Clinical characteristics of elderly patients with proton pump inhibitor-refractory non-erosive reflux disease from the G-PRIDE study who responded to rikkunshito

The G-PRIDE study (UMIN000005880) was a prospective, multi-center, randomized, double-blind, paralleled comparative study that examined the pharmacological effects, efficacy, and safety of drug therapy in patients with PPI-refractory NERD in 55 hospitals in Japan. The design and primary results of the prospective clinical trial have been previously described elsewhere [16]. This report includes prospectively defined subgroup analyses of clinical characteristics of elderly patients with PPI-refractory NERD from the G-PRIDE study who responded to RKT. Briefly, 242 patients with PPI-refractory NERD were enrolled in this study from April 2011 to July 2012. Patients with PPI-refractory NERD were defined as those without endoscopic mucosal breaks and with GERD symptoms (FSSG score ≥8) despite a prior therapy with a standard PPI dose (RPZ: 10 mg/day, omeprazole: 20 mg/day, or lansoprazole: 30 mg/day) for ≥4 weeks. PPI-refractory NERD met the following selection criteria: (1) were >20 years of age; (2) had received standard-dose PPI therapy for ≥4 weeks before the start of this study for the treatment of NERD; (3) had an FSSG score ≥8 after standard-dose PPI therapy for ≥4 weeks; (4) planned to receive RPZ (10 mg/day) treatment for ≥8 weeks; and (5) provided written informed consent regarding study participation. Exclusion criteria were as follows: (1) esophageal mucosal erosion in endoscopy carried out within 6 months before the registration; (2) presence of serious complications (liver, kidney, heart, blood, or metabolic disorders); (3) having undergone resection of the upper digestive tract; (4) confirmed presence of a peptic ulcer (excluding ulcer scar) or malignant tumor of the upper digestive tract; (5) inflammatory bowel disease, irritable bowel syndrome (IBS), esophageal stenosis, or esophageal achalasia; (6) diagnosis of a GI motility disorder by the study investigator; (7) suspected organic hepatic/biliary/pancreatic disorders such as gallstone, hepatitis, and pancreatitis; (8) hemorrhage of the digestive tract, mechanical ileus, or perforation of the digestive tract; (9) taking drugs prohibited for concomitant use (such as anti-ulcer drugs except for rabeprazole, prokinetics, other kampo medicines except for RKT) during the observation period; (10) psychoneurosis; (11) receiving or scheduled to receive an agent that is being developed; (12) lactation, pregnancy, or planned pregnancy during the study or follow-up period; (13) intolerance to oral administration; (14) history of allergy for kampo medicine; and (15) considered ineligible to participate by the chief investigator. Patients were randomized to receive RPZ (10 mg once daily) + RKT (7.5 g/day 3 times) (RKT group) or RPZ (10 mg once daily) + placebo (7.5 g/day 3 times) (PL group) for 8 weeks according to a computer-generated randomization list provided by a statistician from the site management organization (SMO) (Sogo Rinsho Holdings Co., Ltd, Tokyo, Japan). After written informed consent was obtained from the study participants, the patients with PPI-refractory NERD who met the inclusion criteria and did not meet the exclusion criteria were recruited for this study. Patients were randomly assigned to the RKT group [RPZ (10 mg/day) + RKT (7.5 g/t.i.d.) for 8 weeks] or the PL group (RPZ + placebo for 8 weeks). RKT (Tsumura & Co., Tokyo, Japan) was used in the form of a powdered extract obtained by spray drying a hot water extract mixture of the following 8 crude herbs: Atractylodis lanceae rhizoma (4.0 g), Ginseng radix (4.0 g), Pinellia tuber (4.0 g), Hoelen (4.0 g), Zizyphi fructus (2.0 g), Aurantii nobilis pericarpium (2.0 g), Glycyrrhizae radix (1.0 g), and Zingiberis rhizoma (0.5 g). The fingerprint pattern provided by 3-dimensional high-performance liquid chromatography revealed that RKT contains several low molecular compounds (i.e., hesperidin, liquiritin, liquiritigenin, isoliquilitin, isoliquiritigenin, formononetin, glycycoumarin, glycyrrhizin, atractylodin, atractylodinol, 6-shogaol, and 6-gingerol) [17]. Before and after the 4-week and 8-week treatments, GERD symptoms were evaluated using the FSSG questionnaire and GI-related QOL were evaluated using the Gastrointestinal Symptom Rating Scale (GSRS) questionnaire, similar to what was used in earlier clinical reports [18‐20]. The study was performed in accordance with the ethical guidelines for clinical studies and considered the patients’ rights and privacy. The study protocol was approved by the institutional review board of each institution (Additional file 1).

The FSSG questionnaire comprised 12 items: 2 domains, the reflux symptom (RS) domain, in which the sums of the respective scores of items 1, 4, 6, 7, 9, 10, and 12 are calculated; and the acid-related dysmotility symptom (ARD) domain, in which the sums of the respective scores of items 2, 3, 5, 8, and 11 are calculated (Table 1). The scores were calculated according to the frequency of the symptoms as follows: never, 0; occasionally, 1; sometimes, 2; often, 3; and always, 4 as previously reported. The total score is the sum of the RS and ARD scores, and total scores ≥8 indicated probable GERD as previously validated [19].

The GSRS questionnaire is an inquiry table consisting of 15 items for the evaluation of general GI symptoms [20]. Each GSRS item is rated on a 7-point Likert scale from no discomfort to very severe discomfort. Based on a factor analysis, the 15 GSRS items break down into the following 5 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).

Subgroup analysis was pre-planned to perform with respect to each subject’s background factors such as age (≥65 years or ≤64 years), gender (male or female), body mass index (BMI ≥22 or <22), classification as ARD and RS scores of the FSSG. The patients aged ≥65 years with ARD symptoms were selected from the patients with PPI-refractory NERD. A patient who had one or more ARD symptom score of FSSG was defined as a patient with ARD symptoms. Clinical characteristics of elderly patients with PPI-refractory NERD who responded to RKT were assessed by improvement degrees of 12 subscale scores of FSSG and 15 GSRS items after the treatment.

Improvement degree was calculated based on the FSSG or GSRS score before and after treatment, using the following mentioned formula. To compare the effects between the 2 groups, the mean improvement degree was used: (FSSG and GSRS) improvement degree (Δ) = [pre scores] − [post scores]. In addition, rikkunshito responder was defined as those whom the FSSG score has improved 50% or more after the 8-week treatment of RKT, and the differences in clinical characteristics between the RKT responsive group and nonresponsive group were investigated.

The efficacy analysis was based on the full analysis set (FAS) population. The treatment response in each group was evaluated based on changes in the FSSG and GSRS questionnaire scores before and after treatment, using Wilcoxon’s signed rank test. We employed the t-test to compare background factors such as age and BMI. The distributions of gender, current alcohol use, current smoking status, H. pylori infection, gastric mucosal atrophy, gastric mucosa redness, impaired gastroesophageal flap valve (GEFV), and esophageal hiatal hernia (grades B and A) were compared using Fisher’s exact test. Values of P < 0.05 were considered statistically significant. All data are expressed as mean ± standard deviation (S.D.).

Two hundred forty-two patients were randomly assigned to the RKT group (n = 125) or the PL group (n = 117). Twenty-five patients were excluded from the efficacy assessment because they withdrew from the study after registration (RKT group: n = 109, PL group: n = 108). Of the 217 patients, 95 were elderly patients (≥65 years) with ARD (RKT group: n = 52; PL group: n = 43) (Figure 1).

There were no marked differences in baseline demographic or clinical characteristics in the 2 groups except for rate of current smoking. Furthermore, there was no difference in total FSSG or overall GSRS scores before the start of treatment between the 2 groups (Table 2).

Changes in total, ARD, and RS scores of FSSG after treatments in each group are shown in Figure 2A. In both groups, total, ARD, and RS scores of FSSG were significantly decreased after the 4-week treatment compared with that before treatment (P < 0.01). Total, ARD, and RS scores of FSSG in the RKT group but not in the PL group further decreased during 4–8 weeks after treatment. Improvement degrees of total, ARD, and RS scores of FSSG after treatments in each group are shown in Figure 2B. The improvement degrees of the total and ARD scores but not RS score of FSSG were significantly higher in the RKT group than in the PL group after the 8-week treatment.

Out of 5 subscale scores of ARD, improvement degrees of FSSG-02 (abdominal bloating), FSSG-03 (heavy feeling in stomach after meals), and FSSG-05 (sick feeling after meals) scores were significantly higher in the RKT group than in the PL group (Figure 3A). Out of 7 subscale scores of RS, improvement degree of FSSG-06 (heartburn after meals) score was significantly higher in the RKT group than in the PL group (Figure 3B). Improvement degree of FSSG-12 (heartburn when bending over) score tended to be higher in the RKT group than in the PL group (P = 0.054, Figure 3B).

Improvement degrees of GSRS scores after the 8-week treatment in each group are presented in Table 3. Improvement degrees of reflex syndrome score of GSRS tended to be higher in the RKT group than in the PL group (P = 0.064). However, there was no significant difference in the degree of improvement of the overall and 5 subscale scores of GSRS between the 2 groups after the 8-week treatment. Of the 15 GSRS items, improvement degrees of GSRS-03 (acid regurgitation), GSRS-07 (abdominal distension), and GSRS-10 (constipation) scores were significantly higher in the RKT group than in the PL group.

Background factors of RKT responsive group and RKT nonresponsive group were shown in Table 4. Presence of concomitant systemic diseases (hypertension, insomnia, hyperlipidemia and constipation) was the potential predictor of poor response to RKT.