Many studies have investigated the results of laparoscopic gastrectomy [3, 4, 14,15,16]. However, these studies could not provide an answer to whether laparoscopic D2 gastrectomy could be performed with the same quality as in open procedures. The current study reported the outcomes of laparoscopic D2 gastrectomy and confirmed that laparoscopic D2 gastrectomy can be safely performed regarding short-term oncological outcomes.
The operative time is longer in laparoscopic gastrectomy than in conventional open gastrectomy. Changing the instruments, cleaning the cameras, performing a mini-laparotomy and then creating the pneumoperitoneum during the anastomosis are responsible for increasing the operative time . According to the literature, the operative time ranges between 196 min and 370 min for laparoscopic gastrectomy compared with a range from 168 min to 264 min for open gastrectomy [6, 7, 11, 12, 16, 17]. In our study, the median operative time was 240 min [range (200–330 min)], which is longer than that of the study conducted by Lee and colleagues in Korea, which reported a mean operative time of 227 min . Additionally, this time is longer than that of the study conducted in China (CLASS-01 trial) , which demonstrated a mean operative time of 217 min. However, the operative time decreased with subsequent cases after more experience was gained. Moreover, the results of these studies were mainly reported for distal gastrectomies, which had a shorter operative time than total gastrectomies. On the other hand, the operative time in our study favorably compares with the results of studies performed in Japan (JCOG0912 trial) as well as the study conducted in Korea (COACT 1001 trial), which had mean operative times of 278 min and 257.4 min, respectively [3, 8].
Laparoscopic surgery has an advantage over open surgery in minimizing blood loss due to the rapid identification and control of small vessels . In past reports, the estimated blood loss ranged from 82 ml to 333 ml for laparoscopic gastrectomy and 201–440 ml for open procedures [7, 16,17,18,19]. In the present study, the median estimated blood loss was 250 ml, which is considered similar to these studies. However, it is considered significantly higher than the blood loss in other studies conducted in Korea, China, and Japan. The mean blood loss was 153 ml in the KLASS-02 trial and was 105 ml and 115 ml in the CLASS-01 and JCOG0912 trials, respectively [3, 11, 12].
In terms of the adequacy of lymph node yield, in previous studies, the number of harvested lymph nodes ranged from 14–46.5 [3, 5, 8, 11, 12, 20,21,22]. The median number of retrieved lymph nodes in our study was 18, which was the same as that in the study conducted by Brenkman et al.  in the Netherlands. However, it was significantly lower than studies conducted in Asian countries. Lee and colleagues reported a mean number of 46.5 lymph nodes (KLASS-02 trial) . In the CLASS-01 trial conducted in China , the mean number of retrieved lymph nodes was 36.1, while in the JCOG0912 trial performed in Japan, the median number of lymph nodes was 39 . The superiority of these results from Asian studies in contrast to studies in western countries or to our study may be caused by a high qualification system in Asian countries. Additionally, in our study, 8 patients (32%) had received neoadjuvant chemotherapy, which may have influenced the number of harvested lymph nodes. A study in China concluded that neoadjuvant chemotherapy resulted in a reduced lymph node count . This may be a potential reason for the decreased lymph node yield in the current study in comparison to the results from Korea and Japan. Nevertheless, dissection of 18 lymph nodes is considered adequate for tumor clearance and staging. In the current study, the overall rate of adequate lymph node staging (> 15 lymph nodes) was 60% and if we excluded the patients who received neoadjuvant chemotherapy, this rate increased to 70.6%. Furthermore, none of our patients developed nodal recurrence during follow-up. In the current study, the R0 resection in all procedures was 92%, which compared favorably to the study performed by Brenkman et al.  that reported an R0 resection rate of 90%. However, it is unfavorably comparable to the study conducted in Korea, which reported an R0 resection rate of 98.1% .
Laparoscopic gastrectomy is associated with lower morbidity rates than open gastrectomy. In the KLASS-02 trial, the overall complication rates of open and laparoscopic gastrectomy were 24.1% and 16.6%, respectively . The difference was mainly caused by the decrease in local complications, not by systemic complications; in particular, the incidence of fluid collection and intra-abdominal bleeding was significantly lower with laparoscopic gastrectomy than with open surgery. The postoperative morbidity rates of laparoscopic gastrectomy range from 6.4 % to 24.2 % [5, 11, 12, 15, 16, 24]. In the current study, the morbidity rate was 16%, which is within the range reported previously, indicating that the morbidity rate in our study is satisfactory. Anastomotic leakage, which is considered a major complication of gastric surgery, should be evaluated appropriately. Anastomotic leakage rates from previous studies range from 0.2% to 14% [3, 5, 11, 12, 22, 24]. In our study, anastomotic leakage occurred in one case (4%) of total gastrectomy, which is within the range of other studies. Furthermore, D2 lymphadenectomy was performed in all cases in the current study, which is in contrast to most previous studies in which D2 gastrectomy was not routinely performed. Duodenal stump leakage is one of the most severe postgastrectomy complications with rates ranging from 0.4% to 2.4% in previous studies [5, 7, 25, 26]. In the present study, duodenal stump leakage was detected in one patient (4%) with a rate that was 1% higher than previously reported results. In this case, the patient developed sepsis, and reoperation was performed. Drains were inserted, but unfortunately, the patient died from complications. Intra-abdominal collection was found in one patient (4%) in the present study, and it was managed with radiological-guided drainage. This result was unfavorable to the study conducted by Katai et al. in Japan, which reported an intra-abdominal collection in 1.8% of cases . On the other hand, the pancreatic fistula was not reported in the current study, and the incidence of the pancreatic fistula in other studies ranges from 0.4% to 1.9% [3, 11, 12]. Reoperation was indicated in two patients (8%), one of whom underwent distal gastrectomy and then developed duodenal stump leakage; the other patient underwent total gastrectomy and Roux-en-Y esophagojejunostomy and developed leakage from the anastomotic site. Abdominal exploration was performed, and the insertion of drains and a feeding jejunostomy was performed. Two weeks later, the fistula closed with conservative measures. The reoperation rate in our study was higher than that in studies conducted in Japan, Korea, and China, which demonstrated much lower reoperation rates. This is considered to be particularly due to the lack of sufficient experience in comparison to these highly qualified systems in these countries. On the other hand, the reoperation rate in our study was better than that of a study in Italy, which reported a reoperation rate of 9.52% .
According to the literature, mortality rates range from 0% to 6% [5, 11, 22, 24]. In our study, there were two mortalities (8%), which is slightly higher than in previous studies. The reason may be that one of these two mortalities was due to a nonsurgical cause; the patient developed acute myocardial infarction and did not respond to treatment. If we had excluded this patient, the actual surgical mortality would be only one patient (4%) and would be within the range of the previous studies. This patient underwent distal gastrectomy, developed duodenal stump leakage and died from severe sepsis. Although our results in terms of mortality are almost equivalent to other reported results, most of these other studies performed D1 resection rather than D2 resection, which had a higher incidence of morbidities and more technical complexity than D1 resection.
In addition to postoperative complications, conversion to an open procedure is another important index of quality. In the literature, rates of conversion are substantially different between the East and the West. Eastern studies reported a conversion rate ranging from 2.2% to 7% [3, 5, 11, 12, 18, 28]. On the other hand, Ecker et al.  and Brenkman et al.  reported conversion rates of 23.9% and 10%, respectively. In the present study, the conversion rate was 28%, which is significantly higher than that of other studies; however, the reason may be that most of these studies conducted D1 rather than D2 lymphadenectomy. In addition, this may be due to our early experience with this demanding technique; most of the conversions occurred in the first 10 patients; in subsequent patients, the rate of conversion decreased. The main cause for conversion was bleeding, which occurred in two patients (8%); the other causes were adhesions, mass-related causes (the mass was adherent to the pancreas), injury of a visceral organ or technical difficulty (failure to fire the stapler).
In terms of the postoperative period, in the present study, patients started oral fluids after a median period of 4 days, which is considered similar to the study performed in Korea, where patients started oral fluids after 3.7 days . Additionally, this period is considered shorter than the results of two studies done in China that reported periods of 5.5 and 4.7 days prior to starting oral fluids [7, 12]. The early resumption of oral fluids without the need for imaging studies, mainly in cases of distal gastrectomy and a lower incidence of postoperative ileus, may be the reason for our shorter period to resume oral fluids in comparison to these studies.
Patients in our study had a median operative stay of 8 days, which is similar to the studies conducted in Japan, and Korea that reported a mean postoperative stay of 7.5 and 8.1 days, respectively [5, 11]. Patients in the current study had a shorter postoperative stay than those in the CLASS-01 trial, which reported a mean operative stay of 10.8 days .
Concerning the learning curve, proper patient selection resulted in improving the learning curve and decreasing the rates of conversion. Accordingly, the initial stages of laparoscopic gastrectomy should begin with the selection of patients with a body mass index (BMI) < 30 kg/m2 who are suitable for distal gastrectomy and should be conducted under the guidance of expert laparoscopic surgeons . In the present study, conversion rates decreased from 4 conversions in the first 10 patients (40%) to 3 conversions in the last 15 patients (20%), and the median lymph node yield increased from 14 lymph nodes to 20 lymph nodes in the last 15 patients. Moreover, the median operative time decreased by 30 min in the last 15 patients.
There are some limitations in our study. First, we compared our results with the results of open and laparoscopic gastrectomy in the literature. Additionally, the study was limited by the small sample size. Finally, we did not evaluate the long-term outcomes, which might confirm the noninferiority of laparoscopic D2 lymphadenectomy in terms of survival analysis. Despite these limitations, the results are considered satisfactory in comparison to the results of other studies.