The prognosis for patients with gastric cancer depends on the stage of the disease at the time of diagnosis and treatment.¹² After an R0 resection patients with early-stage gastric cancer have a good prognosis with cure rates exceeding 80%. After R0 resection, however, patients with advanced cancers have less than 50% chance of cure. The need for an adjuvant treatment to eradicate residual microscopic disease after potentially curative surgery is clear.³#

The major treatment strategy pursued during the last decades has been postoperative systemic therapy. Hermans et al. reviewed 11 trials reported since 1980.⁴ In these studies, surgery was followed by postoperative chemotherapy or by observation alone. Two thousand and ninety-six patients were studied. The odds ratio was 0.88 (0.78–1.08), which was slightly, but not significantly, in support of adjuvant treatment. Earle and Maroun reported the results of a meta-analysis involving 13 trials performed in non-Asian countries.⁵ In this analysis, the odds ratio for death for patients receiving adjuvant therapy was 0.80 (0.66–0.97) with a relative risk of 0.94 (0.88–1.01). Subgroup analyses showed a trend towards a larger magnitude of the effect when analysis was restricted to trials in which at least two-thirds of patients had node-positive disease. The authors concluded that the survival benefit from these trials in patients undergoing curative resections was small. Panzini et al. reviewed 17 randomized controlled clinical trials testing systemic adjuvant chemotherapy after radical resection of gastric cancer.⁶ A total of 3118 patients were included. They found a significant advantage in survival for adjuvant chemotherapy with the odds ratio of 0.72 (0.62–0.84).#

There are some reports on the survival benefit with adjuvant chemotherapy. Grau et al. reported that adjuvant chemotherapy with mitomycin C alone or with tegafur was associated with long-term survival benefit in resected gastric cancer.⁷ Survival benefit was seen in patients who had received adjuvant chemotherapy compared with the controls (52% vs. 30% alive at the end of the study; p<0.0001). However, most reports concerning mitomycin containing regimens including mitomycin C and ftorafur did not show survival benefit. Nakajima and colleagues reported the results of a large multicenter phase III trial performed in Japan.⁸ Five hundred and seventy-nine patients with serosa-negative gastric cancer were randomized to receive postoperative chemotherapy or observation. The treatment group had intravenous mitomycin and fluorouracil, and oral tegafur and uracil (UFT). Two hundred and eighty-eight patients received treatment and 285 were followed expectantly. At median follow-up of 72months, the overall 5-year survival rate was 82.9% for those in the control group and 85.8% for those in the treatment group (p=0.17). The authors concluded that this regimen had no survival benefit in patients with serosa-negative gastric cancer after curative gastrectomy. They recommended exclusion of T1 cancer from future trials, because curative surgery alone resulted in a high survival rate with no need for adjuvant therapy.#

The uses of several doxorubicin-containing adjuvant combination chemotherapy regimens have been reported. Neri et al. used a regimen of epirubicin, leucovorin, and 5-fluorouracil.⁹ Forty-eight patients received chemotherapy and 55 patients were followed expectantly. After a follow-up period of 36months, 25% of the treated patients and 13% of controls were alive. The median survival for patients receiving chemotherapy (20.4months) was superior to those undergoing observation (13.6months). Krook and colleagues randomized 125 patients to either receive three cycles of 5-fluorouracil and doxorubicin or be observed.¹⁰ Treatment began between 4 and 6weeks after resection. The 5-year survival rate was 33% for the observation arm and 32% for the adjuvant therapy arm. Coombes and colleagues studied 315 patients with curatively resected gastric cancer who were randomized to receive FAM regimen (5-fluorouracil, doxorubicin, and mitomycin) or no postoperative therapy.¹¹ Chemotherapy could be started as late as 6weeks from surgery. At median follow-up of 68months, 164 patients had died, 73 in the treated arm and 91 in the control arm. There was no significant difference in survival between the two arms of the study (p=0.21). There were some evidence that patients with more advanced carcinoma (T3–T4) derived some benefit from treatment (p=0.04). The authors recommended further prospective randomized evaluation in T3–T4 patients. The Southwest Oncology Group also failed to demonstrate an improvement in survival for the treated group.¹² One hundred and ninety-three patients with resected stage I, II, or III gastric carcinoma were accrued in a phase III trial comparing six cycles (1year) of postoperative FAM chemotherapy with observation only. No differences in disease-free survival (p=0.45) and overall survival (p=0.57) between FAM therapy (93 cases) and surgery (100 cases) were observed. However, cases with curative resection, defined in a retrospective review of pathology and surgical reports as cases having no evidence of residual disease in the abdomen and tumor-free margins >1cm, had superior survival compared to cases not meeting these requirements (p<0.001). The authors emphasized surgical quality control and adequate case accrual over a reasonable period in future adjuvant studies. Tsavaris et al. performed a randomized trial of FEM regimen (5-fluorouracil, epirubicin, and mitomycin).¹³ Eighty-four patients were randomized to receive three cycles of FEM (42 patients) or no treatment (42 patients). Chemotherapy was started 2–3weeks after surgery. At the last follow-up at 66months, 27 patients (64%) in chemotherapy group had relapsed or died, compared to 34 patients (81%) in control group. The differences in the relapse and the disease-free and the overall survival rates were not statistically significant. Nitrosourea-containing regimens also did not show survival benefit.¹⁴#

Japanese and Korean researchers have performed a number of trials investigating the use of immunochemotherapy as an adjuvant treatment after curative resection of gastric cancer. A variety of immunotherapeutic agents, such as protein-bound polysaccharide (PSK),¹⁵ Streptococcus pyrogenes preparation (OK432),¹⁶¹⁷ polysaccharide sizofiran (SPG),¹⁸ and Nocardia rubra cell wall skeleton,¹⁹ were used. Results of these investigations suggest that immunotherapy may show a small survival benefit for patients undergoing potentially curative resection. These results need further confirmation.#

The British Stomach Cancer Group randomized 436 patients to adjuvant radiotherapy or chemotherapy with FAM regimen after gastrectomy for adenocarcinoma.²⁰ At 5-year follow-up, no survival advantage was shown for those patients receiving either adjuvant therapy compared to those undergoing surgery alone. The 5-year survival for surgery alone was 20%, for surgery plus radiotherapy 12% and for surgery plus chemotherapy 19%.#

Macdonald et al. randomized 556 patients with resected adenocarcinoma of the stomach or gastroesophageal junction to surgery plus postoperative chemoradiotherapy or surgery alone.²¹ The adjuvant treatment consisted of fluorouracil plus leucovorin, followed by radiation. One month after the completion of radiotherapy, two 5-day cycles of fluorouracil plus leucovorin were given 1month apart. Postoperative adjuvant chemoradiotherapy resulted in a significant improvement in survival rate. The median overall survival in the surgery-only group was 27months, as compared with 36months in the chemoradiotherapy group; the hazard ratio for death was 1.35 (95% confidence interval, 1.09–1.66; p=0.005). The hazard ratio for relapse was 1.52 (95% confidence interval, 1.23–1.86; p<0.001). These data suggest postoperative chemoradiotherapy may be of value.#

The rationale for an intraperitoneal route for chemotherapy administration is based on the high risk of peritoneal metastasis as a component of first failure and the pharmacokinetics of intraperitoneal administration. The literature reveals disagreements over patterns of failure after surgical treatment. Recurrence rates reported after potentially curative surgery are 21.8–48.8%.^22–25 Sites of recurrence can be summarized as follows: local recurrence rates of 22.3–31.8%, peritoneal 13.5–53.5%, hematogenous 17.0–53.8%, and lymphatic 12.3–28.6%.^25–28 These disagreements in the patterns of surgical treatment failure are likely related to the patient cohorts accepted for evaluation, the time at which failure was determined, the method of determination of failure patterns, and the definition of curative surgery. Autopsy series and second-look laparotomy series have reported that up to 50% of patients have clinically evident peritoneal carcinomatosis as a site (sometimes the only site) of failure. Even after extended lymphadenectomy, the peritoneal surface and the liver are major sites of recurrence, although the rate of local-regional relapse is considerably lower compared to a more limited surgery.²⁹ From a pharmacologic perspective there is a strong rationale to use intraperitoneal chemotherapy to treat the patterns of failure that occur after gastric cancer surgery. Drug concentrations within the peritoneal cavity are severalfold to two logs higher than concentrations that can be achieved after oral or intravenous treatment. Also, not only the route but also the timing of adjuvant intraperitoneal chemotherapy administration is crucial for benefits.³⁰ The technique most commonly used is to administer intraperitoneal treatment (with or without hyperthermia) following the resection of all gross disease. In this method intraperitoneal chemotherapy is delivered in the operating room. Alternatively, early postoperative intraperitoneal chemotherapy is given within a few days of resection.#

The initial promising results of Hagiwara and colleagues indicated a marked improvement in survival for patients randomized to intraperitoneal chemotherapy with mitomycin C adsorbed carbon-particle-containing solution compared to those receiving no postoperative antitumor treatment (p<0.01).³¹ Side effects were slight and well tolerated. Rosen and colleagues reported their results using a similar technique.³² Ninety-one patients were randomly assigned to resection followed by observation or resection followed by mitomycin C adsorbed carbon particles given intraperitoneally. The study was stopped prematurely when an interim analysis revealed a marked increase in postoperative complications and increased perioperative mortality in patients given intraperitoneal mitomycin C adsorbed carbon particles. No survival advantage was noted at the time of the interim analysis, and the trial was closed.#

A phase II study indicated that postoperative intraperitoneal chemotherapy using cisplatin and 5-fluorouracil could be given safely.³³ With a median follow-up of 24months, 51% of patients remained alive and free of disease. Toxicity was acceptable. Sautner and colleagues performed a randomized phase III trial using intraperitoneal cisplatin as a single agent.³⁴ Therapy began 2–4weeks after surgery. No difference was found in survival for patients receiving or not receiving intraperitoneal chemotherapy. The authors concluded that intraperitoneal cisplatin as a single agent had no significant impact on recurrence. This study has been criticized in that the timing of the cisplatin infusion was such that postoperative adhesions would exclude intraperitoneal drug from crucial anatomic sites at high risk for cancer seeding.#

Yu and colleagues reported favorable results of a prospective randomized trial in which 248 patients were randomly assigned to receive intraperitoneal therapy with mitomycin C and 5-fluorouracil or to be observed.³⁵ As other mitomycin C containing trials, there was a tendency towards increased mortality with intraperitoneal chemotherapy.#

Continuous hyperthermic peritoneal perfusion (CHPP) for the prophylaxis of peritoneal carcinomatosis is administered during surgery after a primary tumor resection in patients who are considered at high risk for developing peritoneal recurrence. CHPP takes advantage of the favorable pharmacokinetics that can be achieved with intraperitoneal chemotherapy and the established synergistic cytotoxicity of hyperthermia and chemotherapy. Yonemura and coworkers introduced CHPP with mitomycin C and cisplatin as a prophylactic treatment for peritoneal recurrence after curative resection of 79 advanced gastric cancers.³⁶ Survival was compared to that of 81 patients treated during the same period. Prolonged survival was reported in the subgroup of patients with serosal invasion-positive tumors. No difference in morbidity or mortality was reported between the two groups. Fujimoto and colleagues performed a randomized controlled trial in which 141 gastric cancer patients with macroscopic serosal invasion were randomly assigned to receive intraperitoneal hyperthermic chemoperfusion (IHCP) or to observation.³⁷ The peritoneal recurrence rate in the IHCP group was significantly decreased (p<0.0001) compared with that in the control group. The IHCP group showed a significant survival benefit (p=0.0362) compared with the control group. Hamazoe and colleagues reported that the 5-year survival rate (64.2%) of 42 patients in the CHPP group was higher than that (52.5%) of 40 patients in the control group.³⁸ This difference was not significant. Kunisaki and coworkers reported that CHPP had no efficacy as prophylactic treatment for peritoneal recurrence induced by gastric cancer.³⁹#

From the analyses of data on surgical treatment failure, pharmacokinetics of intraperitoneal drugs, and the poor results of other adjuvant therapies, we conducted a prospective randomized trial to clarify the advantages and disadvantages of the early postoperative intraperitoneal chemotherapy in patients with resected advanced gastric cancer. Two hundred and forty-eight patients with biopsy proven gastric cancer were enrolled in the study. Patients were randomized intraoperatively after resection was complete to receive early postoperative intraperitoneal chemotherapy (EPIC) or not to. All patients underwent D2 or more extended lymphadenectomy. In patients randomized to receive EPIC, a 5-day course of intraperitoneal chemotherapy with mitomycin C and 5-fluorouracil was given through the catheters placed during the operation. Results of this trial, which demonstrated a statistically significant improvement in survival, were reported previously.⁴⁰#

Overall 5-year survival rate was 61.1% and 10-year survival 55.8%. Depth of invasion (p<0.0001), lymph node metastasis (p<0.0001), stage (p<0.0001), tumor location (p=0.0001), Borrmann type (p=0.0002), and extent of gastric resection (p=0.0347) significantly influenced survival on univariate analyses. Survival distributions for EPIC adjusted for stage showed statistically significant improvement especially in patients with stage III gastric cancer (p=0.0288; Table 1).#

Postoperative mortality was higher in the study group (4.9%) than in controls (1.1%), but not significantly (p=0.149). Intra-abdominal sepsis (p=0.010) and bleeding (p=0.043) occurred significantly more often in the study group. Patients treated with EPIC stayed in the hospital an average of 6days longer (p<0.001). And, 30.5% of patients who received EPIC experienced a variety of complications attributable to EPIC. Most of these complications could be managed conservatively without any significant sequelae.#

Precise prediction of the risk of recurrence is important in the selection of adjuvant chemotherapy. The proper selection of patients for adjuvant chemotherapy using objective criteria may not only improve the prognosis of gastric cancer patients but also optimize quality of life by eliminating patients in no need of therapy. Because EPIC carries a well-defined morbidity, only patients with a chance for benefit should be selected for treatment.#

Peritoneal recurrence is associated with serosal invasion,²⁵ lymph node involvement,²⁴ infiltrative growth,²⁸ large tumor,⁴¹ gross appearance,⁴² and scirrhous-type stromal reaction.²⁵ Intraoperative peritoneal lavage cytology had been advocated by some researchers because it appeared to be useful for identifying free cancer cells in the peritoneal cavity and it could be a strong indicator of future peritoneal metastasis.⁴³ However, Abe and colleagues reported that peritoneal lavage cytology was insensitive for predicting the development of peritoneal recurrence because such recurrence was seen in 22% of patients with negative cytology and 18% of patients with positive cytology.⁴⁴ A number of sophisticated methodologies using immunohistochemistry or reverse transcriptase–polymerase chain reaction are available for selecting high risk patients. However, cost and availability should also be considered.#

Macroscopic serosal invasion, one of the most important risk factors of peritoneal dissemination of gastric cancer,⁴⁵ is a useful selection factor for the adjuvant intraperitoneal chemotherapy.⁴⁰ When analysis was performed for patients reported to have macroscopic serosal invasion, there were 52 patients in the control group and 69 in the EPIC group. Again, we could identify the difference between the survival distributions. EPIC showed an improved survival rate with a p value of 0.0082 (Fig. 1). Other factors that influenced survival significantly were depth of invasion, lymph node metastasis, stage, tumor location, and Borrmann type (Table 2). When a multivariate analysis by the Cox proportional hazards model was performed, EPIC emerged as the statistically significant and independent prognostic factors (p=0.007; risk ratio=0.453). Analysis of the patterns of failure demonstrated a statistically significant difference between groups. The incidence of peritoneal dissemination after surgery with EPIC was 8.7% and that after surgery only was 26.9% (p=0.041).#

R0 corresponds to complete resection for cure. It is appropriate for cases in which residual tumor cannot be detected by any available diagnostic means. The effectiveness of adjuvant therapy and the possible eradication of micrometastatic foci outside the surgical field can be achieved by the timely administration of chemotherapeutic agents to the predicted recurrence site. On a theoretical basis and the result of our phase III study, perioperative intraperitoneal chemotherapy can be suggested as an adjuvant therapy for gastric cancer. Subgroup analyses of our data indicated significant improvement in survival for those with serosa-positive gastric cancer, but these were the results of retrospective and unplanned analyses. Therefore, these observations may require a larger confirmatory trial.#

Quality of surgical resection affected survival irrespective of adjuvant therapy.¹² Curative surgery should be R0 that includes adequate resection margins and an extended lymphadenectomy. Indication for adjuvant perioperative intraperitoneal chemotherapy must be the macroscopic serosal invasion. Chemotherapeutic agents can be administered intraoperatively for better access of drugs to the peritoneal surface. Hyperthermic intraperitoneal chemotherapy is used to augment the cytotoxic effect of the chemotherapeutic agents. However, the prolonged anesthetic time must be considered and this technique needs more complicated equipments. Also, possible complications of heat must be considered. Early postoperative intraperitoneal chemotherapy is simple and easy. We can use the conventional drugs such as mitomycin C, 5-fluorouracil, and cisplatin. In the future we can consider the use of newer chemotherapeutic agents, including irinotecan, oxaliplatin, and taxanes. With a small number of patients in the trial, only very large differences can be expected to show a statistical benefit. A high priority should be placed on designing a trial with adequate numbers of patients to allow an assessment of benefit.#