Minimally invasive versus open gastrectomy for gastric cancer. A pooled analysis of two European randomized controlled trials

Minimally invasive techniques have shown better short term and similar oncological outcomes compared to open techniques in the treatment of gastric cancer in Asian countries. It remains unknown whether these outcomes can be extrapolated to Western countries, where patients often present with advanced gastric cancer.

With the introduction of neoadjuvant chemotherapy, survival rates for advanced gastric cancer have improved.For advanced gastric cancer, surgery with perioperative chemotherapy, is the preferred treatment.Most guidelines recommend a radical gastrectomy with a D2 lymphadenectomy. 1,2However, in these guidelines, no recommendations are given regarding the use of minimally invasive techniques in the treatment of advanced gastric cancer.
[5] However, in Asian countries gastric cancer has a higher incidence and screening programs ensure most patients are treated in early stages of disease, commonly without systemic therapy and more often with partial gastrectomy. 6,7Patients in Western countries often present with gastric cancer in advanced stages and commonly receive multimodality treatment. 2Furthermore, obesity is more prevalent in the Western population compared with the Asian population and it is elusive whether results from Asian studies regarding minimally invasive techniques can be extrapolated to the Western population.
Recently, two randomized controlled trials were completed in Europe, comparing short term outcomes of minimally invasive gastrectomy (MIG) versus open gastrectomy (OG) for gastric cancer; the STOMACH trial (surgical technique: open vs. minimally invasive gastrectomy after chemotherapy) and the LOGICA trial (laparoscopic vs. open gastrectomy for gastric cancer). 8,9Both trials have a relatively small cohort compared to some Asian trials, additionally the STOMACH trial had to readjust the sample size due to slow accrual.Questions arose whether the trial was underpowered with regard to the conclusions that were drawn.Together these trials result in the largest randomized European cohort of patients undergoing open and MIG for gastric cancer.The aim of this study was to assess short-term outcomes and quality of life in this pooled European cohort.

| Study design
A pooled analysis was made of two recently published Western randomized controlled trials; The STOMACH trial and the LOGICA trial.Primary outcome of this pooled analysis was the number of resected lymph nodes in MIG compared to OG in patients with gastric cancer.With advanced gastric cancer being defined as cT2-4a, N0-3, M0, or cT1N+.][10][11] As the introduction and start of these trials were at the same time, both research groups agreed to use the same case report forms to analyze the pooled data of the completed trials.
The STOMACH trial was a non-inferiority, multicenter, international, randomized trial that took place between January 2015 and June 2018 powered for non-inferiority with regard to quality of oncological resection (radical resection and lymphadenectomy).The LOGICA trial was a superiority, multicenter, Dutch national, randomized trial that took place between February 2015 and August 2018 and powered for superiority with regard to duration of hospital admission.

| Patients
Patients were eligible for participation in both trials if they had histologically proven clinically resectable gastric cancer.In the STOMACH trial patients in whom a total gastrectomy was performed and neoadjuvant chemotherapy was given were included.In the LOGICA trial also patients in whom a distal gastrectomy was performed were included and neoadjuvant chemotherapy was not an inclusion criterion.All patients were at least 18 years of age.If previous surgery of the stomach had occurred patients were excluded from the trials.Quality control took place through video approval by the trial principal investigators or site visits.

| Outcomes
Primary outcome of the STOMACH trial was oncological outcomes measured as the number of resected lymph nodes and radicality.
Primary outcome of the LOGICA trial was length of hospital stay.
Both studies assessed intraoperative outcomes such as operation duration, intraoperative complications, and blood loss, as well as postoperative outcomes such as postoperative morbidity and mortality, disease-free survival, overall survival, and quality of life as secondary outcomes.Quality of life was measured by the EORTC-QLQ-C30 and EORTC-QLQ-STO22 questionnaires at baseline, postoperatively, 3 months, 6 months, and 1 year. 12,13A full overview of the study outcomes, outcome definitions and ways of data collection can be found in the previously published study protocols. 10,11

| Statistics
The two cohorts of patients were pooled and grouped into MIG and OG.Analyses were performed based on the intention-to-treat principle.Baseline characteristics of the two arms were compared by means of the student's t-test, the Mann-Whitney U test and the chi-square test depending on the distribution of the data.All study outcomes except survival were evaluated with the same statistical tests as mentioned above.Survival was evaluated by Kaplan-Meier curves with the log rank test.Quality of life scores were calculated and analyzed at specific time-points and by generalized linear models.
Scores were measured in a scale from 0 to 100 with 100 being the best overall health score.Subgroup analyses were performed for total gastrectomy, patients who underwent neoadjuvant chemotherapy, and advanced gastric cancer.An overview of baseline characteristics and pathologic outcome is depicted in Tables 1 and 2. Intraoperative complications occurred in 14 patients (8.0%) after MIG and in 14 patients (8.9%) after OG (p = 0.788).Intraoperative bleeding was reported when the patient needed a blood transfusion.

| Intraoperative outcomes
An overview of all operative outcomes is depicted in Table 3.  4.

| Oncological outcome
No differences were seen in complete oncological resection (Table 2).

| EORTC-QLQ-C30 separate domain scores
In the EORTC-QLQ-C30 questionnaire, no statistically significant differences were observed between the minimally invasive and open group for the different functional scales; physical functioning, role functioning, emotional functioning, cognitive functioning, and social functioning (Figure 2).
EORTC-QLQ-C30 symptom scales showed no differences between both groups (Figure 2), generalized linear models also revealed no differences in trends over time between the groups.

| EORTC-QLQ-STO22 separate domain scores
The EORTC-QLQ-STO22 is a gastric cancer specific module.The measured domains include; dysphagia, pain, reflux, dietary restrictions, anxiety, dry mouth, taste, and body image.No significant differences were seen between both groups (Figure 2).

| Subgroup analyses
Subgroup analyses were performed of patients who underwent total gastrectomy, distal gastrectomy, had advanced gastric cancer, and who underwent neoadjuvant chemotherapy.In all subgroup analyses, no differences were seen between OG and MIG for postoperative complications, admission duration, lymphadenectomy, radicality, and survival (Table 5).

| DISCUSSION
This pooled European cohort of two randomized controlled trials evaluated postoperative outcomes of MIG versus OG.No differences were seen regarding postoperative morbidity, oncological outcomes, nor quality of life.MIG was associated with a longer operating time but with less intraoperative blood loss.[16][17] MIG has increasingly been adopted worldwide.A recent study demonstrated a 4-fold increase in minimally invasive distal gastrectomy (9% vs. 39%) and a 5.5-fold increase in minimally invasive total gastrectomy (6% vs. 33%). 1816][17]19 The STOMACH and LOGICA trials were the first Western randomized controlled trials on this subject.did not show a significant difference between both groups.A recent pooled analysis of 18 studies comparing MIG with OG for advanced gastric cancer showed a significant difference in hospital stay in favor of MIG. 20However, only one of these studies showed an overall shorter hospital stay in comparison to the outcome of this cohort. 21All other studies reported similar or overall longer hospital stay in both groups.An explanation for these outcomes might be the application of enhanced recovery after surgery protocols in our cohort resulting in overall fast postoperative recovery.A higher use of postoperative analgesia after OG may have resulted in faster recovery as well.This is supported by a secondary analysis of the LOGICA cohort, showing that patients in the MIG group did not need epidural analgesia and had a significantly lower consumption of opioids. 22Long-term complications associated with open abdominal surgery are incisional hernias. 23It is expected that there will be less incisional hernias after MIG in comparison to OG.Long-term complications from the STOMACH and LOGICA cohort has to be awaited.
Gastric cancer is one of the most prevalent malignancies in Asia, but its incidence is low in Western countries. 24In Asia, screening programs for the detection of gastric cancer exist and most diagnoses are made at an early stage of disease resulting in more patients who undergo primary surgery. 25Patients who received neoadjuvant therapy are relatively underrepresented in previous Asian trials, and therefore there was no solid evidence to support MIG for these patients' subgroups. 26In the current study, subgroup analyses for patients who received neoadjuvant therapy and those patients who underwent primary resection showed no difference between MIG and OG regarding oncological outcomes and postoperative outcomes.These results implicate that MIG can be performed safely for patients with advanced gastric cancer, patients who undergo total gastrectomy and patients who are submitted to neoadjuvant therapy.
Several studies have showed similar survival outcomes in MIG and OG, which is in line with the recent analysis of 3-year survival in the STOMACH trial. 20,27,28A recent study from Tsekrekos et al.
showed a better survival after laparoscopic surgery.However, this survival benefit was exclusively seen in patients who underwent distal gastrectomy and no difference was demonstrated in the total gastrectomy group. 29ng-term survival and long-term quality of life of the STOMACH and LOGICA trial has to be awaited.
Before drawing a final conclusion on MIG, data on quality of life, cost-effectiveness and the learning curve of MIG should be taken into account.There was one Asian trial who took quality of life into account; patients with early gastric cancer underwent distal gastrectomy without perioperative chemotherapy who were randomized between MIG and OG. 30 The results of this study demonstrate that at 1-year follow-up, quality of life is comparable between the groups.Based on previous prospective studies, it is not expected that quality of life will deteriorate after 1-year follow-up unless recurrent disease will develop. 31Recent additional analysis of the LOGICA trial regarding cost-effectiveness demonstrates largely comparable cost-effectiveness between MIG and OG at 1-year follow-up.Finally, centralization of gastric cancer treatment and improvement in the learning curve of MIG could improve overall outcomes for minimally invasive techniques in the Western population.
Besides the long-term outcomes of MIG, other developments in gastric cancer surgery should also be carefully monitored.MIG allows for storage of surgical footage, allowing for quality control, and easy implementation of other visualization techniques such as the application of indocyanine green, and these benefits should be taken into account when assessing the optimal technique.[34] For gastric cancer, robot-assisted gastrectomy is performed in less than 5% of procedures. 18Until now, studies from the centers with the largest volumes of robot-assisted gastrectomy demonstrated longer operating time, less blood loss, comparable oncological outcomes, and higher costs as compared to laparoscopic gastrectomy. 35However, a recent randomized trial by Ojima et al. suggested less postoperative complications following robot-assisted gastrectomy compared to laparoscopic gastrectomy. 36Along with the technical advances, the implementation of robotic gastrectomy for gastric cancer will most likely increase in the near future.
Strengths of this study include that it is the largest European randomized cohort of patients on OG versus MIG, including two trials with largely comparable perioperative protocols.Secondly, probably due to surgical quality control, surgical quality was high, with a radical resection rate higher than 92% and a median of 32 resected lymph nodes in both treatment arms.Limitations of this study include patients and investigators were not being blinded for the study arm, it was not powered on important outcomes such as survival and quality of life.Finally, despite randomization more patient in the open group received neoadjuvant chemotherapy, which may have influenced the results.

| CONCLUSION
The pooled data from two Western randomized controlled trials on MIG versus OG did not show any differences in postoperative recovery and complete oncological resection.In addition, quality of life was not different at 1-year follow-up between both groups.
These results support that MIG is a safe alternative to OG for the surgical treatment of patients with advanced gastric cancer.
Six hundred and eighty-six patients were eligible for inclusion, 347 patients were not included in the trial due to a decline of participation, screening failure, language barrier, disseminated disease, or other reasons.A total of 321 were included from both trials; 96 from the STOMACH trial and 225 from the LOGICA trial.Of these, 162 patients (50.5%) were allocated to MIG and 159 patients (49.5%) to OG.A flow chart of the included patients is depicted in Figure1.Baseline characteristics such as age, gender, American Society of Anesthesiologists (ASA) classification were similar between both groups except for body mass index (BMI).Mean BMI was 26.4 ± 4.7 kg/m 2 in the minimally invasive group versus 25.1 ± 3.9 kg/m 2 in the open group (p = 0.005).F I G U R E 1 Flow chart of included patients.van der WIELEN ET AL. | 913 A total of 307 out if 321 patients had clinical locally advanced gastric cancer.There were no significant differences in clinical TNM stage between both groups.In the open group, more patients received neoadjuvant chemotherapy compared to the patients in the minimally invasive group, 136 patients (85.5%) versus 124 patients (76.5%) (p = 0.047).As 10 patients did not proceed to surgery due to metastases or progression of disease at restaging following neoadjuvant treatment, 155 patients underwent MIG and 156 patients OG.In the minimally invasive group 95 patients (58.6%) underwent total gastrectomy, and 92 patients (57.9%) in the open group (p = 0.420).D2 lymphadenectomy was performed in 139 patients (89.7%) in the minimally invasive group and 142 patients (91%) in the open group (p = 0.451).

A
significant difference was seen in blood loss and for the duration of surgery.Median duration of surgery was 228.5 min in the minimally invasive group versus 180 min in the open group (p = 0.005).Median blood loss was 150 mL during MIG versus 260 mL during OG (p < 0.001).
Postoperative recovery was similar between both groups.Median hospital stay was 7.0 in the minimally invasive group and 8.0 in the open group (p = 0.117).In 67 patients (41.4%) in the minimally invasive group a postoperative complication occurred and in 66 patients (41.5%) in the open group (p = 0.978).In patients who experienced postoperative complications, median hospital stay was 10.0 after MIG and 11.0 after OG (p = 0.248).Anastomotic leakage occurred in 14 patients (8.6%) in the minimally invasive group and in 16 patients (10.1%) in the open group (p = 0.705).Postoperative death occurred in 5 patients (3.1%) in the minimally invasive group versus 10 patients (6.3%) in the open group (p = 0.978).An overview of morbidity and mortality is depicted in Table

3. 4 |
Quality of life3.4.1 | Overall health scoresA response compliance was achieved of 80% in the STOMACH trial and 64% in the LOGICA trial for all PROMs.No differences were observed in health-related quality of life (HRQoL) scores between MGI and OG (Figure2).The median EORTC-QLQ-C30 overall health score 1 year postoperatively was 66.7 (50-83.3) in the minimally invasive group and 66.7 (58.3-83.3) in the open group (p = 0.196).An overview of quality-of-life scores is depicted in Figure2.
in both groups (p = 0.661).In the minimally invasive group 146 patients (94.2%) had a radical R0 resection in comparison to 144 patients (92.3%) in the open group (p = 0.786).There was no difference in 1-year survival between both groups.After MIG, 129 patients (83.8%) were still alive 1-year postoperatively versus 129 patients (82.7%) after OG (p = 0.879).T A B L E 1 b Two patients missing in the MIG group.
Health-related quality of life.van der WIELEN ET AL. | 917 T A B L E 5