Antiseizure medication choice in diffuse glioma: A single-center population-based experience
Abstract
Objective
Several antiseizure medications (ASMs) can be considered as first-line treatment for glioma-associated epilepsy (GAE). We aimed to find out, if there are subgroups of glioma patients which may benefit from different first-line ASMs.
Methods
All diffuse glioma (grade 2–4) patients, who were in contact with Helsinki University Hospital Neurology or Oncology Departments during 2013–2015, were recognized from medical records. Follow-up data was retrospectively collected of all GAE patients living in Helsinki for 5 years from diagnosis of glioma, or until death.
Results
A total of 121 patients with GAE fulfilling inclusion criteria were identified. Forty-nine (40.5%) had grade 2 glioma, 18 (14.9%) had grade 3 glioma, and 54 (44.6%) had grade 4 glioma. The most common first ASM was oxcarbazepine (39.2%), followed by levetiracetam (22.5%), carbamazepine (19.3%), phenytoin (10.8%) and valproic acid (8.3%). The first ASM was retained in use in 70 (57.9%) patients until the end of follow-up. Among patients with low-grade glioma, oxcarbazepine was favored, whereas grade 4 glioma patients more often received levetiracetam as first ASM. At the end of follow-up, the retention rate of levetiracetam as first ASM was higher than retention rate of other ASMs in grade 4 glioma patients (p 0.002). Patients who initiated valproic acid as first ASM underwent more ASM changes than patients who initiated other first-line ASM (p 0.005).
Significance
Levetiracetam seems to be a favorable first-line treatment for GAE, especially for patients with grade 4 glioma. For grade 2–3 glioma patients, oxcarbazepine may be a reasonable option.
Plain Language Summary
Grade 2–4 gliomas are central nervous system tumors originating from glial cells, and epilepsy is common in glioma patients. We found that levetiracetam was favored as first antiseizure medication for patients with malignant, grade 4, glioma. For patients with slower-growing tumors, grade 2–3 gliomas, oxcarbazepine may be a reasonable choice for first antiseizure medication. The findings highlight the need for individual assessment in the use of antiseizure medications in glioma patients.
Key points
- The first antiseizure medication was retained in use in 57.9% of the patients until the end of the follow-up.
- Levetiracetam was a favorable first-line option for diffuse glioma-associated epilepsy.
- In grade 2–3 gliomas, oxcarbazepine may be a reasonable first-line antiseizure medication.
1 INTRODUCTION
Seizures are a common complication of glioma. A seizure frequency of 60%–85% is seen in low-grade astrocytomas and oligodendrogliomas. In glioblastoma multiforme, the incidence of epilepsy varies from 30% to 50%.1 In our Helsinki population-based cohort of diffuse glioma (grade 2–4), we found epilepsy in 75.6% of patients.2 Refractory epilepsy is associated with both reduced quality of life and lower cognitive functioning in patients with low grade glioma.3 In people with epilepsy, seizure freedom dramatically improves quality of life (QOL).4 When seizure freedom cannot be achieved, adverse effects of antiseizure medications (ASMs) may have a major impact on QOL.4 Both ASM therapy and anti-tumor treatment including surgical resection, radiotherapy, and chemotherapy are associated with improved seizure control in patients with diffuse glioma.5-10 There are multiple ASMs effective against focal seizures with diverse mechanisms of action and other properties, allowing different ASMs to be selected for different patient subgroups.11, 12
ASM therapy is indicated for glioma patients after the first seizure.10, 13-15 Several factors to be considered in ASM choice are similar in glioma patients and other patients with epilepsy. They include seizure type, adverse effect profile, drug-interaction potential, rapid titration potential, cost and patient-specific features such as age, sex and comorbidities.11 In patients with glioma especially, it is important to take into account the expected or on-going surgical and oncological treatments; potential interactions with chemotherapy and other drugs required along with disease-related neurological deficits and psychiatric symptoms.16 To avoid adverse events based on enzyme induction, the use of carbamazepine, phenytoin and other enzyme-inducing ASMs is discouraged.1, 10, 13, 14 Valproic acid has been considered as one option for first-line ASM in brain tumor patients by some guidelines,13 although otherwise not first choice for focal epilepsy.17 Of other non-enzyme-inducing ASMs, levetiracetam has been studied the most in glioma patients,18 is well tolerated and commonly used.19, 20 Use of valproic acid and/or levetiracetam has been associated with a survival benefit in glioblastoma patients in some studies, but not confirmed in a pooled analysis of prospective clinical studies.21-24 At the moment, there is insufficient evidence to support any ASM choice on the basis of limiting progression or increasing benefit of treatment of glioma.10, 19 There are few comparative studies on suitability of different ASMs in glioma related epilepsy. Levetiracetam and enzyme-inducing ASMs have similar antiseizure effect.25 In a large study, compared to valproate, levetiracetam showed an advantage in efficacy both in low and high-grade glioma with a similar level of toxicity.18 Other large comparative studies are scarce, but reports on use of several different AMSs in glioma-related epilepsy have appeared. Lacosamide and pregabalin have been used with the suggested benefit of rapid titration and easy add-on therapy potential.26, 27 Perampanel has been hypothesized to benefit from its mechanism of action.10, 28, 29 Use of oxcarbazepine has been studied in a small, mixed brain tumor population suggesting non-inferiority to traditional ASMs, carbamazepine, phenobarbital, phenytoin and valproic acid.30
Levetiracetam has shown continued use and efficacy in a 5-year follow-up study.16, 31 The long-term (over 6–12 month) retention rate of other ASMs in the treatment of glioma patients is not well known.16, 19, 22 Although there are some comparative studies between different ASMs, there is no data on whether there are subgroups of diffuse glioma patients who may benefit from different ASMs as first-line ASM.
The purpose of this study was to identify factors contributing to the initial choice of ASM and the reasons for changes of ASM treatment in long-term follow-up of glioma patients as met in clinical practice. We assessed all initiations, need of treatment changes and long-term retention of individual ASMs in a well-characterized, population-based cohort of diffuse glioma patients. The aim of the study was to find out, if it is possible to identify characteristics that separate glioma patients into subgroups that may benefit from different ASMs.
2 MATERIALS AND METHODS
The study was a retrospective observational cohort study. It conformed to the Finnish legislation concerning medical research, and the study permission was granted by the Helsinki University Central Hospital HUCH Neurocenter Institutional Review Board.
2.1 Selection of patients
All cerebral glioma patients in contact with Helsinki University Hospital Neurology or Oncology Departments during 2013–2015 according to the ICD 10 diagnosis codes (C71) were identified. Patients were included, if they had (1) a histopathological diagnosis of World Health Organization grade 2–4 gliomas, (2) histopathological diagnosis of glioma was made after 2004 and (3) the seizure(s) led to the diagnosis or occurred within 5 years after diagnosis of glioma. A neuropathologist made the histopathological diagnosis. Only patients living in Helsinki were included. All glioma patients living in Helsinki were treated at Helsinki University Hospital. Patients may have had additional visits to occupational health care or other units, but neuro-oncological follow-up of the brain tumor including treatment of epilepsy was conducted at Helsinki University Hospital. Patients were excluded, if the histopathological diagnosis was some other tumor than astrocytoma, oligodendroglioma, oligoastrocytoma or glioblastoma, if the diagnosis was pilocytic astrocytoma, or if the primary tumor was outside the brain. Patients with previous epilepsy and patients with unavailable follow-up data due to moving from the Helsinki University Hospital catchment area were also excluded.
2.2 Treatment of patients
Glioma-associated epilepsy (GAE) was diagnosed after the first seizure caused by glioma. ASM was usually recommended after the first seizure. Treatment and magnetic resonance imaging (MRI) follow-up was planned by the neuro-oncological team and clinical follow-up of epilepsy was performed by a neurologist. The purpose of surgical resection was the safe removal of the tumor to an extent possible, without an attempt to delineate or remove the epileptogenic cortex.
During the first years of the study period, Finnish National guidelines suggested carbamazepine or oxcarbazepine as first-line treatment for focal-onset seizures. Levetiracetam was added as a first-line choice in the 3.2.2014 revision. Phenytoin was rarely used as long-term treatment in Finland due to its complicated pharmacokinetics and spectrum of adverse effects. Valproic acid, carbamazepine, oxcarbazepine, phenytoin, clobazam, and clonazepam were fully refundable as first ASMs by The Social Insurance Institution of Finland (KELA). Clobazam and clonazepam were almost exclusively in use as add-on therapy as proposed in glioma-related epilepsy guideline.10 More expensive ASMs, such as levetiracetam, lacosamide, pregabalin, gabapentin, topiramate, lamotrigine, and zonisamide were 40% refunded. These ASMs were fully refunded for yearly costs over 600 euros or if the first-line ASM has not been effective or has caused adverse effects. ASM was initiated intravenously (i.v.) or orally (p.o.) based on clinical evaluation.
2.3 Data collection
Data was collected from electronic medical records for 5 years from the diagnosis of glioma or until death as described in detail previously.2 Included were patient gender, age at diagnosis, presence of other somatic or psychiatric disease, along with abuse of alcohol or drugs. Glioma characteristics included date of glioma diagnosis, location and first symptom of glioma, histopathological diagnosis according to neuropathologist using the classification of brain tumors approved by the WHO 2000 or 2007, presence of IDH mutation, time and type of surgery, oncological treatments, and time of tumor progression. Imaging analysis included conventional T1- and T2-weighted MRI and gadolinium enhancement. In all patients with grade 2–4 glioma, detection of IDH1 R132H mutation at the time of resection or biopsy started in 2011 with immunohistochemistry and identification of IDH 1 and 2 mutations by sequencing started in 2014. IDH mutation was detected later retrospectively according to the clinical need in some cases. In Helsinki University Hospital, the clinical protocol is to start the detection of IDH1 R132H mutation with immunohistochemistry. If this is negative, sequencing is used at least for all patients with grade 2–4 glioma, excluding patients with grade 4 glioma and age over 55. Neuroradiologist evaluated the progression on MRI. Epilepsy data consisted of date of first seizure, seizure types, ASM treatment and seizure-free periods. P.o. ASM use and dosage were documented at the time of diagnosis and at the end of the follow-up. Seizure types were classified according to the ILAE 2017 Classification into focal aware, focal impaired awareness, or focal to bilateral tonic–clonic seizures.32 Reason for the first ASM selection if different from recommended or 100% reimbursed first-line treatment by KELA, all changes to the ASMs and the reasons for ASM changes were documented. Data on the final treatment phase in palliative/hospice care was not collected.
2.4 Statistical analysis
The association between first ASM and clinicopathologic variables was examined using cross tabulation test, chi-square test, Fisher exact test and Phi and Cramer's V. Same tests were used in analysis of ASM retention rate at the end of follow-up, ASM changes and ASM therapy at the end of follow-up. Cross tabulation test, chi-square test and Fisher exact test were used in other analysis. All analysis were performed with SPSS, version 28 (SPSS Inc). A p-value <0.05 was considered statistically significant.
3 RESULTS
3.1 Patient population
We identified all adult (≥18-year-old) patients living in Helsinki who had visited the neurology or oncology departments during 2013–2015 due to diffuse glioma, and who at the time of glioma diagnosis or later received diagnosis of glioma-associated epilepsy (n = 167). One hundred twenty-one patients were included after exclusion due to inaccuracy or very early (prior to 2005) diagnosis of glioma, epilepsy onset over 5 years after glioma diagnosis, lack of follow-up, or previous epilepsy due to other cause. Patient characteristics and selection have been described in detail previously.2 Data on seizure freedom at the end of follow-up was missing in five patients due to transfer to institutional care shortly after diagnosis, not committing to follow-up, alcohol abuse, drug abuse, or severe aphasia.
Characteristics of the patient population are shown in Table 1. A seizure was the presenting symptom of the glioma in 86 (71.1%) patients. If the first symptom was other than seizure, mean time from glioma diagnosis to first seizure was 7.1 months (median 2.0 months, SD 12.0, range 0–56). Most common seizure type was focal to bilateral tonic–clonic (71.7%). During follow-up, 53.7% of patients died. 48 (88.9%) patients with grade 4 glioma died during follow-up, although progression of glioma was observed in only 35 (64.8%) patients. This discrepancy is due to the definition of progression. Progression was only recorded as such if it was observed after a stable post-treatment period or led to a change in treatment. In aggressive grade 4 glioma with continuous progression, the patient may die before or without a change of treatment.
| Characteristics | Grade 2 gliomas | Grade 3 gliomas | Grade 4 gliomas |
|---|---|---|---|
| No. of patients (%) | 49 (40.5) | 18 (14.9) | 54 (44.6) |
| Gender | |||
| Female (%) | 21 (42.9) | 9 (50.0) | 12 (22.2) |
| Male (%) | 28 (57.1) | 9 (50.0) | 42 (77.8) |
| Age, median (years) | 35.0 | 44.0 | 60.5 |
| Seizure as first symptom (%) | 44 (89.8) | 14 (77.8) | 28 (51.9) |
| Tumor location (%) | |||
| Frontal | 25 (51.0) | 8 (44.4) | 20 (37.0) |
| Temporal | 7 (14.3) | 3 (16.7) | 8 (14.8) |
| Parietal | 3 (6.1) | 1 (5.6) | 7 (13.0) |
| Occipital | 0 | 0 | 3 (5.6) |
| Several lobes/large | 14 (28.6) | 6 (33.3) | 16 (29.6) |
| IDH mutation (%) | |||
| Yes | 40 (81.6) | 9 (50.0) | 3 (5.6) |
| No | 4 (8.2) | 8 (44.4) | 49 (90.7) |
| Unknown | 5 (10.2) | 1 (5.6) | 2 (3.7) |
| Seizure typesa (%) | |||
| Focal aware | 28 (57.1) | 9 (50.0) | 28 (51.9) |
| Focal impaired awareness | 13 (26.5) | 6 (33.3) | 16 (29.6) |
| Focal to bilateral tonic–clonic | 39 (79.6) | 11 (61.1) | 36 (66.7) |
| Other | 0 | 0 | 0 |
| Change to more malignant during follow-upb (%) | 12 (24.5) | 2 (11.1) | 0 |
| Progression during follow-upc (%) | 28 (57.1) | 11 (61.1) | 35 (64.8) |
| Tumor treatments during follow-upd (%) | |||
| Resection | 42 (85.7) | 17 (94.4) | 47 (87.0) |
| Biopsy | 11 (22.4) | 1 (5.6) | 7 (13.0) |
| Radiotherapy | 28 (57.1) | 15 (83.3) | 27 (50.0) |
| Radiotherapy alonee | 11 (22.4) | 3 (16.7) | 2 (3.7) |
| Chemotherapy | 28 (57.1) | 14 (77.8) | 42 (77.8) |
| Chemotherapy alonee | 9 (18.4) | 0 | 0 |
| Chemoradiotherapyf | 4 (8.2) | 7 (38.9) | 47 (87.0) |
| No oncological treatments | 9 (18.4) | 0 | 4 (7.4) |
| Death during follow-up (%) | 6 (12.2) | 11 (61.1) | 48 (88.9) |
- a Several seizure types were possible in the same patient.
- b Histopathological verification.
- c Progression leading to a change of treatment or after a stable post-treatment period.
- d Several different treatments were possible during follow-up.
- e No other oncological treatments during follow-up.
- f Simultaneous administration of chemotherapy and radiotherapy.
3.2 Initiation of antiseizure medication
The first p.o. ASMs initiated were oxcarbazepine, levetiracetam, valproate, carbamazepine, or phenytoin (Table 2). The most common first ASM was oxcarbazepine (39.2%), followed by levetiracetam (22.5%), carbamazepine (19.3%), phenytoin (10.8%) and valproic acid (8.3%) (Table 2). The first ASM was retained in use in 70 (57.9%) patients until the end of follow-up.
| Features | OXC patient number (%) | LEV patient number (%) | VPA patient number (%) | CBZ patient number (%) | PHT patient number (%) | p-value | Effect size (phi/Cramer's V) |
|---|---|---|---|---|---|---|---|
| All | 47 (39.2) | 27 (22.5) | 10 (8.3) | 23 (19.2) | 13 (10.8) | ||
| Gender | 0.581 | 0.152 | |||||
| Female | 17 (36.2) | 6 (22.2) | 4 (40.0) | 8 (34.8) | 6 (46.2) | ||
| Male | 30 (63.8) | 21 (77.8) | 6 (60.0) | 15 (65.2) | 7 (53.8) | ||
| Grade | 0.043 | 0.253 | |||||
| 2 | 26 (55.3) | 5 (18.5) | 4 (40.0) | 10 (43.5) | 3 (23.1) | ||
| 3 | 6 (12.8) | 3 (11.1) | 2 (20.0) | 3 (13.0) | 4 (30.8) | ||
| 4 | 15 (31.9) | 19 (70.4) | 4 (40.0) | 10 (43.5) | 6 (46.2) | ||
| IDH mutation | 0.058 | 0.265 | |||||
| Yes | 25 (53.2) | 8 (29.6) | 4 (40.0) | 9 (39.1) | 6 (46.2) | ||
| No | 20 (42.6) | 19 (70.4) | 5 (50.0) | 9 (39.1) | 7 (53.8) | ||
| Unknown | 2 (4.3) | 0 (0.0) | 1 (10.0) | 5 (21.7) | 0 (0.0) | ||
| Location | 0.209 | 0.211 | |||||
| Frontal | 24 (51.1) | 12 (44.4) | 3 (30.0) | 11 (47.8) | 3 (23.1) | ||
| Temporal/parietal/occipital | 13 (27.7) | 4 (14.8) | 3 (30.0) | 8 (34.8) | 3 (23.1) | ||
| Several lobes/large | 10 (21.3) | 11 (40.7) | 4 (40.0) | 4 (17.4) | 7 (53.8) | ||
| Progression during follow-up | 0.759 | 0.128 | |||||
| Yes | 26 (55.3) | 19 (70.4) | 7 (70.0) | 14 (60.9) | 8 (61.5) | ||
| No | 21 (44.7) | 8 (29.6) | 3 (30.0) | 9 (39.1) | 5 (38.5) | ||
| 1st symptom | 0.001 | 0.387 | |||||
| Seizure | 39 (83.0) | 12 (44.4) | 5 (50.0) | 20 (87.0) | 10 (76.9) | ||
| Other | 8 (17.0) | 15 (55.6) | 5 (50.0) | 3 (13.0) | 3 (23.1) | ||
| Death during follow-up | 0.047 | 0.281 | |||||
| Yes | 18 (38.3) | 20 (74.1) | 6 (60.0) | 12 (52.2) | 8 (61.5) | ||
| No | 29 (61.7) | 7 (25.9) | 4 (40.0) | 11 (47.8) | 5 (38.5) | ||
| Number of ASM initiations | <0.001 | 0.386 | |||||
| Years 2005–2010 | 14 (42.4) | 1 (3.0) | 2 (6.1) | 13 (39.4) | 3 (9.1) | ||
| Years 2011–2015 | 33 (37.9) | 26 (29.9) | 8 (9.2) | 10 (11.5) | 10 (11.5) |
- Abbreviations: ASMs: CBZ, carbamazepine; LEV, levetiracetam; OXC, oxcarbazepine; PHT, phenytoin; VPA valproic acid.
Comparing the time periods 2005–2010 and 2011–2015, the initiation of levetiracetam as first ASM increased and that of carbamazepine decreased (Table 2). Among patients with low-grade glioma, oxcarbazepine was favored, whereas grade 4 glioma patients more often received levetiracetam as first ASM. Levetiracetam was preferred as first-line ASM in patients with first symptom other than seizure. At the end of follow-up, the retention rate of levetiracetam as first ASM was higher than the retention rate of other ASMs in grade 4 glioma patients (p 0.002) (Table 3).
| Characteristics | OXC patient number (%) | LEV patient number (%) | VPA patient number (%) | CBZ patient number (%) | p-value | Effect size (phi/Cramer's V) |
|---|---|---|---|---|---|---|
| All patients | 0.055 | 0.258 | ||||
| Yes | 27 (57.4) | 23 (85.2) | 5 (50.0) | 14 (60.9) | ||
| No | 20 (42.6) | 4 (14.8) | 5 (50.0) | 9 (39.1) | ||
| Grade 4 glioma | 0.002 | 0.516 | ||||
| Yes | 7 (46.7) | 18 (94.7) | 3 (75.0) | 4 (40.0) | ||
| No | 8 (53.3) | 1 (5.3) | 1 (25.0) | 6 (60.0) | ||
| Grade 2 or 3 glioma | 0.372 | 0.238 | ||||
| Yes | 20 (62.5) | 5 (62.5) | 2 (33.3) | 10 (76.9) | ||
| No | 12 (37.5) | 3 (37.5) | 4 (66.7) | 3 (23.1) |
- Abbreviations: ASMs: CBZ, carbamazepine; LEV, levetiracetam; OXC, oxcarbazepine; VPA valproic acid.
In 40 patients, choice of first p.o. ASM was other than a 100% refundable ASM or a first-line ASM suggested by national or international guidelines. For 13 patients initiating phenytoin and for 20 patients initiating levetiracetam, the possibility of i.v. administration and/or the need for a rapid effect was the reason for ASM choice. In addition, four patients initiated levetiracetam to avoid complications with treatment of glioma and in three patients the reason was not specified.
3.3 Changes in ASM treatment
In total, 263 initiations of ASMs were made and levetiracetam was the most commonly overall initiated ASM. There were 71 (27.0%) initiations of levetiracetam, 62 (23.6%) initiations of oxcarbazepine, 40 (15.2.%) initiations of valproic acid, 34 (12.9%) initiations of carbamazepine, 19 (7.2%) initiations of phenytoin and 37 (14.1%) initiations of other ASMs. In addition to most common ASMs, changes of ASM and add-on treatment included lacosamide, lamotrigine, pregabalin, topiramate, perampanel, clobazam and clonazepam.
During the 5-year follow-up, 147 ASM changes were made (Figure 1), leading ultimately to discontinuation of 76 previous ASMs. Lack of antiseizure effect was the reason for change in 63 cases (42.9%), adverse effects in 50 cases (34.0%), other reasons in 29 cases (19.7%) and unspecified in five cases (3.4%). Change from levetiracetam to a 100% reimbursed ASM and change from phenytoin to other ASM were the most common reasons in the other reasons group.
When four most initiated ASMs were compared, there was a trend towards an adverse effect being a more common reason for changing oxcarbazepine or carbamazepine than levetiracetam to another ASM. However, the only statistically significant difference in this analysis was that levetiracetam was changed due to other reason more often than other ASMs.
During follow-up, 41 (33.8%) patients received only one ASM (Table 4). Forty-five (37.2%) patients initiated one other ASM, 17 (14.0%) patients initiated two other ASMs, 10 (8.3%) patients initiated three other ASMs, 4 (3.3%) patients four other ASMs and 3 (2.5%) patients initiated six other ASMs. Twenty-eight patients initiated the second ASM and 8 patients initiated the second and the third ASM due to lack of effectiveness. Median of further ASM initiations after first ASM was 1 (mean 1.18). When valproic acid was initiated as first ASM, patients initiated more ASMs during follow-up than when first ASM was oxcarbazepine, levetiracetam or carbamazepine (p 0.005) (Table 4).
| OXC patient number (%) | LEV patient number (%) | VPA patient number (%) | CBZ patient number (%) | p-value | Effect size (phi/Cramer's V) | |
|---|---|---|---|---|---|---|
| ASM initiations after first ASM initiation | 0.005 | 0.272 | ||||
| 0 | 20 (42.6) | 13 (48.1) | 0 | 8 (34.8) | ||
| 1 | 13 (27.7) | 9 (33.3) | 5 (50.0) | 13 (56.5) | ||
| 2 | 7 (14.9) | 2 (7.4) | 0 | 2 (8.7) | ||
| 3 or more | 7 (14.9) | 3 (11.1) | 5 (50.0) | 0 |
- Abbreviations: ASMs: CBZ, carbamazepine; LEV, levetiracetam; OXC, oxcarbazepine; VPA valproic acid.
3.4 Seizure freedom and type of therapy
At some time-point during the follow-up period, 67 patients (55.4%) were seizure-free for at least 12 months. At the end of follow-up, 46 (38.0%) of the patients had been seizure-free during the last year. Sixty-two patients received monotherapy (51.2%), 52 patients were on polytherapy (43.0%), and seven patients did not use any ASMs (5.8%), one patient never starting ASMs.
Twenty-eight patients (40.0%) of 70 patients still using the first ASM at the end of follow-up were seizure-free during the last year of follow-up. Of these patients, 32 were alive and 38 were deceased at the end of follow-up. Twenty-four (75.0%) of the alive patients were seizure-free and 4 (10.5%) of the deceased patients were seizure-free during the last year of follow-up (p < 0.001).
Seizure-freedom at the end of follow-up was more common in patients receiving monotherapy than patients receiving polytherapy (Table 5). Progression correlated positively with use of polytherapy. The duration of epilepsy, resection, radiotherapy, chemotherapy or chemoradiotherapy had no statistically significant correlation with ASM therapy type at the end of follow-up.
| Characteristics | Monotherapy (%) | Polytherapy (%) | No ASM (%) | p-value | Effect size (phi/Cramer's V) |
|---|---|---|---|---|---|
| Grade | 0.644 | 0.100 | |||
| 2 | 25 (51.0) | 20 (40.8) | 4 (8.2) | ||
| 3 | 7 (38.9) | 10 (55.6) | 1 (5.6) | ||
| 4 | 30 (55.6) | 22 (40.7) | 2 (3.7) | ||
| Progression | 0.024 | 0.245 | |||
| Yes | 34 (45.9) | 38 (51.4) | 2 (2.7) | ||
| No | 28 (59.6) | 14 (29.8) | 5 (10.6) | ||
| Death during follow-up | 0.112 | 0.192 | |||
| Yes | 30 (46.2) | 33 (50.8) | 2 (3.1) | ||
| No | 32 (57.1) | 19 (33.9) | 5 (8.9) | ||
| Seizure-free ≥1 year at the end of the follow-up | <0.001 | 0.347 | |||
| Yes | 30 (65.2) | 11 (23.9) | 5 (10.9) | ||
| No | 30 (42.9) | 39 (55.7) | 1 (1.4) |
Altogether there were 24 different polytherapy combinations at the end of follow-up. Polytherapy consisted of two ASMs in 43 patients, three ASMs in six patients, four ASMs in one patient and five ASMs in two patients. At the end of follow-up, the two most common polytherapies were the combinations of levetiracetam with oxcarbazepine (n = 11), and levetiracetam with valproate (n = 9). Levetiracetam combined with other sodium channel blocker than oxcarbazepine (lacosamide or carbamazepine or phenytoin) was received by 7 patients. Valproic acid and oxcarbazepine were used by three patients and valproic acid and other sodium channel blocker (lamotrigine or carbamazepine) by 6 patients.
At the end of follow-up, the mean doses of ASMs in monotherapy were slightly lower than in polytherapy: levetiracetam 1620 mg/d in monotherapy (n = 27) and 2076 mg/d in polytherapy (n = 33), oxcarbazepine 1058 mg/d in monotherapy (n = 18) and 1105 mg/d in polytherapy (n = 19), valproic acid 917 mg/d in monotherapy (n = 6) and 1389 mg/d in polytherapy (n = 27), carbamazepine 605 mg/d in monotherapy (n = 11) and 791 mg/d in polytherapy (n = 11).
4 DISCUSSION
The present study analyzed retrospectively the choice and long-term use of different ASMs in a population-based cohort of diffuse (grade 2–4) glioma patients. The key findings were, that (1) in the majority of patients (57.9%), the first initiated ASM was retained as long-term treatment, (2) in grade 4 glioma patients, levetiracetam was favored as a first ASM and also retained more often in use than oxcarbazepine, and (3) Patients who initiated valproic acid as the first ASM underwent more ASM changes than patients who initiated other ASM as the first ASM.
The first ASM shows efficacy and tolerability in approximately half of patients with newly diagnosed epilepsy.33 In the present study, majority (57.9%) continued the initial ASM at the end of follow-up. In this patient group, seizure freedom was more common in patients who were alive at the end of follow-up than in patients who deceased, as it was in the entire patient cohort.2 These findings indicate that different ASMs can be used in treatment of epilepsy in diffuse glioma patients allowing other patient characteristics to be taken into account, as appropriate individually.
For adult focal epilepsy, oxcarbazepine is considered a first-line option.12 The present study specifically describes the potential of oxcarbazepine use in glioma patients, previous studies on oxcarbazepine use being limited to mixed brain tumor patients.30 Oxcarbazepine is proposed as an option in SNO guideline both as a primary and add on-treatment for GAE.10 The scarce reports of oxcarbazepine use in this patient population may imply that it is more common in Finnish clinical practice than it is in other countries. A Norwegian registry-based study indicates more infrequent use than in our study population.34 17% of patients had psychiatric comorbidity in the present study patient population.2 The relatively low risk for psychiatric and behavioral adverse effects is a significant benefit in glioma patients and favors oxcarbazepine.35
There are, however, several disadvantages to be considered in oxcarbazepine use. Although oxcarbazepine has a relatively low interaction potential, it has some enzyme-inducing and enzyme-inhibiting properties.36 It may reduce the serum levels of certain anticancer drugs, such as imatinib.10, 37 The most common dose-related side-effects of oxcarbazepine are fatigue, headache, dizziness, and ataxia.38 Glioma patients often have cognitive disturbance, fatigue and other side-effects from tumor-directed treatments.10 Sometimes it may be difficult to distinguish side-effects of oxcarbazepine from symptoms caused by oncological or neurosurgical treatment. Oxcarbazepine may cause hyponatremia and changes in liver enzymes.38, 39 Hyponatremia may worsen the symptoms of patients with brain edema and changes in liver enzymes may hamper oncological treatment possibilities. These features of oxcarbazepine implicate favored use in relatively stable periods of glioma, eg. grade 2 (to 3) glioma patients.
Within the study population, the difference in retention rates favored levetiracetam in grade 4 glioma patients. No statistically significant difference could be demonstrated in ASM retention rate between levetiracetam and oxcarbazepine, carbamazepine or valproic acid in patients with grade 2 or 3 glioma, but the levetiracetam group was small in this analysis (n = 8). The present study results support the good tolerability of levetiracetam in glioma patients.18, 40
In the present study, valproic acid was changed to other ASM due to the lack of antiseizure effect slightly more often than oxcarbazepine, levetiracetam or carbamazepine, but the difference was not statistically significant. All patients initiating valproic acid as first ASM initiated at least one other ASM during follow-up, which may also suggest inferior efficacy. This is in accordance with the previous valproate vs. levetiracetam comparative study.18 Valproic acid can cause thrombocytopenia, platelet dysfunction, and coagulation abnormalities, which represent a concern for patients in whom neurosurgery is envisaged.41 In addition, combined use of valproic acid and chemotherapy has been associated with higher hematologic toxicities.19 The findings in the present study are in line with the guidelines suggesting levetiracetam a first-line ASM in glioma patients instead of the previously recommended valproic acid.10, 14
Lamotrigine is considered a first-line option for glioma-associated epilepsy.14 However, the data of its efficacy as monotherapy in tumor-related epilepsy is limited.10 In the present study, lamotrigine was only used in few patients with difficult-to-treat epilepsy. The titration period of lamotrigine is fairly long to reduce the risk of an allergic reaction and it cannot be started in an acute setting as a quarter of patients required in the present study. These characteristics favor other ASMs instead of lamotrigine as first-line ASM for glioma patients requiring urgent initiation of ASM.
Seizure freedom is always the ultimate goal of ASM treatment, but it should not be pursued at all costs.11 In glioma patients, overtreatment can result in aggravation of neurological deficits and symptoms. In the present study, the use of polytherapy correlated with progression of glioma and treatment resistance, reflecting probably the severity of epilepsy.42 Likewise, the ASM doses were bigger in polytherapy regimens than in monotherapy. According to a long-term follow-up study, polytherapy may eventually be more common in users of levetiracetam than alternative ASMs in grade 2 glioma.31 This may be related to effectiveness or incomplete long-term seizure control related to the natural course of disease. In focal epilepsy unspecified by etiology, levetiracetam has shown slightly poorer efficacy in comparison with lamotrigine or carbamazepine in the SANAD and KOMET studies.43, 44 However, in patients over 60 years this was overcome by the favorable tolerability profile of levetiracetam compared to carbamazepine.45 Levetiracetam was also common in polytherapy in the present study. This can indicate lack of efficacy, but may also reflect simply reimbursement effects in Finland: When older drugs have been favored as first ASM, when possible, levetiracetam has more often been initiated as add-on treatment.
It is likely to be impossible to find one ASM which is suitable for all glioma patients. The outcome is uncertain at presentation, both regarding future seizure control, and the speed of glioma progression. However, the patient may benefit from a slightly different approach to the treatment of epilepsy when there is a rapid progression compared to slow or prolonged stable periods in glioma growth and malignization. There may be different phases of treatment and treatment needs. Low potential for interactions, adverse-effect profile, and efficacy of levetiracetam are favorable for glioma patients.31, 46 Levetiracetam is not associated with cognitive impairment but has been associated with psychiatric and behavioral adverse effects.10, 40, 46 Rapid and/or i.v. initiation of ASM treatment may be required and it is preferable to then be able to continue with the ASM started, which is not the case for phenytoin. Adverse effect profile of oxcarbazepine is less favorable for glioma patients especially during surgical or oncological treatment. Taking into account the low interaction potential and the findings of present study, oxcarbazepine as first ASM may be suitable option for grade 2–3 glioma patient with risk of psychiatric adverse effects. For grade 4 glioma patients, the present study suggests that levetiracetam is the preferred option for first-line ASM, instead of oxcarbazepine. This is accordance with observations from other European grade 4 glioma population.47 As in other focal epilepsy, a proportion of patients will be treatment-resistant and then may benefit from diverse ASM combinations to be considered individually.
The present study has the advantage of a representative, well-defined population-based data with few dropouts and long follow-up. The cohort includes different phases of glioma progression and treatment needs. The major limitations of the study are its retrospective nature influencing the reliable recognition of all seizures, and the small patient number. Furthermore, there is an impact on the first ASM selection by the Finnish ASM reimbursement system, which need to be considered in the interpretation of the results. Of the four most commonly initiated ASMs, levetiracetam is the only ASM without 100% reimbursement of KELA as a first-line ASM therapy in glioma patients. The scarce use of other newer non-enzyme-inducing ASMs, including lamotrigine and lacosamide, prevented the analysis of these ASMs. Therapeutic interventions, surgery, radiotherapy and chemotherapy, against glioma are important contributors to seizure control in glioma patients, but the effect is difficult to control in a retrospective study. Although these limitations cause some caution in interpreting the results, the study is a representative cohort study of patients likely to be met in real-world clinical practice.
5 CLINICAL RELEVANCE
Several factors need to be considered in ASM choice for patients with glioma. It depends on the patient which factors ultimately turn out to be the most important. In general, levetiracetam seems to be a favorable first-line treatment for diffuse glioma-associated epilepsy, especially for patients with grade 4 glioma. For grade 2–3 glioma patients at risk of psychiatric adverse-effects, oxcarbazepine may be a reasonable option.
ACKNOWLEDGMENTS
L.O. is grateful to Helsinki University Hospital, Epilepsy Research Foundation in Finland and Maire Taponen foundation for financial support of this study. Open access publishing facilitated by Helsingin yliopisto, as part of the Wiley - FinELib agreement.
FUNDING INFORMATION
This research was supported by Helsinki University Hospital (L.O.), Epilepsy Research Foundation in Finland (L.O.) and Maire Taponen foundation (L.O.). Open access funded by Helsinki University Library.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
ETHICS STATEMENT
We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Open Research
DATA AVAILABILITY STATEMENT
Data cannot be shared openly to protect study subject privacy.
