Abstract
Background and Purpose:
While multiple Asian and a few Western retrospective series have demonstrated the feasibility and safety of robotic-assisted gastrectomy for gastric cancer, its reliability for thorough resection, especially for locoregional disease, has not yet been firmly established, and reported learning curves vary widely. To support wider implementation of robotic gastrectomy, we evaluated the learning curve for this approach, assessed its oncologic feasibility, and created a selection model predicting the likelihood of conversion to open surgery in a United States patient population.
Methods:
We retrospectively reviewed data on all consecutive patients who underwent robotic gastrectomy at a high-volume institution between May 2012 and March 2019.
Results:
Of the 220 patients with gastric cancer selected to undergo curative-intent robotic gastrectomy, surgery was completed using robotics in 159 (72.3%). The median number of removed lymph nodes was 28, and ≥ 15 lymph nodes were removed in 94% of procedures. Surgical time decreased steadily over the first 60–80 cases. Complications were generally minor; 7% of patients experienced complications of grade 3 or higher, with an anastomotic leak rate of 2% and mortality rate 0.9%. Factors predicting conversion to open surgery included neoadjuvant chemotherapy, BMI ≥ 31, and tumor size ≥ 6 cm.
Conclusions:
These findings support the safety and oncologic feasibility of robotic gastrectomy for selected patients with gastric cancer. Proficiency can be achieved by 20 cases and mastery by 60–80 cases. Ideal candidates for this approach are patients with few comorbidities, BMI < 31, and tumors < 6 cm.
Mini-Abstract
We retrospectively reviewed perioperative and oncologic outcomes of robotic gastrectomy for gastric cancer in a Western cohort over the initial period of its use at a high-volume center. This approach can be mastered within 60–80 cases while maintaining sound oncologic outcomes. Guidelines for appropriate selection of robotic cases are proposed.
Introduction
Minimally invasive surgical approaches (MIS) have been widely adopted as an alternative to open surgery for the treatment of selected early gastric cancers because they are associated with faster postoperative recovery and fewer complications. For patients with more advanced disease requiring perioperative systemic treatment, these advantages of MIS may facilitate indicated adjuvant systemic therapy1. Laparoscopic gastrectomy has been established to provide short-term benefits including shorter hospital stay, fewer complications, lower analgesic requirements, reduced blood loss, and improved overall quality of life compared with open gastrectomy1–5. Multiple retrospective studies, prospective studies, randomized clinical trials, and meta-analyses from Asia, Europe, and the United States have shown that minimally invasive and open gastrectomy have oncologically equivalent outcomes with fewer complications and shorter recovery time1,6–16. However, the restricted freedom of movement inherent within laparoscopic gastrectomy make it technically challenging to master17,18.
The introduction of robotics for gastric adenocarcinoma helps to overcome the difficulties of laparoscopic gastrectomy by providing three-dimensional, high-definition, stable, and magnified views of the operative field and instruments that can move with 7 degrees of freedom. The feasibility and safety of robotic-assisted gastrectomy for gastric cancer has been demonstrated in one prospective19 and several retrospective studies20 by showing similar rates of overall complications, mortality, and number of harvested lymph nodes to those of laparoscopic gastrectomy. Limited conclusions can be drawn regarding the reliability of robotic gastrectomy for gastric cancer, particularly in Western populations, because of considerable variability among these studies in inclusion criteria, surgeon experience, type of reconstruction performed, and the outcomes evaluated.
An important question regarding robotic procedures for gastric cancer is the duration and variability of time required to learn the technique. Robotic surgery is assumed to be easier to learn than laparoscopic gastrectomy due to ergonomic and technical advantages. Some Asian studies have suggested that learning the robotic approach may require as few as 20 to 2521,22 or up to 95 cases17; this wide variability may result from disparities in the stage of presentation and the incidence of various subtypes of gastric cancer. Differences in comorbidities and body mass index23 between Asia and the West may also affect the learning curve. Given the high degree of interest from surgeons to learn and appropriately apply this technique, the purpose of this analysis is to describe our experience adopting robotic gastrectomy for selected patients at a high-volume center in the U.S. Within this selected cohort, we also aimed to estimate the learning curve, assess the surgical safety and oncologic feasibility of robotics for curative-intent gastrectomy, and create a model to predict the likelihood of conversion to open surgery to guide patient selection in the West. This information may support the broader implementation of robotic gastrectomy beyond high-volume centers.
Methods
Following IRB approval, data were collected from our institution’s prospective database for patients undergoing robotic gastrectomy from 2012–2019. All patients eligible for curative intent gastrectomy by an MIS approach were offered a robotic gastrectomy with the understanding that patient safety, oncologic resection factors, and prolonged operative times would result in a low threshold for conversion to an open resection. Patients receiving neoadjuvant chemotherapy, or with elevated BMI or comorbidities that were determined to be acceptable candidates for surgery were nonetheless considered for a robotic approach. Relative contraindications included extra-gastric organ invasion, poor performance status, history of major intraabdominal surgeries, and large tumor size (> 6 cm) or extensive linitis plastica. Although younger, earlier-stage patients were preferentially selected for the robotic approach at the start of the learning curve, after about 30 cases, all patients who otherwise met above inclusion criteria were offered the robotic approach. All procedures were performed by the same high-volume gastric surgeon, who had performed over 100 laparoscopic gastrectomies prior to beginning robotic surgery and had no other prior robotic experience. All patients who went to the operating room for a planned robotic assisted gastrectomy were included in the analysis. Patients were excluded if their resection was palliative. Patient records were reviewed to determine clinicopathologic features and outcome. Complications within 30 days postoperatively were recorded and reviewed by an MD according to the Clavien-Dindo classification. Complications were reported as the highest grade complication noted for each patient.
Definitions
Length of stay was defined as the number of days from the beginning of the operation to the time of discharge, where ≤ 24 h was classified as 1 day. Tumor size was the maximum tumor dimension on final pathology of the resected specimen. Receipt of neoadjuvant chemotherapy or chemoradiotherapy prior to operation was recorded, but specific systemic treatments were not included in the analysis. The da Vinci robot system (Si or Xi) was used for all procedures.
Surgical Details
The technique for MIS gastrectomy as performed by these authors has been previously described24. Briefly, robotic gastrectomy is performed with the patient positioned supine on a split-leg table in a reverse Trendelenburg position. Port placement includes the camera port at the umbilicus, 2 8-mm Da Vinci ports on the left side, a 12-mm port in the right mid-clavicular line, and a 5-mm assistant port further laterally on the right side. Placement of a Nathanson liver retractor, via a small subxiphoid stab-wound incision, facilitates retraction of the left lateral lobe of the liver. First, the greater omentum is retracted cephalad and the lesser sac is entered. The omentectomy is carried up toward the spleen to collect the level 2, 4, and 6 lymph nodes. The right gastroepiploic vessels and corresponding lymph nodes are identified and dissected circumferentially at the level of the superior border of the pancreas at their point of origin from the gastroduodenal vessels with a stapler or clips. The lymphatic tissue along the hepatic proper and common hepatic artery is swept medially toward the specimen and a window is created at the level of the pylorus. The posterior aspect of the pylorus and proximal duodenum is gently elevated, an endovascular stapler is introduced, and the proximal duodenum is stapled distal to the pylorus. The dissection is continued along the common hepatic artery toward the celiac axis and proximal splenic artery. The left gastric vein and artery are identified at the celiac axis and all surrounding lymph nodes carefully swept up en bloc with the specimen. The vessels are then divided. The level 1 and 3 lymph nodes are dissected with the proximal stomach up to the right crus of the diaphragm and esophagus. The distal esophagus or stomach (depending on the tumor) is then divided with a stapler. At this point, the specimen is removed via the umbilical port site. A Billroth II reconstruction is performed for tumors below the incisura of the stomach and a Roux-en-Y esophagojejunostomy for more proximal or gastroesophageal junction tumors. A segment of jejunum approximately 30–40 cm downstream from the ligament of Treitz is selected based on mobility and tension-free reach to the transected stomach or esophageal hiatus and is used for reconstruction. After the jejunojejunostomy creation, the Roux limb is prepared for the esophagojejunostomy via a circular 25 French EEA stapler or linear stapled anastomosis for total gastrectomy, and by linear stapled anastomosis for Billroth II or Roux-en-Y reconstruction of subtotal gastrectomy.
Statistical Methods
Patients whose robotic operation was completed were included in the learning curve analyses. Learning curve for the entire cohort of completed robotic gastrectomies was assessed by 2 outcomes: time to transection of the pylorus, selected as a surrogate because it is a standard part of distal, subtotal, and total gastrectomies and is not affected by the extent of planned resection or type of reconstruction, and total robot time (dock to undock). Learning curve was also assessed for each type of reconstruction (Roux-en-Y or Billroth-II) using time from incision to closure of all incisions as the endpoint.
All patients were included in the predictors of conversion analysis. A classification tree approach, based on rigorous statistical methodology25, was chosen instead of regression because of the strong interactions between predictors; the strength of association between the outcome and a given predictor was highly sensitive to the value of other predictors26. Data were not sufficient to include interaction terms in a regression model, and classification trees handle interactions more effectively, although they do not provide odds ratios or allow assessment of variable independence27. Classification trees do, however, provide thresholds for creating subgroups, which may differ among parts of the tree. The tree was cross-validated to prevent overfitting and used to identify terminal groups, each of which was associated with a risk of conversion. All patients who fell into a terminal group that had a risk of conversion higher than the prevalence of conversion in our cohort were considered “high risk.” Based on the “high risk” and “low risk” groups, the positive and negative predictive value of the model was calculated.
All analyses were performed in R and the classification tree was developed using Rpart library.
Results
Between May 2012 and March 2019, 220 consecutive patients were taken to the operating room for a planned robotic curative-intent gastrectomy. Patient characteristics are described in Table 1; characteristics of patients who underwent open gastrectomy are provided for reference in Supplemental Table 1. Surgery was completed using robotics in 159 patients (72.3%) and was converted to open gastrectomy in 61 (27.7%). No conversions occurred emergently, and none occurred for reasons of intraoperative bleeding. The decision to convert was typically made based on limited linitis plastica, making manipulation of the stomach cumbersome, or inability to visually assess the proximal extent of tumor involvement to guide resection margin, such as in poorly differentiated tumors and those with more extensive submucosal spread, particularly following neoadjuvant chemotherapy. Postoperative complications occurred in 19% of patients, and there was a 5.5% readmission rate (n = 12 patients) within 30 days postoperatively.
Table 1. Clinical, operative, and pathologic characteristics.
Categorical data are n (%) and continuous data are median (range). Complication grades are according to the Clavien-Dindo classification.
| All patients (n = 220) | ||
|---|---|---|
| Age | 59.0 (22–83) | |
| Male gender | 111 (50.5%) | |
| BMI | 25.8 (16.9–44.9) | |
| Length of stay (days) | 5 (3–38) | |
| Neoadjuvant radiation | 0 (0.0%) | |
| Total gastrectomy | 91 (41.4%) | |
| Roux-en-Y | 133 (60.5%) | |
| D2 | 211 (95.9%) | |
| Time to pylorus transection (min) | 48 (15–111) | |
| Total robotic time (min) | 177 (30–350) | |
| Total operative time (min) | 205 (79–390) | |
| Lymph nodes retrieved | 28 (4–93) | |
| T4a | 22 (10.0%) | |
| Mixed | 51 (24.3%) | |
| Not reported | 2 (0.9%) | |
| Negative final margin | 217 (98.6%) | |
| Duodenal stump leak | 0 (0%) | |
| Readmission within 30 days postoperatively | 12 (5.5%) | |
Learning Curve
For all completed cases (n = 159), time to transection of pylorus decreased by 2.3 minutes for each 10 additional cases (p < 0.001, Fig. 1A) saving over 20 minutes over the course of the learning curve (Fig. 1B), and plateaued at approximately 40–45 minutes after 60–80 cases, compared with approximately 60–65 minutes in the initial 20. Total robot time decreased at similar rates for both types of reconstruction; by 3.3 minutes for each 10 additional cases for Billroth-II (p = 0.001, Figure 1C) and 3.6 minutes for each 10 additional cases for Roux-en-Y (p = 0.003, Fig. 1D).
Figure 1. Robotic gastrectomy learning curve.
A, Time to pylorus for all completed cases, by 20 cases. B, Time to pylorus for all completed cases, by case, color-coded by reconstruction type. C, Total robot time for Billroth-II reconstructions, by 20 cases. D, Total robot time for Roux-en-Y reconstructions, by 20 cases.
Oncologic Adequacy of Robotic Gastrectomy
The median number of lymph nodes retrieved was 28, and ≥ 15 nodes were retrieved in 94% of patients. The one case with < 15 lymph nodes retrieved was a prophylactic total gastrectomy for a CDH1 mutation. Three patients had microscopically positive margins on final histopathological examination of resected specimens following negative intraoperative frozen section analysis.
Factors Predicting Conversion
Factors predicting conversion included neoadjuvant chemotherapy (p < 0.001), higher BMI (p = 0.005), and larger tumor size (p = 0.04). Using data on all patients in the database, a statistically generated classification tree was created using the factors that were predictive of conversion to stratify patients into conversion risk groups (Fig. 2). The strongest predictor of conversion was neoadjuvant chemotherapy. Increasing BMI was the second strongest predictor of conversion in both the neoadjuvant chemotherapy and the non-chemotherapy groups, although the BMI threshold was lower in the chemotherapy group (31 vs. 37). Among patients who did not receive neoadjuvant chemotherapy, tumor size did not predict conversion and BMI alone was the risk stratifying factor.
Figure 2. Classification tree for risk of conversion to open gastrectomy.
Each branch point represents a risk factor (neoadjuvant chemotherapy, higher BMI, and larger tumor size).
Those with a BMI ≥ 37 had a 57% risk of conversion, while those with a BMI < 37 had a 14% risk of conversion. The combination of neoadjuvant chemotherapy and BMI ≥ 31 yielded a 78% risk of conversion. For those who received neoadjuvant chemotherapy and had a BMI < 31, risk of conversion could be stratified further based on tumor size, where tumor size ≥ 6 cm carried a 67% risk of conversion, and tumor size < 6 cm had a 40% risk.
Discussion
Our analysis confirms that in a Western cohort, robotic gastrectomy is safe, achieves oncologic goals for the treatment of properly selected patients with gastric cancer, and can be learned over a defined number of cases.
Oncologic Adequacy of Robotic Gastrectomy
Data on the long-term outcomes of robotic-assisted gastrectomy are limited but suggest acceptable survival and recurrence rates, with no difference in rates of negative margins and completeness of lymphadenectomy28. In our study, lymph node retrieval met oncologic goals, with a median number of 28 removed nodes and all procedures performed for the treatment of existing gastric cancer yielding 15 or more lymph nodes. In the one case in which a distal margin was found to be positive intraoperatively, the distal margin (adjacent to the common bile duct) was also involved. Because achieving a negative margin would have required a more radical procedure, the operation was completed as planned and adjuvant therapy was administered for the remaining disease. In this situation, the margin status was not felt to be related to the approach but rather to tumor biology. In the other two cases of positive margins, the intraoperative frozen sections were negative, and the final pathologic results revealed additional tumor at or near the margin. In one case, the patient had advanced disease with positive lymph nodes. Following adjuvant chemotherapy and radiation, there were no signs of local recurrence by two years postoperatively.
Selection Criteria
To help guide selection of patients for the robotic approach in this Western cohort, one surgeon trained in the robotics offered this approach to all patients in whom complete resection was deemed possible, including those with high BMI (> 35) and those who underwent neoadjuvant treatment. This approach allowed us to define optimal criteria without affecting safety or oncologic outcomes, and patients were informed that the threshold to convert would be low if a safer and better cancer operation would be possible with an open approach, such as for larger tumors or for submucosal tumors where the proximal margin could not be well visualized, explaining the high conversion rate in this series.
Our analysis of all robotic cases in the database identified 3 factors most highly predictive of conversion to an open gastrectomy: neoadjuvant chemotherapy, higher BMI, and larger tumor size, with receipt of neoadjuvant chemotherapy being the strongest predictor. Larger tumor size affected conversion due to the bulky nature of the tumor and the resulting difficulty in manipulating the stomach and surrounding vessels without compromising a safe oncologic resection. Neoadjuvant chemotherapy’s influence on conversion likely relates to the resulting changes in peri-tumoral connective tissue, enlarged liver size, and obfuscation of tumor margins. This loss of visual definition of the proximal border of the tumor after systemic chemotherapy, especially in patients with poorly differentiated and signet ring cell tumors, was the most common reason for conversion to an open approach. These cases ones are now approached robotically in only a few select circumstances.
The highest risk cohort were patients who received neoadjuvant chemotherapy and had a BMI ≥ 31, a population that is highly prevalent in Western practice, but rare in Asia, where BMIs are generally lower and neoadjuvant chemotherapy is not standard practice. In these groups, tumor size did not influence risk of conversion. Interestingly, patients with tumors < 6 cm had a lower risk of conversion than patients who did not receive chemotherapy but who had a BMI ≥ 37 (40% vs. 57%), suggesting that body habitus influences the ability to perform a successful minimally invasive operation for gastric cancer cases. Patients with higher BMI were less amenable to a robotic approach because of concern over tension in the Roux limb of the esophagojejunal anastomosis due to excess intra-abdominal adipose tissue. In these cases, an open approach is necessary to allow adequate transillumination, and often Doppler ultrasound, of the mesenteric vessels to facilitate lower-tension anastomosis.
Because operating time at the beginning of adoption of robotic-assisted gastrectomy is longer than for open gastrectomy,29 patient selection should be more stringent early in a surgeon’s experience if their experience with MIS approaches is limited. Ideal candidates for robotic-assisted gastrectomy are those with few medical comorbidities, low or normal BMI, small tumors, distal tumors, and those who have not received neoadjuvant chemotherapy, although these are not absolute contraindications. Distal tumors are particularly amenable to early learning of the robotic approach due to the need for only one anastomosis rather than two for Roux-en-Y reconstructions. Robotic-assisted gastrectomy is also suitable for patients with a CDH1 mutation, as these patients are recommended to undergo prophylactic total gastrectomy due to their approximately 55–70% lifetime risk of developing diffuse gastric adenocarcinoma30. Relative contraindications may include significant intra-abdominal adhesions, large tumor size, or invasion into adjacent organs. As for any minimally invasive surgery, dense intra-abdominal adhesions can prevent clear visualization of important structures and therefore compromise the operation. As a surgeon gains experience with the robotic platform and instrumentation improves, individual surgeons’ inclusion criteria will likely expand to include patients with more advanced disease and higher BMI.
Learning Curve
For this Western cohort of patients, the steepest part of the learning curve was achieved during the first 20 cases, reflected mostly in decreased time to set up and dock the robot. By 60–80 cases mastery was achieved for subtotal and total gastrectomy as reflected by further and sustained decreases in time, despite the undertaking of more complex cases. Other studies, mostly from Asia, have reported slightly shorter learning curves for robotic surgery, ranging from 40–60 cases29,31,32. This difference may result from various factors, such as the lower BMI of all Asian cohorts reported. Our study found that higher BMI complicates surgery and significantly increases the risk of conversion to an open approach. For the learning curve and time analysis, all procedures undergoing Roux-en-Y reconstructions were analyzed separately from those undergoing Billroth II reconstructions because of the additional anastomosis and time needed for creation of a Roux-en-Y. Total robot time decreased at a similar rate for Roux-en-Y and Billroth-II reconstruction cases (3.6 vs. 3.3 minutes per 10 cases). In addition, the surgeon’s threshold to take on more complex cases with higher BMIs and more advanced tumors increased with experience, which may confound the learning curve data for some of the later cases, but this learning experience is representative of real-life practice and experience.
We chose not to use cumulative sum plots to assess the learning curve, as our data are not compared with an external standard. Such an analysis would be misleading in this case, when the standard would be internal, such as an average over the course of the study.
Although no study has prospectively compared the learning curves of surgeons for robotic versus laparoscopic gastrectomy at the beginning of their minimally invasive experience, it has been suggested that with the appropriate formal simulation training on the robotic platform in both dry and wet labs, experienced open surgeons may be able to transition directly to the robotic platform without laparoscopic experience33.
Limitations
This study’s findings were unique to the learning curve and operative selection criteria for Western gastric cancer patients, for whom considerations likely differ from those for Asian patients with regards to body habitus, comorbidities, utilization of neoadjuvant chemotherapy, and perhaps tumor epidemiology such as a higher proportion of diffuse-type tumors. Additional studies are needed to fully appreciate the clinical benefits of the robotic approach, particularly as they relate to recovery and initiation of adjuvant treatment, which may affect recurrence-free survival (RFS), as well as long-term oncologic outcomes, complications, and quality of life.
Supplementary Material
Table 2. Descriptive characteristics grouped by conversion status.
Categorical data are n (%) and continuous data are median (range).
| Completed robotic (n = 159) | Converted to open (n = 61) | P value | |
|---|---|---|---|
| Age (years) | 59 (22–83) | 62 (22–81) | 0.411 |
| Gender (male/female) | 73/86 | 38/23 | 0.043 |
| BMI | 25.3 (16.9–39.2) | 26.4 (17.2–44.9) | 0.047 |
| Length of stay (days) | 5 (3–38) | 6 (4–38) | 0.019 |
| Neoadjuvant chemotherapy (yes/no) | 41/118 | 37/24 | < 0.001 |
| Operative characteristics | |||
| Procedure type (distal/total gastrectomy) | 98/61 | 31/30 | 0.192 |
| Type of reconstruction (Billroth-II/Roux-en-Y) | 79/80 | 8/53 | < 0.001 |
| Lymphadenectomy (D1/D2) | 8/151 | 1/60 | 0.603 |
| Time to pylorus transection (min) | 47.5 (15–111) | 54 (24–96) | 0.421 |
| Total robotic time (min) | 179.5 (35–350) | 87.5 (30–201) | < 0.001 |
| Total operative time (min) | 209 (126–390) | 181 (79–329) | < 0.001 |
| Final margin status (negative/positive) | 156/3 | 61/0 | 0.562 |
| Lymph nodes retrieved | 27 (4–78) | 30 (7–93) | 0.110 |
| T4a | 12 (7.6%) | 10 (16.4%) | |
| Not reported | 5 (3.1%) | 5 (8.2%) |
Synopsis.
We retrospectively reviewed perioperative and oncologic outcomes of 220 consecutive robotic gastrectomies for gastric cancer at a single United States institution over the initial period of adoption. This approach can be mastered within 60–80 cases while maintaining oncologic feasibility.
Acknowledgment
The authors acknowledge Jessica Moore for her valuable contributions in reviewing and revising this manuscript. We would also like to acknowledge Mike Patane, who was instrumental in refining this technique in the early stages of adoption, and Drs. Daniel Coit and Murray Brennan for their support and encouragement in pursuing this expanding technical field in the pursuit of optimal patient care.
Financial disclosures:
Mark Schattner is a consultant for Boston Scientific; no other authors have financial relationships to disclose. This work was financially supported in part by the NIH/NCI Cancer Center Support Grant P30 CA008748.
References
- 1.Kelly KJ, Selby L, Chou JF, et al. Laparoscopic Versus Open Gastrectomy for Gastric Adenocarcinoma in the West: A Case-Control Study. Ann Surg Oncol. 2015;22(11):3590–3596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Yang SY, Roh KH, Kim YN, et al. Surgical Outcomes After Open, Laparoscopic, and Robotic Gastrectomy for Gastric Cancer. Ann Surg Oncol. 2017;24(7):1770–1777. [DOI] [PubMed] [Google Scholar]
- 3.Li HZ, Chen JX, Zheng Y, Zhu XN. Laparoscopic-assisted versus open radical gastrectomy for resectable gastric cancer: Systematic review, meta-analysis, and trial sequential analysis of randomized controlled trials. J Surg Oncol. 2016;113(7):756–767. [DOI] [PubMed] [Google Scholar]
- 4.Zou ZH, Zhao LY, Mou TY, et al. Laparoscopic vs open D2 gastrectomy for locally advanced gastric cancer: a meta-analysis. World J Gastroenterol. 2014;20(44):16750–16764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee HJ, Hyung WJ, Yang HK, et al. Short-term Outcomes of a Multicenter Randomized Controlled Trial Comparing Laparoscopic Distal Gastrectomy With D2 Lymphadenectomy to Open Distal Gastrectomy for Locally Advanced Gastric Cancer (KLASS-02-RCT). Ann Surg. 2019;270(6):983–991. [DOI] [PubMed] [Google Scholar]
- 6.Hu Y, Ying M, Huang C, et al. Oncologic outcomes of laparoscopy-assisted gastrectomy for advanced gastric cancer: a large-scale multicenter retrospective cohort study from China. Surg Endosc. 2014;28(7):2048–2056. [DOI] [PubMed] [Google Scholar]
- 7.Hur H, Lee HY, Lee HJ, et al. Efficacy of laparoscopic subtotal gastrectomy with D2 lymphadenectomy for locally advanced gastric cancer: the protocol of the KLASS-02 multicenter randomized controlled clinical trial. BMC Cancer. 2015;15:355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Inaki N, Etoh T, Ohyama T, et al. A Multi-institutional, Prospective, Phase II Feasibility Study of Laparoscopy-Assisted Distal Gastrectomy with D2 Lymph Node Dissection for Locally Advanced Gastric Cancer (JLSSG0901). World J Surg. 2015;39(11):2734–2741. [DOI] [PubMed] [Google Scholar]
- 9.Katai H, Mizusawa J, Katayama H, et al. Short-term surgical outcomes from a phase III study of laparoscopy-assisted versus open distal gastrectomy with nodal dissection for clinical stage IA/IB gastric cancer: Japan Clinical Oncology Group Study JCOG0912. Gastric Cancer. 2017;20(4):699–708. [DOI] [PubMed] [Google Scholar]
- 10.Kim HH, Han SU, Kim MC, et al. Prospective randomized controlled trial (phase III) to comparing laparoscopic distal gastrectomy with open distal gastrectomy for gastric adenocarcinoma (KLASS 01). J Korean Surg Soc. 2013;84(2):123–130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kim HH, Hyung WJ, Cho GS, et al. Morbidity and mortality of laparoscopic gastrectomy versus open gastrectomy for gastric cancer: an interim report--a phase III multicenter, prospective, randomized Trial (KLASS Trial). Ann Surg. 2010;251(3):417–420. [DOI] [PubMed] [Google Scholar]
- 12.Kim KH, Kim MC, Jung GJ, Kim HH. Long-term outcomes and feasibility with laparoscopy-assisted gastrectomy for gastric cancer. J Gastric Cancer. 2012;12(1):18–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kim W, Kim HH, Han SU, et al. Decreased Morbidity of Laparoscopic Distal Gastrectomy Compared With Open Distal Gastrectomy for Stage I Gastric Cancer: Short-term Outcomes From a Multicenter Randomized Controlled Trial (KLASS-01). Ann Surg. 2016;263(1):28–35. [DOI] [PubMed] [Google Scholar]
- 14.Kim YW, Baik YH, Yun YH, et al. Improved quality of life outcomes after laparoscopy-assisted distal gastrectomy for early gastric cancer: results of a prospective randomized clinical trial. Ann Surg. 2008;248(5):721–727. [DOI] [PubMed] [Google Scholar]
- 15.Strong VE, Devaud N, Allen PJ, Gonen M, Brennan MF, Coit D. Laparoscopic versus open subtotal gastrectomy for adenocarcinoma: a case-control study. Ann Surg Oncol. 2009;16(6):1507–1513. [DOI] [PubMed] [Google Scholar]
- 16.Vinuela EF, Gonen M, Brennan MF, Coit DG, Strong VE. Laparoscopic versus open distal gastrectomy for gastric cancer: a meta-analysis of randomized controlled trials and high-quality nonrandomized studies. Ann Surg. 2012;255(3):446–456. [DOI] [PubMed] [Google Scholar]
- 17.Kim HI, Park MS, Song KJ, Woo Y, Hyung WJ. Rapid and safe learning of robotic gastrectomy for gastric cancer: multidimensional analysis in a comparison with laparoscopic gastrectomy. Eur J Surg Oncol. 2014;40(10):1346–1354. [DOI] [PubMed] [Google Scholar]
- 18.Jung DH, Son SY, Park YS, et al. The learning curve associated with laparoscopic total gastrectomy. Gastric Cancer. 2016;19(1):264–272. [DOI] [PubMed] [Google Scholar]
- 19.Kim HI, Han SU, Yang HK, et al. Multicenter Prospective Comparative Study of Robotic Versus Laparoscopic Gastrectomy for Gastric Adenocarcinoma. Ann Surg. 2016;263(1):103–109. [DOI] [PubMed] [Google Scholar]
- 20.Shen WS, Xi HQ, Chen L, Wei B. A meta-analysis of robotic versus laparoscopic gastrectomy for gastric cancer. Surg Endosc. 2014;28(10):2795–2802. [DOI] [PubMed] [Google Scholar]
- 21.Kang BH, Xuan Y, Hur H, Ahn CW, Cho YK, Han SU. Comparison of Surgical Outcomes between Robotic and Laparoscopic Gastrectomy for Gastric Cancer: The Learning Curve of Robotic Surgery. J Gastric Cancer. 2012;12(3):156–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Nakamura K, Katai H, Mizusawa J, et al. A phase III study of laparoscopy-assisted versus open distal gastrectomy with nodal dissection for clinical stage IA/IB gastric Cancer (JCOG0912). Jpn J Clin Oncol. 2013;43(3):324–327. [DOI] [PubMed] [Google Scholar]
- 23.Strong VE, Song KY, Park CH, et al. Comparison of gastric cancer survival following R0 resection in the United States and Korea using an internationally validated nomogram. Ann Surg. 2010;251(4):640–646. [DOI] [PubMed] [Google Scholar]
- 24.Kelly KJ, Strong VE. Robotic Utilization in Gastric Cancer Surgery. In: Hochwald SN, Kukar M, eds. Minimally Invasive Foregut Surgery for Malignancy: Principles and Practice. New York: Springer; 2015:261–268. [Google Scholar]
- 25.Steyerberg E. Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating. New York: Springer; 2009. [Google Scholar]
- 26.Chun FK, Karakiewicz PI, Briganti A, et al. A critical appraisal of logistic regression-based nomograms, artificial neural networks, classification and regression-tree models, look-up tables and risk-group stratification models for prostate cancer. BJU Int. 2007;99(4):794–800. [DOI] [PubMed] [Google Scholar]
- 27.Adams ST, Leveson SH. Clinical prediction rules. BMJ. 2012;344:d8312. [DOI] [PubMed] [Google Scholar]
- 28.Nakauchi M, Suda K, Susumu S, et al. Comparison of the long-term outcomes of robotic radical gastrectomy for gastric cancer and conventional laparoscopic approach: a single institutional retrospective cohort study. Surg Endosc. 2016;30(12):5444–5452. [DOI] [PubMed] [Google Scholar]
- 29.Kim MS, Kim WJ, Hyung WJ, et al. Comprehensive Learning Curve of Robotic Surgery: Discovery From a Multicenter Prospective Trial of Robotic Gastrectomy. Ann Surg. 2019. [DOI] [PubMed] [Google Scholar]
- 30.Strong VE, Gholami S, Shah MA, et al. Total Gastrectomy for Hereditary Diffuse Gastric Cancer at a Single Center: Postsurgical Outcomes in 41 Patients. Ann Surg. 2017;266(6):1006–1012. [DOI] [PubMed] [Google Scholar]
- 31.Kunisaki C, Makino H, Yamamoto N, et al. Learning curve for laparoscopy-assisted distal gastrectomy with regional lymph node dissection for early gastric cancer. Surg Laparosc Endosc Percutan Tech. 2008;18(3):236–241. [DOI] [PubMed] [Google Scholar]
- 32.Yoo CH, Kim HO, Hwang SI, Son BH, Shin JH, Kim H. Short-term outcomes of laparoscopic-assisted distal gastrectomy for gastric cancer during a surgeon’s learning curve period. Surg Endosc. 2009;23(10):2250–2257. [DOI] [PubMed] [Google Scholar]
- 33.Coratti A, Annecchiarico M, Di Marino M, Gentile E, Coratti F, Giulianotti PC. Robot-assisted gastrectomy for gastric cancer: current status and technical considerations. World J Surg. 2013;37(12):2771–2781. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.