Robotic gastrectomy for gastric cancer: feasibility and outcomes of a newly established program

INTRODUCTION

Gastric cancer is the fifth most diagnosed malignancy worldwide, with significant variation in incidence across geographic regions. Eastern Asia reports the highest rates of gastric cancer, whereas all regions of Africa have among the lowest. Despite a global decline in incidence - suspected to be linked to improved food preservation methods and decreasing prevalence of Helicobacter pylori infection - gastric cancer remains the fifth leading cause of cancer-related mortality globally^[1].

The surgical management of gastric cancer, along with the technologies used to perform it, has evolved considerably over time. While ongoing clinical trials such as CRITICS, SENORITA, TOPGEAR, PILGRIM, and RENAISSANCE^[2-6] continue to refine and optimize treatment strategies, the standard approach for T1b–T4 tumors remains partial or total gastrectomy with an adequate D2 lymph node dissection, or a minimum of 16 lymph nodes evaluated in the specimen as outlined in current NCCN guidelines^[7]. Historically, this procedure was carried out via an open approach, and is inherently complex due to meticulous dissection, intestinal reconstruction, and thorough nodal evaluation.

In recent decades, the emergence of minimally invasive techniques has challenged the traditional open approach by offering potential benefits such as reduced postoperative pain, faster recovery, and shorter hospital stays. However, performing a gastrectomy laparoscopically significantly increases the technical demands of the procedure. These challenges hindered the widespread adoption of laparoscopic gastrectomy in many surgical centers^[8].

The introduction of robotic-assisted surgery for gastric cancer in 2002 marked a pivotal advancement^[9]. The robotic platform offers enhanced dexterity through multi-directional articulating instruments, improved instrument stability, tremor filtration, and superior three-dimensional camera visualization and 10× magnification. These features help overcome many of the limitations associated with conventional laparoscopy and have facilitated the more widespread use of minimally invasive techniques in complex gastric cancer surgery.

This study retrospectively evaluates the peri-operative and oncologic outcomes of robotic vs. open gastrectomy for gastric cancer, as performed by a single institution transitioning to robotic surgery without prior experience in complex laparoscopy.

METHODS

RESULTS

DISCUSSION

In this single-center retrospective analysis comparing robotic and open gastrectomy for gastric adenocarcinoma, 56 cases were reviewed to assess differences between minimally invasive and traditional surgical approaches. Accordingly, this study is best interpreted as an evaluation of early program implementation rather than a definitive comparative effectiveness analysis. Although speculative, the open cohort may have represented a higher-risk subset, as early implementation of a robotic program often favors patient selection with fewer comorbidities, risk factors, or prior abdominal operations to minimize conversion risk and support a safer learning curve. Nonetheless, baseline demographic and clinical characteristics were largely comparable between groups, with the exception of diabetes, which was more prevalent in the open cohort (44.8% vs. 18.5%, P = 0.047). Additionally, a greater proportion of patients in the robotic group received neoadjuvant chemotherapy (70.4% vs. 44.8%, P = 0.064) and neoadjuvant radiation (25.9% vs. 10.3%, P = 0.171), though these differences did not reach statistical significance. No additional differences were observed in comorbidities, prior intra-abdominal surgery, or type of gastrectomy performed.

Three robotic cases (11.1%) required conversion to an open procedure. One conversion was necessitated by dense adhesions in a patient with prior open cholecystectomy and inguinal hernia repair. The second occurred due to intra-operative bleeding and limited visualization despite no prior abdominal surgery. The third involved extensive fibrosis and poor visualization in a patient with a history of jejunostomy tube placement. These cases underscore the technical challenges inherent to the early adoption of robotic gastrectomy, particularly in patients with complex surgical histories or altered anatomy. The observed conversion rate in this series (11.1%) was slightly higher than national averages reported in larger cohorts, which range from 3.8% to 7.8%^[10,11]. Prior studies have identified factors such as large tumor size, total gastrectomy, and bulky nodal disease or adjacent organ invasion as predictors of conversion^[10,11]. Notably, other institutional experiences have demonstrated that the standardization of operative techniques and accumulation of surgeon experience can substantially reduce conversion rates, from 46.0% to 15.8%^[12], underscoring the impact of structured program development on procedural success.

In the present study, several peri-operative outcomes were observed to favor the robotic approach over open gastrectomy, including reduced EBL, fewer operative complications, and shorter postoperative hospital stay. EBL was significantly lower in the robotic cohort compared with the open group (100 vs. 175 mL, P = 0.025). Although our findings did not replicate the extremely low blood loss reported in some series (as low as 25 mL)^[13], they are consistent with results from a randomized controlled trial comparing robotic and open gastrectomy, which demonstrated mean blood loss values of approximately 124 and 276 mL, respectively^[14]. These observed results are consistent with larger series as well^[15-17], suggesting the ability to safely perform gastrectomy using a robotic approach, even during early adoption phase.

Interestingly, the observed reduction in operative complications in the robotic group was not driven by wound-related morbidity but rather by differences in specific intraoperative and technical complications. In our cohort, injuries to adjacent organs accounted for the majority of complications and occurred exclusively in the open group, suggesting a potential advantage of the robotic platform in facilitating precise dissection and improved visualization in confined operative planes. Similarly, anastomotic leaks were observed only in the open cohort, although the absolute number of events was small.

These findings differ somewhat from prior large database studies, which have primarily demonstrated reductions in systemic and postoperative complications - such as surgical site infections, pulmonary complications, and transfusion requirements - rather than intraoperative injuries^[18]. In the pooled analysis by Solaini et al., overall complication rates for robotic gastrectomy were comparable to those observed in our study, with low reported rates of anastomotic leak (~2%-3%) and no clear difference between approaches^[10]. Likewise, Garsot et al. found similar rates of anastomotic and major complications between open and minimally invasive approaches^[19].

The predominance of adjacent organ injuries and anastomotic complications in the open cohort of our study may reflect technical factors, including exposure and visualization, but may also be influenced by case selection and the learning curve associated with early adoption of a robotic program. Therefore, while our findings suggest that robotic gastrectomy may reduce certain intraoperative technical complications, these results should be interpreted with caution and validated in larger cohorts.

Additionally, robotic gastrectomy was associated with a much shorter postoperative hospital stay (7 [4-21] vs. 9 [5-38] days, P = 0.025). The IMIGASTRIC registry demonstrated a median hospital stay of 8 days for robotic cases compared to 11 days for open surgery (P < 0.0001)^[16], while the recent network meta-analysis by Manara et al. (2024) similarly found that robotic total gastrectomy was associated with faster postoperative recovery relative to open procedures^[17]. Nakauchi et al. (2021), in a multi-center U.S. analysis, also reported that minimally invasive approaches - including robotic - were associated with reduced perioperative morbidity and shorter hospitalization without compromising oncologic outcomes^[15]. These consistent findings likely stem from the reduced surgical trauma inherent to minimally invasive techniques, which utilize smaller incisions, magnified three-dimensional visualization, and enhanced instrument dexterity to minimize tissue disruption and accelerate recovery.

In contrast, the open approach had shorter operative time when compared to its robotic counterpart (median 255 vs. 320 min, P = 0.012). This finding aligns with the randomized controlled trial by Ribeiro et al., which reported mean operative durations of 354 min for robotic and 214 min for open gastrectomy^[14]. Prolonged operative times for robotic surgery have also been consistently observed in other studies and meta-analyses^[16,20].

When evaluating lymphadenectomy, the open and robotic group demonstrated very similar results (20 [8-52] vs. 19 [7-50], P = 0.328). Our results do contrast with recent multi-center analyses demonstrating equivalent or superior lymph-node retrieval with robotic gastrectomy. Trastulli et al. (2023) reported comparable nodal yields and lower postoperative morbidity in the robotic group (16.2% vs. 24.1%, P = 0.017)^[16], while Hirata et al. (2023) found a higher median node count and greater likelihood of achieving ≥ 16 nodes in robotic cases (19 vs. 16, P < 0.001)^[21]. These discrepancies likely reflect institutional learning curves, procedural standardization, and variation in the routine adoption of formal D2 dissection.

Emerging strategies such as sentinel lymph-node (SLN) navigation surgery may further refine the balance between oncologic adequacy and surgical invasiveness. The SENORITA randomized trial reported no significant difference in 5-year disease-free, overall, or disease-specific survival between standard laparoscopic gastrectomy and SLN navigation surgery^[22], suggesting that tailored nodal dissection strategies may play an increasing role in early-stage disease.

When comparing long-term outcomes, the 5-year estimated overall survival for the entire cohort was 59.8% (95%CI, 44.1-75.5), with a higher, though not statistically significant, rate observed in the robotic group compared with the open group (66.1% vs. 52.3%, P = 0.166). These findings are reassuring and consistent with previously published series reporting 5-year survival rates of approximately 50%-65%, with robotic gastrectomy demonstrating comparable or, in some cases, superior outcomes relative to the open approach^[10,19]. However, our study is underpowered to detect differences in long-term survival.

Overall, our findings underscore that the transition from open to robotic gastrectomy involves a distinct and measurable learning curve. Importantly, these results demonstrate that robotic gastrectomy can be implemented safely, with perioperative and oncologic outcomes comparable to the open approach. Although minimally invasive techniques may offer advantages in postoperative recovery, achieving consistent oncologic equivalence - particularly in lymph-node harvest - depends on accumulated surgical experience, procedural standardization, and institutional support. While some outcomes were favorable in the robotic cohort, these findings are best interpreted as supporting feasibility and safety rather than establishing superiority. The evolution of robotic prostatectomy provides an instructive parallel: while early laparoscopic prostatectomy offered clinical benefits, its technical complexity limited widespread adoption until the introduction of robot-assisted platforms, which reduced operative times and improved oncologic outcomes as experience accrued^[23].

Drawing from this precedent, the gastric cancer surgical community can benefit from implementing structured, competency-based training models to accelerate safe and effective adoption of robotic techniques. A modular curriculum incorporating simulation, mentorship, and proctoring could help mitigate early technical challenges and shorten the learning curve. Adapting evidence-based frameworks from urologic robotic surgery education - such as multimodal simulation, stepwise supervised progression, and video-based performance review - may further promote earlier attainment of proficiency while maintaining patient safety^[24].