Actual 3-year disease-free survival. AGC indicates advanced gastric cancer; EGC, early gastric cancer.
Overall actual 3-year disease-free survival. Five deaths were excluded from the analysis (from a traffic accident in 1, fulminant hepatitis in 1, multiple organ failure in 2, and suicide in 1).
Hwang S, Park DJ, Jee YS, Kim M, Kim H, Lee H, Yang H, Lee KU. Actual 3-Year Survival After Laparoscopy-Assisted Gastrectomy for Gastric Cancer. Arch Surg. 2009;144(6):559-564. doi:10.1001/archsurg.2009.110
To analyze 3-year actual disease-free survival after laparoscopy-assisted gastrectomy for gastric cancer on the assumption that 3-year disease-free survival may represent 5-year overall survival.
Department of surgery of a university hospital.
A total of 197 patients who underwent laparoscopy-assisted gastrectomy for gastric cancer from May 1998 to September 2007 and who were followed up for more than 3 years.
Main Outcome Measures
Feasibility and long-term survival rate with survival analysis by the Kaplan-Meier method.
Subtotal and total gastrectomies were performed in 178 and 19 patients, respectively. The scope of the lymph node dissections were D1 + β (n = 152) and D2 (n = 45). There were 153, 28, 8, 6, 1, and 1 patients in stages Ia, Ib, II, IIIa, IIIb, and IV, respectively. The median follow-up was 45 months (range, 1-113 months), and there were 7 recurrences. Multivariate analysis of disease-specific survival showed that depth of invasion and lymph node metastasis influenced the prognosis independently. The actual 3-year disease-free survival rate for all patients was 96.9%. The 173 patients with early gastric cancer and 24 with advanced gastric cancer showed 98.8% and 79.1% actual 3-year disease-free survival rates, respectively.
Laparoscopy-assisted gastrectomy is acceptable oncologically in early gastric cancer if 3-year disease-free survival represents 5-year overall survival. Laparoscopy-assisted gastrectomy may also play an important role in the treatment of advanced gastric cancer.
Despite the universal decline in gastric cancer incidence and mortality, particularly in the United States and Western Europe, gastric cancer is still the second most common cancer worldwide.1,2 In Far Eastern countries such as Korea, Japan, and China, and also in many developing countries, gastric cancer is the most prevalent malignant neoplasm and the leading cause of cancer death.3,4 The technical aspects of its treatment are constantly evolving to reduce the particularly high preoperative and postoperative mortality and morbidity,5- 7 and minimally invasive surgery should be considered a tool worth evaluating for optimizing therapy for gastric cancer. Recently, laparoscopic approaches have been introduced to treat patients with early gastric cancer.8- 10
Since the first report in 1992 of laparoscopy-assisted distal gastrectomy (LADG) with Billroth II anastomosis for a case of peptic ulcer disease,11 laparoscopy-assisted gastrectomy (LAG) and lymph node dissection with curative intent have been recommended and practiced for the treatment of early gastric cancers in Korea and Japan.12,13 Those studies demonstrated that LADG is less invasive than open surgery, thus leading to faster recovery.13 However, only a few reports concerning the prognosis of patients who underwent LAG for early gastric cancer or LAG outcome data have been published.12,14,15
Recurrence rates after gastrectomy for gastric cancer remain high, ranging from 40% to 80%, depending on the series.16,17 Most recurrences occur within the first 3 years.18 Moreover, most patients with recurrent gastric cancer die within 1 to 2 years. Therefore, actual 3-year disease-free survival has important implications in terms of curability.
We report the 3-year survival of a retrospective series of 197 patients with gastric adenocarcinoma who underwent LAG and whose treatment included minimally invasive surgery used a surgical tool with “intention-to-treat” laparoscopy.
A review of our prospective gastric cancer database and electronic medical records from May 1998 to September 2007 found 197 patients with gastric cancer who underwent LAG in the Department of Surgery at Dong-A University and Seoul National University Bundang Hospital. All patients were followed up for more than 3 years after their operations. All surgeries were performed by a single surgeon (H.-H.K.). Endoscopic ultrasonography and stomach computed tomography were performed routinely to discriminate early from advanced gastric cancer. We performed LAG in patients diagnosed as having cT1N0M0 to cT2N0M0 gastric cancer, regardless of sex, age, body mass index, previous abdominal operative history, or comorbidities. Of the 197 patients who underwent LAG, 178 had LADG and 19 had laparoscopy-assisted total gastrectomy (LATG).
Depending on the extent of lymph node dissection, modified surgery is classified as D1 + α or D1 + β dissection. The No. 7 lymph nodes are dissected during a D1 + α procedure, regardless of the location of the lesions; in addition, No. 8a lesions are dissected in cases in which lesions are located in the lower third of the stomach. The No. 7, 8a, and 9 lymph nodes are dissected during a D1 + β procedure. In this study, there were 2 types of lymphadenectomy. Patients with T1sm lesions invading to the submucosa underwent D1 + β dissection; those with lesions classified as T2 or greater underwent D2 dissection.
For LAG, after a pneumoperitoneum was established by the open technique, 6 ports were placed. The procedure began by dividing the greater omentum 3 to 4 cm from the gastroepiploic arcade toward the branch of the short gastric vessels, including lymph nodes 4d and 4sb in LADG. In LATG, all short gastric vessels were divided with dissection of their lymph nodes (group 4sa), followed by division of the gastrophrenic ligament. Division of the gastrocolic ligament was continued distally toward the first portion of the duodenum. The right gastroepiploic vein and artery were individually divided after clipping to clear lymph node 6. After the right gastric artery, proper hepatic artery, and anterior surface of the portal vein had been exposed for clearing lymph nodes 5 and 12a, the right gastric artery was divided at its origin by means of vessel ligating clips (Hem-O-Lok; Weck Closure Systems, Teleflex Inc, Research Triangle Park, North Carolina). Caudal retraction of the pancreas with a 5-mm fan retractor facilitated exposure of the common hepatic artery, splenic artery, and celiac trunk. Lymph nodes 8a, 9, and 11p were dissected along each artery with an ultrasonic scalpel (Harmonic; Ethicon Inc, Somerville, New Jersey) or by electrocautery. In addition, the right cardiac node (node 1), left cardiac node (node 2), splenic hilar node (node 10), and the nodes along splenic vessels (No. 11 nodes) were dissected in LATG. The left gastric artery and vein were exposed and divided individually with vessel ligation clips, allowing dissection of the No. 7 lymph nodes. Additional dissection of the No. 9 lymph nodes was added proximally. Lymph nodes 1 and 3 were dissected, and the upper third of the lesser curvature was skeletonized. After full mobilization of the stomach, a 4- to 5-cm transverse incision was made for removal of the specimen and for reconstruction. Billroth I or II or Roux-en-Y reconstruction was performed in LADG, and Roux-en-Y esophagojejunostomy was performed in LATG.
Follow-up investigations were scheduled in our department at 3-month intervals for the first 2 years, then every 6 months for the next 3 years, and annually until the patient's death. The follow-up program consisted of a physical examination, laboratory blood tests, endoscopy, and ultrasonography or computed tomography. Recurrence was diagnosed from clinical, radiologic, or endoscopic signs of disease.
Data were collected by reviewing electronic medical records and our gastric cancer database. Patient characteristics, operative findings, clinicopathological measures, and postoperative progress including complications were recorded and analyzed.
Statistical analysis was performed by SPSS software (basic and advanced program, version 11.0; SPSS Inc, Chicago, Illinois). All values are expressed as the mean and standard error of the mean. Overall and disease-free survival rates were assessed by Kaplan-Meier analysis and compared by the log-rank test. Significance was taken as P < .05.
The 197 patients treated in this series consisted of 112 men and 85 women with a mean age of 58.3 years (SEM, 12.6 years; range, 24-87 years) and a mean body mass index of 22.7 (SEM, 2.8) (calculated as weight in kilograms divided by height in meters squared).
Short-term outcomes, including operative details and postoperative course, are shown in Table 1. All procedures were completed successfully without conversion. The mean estimated blood loss was 58.3 mL. Patients were hospitalized for a median of 8.9 days. The mean (SEM) time until the start of oral intake was 3.9 (1.3) days.
Complications occurred in 18 of the 197 patients (9.1%). Major postoperative complications consisted of 3 cases of postoperative bleeding, 2 of gastrointestinal tract bleeding, 3 wound infections, 3 cases of intra-abdominal fluid collection, and 1 case each of ascites due to underlying liver cirrhosis, esophagojejunostomy anastomotic leakage, duodenal stump leakage, splenic infarction, delayed gastric emptying, leakage of lymphatics, and respiratory complication. Among them, 3 cases of postoperative bleeding, 1 of duodenal stump leakage, 1 of anastomotic leakage, and 1 of splenic infarction were treated by reoperation. The other patients recovered with nonoperative management. One patient (0.5%) died 1 month after the operation. The patient developed bleeding, which required reoperation, and subsequently died of multiple organ failure.
Among the 197 patients, 24 had advanced gastric cancer (12.2%). The stages of the tumor were IA in 153 (77.7%) patients, IB in 28 (14.2%), II in 8 (4.1%), IIIA in 6 (3.0%), IIIB in 1 (0.5%), and IV in 1 (0.5%).
Table 2 shows recurrences after operation. There were 7 cancer recurrences (3.5%): 2 ovarian metastases, 3 bony recurrences, 1 lymph node metastasis, and 1 lung metastasis during the median follow-up period of 45 months (range, 1-112 months). All recurrences occurred within 34 months of operation. The mean time to recurrence was 18.7 (SEM, 10.4) months (range, 8-34 months).
The disease-free actual 3-year survival rates were 98.8% (171 of 173) and 79.2% (19 of 24) for early and advanced gastric cancers, respectively (P < .001) (Figure 1). The overall actual 3-year survival rate was 96.9% (Figure 2). The factors affecting the disease-free survival of the patients were the presence of lymph node metastasis (P < .001), tumor size of 5 cm or larger (P = .02), and depth of invasion (early vs advanced gastric cancer; P < .001; Table 3). On multivariate analysis, the depth of invasion (P = .02; odds ratio, 5.94; 95% confidence interval, 1.26-27.8) and presence of lymph node metastasis (P < .001; odds ratio, 21.3; 95% confidence interval, 2.32-196.5) were independent factors affecting overall survival (Table 4).
Despite a declining incidence, gastric cancer remains an important cause of cancer death.19 This high incidence of gastric cancer, coupled with strong public and physician cooperation in screening, has enabled the Koreans to institute a national effort to diagnose gastric cancers.
Surgical resection for gastric cancer is the only therapeutic modality for curability.20 Recently, several studies have demonstrated immediate postoperative advantages of laparoscopic surgery over conventional open surgery for resection of gastric cancers.13 Furthermore, short-term oncologic outcomes were not compromised by the laparoscopic approach.9,12,21 The Japanese Gastric Cancer Association22 has provided indications for LADG, including (1) mucosal cancer without preoperatively diagnosed lymph node metastases and (2) cancer with submucosal invasion and without preoperatively diagnosed lymph node metastases. However, in our institution, gastric cancers in categories less than cT2aN0M0 are treated by LAG.
The use of LAG has been criticized for its radical nature, and some of these criticisms are apt in this study. However, Kim et al23 recently confirmed that LADG with extraperigastric lymph node dissection (No. 7, 8, and 9) proved to be a feasible and acceptable surgical technique for early gastric cancer. The optimal extent of lymph node dissection for early gastric cancer remains controversial; however, most surgeons consider D1 + β gastrectomy to be the standard and optimal surgical procedure for patients with early gastric cancer24- 26 and D2 gastrectomy for patients with advanced gastric cancer.27 Song et al28 showed that the number of lymph nodes retrieved laparoscopically is adequate for staging, and LAG with D2 gastrectomy is a technically feasible procedure when compared with open surgery.
The patterns of recurrence also did not appear to be different between laparoscopic and open gastrectomy.15 However, LAG for the treatment of malignant neoplasms remains controversial because of the lack of large-scale study data on short-term and long-term outcomes.
The depth of cancer invasion through the gastric wall and the presence or absence of lymph node involvement are 2 important factors influencing survival in resectable gastric cancer.29,30 Lee et al31 reported that early gastric cancer with lymph node metastasis has a higher possibility of recurring, especially when there are more than 6 positive lymph nodes. They also reported that the median interval between gastrectomy and recurrence was 18.5 months, that 76.1% of cases recurred within 36 months, and that 23.9% recurred after 36 months. In our series, 7 among 197 cases recurred within the median follow-up period of 45 months. All recurrences occurred within 34 months.
The rationale for minimally invasive treatment is the low incidence of lymph node metastasis for mucosal cancer (T1m). Recent reports of the histopathological features of more than 13 000 mainly Japanese patients with early gastric cancer established that only 2% (range, 0%-4.8%) of patients with mucosal cancer have positive lymph nodes.32 However, when the tumor invades the submucosal layer (T1sm), this rate increases to about 20% (range, 15%-25%). Interestingly, the metastasis is not confined to the perigastric nodes (N1 level) only, but in about 5% (range, 2.8%-6.4%) of patients with submucosal cancers it will also involve the extraperigastric lymph nodes (N2 level). A risk factor for N2 disease is a submucosal lesion larger than 2 cm.33 This finding is important clinically; there is a clear consensus that R0 resection is the most important independent treatment-related prognostic factor, and complete removal of metastatic N2 nodes would only be achieved by the more radical D2 lymph node dissection.34,35
The prognosis of patients with early gastric cancer is known to be excellent, with 5-year survival rates of 90% or more.36,37 However, lymph node metastasis, large tumors, submucosal invasion, and lymphatic involvement of cancer cells remain risk factors for recurrence of early gastric cancer.38- 40 Kwon et al41 reported that 3-year disease-free survival can be substituted for 5-year survival in curatively resected gastric cancer, and, considering the previously reported results,42 this study showed that patients with early gastric cancer treated by LAG had an excellent 3-year survival rate, despite the short median follow-up period of 45 months. Moreover, mortality and morbidity rates appear to be comparable to those in conventional surgery.
The survival rates after gastrectomy for gastric cancer in published series are summarized in Table 5.5,12,43- 45 Our survival results after LAG were slightly better than those of other open gastrectomy or LAG series, although accurate comparison is not possible because of the relatively shorter follow-up period and smaller patient numbers with stage IIIA, IIIB, and IV tumors compared with other series.
We demonstrated the following clinicopathological factors to be correlated with recurrence: (1) muscularis propria invasion, (2) lymph node metastasis, and (3) large tumor size (≥5 cm). This observation is consistent with previous reports by other investigators.30,39 However, multivariate analysis showed that the first 2 factors, but not the last one, are independent risk factors for recurrence. Therefore, we believe that LAG is feasible in patients with node-negative early gastric cancer. However, it is difficult to discriminate positive nodes from normal nodes and early from advanced gastric cancer preoperatively or intraoperatively. There have been various attempts to improve the accuracy of preoperative and intraoperative diagnosis, with generally unsatisfactory results.46 Therefore, accurate diagnostic modalities for determining preoperative tumor depth and lymph node metastasis should be evaluated so that LAG can be performed more safely.
In conclusion, LAG with regional lymph node dissection for early gastric cancer is considered a safe and curative procedure. A large-scale multicenter randomized trial is needed to document with certainty the benefits and oncologic safety of laparoscopic vs open gastrectomy in both early and advanced gastric cancers.
Correspondence: Hyung-Ho Kim, MD, Department of Surgery, Seoul National University Bundang Hospital, 300 Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi, Korea (firstname.lastname@example.org).
Accepted for Publication: May 22, 2008.
Author Contributions:Study concept and design: Hwang, Park, Yang, and K. U. Lee. Acquisition of data: Jee, M.-C. Kim, H.-H. Kim, and H.-J. Lee. Analysis and interpretation of data: Hwang. Drafting of the manuscript: Hwang and M.-C. Kim. Critical revision of the manuscript for important intellectual content: Park, Jee, H.-H. Kim, H.-J. Lee, Yang, and K. U. Lee. Statistical analysis: Jee. Obtained funding: H.-H. Kim. Administrative, technical, and material support: K. U. Lee. Study supervision: Hwang, Park, M.-C. Kim, H.-H. Kim, H.-J. Lee, and Yang.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grant 0520310-3 from the National Research and Development Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea.