Gastric cancer remains a leading cause of cancer death worldwide. It is estimated that 26,370 people will be diagnosed with gastric cancer in the United States in 2013.¹ Unfortunately, only 25% of patients present with localized disease, while 30% present with regional disease and 34% have distant disease at the time of diagnosis.² Therefore, only a small percentage of patients with gastric cancer present with potentially curable disease. For the remainder of patients, the options are noncurative resection or chemotherapy with or without radiotherapy. Given this low rate of curative disease, it is critical for surgeons to be knowledgeable about management options for these patients when cure is not possible.

Prior to considering any intervention when faced with a patient with incurable gastric cancer, it is imperative to distinguish between noncurative and palliative interventions. While oftentimes used interchangeably, they are not synonymous. Evidence of the challenges in making this distinction was provided by McCahill et al³ in a survey of the membership of the Society of Surgical Oncology. In this study, the authors found that while 41% of surgeons defined a procedure as palliative based upon the preoperative intent of the procedure, 27% defined the procedure based upon the postoperative evaluation. Surgeons in this group waited for the results of the operation to determine whether it was palliative or curative. One third of surgeons based their definition of a palliative procedure based upon the patient’s prognosis. According to the World Health Organization, a palliative procedure is an intervention that “provides relief from pain and other distressing symptoms” and “intends neither to hasten nor to postpone death.”⁴ As is highlighted below, the outcomes following noncurative versus palliative interventions for incurable gastric cancer are not the same. Surgeons and proceduralists offering interventions in this setting need to be clear and, most importantly, provide clarity to patients (and their families) about the intent of the intervention, likelihood of achieving the anticipated outcomes, and potential risks of the intervention.

To add to the challenge of caring for patients with advanced gastric cancer, it is sometimes difficult to know when an advanced gastric cancer is curable or not. Given that complete surgical resection (with adjuvant chemoradiation)⁵ has been shown to offer the best chance for prolonged survival, the starting point for a surgeon is to determine whether a gastric cancer is resectable with curative intent. Some characteristics of clearly unresectable and therefore incurable gastric cancer include presence of distant metastases, invasion of a major vascular structure (e.g., aorta, celiac axis), bulky lymphadenopathy at the level of the pancreatic head which would require pancreaticoduodenectomy to completely resect the disease, or lymphadenopathy not typically included in standard lymphadenectomy based upon the location of the primary lesion. While these characteristics are generally able to be determined on current preoperative imaging, consider the following, less clear case: 74-year-old man with biopsy-proven poorly differentiated adenocarcinoma, diffuse type, with signet ring cell features and diffuse mucosal friability and nodularity found in the proximal body and extending to the antrum. Additional staging studies reveal no distant disease. Assuming that this patient did not require gastrectomy for control of bleeding or obstruction, when would a gastrectomy in the setting be considered curative versus palliative? Would the intention of the procedure differ if it was known preoperatively that an R0 resection could not be performed?

This chapter examines the outcomes following noncurative gastrectomy, including its impact on prevention of tumor-related complications and survival as well as the morbidity and mortality associated with these resections. The impact of noncurative gastrectomy on palliative outcomes will then be reviewed, including the effect on quality of life and symptom relief and need for additional palliative interventions. Lastly, nonresectional palliative interventions for intestinal obstruction and bleeding are examined.

Prevention of Tumor-Related Complications

Even when noncurative, many surgeons may consider gastrectomy as the preferred management for selected patients with locally advanced and/or metastatic gastric cancer, since it can (1) prevent potential complications of the tumor and (2) may prolong survival. With regard to the first issue, namely prevention of potential complications from the gastric cancer, a study by a group from Memorial Sloan-Kettering Cancer Center (MSKCC) sought to address this question by examining the outcomes of 165 laparoscopically detected stage M1 gastric cancer patients who did not initially undergo gastrectomy.⁶ Of the 97 patients who continued treatment at MSKCC following diagnosis of M1 disease, 48 (50%) required palliative intervention. The median interval between the staging laparoscopy and the first palliative intervention was 4 months. The most common, primary tumor-related complication requiring intervention was intestinal obstruction, followed by the need for enteral access due to oral diet intolerance and bleeding. Eighty-eight percent of patients did not require laparotomy and 58% did not require a stomach-related procedure. Only one patient developed gastric perforation of their tumor during chemotherapy. The authors concluded that although the need for palliative interventions was common in this selected group of patients with stage IV gastric cancer, a vast majority of palliative interventions were provided without laparotomy. Additionally, they found a very low incidence of either bleeding or perforation which would have been prevented by noncurative gastrectomy. A smaller study by Sarela et al⁷ of 55 patients with M1 disease initially managed nonoperatively found that only 25% of patients (N = 14) required an intervention for management of a tumor-related complication. Similar to the findings from the MSKCC study, they reported that a small minority of patients (9%) required laparotomy for management of their tumor-related complication and no patient required gastrectomy. More recently, Schmidt et al⁸ studied a group of 279 patients with gastric cancer, 169 of whom had clearly incurable disease at presentation and were managed nonoperatively and 110 patients underwent surgical exploration but were determined to have incurable disease and received either noncurative resection (N = 46) or nonresectional surgical intervention (e.g., gastrojejunostomy, diagnostic laparoscopy/laparotomy) (N = 64). Of the 169 patients managed nonoperatively, only 3 patients (1.8%) ultimately required a surgical procedure for management of a tumor-related complication. A similar number of patients who were initially explored but did not receive noncurative resection required operative intervention during the course of follow-up (3/64, 4.7%). One patient in both groups required operation for management of perforation and two patients in both groups required operative intervention for tumor-related obstruction. Based upon these findings, the authors concluded that complications resulting from incurable gastric cancers are not common enough to justify preemptive surgery. Therefore, based upon the current literature, noncurative gastrectomy for prevention of tumor-related complications is not warranted, given that the majority of tumor-related complications (e.g., obstruction) can be managed nonoperatively and there is a very low risk of complications that would have been prevented by noncurative gastrectomy.

Impact on Survival

A second argument in favor of noncurative gastrectomy is the potential for prolonged survival. One of the oldest studies to examine the potential survival benefit of noncurative gastrectomy was published by Lawrence and McNeer in 1958.⁹ The studies reported an average survival of 4.6 months among patients who were explored but did not undergo resection or bypass compared to 8.2 months in patients who underwent noncurative total gastrectomy and 9.5 months in patients who had a noncurative subtotal gastrectomy. A recent multicenter study in France compared survival in 677 patients who underwent noncurative gastrectomy to 532 patients who managed without resection.¹⁰ Median survival was 11.9 months in the resection group compared to 8.5 months in the nonresection group (p = 0.001). This prolonged survival in the noncurative gastrectomy group is notable in light of recent Surveillance, Epidemiology, and End Results data in which the average survival for all patients with stage IV gastric cancer is 4 months, with only 4% surviving 5 years.² However, palliative chemotherapy is associated with a median survival of 10 months.¹¹

Unfortunately, there are no randomized controlled trials available that compare survival following nonoperative treatment of incurable gastric cancer to survival following noncurative resection. The existing literature consists primarily of retrospective analyses from single centers in which highly selected patients underwent noncurative gastrectomy. Table 101-1 contains a summary of survival outcomes from several of the larger available series which specifically examine survival following noncurative resections. As evident in this table, there is wide variability in overall survival following noncurative gastrectomy, from 8.1 to 16.4 months compared to 4.1 to 11.2 months following various nonresectional and/or nonoperative therapies. In all of the studies shown, survival is improved following noncurative gastrectomy. Rather than definitively illustrating a survival advantage of noncurative gastrectomy, these studies reveal the accuracy with which surgeons are able to identify patients most likely to benefit from noncurative gastrectomy. A large number of confounding variables likely account for the survival advantage seen among patients who underwent noncurative gastrectomy compared to those who received either nonresectional surgical interventions or nonoperative interventions. Such variables include patient factors, such as comorbidities and performance status, tumor-related factors such as extent of disease, and treatment factors such as receipt of systemic chemotherapy.

In an attempt to identify patients most likely to experience a survival benefit following noncurative gastrectomy, several authors have sought to examine factors associated with improved survival following noncurative gastrectomy. The largest study to address this question is the multicenter French registry cited above.¹⁰ The authors identified factors associated with poor survival following noncurative gastrectomy, which are summarized in Table 101-2. The two patient-related factors associated with poor survival were higher ASA score and malnutrition. The tumor-related factors related to reduced survival were signet ring cell morphology and more advanced staged tumors (e.g., higher TNM stage, solid organ metastasis). Failure to receive postoperative chemotherapy was the only treatment-related factor associated with poorer survival after noncurative gastrectomy. Of these, the authors identified four independent pre- and perioperative variables associated with poor survival: localized peritoneal carcinomatosis, solid organ metastasis, signet ring cell histology, and diffuse carcinomatosis. Based upon their data, the authors concluded that noncurative gastrectomy does not confer a survival benefit in poor surgical-risk patients (i.e., ASA III–IV), those with signet ring cell histology, or in patients with solid organ metastasis with or without peritoneal carcinomatosis. Similar to the results of Mariette et al,¹⁰ Chang et al¹⁹ and Park et al²¹ also reported the significant impact of receipt of chemotherapy on outcome in stage IV gastric cancer patients. In the study published by Chang et al¹⁹ of 257 patients, a survival advantage of gastrectomy was only found in patients who underwent resection and received chemotherapy compared to those who underwent either nonresectional or no surgical intervention and received chemotherapy. This survival advantage was no longer found when they compared patients who received resection without chemotherapy to those who underwent nonresectional or nonoperative management without chemotherapy (Table 101-1).