Abstract
OBJECTIVE: Lithium augmentation (LA) is an effective treatment for patients with major depression after inadequate antidepressant response, but therapeutic outcomes vary considerably between individuals. Molecular studies could yield novel insights into treatment prediction to enable personalized therapy choices. Here, we investigated the effects of polygenic risk scores (PRS) for schizophrenia (SCZ), major depressive disorder (MDD), and bipolar disorder (BIP) on clinical outcomes following LA. METHODS: Recent GWAS summary statistics were used to construct disorder-specific PRS in lithium-augmented MDD patients who participated in a prospective study after poor response to at least one antidepressant drug. Depressive symptoms were assessed for four weeks or longer using the Hamilton Depression Rating Scale (HAMD). Hazard ratios (HR) of favorable outcomes, response (≥ 50% reduction in HAMD composite scores) and remission (HAMD ≤ 7), were estimated by Cox proportional hazards regression models adjusted for ancestry, demographic, and clinical covariates. RESULTS: In 193 patients, BIP-PRS was positively associated with both response (HR = 1.29, 95% CI = 1.02-1.63, p = 0.03, Nagelkerke R(2) = 2.51%) and remission (HR = 1.52, 95% CI = 1.14-2.04, p = 0.004, Nagelkerke R(2) = 4.53%) after LA. Our data further suggest that individuals who carry a lower polygenic burden for MDD tend to respond better to LA (HR = 0.81, 95% CI = 0.66-1.00, p = 0.048, Nagelkerke R(2) = 1.99%). No associations were observed between SCZ-PRS and either clinical outcome (p > 0.05). CONCLUSIONS: Our findings indicate that individuals at higher polygenic risk for BIP and lower polygenic risk for MDD are more likely to benefit from augmentation with lithium. If replicated, PRS may inform future efforts to establish clinical prediction models for LA outcomes in unipolar depression.