A total of 33 patients with gastric cancer (mean age 70 years; male/female, 25/8) who had undergone ESD at Nagasaki University Hospital (Nagasaki, Japan) between January 2010 and September 2011 were included in the present study as primary subjects (Step 1). The primary subjects met the following inclusion criteria: i) patients with gastric cancer subjected to ESD, ii) patients aged >20 and <81 years, iii) patients for whom oral administration was possible, and iv) patients who provided written informed consent regarding their participation in the study. The exclusion criteria were the following: i) presence of an additional type of cancer, ii) patients who needed chemotherapy, iii) patients with serious complications (liver, kidney, heart, blood or metabolic disorders), iv) patients with alcohol or drug dependence, v) patients under treatment for a psychological disease, vi) patients with ulcerative colitis, Crohn’s disease or irritable bowel syndrome (IBS), vii) women who were pregnant or wished to become pregnant during the study or the follow-up period, as well as lactating women, viii) patients who had received traditional Japanese medicine including the test drug within one month prior to the administration of the test drug, ix) patients with a history of hypersensitivity to traditional Japanese medicine including the test drug, and x) patients who were not considered eligible for inclusion in the present study by the chief investigator.

This study (UMIN000002302) was a prospective, randomized, parallel, comparative study to examine the pharmacological effects, efficacy and safety of drug therapy in gastric cancer patients subjected to ESD. This study was conducted based on ethical guidelines for clinical studies, taking into consideration the human rights and privacy of the patients. The protocol of this study was approved by the Institutional Review Board of Nagasaki University Hospital.

The study procedures followed are shown in Fig. 1. After obtaining written informed consent regarding participation prior to ESD, patients who fulfilled all the inclusion criteria were included in this study. The GI symptoms of the patients were assessed using the Gastrointestinal Symptom Rating Scale (GSRS) 6–8 days following ESD. Gastric emptying was evaluated using the [¹³C]-labeled acetate breath test (described in detail later) that consisted of the administration of a liquid meal [OKUNOS-A (200 ml) containing ¹³C-sodium acetate (100 mg)] and the determination of the peak time of the 13C% dose-excess curve (T-max) after the evaluation of GI symptoms. The Japanese Ministry of Health, Labour and Welfare has approved rikkunshito for the treatment of abdominal pain and indigestion. Thirteen patients who scored ≥3 more than the average GSRS scores for abdominal pain or indigestion 6–8 days after ESD were included in the Step 2 study, and were randomized to the PPI group [standard treatment with rabeprazole monotreatment (20 mg/day), twice/day (b.i.d); n=5] or the PPI + rikkunshito group [rabeprazole (20 mg/day, b.i.d) combined with rikkunshito (7.5 g/day), three times/day (t.i.d); n=8]. The respective GSRS scores of the two groups after 4 and 8 weeks of treatment were compared with baseline values.

The GI symptoms were evaluated using the GSRS, which is an inquiry table consisting of 15 items for the evaluation of general GI symptoms (15). Each GSRS item is rated on a 7-point Likert scale ranging from no discomfort to very severe discomfort. Based on a factor analysis, the 15 GSRS items break down into the following five scales: abdominal pain (abdominal pain, hunger pain and nausea), reflux syndrome (heartburn and acid regurgitation), diarrhea syndrome (diarrhea, loose stools and urgent need for defecation), indigestion syndrome (borborygmus, abdominal distension, eructation and increased flatus) and constipation syndrome (constipation, hard stools and a feeling of incomplete evacuation).

In the Step 2 study, the overall GSRS score and subscales after 4 and 8 weeks of treatment were compared against baseline values to evaluate the potential improvement of GI symptoms in each group.

The gastric emptying test was performed according to the protocol of the ¹³C-acetate breath test standardized by the Japan Society of Smooth Muscle Research (16). Briefly, a commercially available diet (OKUNOS-A; Horika Foods, Tokyo, Japan) was used as the test meal. A 200 ml portion of the diet contained 9.8 g protein, 5.2 g fat, 28.6 g carbohydrate and 200 kcal. The test meal was enriched with 100 mg ¹³C-sodium acetate (Wako Pure Chemicals Industries, Osaka, Japan). Breath samples were collected into 300 cc packs (Otsuka Pharmaceutical Co., Tokyo, Japan). The analysis of the ¹³CO₂/¹²CO₂ enrichment in the breath samples was performed using an infrared spectrophotometer (UBiT-IR200; Otsuka Electronics Co., Tokyo, Japan). The T-max of gastric emptying was calculated using Excel software (Star Medical, Tokyo, Japan). The cumulative excretion of ¹³CO₂ (as a percentage of the ingested dose) was also calculated.

Once baseline measurements had been taken after an overnight fast, subjects consumed the test meal within 15 min. Upon completion of the meal (t=0 min), sequential postprandial measurements of gastric emptying were taken. Expired breath samples were collected at t=0 min and every 5 min for the first half hour after the meal was consumed, and at 15-min intervals from t=30 min until 90 min for the detection of ¹³CO₂.

Safety and tolerability were assessed by recording all adverse events, and changes in hematological and clinical laboratory variables were measured during the screening visit and after the post-dose esophagogastroduodenoscopy. An adverse event was defined as any unfavorable or unintended sign, whether or not it was considered to be causally related to the drugs used in this study.

Treatment compliance was defined as the percentage of the test drug used. A treatment compliance of ≥66.6% was considered acceptable.

The treatment response in each group was evaluated based on changes in the GSRS or GSRS sub-item scores prior to and following treatment using the Wilcoxon signed-rank test. Furthermore, the improvement rate in each subject was calculated from these scores, and the mean values were compared between the two groups using the Wilcoxon rank-sum test. This test was also used to compare background factors such as age and body mass index (BMI). The distribution of gender was compared using Fisher’s exact test. P<0.05 was considered to indicate a statistically significant difference. All data are expressed as the mean ± standard deviation (SD).

Patient characteristics including mean age, male/female ratio, BMI and clinicopathological characteristics of EGC are shown in Table I. The standard guideline criteria and expanded criteria for ESD were established by the Japanese Gastric Cancer Association (17,18). The standard guideline criteria were defined as differentiated mucosal cancer of ≤2 cm without ulcer. The expanded criteria encompassed non-ulcerative differentiated cancer >2 cm, ulcerative differentiated cancer <3 cm, non-ulcerative undifferentiated cancer <2 cm and differentiated, submucosal invading (limited to 500 μm below the lamina propria) cancer <3 cm in diameter. In 11 patients the lesion fulfilled the standard guideline criteria, in seven patients the lesion fulfilled the expanded criteria, and in 15 patients the lesion was an adenoma or atypical duct hyper-plasia that did not comply with either the standard guideline or expanded criteria. The lesions were removed by en bloc resection in all patients. The incidence of local tumor recurrence was 3% (complete removal:incomplete removal, 32:1).

Of the 33 patients who had undergone ESD, the evaluation of gastric emptying was possible in 32 patients. The reference range of T-max was derived from 63 healthy volunteers by the Japan Society of Smooth Muscle Research; the mean ± SD of the T-max was 43.9±10.3 min (16). Gastric emptying following ESD in the patients with EGC in the present study is shown in Fig. 2. The mean T-max 6–8 days after ESD was 75.4±13.6 min which was significantly increased compared with that in healthy subjects (43.9±10.3 min).

GI symptoms were evaluated in 29 of the 33 patients subjected to ESD. Of these 29 patients, 25 were scored ≥3 (≥mild discomfort) for at least one of the 15 GSRS items. The ratio of patients with GI symptoms (≥mild discomfort) according to lesion location is shown in Fig. 3. The lesion location was defined by anatomically dividing the stomach into three portions according to the Japanese classification of gastric carcinoma: the upper (U), middle (M) and lower (L) parts (17,18). Constipation (mean of U, M and L, 56%), a sense of incomplete evacuation (44%), hard stools (33%), abdominal distension (32%) and abdominal pain (30%) were reported in ≥30% of the patients irrespective of the lesion location. The incidences of nausea and a sense of incomplete evacuation were higher in the patients whose lesion was located in the U portion (33 and 67%, respectively) compared with those in the patients whose lesion was located in the M (0 and 44%, respectively) or L portions (14 and 21%, respectively). The incidence of upper abdominal pain was higher in the patients whose lesion was located in the L portion (50%) than in those whose lesion was located in the U (17%) or M portions (22%).

Thirteen patients who scored ≥3 more than the average scores for abdominal pain or indigestion 6–8 days after ESD were included in the Step 2 study. The overall scores and the five subscale scores of GSRS after 4 and 8 weeks of treatment with a PPI alone (rabeprazole 20 mg/day, b.i.d, n=5) or a PPI (rabeprazole 20 mg/day, b.i.d, n=8) combined with rikkunshito (7.5 g/day, t.i.d) are shown in Table II. The overall scores and five subscale scores of GSRS did not change after 4 and 8 weeks of treatment with the PPI alone, while a significant improvement of overall GSRS scores was observed after 4 and 8 weeks of treatment with PPI plus rikkunshito (P=0.1313). Moreover, the mean abdominal pain score was significantly decreased after 8 weeks of treatment with PPI plus rikkunshito (Fig. 4), indicating the beneficial effects of rikkunshito, particularly against abdominal and hunger pains (Fig. 5).