0035 Characterization of Metabolites in a Randomized, Controlled Trial of Clarithromycin for Pathological Sleepiness

Abstract Introduction Pathologic sleepiness in CDH is associated with the presence of an endogenous, cerebrospinal fluid peptide that can positively, allosterically modulate GABA-A receptors. The macrolide antibiotic clarithromycin has been shown to benefit patients with hypersomnolence that is refractory to standard of care agents and is a negative allosteric modulator of GABA-A receptors. One explanation of its wake-promoting effects is the alteration of gastrointestinal microbiota composition. Here we characterize the metabolomics of serum and stool samples collected as part of a randomized, controlled trial of clarithromycin. Methods Ten patients and 10 Healthy controls had serum and stool samples taken on baseline study day and after a 2 week trial of clarithromycin dosed at 500mg bid, A total of 60 serum and stool samples were analyzed using Mass Spectrometry with a metabolomics panel. A total of 214 metabolites were detected, 122 in serum and 148 in stool with 56 overlapping metabolites. Results Samples were normalized by QC samples, log transformed and scaled separately for serum and stool. When comparing the 3 groups at the metabolite level, univariate analyses revealed 7 in serum and 4 in stool that had different levels between at nominal significance (p< 0.05). The 4 in stool were 2-Deoxyadenosine, 3-Hydroxybenzaldehyde, Imidazole acetic acid and N-Acetyl aspartate. The first 2 were driven by lower values in the Treatment group vs Healthy Controls, and the latter 2 were the opposite. The 7 in serum were (R)-2,3-Dihydroxy-isovalerate, Homoserine, L-Allothreonine, L-Threonine, O-Acetyl-L-Serine, Palmitoyl carnitine, Riboflavin. Paired log2 fold-change analysis in serum between treatment and baseline revealed nominally significant decrease in the expression of Theobromine (P=0.4). In stool, the analysis revealed changes in 3-Hydroxybutyric acid, 2-Deoxyadenosine,3-2-Hydroxyethylindole,3-Hydroxy-2-Methylpyridine, Glycerol 2-phosphate, N-Acetylaspartate, Nicotinic Acid, Mononucleotide, Riboflavin, and Theobromine. PCA loadings of serum and stool samples, overall and in each group, revealed the first 2 PC’s consisting of 3-Hydroxybutyric acid and (R)-2,3-Dihydroxy-isovalerate Conclusion Despite similar profiles in serum and stool in patients, before and after treatment and healthy controls; subtle differences in particular metabolites may need to be further studied. Alternative pathways not currently captured in metabolomics in this study may be better suited to capture evidence of alternative mechanisms of clarithromycin efficacy. Support (if any) NS111280