Abstracts

RATIONALE: Urinary lithiasis was a rare side effect of anticonvulsants in children and young adults. Recently I experienced a few patients with epilepsy who suffered from urolithiasis and tried to find out its risk factors. Although alkaline urinary pH is known to be one of them, the relationship between anticonvulsants and urinary pH has never been investigated. To address this issue, a large number of urinary pH was studied from patients who were on anticonvulsants monotherapy.
METHODS: I reviewed the medical records of all epilepsy patients who were administrated anticonvulsants for more than six months at Department of Pediatrics, Otsu Red Cross Hospital, Japan from 1995 to 2001. All urine was collected when patients visited the hospital for the regular examination. The blood concentration of anticonvulsants was usually measured at the same time. Urinary pH was examined by the test tape containing methyl red and bromothymol blue. Urinary sediments were also inspected by the microscope. Pyuria, bacteriuria, hematuria,and proteinuria were excluded from this study. Urine from patients with anticonvulsants polytherapy were not included. Total 800 urinary samples from 177 patients aged from five months to 28 years were enrolled. The relationship between anticonvulsants and urinary pH was analyzed by the chi-square test for independence.
RESULTS: I classified urinary pH as the acid group (pH [lt]/= 5.5), the normal group (pH between 6.0 and 7.5), and the alkaline group (pH [gte] 8.0). Anticonvulsants administrated in this study were carbamazepine (431 urinary examinations), valproate (233), phenobarbital (58), zonisamide (40), sulthiame (22), and phenytoin (16). The acid group in valproate was 4.8 % (10/233) and markedly smaller ( p [lt] 0.0001) than that in other anticonvulsants: carbamazepine 19.0 % (82/431), phenobarbital 22.4 % (13/58), zonisamide 15.0 % (6/40), sulthiame 4.5 % (1/22), and phenytoin 12.5 % (2/16). On the other hand the alkaline group in valproate was 21.5 % (50/233) and significantly greater (p = 0.0285) than that in other anticonvulsants: carbamazepine 14.4 % (62/431), phenobarbital 10.3 % (6/58), zonisamide 15.0 % (6/40), sulthiame 13.6 % (3/22), and phenytoin 6.3 % (1/16). There was no difference in age, sex, and the blood concentration among three urinary pH groups in valproate.
CONCLUSIONS: Patients with valproate monotherapy are less likely to have acid urine but more likely to have alkaline urine than other anticonvulsants. This difference is independent from age, sex, and the blood concentration of valproate. There have been no reports of urinary lithiasis in valproate monotherapy. However, this effect might play a roll in patients with polytherapy containing valproate because polypharmacy is also described as one of the risk factors for urolithiasis. The mechanism of the effect of valproate on urinary pH needs to be investigated.
Support: There is no financial support for this work.