A Polymorphism Cluster at the 2q12 locus May Predict Response to Piromelatine in Patients with Mild Alzheimer's Disease.

Background: Piromelatine is a novel melatonin MT1/2/3 and serotonin 5-HT-1A/1D receptors agonist developed for mild Alzheimer's disease (AD). In a randomized, placebo-controlled, dose-ranging study (ReCognition) of piromelatine (5, 20, and 50 mg daily for 6 months) in participants with mild dementia due to AD (n=371, age 60-85 years), no statistically significant differences were found between the piromelatine and placebo-treated groups on the primary (i.e., computerized neuropsychological test battery (cNTB)) and secondary outcomes (ADCS-CGIC, ADCS-MCI-ADL, ADAS-cog14, NPI, and Pittsburgh Sleep Quality Index (PSQI)) nor were there safety concerns (https://clinicaltrials.gov/ct2/show/NCT02615002).

Objectives: This study was aimed at identifying genetic markers predicting piromelatine treatment response using a genome-wide association approach (GWAS).

Design: Variant genotyping of a combined whole genome and whole exome sequencing was performed using DNA extracted from lymphocytes from consenting participants. The general case-control allelic test was performed on piromelatine-treated participants, taking "responders" (i.e., >0.125 change from baseline in the cNTB) as cases and "non responders" as controls, using a Cochran-Armitage trend test.

Setting: 58 outpatient clinics in the US.

Participants: 371 participants were randomized in the trial; 107 provided informed consent for genotyping.

Results: The GWAS sample did not differ from the full study cohort in demographics, baseline characteristics, or response to piromelatine. Six single-nucleotide polymorphisms (SNPs) in chromosome 2q12 (2:107,510,000-107,540,000) were associated with response (p-value < 1x10 -4 each). Post hoc analyses suggested that the carriers of the 2q12 polymorphism cluster (27% of the GWAS sample) improved significantly on the cNTB on piromelatine as compared to placebo but significantly worsened on the ADAS-Cog14 and PSQI. By contrast, "non-carriers" improved significantly with piromelatine compared to placebo on the ADAS-Cog14 ( 2.91 (N=23) with piromelatine 20 mg vs 1.65 (N=19) with placebo (p=0.03)) and PSQI.

Conclusions: The 2q12 (2:107,510,000-107,540,000) 5-6 SNPs cluster may predict efficacy of piromelatine for mild AD. These findings warrant further investigation in a larger, prospective early-stage AD clinical trial for patients who are non-carriers of the 2q12 polymorphism cluster.