- Top of page
- SUBJECTS AND METHODS
Summary: Purpose: Antiepileptic drugs may affect the serum thyroid hormone concentrations. The aim of this study was to evaluate thyroid function in 78 girls taking carbamazepine (CBZ), oxcarbazepine (OXC), or valproate (VPA) monotherapy for epilepsy and after withdrawal of the treatment.
Methods: Forty-one girls taking VPA, 19 taking CBZ, and 18 taking OXC for epilepsy, as well as 54 healthy age-matched controls, aged 8 to 18 years, participated in the study. All the girls were examined clinically, and their pubertal stage was assessed. Blood samples were obtained for thyroid hormone and antibody assays. These examinations were repeated after a mean follow-up of 5.8 years to assess thyroid function, and 64 (82%) of 78 patients and 42 (78%) of 54 controls agreed to participate in the second evaluation.
Results: In the first evaluation, the mean serum thyroid hormone concentrations were lower in the girls taking CBZ [thyroxine (T4), 70.2; SD, 10.9 nM; and free thyroxine (FT4), 11.5; SD, 1.8 pM] or OXC (T4, 74.9; SD, 16.4 nM; and FT4, 11.3; SD, 1.8 pM) than in the control girls (T4, 96.6; SD, 15.1 nM, and FT4, 14.4; SD, 1.5 pM; p < 0.001, all comparisons). However, thyrotropin (TSH) concentrations were normal in the girls taking CBZ or OXC. Sixty-three% of the girls taking CBZ and 67% of the girls taking OXC had serum T4 and/or FT4 levels below the lower limit of the reference range. The VPA-treated girls with epilepsy had normal serum T4 and FT4 concentrations, but slightly increased TSH levels (3.3; SD, 1.5 mU/L; p < 0.01) compared with the control girls (2.5; SD, 1.0 mU/L). Normal serum hormone concentrations were restored in the patients who discontinued the medication.
Conclusions: Both CBZ and OXC reduce serum thyroid hormone concentrations in girls with epilepsy. Conversely, VPA is associated with normal serum thyroid hormone and increased thyrotropin levels. However, our results suggest that the changes in serum thyroid hormone and thyrotropin levels are reversible after withdrawal of the medication.
It is well known that the traditionally used antiepileptic drugs (AEDs) may affect thyroid function (1–5). The effects of AEDs on thyroid function in children with epilepsy have been investigated less extensively, however. Moreover, no data exist on thyroid function in children treated with oxcarbazepine (OXC).
Carbamazepine (CBZ) is the most widely used first-line AED in partial-onset epilepsy of both adults and children (6). OXC is the 10-keto analogue of CBZ and has anticonvulsant efficacy comparable to that of CBZ in partial seizures with or without generalization (7). Despite this, CBZ and OXC have different metabolic pathways in the liver: CBZ is oxidized and OXC is metabolized mainly by reduction to its active metabolite, 10,11-dihydro-10-hydroxy-carbamazepine.
CBZ therapy has certain effects on thyroid function: it decreases the serum thyroid hormone levels, but allows the serum thyrotropin (TSH) concentrations and TSH responses to thyrotropin-releasing hormone to remain normal (2,4,8). It has been speculated that these low serum thyroid hormone levels are due to the induction of the hepatic P450 enzyme system by CBZ. Despite its close structural resemblance to CBZ, neither OXC nor its active monohydroxy metabolite induces the oxidative P450 enzyme system to a similar extent as does CBZ (9). Low serum thyroid hormone concentrations induced by CBZ have been found to normalize during the first 6 months after replacement of CBZ with OXC in adult patients with epilepsy (10). However, OXC has been found to induce hepatic enzymes when given at high doses to adult patients with epilepsy (11), and low serum thyroid hormone concentrations were reported in men with epilepsy receiving long-term treatment with OXC (12).
Valproate (VPA) is widely used for the treatment of both partial and generalized epilepsy in childhood and adolescence (13). The reports on the effects of VPA on serum thyroid hormone concentrations have been controversial (1,5,8,12,14). However, VPA medication has been reported to have effects on the metabolic and endocrine function in women and girls with epilepsy (15–17).
The aim of this study was to evaluate thyroid function during pubertal development in girls and female adolescents taking CBZ, OXC, or VPA medication for epilepsy and years after withdrawal of the medication.
SUBJECTS AND METHODS
- Top of page
- SUBJECTS AND METHODS
This study, which consisted of two cross-sectional evaluations, was carried out at Oulu University Hospital. It was performed according to the provisions of the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Medicine, University of Oulu. Informed consent was obtained from all the patients and control subjects and/or their parents.
The study series comprised 78 girls with epilepsy, aged 8 to 18 years, who were consecutive patients visiting the Outpatients Departments of Pediatrics or Neurology at Oulu University Hospital (16,17). The inclusion criteria were epilepsy and AED monotherapy. Eighteen girls were excluded because of brain damage because of preterm birth, hydrocephalus, progressive brain disease, epilepsy surgery, or trauma, or because they had received steroid treatment for intractable epilepsy. In addition, one girl was excluded because of use of oral contraceptives (OCs), and two girls refused to participate. Epilepsy type was classified according to the recommendations of the International League Against Epilepsy (18). The AED was chosen according to the type of epilepsy; CBZ or OXC was used particularly for the treatment of patients with partial seizures with or without generalization, and VPA was used primarily for patients with primary generalized seizures. The 54 age-matched control subjects were recruited from local primary and secondary schools. The reproductive endocrine function and growth of these girls have been reported earlier (16,17).
The second evaluation of thyroid function in these patients was performed after a mean of 5.8 years (SD, 1.0 year). Altogether 64 (82%) of 78 patients and 42 (78%) of 54 controls agreed to participate. One patient had died. Four patients initially treated with CBZ, three patients with OXC, and seven patients with VPA refused the follow-up examination.
Blood samples were taken at 8 to 10 a.m. after an overnight fast and after the application of a local anesthetic (Emla; Astra, Södertälje, Sweden) on the skin, if needed. All the subjects were examined clinically by one of the authors. The stage of puberty was assessed according to Tanner and Whitehouse (19). Bone age was evaluated by the method of Greulich and Pyle based on a radiograph of the left hand and wrist (20). The subjects were clinically classified into three pubertal stages. They were classified as prepubertal if no clinical signs of puberty were noted [i.e., breast 1 (B1) and pubic hair 1 (PH1)], and as pubertal if they had clinical signs of puberty (B2-4 or PH2-4), whereas subjects with full maturation (B5PH5 or B4PH4 and adult bone age) were classified as postpubertal.
The serum samples were stored frozen at −20°C until analyzed. Serum T4 concentrations were measured by radioimmunoassays with kits obtained from Orion Diagnostica (Turku, Finland). The sensitivity of the T4 assay was <5 nM. The coefficient of intraassay variation was 4.7%, whereas that of interassay variation was 4.9%. The serum FT4 and free triiodothyronine (FT3) concentrations in the first evaluation were assayed with the Chiron Diagnostics automated chemiluminescence system by using an ACS:180 analyzer (Medfield, MA, U.S.A.). The sensitivity of the FT4 assay was 1.3 pM, and the intra- and interassay variations were 2.5% and 5.1%, respectively. The respective values for the FT3 assay were 1.6 pM, 1.4%, and 2.0%. The FT4 and FT3 concentrations in the second evaluation were determined by using immunologic chemiluminescence methods and a Bayer ADVIA Centaur analyzer (Tarrytown, NY, U.S.A.). The sensitivity of the FT4 assay was 2.0 pM, the intraassay variation was 3.9%, and the interassay variation was 4.9%. The respective values for the FT3 assay were 1.0 pM, 1.6%, and 1.8%. Serum TSH concentrations were measured with a two-site fluoroimmunometric method with kits obtained from Wallac Oy (Turku, Finland). The sensitivity of the TSH assay was <0.03 mU/L, and the intra- and interassay variations were 2.0% and 3.5%, respectively. The serum TPO-ab and TG-ab levels were measured with radioimmunoassays by using kits from Brahms Diagnostica (Berlin, Germany). In both tests, antibody levels >60 U/ml were considered positive, as recommended by the manufacturer. The serum CBZ and VPA concentrations were analyzed with a fluorescence polarization immunoassay system by using the AxSym analyzer (Abbott Diagnostic Division, Irving, TX, U.S.A.). The sensitivity of the CBZ assay was 0.5 μM, the intraassay variation was 1.2%, and the interassay variation was 3.2%. The respective values for the VPA assay were 5.0 μM, 1.8%, and 2.7%. The serum concentrations of monohydroxycarbazepine (the active metabolite of OXC) were determined by high-performance liquid chromatography (HPLC) (21). The sensitivity of the assay was 0.1 μM, and the intraassay and interassay variations were <5%.
The differences between two groups were analyzed by t test as appropriate. The differences in the mean serum concentrations of the thyroid hormones between the three patient groups and the control girls were analyzed with one-way analysis of variance (ANOVA) with an LSD post hoc test. If homogeneity of variance was not assumed, ANOVA was not usable, and the results were confirmed by the Kruskal–Wallis test. To find out the effect of age on the changes of hormone levels between two evaluations in the different groups, repeated measures of ANOVA were performed with age as a covariate. The changes of serum hormone concentrations between the two evaluations were measured by ANOVA.
- Top of page
- SUBJECTS AND METHODS
Initially, 19 girls were receiving CBZ treatment; 18 girls, OXC treatment; and 41, VPA. The clinical characteristics of the girls participating in the study are given in Table 1. All these patients had drug concentrations at a therapeutic level. At the second visit, 37 (58%) of the 64 patients were off medication: 10 of the CBZ, 10 of the OXC, and 17 of the VPA group. Only three patients continued on CBZ, two on OXC, and six on VPA monotherapy. The medication of the other patients had been replaced by some other AED (n = 13) or polytherapy (n = 3, one VPA and CBZ and two VPA and lamotrigine). Their results are not shown in detail.
Table 1. Clinical characteristics of the 78 girls receiving antiepileptic monotherapy
|Medication||N||Mean age||Type of epilepsy||Duration of medication;||Mean dose/day|
|(range), yr||G||P||PG||mean (range), yr||(range), mg|
|1. Carbamazepine||19||12.7 (8.4–18.9)|| 0|| 8||11||4.1 (0.8–12.8)||489 (300–700) |
|2. Oxcarbazepine||18||12.7 (8.5–17.1)|| 2|| 9|| 7||1.8 (0.9–3.6) ||867 (450–1,500)|
|3. Valproate||41||12.5 (8.4–18.5)||24||10|| 7||3.0 (0.8–10.3)||720 (300–1,200)|
First cross-sectional evaluation
The girls taking CBZ or OXC had significantly lower serum T4 and FT4 levels than the control girls, but no significant differences were found in the serum TSH concentrations (Table 2). Of the 19 girls receiving CBZ treatment, seven were prepubertal, eight pubertal, and four postpubertal, and of the 18 girls taking OXC, eight were prepubertal, five pubertal, and five postpubertal. The serum thyroid hormone levels were similar in the different pubertal groups.
Table 2. Serum thyroid hormone and thyrotropin levels in girls with epilepsy and age-matched control girls at the first evaluation
|Medication||T4 (nM)||FT4 (pM)||FT3 (pM)||TSH (mU/L)|
|Carbamazepine||70.2 (10.9)a||11.5 (1.8)a||5.2 (0.6)||2.3 (0.8) |
| (n = 19)|| |
|Oxcarbazepine||74.9 (16.4)a||11.3 (1.8)a||5.1 (0.5)||2.1 (0.9) |
| (n = 18)|| |
|Valproate||92.8 (18.1) ||14.2 (1.6) ||5.7 (0.8)||3.3 (1.5)b|
| (n = 41)|| |
|Control girls||96.6 (15.1) ||14.4 (1.5) ||5.5 (0.7)||2.5 (1.0) |
| (n = 54)|| |
The mean serum thyroid hormone T4 and FT4 levels in the eight girls taking a high (≥20 mg/weight kg) dose of OXC did not differ from those in the 10 girls taking a low (<20 mg/kg) dose. The duration of OXC therapy did not affect the serum thyroid hormone levels.
Forty-one girls with epilepsy were receiving VPA treatment, with a mean duration of 3.0 years (range, 0.8 to 10.3 years). Of these girls, 16 were prepubertal, 11 pubertal, and 14 postpubertal. The serum T4 and FT4 concentrations did not differ in the girls taking VPA from those of the control girls, but the serum TSH levels were significantly higher than the TSH levels in the control girls (p < 0.01; Table 2). Twelve (63%) of 19 girls taking CBZ and 12 (67%) of 18 taking OXC had serum T4 and/or FT4 levels below the lower limit of the reference level in the respective pubertal groups (T4 under/or 85.9 nM in prepubertal; 58.3 nM in pubertal; 67.4 nM in postpubertal girls; FT4 under/or 12.4 pM in prepubertal; 10.1 pM in pubertal; 11.5 pM postpubertal girls) (Fig. 1), whereas only four (10%) of 41 girls taking VPA had T4 and/or FT4 levels below these reference levels. The lower level of the reference range was defined as a mean serum concentration –2 SD of the respective values of the healthy control girls.
Figure 1. Frequency of serum thyroxine (T4) and/or free thyroxine (FT4) levels below the reference range in girls taking antiepileptic drugs (CBZ, carbamazepine; OXC, oxcarbazepine; VPA, valproate) and controls.
Download figure to PowerPoint
All the girls with epilepsy receiving different drug treatments were clinically euthyroid. One of the girls treated with CBZ had positive TPO-ab levels at the first evaluation, and one of the VPA-treated girls had positive TPO-ab levels, whereas none of the OXC-treated girls had elevated ab levels. Three (4%) control girls had positive TPO- and TG-ab levels, and they were also clinically euthyroid, and their serum thyroid hormone levels were similar to those recorded in the other control subjects. During the follow-up period, one patient had had autoimmune thyroiditis ˜3 years before the second evaluation, but she was euthyroid at the second evaluation. In one of the control girls who had had positive antibody levels at the baseline, hypothyroidism developed, and another had Hashimoto thyroiditis. These two control girls were excluded from the analysis at the second evaluation.
Second cross-sectional evaluation
The clinical characteristics and laboratory results of the 37 patients off medication and the control subjects at the second evaluation are presented in Table 3. The mean time after AED discontinuation was 4.8 (SD, 1.1) years. The serum concentrations of TSH, FT4, or free triiodothyronine (FT3) of the patients off medication did not differ from those of the control subjects. No interaction was observed between age and the change of the serum TSH level (p = 0.09) or the serum FT4 level (p = 0.3) in the different groups. The decreased serum FT4 levels observed at the first evaluation in the girls taking CBZ and OXC had returned to normal by the time of the second evaluation after discontinuation of the medication. The mean increase of the serum FT4 level in the patients off medication was 4.1 pM (SD, 2.0) in the CBZ group, 3.9 pM (SD, 3.1) in the OXC group, and 0.1 pM (SD, 2.1) in the control group; p < 0.0005. The corresponding values of the patients off medication in the VPA group were 1.7 pM (SD, 2.9), and those of the control subjects, 0.1 pM (SD, 2.1); p < 0.04. The elevated serum TSH levels of the patients taking VPA at the first evaluation had decreased after withdrawal of the medication. The mean serum TSH concentration was still high in the patients continuing on VPA 4.4 mU/L (SD, 3.7) compared with the patients off medication but initially taking VPA, 2.1 mU/L (SD, 0.8), or the control subjects, 2.4 mU/L (SD, 1.1; Fig. 2).
Table 3. Clinical characteristics and laboratory results of patients off medication and control subjects at the second evaluation
| || N|| Age (yr)|| AED therapy||Duration of AED (yr)||Time without TSH (mU/L)|| FT4 (pM)|| FT3 (pM)|
|Patients Original medication|| |
|CBZ||10||17.5 (1.8)||4.6 (1.8)||5.0 (1.1)||1.9 (0.7)||15.6 (2.4)||5.5 (0.4)|
|OXC||10||18.2 (2.2)||3.4 (1.0)||4.8 (0.9)||2.1 (0.9)||14.9 (3.7)||5.2 (0.9)|
|VPA||17||17.5 (2.6)||4.8 (2.3)||4.7 (1.2)||2.1 (0.8)||15.8 (2.4)||5.3 (0.6)|
|All||37||17.7 (2.3)||4.4 (2.0)||4.8 (1.1)||2.1 (0.8)||15.5 (2.7)||5.3 (0.6)|
|Controls||40||17.4 (3.4)|| || ||2.4 (1.1)||14.7 (2.0)||5.3 (0.9)|
Figure 2. Mean serum thyrotropin (TSH) concentrations in patients off medication at the second evaluation initially treated with carbamazepine (CBZ), oxcarbazepine (OXC), or valproate (VPA) and ˜5 years after cessation of the medication, in patients still taking VPA at the second evaluation, and in control subjects. The changes of serum TSH between the two evaluations were ap <0.05 compared with control subjects or patients off medication initially taking OXC and ap <0.006 compared with patients off medication initially taking VPA or CBZ; bp < 0.001 compared with control subjects and bp = 0.01 with patients off medication initially taking OXC; the analysis of variance, post hoc LSD.
Download figure to PowerPoint
Two of the 10 girls taking CBZ, two of the 10 girls taking OXC, six of the 17 girls taking VPA, and 11 of the 40 control subjects were using OCs. The use of OCs did not affect the TSH or FT3 levels, but had an effect on the FT4 level. The FT4 concentrations in the girls without OC are presented in Fig. 3.
Figure 3. Mean serum free thyroxine (FT4) concentrations in two evaluations of patients off medication in the three initial medication groups and control subjects with no oral contraceptives; the change of serum FT4 between the two evaluations ap < 0.005 compared with the change in control subjects, and ap < 0.03 compared with the patients off medication initially taking valproate; analysis of variance, post hoc LSD.
Download figure to PowerPoint
- Top of page
- SUBJECTS AND METHODS
All of the three AEDs studied affected thyroid function during pubertal development in girls with epilepsy. Both CBZ- and OXC-treated girls had low serum T4 and FT4 concentrations, but the serum TSH levels remained unchanged, and the girls were clinically euthyroid. Conversely, the serum TSH levels were slightly increased, but the T4 and FT4 concentrations were normal among the VPA-treated girls. These altered serum hormone levels returned to normal after discontinuation of medication. To our knowledge, no previous long-term controlled follow-up studies on thyroid function exist for patients with epilepsy during and after AED treatment from childhood and adolescence until adulthood.
More than half of the girls taking CBZ or OXC had serum T4 and/or FT4 below the reference range. These findings are similar to and of same magnitude as those earlier reported in CBZ-treated children and adults (2,4,8,22,23). To our knowledge, however, no previous studies exist on the effects of OXC on thyroid function in children. CBZ is known to induce the hepatic P450 enzyme system, and it has been suggested that it may increase the metabolism of thyroid hormones by this mechanism. The girls in the present study had received OXC for epilepsy for ≥0.9 years, the mean duration of OXC therapy was 1.8 years, and they had low serum thyroid hormone levels. Serum thyroid hormone concentrations were equally low in the girls taking high or low doses of OXC. Recently, low serum thyroid hormone levels also were reported in men taking long-term OXC for epilepsy. No correlation was noted between serum thyroid hormone concentrations and γ-glutamyl transferase (GGT) levels, which were used as markers of liver enzyme induction (12). T4 and FT4 levels were found to be similarly low in men taking either high or low doses of OXC. These results suggest that OXC-associated changes in serum thyroid hormone concentrations also are seen with low doses of the drug. Conversely, replacement of CBZ with OXC in men with epilepsy resulted in restoration of normal thyroid hormone levels during a 6-month follow-up (10). These results further support the view that hepatic P450 enzyme induction is not the main or the only reason for decreased thyroid hormone concentrations during CBZ or OXC treatment. Surks and DeFesi (24) showed that patients taking CBZ medication had low FT4 and FT3 levels when serum samples were analyzed by a commercial procedure from diluted serum, but not when an ultrafiltration method from undiluted serum was used. Thus the low serum FT4 levels in the CBZ-treated girls in the present study may be related to the method used for the FT4 assay. Moreover, the low serum FT4 concentrations observed in the OXC-treated patients also may be ascribed to the same reason. However, the decreases found in serum T4 levels in girls taking CBZ or OXC are evidently real and reflect competitive displacement of T4 from TBG by these drugs (24).
VPA is widely used in childhood epilepsy. In the present study, serum TSH levels were increased in girls treated with VPA. The serum T4 and FT4 concentrations were within the normal range, and the mechanism and clinical significance of the elevated serum TSH levels during VPA treatment remain unknown. In earlier reports, the effects of VPA on serum thyroid hormone levels have been inconsistent (1,5,12,14,25). Verrotti et al. (8) did not find any alterations in thyroid hormone metabolism, and the TSH response to thyroid-releasing hormone also was normal. However, findings similar to the present ones with slightly increased serum TSH levels and normal serum thyroid hormone concentrations have previously been reported in adults taking VPA for epilepsy (26).
Despite the low serum thyroid hormone concentrations, both CBZ- and OXC-treated girls had normal serum TSH levels and were clinically euthyroid. Positive serum TPO-ab and TG-ab levels were equally frequent as in the control subjects. Thus the possible clinical implications of the altered serum thyroid hormone concentrations in these girls remain unknown. However, it is reassuring that, according to the present results, the AED-associated changes in serum thyroid hormone levels are not permanent, and normal serum thyroid hormone concentrations are restored after the discontinuation of AEDs. All the girls with epilepsy were euthyroid, and CBZ, OXC, and VPA had no effect on their linear growth or pubertal development (16). Moreover, the weight of the girls taking CBZ or OXC was similar to the weight of the control girls. These observations imply that the low serum thyroid hormone concentrations during CBZ or OXC medication are not associated with clinical or even subclinical hypothyroidism with effects on growth.
In conclusion, all AEDs studied had an effect on thyroid function in girls during pubertal development. However, these hormonal changes had no effect on growth or pubertal development, and normal serum thyroid hormone concentrations were restored after discontinuation of the medication. It appears that measurement of serum TSH concentration is the most reliable means to evaluate the true thyroidal state of patients taking CBZ, OXC, or VPA for epilepsy.