Influence of CYP2C9 phenotypes on phenytoin plasma concentration in neurosurgical Brazilian patients.

Aims: To investigate the association of CYP2C9 metabolic phenotypes with phenytoin plasma concentration ([PTH]) in neurosurgical patients from the Brazilian Public Health System.

Methods: Patients (n = 170) were treated with phenytoin (300 mg/day) perioperatively as prophylaxis for postoperative seizures. Two to 10 days after surgery, a blood sample was collected for quantification of [PTH] and genotyping of CYP2C9*2 and *3 alleles. CYP2C9 metabolic phenotypes, NM (normal), IM (intermediate), and PM (poor) metabolizer, were inferred from CYP2C9 diplotypes. Linear regression modeling was applied to identify predictors of [PTH].

Results: Wide (22-fold) interindividual variation in [PTH] was observed (2.2-47.5 mg/l). [PTH] associated significantly (Kruskal-Wallis P < 0.005) with CYP2C9 phenotypes and there was a significant trend (Jonckheere-Terpstra test, P < 0.0001) for [PTH] increase in the order NM < IM < PM. [PTH] was within the target therapeutic range (10-20 mg/l) in 34.7% of patients, while 39.4% and 25.9% had [PTH] below and above the range, respectively. CYP2C9 phenotypes associated significantly (chi-square P = 0.004) with the distribution of patients in [PHT] therapeutic categories and the Cramér's V test pointed to moderate magnitude of the effect of CYP2C9 phenotypes (V = 0.211).

Conclusion: Diplotype-predicted CYP2C9 metabolic phenotypes are associated significantly with [PTH] in neurosurgical Brazilian patients receiving phenytoin for postsurgery seizure prophylaxis. [PHT] increased progressively in the phenotype order NM < IM < PM, and all PM patients had [PHT] above the target therapeutic range, consistent with the CPIC guideline 'strong' recommendation for phenytoin dosing adjustments in PMs.