- Email: [email protected]
Add-on lacosamide: A retrospective study on the relationship between serum concentration, dosage, and adverse events
Epilepsy & Behavior 22 (2011) 548–551
Contents lists available at SciVerse ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Add-on lacosamide: A retrospective study on the relationship between serum concentration, dosage, and adverse events Barbara Hillenbrand, Ilona Wisniewski, Uta Jürges, Bernhard J. Steinhoff ⁎ Kork Epilepsy Centre, Kehl-Kork, Germany
a r t i c l e
i n f o
Article history: Received 11 July 2011 Revised 24 August 2011 Accepted 26 August 2011 Available online 1 October 2011 Keywords: Lacosamide Therapeutic range Serum concentration Weight-related dosage Sodium channel blockers
a b s t r a c t We performed a retrospective study in patients with poorly controlled epilepsy treated with add-on lacosamide (LCM) to investigate the relationship of LCM-related adverse events with LCM serum concentration and weightdependent dosage. We collected serum concentrations, weight-related dosages, and occurrences of the seven most frequent adverse events according to the randomized double-blind, placebo-controlled trials. Seventy of 131 patients could be sufficiently evaluated. LCM serum concentrations and weight-related dosages in patients with and without typical adverse events did not differ significantly. Closer analysis of the data suggested that dizziness as the leading adverse event occurred significantly more often if LCM was combined with classic sodium channel blockers. There was a significant correlation between LCM serum concentrations and co-medication, so there is still evidence for dependent variables that might have a relevant impact in individual cases. However, our data do not allow definition of a safety range for LCM. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Lacosamide (LCM) was officially approved in Germany for add-on treatment of adults with partial-onset seizures in 2008. The efficacy and tolerability of LCM had been evaluated in three placebo-controlled randomized, double-blind studies in adults with difficult-to-treat partial-onset seizures. In these studies, LCM maintenance dosages were 200, 400, and 600 mg/day [1], 200 and 400 mg/day [2], and 400 and 600 mg/day, respectively [3]. A twice-daily dosing regimen was used in all studies. According to a pooled analysis of all three studies [4] the most frequent adverse event in these randomized trials was dizziness, which was reported by 16, 30, and 53% of patients at doses of 200, 400, and 600 mg/day, respectively, compared with 8% for placebo. Other adverse events that showed some relationship to dose include nausea and vomiting, abnormal coordination, tremor, visual disturbances, and fatigue. Somnolence was uncommon, even at high doses. The incidence of adverse events was markedly higher during titration than during maintenance, suggesting that a slower titration than in the clinical trials may be beneficial in some instances. Lacosamide serum concentration may be measured by highperformance liquid chromatography coupled with mass spectroscopy (LC-MS) [5]. Therapeutic drug monitoring (TDM) and weight-related dosages may be useful for a rational therapeutic regimen and were used particularly with first- and second-line AEDs [6]. Although serum
concentrations may be measured for most of the new AEDs [5], in most instances, the therapeutic relevance is less well defined although therapeutic ranges are given for almost every new AED in the literature [7]. The less reliable data on new AEDs result partly from a lack of reliable investigations about a potential therapeutic range and do not necessarily mean that TDM is not helpful [6]. Lamotrigine (LTG) and topiramate (TPM) are examples where TDM appears to be helpful either in reflecting changes during hormonal contraception or pregnancy for LTG [8–10] or in defining a safety range that should be intended to avoid adverse events of TPM [11]. As the leading side effects of LCM are typical neurotoxic effects such as dizziness, we thought that there might be a similar range for this new AED. We therefore used the methodology of Fröscher et al. [11] to investigate the relationship between serum concentration and adverse events in the case of LCM as add-on-therapy. A significant correlation would allow the specification of an approximate upper limit for a “therapeutic range” of LCM serum concentrations. Furthermore, we examined whether adverse events of LCM correlate with the weight-related dosage. Recent literature indicated that add-on LCM might be better tolerated in combination with AEDs that do not act via the classic voltage- and use-dependent blockade of sodium channels [12,13]. Therefore we additionally addressed whether the mode of action of the concomitant AEDs influenced our findings. 2. Methods
⁎ Corresponding author at: Epilepsiezentrum Kork, Landstrasse 1, 77694 Kehl-Kork, Germany. Fax: + 49 7851842555. E-mail address: [email protected] (B.J. Steinhoff). 1525-5050/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2011.08.032
We retrospectively assessed the data on 131 adolescent and adult inpatients and outpatients with poorly controlled epilepsy who
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
had been treated with LCM over an approximately 2-year period (between September 2008 and December 2010) at our Epilepsy Centre. Inclusion criteria were as follows: Uptitration was undertaken at 50- or 100-mg increments weekly, serum concentrations and dosages (mg/ kg) of LCM and the concomitant anticonvulsants (AEDs) were assessed after a 6-month period (patients without adverse events) and at the time of an adverse event that had to be clearly documented and defined. For all patients without an adverse event, blood samples were taken 0.5 to 4 hours after the morning dose (many patients attending our outpatient clinic live quite a distance from the hospital); therefore, blood examinations could not be done before drug intake. For patients with an adverse event, blood samples were taken during the clinical manifestation of the side effect, which usually also occurred 0.5 to 4 hours after drug intake, but not necessarily exclusively in the morning. We listed the most frequent adverse events as dizziness, visual disturbance/diplopia, abnormal coordination, fatigue, nausea/vomiting, headache, and tremor. Age, gender, seizure types, compliance, and concomitant medications were also documented. In the case of combination therapy, serum concentrations and dosages of the following anticonvulsants were measured: LTG, levetiracetam (LEV), oxcarbazepine (OXC; in cases of OXC treatment we measured MHD, i.e., the monohydroxy derivative of OXC [14]), phenobarbital (PB), carbamazepine (CBZ), phenytoin (PHT), valproic acid (VPA), clobazam (CLB), zonisamide (ZNS), topiramate (TPM, pregabalin (PGB), sultiame (STM), rufinamide (RUF), and eslicarbazepine acetate (ESL) (as MHD is also the major metabolite of ESL, though in another composition of the enantiomers [15], we measured MHD concentrations in cases of ESL treatment). Lacosamide serum concentrations were determined by LC-MS. Serum concentrations of LEV, ZNS, TPM, PGB, and CLB were also determined by LC-MS. MHD, LTG, STM, and RUF were determined by classic high-performance liquid chromatography with UV detection (HPLC). CBZ, VPA, PHT, and PB were determined by fluorescence polarization immunoassay (FPIA) [5]. The lack of homogenicity of the data in the other cases in this retrospective analysis (in the majority of these cases, blood samples were not taken at the time of the adverse event) allowed inclusion of only the findings for 70 patients. The demographic and clinical characteristics of the patients are given in Table 1. Thirty-seven patients were treated with LCM in combination with one additional anticonvulsant, 24 patients in combination with two anticonvulsants, and 10 patients in combination with three additional anticonvulsants (Table 2). 2.1. Statistical analysis Statistical methods included descriptive statistics for specification of the total sample, and the Mann–Whitney U test for differences between patients without those with adverse events relating to LCM serum concentrations and LCM dosages.
Table 1 Demographic and clinical characteristics of patients (n = 70). Age, years Median Range Gender (M/F) Focal seizures Without impairment of consciousness With impairment of consciousness Evolving into bilateral convulsive seizure Compliance during the study (n = 70 patients)a Yes No Uncertain Mean number (range) of AEDs taken before beginning LCM therapy a
37.4 17–68 33/37 58 66 68 66 (94.3%) 0 (0%) 4 (5.7%) 9.4 (4–19)
Retrograde interview about the reliability of the intake of medication.
549
Table 2 Concomitant antiepileptic drugs at the time of the adverse event or 6 months (n = 70 patients).
Lamotrigine Levetiracetam Oxcarbazepine Phenobarbital Carbamazepine Phenytoin Valproic acid Clobazam Zonisamide Topiramate Pregabalin Sultiame Rufinamide Eslicarbazepine acetate a
n
Dose (mg)
Serum concentration (μg/mL)
22 19 17 12 10 7 7 7 5 4 2 1 1 1
455 (250–800)a 2645 (750–4000) 2100 (600–4800) 169 (50–400) 1790 (1200–2700) 311 (225–400) 1957 (1000–2900) 21.4 (15–30) 380 (200–500) 338 (150–600) 400 (300–500) 400 2000 2000
8.2 (3.1–18.2) 22.9 (1.8–62.3) 26.8 (7.2–42.2) 31.1 (12.9–44.3) 11.9 (8.6–16.0) 18.1 (7.6–30.5) 75.0 (62.6–96.1) Not determined in all cases 18.5 (12.5–18.6) 7.7 (3.3–14.9) 5.0 (2.0–7.9) 8.5 11.4 32.6
Median (range).
To examine the influence of the co-medication on the occurrence of adverse events, we performed three independent multivariate linear regressions using the binomial variable dizziness as the dependent variable and total number of anticonvulsants, number of sodium blockers, and number of other types of drugs as categorical predictors. Three logistic linear regression models were also made, again using “dizziness” as the binomial dependent variable, with total number of anticonvulsants (model 1), number of sodium blockers (CBZ, OXC, ESL, PHT, and LTG) (model 2), and number of other types of drugs (model 3) as categorical predictors, and serum concentrations and dosages of LEV, PB, VPA, CLB, ZNS, TPM, PGB, STM, and RUF as continuous variables. 3. Results Thirty-two of 70 patients (46%) experienced one or more adverse events (Table 3). The LCM serum concentration and the LCM weightrelated dosage at the time of these adverse events were compared with the corresponding values of patients who had experienced no adverse event over a 6-month period. LCM serum concentrations and LCM weight-dependent dosages (mg/kg) did not differ significantly between patients without an adverse event and patients with typical clinical adverse events (Table 4). As described above we performed three independent multivariate linear regressions to examine the influence of the co-medication on the occurrence of adverse events. Intake of classic sodium blockers was associated with a significant P value (0.03) for the occurrence of “dizziness” (Table 5). Each model—“anticonvulsant medications excluding sodium channel blockers,” “sodium channel blockers only,” and “all anticonvulsants”— was compared with the dosage (mg/kg) and serum concentration of LCM. In all subgroups, P values were significant for serum concentration
Table 3 Adverse events in patients treated with lacosamide. Adverse event
n
% of 70 patients
Dizziness Visual disturbance/diplopia Abnormal coordination Fatigue Nausea/vomiting Headache Tremor
17 14 5 4 4 4 1
24 20 7 6 6 6 1
550
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
Table 4 Lacosamide serum concentration at first observation of an adverse event: Comparison with LCM serum concentrations of 38 patients without adverse events after 6 months. Adverse event
Dizziness Visual disturbance/diplopia Abnormal coordination Fatigue Nausea/vomiting Headache Tremor
Patients with adverse events
Patients without adverse events
n
Serum concentration (μg/mL)
n
Serum concentration (μg/mL)
17 14 5 4 4 4 1
6.7 ± 4.3 5.98 ± 4.67 6.24 ± 2.6 6.24 ± 2.67 6.7 ± 4.3 4.45 ± 1.38 5
38 38 38 38 38 38 38
5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2
(P =0.01, 0.02, and 0.01, respectively). Additionally, in subgroup 1 (“anticonvulsant medications excluding sodium channel blockers”), there was a significant effect for LCM dosage as well (P = 0.04). Furthermore, there was a trend toward a dosage effect in subgroup 3 (“all anticonvulsants”). Confidence intervals (CIs) and Wald statistics for each test are also listed (Table 6). 4. Discussion Lacosamide serum concentrations and weight-dependent dosages (mg/kg) did not differ significantly between patients without an adverse event and patients with typical clinical adverse events (dizziness, visual disturbance/diplopia, abnormal coordination, nausea/vomiting, headache, tremor). This means that there was no correlation between serum concentration levels or weight-related dosages and clinical tolerability of LCM. Therefore, our data do not allow definition of a safety range for LCM. No previous studies have investigated this relationship. The “therapeutic” range usually reflects both efficacy and tolerability and defines serum concentrations or weight-related dosages that correspond to a high probability of efficacy and good tolerability [7]. We addressed only the tolerability aspect and followed the approach of Fröscher et al. [11], who used this methodology for TPM. Certainly it is more important to establish such a safety corridor to avoid toxic serum levels or dosages. As we were unable to find such a range it is not very probable that a therapeutic range in terms of efficacy can be defined. The latter is less important because it would not be customary to increase a dosage to a so-called therapeutic level if a patient was already free of seizures. Whereas Fröscher et al. [11] were able to define a serum concentration of TPM that was clearly associated with an elevated risk of typical adverse events, we could not identify such a level for LCM, which means that the interindividual variability of the correspondence between serum concentrations and clinical effects is much wider. To examine the influence of the co-medication on the occurrence of adverse events, we performed three independent multivariate linear regressions. LCM acts on sodium channels too, but its effect differs from that of classic blockers of voltage- and use-dependent sodium channels, because LCM selectively enhances the slow inactivation of voltagegated sodium channels without affecting fast inactivation [4,16,17]. Interestingly we could show a positive correlation between occurrences of the most frequent adverse event “dizziness” and classic sodium
Table 5 Results of the linear regression analysis to examine the influence of the co-medication on “dizziness”. Co-medication
SE
P
Confidence interval − 95%
+ 95%
Sodium blocker Other Total
0.578481 0.434352 0.373773
0.031793 0.831764 0.159474
− 2.37580 − 0.94359 − 0.20674
− 0.108200 0.759042 1.258422
Mann–Whitney U
P
262 254 69 61 61 68 Not available
0.27 0.80 0.32 0.52 0.52 0.73
channel blockers which is in line with recent post hoc analyses of the pivotal regulatory trials [12] and postmarketing observations [13], although not every postmarketing study confirmed such a correlation [18]. Whether or not this reflects a synergistic (negative) effect remains a matter of speculation. One has to remember that at the least, combinations of classic sodium channel blockers with sedating potential, like CBZ and OXC, may be associated with an elevated risk of dizziness in combination with AEDs other than LCM and, especially, with AEDs with different modes of action. Still this finding encourages further investigation of the hypothesis of rational polytherapy. Is therapeutic drug monitoring of LCM meaningless? We do not think so. Although our data indicate that individual LCM serum concentrations do not correspond to a predictable risk of adverse events, the measurement of LCM levels may nevertheless be helpful in certain clinical circumstances. Although we did not observe statistically significant correlations between serum concentrations/dose-related dosages and the leading adverse events with add-on LCM therapy, except for dizziness with the combination of LCM and classic sodium channel blockers (Table 5), a different statistical approach revealed statistically significant or at least close-to-significant relationships between serum concentrations and, to a lesser extent, dose-related dosages and the medications (Table 6). One example of the importance of serum levels is LTG. A therapeutic range has been proposed [7]. However, it is clinically less relevant than the measurement of individual serum levels of LTG during hormonal contraception or pregnancy [8–10]. Furthermore, because of the synergistic effect of VPA and LTG, therapeutic drug monitoring may be very helpful if this combination is established or discontinued because of adverse events that result from the complex interaction of these drugs, especially the VPA-induced increase in serum LTG concentration [19]. We cannot foresee whether the clinical circumstances will be similar for LCM, but one could imagine that this might occur if LCM is combined with enzyme-inducing AEDs. In vitro data generally suggest that LCM itself has a low interaction potential. The data indicate that the enzymes CYP1A2, CYP2B6, and CYP2C9 are not induced and that CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2D6, and CYP2E1 are not inhibited by LCM at plasma concentrations observed in clinical trials. LCM seems not to inhibit or induce CYP2C19 and 3A4 enzymes to a clinically relevant extent. It has been found not to influence the serum concentrations of CBZ, CBZ-10,11-epoxide, PHT, VPA, LTG, MHD, TPM, ZNS, LEV, GBP, metformin, digoxin, omeprazole, ethinylestradiol, and levonorgestrel [4]. However, there might be clinically relevant interactions vice versa that might require therapeutic drug monitoring of LCM: Strong inhibitors of CYP2C9 (e.g., fluconazole) and CYP3A4 (e.g., itraconazole, ketoconazole, ritonavir, clarithromycin) may potentially lead to increased systemic exposure to LCM [4]. Other pharmacokinetic studies have shown that concomitant administration of enzyme-inducing AEDs is associated with an approximate 25% reduction in serum LCM levels, although in a formal pharmacokinetic study, LCM kinetics were unaffected by co-administration with CBZ at a daily dosage of 400 mg [4]. These data may hint at clinical situations in which therapeutic drug monitoring may be helpful in understanding clinical changes and reacting accordingly.
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
551
Table 6 Linear regression models for three medication subgroups: (1) Anticonvulsant medication excluding sodium channel blockers, (2) sodium channel blockers only, and (3) all anticonvulsants. Level of effect
Intercept mg/kg LCM Serum concentration of LCM Number of others Number of others Number of others Intercept mg/kg LCM Serum concentration of LCM Number of sodium channel blockers Number of sodium channel blockers Number of sodium channel blockers Intercept mg/kg LCM Serum concentration of LCM All anticonvulsants All anticonvulsants All anticonvulsants
SE
df
P
1. Anticonvulsant medications excluding sodium channel blockers 0.906848 1 0.344147 0.269576 1 0.044431 0.149962 1 0.012210 2 0.768201 0 0.447014 0.733532 1 0.446444 0.510076 2. Sodium channel blockers 1.056217 1 0.191866 0.278643 1 0.142276 0.148906 1 0.025070 2 0.331367 0 0.778883 0.115893 1 0.507083 0.433910 3. All anticonvulsants 0.907978 1 0.348766 0.268751 1 0.067094 0.348220 1 5.953519 0.014688 2 0.232276 0 0.424827 0.733532 1 0.494912 0.510076
References [1] Ben-Menachem E, Biton V, Jatuzis D, Abou-Khalil B, Doty P, Rudd GD. Efficacy and safety of oral lacosamide as adjunctive therapy in adults with partial-onset seizures. Epilepsia 2007;48:1308–17. [2] Chung S, Sperling MR, Biton V, et al. Lacosamide as adjunctive therapy for partialonset seizures: a randomized controlled trial. Epilepsia 2010;51:958–67. [3] Halász P, Kälviäinen R, Mazurkiewicz-Bedzinska M, et al. Adjunctive lacosamide for partial-onset seizures: efficacy and safety results from a randomized controlled trial. Epilepsia 2009;50:443–53. [4] Curia G, Biagini G, Perucca E, Avoli M. Lacosamide: a new approach to target voltage-gated sodium currents in epileptic disorders. CNS Drugs 2009;23:555–68. [5] Jürges U, Steinhoff BJ. Therapeutic drug monitoring of antiepileptic drugs. Z Epileptol 2011;24:26–33. [6] Steinhoff BJ, Fröscher W. Importance of therapeutic drug monitoring of anticonvulsants: an overview. Z Epileptol 2011;24:6–11. [7] Patsalos PN, Berry DJ, Bourgeois BF, et al. Antiepileptic drugs: best practice guidelines for therapeutic drug monitoring—a position paper by the subcommission on therapeutic drug monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia 2008;49:1239–76. [8] Sabers A. Pharmacokinetic interactions between contraceptives and antiepileptic drugs. Seizure 2008;17:141–4. [9] Sabers A, Ohman I, Christensen J, Tomson T. Oral contraceptives reduce lamotrigine plasma levels. Neurology 2003;26:570–1.
Confidence interval − 95%
+ 95%
− 2.63527 − 1.07021 0.08189
0.919508 − 0.013488 0.669725
− 0.72395 − 1.16910
1.028308 0.580933
− 3.44859 − 0.95500 0.04175
0.691702 0.137259 0.625449
− 2.66789 − 0.73226
0.385273 1.255463
− 2.63037 − 1.01884 0.627935
0.928842 0.034647
− 0.93946 − 1.71095
0.725830 0.229071
Odds ratio
N/A
2.38
3.61
[10] Tran TA, Leppik IE, Blesi K, Sathanandan ST, Remmel R. Lamotrigine clearance during pregnancy. Neurology 2002;59:251–5. [11] Fröscher W, Schier KR, Hoffmann M, et al. Topiramate: a prospective study on the relationship between concentration, dosage and adverse events in epileptic patients on combination therapy. Epileptic Disord 2005;7:237–48. [12] Sake J-K, Hebert D, Isojärvi J, et al. A pooled analysis of lacosamide clinical trial data grouped by mechanisms of action of concomitant antiepileptic drugs. CNS Drugs 2010;24:1055–68. [13] Novy J, Patsalos PN, Sander JW, Sisodiya SM. Lacosamide neurotoxicity associated with concomitant use of sodium channel-blocking antiepileptic drugs: a pharmacodynamic interaction? Epilepsy Behav 2011;20:20–3. [14] Steinhoff BJ. Oxcarbazepine extended-release formulation in epilepsy. Expert Rev Clin Pharmacol 2009;2:155–62. [15] Almeida L, Soares-da-Silva P. Eslicarbazepine acetate (BIA 2–093). Neurotherapeutics 2007;4:88–96. [16] Beyreuther BK, Freitag J, Heers C, Krebsfänger N, Scharfenecker U, Stöhr T. Lacosamide: a review of preclinical properties. CNS Drug Rev 2007;13:21–42. [17] Errington AC, Stöhr T, Heers C, Lees G. The investigational anticonvulsant lacosamide selectively enhances slow inactivation of voltage-gated sodium channels. Mol Pharmacol 2008;73:157–69. [18] Stefan H, Kerling F, Steinhoff BJ. First clinical experiences with lacosamide treatment. Z Epileptol 2010;23:98–103. [19] Steinhoff BJ. How to replace lamotrigine with valproate. Epilepsia 2006;47: 1943–4.
Contents lists available at SciVerse ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Add-on lacosamide: A retrospective study on the relationship between serum concentration, dosage, and adverse events Barbara Hillenbrand, Ilona Wisniewski, Uta Jürges, Bernhard J. Steinhoff ⁎ Kork Epilepsy Centre, Kehl-Kork, Germany
a r t i c l e
i n f o
Article history: Received 11 July 2011 Revised 24 August 2011 Accepted 26 August 2011 Available online 1 October 2011 Keywords: Lacosamide Therapeutic range Serum concentration Weight-related dosage Sodium channel blockers
a b s t r a c t We performed a retrospective study in patients with poorly controlled epilepsy treated with add-on lacosamide (LCM) to investigate the relationship of LCM-related adverse events with LCM serum concentration and weightdependent dosage. We collected serum concentrations, weight-related dosages, and occurrences of the seven most frequent adverse events according to the randomized double-blind, placebo-controlled trials. Seventy of 131 patients could be sufficiently evaluated. LCM serum concentrations and weight-related dosages in patients with and without typical adverse events did not differ significantly. Closer analysis of the data suggested that dizziness as the leading adverse event occurred significantly more often if LCM was combined with classic sodium channel blockers. There was a significant correlation between LCM serum concentrations and co-medication, so there is still evidence for dependent variables that might have a relevant impact in individual cases. However, our data do not allow definition of a safety range for LCM. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Lacosamide (LCM) was officially approved in Germany for add-on treatment of adults with partial-onset seizures in 2008. The efficacy and tolerability of LCM had been evaluated in three placebo-controlled randomized, double-blind studies in adults with difficult-to-treat partial-onset seizures. In these studies, LCM maintenance dosages were 200, 400, and 600 mg/day [1], 200 and 400 mg/day [2], and 400 and 600 mg/day, respectively [3]. A twice-daily dosing regimen was used in all studies. According to a pooled analysis of all three studies [4] the most frequent adverse event in these randomized trials was dizziness, which was reported by 16, 30, and 53% of patients at doses of 200, 400, and 600 mg/day, respectively, compared with 8% for placebo. Other adverse events that showed some relationship to dose include nausea and vomiting, abnormal coordination, tremor, visual disturbances, and fatigue. Somnolence was uncommon, even at high doses. The incidence of adverse events was markedly higher during titration than during maintenance, suggesting that a slower titration than in the clinical trials may be beneficial in some instances. Lacosamide serum concentration may be measured by highperformance liquid chromatography coupled with mass spectroscopy (LC-MS) [5]. Therapeutic drug monitoring (TDM) and weight-related dosages may be useful for a rational therapeutic regimen and were used particularly with first- and second-line AEDs [6]. Although serum
concentrations may be measured for most of the new AEDs [5], in most instances, the therapeutic relevance is less well defined although therapeutic ranges are given for almost every new AED in the literature [7]. The less reliable data on new AEDs result partly from a lack of reliable investigations about a potential therapeutic range and do not necessarily mean that TDM is not helpful [6]. Lamotrigine (LTG) and topiramate (TPM) are examples where TDM appears to be helpful either in reflecting changes during hormonal contraception or pregnancy for LTG [8–10] or in defining a safety range that should be intended to avoid adverse events of TPM [11]. As the leading side effects of LCM are typical neurotoxic effects such as dizziness, we thought that there might be a similar range for this new AED. We therefore used the methodology of Fröscher et al. [11] to investigate the relationship between serum concentration and adverse events in the case of LCM as add-on-therapy. A significant correlation would allow the specification of an approximate upper limit for a “therapeutic range” of LCM serum concentrations. Furthermore, we examined whether adverse events of LCM correlate with the weight-related dosage. Recent literature indicated that add-on LCM might be better tolerated in combination with AEDs that do not act via the classic voltage- and use-dependent blockade of sodium channels [12,13]. Therefore we additionally addressed whether the mode of action of the concomitant AEDs influenced our findings. 2. Methods
⁎ Corresponding author at: Epilepsiezentrum Kork, Landstrasse 1, 77694 Kehl-Kork, Germany. Fax: + 49 7851842555. E-mail address: [email protected] (B.J. Steinhoff). 1525-5050/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.yebeh.2011.08.032
We retrospectively assessed the data on 131 adolescent and adult inpatients and outpatients with poorly controlled epilepsy who
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
had been treated with LCM over an approximately 2-year period (between September 2008 and December 2010) at our Epilepsy Centre. Inclusion criteria were as follows: Uptitration was undertaken at 50- or 100-mg increments weekly, serum concentrations and dosages (mg/ kg) of LCM and the concomitant anticonvulsants (AEDs) were assessed after a 6-month period (patients without adverse events) and at the time of an adverse event that had to be clearly documented and defined. For all patients without an adverse event, blood samples were taken 0.5 to 4 hours after the morning dose (many patients attending our outpatient clinic live quite a distance from the hospital); therefore, blood examinations could not be done before drug intake. For patients with an adverse event, blood samples were taken during the clinical manifestation of the side effect, which usually also occurred 0.5 to 4 hours after drug intake, but not necessarily exclusively in the morning. We listed the most frequent adverse events as dizziness, visual disturbance/diplopia, abnormal coordination, fatigue, nausea/vomiting, headache, and tremor. Age, gender, seizure types, compliance, and concomitant medications were also documented. In the case of combination therapy, serum concentrations and dosages of the following anticonvulsants were measured: LTG, levetiracetam (LEV), oxcarbazepine (OXC; in cases of OXC treatment we measured MHD, i.e., the monohydroxy derivative of OXC [14]), phenobarbital (PB), carbamazepine (CBZ), phenytoin (PHT), valproic acid (VPA), clobazam (CLB), zonisamide (ZNS), topiramate (TPM, pregabalin (PGB), sultiame (STM), rufinamide (RUF), and eslicarbazepine acetate (ESL) (as MHD is also the major metabolite of ESL, though in another composition of the enantiomers [15], we measured MHD concentrations in cases of ESL treatment). Lacosamide serum concentrations were determined by LC-MS. Serum concentrations of LEV, ZNS, TPM, PGB, and CLB were also determined by LC-MS. MHD, LTG, STM, and RUF were determined by classic high-performance liquid chromatography with UV detection (HPLC). CBZ, VPA, PHT, and PB were determined by fluorescence polarization immunoassay (FPIA) [5]. The lack of homogenicity of the data in the other cases in this retrospective analysis (in the majority of these cases, blood samples were not taken at the time of the adverse event) allowed inclusion of only the findings for 70 patients. The demographic and clinical characteristics of the patients are given in Table 1. Thirty-seven patients were treated with LCM in combination with one additional anticonvulsant, 24 patients in combination with two anticonvulsants, and 10 patients in combination with three additional anticonvulsants (Table 2). 2.1. Statistical analysis Statistical methods included descriptive statistics for specification of the total sample, and the Mann–Whitney U test for differences between patients without those with adverse events relating to LCM serum concentrations and LCM dosages.
Table 1 Demographic and clinical characteristics of patients (n = 70). Age, years Median Range Gender (M/F) Focal seizures Without impairment of consciousness With impairment of consciousness Evolving into bilateral convulsive seizure Compliance during the study (n = 70 patients)a Yes No Uncertain Mean number (range) of AEDs taken before beginning LCM therapy a
37.4 17–68 33/37 58 66 68 66 (94.3%) 0 (0%) 4 (5.7%) 9.4 (4–19)
Retrograde interview about the reliability of the intake of medication.
549
Table 2 Concomitant antiepileptic drugs at the time of the adverse event or 6 months (n = 70 patients).
Lamotrigine Levetiracetam Oxcarbazepine Phenobarbital Carbamazepine Phenytoin Valproic acid Clobazam Zonisamide Topiramate Pregabalin Sultiame Rufinamide Eslicarbazepine acetate a
n
Dose (mg)
Serum concentration (μg/mL)
22 19 17 12 10 7 7 7 5 4 2 1 1 1
455 (250–800)a 2645 (750–4000) 2100 (600–4800) 169 (50–400) 1790 (1200–2700) 311 (225–400) 1957 (1000–2900) 21.4 (15–30) 380 (200–500) 338 (150–600) 400 (300–500) 400 2000 2000
8.2 (3.1–18.2) 22.9 (1.8–62.3) 26.8 (7.2–42.2) 31.1 (12.9–44.3) 11.9 (8.6–16.0) 18.1 (7.6–30.5) 75.0 (62.6–96.1) Not determined in all cases 18.5 (12.5–18.6) 7.7 (3.3–14.9) 5.0 (2.0–7.9) 8.5 11.4 32.6
Median (range).
To examine the influence of the co-medication on the occurrence of adverse events, we performed three independent multivariate linear regressions using the binomial variable dizziness as the dependent variable and total number of anticonvulsants, number of sodium blockers, and number of other types of drugs as categorical predictors. Three logistic linear regression models were also made, again using “dizziness” as the binomial dependent variable, with total number of anticonvulsants (model 1), number of sodium blockers (CBZ, OXC, ESL, PHT, and LTG) (model 2), and number of other types of drugs (model 3) as categorical predictors, and serum concentrations and dosages of LEV, PB, VPA, CLB, ZNS, TPM, PGB, STM, and RUF as continuous variables. 3. Results Thirty-two of 70 patients (46%) experienced one or more adverse events (Table 3). The LCM serum concentration and the LCM weightrelated dosage at the time of these adverse events were compared with the corresponding values of patients who had experienced no adverse event over a 6-month period. LCM serum concentrations and LCM weight-dependent dosages (mg/kg) did not differ significantly between patients without an adverse event and patients with typical clinical adverse events (Table 4). As described above we performed three independent multivariate linear regressions to examine the influence of the co-medication on the occurrence of adverse events. Intake of classic sodium blockers was associated with a significant P value (0.03) for the occurrence of “dizziness” (Table 5). Each model—“anticonvulsant medications excluding sodium channel blockers,” “sodium channel blockers only,” and “all anticonvulsants”— was compared with the dosage (mg/kg) and serum concentration of LCM. In all subgroups, P values were significant for serum concentration
Table 3 Adverse events in patients treated with lacosamide. Adverse event
n
% of 70 patients
Dizziness Visual disturbance/diplopia Abnormal coordination Fatigue Nausea/vomiting Headache Tremor
17 14 5 4 4 4 1
24 20 7 6 6 6 1
550
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
Table 4 Lacosamide serum concentration at first observation of an adverse event: Comparison with LCM serum concentrations of 38 patients without adverse events after 6 months. Adverse event
Dizziness Visual disturbance/diplopia Abnormal coordination Fatigue Nausea/vomiting Headache Tremor
Patients with adverse events
Patients without adverse events
n
Serum concentration (μg/mL)
n
Serum concentration (μg/mL)
17 14 5 4 4 4 1
6.7 ± 4.3 5.98 ± 4.67 6.24 ± 2.6 6.24 ± 2.67 6.7 ± 4.3 4.45 ± 1.38 5
38 38 38 38 38 38 38
5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2 5.12 ± 2.2
(P =0.01, 0.02, and 0.01, respectively). Additionally, in subgroup 1 (“anticonvulsant medications excluding sodium channel blockers”), there was a significant effect for LCM dosage as well (P = 0.04). Furthermore, there was a trend toward a dosage effect in subgroup 3 (“all anticonvulsants”). Confidence intervals (CIs) and Wald statistics for each test are also listed (Table 6). 4. Discussion Lacosamide serum concentrations and weight-dependent dosages (mg/kg) did not differ significantly between patients without an adverse event and patients with typical clinical adverse events (dizziness, visual disturbance/diplopia, abnormal coordination, nausea/vomiting, headache, tremor). This means that there was no correlation between serum concentration levels or weight-related dosages and clinical tolerability of LCM. Therefore, our data do not allow definition of a safety range for LCM. No previous studies have investigated this relationship. The “therapeutic” range usually reflects both efficacy and tolerability and defines serum concentrations or weight-related dosages that correspond to a high probability of efficacy and good tolerability [7]. We addressed only the tolerability aspect and followed the approach of Fröscher et al. [11], who used this methodology for TPM. Certainly it is more important to establish such a safety corridor to avoid toxic serum levels or dosages. As we were unable to find such a range it is not very probable that a therapeutic range in terms of efficacy can be defined. The latter is less important because it would not be customary to increase a dosage to a so-called therapeutic level if a patient was already free of seizures. Whereas Fröscher et al. [11] were able to define a serum concentration of TPM that was clearly associated with an elevated risk of typical adverse events, we could not identify such a level for LCM, which means that the interindividual variability of the correspondence between serum concentrations and clinical effects is much wider. To examine the influence of the co-medication on the occurrence of adverse events, we performed three independent multivariate linear regressions. LCM acts on sodium channels too, but its effect differs from that of classic blockers of voltage- and use-dependent sodium channels, because LCM selectively enhances the slow inactivation of voltagegated sodium channels without affecting fast inactivation [4,16,17]. Interestingly we could show a positive correlation between occurrences of the most frequent adverse event “dizziness” and classic sodium
Table 5 Results of the linear regression analysis to examine the influence of the co-medication on “dizziness”. Co-medication
SE
P
Confidence interval − 95%
+ 95%
Sodium blocker Other Total
0.578481 0.434352 0.373773
0.031793 0.831764 0.159474
− 2.37580 − 0.94359 − 0.20674
− 0.108200 0.759042 1.258422
Mann–Whitney U
P
262 254 69 61 61 68 Not available
0.27 0.80 0.32 0.52 0.52 0.73
channel blockers which is in line with recent post hoc analyses of the pivotal regulatory trials [12] and postmarketing observations [13], although not every postmarketing study confirmed such a correlation [18]. Whether or not this reflects a synergistic (negative) effect remains a matter of speculation. One has to remember that at the least, combinations of classic sodium channel blockers with sedating potential, like CBZ and OXC, may be associated with an elevated risk of dizziness in combination with AEDs other than LCM and, especially, with AEDs with different modes of action. Still this finding encourages further investigation of the hypothesis of rational polytherapy. Is therapeutic drug monitoring of LCM meaningless? We do not think so. Although our data indicate that individual LCM serum concentrations do not correspond to a predictable risk of adverse events, the measurement of LCM levels may nevertheless be helpful in certain clinical circumstances. Although we did not observe statistically significant correlations between serum concentrations/dose-related dosages and the leading adverse events with add-on LCM therapy, except for dizziness with the combination of LCM and classic sodium channel blockers (Table 5), a different statistical approach revealed statistically significant or at least close-to-significant relationships between serum concentrations and, to a lesser extent, dose-related dosages and the medications (Table 6). One example of the importance of serum levels is LTG. A therapeutic range has been proposed [7]. However, it is clinically less relevant than the measurement of individual serum levels of LTG during hormonal contraception or pregnancy [8–10]. Furthermore, because of the synergistic effect of VPA and LTG, therapeutic drug monitoring may be very helpful if this combination is established or discontinued because of adverse events that result from the complex interaction of these drugs, especially the VPA-induced increase in serum LTG concentration [19]. We cannot foresee whether the clinical circumstances will be similar for LCM, but one could imagine that this might occur if LCM is combined with enzyme-inducing AEDs. In vitro data generally suggest that LCM itself has a low interaction potential. The data indicate that the enzymes CYP1A2, CYP2B6, and CYP2C9 are not induced and that CYP1A1, CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2D6, and CYP2E1 are not inhibited by LCM at plasma concentrations observed in clinical trials. LCM seems not to inhibit or induce CYP2C19 and 3A4 enzymes to a clinically relevant extent. It has been found not to influence the serum concentrations of CBZ, CBZ-10,11-epoxide, PHT, VPA, LTG, MHD, TPM, ZNS, LEV, GBP, metformin, digoxin, omeprazole, ethinylestradiol, and levonorgestrel [4]. However, there might be clinically relevant interactions vice versa that might require therapeutic drug monitoring of LCM: Strong inhibitors of CYP2C9 (e.g., fluconazole) and CYP3A4 (e.g., itraconazole, ketoconazole, ritonavir, clarithromycin) may potentially lead to increased systemic exposure to LCM [4]. Other pharmacokinetic studies have shown that concomitant administration of enzyme-inducing AEDs is associated with an approximate 25% reduction in serum LCM levels, although in a formal pharmacokinetic study, LCM kinetics were unaffected by co-administration with CBZ at a daily dosage of 400 mg [4]. These data may hint at clinical situations in which therapeutic drug monitoring may be helpful in understanding clinical changes and reacting accordingly.
B. Hillenbrand et al. / Epilepsy & Behavior 22 (2011) 548–551
551
Table 6 Linear regression models for three medication subgroups: (1) Anticonvulsant medication excluding sodium channel blockers, (2) sodium channel blockers only, and (3) all anticonvulsants. Level of effect
Intercept mg/kg LCM Serum concentration of LCM Number of others Number of others Number of others Intercept mg/kg LCM Serum concentration of LCM Number of sodium channel blockers Number of sodium channel blockers Number of sodium channel blockers Intercept mg/kg LCM Serum concentration of LCM All anticonvulsants All anticonvulsants All anticonvulsants
SE
df
P
1. Anticonvulsant medications excluding sodium channel blockers 0.906848 1 0.344147 0.269576 1 0.044431 0.149962 1 0.012210 2 0.768201 0 0.447014 0.733532 1 0.446444 0.510076 2. Sodium channel blockers 1.056217 1 0.191866 0.278643 1 0.142276 0.148906 1 0.025070 2 0.331367 0 0.778883 0.115893 1 0.507083 0.433910 3. All anticonvulsants 0.907978 1 0.348766 0.268751 1 0.067094 0.348220 1 5.953519 0.014688 2 0.232276 0 0.424827 0.733532 1 0.494912 0.510076
References [1] Ben-Menachem E, Biton V, Jatuzis D, Abou-Khalil B, Doty P, Rudd GD. Efficacy and safety of oral lacosamide as adjunctive therapy in adults with partial-onset seizures. Epilepsia 2007;48:1308–17. [2] Chung S, Sperling MR, Biton V, et al. Lacosamide as adjunctive therapy for partialonset seizures: a randomized controlled trial. Epilepsia 2010;51:958–67. [3] Halász P, Kälviäinen R, Mazurkiewicz-Bedzinska M, et al. Adjunctive lacosamide for partial-onset seizures: efficacy and safety results from a randomized controlled trial. Epilepsia 2009;50:443–53. [4] Curia G, Biagini G, Perucca E, Avoli M. Lacosamide: a new approach to target voltage-gated sodium currents in epileptic disorders. CNS Drugs 2009;23:555–68. [5] Jürges U, Steinhoff BJ. Therapeutic drug monitoring of antiepileptic drugs. Z Epileptol 2011;24:26–33. [6] Steinhoff BJ, Fröscher W. Importance of therapeutic drug monitoring of anticonvulsants: an overview. Z Epileptol 2011;24:6–11. [7] Patsalos PN, Berry DJ, Bourgeois BF, et al. Antiepileptic drugs: best practice guidelines for therapeutic drug monitoring—a position paper by the subcommission on therapeutic drug monitoring, ILAE Commission on Therapeutic Strategies. Epilepsia 2008;49:1239–76. [8] Sabers A. Pharmacokinetic interactions between contraceptives and antiepileptic drugs. Seizure 2008;17:141–4. [9] Sabers A, Ohman I, Christensen J, Tomson T. Oral contraceptives reduce lamotrigine plasma levels. Neurology 2003;26:570–1.
Confidence interval − 95%
+ 95%
− 2.63527 − 1.07021 0.08189
0.919508 − 0.013488 0.669725
− 0.72395 − 1.16910
1.028308 0.580933
− 3.44859 − 0.95500 0.04175
0.691702 0.137259 0.625449
− 2.66789 − 0.73226
0.385273 1.255463
− 2.63037 − 1.01884 0.627935
0.928842 0.034647
− 0.93946 − 1.71095
0.725830 0.229071
Odds ratio
N/A
2.38
3.61
[10] Tran TA, Leppik IE, Blesi K, Sathanandan ST, Remmel R. Lamotrigine clearance during pregnancy. Neurology 2002;59:251–5. [11] Fröscher W, Schier KR, Hoffmann M, et al. Topiramate: a prospective study on the relationship between concentration, dosage and adverse events in epileptic patients on combination therapy. Epileptic Disord 2005;7:237–48. [12] Sake J-K, Hebert D, Isojärvi J, et al. A pooled analysis of lacosamide clinical trial data grouped by mechanisms of action of concomitant antiepileptic drugs. CNS Drugs 2010;24:1055–68. [13] Novy J, Patsalos PN, Sander JW, Sisodiya SM. Lacosamide neurotoxicity associated with concomitant use of sodium channel-blocking antiepileptic drugs: a pharmacodynamic interaction? Epilepsy Behav 2011;20:20–3. [14] Steinhoff BJ. Oxcarbazepine extended-release formulation in epilepsy. Expert Rev Clin Pharmacol 2009;2:155–62. [15] Almeida L, Soares-da-Silva P. Eslicarbazepine acetate (BIA 2–093). Neurotherapeutics 2007;4:88–96. [16] Beyreuther BK, Freitag J, Heers C, Krebsfänger N, Scharfenecker U, Stöhr T. Lacosamide: a review of preclinical properties. CNS Drug Rev 2007;13:21–42. [17] Errington AC, Stöhr T, Heers C, Lees G. The investigational anticonvulsant lacosamide selectively enhances slow inactivation of voltage-gated sodium channels. Mol Pharmacol 2008;73:157–69. [18] Stefan H, Kerling F, Steinhoff BJ. First clinical experiences with lacosamide treatment. Z Epileptol 2010;23:98–103. [19] Steinhoff BJ. How to replace lamotrigine with valproate. Epilepsia 2006;47: 1943–4.