The percentages of patients with PU history (26.1% vs 15.1%, P = 0.004), chronic renal failure (9% vs 1.7%, P < 0.001), or taking non-steroidal anti-inflammatory drugs (NSAIDs) (14.4% vs 3.1%, P < 0.001) in the ulcer group were significantly higher than those in the control group. The percentages of patients Selleckchem Doxorubicin taking cotreatment of anti-acid (histamine H2-receptor antagonists or proton-pump inhibitors) (32.4% vs 70.3%, P < 0.001), ARBs or angiotensin-converting

enzyme inhibitors (ACEIs) (45% vs 58.1%, P = 0.01), or statins (41.4% vs 53.7%, P = 0.02) in the ulcer group were significantly lower than in the control group (Table 1). Similar significant results were observed in the bleeding group, and the same factors for ulcer were significantly different between the bleeding group and the controls. In addition,

the percentage of patients taking β(α)-blocker (17.8% vs 35.5%, P = 0.02) cotreatment was significantly lower than in the control group (Table 1). The candidate 29 SNPs of 24 genes associated with small bowel or ulcer bleeding identified by genome-wide preliminary analysis were evaluated in 593 patients; however, only the CHST2 2082 SNP was significantly associated with ulcer and ulcer bleeding (Table 2). The allele frequencies of SLCO1B1 and CHST2 2082 SNP and the haplotype frequencies of SLCO1B1 are shown in Table 2. The allele frequencies of the polymorphisms did not deviate significantly from those expected under Hardy–Weinberg equilibrium. The frequency of the CHST2 2082 T allele was significantly higher in both the ulcer group (60% vs 46.4%, P = 0.01) and the bleeding group (70.7% vs 46.4%, P = 0.003) compared Sorafenib supplier to the controls (Table 2). The haplotype frequencies of SLCO1B1*1a (388A and 521T, wild type), *1b (A388G and 521T), *5 (388A

and T521C), and *15 (A388G and 521C) in the controls were 10.6, 62.8, 0, and 26.6%, respectively (Table 2). The frequency of the SLCO1B1*1b haplotype was highest and significantly higher in the ulcer group (74.3% vs 62.8%, P = 0.02) compared to the controls (Table 2). Among the patients taking stain, ARB, or ACEI, the frequencies of the SLCO1B1*1b haplotype were significantly higher not only in the ulcer group (77.9% vs 63.1%, P = 0.02) but also in the bleeding group (87.1% vs 63.1%, P = 0.006) N-acetylglucosamine-1-phosphate transferase compared to the controls (Table 3). History of PU (adjusted OR 2.52, 95% CI 1.39–4.55), chronic renal failure (7.63, 2.34–24.9), cotreatment with NSAIDs (6.62, 2.63–16.7), anti-acid (0.18, 0.11–0.31), and SLCO1B1*1b (2.20, 1.24–3.89) were significantly associated with ulcer after adjustment of significant factors in the univariate analysis (Table 4). Age > 80 years (adjusted OR 3.60, 95% CI 1.51–8.59), PU history (4.20, 1.71–10.3), chronic renal failure (6.42, 1.29–32.0), cotreatment with NSAIDs (8.57, 2.20–33.4), anti-acid (0.05, 0.02–0.15), β(α)-blockers (0.33, 0.12–0.90), and CHST2 2082 T allele (2.57, 1.07–6.