Maurice Dickins PhD*
Chao Chen PhD**
* Senior Scientist, Drug Metabolism and Pharmacokinetics, GlaxoSmithKline, Ware, Hertfordshire, United Kingdom
** Section Head, Pharmacokinetics, Department of Clinical Pharmacology and Experimental Medicine, GlaxoSmithKline, Greenford, Middlesex, United Kingdom
Lamotrigine (3,5-diamino-6-[2,3-dichlorophenyl]-1,2,4-triazine, Lamictal) is synthesized by reacting thionyl chloride with 2,3-dichlorobenzoic acid, yielding an acid chloride derivative, which is converted to the corresponding ketonitrile in the presence of cuprous cyanide. Condensation of the ketonitrile with aminoguanidine under strongly acidic conditions (6 mol/L HNO3) produced an amidinohydrazone product that readily cyclized in basic conditions to yield lamotrigine (1). Lamotrigine is a weak base with a negative log of dissociation constant (pKa) of 5.5 and a molecular weight of 256.09 (2). In contrast, metabolites of lamotrigine substituted at the N-2 nitrogen such as the N-2 glucuronide are much stronger bases (pKa = 10.6) because of the presence of a quaternary N atom. Lamotrigine is poorly soluble in water (0.17 mg/mL at 25°C) with a log octanol/water partition coefficient (Log P) of 1.19 at pH 7.6. Isethionate and mesylate salts of lamotrigine have been synthesized with greater aqueous solubility.
The metabolism of lamotrigine occurs predominantly by attack at the N-2 nitrogen atom of the molecule, although there are species differences both in the nature and the extent of metabolism (3). The major human metabolite seen in vivo was the aromatic N-2 glucuronide (4), whereas in the dog, the N-2 methyl derivative was the primary metabolite. In rodents, the N-2 oxide was a significant metabolite, but a substantial proportion of the drug was excreted as the parent molecule. Lamotrigine also undergoes significant metabolism in the guinea pig to the N-2 glucuronide (5); both the rabbit and cynomolgus monkey also produced substantial amounts of the N-2 glucuronide in addition to the aliphatic N-5 glucuronide (6) (Figure 35.1).
In healthy volunteers, the total dose after oral administration of [14C]-labeled lamotrigine over a 1-week period was essentially recovered in the urine, with 71% of the dose as the N-2 glucuronide, 9% as the N-5 glucuronide, 10% as unchanged lamotrigine and 0.1 % as the N-2 methyl metabolite (1). In patients, a similar amount (43% to 87% of the dose) was excreted in the urine, mainly as the N-2 glucuronide (7). The majority (60%) of an intravenous dose of lamotrigine in the guinea pig was recovered in the urine as the N-2 glucuronide metabolite (5). In contrast, 50% of the dose in rats and mice was excreted in the urine as unchanged parent drug, whereas approximately 50% of the dose in the dog was N-2-methyl lamotrigine (1).
Lamotrigine is metabolized in the rat both in vivo and in isolated hepatocytes to a glutathione conjugate by epoxidation of the dichlorophenyl ring, although the actual sites of metabolism were not determined (8). In vivo metabolites excreted in the bile were the arene oxide intermediate (found as an unstable glutathione conjugate) and the dehydrated glutathione adduct of the parent drug together with the cysteinylglycine and N-acetylcysteine adducts of lamotrigine. Rat hepatocytes produced the glutathione adduct and the N-2 oxide. However, neither rat nor human liver microsomes catalyzed NADPH (reduced form of nicotinamide-adenine dinucleotide phosphate)-dependent irreversible binding to microsomal protein.
FIGURE 35.1. Structures of lamotrigine and its major metabolites.
N-2 Methyl Metabolite
Metabolism of lamotrigine to the N-2 methyl metabolite in the dog was hardly detectable in all other species investigated, including humans. The reaction was catalyzed by a nonspecific methyltransferase found in dog liver (1). Hepatic cytosol preparations from the dog (but not other species) were shown to metabolize lamotrigine to the N-2 methyl derivative in the presence of the cofactor S-adenosyl methionine. There is a literature precedent for this class of reaction being specific to the dog; the tetrahydroisoquinoline compound SK&F 64139 and related compounds also undergo N-methylation in this species (9).
Several studies have demonstrated the metabolism of lamotrigine to its N-2 glucuronide metabolite using hepatic microsomal preparations. Liver microsomes from rabbits (10), guinea pigs (5), and humans (10,11) all catalyzed the formation of the N-2 glucuronide in the presence of the cofactor uridine diphosphate (UDP) glucuronic acid. Isolated hepatocytes from guinea pigs (5) and humans (12) also catalyzed the metabolism of lamotrigine to this metabolite. These findings are in agreement with the known in vivo metabolic profile of lamotrigine for these species and suggest that the liver is a major site of metabolism to the N-2 glucuronide.
Michaelis constant (Km) and maximum velocity (Vmax) values for the formation of lamotrigine N-2 glucuronide are similar using both guinea pig and human liver microsomes as the enzyme source: Km = 2.1 mmol/L (guinea pig) (5) and 2.6 mmol/L (human) (10); Vmax = 252 pmol/min/mg protein (guinea pig) (5) and 650 pmol/min/mg protein (human) (10). More recent data with a greater number of human liver samples (n = 12) produced values of Km = 5.5±5.2 mmol/L and Vmax = 960±380 pmol/min/mg protein (11). In this latter study, some liver samples from patients with liver cirrhosis (n = 10) did not show an appreciable
change in the kinetics of lamotrigine N-2 glucuronidation (Km = 4.3±2.0 mmol/L; Vmax = 850±540 pmol/min/mg protein). However, the clearance of lamotrigine in vivo was reduced in diseased patients compared with those with normal liver function (see the later discussion of hepatic dysfunction). A reduction in the dose of lamotrigine for patients with liver cirrhosis was advised.
Glucuronidation is a major conjugation reaction that is catalyzed by a number of different isoforms of UDP-glucuronosyltransferase (UGT) (13,14). N-glucuronidation is a now a well-established pathway in the human metabolism of drugs with a tertiary amine group (15). Although the substrate specificity of many UGTs is unclear, UGT1A4 has been implicated in the formation of quaternary ammonium-linked glucuronides including lamotrigine (16). A panel of expressed UGT isoforms has been used to investigate the metabolism of 1-phenylimidazole, a model substrate for the quaternary nitrogen-linked glucuronidation reaction (17). Only UGTs 1A3 and 1A4 were capable of catalyzing the formation of 1-phenylimidazole N+ glucuronide. However, a significant contribution of UGT1A3 to the overall metabolism of tertiary amines by human liver preparations is unlikely (18), given that the expression of UGT1A3 is very low in human liver (19). Lamotrigine has also been cited as a UGT1A4 substrate in inhibition experiments with the anticonvulsant agent retigabine (20). In this study, retigabine was metabolized to an aliphatic N-glucuronide by human liver microsomes and several expressed UGT isoforms including UGT1A1 and 1A4. However, retigabine N-glucuronidation by human liver microsomes was inhibited by lamotrigine and not by bilirubin, a substrate for UGT1A1.
The evidence taken together suggests that lamotrigine N-2 glucuronidation, the major route of metabolism in humans, is catalyzed by human UGT1A4. This pathway is inhibited by the anticonvulsant drug valproate (21, 22, 23), and it is inducible by other anticonvulsants (22,23). Rifampicin has also been shown to increase the clearance of lamotrigine (24). Direct evidence of induction of UGT was shown by measurement of increased production of lamotrigine N-2 glucuronide after rifampicin treatment.
HUMAN PHARMACOKINETICS OF LAMOTRIGINE
Lamotrigine is available for oral administration as conventional and chewable or dispersible tablets. Although the availability of the formulations and their strengths vary among the regional markets, the conventional tablets exist in 25-, 50-, 100-, 150-, 200-, and 250-mg strengths and the chewable or dispersible tablets exist in 2-, 5-, 25-, and 100-mg strengths. Gelatin oral capsules were used during early clinical testing. Bioequivalency has been demonstrated among the capsules, the conventional tablets, and the chewable or dispersible tablets. Bioequivalency has also been established when the chewable or dispersible tablets are swallowed whole, chewed, or ingested after dispersion in a small amount of liquid (25). Treatments are considered bioequivalent when the 90% confidence intervals for the geometric least-square mean ratios of the area under the systemic concentration-time profile and of the maximum systemic concentration between the treatments are between 0.8 and 1.25. Because data on individual bioequivalence are currently lacking, patients need close clinical monitoring for toxicity or seizure worsening when they are switched between formulations.
Oral absorption is rapid, with the peak plasma concentration typically appearing between 1 and 3 hours after administration (Table 35.1). The average maximum plasma concentration increased linearly from 0.4 to 3.16 µg/mL after a 30- to 240-mg dose and from 0.58 to 4.63 µg/mL after a 50- to 400-mg dose (26, 27, 28). The steady-state maximum plasma concentration also increased dose proportionally from 3.45 to 9.44 µg/mL in healthy subjects receiving once-daily doses of 50 to 150 mg of lamotrigine in addition to concurrent sodium valproate and 0.96 to 3.00 µg/mL in patients receiving twice-daily doses of 50 to 150 mg lamotrigine in addition to concurrent enzyme-inducing anticonvulsants (29,30). Time to the maximum concentration is not altered by concurrent administration of sodium valproate or enzyme-inducing anticonvulsants, and it is dose independent after a single dose in the absence of any other medication or at steady state with these concurrent medications (26,28, 29, 30).
Oral absorption is complete. A single 75-mg dose showed an absolute bioavailability of 98±0.05% when the drug was given to eight healthy men and women who also received an equivalent intravenous dose (31). The nearcomplete oral absorption was supported by an average of 94% urinary recovery of radioactivity from six healthy subjects who ingested 240 mg [14C]-labeled lamotrigine (32). This means that lamotrigine is unlikely to show variable absorption that can sometimes be caused by factors including concurrent medication or meals. Indeed, whereas food consumed at the time of tablet administration slightly delays the occurrence of the peak plasma concentration, it does not alter the peak concentration or the area under the concentration-time curve (27). Therefore, there is no need to control mealtimes when one is taking lamotrigine.
After intravenous or oral administration, the plasma concentration of lamotrigine is best characterized by a one-compartment open model with first-order elimination (26,31, 32, 33,34). Weight-normalized volume of distribution after intravenous administration and weight-normalized apparent volume of distribution after oral administration have been reported to be 1.14 and mostly between 0.9 to 1.5 L/kg, respectively (Table 35.1). Apparent volume of distribution is not altered by concurrent administration of sodium valproate or enzyme-inducing anticonvulsants, and it is dose independent after a single dose in the absence of any other medication or at steady state with these other medications (26,28, 29, 30).
TABLE 35.1. MEAN PHARMACOKINETIC PARAMETER VALUES OF LAMOTRIGINE
Data from in vitro experiments show that lamotrigine is approximately 55% protein bound in human plasma at total concentrations from 1 to 10 µg/mL. The low degree of protein binding suggests that protein binding is unlikely to cause interactions between lamotrigine and other drugs. The binding of lamotrigine to plasma proteins does not change in the presence of therapeutic concentrations of phenytoin, phenobarbital, or valproate (25). These in vitro results are supported by clinical findings. Plasma protein binding of lamotrigine in patients receiving 150 to 300 mg/day of lamotrigine in the presence of a variety of concurrent medications has been reported to be about 55% (35). Although serum free fraction in a woman maintained on lamotrigine during the 3 months after labor was unchanged (34%), the free fraction in the breast-fed child decreased from 43% at birth to 32% 3 months later (36).
Little information is available on tissue distribution of lamotrigine in humans. In pediatric and young adult patients receiving adjunctive lamotrigine therapy, the cerebrospinal fluid:plasma concentration ratio is 43% (37). Therefore, the cerebrospinal concentration is likely to be similar to plasma-free concentration. The brain tissue and plasma free concentrations found in an epileptic patient undergoing a frontal topectomy at 4 hours after the last lamotrigine dose were 4.2 µg/g and 2.64 µg/mL, respectively, a finding suggesting tissue binding in the brain (38). In a woman undergoing labor, the concentrations in the umbilical cord and plasma were 4.0 and 3.3 µg/mL, respectively (39). The saliva: plasma concentration ratio has been reported to be 0.46 in healthy subjects receiving a single 120- or 240-mg dose or 0.56 in epileptic patients receiving adjunctive therapy at 150 to 300 mg/day (31,35,40). Therefore, the saliva concentration appears to be comparable to the plasma free concentration. Although the strong correlation between saliva and plasma concentrations suggests the potential to use saliva for noninvasive monitoring of the systemic concentrations of lamotrigine, data on the reliability of this approach in clinical practice are currently lacking.
Routes of Elimination
Lamotrigine is metabolized in vitro by glucuronic acid conjugation, catalyzed by UGT (10). After oral administration of 240 mg of [14C]lamotrigine (15 µCi) to six healthy subjects, 94% of the radioactivity was recovered in the urine, and 2% was recovered in the feces. The radioactivity in the urine consisted of unchanged lamotrigine (10%), a 2-N-glucuronide (76%), a 5-N-glucuronide (10%), a 2-N-methyl metabolite (0.14%), and other unidentified minor metabolites (4%) (25).
Lamotrigine is excreted in breast milk. Concentrations of 3.6 to 9.6 µg/mL were found in the milk from a woman maintained on 200 to 300 mg/day treatment during the 3 months after labor. These concentrations were between the total and free serum concentrations measured at the same times. It was estimated that 2 to 5 mg/day was excreted in the milk. The serum concentrations in the breast-fed baby were between 0.75 and 2.79 µg/mL (36). In another case, concentrations of roughly 13 µmol/L were found in the milk of a woman who was receiving 200 mg/day lamotrigine and whose plasma concentrations were 22 µmol/L. The plasma concentration in the breast-fed baby was between 5 and 6 µmol/L (39). Therefore, by breast-feeding, a mother receiving therapeutic doses of lamotrigine can pass a significant amount of the drug to the baby.
Clearance and Half-Life
Clearance after intravenous administration and apparent clearance after oral administration of a single dose have been reported to be 36.4 and typically between 30 and 40 mL/min, respectively (Table 35.1). The plasma half-life usually averages between 24 and 35 hours (Table 35.1). Plasma concentration is dose proportional, and half-life is dose independent after a single dose in the range of 30 to 400 mg (26,28).
There is conflicting information on whether lamotrigine induces its own metabolism. Oral clearance calculated in a single-dose study was almost identical to that obtained at the steady state on day 7 of a multiple-dose study (26). Another study noted a 25% decrease in half-life and a 37% increase in oral clearance on day 14 of 150 mg twice daily compared with values obtained in the same subjects after a single dose (25). Population pharmacokinetic modeling using concentrations collected in monotherapy trials revealed a small (17.3%) but statistically significant increase in oral clearance during long-term administration (41).
Comedicated Epileptic Patients
Valproate decreases oral clearance and increases the half-life of lamotrigine. Weight-normalized oral clearance averaged 0.17 to 0.20 mL/min/kg, and half-life averaged 68 to 74 hours when healthy subjects received steady-state lamotrigine doses of 50, 100, or 150 mg/day in addition to valproate treatment of 500 mg twice daily. Steady-state concentration of lamotrigine was dose proportional (29). In another study, subjects received 100-mg single doses of lamotrigine with or without six doses of 200 mg valproate given
at 8-hour intervals. Lower clearance and longer half-life were observed when lamotrigine was administered with valproate. Results from this second study should be treated with caution because valproate concentrations were not at the steady state for the entire duration of the study (21).
Lamotrigine oral clearance is higher and half-life is shorter in patients maintained on anticonvulsants known to induce liver enzymes, including carbamazepine, phenytoin, phenobarbital, and primidone. In patients receiving daily doses of 100 to 300 mg in addition to these enzyme inducers and not valproate, lamotrigine oral clearance averaged 1.20 to 1.25 mL/min/kg, and half-life averaged 12.3 to 15.1 hours. Plasma concentration was again proportional to dose (30). Oral clearance and half-life in patients receiving both enzyme inducers and sodium valproate lie between the corresponding values in the populations receiving enzyme inducers or sodium valproate alone (25,33).
The effects of sodium valproate and enzyme-inducing anticonvulsants on the oral clearance of lamotrigine identified in small studies have been confirmed in population-based pharmacokinetic analysis using plasma concentration data collected in large patient trials (42,43). In addition, methsuximide and oxcarbazepine have also been found to lower lamotrigine blood concentration (44,45).
A few data suggest that topiramate lowers lamotrigine serum concentration (46) and sertraline elevates lamotrigine blood level (47). In healthy male subjects, felbamate, at 1,200 mg twice daily, or bupropion, 150 mg twice daily, did not cause clinically relevant change in the pharmacokinetics of lamotrigine (48,49).
The pharmacokinetics of lamotrigine has been studied in pediatric patients receiving sole and adjunctive therapies. In one study (50), 12 patients aged 3 to 11 years received a 2 mg/kg single dose. Oral clearance and apparent volume of distribution averaged 16.6 mL/min (0.64 mL/min/kg) and 37.7 L (1.5 L/kg), both greater than the corresponding values in adults; and half-life averaged 32 hours, similar to the adult value. Although clearance was greater in the older (≥6 years, and heavier) patients (19.0 versus 12.6 mL/min), weight-normalized clearance was lower in the same patients (0.55 versus 0.82 mL/min/kg). This probably reflects the smaller liver size relative to body weight in the older patients. In two other studies, patients aged 10 months to 5 years or 5 to 11 years was each given a 2-mg dose while they received other anticonvulsants (25). Among the younger children, clearance averaged 3.6 or 0.47 mL/min/kg, and half-life averaged 7.7 or 45 hours in those receiving enzyme inducers or those receiving sodium valproate. Among the older children, clearance averaged 2.5 or 0.24 mL/min/kg, and half-life averaged 7.0 or 66 hours in those receiving enzyme inducers or in those receiving sodium valproate. When these results are compared with the values obtained from children receiving lamotrigine monotherapy (50), it becomes clear that the anticonvulsants carbamazepine, phenytoin, phenobarbital, and primidone, which induce lamotrigine metabolism in adults, and sodium valproate, which inhibits lamotrigine metabolism in adults, have similar effects in children. Further, as in monotherapy (50), weight-normalized clearance is higher in younger children receiving adjunctive lamotrigine therapy.
The effects of the concurrent antiepileptic therapy on lamotrigine oral clearance have been confirmed in population pharmacokinetic analyses including data from seven clinical studies in children >2 years old (51,52). The population analyses have also revealed that weight has more impact on clearance than age. In fact, when clearance is described by a linear model that adequately takes into account the weight effect, clearance is no longer a function of age in the population studied. In other words, children >2 years old with the same weight are expected to have the same clearance even if they are at different ages. Because lamotrigine is predominantly eliminated by glucuronidation, this finding is not inconsistent with the suggestions that the adult level of glucuronidation capacity is achieved by 3 years of age (53,54).
A report comparing lamotrigine pharmacokinetics in 12 elderly subjects receiving a 150-mg single dose in one study with that in 12 nonelderly adults receiving the same dose in a different study concluded that apparent clearance was 37% lower and half-life was 6.3 hours longer in the elderly (55). The mean clearance in the elderly (0.39 mL/min/kg) lies within the range of the mean clearance values (0.31 to 0.65 mL/min/kg) obtained from nine studies in nonelderly adults after single doses of 30 to 450 mg (GlaxoSmithKline,unpublished data). The similarity in pharmacokinetics between young and elderly subjects was confirmed in a population pharmacokinetic analysis performed on data from 163 subjects receiving lamotrigine monotherapy, among whom 25 were aged 65 to 76 years (41). After a single dose, apparent clearance decreased by 12% from 35 mL/min at the age of 20 years to 31 mL/min at 70 years. The corresponding decrease after 48 weeks of treatment was 10% from 41 to 37 mL/min. Therefore, the pharmacokinetics of lamotrigine in elderly patients does not differ markedly from that in young subjects.
Few data have been reported on the possible difference in lamotrigine clearance among races. Nonlinear mixed-effect pharmacokinetic modeling of the routine blood monitoring data showed that clearance was 28.7% lower in Asians (n = 5) of unspecified geographic origin than in whites (n = 158), and clearance was 25% lower in nonwhites of undocumented
ethnic origin (n = 53) than in whites (n = 464). Both findings were statistically significant. Because the clinical trial experience of lamotrigine use was gained primarily from the white population, caution should be exercised when administering the drug to patients of other races.
Pharmacokinetic differences between the sexes have been investigated by population pharmacokinetic analysis using plasma concentrations obtained in patient trials. Results of nonlinear mixed-effect modeling of pharmacokinetics conclude a lack of statistically significant or clinically relevant difference in clearance, once corrected by weight using proper functions, between men and women receiving lamotrigine monotherapy or adjunctive therapy (41,42). Linear regression showed that lamotrigine concentration was slightly (13.7%) lower in women than in men of equal weight and receiving the same dose (43). Initial assessment in children receiving adjunctive therapy revealed a statistically significant 12% difference in weight-corrected clearance between boys and girls. The difference was not confirmed in a subsequent analysis including more patients (52). Therefore weight-corrected clearance is not markedly different between male and female patients.
Limited evidence suggests an increase in lamotrigine clearance during pregnancy (39).
The pharmacokinetics of lamotrigine in subjects with renal function impairment, including those requiring hemodialysis, has been described. In one report, 10 patients with renal failure and 11 normal subjects each took a 200-mg single dose of lamotrigine (56). Although renal clearance of lamotrigine in the patients was much lower than that in the healthy persons, the difference did not cause a clinically relevant difference in total clearance because renal elimination of lamotrigine has only a small contribution to the overall elimination. Clearance of lamotrigine was similar between the two populations, averaging 0.51 mL/min/kg among the patients or 0.51 mL/min/kg among the healthy subjects. Mean half-life was 36 hours among the patients, longer than the 28-hour average among the healthy subjects, a finding reflecting a larger volume of distribution in the patients. In another study, six healthy subjects, 14 patients with moderate to severe renal impairment not requiring hemodialysis, and six patients undergoing hemodialysis each took a 100-mg single dose of lamotrigine (57). Mean clearance of 27.9 mL/min was lower and mean half-life of 50.7 hours was longer in the patients who were not undergoing dialysis compared with the corresponding values of 38.5 mL/min and 25.7 hours in the healthy controls. Mean half-lives in the dialyzed patients were 59.6 hours off dialysis and 15.5 hours during dialysis. Dialysis clearance on two occasions averaged 42.4 and 44.6 mL/min/kg. On average, 17% of the drug was removed by each 4-hour dialysis session. Although hemodialysis may be used as an effective way to reduce the body load of lamotrigine, there does not seem to be a need for a supplement dose immediately after a dialysis session conducted under the conditions used in this study. The daily therapeutic dose for a given patient requiring dialysis depends on the overall clearance, which can be influenced by the dialysis conditions. In light of the findings reported in this study, daily doses required for patients with renal dysfunction may be lower than those for patients with normal renal function.
The pharmacokinetics of lamotrigine after a single 100-mg dose was compared between 24 subjects with liver cirrhosis and 12 healthy subjects. The median oral clearance was 0.31, 0.24, or 0.10 mL/min/kg in patients with grade A, B, or C (Child-Hugh classification) hepatic impairment, respectively, compared with 0.34 mL/min/kg in the healthy controls. Median half-life was 36, 60, and 110 hours in the corresponding patient groups and 32 hours in the controls. It is advised that doses prescribed for patients with liver cirrhosis should be lower than those for patients with normal liver function (25).
Unconjugated hyperbilirubinemia (Gilbert's syndrome) is a disorder of bilirubin metabolism. It is caused by the functional impairment of UGT, which is responsible for the metabolism of bilirubin. Because lamotrigine is eliminated primarily by glucuronidation catalyzed by UGT, the pharmacokinetics of lamotrigine in patients with Gilbert's disease were compared with that in healthy subjects (58). Lamotrigine clearance was 32% lower and half-life was 37% longer in the patients with Gilbert's syndrome than in the healthy controls. The lamotrigine:glucuronide ratio in urine collected during the 168 hours after dose was 0.116 for the healthy subjects and 0.146 for the patients with Gilbert's syndrome, a finding providing mechanistic evidence that lamotrigine glucuronidation is partially impaired in the patients with Gilbert's syndrome. Treating patients with Gilbert's syndrome with lamotrigine should be done carefully.
Relationship between Serum Concentration and Dose
Lamotrigine has linear pharmacokinetics in the dose ranges tested when it is given alone, with enzyme-inducing anticonvulsants,
or with sodium valproate (26, 27, 28, 29, 30). The maximum concentration, the area under the concentration-time curve, the steady-state average concentration, and the steady-state trough concentration are all proportional to dose. This means that the values of these parameters at any given dose are readily predictable based on the values obtained at any other dose.
Relationship between Serum Concentration and Effect
Lamotrigine was predicted from animal experiments to be clinically effective in the plasma concentration range of 1.5 to 3 µg/mL (40). In this concentration range, lamotrigine has been shown to reduce photosensitivity and interictal spikes in small numbers of patients with epilepsy (34,59). A trend of dose-response correlation observed in one study employing three treatment groups (placebo, 200 to 300 mg/day, and 400 to 500 mg/day) suggested an underlining concentration-response relationship within the dose range included (60). However, the concentration-effect relationship of lamotrigine for seizure control is yet to be established.
Most of the reported efficacy trials required a target maintenance dose or a narrow range of doses, allowed flexibility on the basis of efficacy and tolerance within a dosage window, or adapted a dosing strategy to achieve a trough concentration within the range of 1 to 4 µg/mL. These trial design characteristics are not suitable for identifying any concentration-effect relationship. Attempts, with limited success, were made in many of these trials to identify the correlation between plasma concentration and the measurement of the primary efficacy end point, which is usually seizure frequency, or its change from baseline, during the maintenance phase of the trial.
In a placebo-controlled fixed-dose crossover trial (61), daily doses of lamotrigine were 400 mg for patients receiving enzyme inducers without valproate and 200 mg for patients receiving both. The mean concentration of lamotrigine was 2.2 µg/mL in responders and 2.4 µg/mL in nonresponders. In several reports, lamotrigine dose was determined individually to achieve a peak concentration of 3 µg/mL (33,62) or a trough concentration between 1.5 and 2 µg/mL (63). In all these cases, mean trough concentrations were nearly identical in the responders and the nonresponders. In three studies in which lamotrigine dose adjustment based on clinical effects was allowed, there was a lack of either a correlation between lamotrigine concentration and efficacy (64) or a difference in concentration between responders and nonresponders (65,66). Yet two other reports with dosages targeting concentration ranges of 1 to 3 µg/mL (67) or 1.2 to 2.5 µg/mL (68) showed slightly higher concentrations in the responders and significant correlation between concentration and efficacy. However, these results do not establish any causal relationship, and the narrow concentration range does not allow an adequate definition of the concentration-response relationship. Lamotrigine concentration-effect relationship for seizure control is yet to be established.
One retrospective investigation revealed a large proportion of patients being maintained on lamotrigine concentrations outside the range of 1 to 4 µg/mL and questioned the adequacy of this commonly targeted range (43). Although the exact concentration range, the concentration distribution, and the effectiveness of lamotrigine treatment in these patients were not included in the report, the data indicate potential therapeutic benefits at concentrations outside the range of 1 to 4 µg/mL. In a prospective dosetitration adjunctive therapy study (69), the median trough concentration of lamotrigine at which monthly seizure frequency fell by ≥50% in the responders was 7.9 (2.1 to 15.4) mg/L; and the median trough plasma concentration at which lamotrigine-related side effects appeared was 16 (range, 7.9 to 19.4) mg/L. The therapeutic window was therefore proposed to be 7.9 to 16 mg/L. A retrospective survey has concluded that a proposed therapeutic concentration range of 3 to 14 mg/L is widely accepted and is increasingly applied in clinical practice (70).