OC-081 Genetic Polymorphisms and Patterns of FGF19 Response to Meals in Primary Bile Acid Diarrhoea

Introduction Primary bile acid diarrhoea (PBAD) is associated with low median serum fibroblast growth factor 19 (FGF19) levels. Our group has suggested that PBAD results from impaired negative feedback by FGF19 on the enzyme CYP7A1, leading to over production of bile acids, which spill over into the colon leading to diarrhoea. We aim to characterise abnormalities in FGF19 in PBAD, in particular the response of the hormone to meal stimuli and to describe phenotypic patterns of response. We have explored possible associations of genetic polymorphisms with both PBAD and specific phenotypes. Methods Subjects with PBAD were prospectively recruited. After an overnight fast, blood was sampled every 90 min for 6 h. Meals were provided at 9:00 and 12:00. Serum FGF19 was quantified by ELISA using a commercially available kit. Three phenotypes of meal stimulated FGF19 response are described: 1. Low-low (L-L) all five serum FGF19 levels <400 pg/ml. 2. Low-high (L-H) sample 1 400 pg/ml sample 4 or 5 >400 pg/ml. 3. High-high (H-H) samples 1 and 3 or 5 >300 pg/ml. In a separate part of the study genomic DNA was extracted from 91 subjects with PBAD and 120 diarrhoea control subjects. Eight SNPs were analysed, two from the FXR gene (rs61755050, rs56163822), two from the FGF19 gene (rs1789170, rs948992) as well as SNPs from the genes for OSTα, klothoβ, FGFR4 and ASBT (rs939885, rs17618244, rs376618, rs188096). Genotyping was performed using Taqman SNP genotyping assays. Results 18 subjects underwent serial FGF19 sampling. Seven were defined as L-L response, five as L-H and four as H-H. Two subjects did not match criteria for the subtypes. The seven L-L subjects had lower SeHCATs compared with L-H and H-H combined (p<0.05). Five subjects had SeHCATs of <5%, all but one were L-L. Overall, subjects with lower SeHCATs had lower FGF19 levels. Area under the curve was lower for those with SeHCAT <5% than those above (median 701 pg/ml h vs 1655 pg/ml h p=0.023). The FGF19 level at 1.5 h correlated with subjects' SeHCAT (r=0.49, p=0.04). The ASBT SNP was associated with L-L subjects, present in four out of seven (p<0.05). No other significant SNP associations were found within the phenotypes described nor between PBAD as a whole and diarrhoea controls. Conclusion Serial FGF19 sampling in PBAD identified three responses to meal stimulus. Most individuals with severe PBAD have L-L response of FGF19, and most of these have a polymorphism in the gene for ASBT. L-H and H-H phenotypes are seen in those with higher SeHCATs. The L-H pattern resembles that previously reported in healthy individuals. These patterns may represent a defect in the receptor component of the feedback mechanism for BA synthesis. Competing interests None declared.