Sodium Valproate-Associated Thrombocytopenia in Antiepileptic Drug Polypharmacy: A Case Report and Review of the Literature

Introduction

Psychiatric disorders, although relatively uncommon in patients with epilepsy (prevalence: 5.6%), can substantially affect medication adherence and contribute to treatment complications [1]. Current guidelines emphasize the importance of balancing the effectiveness and safety of antiepileptic drugs (AEDs) when selecting therapy, particularly with respect to hematologic adverse reactions such as drug-induced thrombocytopenia (DITP) [2,3]. Sodium valproate (VPA) and levetiracetam (LEV) are widely prescribed AEDs. In clinical practice, LEV is often used as adjunctive therapy in patients who do not respond adequately to VPA monotherapy. Combination therapy demonstrates significantly greater clinical efficacy compared with VPA alone [4–6]. However, VPA is frequently associated with dose-dependent bone marrow suppression and thrombocytopenia [7–9]. Reports of VPA-induced thrombocytopenia have accumulated since the introduction of this drug, and Table 1 summarizes key published studies.

DITP is a hematologic disorder characterized by a substantial reduction in peripheral blood platelet (PLT) count (<100×109/L) after administration of specific medications. It increases the risk of bleeding complications, including petechiae, mucosal hemorrhage, and potentially fatal intracranial hemorrhage. Although rare, several classes of drugs have been implicated in DITP, including anticoagulants (e.g., heparin), antibiotics (e.g., sulfonamides, vancomycin), nonsteroidal anti-inflammatory drugs (e.g., aspirin, ibuprofen), antiepileptics (e.g., valproate, carbamazepine), and diuretics (e.g., hydrochlorothiazide) [10–12]. The pathogenesis of DITP is complex and may involve drug-induced inhibition of PLT production and/or immune-mediated PLT destruction [13,14]. Early recognition and discontinuation of the causative agent are essential because delayed intervention may result in severe morbidity or mortality. This clinical challenge underscores the importance of pharmacovigilance and mechanistic insights to mitigate VPA-associated hematologic risks.

Case Report

A 31-year-old man with a history of long-term drug-resistant epilepsy presented with psychiatric abnormalities and required hospitalization. The patient had previously received monotherapy with carbamazepine, sodium valproate, and levetiracetam, but he had not achieved adequate seizure control. Notably, prior VPA or LEV monotherapy had not produced clinically significant adverse drug effects. On admission, diagnoses were epileptic psychosis and mild intellectual disability. Psychiatric symptoms were assessed using the Positive and Negative Symptom Scale (PANSS), which yielded a total score of 70, consistent with moderate psychopathological severity. Initial laboratory evaluation demonstrated pancytopenia: red blood cell (RBC) count was 3.36×1012/L (reference range: 4.3–5.8×1012/L), hemoglobin (Hb) level was 112 g/L (reference range: 120–160 g/L), and PLT count was 104×109/L (reference range: 100–350×109/L).

The patient was initially treated with clonazepam 0.5 mg 3 times daily and aripiprazole 2.5 mg once daily for psychiatric symptoms, together with VPA 500 mg 3 times daily for seizure control. Due to inadequate response, the dosages were adjusted as follows: clonazepam was increased to 1 mg 3 times daily and aripiprazole to 5 mg once daily; LEV 125 mg once daily was introduced as adjunctive therapy. Despite these changes, the patient experienced a seizure on day 8, which prompted an increase in LEV dosage to 250 mg twice daily. Routine blood testing on day 14 revealed thrombocytopenia (PLT count 80×109/L), although the patient exhibited no clinical signs of bleeding, such as subcutaneous ecchymosis. A comprehensive differential diagnosis was performed to exclude other potential causes of thrombocytopenia. Laboratory testing ruled out vitamin B12 and folate deficiency, given that serum levels were within normal ranges. No evidence of active infection was identified; no fever, leukocytosis, or positive blood cultures were observed. Autoimmune disorders were also considered unlikely due to the absence of relevant symptoms. Other medications, including antipsychotics and benzodiazepines, were reviewed, but these are not commonly associated with thrombocytopenia. The attending physician initially suspected DITP secondary to LEV and therefore discontinued the drug; leucogen tablets were prescribed to elevate PLT counts. Clinical monitoring demonstrated gradual recovery of RBC and Hb levels, but a persistent decline in PLT count, which reached a nadir of 47×109/L on day 18 (Figure 1). In accordance with the clinical pharmacist’s recommendation, iron dextran was administered to address concurrent anemia, and VPA was discontinued due to its potential hematologic adverse effects. The PLT count subsequently increased to 93×109/L after 3 days. Both PLT count and Hb level eventually returned to their normal ranges. During the treatment period, the patient continued oral clonazepam and aripiprazole. At discharge, hallucinations and delusional symptoms had resolved (PANSS score <60), and no further seizures were reported. The normalization of hematologic parameters after discontinuation of VPA supports the hypothesis that VPA was the primary contributor to thrombocytopenia. Furthermore, neither VPA nor LEV monotherapy had previously produced this condition; combination therapy induced progressive thrombocytopenia in this patient, suggesting that LEV can potentiate VPA-induced thrombocytopenia.

Discussion

This case highlights the complex hematologic risks associated with VPA therapy in patients exhibiting pharmacoresistant epilepsy. Our patient’s thrombocytopenia was initially attributed to LEV. However, resolution occurred only after VPA was discontinued, indicating that VPA was the more likely cause. Further analysis of the association between the adverse drug reaction and the medications was conducted using the Naranjo Assessment Scale. The score for LEV was 1 (Table 2), whereas the score for VPA was 6 (Table 3), supporting a stronger association between DITP and VPA.

VPA is widely recognized for its efficacy in seizure control; however, its hematologic toxicity, ranging from myelosuppression to immune-mediated PLT destruction, remains a critical concern, particularly in patients receiving polypharmacy with AEDs or those with a history of hypersensitivity reactions. The hematologic adverse effects of VPA are consistent with its known mechanisms of toxicity. As a histone deacetylase inhibitor, VPA may impair hematopoiesis by altering chromatin remodeling in hematopoietic stem cells, potentially exacerbating myelosuppression in susceptible individuals [14,15]. Hematologic toxicity is relatively common and typically occurs when serum concentrations exceed 100 μg/mL. The onset and severity of symptoms may vary and manifest as recurrent, transient, or persistent. In most cases, although thrombocytopenia is clinically significant, symptoms can be ameliorated by dose reduction, and complete discontinuation is rarely necessary [16].

In our case, DITP was initially attributed to LEV because the patient had been receiving a stable dose of VPA (500 mg 3 times daily) for greater than 6 months before admission. However, drug concentration monitoring on day 13 revealed a VPA serum level of 112.4 μg/mL (therapeutic range: 50–100 μg/mL), which slightly exceeded the maximum recommended concentration. After LEV discontinuation, the patient’s PLT count did not recover. The clinical pharmacist therefore concluded that the dose-dependent adverse effect of VPA was a contributing factor to the thrombocytopenia and recommended its discontinuation. Within 3 days of discontinuation, the PLT count increased to 93×109/L, confirming VPA as the causative agent. Notably, the patient had not exhibited thrombocytopenia during VPA monotherapy. The hematologic abnormality emerged only when LEV was co-administered, suggesting that LEV potentiated VPA-associated thrombocytopenia.

VPA and LEV are metabolized through distinct pathways [17,18]. Theoretically, LEV is not expected to affect VPA metabolism by inhibiting or inducing metabolic enzymes. In the present case, the patient’s liver function test results remained stable throughout the observation period; alanine transaminase ranged from 21.1 to 22.5 U/L and aspartate aminotransferase ranged from 31.9 to 40.0 U/L, all within normal reference ranges. However, both VPA and LEV are excreted via the kidneys. In patients with renal insufficiency, concurrent use of these drugs may lead to the accumulation of VPA and its metabolites due to competitive inhibition within the excretion pathway. Our patient’s serum creatinine level was 100.4 μmol/L during concomitant administration and decreased to 77.4 μmol/L after discontinuation of the drugs. Additionally, VPA metabolism may be reduced in individuals with variable liver enzyme function [19,20]; concomitant LEV could indirectly influence its pharmacokinetics, potentially increasing VPA levels in certain patients. Therefore, the elevated serum VPA concentration and occurrence of thrombocytopenia in this case may have been multifactorial. Nevertheless, the findings suggest that LEV can increase the risk of VPA-induced thrombocytopenia, highlighting the need for vigilance regarding potential drug interactions between these 2 agents. It is important to note that the PLT count nadir in this case was delayed relative to the time of drug adjustment, emphasizing the importance of longitudinal monitoring because hematologic changes may lag behind clinical manifestations. The primary challenge when diagnosing thrombocytopenia in such patients lies in distinguishing VPA-induced bone marrow toxicity from other etiologies, including LEV-associated DITP or nutritional deficiencies. For severe and persistent cytopenias, treatment discontinuation is recommended [21], although clinical decisions must balance seizure control against hematologic risk. After VPA cessation, adjunctive use of LEV, a non-enzyme-inducing AED with a generally favorable hematologic profile, may provide a safer alternative for seizure management.

The management of drug-induced hematotoxicity requires a comprehensive and multidisciplinary approach, particularly in complex clinical settings where causality is not immediately evident. In the present case, integration of expertise from clinical pharmacology, hematology, and psychiatry was essential for accurate diagnosis and timely intervention. Multidisciplinary consultation helps to distinguish drug-induced hematotoxicity from other potential causes; it also facilitates individualized treatment strategies that balance therapeutic efficacy with patient safety. Our case highlights the importance of close collaboration among specialists, especially in the treatment of patients receiving polypharmacy or those with atypical clinical presentations. Future studies should investigate standardized protocols for multidisciplinary management of drug-induced hematotoxicity to improve diagnostic accuracy and clinical outcomes.

Conclusions

We reported a case of thrombocytopenia in a patient with an epileptic mental disorder who was receiving combination therapy with VPA and levetiracetam. This report highlights the need for pharmacovigilance in the management of epilepsy. Future studies should explore biomarkers that predict VPA toxicity to improve risk stratification in complex clinical settings.