorter for experienced surgeon who had performed LG, which is forecasted to be only 10 to 20 cases [12, 13, 18, 26]. A surgeon experienced in laparoscopic surgery can conduct robotic surgery securely even in their first case [16]. Several studies investigated in this meta-analysis compared the initial and later experiences of robotic surgery [12, 13, 18, 26]. The later cases performed by the same surgical team could progress toward shortening operation times.

Postoperative morbidity is the main indicator for assessing the safety and feasibility of one procedure. It is widely accepted that laparoscopic surgery for gastric cancer is safer and could have fewer complications than open surgery [45]. Our meta-analysis demonstrated a comparable complication rate in RG versus LG group, and the low heterogeneity regardless of overall, surgical or medical complications encourages us to believe that RG indeed is as safe as LG. Improvements such as three-dimension images and tremor filtering could theoretically contribute to safer implementations of the robotic system for gastrectomy and lymphadenectomy. According to the multivariate analyses in the Suda study, the application of RG was an important independent protective factor in regards to the postoperative complication [37]. Tokunaga et al. [46, 47] reported the incidences of overall adverse events after RG which were 14.2% and 22.2% based on their two-phase II studies, which are comparable to the rates of 19–27% in previous studies of LG [48, 49].

Obesity is one of the most significant health problems today and rates are still increasing around the world. Some studies claim obesity causes increased blood loss, operation time, and wound infection rate et al. [50, 51], whereas others did not observe any negative effect on surgical outcomes [52]. Recently, Harr et al. [10] showed that the benefits of robotic methods were more evident in high versus normal BMI patients when performing a colostomy. The authors concluded that robotic surgery might overcome the difficulties associated with thick abdominal walls and excessive intra-abdominal fat, thanks to improved visualization, instrumentation, and ergonomics [10]. However, compared to other operations such as the colorectal or prostatic surgeries, which are in relatively narrow regions, the superiority of da Vinci over the laparoscopy may not be obvious, in that gastric surgery is conducted in the upper abdomen of a relatively spacious location. In our study, the overall mean operation time of RG and LG were similar in the overweight subgroups, contrasting with those in the non-overweight subgroups, which implied RG to be superior to LG when used on overweight patients. However, the sample size of the overweight subgroups was not large enough to be conclusive.

The traditional straight forceps in LG fail to enable surgeons to reach deep-seated vessels and other areas, like the supra pancreatic one, in which the dissection of lymph nodes around the splenic hilum, splenic artery, and hepatic artery areas is deemed extremely hard. The tremor filtering, wristed instruments, as well as stable exposure and high-solution image can help surgeons thoroughly retrieve the lymph nodes around the delicate areas [21]. According to one included study, the amount of RLNs was considerably higher with robotic surgery in the splenic hilum and splenic artery areas [29]. Our meta-analysis shows adequate RLNs with means of 35.4 and 36.1 in the LG and RG groups, respectively. The mean number of RLNs of RG was more than that of LG with a marginal difference observed in the pooled data, even though most studies had been done during initial implementation of the robotic technique. Therefore, we believe that robotic technique could be superior to the conventional laparoscopic technique for lymphadenectomy. Since the history of the clinical application of RG is a short one, few reports have compared long-term survival outcomes with other methods. Coratti et al. [53] demonstrated that the 5-year survival rate after RG stratified with Stage IA, IB, II, and III was 100%, 84.6%, 76.9%, and 21.5%, respectively. Pugliese et al. [11] reported a cumulative overall 5-year survival rate of 78% with a mean follow up of 30 months (range 2–86) after RG for gastric cancer.

The application of robotic surgery remains controversial, mainly due to the considerable expense. The total difference in cost between the LG and RG groups has been predicted to be around 3900 USD [18, 31, 38], which is mainly derived from the robotic system itself. According to the opinions of some investigators, the higher cost of robotic surgery is not enough to justify the theoretical advantages of this technology [54]. If RG can reduce complications and shorten hospital stay, the higher costs of the robotic system would be partially offset. Based on this, it is essential for robotic operators to inspect whether the potential advantages of the robotic approach justifies its high cost in the treatment of gastric cancer.

Our research has the following limitations: (1) Selection bias: As no RCT was available to be included in the meta-analysis due to the higher cost of robotic surgery, selection biases are inevitable in surgical abstention which should be carefully interpreted. (2) Clinical heterogeneity: The homogeneity test for the continuous variables exhibited substantial heterogeneity due to the inherent flaws of a retrospective study, the uneven surgical skills of the different surgeons, as well as regional differences, etc. More importantly, for surgeons in the East, radical distal gastrectomy for middle and distal gastric cancer is popular [55], while the distal subtotal is preferred in the West [56]. Thus we cataloged distal gastrectomy and subtotal gastrectomy as a subgroup. Though it brings some interesting results due to the expansion of sample size, such a combination would result in clinical heterogeneity. (3) Regional difference: The majority of the included studies came from East Asia, because East Asia has the highest prevalence of gastric cancer, while gastric cancer is relatively uncommon in Western countries. Besides, in East Asia, particularly Korea, Japan and some areas of China, the proportion of early gastric cancer has increased as a result of the improved surveillance of gastric cancer in these regions [57, 58]. On the other hand, although increasing evidence continues to show no difference between patients undergoing open or laparoscopic surgery for oncologic outcomes, the Japanese Gastric Cancer Association still classifies minimal invasive surgery as investigational treatment and only recommends minimal invasive surgery for early stage gastric cancer patients [55]. Therefore, the cases in our studies, especially those from East Asia, were mainly early stage cases. All of the above limitations must be kept in mind when interpreting the results of our study.

Except for the longer operation time and higher costs, RG for the patients with gastric cancer was not inferior to LG. Besides; RG holds the potential benefits for larger numbers of lymph node dissection and reduced intraoperative blood loss. Further prospective studies are needed in order to confirm these advantages. In addition, long-term results are needed, particularly for the oncological adequacy of robotic gastric cancer resections.