402. Integrating clinical and genomic determinants of valproate-associated side effects in a Taiwanese population

Abstract

Abstract Background Valproic acid (VPA) is widely prescribed for migraine, seizure disorders, and bipolar disorder; however, adverse effects such as hyperammonemia, liver toxicity, and thrombocytopenia may limit its clinical use. Aims & Objectives To identify clinical and genetic factors associated with VPA-induced adverse effects in order to improve patient safety and support personalized therapeutic decision-making. Method Hospitalized adults at Kaohsiung Veterans General Hospital (KSVGH), Taiwan, who received VPA with serum drug level monitoring and had available DNA samples from the KSVGH Biobank were enrolled. Genotyping was performed using the Taiwan Biobank C2-58 Axiom Genome-Wide TWB 2.0 Array. Multivariable logistic regression was used to identify clinical factors associated with hyperammonemia, liver toxicity, and thrombocytopenia. Genome-wide association studies (GWAS) and MAGMA gene-based analyses were conducted to evaluate genetic associations with serum ammonia levels, liver enzymes, and platelet counts. Results Among 151 patients, hyperammonemia occurred in 18 (11.9%), liver toxicity in 7 (4.6%), and thrombocytopenia in 44 (29.1%). Hyperammonemia was independently associated with stroke history (adjusted odds ratio [aOR] 8.1, 95% CI 2.5–26.8), concomitant topiramate use (aOR 24.1, 95% CI 3.3–177.8), and higher baseline alanine aminotransferase levels (aOR 1.0 per 1 U/L increase). The prediction model showed good discrimination (AUC = 0.79). Psychiatric disorders were associated with liver toxicity, while VPA indication, lacosamide use, and pre-existing liver disease were associated with thrombocytopenia. GWAS identified candidate loci associated with serum ammonia levels, including NCKAP5, AHMO, and HECW1, but no significant genetic associations were observed for liver toxicity or thrombocytopenia. Discussion & Conclusions VPA-induced hyperammonemia appears to be influenced by both clinical factors and genetic susceptibility, whereas liver toxicity and thrombocytopenia are primarily driven by clinical and treatment-related factors. Hyperammonemia was strongly associated with stroke history, topiramate use, and impaired baseline liver function, with persistently elevated ammonia levels during follow-up. Prior evidence suggests that topiramate may contribute to hyperammonemia through indirect interference with carbamoyl phosphate synthetase 1 in the urea cycle, while hepatic dysfunction may further impair ammonia clearance. Although candidate genetic loci associated with ammonia levels were identified, larger studies are needed to clarify genetic contributions to other VPA-related adverse effects.