To the Editor: Concomitant carbamazepine and isoniazid therapy, with or without rifampin, has produced increases in carbamazepine serum concentrations (1–4), causing symptoms of toxicity. However, the significance of a potential interaction between rifampin and carbamazepine has received little attention.
Ms. A, a 44-year-old woman, was admitted to the psychiatric ward with symptoms of hypomania. She had a 10-year history of bipolar affective disorder that had been adequately controlled with carbamazepine, 200 mg b.i.d., and trifluoperazine, 5 mg b.i.d. Before admission, because of suspected tuberculosis, Ms. A had been given 600 mg/day of rifampin, 300 mg/day of isoniazid, and 50 mg/day of pyridoxine. Her serum carbamazepine level 2 weeks after admission was 3.1 mg/liter (therapeutic range=4–10); thus, her dose was increased to 200 mg t.i.d. Subsequent serum measurements (1 week apart) provided the following readings: 2.8, 2.5, and 2.3 mg/liter. The results of liver and renal function tests were normal.
Although our literature search retrieved no evidence of a drug interaction involving antituberculosis agents and a decrease in serum carbamazepine levels, the most likely cause of the continued low levels was a combination of carbamazepine auto-induced metabolism and the concomitant antituberculosis agents Ms. A was taking. The antituberculosis drugs were discontinued, and 5 days later, Ms. A’s serum carbamazepine level was 4.9 mg/liter. Ms. A was discharged with a carbamazepine dose of 400 mg b.i.d., having fully recovered from her hypomanic episode. Two months after discharge, her carbamazepine level was found to be 4.6 mg/liter, and her mood was euthymic.
Rifampin is a known inducer of many cytochrome P-450 enzymes and has been shown to reduce the plasma concentrations of many drugs used in psychiatry, such as triazolam, zopiclone, zolpidem, and midazolam (5). It has also been reported that both rifampin and carbamazepine are inducers of the same liver isoenzymes (CYP 2C9, CYP 2C19, CYP 3A4) (6). In our review of the literature, not one report was found describing a possible effect of rifampin in further inducing carbamazepine’s auto-induced metabolism, leading to progressively low blood concentrations of the mood stabilizer.
This case highlights the importance of monitoring carbamazepine levels, particularly when carbamazepine is co-administered with drugs known to be either inducers or inhibitors of hepatic enzymes. Because predicting the clinical significance of the interaction is difficult, it is wise to watch not only for signs of carbamazepine toxicity but also for signs of inadequate symptom control. Close monitoring of serum carbamazepine levels after addition or deletion of antituberculosis drugs is highly advisable so that the carbamazepine dose can be adjusted accordingly.