1 Serum concentrations of testosterone, androstenedione, dehydroepiandrosterone sulphate (DHAS), sex hormone binding globulin (SHBG) and luteinising hormone (LH) were measured before, during and after 21 days treatment with carbamazepine (CBZ) 400 mg daily in six healthy male subjects. 2 Induction of hepatic microsomal enzyme activity was confirmed by an increase in antipyrine clearance (P < 0.02) and a fall in circulating CBZ concentrations from the seventh to the fourteenth CBZ dose (P < 0.05). 3 Within 7 days of starting CBZ there was a rise in SHBG (P < 0.05) and a fall in testosterone, free testosterone fraction, DHAS and androstenedione (P < 0.05). 4 Testosterone, free testosterone fraction and androstenedione levels rose towards baseline by the end of the treatment period while DHAS concentration remained low (P < 0.05). 5 The rise in SHBG and increased androgen catabolism is most likely to be secoi-dary to induction of hepatic monooxygenase activity by CBZ. 6 These changes may be implicated in the production of sexual dysfunction encountered in some epileptic patients on chronic anticonvulsant therapy.

Steady state carbamazepine (CBZ) plasma concentrations were similar in 15 epileptics receiving monotherapy and in 24 patients taking CBZ in combination with one other anticonvulsant. The ratio of CBZ 10, 11-epoxide (CBZ-E) to the parent drug was significantly higher (P < 0.01) in those patients taking concomitant phenytoin (n = 9), phenobarbitone or primidone (n = 9), and valproic acid (n = 6) than in the patients receiving CBZ alone. In the monotherapy group, there was a significant correlation between CBZ-E/CBZ ratio and the concentration of the parent drug (P < 0.05). If CBZ-E has equipotent anticonvulsant properties to CBZ in man as is the case in animal models, routine CBZ-E concentrations may provide further refinement in the therapeutic drug monitoring of CBZ.

The effect of short-term hepatic enzyme induction with carbamazepine (CBZ) on circulating thyroid hormone concentrations was studied in 10 healthy male subjects. CBZ 400 mg per day was given for 21 days in 6 subjects and for 14 days in a further 4. In the former group the effect of therapy on the pituitary/thyroid axis was also assessed by measuring thyroid stimulating hormone (TSH) response to thyrotrophin-releasing hormone. CBZ therapy resulted in induction of hepatic monooxygenase activity, evidenced by a fall in antipyrine half-life (11.1 +/- 0.7 to 7.6 +/- 0.7 h; p less than 0.001), and a rise in antipyrine clearance (0.72 +/- 0.06 to 0.98 +/- 0.1 ml min-1 kg-1; p less than 0.001). A significant fall in total serum thyroxine (T4) (81.9 +/- 2.9 to 75.1 +/- 2.9 nmol l-1), and triiodothyronine (T3); (1.59 +/- 0.07 to 1.37 +/- 0.05 nmol l-1) and free T4 (16.03 +/- 0.82 to 14.2 +/- 0.8 pmol l-1) was seen after CBZ therapy. (all p less than 0.05). No significant change in reverse T3 or thyroid binding globulin occurred. In the 6 subjects studied for 21 days, maximal changes were found following 14 days' treatment. Basal and stimulated TSH remained unaltered. These effects on circulating thyroid hormone concentrations are likely to be secondary to hepatic enzyme induction leading to accelerated nondeiodinative hepatic hormone disposal. The reason for the failure of pituitary TSH secretion to rise in response to the fall in circulating T4 and T3 is unclear but may have implications for chronic treatment with CBZ in epileptic patients.

1 Simultaneous cerebrospinal fluid (CSF), total and free plasma valproic acid (VPA) concentrations were measured in 17 patients receiving two weight-adjusted VPA doses as seizure prophylaxis prior to diagnostic myelography or cistemography. 2 Free drug concentrations were similar when measured by equilibrium dialysis (ED) at 37°C for 24 h (Dianorm) or by a novel ultrafiltration (UF) method (EMIT freelevel system 1, SYVA) (ED: 2.3-35.5 mg-1; UF: 1.3-33.6 mg-1; r = 0.78, P < 0.002). 3 There was wide variation in total VPA concentration (39-154 mg -l) and in free fraction (ED: 3.3-25.6%; UF: 5.9-24%). Concentration dependent protein binding was not demonstrated. 4 CSF VPA varied between 4.2 and 25.6 mg -l and was accurately reflected by free plasma VPA concentrations (ED: r = 0.75, P < 0.005: UF: r = 0.93, P < 0.001). CSF concentration also correlated with the total plasma VPA (r = 0.76, P < 0.005). 5 The Emit freelevel system 1 provides a rapid measure of unbound VPA in the plasma which may be suitable for routine clinical use.

1 Antipyrine metabolism, daily urinary 6-,3-hydroxycortisol excretion, carbamazepine (CBZ) halflives and leucocyte delta-aminolaevulinic acid synthase (ALA.S) activities were measured following 2 weeks' treatment with CBZ 400 mg and 600 mg once daily in eight healthy male volunteers. 2 Dose-dependent induction of antipyrine metabolism was demonstrated but cortisol hydroxylation appeared maximally induced by the 400 mg dose. 3 CBZ half-lives fell significantly in both studies (P < 0.01 in each case) but a greater fall was seen with the higher dose (P < 0.01). Plasma CBZ concentrations were higher following the eighth doses (P < 0.01) in both studies. 4 Leucocyte ALA.S activity increased by a mean of 657% following 1 week's treatment with 400mg CBZ and 1145% on the 600 mg dose. In both studies ALA.S activities fell towards baseline during the second week of treatment. 5 CBZ possesses potent dose-dependent hetero-and auto-inducing properties. 6 Leucocyte ALA.S activity may represent a novel in vivo index of extrahepatic enzyme induction in man.