Treatment of gastric cancer involves a multidisciplinary approach to achieve long-term outcome, including surgery, chemotherapy, and radiation therapy. Most patients present with advanced disease and are not candidates for a curative approach. Palliative chemotherapy is recommended for symptom control and for short-term advances in survival. Surgery combined with different chemotherapy and chemoradiation options improves survival. Initial studies focused on adjuvant chemoradiation and showed improved survival. More recent trials have demonstrated that perioperative chemotherapy before and after surgery provides a survival advantage. Such an approach may also downstage marginal patients who can then be selected to undergo curative resection and complete adjuvant chemotherapy.
Key points
- •
Curative treatment for gastric cancer involves either total gastrectomy or subtotal gastrectomy with extended D2 lymph node dissection for patients who present with more localized disease.
- •
After resection, adjuvant chemoradiation has improved long-term survival in gastric cancer, especially for patients who did not have a complete lymph node dissection.
- •
Perioperative chemotherapy and radical gastrectomy improves survival and may downstage select patients from unresectable to resectable.
- •
Chemotherapy for advanced gastric cancer provides palliative benefits, and may provide some short-term survival advantages.
Introduction
Despite modern advances in diagnostics, surgical techniques, radiation therapy, and chemobiologic therapy, gastric cancer remains a highly fatal disease. Based on SEER data, the 5-year relative survival for gastric cancer (all stages) from 2002–2008 is 27%. The presence of micrometastatic disease, even in apparent early stage disease, contributes to the poor overall survival and the need for improved chemotherapeutics. Most gastric cancers are adenocarcinoma histology (>90%). There has been a migration of gastric cancers over the past century. In the early 1900s most gastric cancers were distal and involved the gastric antrum and body. The incidence of distal gastric cancers has been declining steadily, whereas the incidence of proximal gastric cancers and esophagogastric junction (EGJ) cancers has risen. There has also been a corresponding increase in the number of adenocarcinomas of the distal esophagus. The rate of increase is alarmingly high and should be considered an epidemic. In regards to gastric, EGJ, and distal adenocarcinomas of the esophagus, their natural history, response to treatment, and prognosis are very similar suggesting a shared pathogenesis. Some of the possible causes include infectious pathogens (ie, Helicobacter pylori ); Barrett esophagus; and obesity. As the world continues to struggle with the obesity epidemic, there will continue to be an increase in these malignancies.
Surgical resection remains the primary therapy for gastric cancers when the goal of care is curative. More than 50% of gastric cancers have metastasized to regional lymph nodes at the time of surgery. Survival rates are correlated with the degree of lymph node involvement. Five-year survival for N0 disease is around 50%, whereas survival for N3 disease is only 10%. The poor survival with advanced disease confirms the importance of developing more active systemic therapy, which can be incorporated into a multidisciplinary approach to improve outcomes. There is currently no agreed on approach (neoadjuvant vs adjuvant) for locally advanced gastric cancer. This is in part related to the numerous areas of controversy, which have not been addressed in large multicenter randomized controlled phase III trials. Some of the issues have arisen because classifications of gastric cancers and EGJ cancers have changed with time. In addition, this overlap has diluted the data pool because most of the major clinical trials have included a sizable population of distal esophageal, EGJ, and gastric cancers. In the most recent version of TNM staging system (version 7), EGJ and gastric cardia tumors (Siewert III, arise in cardia within 5 cm of EGJ and extend into esophagus or EGJ junction ) are classified as esophageal cancer rather than gastric cancer.
Introduction
Despite modern advances in diagnostics, surgical techniques, radiation therapy, and chemobiologic therapy, gastric cancer remains a highly fatal disease. Based on SEER data, the 5-year relative survival for gastric cancer (all stages) from 2002–2008 is 27%. The presence of micrometastatic disease, even in apparent early stage disease, contributes to the poor overall survival and the need for improved chemotherapeutics. Most gastric cancers are adenocarcinoma histology (>90%). There has been a migration of gastric cancers over the past century. In the early 1900s most gastric cancers were distal and involved the gastric antrum and body. The incidence of distal gastric cancers has been declining steadily, whereas the incidence of proximal gastric cancers and esophagogastric junction (EGJ) cancers has risen. There has also been a corresponding increase in the number of adenocarcinomas of the distal esophagus. The rate of increase is alarmingly high and should be considered an epidemic. In regards to gastric, EGJ, and distal adenocarcinomas of the esophagus, their natural history, response to treatment, and prognosis are very similar suggesting a shared pathogenesis. Some of the possible causes include infectious pathogens (ie, Helicobacter pylori ); Barrett esophagus; and obesity. As the world continues to struggle with the obesity epidemic, there will continue to be an increase in these malignancies.
Surgical resection remains the primary therapy for gastric cancers when the goal of care is curative. More than 50% of gastric cancers have metastasized to regional lymph nodes at the time of surgery. Survival rates are correlated with the degree of lymph node involvement. Five-year survival for N0 disease is around 50%, whereas survival for N3 disease is only 10%. The poor survival with advanced disease confirms the importance of developing more active systemic therapy, which can be incorporated into a multidisciplinary approach to improve outcomes. There is currently no agreed on approach (neoadjuvant vs adjuvant) for locally advanced gastric cancer. This is in part related to the numerous areas of controversy, which have not been addressed in large multicenter randomized controlled phase III trials. Some of the issues have arisen because classifications of gastric cancers and EGJ cancers have changed with time. In addition, this overlap has diluted the data pool because most of the major clinical trials have included a sizable population of distal esophageal, EGJ, and gastric cancers. In the most recent version of TNM staging system (version 7), EGJ and gastric cardia tumors (Siewert III, arise in cardia within 5 cm of EGJ and extend into esophagus or EGJ junction ) are classified as esophageal cancer rather than gastric cancer.
The role of surgery in the multidiciplinary treatment of gastric cancer
Extent of Gastric Resection
Resection of the stomach has been the primary weapon in the attempt to cure gastric cancer for over a century. The advantages and potential benefits of gastrectomy allow surgery to continue to be the mainstay in the multimodality treatment of gastric adenocarcinoma. A multitude of studies have attempted to delineate the optimal extent of organ resection and lymphadenectomy in terms of potential value versus perioperative morbidity and mortality. Much debate has occurred in regards to the extent of gastric resection to reduce perioperative complications and death rates, while at the same time optimizing the patient’s chance for long-term survival. Gastric body and antrum tumors are best managed by distal subtotal gastrectomy with reconstruction, and cancers arising in the fundus are routinely managed with total gastrectomy. There is some variation with regard to proximal tumors, those arising near the gastric cardia or gastroesophageal junction. In most cases, operative approach entails either transhiatal or thoracoabdominal esophagogastrectomy. Most gastroesophageal junction cancers are treated like esophageal cancers and are not the focus of this article. The debate centers around the management of tumors arising in the body and distal half of the stomach.
McNeer and colleagues published their results in 1974 after performing routine total gastrectomy for all patients with cancer of the stomach, demonstrating a high morbidity and mortality associated with the procedure. This raised the question as to whether such a radical operation was indeed necessary for all patients with gastric adenocarcinoma. Subsequently, there have been several large studies comparing outcomes data between total and subtotal gastrectomy. A multicenter randomized trial was conducted by Gouzi and colleagues and published in 1989. This French subtotal versus total gastrectomy trial was conducted between 1980 and 1985 and included 169 patients with distal gastric adenocarcinoma who were randomized to undergo either a distal subtotal gastrectomy or total gastrectomy. The trial showed a slightly lower perioperative mortality rate in the total gastrectomy group; however, 5-year survival outcomes data were virtually identical at 48% for both groups. The Italian Gastrointestinal Tumor Study Group conducted a similar study between 1982 and 1993, accruing patients with carcinoma confined to the gastric antrum. A total of 618 patients from 31 Italian institutions were included in the study, and were randomized to subtotal gastrectomy or total gastrectomy. D2 lymphadenectomy and omental bursectomy were recommended for both groups but not mandated by the trial design. There was median follow-up of 72 months, and the 5-year survival was found to be 65.3% for the subtotal gastrectomy group and 62.4% for the total gastrectomy group. These two landmark studies solidified the theory that subtotal gastrectomy could be performed safely and effectively, with similar long-term survival to total gastrectomy. In addition, multiple publications have documented significantly better outcomes with regard to patient satisfaction and quality of life after subtotal gastrectomy compared with total gastric resection.
Extent of Lymph Node Dissection and Splenectomy
Gastric cancer has a strong predilection to spread to local and regional lymph nodes as the first site of metastatic disease. For this reason, as with many gastrointestinal cancers, lymph node harvesting is an imperative part of a proper cancer operation. Lymph node staging provides prognostic information to guide further therapy and to guide patient expectations. Lymph node dissection also has therapeutic benefits that improve locoregional control, and which may potentially improve long-term survival. The debate over the past few decades has revolved around the extent of lymph node dissection necessary for adequate staging and treatment of patients with gastric cancer, ranging from removal of only perigastric nodes (D1) to removal of the distant para-aortic nodes (D3).
The lymph node stations of the foregut in the region of the stomach are identified as D1 to D3 by the Japanese Gastic Cancer Association. The first level nodes (D1) are stations 1 to 6 and are referred to as the perigastric lymph nodes. Second level lymph nodes (D2) include stations 7 to 12 and are the regional nodes. The third level (D3) lymph nodes include stations 13 to 16, which are the para-aortic lymph nodes. There is a vast body of literature comparing a more aggressive D2 lymphadenectomy with the more conservative D1 resection. A large proportion of this data comes from the Japanese, where at this time the completion of D2 lymphadenectomy is considered standard of care. Such a standard has been met with some reluctance by surgeons in the United States and Europe, and this is a topic of debate among surgical oncologists. In part, many Western surgeons believe that gastric cancer in Western patients is a different disease than what is seen in the East, making comparisons in outcome based on technique alone very difficult and potentially misleading. Tumor location, tumor histology, patient demographics, and comorbidities are all variables that potentially make gastric cancer seen in the West a biologically more aggressive disease. Nonetheless, many surgical oncologists in the West are now trained to perform an extended lymphadenectomy, and based on current American Joint Committee on Cancer staging (seventh edition), a proper gastric cancer operation needs to harvest at least 15 lymph nodes, which usually requires a D2 lymphadenectomy.
The first large randomized clinical trial comparing D1 with D2 lymphadenectomy was conducted from 1986 to 1995 by Cuschieri and colleagues (British Surgical Cooperative Group) and published in 1996 with long-term data published in 1999. At the time of laparotomy, patients were assessed for resectability, and if resectable were then randomized to either D1 or D2 lymphadenectomy. A total of 737 cases were registered during the study period. Four hundred subjects were included for randomization after intraoperative assessment of eligibility. Median follow-up was 6.5 years. The 5-year overall survival for the D1 group was 35% compared with 33% for the D2 group. There was no difference between disease-free or recurrence-free survival. Morbidity and mortality were higher in the D2 group; however, this seemed to correlate with the much higher likelihood of concomitant pancreatectomy or splenectomy at the time of resection. Splenectomy and pancreatectomy performed during lymphadenectomy proved to be independent predictors of poor survival, perhaps negating any potential benefit that would have been seen with a D2 lymphadenectomy.
The Dutch D1 versus D2 lymphadenectomy trial took place between 1989 and 1993, with results published by Bonenkamp and colleagues in 1995. A total of 996 patients, after exclusions, were randomized to either D1 or D2 lymphadenectomy. Two hundred eighty-five of these patients were found to have distant metastatic disease and were excluded. The remaining 711 subjects underwent the randomly assigned operation and lymphadenectomy. Morbidity (25% D1 vs 43% D2) and in-hospital mortality (4% D1 vs 10% D2) were statistically significantly higher in the D2 group. These findings were comparable with those reported by the British trial. Rates of pancreatic and splenic resections were again much higher in the D2 group. Similar to the British trial, no statistically significant difference was found between the D1 and D2 groups in terms of survival. Five-year survival in the D1 group was 45% and 47% for the D2 group. Fifteen-year follow-up data published by Songun and colleagues in 2010 has reported a benefit to completion of D2 gastrectomy in patients with more advanced stage cancer with more extensive nodal involvement.
The German Gastric Cancer Study was published in 1998 by Siewert and colleagues and included 1654 patients conducted in a prospective fashion. Long-term results of D1 versus D2 lymphadenectomy were compared. D1 lymphadenectomy was defined as resection of less than or equal to 25 lymph nodes, and D2 defined as removal of 25 or more nodes. The results of this study demonstrate a statistically significant difference in long-term survival in patients with stage II (T2N1) disease for whom D2 lymphadenectomy was performed. These results were accomplished without significant increase in perioperative morbidity or mortality in the D2 group. When comparing these data with the long-term outcomes published by Songun from the Dutch trial, it seems that patients with positive nodes are the ones who benefit the most from the lymph node dissection. Because the surgeon does not know this prospectively, and considering the prevalence of lymph node metastases is high in patients with gastric cancer, in general, all such patients should undergo an extended lymphadenectomy. To address this challenge, Japanese surgeons have developed a predictive tool that gives a risk stratification to determine which patients and which nodal stations are most likely to harbor metastases based on certain tumor characteristics.
The Importance of the Maruyama Index
The Maruyama program was produced by Kampschoer and colleagues and published in 1989. It was initially comprised of a database of 3843 cases of gastric cancer collected prospectively from 1969 to 1984 at the National Cancer Center, Tokyo, Japan, and has since been expanded. The program allows new cases to be matched based on seven preoperatively identifiable characteristics, including histologic type, tumor location, tumor diameter, depth of tumor invasion, age and gender of patient, and Borrmann morphologic type. It matches new cases of gastric cancer to cases within the database with similar characteristics, and based on known results gives a nodal metastasis risk for each individual lymph node station and overall survival in the form of a numerical score. The validity of the predictions made by this program has been tested using Dutch, Italian, and German populations, with published results proving it to be accurate.
Blinded reanalysis and autopsy data from the Dutch D1-D2 trial has been reported evaluating the effect of Maruyama index in this population. The authors demonstrated that a Maruyama index of less than five results in significantly superior long-term survival, and a low index score is an independent predictor of disease-free and overall survival. Long-term (>11 years) follow-up data from autopy-based analysis of these subjects revealed that distant-only failure was no different between the Maruyama index categories (15% for MI <5 vs 13% for MI >5), but regional failure was significantly different between the groups (8% for MI <5 vs 21% for MI >5). The authors concluded that performing “low Maruyama index” surgery provides significant benefit in terms of local-regional control and survival, but does not provide long-term prevention of distant metastatic disease. This literature suggests that planning a “low Maruyama index” surgery can be considered more crucial to overall and disease-free survival than simply the D level of lymphadenectomy completed alone. Many suggest that this piece of data should be calculated for all patients undergoing gastric resection and lymphadenectomy for adenocarcinoma of the stomach.
The role of minimally invasive techniques in the treatment of gastric cancer
The use of diagnostic laparoscopy and laparoscopic ultrasound as adjuncts in the metastatic work-up of foregut cancers including gastric adenocarcinoma is now widely accepted. In patients with advanced disease, diagnostic laparoscopy is highly accurate for identification of occult metastases not visualized on preoperative computed tomography scanning. A study by Burke and colleagues at Memorial Sloan Kettering Cancer Center in 1997 reported staging laparoscopy to be 94% accurate with sensitivity of 84% and specificity of 100%. A more recent study published in 2009 from Romania found strikingly similar statistics, including accuracy of 95.5%, and sensitivity and specificity of 89% and 100%, respectively.
Similarly, laparoscopic gastric resection and lymphadenectomy has become a viable option in the surgical treatment of early gastric cancer. However, its use is still controversial in terms of overall oncologic quality, safety, and cost effectiveness. A large meta-analysis published by Zeng and colleagues in 2012 examined results of studies comparing laparoscopic-assisted versus open distal gastrectomy for early gastric cancer. Twenty-two studies with a total of 3411 participants were included in the meta-analysis. They reported similar numbers of lymph nodes retrieved between the groups. Postoperative morbidity was significantly less after laparoscopic resection, including intraoperative blood loss, postoperative analgesic consumption, and hospital duration. Long-term survival was not statistically significant between the groups, with reported 5-year survival rates ranging between 92% and 98% for each group. The surgical techniques required for adequate laparoscopic gastric resection and lymphadenectomy, however, are operator dependant, time consuming, and have a steep learning curve as with any technically complex laparoscopic operation. This poses a challenge to cancer surgeons, because gastric cancer is not that common, limiting the number of patients available to gain experience with laparoscopic oncologic gastrectomy and lymphadenectomy.
Endoscopic mucosal resection has also emerged as an acceptable option for selected Tis and T1 lesions. Unfortunately, such early disease is rarely seen in Western patients. In Eastern countries, gastric cancer is much more prevalent, screening programs are routine, and early cancers are indeed detected frequently. Mounting evidence has shown that endoscopic mucosal resection is a safe and effective option for very early gastric cancers with favorable characteristics, because the adequacy of this modality depends on the absence of lymph node metastasis. This has proved to be a safe procedure, with evidence published by Ono and colleagues in 2001. This was an 11-year case series including 445 patients. They reported no gastric cancer- or treatment-related deaths during a median follow-up period of 38 months. Endoscopic mucosal resection is, similar to laparoscopic gastrectomy, a technically challenging procedure and should be performed at only specialized centers with high volume and adequately trained operators.
The role of chemotherapy in the multidisciplinary treatment of gastric cancer
The role of chemotherapeutics in gastric cancer is constantly evolving. As understanding of the molecular nature of the various subtypes of gastric cancer improves, therapies will also continue to evolve and it is hoped improve outcomes and lessen the toxicity associated with therapy. There is current evidence to support the use of systemic chemotherapy in the perioperative setting and in the metastatic setting. Some of the pivotal trials, which have depicted the benefit of systemic therapies in gastric adenocarcinoma, are highlighted next. Initially discussed are perioperative treatment trials followed by adjuvant trials and finally the treatment of metastatic disease.
Treatment biases remain and treatment does vary geographically. One can refer to the surgery section for more information on the different surgical techniques and degree of lymph node resection to see how this likely affects outcome and response to chemoradiotherapy. In the United States, most patients are treated initially with surgery followed by adjuvant chemoradiotherapy based on the results of the Intergroup 0116 trial. In Northern Europe they favor a perioperative or neoadjuvant approach based on the result of the MAGIC trial and French FNLCC trial. In Japan and southern Europe they tend to favor adjuvant chemotherapy alone without radiation therapy. One can argue that the patient population and the molecular characteristics of the tumor and patients also may very geographically. Clinicians need to improve the understanding of the molecular characteristics of the tumors too not only create novel targeted agents, but also to be able to appropriately test these agents and treatment strategies appropriately. Highlighted next are the major studies that have evaluated the role of the various perioperative treatment strategies.
Preoperative and neoadjuvant chemotherapy
There are several potential advantages to the neoadjuvant approach. One could downstage the patient or convert borderline resectable cases. There is perceived improved tolerance and ability to deliver the entire treatment program. For patients who are destined to develop metastatic disease one could potentially spare surgery if the patient develops metastatic disease during the chemotherapy portion of treatment. For chemotherapy-sensitive disease it allows early exposure to systemic chemotherapy and potentially improved outcomes. The major downside is delaying surgery and allowing a potentially resectable case to become unresectable.
There have been three large randomized clinical trials that have evaluated the perioperative and neoadjuvant approach to multimodality therapy in gastric adenocarcinomas. Two of the trials have shown a clear survival advantage to this approach (MAGIC and FRENCH FNLCC/FFCD) and have shaped the way gastric cancer is managed in most of Western Europe. The EORTC trial failed to show a survival advantage, although it was not powered to show a survival advantage, given poor accrual. Two additional trials have incorporated an oral fluoropyrimidine, which also failed to show a survival advantage.
The MAGIC trial was conducted in the United Kingdom and included 503 patients, 74% with gastric, 11% with distal esophageal, and 155 with EGJ adenocarcinomas. Patients had to have at least T2 or higher disease. The trial randomized patients to surgery alone versus surgery and perioperative chemotherapy. The chemotherapy treatment program included three preoperative and three postoperative cycles of ECF (epirubicin, 50 mg/m 2 Day 1; cisplatin, 60 mg/m 2 Day 1; 5-fluorouracil [5-FU], 200 mg/m 2 daily), each cycle lasting 21 days.
The results showed a significant improvement in 5-year survival with perioperative chemotherapy 36% versus surgery alone 23% (hazard ratio [HR], 0.75; 95% confidence interval [CI], 0.6–0.93). Progression-free survival (PFS) was also improved. The patients receiving chemotherapy also had a higher rate of potentially curative surgery (79% vs 70%). The final pathologic stage also suggested that patients were able to be downstaged with preoperative chemotherapy (T1/T2 52% vs 37%; N0/N1 84% vs 71%). Local (14% vs 21%) and systemic (24% vs 37%) failure rates were reduced with perioperative chemotherapy.
The ECF chemotherapy regimen seemed to be relatively well tolerated. The most common significant toxicity was neutropenia at a rate of 23%. Other grade 3/4 toxicities were noted in less than 12% of the patients. However, only 42% of patients were able to complete the entire treatment program including all recommended chemotherapy, surgery, then more chemotherapy.
The other large randomized trial that showed a survival advantage to perioperative chemotherapy, is the French FNLCC/FFCD trial. It randomized 224 patients with stage II or greater gastric, EGJ, and distal esophageal cancers to either surgery alone or surgery and perioperative chemotherapy. The chemotherapy program included infusional 5-FU, 800 mg/m 2 daily for 5 days, plus cisplatin, 100 mg/m 2 on Day 1 or 2 every 28 days. Two to three cycles were administered preoperatively and three to four cycles postoperatively for a total of six cycles. The results showed an increase in the number of patients achieving an R0 resection when they received preoperative chemotherapy (84% vs 73%). Five-year survival and disease-free survival were both significantly improved in the chemotherapy group (38% vs 24% and 34% vs 19%, respectively).
Some of the major criticisms of the trial mirrored those of the other perioperative treatment trials and included a large proportion of EGJ and distal esophageal cancers. In addition, only around 50% of the patients who were randomized to chemotherapy received postoperative chemotherapy.
The EORTC 40954 is the third large randomized trial that evaluated the role of perioperative chemotherapy, although this trial failed to show a survival advantage to chemotherapy in addition to surgical resection. It randomized 144 patients with stage III or IV (M0) disease and included both gastric and EGJ cancers to surgery alone versus neoadjuvant chemotherapy followed by surgery. There was no postoperative chemotherapy in this trial. Of note, a large number of patients in these trials were unable to receive postoperative chemotherapy. The chemotherapy program in this trial was two 48-day cycles of cisplatin administered on Days 1, 15, and 29, and leucovorin and 24-hour infusional FU on Days 1, 8, 15, 22, 29, and 36.
With median follow-up of 4.4 years the HR is 0.84 (CI, 0.52–1.35). There was a significant difference in the percent of patients able to achieve a complete resection favoring chemotherapy (82% vs 67%). The incidence of postoperative complications was also higher in the chemotherapy group (27% vs 16%).
The criticism of this trial is that it again included patients with EGJ cancers and the study was underpowered to show a survival advantage secondary to poor accrual (planned 360 patients). One of the problems is that in the United States most patients are treated based on the results of the Intergroup trial and are not offered perioperative chemotherapy.
Related posts:
Multidisciplinary Care of the Patient with Head and Neck Cancer
Multidisciplinary Management of Esophageal Cancer
Colorectal Cancer Metastases
Multimodality Treatment of Pleural Mesothelioma
Multidisciplinary Management of Patients with Localized Bladder Cancer
Advances in the Multimodality Management of High-risk Prostate Cancer
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
