Gastric cancer is a global health issue. Worldwide, it is the second most common cause of cancer-related mortality.[1] In the United States, gastric cancer is the eighth most common cause of cancer- related death.[2] Cancer of the gastric antrum has decreased in incidence in the United States and western Europe; however, this has been countered by an increased incidence of cancer of the gastric cardia.[3-5] In the Far East, particularly Japan, there has been a notable improvement in mortality associated with this disease.[ 6] This has been partially attributed to early detection of tumors by screening and surveillance programs. Cancers of the distal stomach are more common in Asia, and exhibit the following characteristics: (1) They are associated with Helicobacter pylori infection, (2) they follow a chronic atrophic gastritis/metaplasia/dysplasia sequence, and (3) upon histology, they are usually the intestinal type.[1,7] On the other hand, cancers of the proximal stomach have been associated with improved socioeconomic conditions and are less often associated with H pylori.[1,7] Proximal tumors are associated with early hematogenous spread as compared to slower progression and locoregional spread in the case of distal tumors.[7] There has been a recent increase in the incidence of proximal gastric and gastroesophageal junction cancers. The reasons for this increase are unknown.[ 3-5] Siewert and Stein classified adenocarcinoma of esophagogastric junction (AEG) into three distinct entities based on anatomic origin: (1) adenocarcinoma of the distal esophagus invading the gastroesophageal junction (AEG type I tumors), (2) carcinoma of the gastric cardia immediately at the gastroesophageal junction (AEG type II tumors), and (3) subcardial gastric carcinoma infiltrating the gastroesophageal junction (AEG type III tumors).[8] AEG type II and III tumors are more likely to be undifferentiated and have a worse prognosis.[1] Thus, although gastric cancer is referred to as one entity, there is heterogeneity due to epidemiologic, clinical, and therapeutic differences between cardiac, antral, and gastroesophageal junction cancers. In the United States and Europe, potentially curative resections (with negative margins, R0) are possible in only about 50% of newly diagnosed gastric cancer patients.[9-11] For patients with locally advanced tumors (stages II, IIIA, or IIIB) even after gastric resection with curative intent, the recurrence rate is as high as 40% to 65%.[12] As most cases are diagnosed at an advanced stage, there is clearly a need to develop innovative treatment strategies that will downstage the tumor, increase the R0 resection rate, and decrease the risk of recurrence after surgery. Adjuvant chemoradiation is commonly used postoperatively for gastric cancer. The ability to deliver adjuvant chemoradiation is adversely influenced by the 25% to 46% postoperative morbidity seen after gastrectomy with lymphadenectomy.[13,14] Only 64% of patients completed adjuvant chemoradiotherapy in the important Intergroup 0116 trial.[15] Neoadjuvant therapy for gastric cancer is emerging as a promising approach. This article discusses the rationale for neoadjuvant therapy and the available therapeutic options. A brief overview of current staging and operative and adjuvant strategies is presented below as a preface to this discussion. Staging Strategies
National Comprehensive Cancer Network (NCCN) guidelines recommend a multidisciplinary approach to staging that includes a complete history and physical examination, laboratory studies (complete blood count, platelet count, and serum biochemistry), esophagogastroduodenoscopy, chest radiography, and computed tomography (CT) of the abdomen and pelvis. Laparoscopy has also been added to these guidelines as a category 2B recommendation (nonuniform consensus among panel members).[16] Table 1 depicts the various staging modalities available for gastric cancer, along with accuracy, advantages, and disadvantages. The use of positronemission tomography (PET) for staging, detecting recurrence, and evaluating response to therapy in gastric cancer is evolving. A recent study in esophageal cancer suggests that quantitative measurements of tumor ¹⁸F-fluorodeoxyglucose (FDG) uptake may predict histopathologic tumor response and patient outcome as early as 2 weeks after initiation of preoperative chemotherapy.[ 17] This may be a useful investigational tool in gastric cancer as part of a neoadjuvant strategy. The recent availability of integrated CT/PET imaging is likely to refine staging of these tumors in the future. Japanese investigators use a computer- based program (the Maruyama program) to predict lymph node stations at risk for metastasis in particular cases. The cases are sorted on the basis of similarity with seven specified variables: age (± 5 years), sex, Borrmann type of the tumor, greatest dimension of the tumor as measured on the luminal surface (± 2.5 cm), location of the tumor, estimated tumor depth, and histology. Comparison is made with a database of 3,843 gastric cancer patients at the National Cancer Center Hospital in Tokyo who had been treated by D2 or more extensive lymphadenectomy. The program then predicts the percentage likelihood of disease at each of the 16 lymph node stations around the stomach on the basis of actual patient experience.[18] This information could be used prospectively to enroll higher-stage patients into neoadjuvant programs. The Role of Surgery Surgical resection remains the only curative therapy for this disease. The debate continues regarding the optimal extent of lymph node dissection required for maximal therapeutic benefit. Extended lymphadenectomy remains the standard of care in the Far East at the cost of acceptable postoperative morbidity.[19] Similar results could not be reproduced in Western trials and no survival advantage was attributed to D2 dissections in the Dutch Gastric Cancer Group trial.[14] Similar results were noted in the Medical Research Council (MRC) trial from the United Kingdom.[13] The results of randomized control trials comparing D1 with D2 dissections are depicted in Table 2. These randomized trials did not reveal any survival advantage for D2 dissection. Morbidity and mortality associated with D2 dissection was significantly higher in these randomized trials. Five-year survival after D1/D2 dissection remains dismal at 30% to 45%. Thus, no phase III study in the Western population has proved the therapeutic advantage of D2 dissection. Cochrane database review done by McCulloch et al revealed survival benefit of D2 dissections in T3 or more advanced lesions.[20] Subgroup analysis of the MRC trial suggests a survival advantage for stage II and III patients or those with N1 disease.[13] The Dutch Gastric Group trial indicated a survival benefit for N2 patients.[ 14] Siewert et al, in a large prospective multicenter observational trial, found that the pathologic subgroup of pT2, N1 and pT3, N0 had a significant survival benefit with extended lymph node dissection.[21] The extent of gastrectomy may also impact on postoperative morbidity and quality of life. Due to high incidence of locoregional failure, multifocal disease, and submucosal pattern of spread, total gastrectomy with at least a 5-cm gastric margin is recommended for proximal gastric tumors.[9] Subtotal gastrectomy is often the procedure of choice for distal (antral) tumors.[22] Although total gastrectomy is superior in terms of margin-free resection, this procedure is associated with postoperative morbidity of dumping syndrome, malabsorption, anemia, osteoporosis, and malnutrition. Quality- of-life impairment results from these difficulties. Davies et al demonstrated that gastric cancer patients who underwent subtotal gastrectomy enjoyed superior quality of life at the end of 1 year as compared with total gastrectomy patients.[23] Proximal gastric resection is associated with a higher risk of anastomotic leakage and bile-acid reflux. This complication may be prevented by jejunal interposition.[9] Based on the above observations, it may be worthwhile to investigate neoadjuvant therapy followed by proximal gastrectomy as a possible alternative to total gastrectomy for proximal gastric cancers. A recent trial by Bosing et al revealed a dismal 14% 5-year survival in patients with AEG type III tumors, despite surgery.[24] Linitis plastica and signet-ring histology are other gastric cancer subtypes that have a poor outcome with surgical therapy alone.[25] Thus, patients with stages II, IIIA, or IIIB (locally advanced) proximal tumors or those with aggressive histology are likely to benefit from neoadjuvant therapy, as surgery alone is usually not curative in these cases. The Role of Adjuvant Therapy The use of postoperative adjuvant chemoradiotherapy for gastric cancer gained popularity after a randomized trial by Macdonald et al.[15] The trial showed improved survival in the adadjuvant therapy arm receiving 45 Gy of radiotherapy and three cycles of fluorouracil (5-FU) and leucovorin compared to surgery alone. Median overall survival in the surgery-only group was 27 months, as compared with 36 months in the chemoradiotherapy group (P = .005). Critics of this trial note that over half of the patients received inadequate lymphadenectomy (most had D0 dissection); hence, adjuvant therapy may have been advantageous for these patients, many of whom possibly had residual nodal disease. In the Western world, adjuvant therapy is considered standard of care since the rate of R0 resection (microscopically negative postoperative margins) remains suboptimal. The role of adjuvant therapy after D2 dissection is unproven. Neoadjuvant Therapy
Neoadjuvant therapy has a number of theoretical advantages. First, neoadjuvant therapy may shrink and "downstage" the tumor, hence allowing complete resection of cancers previously considered inoperable. Second, radiation treatment volumes are more precise and tissues are better oxygenated when radiation is administered preoperatively. Third, postoperative morbidity and significant weight loss of up to 10% precludes administration of adjuvant therapy in more than 30% of patients.[26] Neoadjuvant therapy, including chemotherapy and radiation, is better tolerated and feasible in most patients. This strategy offers patients the maximum benefit of all available modalities of treatment. Fourth, neoadjuvant therapy provides a unique mechanism to assess in vivo response to chemotherapeutic agents and an opportunity to study molecular changes with therapy. Neoadjuvant therapies may include preoperative chemotherapy, preoperative external-beam radiation therapy (EBRT), concurrent chemoradiotherapy (chemo-RT), and intraperitoneal chemotherapy. These modalities are discussed below. Preoperative Chemotherapy
Selected phase II trials using preoperative chemotherapy are shown in Table 3. Studies that reported median overall survival and included either D2 dissection or comprehensive staging workup were selected. These trials demonstrated that chemotherapy delivered in the preoperative setting is feasible and well-tolerated. These small trials suggest an improved R0 resection rate and a favorable median survival compared with historical controls (median survival of patients who underwent surgery alone in the Intergroup 0116 trial was 27 months).[15] An observation by Lowy et al was that the chemotherapy responders had a twofold improvement in overall survival compared to nonresponders.[27] Although promising, a final answer regarding the efficacy of neoadjuvant chemotherapy can only be provided by a well-powered phase III study. Currently, three large phase III trials investigating the role of neoadjuvant therapy are under way. These include the MRC Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial, the Swiss Study SAKK 43/99, and the multinational European trial EORTC (European Organisation for Research and Treatment of Cancer) 40954.[13,28] The MAGIC trial accrued 503 gastric cancer patients between 1994 and 2002.[13] Patients had stage II or more advanced gastric cancer and were randomized to perioperative chemotherapy with three preoperative and three postoperative cycles of epirubicin (Ellence), cisplatin, and 5-FU (the CSC [chemotherapy->surgery->chemotherapy] arm) or surgery alone. The preliminary results of the trial are depicted in Table 4. The R0 resection rate was significantly higher in the CSC arm with no significant difference in postoperative complications or length of postoperative hospital stay. Progression-free survival was superior in the CSC arm as compared with the surgery-alone arm (hazard ratio = 0.70, 95% confidence interval [CI] = 0.56-0.88, P = .002). The trend toward improved survival awaits trial completion. Two elements of the study design of the MAGIC trial deserve mention. First, about 11% of patients had distal esophageal cancer. A prior Intergroup study in esophageal cancer revealed no survival advantage with preoperative chemotherapy; however, there are limited data in gastric cancer.[29] This may confound study results. In addition, neither endoscopic ultrasonography nor diagnostic laparoscopy was performed to accurately assess stage of disease preoperatively. Despite these problems, the MAGIC trial is the first well-powered, randomized, controlled phase III trial that demonstrates the safety and feasibility of perioperative chemotherapy. Final efficacy results of this study are eagerly anticipated. The Swiss study SAKK 43/99 compares preoperative docetaxel (Taxotere), cisplatin, and 5-FU followed by surgery, vs surgery followed by the same chemotherapy regimen. The EORTC 40954 trial compares surgery plus preoperative cisplatin, leucovorin, and 5-FU with surgery alone. The results of these trials will help establish whether there is a benefit of neoadjuvant chemotherapy and may provide information regarding the superiority of specific neoadjuvant chemotherapy regimen. Preoperative EBRT
The rationale for neoadjuvant radiation is reduction of local and regional recurrences. External-beam radiation therapy also has a palliative role for the management of bleeding tumors. The location and extent of tumor is best imaged in the preoperative setting. As mentioned earlier, delivery of radiation therapy preoperatively may enhance efficacy, reduce treatment volumes, and decrease toxicity. Almost one-third of the radiation fields had to be redesigned in the postoperative adjuvant setting in the Intergroup 0116 trial.[15] Preoperative EBRT allows the patient a better chance of receiving all available therapies, as postoperative morbidity after gastrectomy is not insignificant. Table 5 shows trials comparing preoperative radiation plus surgery with surgery alone. These studies used EBRT doses between 20 and 40 Gy. However, more current combinedmodality regimens utilize EBRT doses up to 45 Gy (Table 6). Zhang et al demonstrated a significant improvement in 5-year survival after 40 Gy was given preoperatively.[52] There did not appear to be an increase in postoperative morbidity related to the therapy; these studies indicated a trend toward improved survival.