Incidence and Predictors of Antiepileptic Drug–Induced Thrombocytopenia: A Prospective Cohort Study from a Tertiary Care Hospital in Lahore, Pakistan

Abstract

Background: Antiepileptic drugs (AEDs) are the foundation of epilepsy treatment, but hematological toxicity including thrombocytopenia is a frequent, and sometimes unrecognized, side effect, especially of valproic acid. The incidence and independent risk factors for this harmful effect in the south Asian tertiary care setting are still limited.

Objective: To examine the cumulative incidence and risk factors for AED-induced thrombocytopenia in treatment naive epileptic patients and primary hypothesis: To test the hypothesis that the use of valproic acid was independently associated with thrombocytopenia.

Methods: The prospective cohort study was carried out in Neurology and Haematology Department of a tertiary care hospital in Lahore, Pakistan for one year. A minimum of 195 participants was calculated to be needed using Cochran's formula and 196 adults with no prior use of AED who started on monotherapy (valproic acid, levetiracetam, carbamazepine or phenobarbital) were recruited for a study period of 24 weeks, with serial platelet counts. Independent predictors were determined by multivariable logistic regression, Cox proportional hazards regression and Kaplan-Meier survival analysis (p<0.05).

Results: The cumulative incidence of thrombocytopenia was 26.0% (51/196), and was highest with valproic acid (41.3%) compared with other AEDs combined (16.5%; relative risk 2.50, 95% CI 1.47–4.27, p<0.001). Mean platelet count declined significantly from 279.4±61.2×10⁹/L at baseline to 224.8±78.5×10⁹/L at week 24 (paired t=9.42, p<0.001). The use of valproic acid (adjusted OR 2.85, 95% CI 1.52–5.34, p=0.001), higher dosage, longer duration of therapy, older age, polytherapy and lower baseline platelet count were all independent predictors of thrombocytopenia on multivariable logistic regression. Valproic acid had a significantly shorter time to onset (6 vs 12 weeks, log rank p=0.002).

Conclusions: AED-induced thrombocytopenia is frequent, especially when using valproic acid, and can be predicted from easily available clinical factors. To prevent bleeding problems, regular platelet counts should be performed and may be necessary during the initial 8–12 weeks of valproic acid treatment.