Introduction

Aspirin had long been regarded as safe beyond the duodenum because of its rapid absorption and lack of an enterohepatic recirculation [1]. However, in recent studies, small bowel injury and enteropathy associated with low dose aspirin (LDA) are increasingly being recognized [2-5]. In a recent Japanese retrospective cohort study of patients taking LDA, the incidence of overt GI bleeding and suspected of small bowel blood loss was 3.9% (upper GI 3.1%, colon 0.7%, and small bowel 0%) and 1.4%, respectively [6]. In another retrospective study of patients undergoing LDA therapy following percutaneous coronary intervention (PCI), the incidence of GI bleeding within 30 days after PCI was 4.3% including occult bleeding (1.9%), small bowl ulcer (0.5%), and overt not upper GI bleeding(0.5%) indicating that attention must be paid to small bowel and lower GI bleeding as well as upper GI bleeding [7].

LDA is now widely used for primary and secondary prevention of cardiovascular events [8-10]. The thienopyridine derivatives, ticlopidine or clopidogrel are also regarded as mandatory following PCI at least until the coronary stents are fully endothelialized which requires approximately 3 months for bare metal stents and up to 1 year for drug-eluting stents [11]. One concern regarding prolonged antiplatelet or anticoagulant (antithrombotic) therapy is the risk of gastrointestinal (GI) bleeding including bleeding from the small intestine [10-13]. However, the pathogenesis and risk factors for small bowel damage induced by aspirin are not also well understood making prevention difficult.

In this era of personalized medicine, the technology allows one to identify genetic risk factors in relation to side effects of medical therapy. To date, there are few studies of the association between genetic polymorphisms and the risks of aspirin-induced ulcer or its complications[14-17]. Individuals with two single nucleotide polymorphisms (SNPs) of cyclooxygenase-1 (COX-1), A-842G and C50T, exhibit increased sensitivity to aspirin and lower prostaglandin synthesis capacity but the polymorphism lacked statistical significance in relation to an association with bleeding peptic ulcer [18]. Our recent Japanese study suggested that the SLCO1B1 521TT genotype and the SLCO1B1 *1b haplotype were significantly associated with the risk of peptic ulcer and ulcer bleeding in patients taking low dose enteric-coated aspirin [17]. However, data on the related factors with small intestinal or lower GI events among the patients taking LDA are lacking. Particularly, to our knowledge the association between genetic markers and risk of small bowel bleeding induced by NSAIDs including LDA are still lacking. Therefore, the aim of the present study was to identify the genetic markers related with small bowel bleeding among the patients taking LDA.

Materials and Methods

Study subjects consisted of patients taking 100 mg of enteric coated aspirin (Bayer Health Care, Osaka, Japan) and suspected of bleeding from the small intestine and controls. All patients with at least a one year history of aspirin and or at least a 3 month history of anti-thrombotics use were entered. The study was conducted in accordance with the Declaration of Helsinki from 2007 January to 2012 April at the hospital of Kawasaki Medical School and Sakakibara Heart Institute of Okayama, Japan. The permission was granted by the Ethics Committees of both hospitals, and written informed consent was obtained from each patient.

Results

Discussion

Our study attempted to identify genetic risk factors associated with small bowel bleeding. Although 27 candidate SNPs of 23 genes were identified by DMET™ Plus GeneChip array, in our validation study, LDA associated small bowel bleeding occurred more often in the patients carrying the GG homo-genotypes of CYP4F11 (rs1060463) or CYP2D6 (rs28360521) , the patients carrying T allele of CYP24A1(rs4809957), or G allele of GSTP1 (rs1695). The role of the other SNPs must await additional studies with larger numbers of samples than available from our array study.

Human CYP4F11 is one of the orphan cytochrome P450 with limited information regarding heterologous expression and functional characterization. It has been reported to be located in a cluster with the other five members of the Cytochrome P450 4F (CYP4F) subfamily on chromosome 19p13.1-2; its mRNA is expressed mainly in human liver [19]. Cytochrome P450 4F (CYP4F) enzymes are involved in cellular protection and metabolism of numerous small molecules, including drugs, toxins, and eicosanoids [20]. CYP4F11 appears to be involved in arachidonic acid and fatty-acid metabolism and catalyzes ω-hydroxylation of leukotriene B4, lipoxins A4 and B4, and hydroxyeicosatetraenoic acids (HETEs) (ie, metabolites with roles in many biological processes including platelet aggregation). Although CYP4F11 activities are much lower than those of CYP4F3 [21], CYP4F11 may have a role in platelet aggregation which would provide a theoretically basis for it being related to small bowel bleeding induced by LDA. However, its function is still not well characterized and there was no study investigating the biologic basis of CYP4F11 SNPs in this study. The SNP of CYP4F11 (rs1060463) was the most significantly associated with small bowel bleeding among the identified SNPs. However, in multivariate analysis it lost significance as well as taking NSAIDs which is well known risk factor for small bowel injury and bleeding. There may be some interactional association between NSAIDs use and the CYP4F11 SNP.

CYP2D6 plays a critical role in drug metabolism and synthesis of cholesterol, steroids and other lipids. This protein is localized to the endoplasmic reticulum and is known to metabolize as many as 20% of clinically prescribed drugs such as psychotropics, antihistamine, beta-blockers etc., but not aspirin or NSAIDs [22]. The CYP2D6 gene is polymorphic and almost 80 variant alleles have been reported. CYP2D6*3, CYP2D6*4, and CYP2D6*5 are the major alleles involved in the poor metabolizers (PM) that have little enzymatic activity, while CYP2D6*10 is a variant allele of the intermediate metabolizer (IM) phenotype. The frequency of PMs is less than 1% and IM phenotype is markedly dominant in Asia including the Japanese population. In the present study, the SNP associated with small bowel bleeding was CYP2D6 rs28360521, which is found -2182 5’-UTR of the CYP2D6 gene. This SNP is in linkage disequilibrium with the entire gene (as well as 2 other more distant genes: NDUFA6 and NAGA) [23,24]. The rs28360521 may be considered as potentially etiologic, although it is unknown how it might play a role in small bowel bleeding induced by LDA.

We first reported the possible association of CYP4F11 and CYP2D6 SNPs with LDA induced small bowel bleeding, and there is no report indicating the association of the SNPs with not only small bowel bleeding, but also upper GI bleeding. Moreover, we failed to confirm the significant association of the previously identified SNPs related with LDA induced peptic ulcer with small bowel bleeding (Data is not shown). The pathogenesis of small intestinal damage induced by NSAIDs including aspirin is not well understood. Based on the animal studies and in vitro studies, the same mechanisms in the stomach such as topical irritant properties of NSAIDS inducing direct mucosal toxicity, mitochondrial damage, breakdown of intercellular integrity has been assumed [25]. However, enterohepatic recirculation, which is prevented by ligation of bile ducts seems to be more important component of the mechanism underlying the pathogenesis of small intestinal damage than suppression of prostaglandin synthesis [26,27]. Moreover, enteric bacteria is thought to cause and exacerbate the intestinal damage elevating intestinal permeability by NSAIDs, activating neutrophils, generating induction of inducible nitric oxide synthesis [28-30]. The further studies how the SNPs are involved in LDA induced enteropathy or bleeding are required.

In the present study, active smoking, non-cardiac vascular diseases including cerebrovascular diseases warfarin, and NSAIDs were significantly associated with small bowel bleeding among the patients taking LDA. Drug combinations involving antiplatelets and anticoagulants are known to lead to a greatly increased risk of GI bleeding, and the prescribing of aspirin with warfarin was reported to increase risk of GI bleeding than that observed with each drug alone [31]. There is no report indicating association of small bowel bleeding with smoking. Moreover, there is no evidence indicating that the prevalence of GI bleeding was more frequent in aspirin users with non-cardiac vascular diseases compared to those with cardiac disease, and the risk of GI bleeding seems to be similar according to the previous studies although there is no data comparing the risk of small bowel bleeding [32]. The significant results of underlying disease treated by LDA possibly were caused by selection bias. Wallace JL et al [30] recently reported PPIs significantly exacerbated NSAIDs-induced intestinal ulceration and bleeding in the rat. However, in the present study there was no association between PPI use and small bowel bleeding. The limitation of our study is case control and the small number of patients with small bowel bleeding, and the large scale clinical cohort studies are absolutely required to investigate the possibility of this important issue.

Although our data needs to be validated and extended in a larger cohort, this exploratory study suggests that CYP4F11 and CYP2D6 SNPs may identify patients at increased risk for aspirin-induced small bowel bleeding.

Acknowledgments

We thank Ms. Maki Nomura and Ms. Tomoko Yobimoto (Kawasaki Medical School, Okayama, Japan) for assistance of laboratory work.