Genome-wide association study of antidepressant induced treatment emergent suicidal ideation (TESI)

TESI induced by antidepressants in a major depression is a serious side effect and has led to a black-box warning by the FDA. Associations of genetic variants within GRIK2 and GRIA3 have been found in the STAR*D sample and could partly be replicated in our Munich Antidepressant Response Signature (MARS) project. To extend these findings we performed a genome-wide association (GWA) study. TESI was evaluated in the MARS project N=397 and defined by an increase of suicidal thoughts anytime during hospitalisation in patients without suicidal thoughts at hospital admission using the suicide item (item 3) of the Hamilton Depression Rating Scale. GWA was analysed by comparing the genotype frequencies of 405.383 SNPs in patients with TESI N=32 to patients without increase in suicidal ideation N=329 and to patients with no suicidal ideation through-out hospitalization N=79. Best results were replicated in a second independent German sample N=530 of depressed patients with N=45 TESI positive, N=148 completely lacking suicidal ideation and N=416 without increase in suicidal ideation. 100 SNPs were associated with TESI with a Fisher-Product Method p<0.001 (over allelic and genotypic tests) for the narrow as well as the broader non-TESI definition. In the replication sample 16 of the selected 100 SNPs achieved a p<0.05. This subset could predict TESI with >90% in both samples. We identified a subset of SNPs associated with TESI, which could in part be replicated in an independent sample.