The present study demonstrates a significant reduction in length of hospital stay, intraoperative blood loss and inflammatory stress response measured by CRP in patients undergoing colorectal resection for cancer with RCS compared to the LCS. A multivariate regression analysis with adjustment for ASA-score, T-stage and neoadjuvant chemotherapy demonstrated an additional reduction in the operative time and the amount of harvested lymph nodes favoring RCS. No difference was found in surgical or medical morbidity, time to first flatus or stool, conversion rate to open surgery, postoperative leukocyte count or the microradical resection rate between the two groups in multivariate analyses.
Comparison of RCS and LCS for malignant disease remains poorly investigated. In 2012, a prospective randomized controlled trial examined the length of stay as a primary outcome and reported no difference between the two surgery techniques (2). The ROLARR study from 2017 demonstrated that the risk of conversion was higher for obese patients (BMI > 30 kg/m2) undergoing surgery for rectal cancer (7).
Recent systematic reviews and meta-analyses comparing surgical efficacy and safety of RCS and LCS favor RCS concerning several intra- and postoperative outcomes. These outcomes include lower conversion rates, intraoperative blood loss, decreased overall morbidity, earlier hospital discharge, and earlier establishment of bowel function (8, 13, 14, 21–23). In this study, there was a non-significant conversion rate of 1.4% in the RCS group and 3.9% in the LCS group, p = 0.187. Solaini et al. reported a significant conversion rate to open surgery in LCS resections (RR 1.7; 95% CI (1.1–2.6), p = 0.02). This finding has also been confirmed in other systematic reviews and meta-analyses (3, 8, 24–27). Despite more patients in our population receiving combined chemo-/radiotherapy, no difference in conversion rates could be detected between the surgical groups. However, conversion rates for patients receiving neoadjuvant radiotherapy were higher. Factors that may complicate the surgical procedure and induce this higher conversion rate include radiotherapy to the pelvic floor causing fibrosis, edema, inflammation and necrosis (28). Neoadjuvant radiotherapy is associated with a higher risk of postoperative surgical complications and delayed perineal wound healing following abdominoperineal resection (29). None of these complications were overrepresented in our study.
In contrast to other studies, operative time was no longer for RCS than LCS (4, 10, 24–26, 30). However, previous meta-analyses have been conducted on predominantly observational studies whereby there is a risk of misinterpretation of total procedure times and surgical times (31–33). The widely criticized fact by RCS is the setup and docking time of the robotic console. The prolonged surgical times cannot be attributed to simple factors but rather the joint effort of the limited number of certified RCS surgeons, dedicated operating nurses and/or anesthesiology team. Improved RCS technological development and a transition from DaVinci Si ® to the Xi ® model and a dedicated robotic team can reduce total operation time. Several studies have confirmed a reduction in both docking and total operating times using the DaVinci Xi ® model, with an average of 21 cases needed to reach a statistically significant reduction in the docking time (34–36). The technological development of RCS and the da Vinci Xi ® model's introduction allows the surgeon more freedom and the ability to perform even technically demanding procedures that previously have been difficult to perform laparoscopically. The stable and precise high definition camera, which the surgeon independently can maneuver and the higher degree of free movement of robotic arm joints contributes to better hemostasis (37). Although both surgical modalities are minimally invasive, RCS is presumed to be associated with a gentler manipulation with organs.
Postoperative CRP levels were significantly lower in RCS and can predict the inflammatory stress response induced by surgery. There is sparse literature reporting on the systemic inflammatory response in RCS compared to LCS (38). Previous studies have mostly compared the systemic inflammatory response in RCS to open colorectal surgery. RCS was associated with a lower inflammatory stress response compared to open surgery (39, 40). A prospective, non-randomized study comparing RCS (n = 30) and LCS surgery (n = 120) for early gastric adenocarcinoma reported a lower postoperative CRP and interleukin-6 response in the LCS group. In this study, there was an unequal distribution of patients, lack of randomization and usage of older da Vinci robotic ® technology. These factors may have contributed to a lower postoperative inflammatory response in the LCS group.
Regarding oncological outcomes, the microradical resection rate was not statistically different between groups. Our study showed a significantly lower amount of harvested lymph nodes in LCS group from the multivariate analysis. Many existing reports examine the rate of harvested lymph, and the majority report no differences between the two operation methods (2, 10, 25–27, 41, 42). However, a large Danish register-based observational study including a total of 8104 LCS and 511 RCS procedures for colorectal cancer showed the risk of achieving a microradical resection in colon cancer was significantly higher using LCS, and higher but non-significant for rectal cancer in patients undergoing RCS (43).
The most important limitation of this study is the retrospective design. Patients were preoperatively balanced between the two surgical procedure groups regarding age distribution, BMI and ASA-score. There was a selection of patients who had received neoadjuvant oncological treatment in favor of LCS. These patients did not have a significantly increased rate of conversion or postoperative morbidity in either of the surgical groups. To minimize the risk of selection bias, we performed a multivariate regression analysis adjusting for clinically relevant confounders. Apart from time to first stool, none of the univariate analyses were non-significant by these adjustments.