JLE

Epileptic Disorders

MENU

Prospective study on concentration-efficacy and concentration-toxicity: correlations with lamotrigine serum levels Volume 4, issue 1, March 2002

Figures

See all figures

The relationship between lamotrigine (LTG) serum concentration and clinical response in patients with epilepsy is controversial.

Although the existence of a positive correlation between LTG serum concentration and clinical efficacy (reduction in seizure frequency) is supported by many studies [2-11], numerous other investigators, found non such correlation [12-20]. The findings regarding the correlation between LTG serum levels and adverse drug effects are also controversial (positive correlation: 21-25; no correlation: 2, 9, 17, 26-28). This situation prompted us to set up a prospective study where LTG was added on to patients with poorly controlled epilepsy. The aim of the study was to determine the serum concentration of patients with seizure reduction of at least 50% after LTG addition. This was done to assess if there exists a target range of LTG serum levels regarding its clinical efficacy. Furthermore, we wanted to study a possible correlation between the occurrence of LTG adverse effects and its serum concentration.

Patients and methods

Relationship between LTG-serum concentration and efficacy

All patients (mostly outpatients) were exposed to LTG for the first time. Follow-up investigations were done monthly and the LTG dose was increased slowly with respect to seizure frequency and adverse effects. As a rule, LTG was given t.i.d. or b.i.d. The observation period was continued until the patients had reached a seizure reduction of at least 50% (comparison of the last three months during LTG treatment with the last three months before starting LTG treatment). The dose of concomitant antiepileptic drugs had to be kept constant; therefore the observation period was limited. Fifteen adult patients with epilepsy were evaluated with respect to seizure reduction by LTG add-on treatment. Patients who did not reach a seizure reduction of at least 50% were not evaluated further, because the aim of our investigation was to determine which LTG serum concentration was associated with a seizure reduction of at least 50% in those patients who were responsive.

The LTG levels at the end of the successful three month-period were regarded as indicators of a target range. One patient received LTG monotherapy, 14 patients received one or two concomitant antiepileptic drugs (valproic acid (VPA) 6 patients, carbamazepine (CBZ) 4 patients, phenobarbital/primidone (PB/PRM) 4 patients, phenytoin (PHT) 3 patients, bromide (Bro) 1 patient, methsuximide (MSM) 1 patient, vigabatrin (VGB) 1 patient). Nine patients had focal seizures with or without generalization, 6 patients had generalized seizures (group I /table I).

Relationship between of LTG-concentration and toxicity

Because of the small number of patients in group I, the correlation of adverse effects and LTG serum concentration was investigated in a second group of patients with focal epilepsy, independently from the degree of seizure reduction. All patients of group II were outpatients of only one centre (Ravensburg), and they were treated continuously by one of the authors. Any selection of patients was avoided by including all outpatients who had been treated with LTG between February 1996 and May 1999 (with the exception of one patient with renal insufficiency). The only inclusion criterion was the treatment with LTG. In both patient groups, the study started before monotherapy with LTG was officially approved in Germany.

Group II consisted of 33 women and 30 men, with an average age (median) of 35 years (range 16-72 years). Patients were asked at every follow up visit by the treating physician whether they had noted any adverse effect. The results of the clinical examination (e.g. in case of tremor) were also recorded. The determination of the cause-effect relationship in assessing adverse effects was complicated by the fact that most patients were on a combination therapy. Since not all patients were evaluated from the start of LTG treatment (some patients dropped out earlier, for example due to a rash or other intolerance appearing at low dosage), there may be a certain selection bias. Furthermore, patients with no adverse effects during the study period, could have presented initial adverse effects, e.g. dizziness, which subsequently disappeared. We believe that these selection biases have no significant influence on our results.

Serum LTG levels were obtained during each follow up visit. Within three years, 153 LTG determinations (1-12 determinations per patient) were performed.

During the first investigation 63 patients received an absolute LTG dosage of 304 ± 159 mg (range 50-650 mg). With regard to the body weight, the mean dosage was 4.4 ± 2.4 mg/kg (range 0.5-12.5 mg/kg). The mean value for the LTG serum concentration was 5.3 ± 5 mug/ml (range 0.6-22.1 mug/ml). Seven patients of group II had been treated with LTG monotherapy, 32 patients with one additional drug (CBZ 15 patients, VPA 7 patients, PB/PRM 4 patients, PHT 3 patients, gabapentin (GBP) 3 patients). Twenty-one patients received two additional drugs, 3 patients had three additional drugs. The drugs most frequently given in the 56 patients receiving polytherapy were CBZ 16 patients, PB/PRM 10 patients, VPA 8 patients, GBP 4 patients, vigabatrin (VGB) 4 patients.

The concentration of LTG in serum was determined by HPLC (high-performance liquid chromatography) in all patients of groups I and II (in group I patients by an HPLC method of Jürgens [29, 30] or Wad [31]; in group II patients by the HPLC method of Jürgens [29, 30]), exclusively. All centres participate regularly in national or international quality assurance programs. The levels of CBZ, PB, PHT and VPA were determined by FPIA (fluoresence polarization immunoassay, Abbott AxSYM®).

The results are given as means or medians, and ranges. The differences in LTG levels between patient groups were compared using the t-test, the Mann-Whitney U-test or Chi2-test, depending on the statistical assumptions.

Results

Seizure frequency (group I)

The range of the LTG serum concentrations of the 15 patients included in Group I was 1.3-7.1 mug/ml, median = 3.6 mug/ml; and the interval between drug intake and blood collection was 1.2-20 hrs, median = 7 hrs (table I). The daily LTG dosage of the successfully treated patients was 100-400 mg (median = 200 mg; mean = 240 ± 109 mg or 3.4 ± 1.8 mg/kg/day).

Adverse effects (groups I and II)

In group I, three of the 15 successfully treated patients had adverse effects (tremor, LTG level 6.3 mug/ml; dizziness, LTG level 6.8 mug/ml; exanthema, LTG level 2.7 mug/ml) by the time they had reached the intended seizure reduction. Twenty-one patients of group II (n = 63 patients, 153 LTG serum concentrations measured) reported one or more adverse effects during LTG treatment. Overall, 16 different adverse effects were noticed (table II). If a patient had the same adverse effect at different times, it was indicated in table II only once.

Figure 1 shows all LTG serum levels in patients included in group II. Patients are arranged in numerical order on the abscissa. Serum levels of every patient are shown on an imaginary vertical line above the patient number; the crosses represent LTG levels associated with an adverse effect. The increasing number of crosses with high LTG serum concentrations is illustrated, as well as the broad overlapping of serum concentrations associated with adverse effects and serum concentrations not associated with adverse effects. The highest LTG serum concentration without an adverse effect was 20.1 mug/ml, the highest observed LTG serum concentration was 22.1 mug/ml. Adverse effects are reported by many patients with low LTG levels: 45% of serum levels accompanied by an adverse effect were lower than 5 mug/ml, 50% were lower than 10 mug/ml. Between 5 to 13 mug/ml the frequency of LTG levels associated with adverse effects increases only slowly, above 13 mug/ml there is a steep increase.

If one compares the results of the first examination (with one LTG serum concentration for every patient), the difference between the median of serum concentrations associated with an adverse effect (n = 10 patients, median = 4.3 mug/ml LTG (range 1.1-22.1 mug/ml) and the median of serum concentrations not associated with an adverse effect (n = 53 patients, median = 4.0 mug/ml LTG (range 0.6-20.1 mug/ml)), the difference was not significant (P = 0.56, Mann-Whitney U-test). The dosage of LTG also was similar in both groups (P = 0.61, Mann-Whitney U-test). Dosage in patients with an adverse effect: median = 4.4 mg/kg (range 1.1-9 mug/ml), dosage in patients without an adverse effect: median = 4.3 mg/kg (range 0.5-12.5 mg/kg).

Clinical experience suggests a correlation between LTG serum concentration and side effects such as dizziness, tremor and double vision. Four out of 21 patients with adverse effects had dizziness, 4 had tremor, and double vision was reported in only one patient. The LTG serum concentration in 4 patients with dizziness was compared with the LTG serum concentration (first determination) of 42 patients who had no adverse effects at any time. For every patient with "dizziness" at different time points, the serum level at the first appearance of "dizziness" was used. The result was not statistically significant, but there was a strong tendency towards a difference in the group means (p = 0.067, t-test). The mean LTG serum concentration in 4 patients with dizziness was 8.1 ± 4.7 mug/ml. The mean LTG serum concentration in 42 patients (42 LTG levels) without an adverse effect at any time was 4.8 ± 4.1 mug/ml. Patients with dizziness received a mean LTG dosage of 462.5 ± 110.9 mg LTG (6.7 ± 1.9 mg/kg), the 42 patients without any adverse effect received a mean LTG dosage of 304.8 ± 159.6 mg (4.5 ± 2.5 mg/kg). The difference in the LTG dosage was not significant (P = 0.067, Mann-Whitney U-test), but there was a strong trend towards a difference (the 4 patients with dizziness had drug combinations with PHT or PRM or CBZ or CBZ+VPA).

The LTG serum concentration of the 4 patients with tremor (4 values, one value per patient) was also compared with 42 serum levels, obtained from 42 patients, not associated with an adverse effect at any time. The difference was highly significant (P = 0.003, T-test). The mean LTG serum level of the 4 patients with tremor was 14 ± 2.8 mug/ml versus 4.8 ± 4.1 mug/ml in the 42 patients without adverse effects. The mean LTG dosage in patients with tremor was 400 ± 115.5 mg (5 ± 1.7 mg/kg). The 42 patients without adverse effect received a mean LTG dosage of 304.8 ± 156.6 mg (4.5 ± 2.5 mg/kg). The difference in the LTG dosage was not significant (P = 0.212 for the absolute dose, P = 0.58 for the relative dose mg/kg, Mann-Whitney U-test). LTG serum concentration and drug combinations for the 4 patients with tremor are shown in table III.

Discussion

In the first part of this study on LTG serum level monitoring, we studied the correlation between seizure reduction by LTG and its serum level. The range of the LTG serum concentration in 15 patients with a 50% or greater reduction of seizure frequency was 1.3-7.1 mug/ml. The number of patients evaluated is relatively small because the addition of LTG in many patients with refractory partial epilepsies did not result in a seizure reduction of at least 50%. The contribution of new antiepileptic drugs such as LTG to the control of seizures in these patients has been modest [32]. The limited number of patients did not allow the investigation of subgroups with identical dosage schedules and identical intervals between drug intake and collection of blood, and subgroups with respect to factors which might have had an influence on the target range such as seizure type, seizure frequency and drug combinations. Therefore, our target range values are only approximate values. Nevertheless, our results are in agreement with the findings of Stolarek et al. [7]. The target ranges of LTG serum concentration reported in the literature range from 0.5 (lower limit) to 16 mug/ml (upper limit; 4-6, 21, 23-25, 33-36); a target range between 1 and 4 mug/ml was initially suggested [33, 37]. Newer studies suggest the much higher upper limit of 16 mug/ml [23, 24]. In this connection, one has to bear in mind that in some studies these figures refer only to seizure reduction [34-36] or to toxicity [32], whereas in other studies (e.g. 4) both aspects are considered.

As mentioned above several investigators have not found a correlation between LTG serum level and efficacy at all [17].

Several studies demonstrated a positive correlation between LTG dose and seizure reduction [4, 17]. A significant positive correlation between LTG dose and LTG serum concentration has been demonstrated in many studies [4, 5, 12, 34, 38-43]. From these results, a positive correlation between LTG serum concentration and seizure reduction can also be expected. Only exceptionally did investigations involving the correlation between LTG dose and LTG serum concentration not result in a positive correlation [10, 42, 44, 45]. In agreement with the findings of Schapel et al. [6], we have shown that the median LTG serum concentration necessary to reduce seizure frequency is distinctly lower than the median LTG serum concentration in patients with adverse effects. In the study of Schapel et al., the median trough serum LTG level at which the median seizure frequency fell by at least 50% in the responders was 7.9 mug/ml (2.1-15.4). The median serum LTG level at which LTG dose-related side-effects, including ataxia, diplopia, dysarthria, headache and vomiting appeared was 16.0 mug/ml (7.9-19.4 mug/ml). In our patients, we observed a steep increase of the frequency of adverse effects above a serum concentration of 13 to 14 mug/ml; Besag et al. [21] demonstrated a similar distinct increase in toxicity in patients with an LTG serum level above 15 mug/ml. Mayer et al. [23, 24] reported adverse effects at an LTG serum concentration of 16 mug/ml; the manifestation of adverse effects depended on the antiepileptic co-medication. In the case of LTG monotherapy or an LTG/VPA combination, higher LTG levels have been tolerated than with other combinations [23, 24].

The quality and frequency of LTG-related adverse effects in our patients is in accordance with other literature reports [23, 46, 47], with the exception of skin rash which was very rare in our patients (one patient in group I).

The comparison between serum levels associated and those not associated with adverse effects (first examination of every patient) shows a trend but not a significant difference. A statistical significance could be demonstrated in the case of "tremor". The mean LTG serum concentration in 4 patients with "tremor" (mainly action tremor in the upper limbs), was 14 mug/ml. Patients without adverse effects at any time had a mean LTG serum concentration of 4.8 mug/ml. All 4 patients with "tremor" received VPA besides LTG. Because VPA and LTG in monotherapy are able to induce tremor, we cannot be certain if these 4 patients would have had "tremor" if they had received the same LTG concentration with LTG monotherapy. In single cases with an exact observation of the long time course of symptoms such as tremor, it is easier to relate adverse effects to the different components of a combination therapy, e.g. one of the 4 patients with "tremor" (Table III, Et) had an action tremor of the upper limbs and the head with an LTG level of 17.5 mug/ml and a VPA level of 59 mug/ml. Four months earlier, the same patient had no action tremor. At this time the LTG level was lower (13.6 mug/ml), whereas the VPA concentration was even higher (97 mug/ml).

The potential role of therapeutic drug monitoring (TDM) for lamotrigine has been variably assessed due to the wide interindividual variability in the serum concentration at which therapeutic and toxic effects of LTG are observed. LTG is regarded as an "attractive candidate" for therapeutic drug monitoring by Perucca [48], whereas Culy und Goa [37] conclude that the wide range of optimal plasma lamotrigine concentrations (1 to 15 mug/ml) would imply that drug monitoring would be unwarranted. According to Binnie [49] and Kilpatrick [17], routine TDM with LTG is currently not indicated with the exception of checking compliance. Zagnoni et al. [50], however, recommend monitoring of LTG serum concentrations just because of wide intraindividual variability in the serum concentration in patients receiving a constant daily dosage. Our TDM experience with LTG suggests a relationship between serum LTG levels and clinical response which is similar to older antiepileptic drugs. To define a "target range" for LTG should be as useful and as limited as for CBZ or PHT [51, 52].

Our results suggest a broad target range of LTG of about 1 to 13 mug/ml (especially if LTG is used as an add-on drug). This range includes a wide safety margin between the serum concentration which is necessary to reduce seizures and the toxic range. With respect to seizure reduction achieved with LTG concentrations of about 1 to 7 mug/ml and a high risk of adverse effects not before 13 mug/ml, LTG seems to be a rather safe drug, with a broad "therapeutic window".

Further studies with more patients should affirm this data and would confirm whether or not the "target range" of LTG is identical in both monotherapy and polytherapy.

Received August 28, 2001 ; Accepted January 21, 2002

CONCLUSION

Acknowledgements: We are especially thankful to Dr A. Flierl, Pr. Dr H. Stefan (Department of Neurology, University of Erlangen), Dr A. Bredel-Geissler, Priv.-Doz. Dr B. Tettenborn (Department of Neurology, University of Mainz) for the examination of two patients of group I. The authors also thank the following people for their technical support in this study: Dr J. Viering, Labor Dr. Gärtner, Weingarten, Mr. M. Herbold and his colleagues (Ravensburg) and Mrs. D. Herforth (Ravensburg).