Tumors confined to the mucosa and submucosa (T1) are referred to as early gastric cancers (EGC) regardless of lymph node status. Overall, these patients do very well, experiencing cure rates exceeding 80–90 % after surgery alone. The number of lymph node metastasis is an important prognostic factor. In a report from Japan, the 5-year survival rate was intimately linked to nodal disease burden, which decreases from 85 % for N0 disease to 61 % for N1 involvement, 31 % for N2 disease, 10 % for N3 disease, and only 2 % for N4 disease [19].

Despite the changes in incidence, screening, and diagnosis, the paradigm that guides operative management for gastric cancer has not changed over the last century – complete surgical resection is the only potentially curative treatment for localized disease. An important change occurred in the early 1990s when reports from large series from Japan showed the efficacy of using local resection techniques such as an endoscopic mucosal resection (EMR) without a lymphadenectomy or gastrectomy for a select group of patients who were at very low risk of having nodal metastases. The proposed selection criteria include: (1) the tumor is small (≤3 cm in diameter), (2) well differentiated and without lymphovascular invasion, and (3) superficially elevated and/or depressed but without ulceration or definitive submucosal invasion [20, 21]. Given the low likelihood of identifying such lesions in a Western population, its applicability to this cohort appears to be limited. Furthermore, long-term outcome with EMR is lacking. Thus, EMR should only be performed under the auspice of a clinical trial or be limited to medical centers that are highly experienced with such a technique.

If the tumor has spread into the submucosa, the incidence of lymph node metastasis increases, and more definitive surgery (i.e., gastrectomy) must be performed to ensure an R0 resection. A subject of great controversy in the past was whether all gastric cancer patients should have a total gastrectomy or a subtotal gastrectomy. Two prospective randomized trials, one from France and a second from Italy conducted over two decades ago, concluded the following: both procedures have similar 5-year survival rates, but a subtotal gastrectomy is a technically easier operation, is associated with less perioperative morbidity and mortality, allows for better postoperative nutritional status and quality of life, and hence is preferred over total gastrectomy, when feasible [22, 23]. Therefore, our current practice is to perform a subtotal gastrectomy for tumors of the distal stomach and body (Fig. 9.5a–b), depending on the tumor size and margin status and a total gastrectomy for most cancers of the proximal stomach and large tumors in the fundus. A Billroth II gastrojejunostomy (BII) is the preferred method of reconstruction with a caveat that the afferent limb should not be longer than 25 cm in order to avoid the acute afferent loop syndrome (ALS), a potentially serious condition that usually requires emergent surgical intervention (Fig. 9.5a).

As an aside, patients who underwent a subtotal gastrectomy with a BII reconstruction may have a number of gastric motility disorders postoperatively. Two worth mentioning include gastroparesis and afferent loop syndrome. Symptoms for both include upper abdominal pain, nausea and vomiting, early satiety, and anorexia, although ALS tends to present with an acute abdomen. Patients are unable to tolerate per oral intake and require long-term nasogastric tube resulting in a prolonged postoperative recovery time and hospital stays. The causes of postoperative gastroparesis are unclear but may be related to damage of the gastric electrical pacemaker, gastrointestinal dysfunction, and vagus resection following subtotal gastrectomy [24]. As mentioned above, ALS is due to an obstruction of the afferent limb and is mainly due to technical errors (i.e., creation of a long afferent limb), but also due to twisting or kinking of the afferent limb or swelling at the anastomosis of the afferent limb to the stomach. Regardless, the solution to the above two conditions is to create a Braun enteroenterostomy (Fig. 9.5a), either at the time of the initial operation or when these complications occur. The Braun enteroenterostomy helps to divert contents of the afferent limb (bile) to the distal bowel.

For cancers in the proximal stomach, a total gastrectomy with a Roux-en-Y reconstruction is performed (Fig. 9.6), although a subtotal gastrectomy can still be considered if a negative margin can be achieved. The distance between the esophagojejunostomy anastomoses and the jejunojejunostomy anastomoses should be 40–60 cm in order to avoid symptomatic bile reflux.

For the majority of gastric cancers, 5 cm is considered a safe proximal margin. The presence of a microscopic positive margin should be managed in the context of the extent of disease being treated. In patients without significant lymph node metastasis (≤5 positive lymph nodes out of 15 pathologically examined), survival is significantly decreased by leaving a positive margin. However, if patients that have five or more positive lymph nodes, the presence of a positive margin will have little influence on disease-related survival [25]. This brings us to our next consideration in the operative management of gastric tumors – regional control and the extent of lymph node resection.

In order to accurately depict a tumor as N0 within the AJCC/TNM classification, a minimum of 15 lymph nodes is required to be examined by pathology. The Japanese Research Society for the Study of Gastric Cancer has defined a standard for the extent of lymph node dissection and labeled them D1-4 (Fig. 9.7). D-dissection should be distinguished from R-resection; D-dissection describes the extent of lymphadenectomy, while R-resection describes the degree of residual tumor left behind after resection. An R0 resection is a complete resection without evidence of gross or microscopic disease; an R1 is one that has no gross disease, but has evidence of microscopic residual disease, and an R2 is one that has gross residual disease.

A D1 dissection includes a subtotal or total gastrectomy and removal of only the perigastric nodes along the lesser and greater curvature of the stomach (Stations 1–6). A D2 dissection adds the removal of nodes along the left gastric artery, common hepatic artery, celiac trunk, and splenic artery and hilum (Stations 7–12). With some proximal T4 tumors, resection of the spleen and pancreatic tail is necessary to achieve adequate nodal clearance along the splenic artery and hilum, but in most cases the spleen and distal pancreas can and should be preserved (Fig. 9.8). Given the distinction between D-dissection and R-resection, one can see how it is possible to have an R1 resection (microscopic disease) with a D2 dissection.

Prophylactic distal pancreatectomy and splenectomy should not be routinely performed; a splenectomy should only be done when the hilum is involved. D3 and D4 resections involve resecting more distant and periaortic nodes, and recent data has shown that this extent of lymphadenectomy does not improve survival over a D2 dissection [26, 27].

Just as total gastrectomy (in comparison to subtotal) is associated with greater postoperative morbidity and mortality, so can a more extensive lymph node dissection [25]. A randomized Dutch trial of 1,078 patients with gastric cancer (711 patients treated with curative intent), for which the majority had pathologic T1 to T3 tumors, found no long-term overall survival benefit from an extended D2 versus conventional D1 lymph node dissection. It was thought that the higher postoperative mortality rate associated with a D2 dissection likely offsets the long-term survival advantage with a D2 dissection [28]. In the final report from this landmark trial, which was exemplified by the rigorous quality control including training of Dutch surgeons by an expert counterpart from Japan in performing a D2 dissection, the overall survival for patients undergoing a D1 dissection was 30 % at 11 years compared to 35 % for those who underwent a D2 dissection (P = 0.53) [26]. However, when the investigators stratified patients postoperatively by lymph node stage, those with N2 disease who had a D2 resection had a statistically significant improvement in 11-year survival rates of 21 % compared with 0 % in the D1 group. At a median follow-up of 15 years, there was a lower locoregional recurrence rate in the D2 lymphadenectomy group than the D1 group, though this group had a higher rate of postoperative morbidity, mortality, and reoperation [29]. A meta-analysis of 12 randomized trials comparing D1 and D2 demonstrated no significant difference in OS (HR = 0.92, 95 % CI, 0.77–1.10, P = 0.36), but a subgroup analysis of patients who underwent a D2 dissection without a splenectomy and/or pancreatectomy demonstrated a trend towards OS benefit [30]. Therefore, D2 resection is a reasonable approach when the likelihood of the patient having locally advanced nodal disease exists. In Japan, a D2 dissection is considered a routine practice. However, D2 should only be performed by experienced and trained surgeons. To complete the operation, a feeding jejunostomy should be considered for postoperative feeding.

Laparoscopic techniques have become an integral part of surgical practice over the past several decades. For gastric cancer, multiple retrospective studies have reported the advantages of laparoscopic gastrectomy (LG) over open gastrectomy (OG) [31–34]. A recent meta-analysis of 15 non-randomized comparative studies has also shown that although LG had a longer operative time than OG, it was associated with lower intraoperative blood loss, overall complication rate, fewer wound-related complications, quicker recovery of gastrointestinal motility with shorter time to first flatus and oral intake, and shorter hospital stay [34]. A randomized prospective trial comparing laparoscopic assisted with open subtotal gastrectomy reported that LG had a significantly lower blood loss (229 ± 144 ml versus 391 ± 136 ml; P < 0.001), shorter time to resumption of oral intake (5.1 ± 0.5 days versus 7.4 ± 2 days; P < 0.001), and earlier discharge from hospital (10.3 ± 3.6 days versus 14.5 ± 4.6 days; P < 0.001) [35] (Table 9.2).

Laparoscopic subtotal gastrectomy appears to offer similar oncologic outcomes as open subtotal gastrectomy as reported in multiple retrospective studies [34, 41–44] as well as in six randomized controlled trials [35–40] (Table 9.2). As an aside, there are no clinical trials comparing laparoscopic total gastrectomy with open total gastrectomy. Unfortunately, most of these clinical trials had small number of patients; the highly quoted study by Huscher et al. reported only 59 patients [35]. To address the concern of insufficient power, the Korean Laparoscopic Gastrointestinal Surgery Study (KLASS) Group is conducting a large, multicenter phase III trial of 1,415 patients from 12 institutions to compare overall survival, disease survival, morbidity, mortality, quality of life, inflammatory and immune responses, and cost-effectiveness between laparoscopic distal gastrectomy and open distal gastrectomy on early stage gastric cancer (cT1N0M0-cT2aN0M0) (Clinical Trials.gov ID: NCT00452751) [40]. A separate trial comparing laparoscopic surgery with open surgery in advanced gastric cancer is also underway (KLASS 02). Results of these studies will not be available until the distant future.