Keywords: Aged - Antiepileptic Drugs - Epilepsy - Inpatients - Seizures
Palavras-chave: Idoso - Anticonvulsivantes - Epilepsia - Paciente Internados - Convulsões
INTRODUCTION
People aged over 60 years represent the fastest-growing population group in the world[1 ]. The burden of neurological disorders such as stroke, neurodegenerative, and neoplastic
disorders, which are the most frequent etiologies of epilepsy in older adults, increases
as the population ages[2 ]. Moreover, 25 to 30% of all new-onset seizures occur in individuals older than 60
years[3 ],[4 ]. Prescription of antiepileptic drugs (AEDs) has been increasing as epilepsy diagnoses
in older people escalate. This demographic shift poses important challenges to prescription
safety. Although there is some evidence for prescribing AEDs to older individuals,
the inclusion of this population in clinical trials is uncommon[5 ]. Older people often have multiple comorbidities in addition to other impairments
such as altered physiology and pharmacokinetics, polypharmacy, and atypical disease
presentations. Furthermore, the simultaneous use of AEDs and other medicines may result
in drug-drug interactions, dose-related side effects, and idiosyncratic drug reactions,
which hamper the appropriate treatment of epilepsy[6 ]. Newer AEDs were introduced in the market to increase efficacy and reduce side effects,
and epilepsy experts regard several of them as more appropriate options for older
people with epilepsy[7 ],[8 ]. Older AEDs, such as phenobarbital and phenytoin, have similar efficacy compared
with the newer ones, but they usually have a less favorable safety profile and are
considered suboptimal AEDs for older people with epilepsy (EPWE)[9 ],[10 ],[11 ].
A prior study was conducted to investigate prescribing patterns of AEDs for older
adult inpatients with late-onset epilepsy and found that the most common AED prescribed
to that population was phenytoin (53%)[12 ], which is consistent with several studies also showing phenytoin as the most commonly
prescribed AED (50-70%)[9 ],[13 ]. Notwithstanding guidelines and expert consensus opinions advocating the use of
newer AEDs in the treatment of epilepsy among older people, these patients still receive
suboptimal treatment in clinical practice. The aim of this study was: first, to assess
changes in prescribing patterns of AEDs among older adult inpatients with late-onset
epilepsy between 2009-2010 and 2015-2019; and second, to interpret any unexpected
patterns of AED prescription over the 2015-2019 period.
METHODS
This prospective, observational, single-center study was conducted on patients aged
≥60 years who were consecutively admitted to Hospital São Rafael, a general tertiary
teaching hospital with 356 beds located in Salvador, state of Bahia, Brazil, between
November 2015 and May 2019.
The study participants included older adult inpatients with epileptic seizures that
either led to their hospital admission or occurred during a period of care in the
emergency room or during hospitalization, and who had their first unprovoked seizure
at 60 years of age or older. All patients were under the care of a neurology team.
Exclusion criteria were: (1) older adult inpatients who had been admitted with a diagnosis
of seizure that was not confirmed or who were later diagnosed with other paroxysmal
neurological disorders such as syncope, delirium, or transient ischemic attack; (2)
patients with isolated acute symptomatic seizures, including those who had seizures
in the setting of an acute traumatic brain injury and were followed up by the neurosurgery
team, except those with remote and/or progressive seizures (epilepsy) who also had
a seizure due to an acute cause at the time of the study; (3) previously included
patients who had been readmitted to the hospital, even if they had had other seizures;
and (4) patients on AEDs for conditions other than epilepsy such as neuropathic pain,
mood disorders, and migraine.
To investigate trends in prescribing patterns of AEDs, the AEDs prescribed to the
study cohort of older adult inpatients with late-onset epilepsy between 2015 and 2019
(Y2) were identified and these data were compared to results of the previous retrospective
study on a similar population conducted between 2009 and 2010 (Y1)[12 ]. Patient characteristics were also analyzed to identify any risk factor for receiving
suboptimal therapy.
Diagnostic criteria
The diagnosis of epilepsy followed the recommendations of the International League
Against Epilepsy (ILAE) Official Report[14 ]. The diagnosis of status epilepticus (SE) included the revised concepts, definition,
and classification from the ILAE[15 ].
The characteristics of the seizures were described, according to the latest recommendations,
based on the Operational Classification of Seizure Types by the ILAE[16 ].
An acute symptomatic seizure was diagnosed according to the ILAE recommendations[17 ].
To ascribe an acute symptomatic seizure to metabolic imbalance, the cutoff values
that are most likely to be associated with seizures were used, namely: sodium <115
mg/dL (or >145 mmol/L), calcium <5.0 mg/dL, magnesium <0.8 mg/dL, glucose fasting
<36 mg/dL (or >~450 mg/dL, when associated with ketoacidosis), urea >214 mg/dL, and
creatinine >10 mg/dL[17 ],[18 ].
Seizure etiology was categorized as (1) symptomatic (known cause), including a) acute
seizures (e.g., stroke, central nervous system [CNS] infection, metabolic disorder,
and autoimmune disease), b) remote seizures (e.g., post-stroke, posttraumatic, and
postencephalitic), and c) progressive symptomatic (brain tumor and dementia), and
as (2) unknown cause[15 ],[19 ].
Late-onset epilepsy was conventionally defined as epileptic seizures first occurring
in subjects aged 60 years or older[20 ],[21 ]. An unprovoked seizure was defined as a seizure occurring in the absence of precipitating
factors and that may have been caused by static or progressive injury[22 ].
Patient multimorbidity was defined as the coexistence of two or more chronic conditions[23 ].
Investigation of seizure etiology included laboratory tests (glucose, urea, magnesium,
sodium, calcium, and creatinine). Routine electroencephalogram (EEG), 24-h video-EEG,
and brain imaging data (computed tomography [CT], magnetic resonance imaging [MRI],
MRI angiography [MRA], and positron emission tomography [PET]) were also collected
when necessary. Some patients underwent cerebrospinal fluid analysis. The following
EEG results were considered abnormal: diffuse or focal slowing abnormalities and interictal
or ictal paroxysms. This definition excluded records with nonspecific abnormalities.
Analysis methods
The following data were analyzed: (1) demographic data, including age and sex; (2)
hospitalization data, including intensive care unit (ICU) admission and length of
stay (LOS); (3) characteristics of the seizure disorders during the index seizure,
such as SE, acute symptomatic seizure, single unprovoked seizure, epilepsy, seizure
type, and age at onset of the first seizure; (4) prescription of AEDs at the time
of the index seizure and/or at discharge; (5) potentially proconvulsant comedications;
(6) seizure etiology; and (7) clinical and neurological comorbidities.
AED regimens were classified for this population into two subgroups according to evidence-based
clinical recommendations to keep the same categorization of our previous study on
a similar population conducted between 2009 and 2010 (Y1): a) group 1 included appropriate
AEDs (carbamazepine, oxcarbazepine, valproic acid, gabapentin, clobazam, lamotrigine,
levetiracetam, topiramate, and lacosamide); and b) group 2 comprised suboptimal AEDs
(phenytoin and phenobarbital)[12 ]. AED regimens included both monotherapy and polytherapy (combination therapy with
two or more AEDs)[12 ].
Statistics
Descriptive statistics for the cohort of late-onset epilepsy were tabulated. Quantitative
variables with normal distribution were expressed as means and standard deviations.
Variables with non-normal distribution were expressed as medians and interquartile
ranges. Normal variables were identified by the analysis of histograms and box plots
and by the Kolmogorov-Smirnov test. Categorical variables were reported as frequencies
and percentages.
Chi-square analysis was used to investigate changes in the initial AED treatments
that patients received between Y1 (year) (2009-2010) and Y2 (2015-2019). To ensure
these bivariate results were not due to changes in patient characteristics over time,
logistic regression models were also used to control for patient characteristics (age,
sex, and comorbidities).
For bivariate comparisons, Student’s t- test was used for numerical variables with normal distribution and the Mann-Whitney
test for those with non-normal distribution. Categorical variables were compared by
Pearson’s chi-square or Fisher’s exact test when necessary.
A backward stepwise multiple logistic regression was performed for data analysis,
which was conducted in three blocks to improve the power of the model, with demographic
variables followed by seizure etiologies and comorbidities.
This study considered p≤0.05 as statistically significant for univariate and multivariate
analysis.
SPSSâ Statistics (v. 25, Chicago, IL, USA), R Program (v.3.4.4), and Microsoft Excelâ
2016 were used to perform statistical analyses.
Ethics
The Research Ethics Committee of Hospital São Rafael approved this study on November
24, 2014 (no. 904.379, version 5).
RESULTS
Demographic data
A total of 150 older adult patients met the study inclusion criteria, 16 of whom were
excluded because they were later recognized to have isolated acute symptomatic seizures;
however, patients with acute symptomatic seizures in combination with another seizure
etiology, such as remote or progressive causes, were enrolled in this study. The present
cohort comprised 134 patients that were followed up during their hospitalization.
Patients were mostly men (53%) and married (62.7%); 91.8% of patients had private
health insurance coverage.
The mean age of the cohort was 77.2±96.6 years, with a median age of 78 years (interquartile
range [ IQR ], 62-94 years). Mean age at the first-time seizure was 76.5±9.8 years.
A total of 87 (64.9%) patients were admitted to the ICU. The median LOS was 11 days,
with an interquartile range of 5-21 days.
Clinical characteristics
The cause of epilepsy was determined for nearly all patients, including: cerebrovascular
disorders in 75 (56%) patients, dementia in 33 (24.6%) patients, and brain tumors
in nine (6.7%). Three (2.2%) patients had a combination of two of these conditions
(stroke and dementia). Epilepsy of unknown cause was found in 14 (10.5%) patients
of this population. Focal onset seizures were the most common type, occurring in 68.6%
of patients, while seizures of unknown type occurred in 3%. In addition, we found
a great proportion of patients who presented with SE (33.6%), most of whom had convulsive
SE (75.5%).
The mean number of comorbidities was high, 5.4±2.1 (95% confidence interval [95%CI]
5.04 to 5.77). The most commonly identified clinical comorbidities were hypertension
(83.6%), dyslipidemia (58%), and systemic infections (50%). Data are summarized in
[Table 1 ].
Table 1
Most common neurological and clinical comorbidities among 134 older adult inpatients
in a tertiary center.
Comorbidities
Frequency
Hypertension
112 (83.6)
Cerebrovascular disorders
84 (62.7)
Ischemic stroke
72 (53.7)
Dyslipidemia
78 (58%)
Systemic infections
67 (50)
Diabetes mellitus
57 (42.5)
Psychiatric disorders
47 (35)
Cardiac arrhythmias
41 (30.6)
Toxic and metabolic disorders
39 (29)
Dementia
35 (26)
Kidney failure
32 (24)
Non-brain tumors
27 (20.2)
Sepsis
24 (18)
Hypothyroidism
24 (18)
Delirium
21 (15.7)
Toxic-metabolic encephalopathy
15 (11.2)
Movement disorders
13 (9.7)
Brain tumors
9 (6.7)
Data are expressed as number of patients, with percentages (%) in brackets.
Routine EEG was performed in 91.8% of patients, and 77 (57.5%) of them fulfilled the
predefined criteria for abnormal records.
Potentially proconvulsant comedications were used by 50% of patients and are shown
in [Table 2 ].
Table 2
Potentially proconvulsant comedications in addition to antiepileptic drugs among 134
older adult inpatients.
Comedication
Frequency
Cephalosporin 3rd and 4th generation (35)
32 (23.9)
Quetiapine (36)
31 (23.1)
Beta blocker (35)
24 (17.9)
Carbapenem (35)
20 (14.9)
Insulin (35)
19 (14.2)
Diuretic (thiazide) (35)
13 (9.7)
Penicillin derivative (35)
11 (8.2)
Antiarrhythmic (35)
9 (6.7)
Levodopa (35)
8 (5.9)
Olanzapine (36)
8 (5.9)
Opioid pain killer (35)
7 (5.2)
Haloperidol (35)
5 (3.7)
Phenothiazine (35)
4 (2.9)
Risperidone (36)
1 (0.7)
Quinolone (35)
1 (0.7)
TCAs (amitriptyline) (35)
1 (0.7)
TCAs: tricyclic antidepressants. Data are expressed as number of patients, with percentages
(%) in brackets.
Prescribing patterns of antiepileptic drugs
A suboptimal AED regimen was prescribed to 69 patients (51.5%; 95%CI 42.7-60.2), and
there were no sex disparities; these patients received phenytoin and/or phenobarbital
alone or in combination with any other AED ([Figure 1 ]). Patients were preferentially treated with monotherapy (53.7%; 95%CI 44.9-62.3).
Polytherapy with two or three AEDs was prescribed to 41.8% and 3.7% of patients (95%CI
33.4 to 50.6%; 95%CI 1.4-8.9), respectively ([Figure 2 ]). Phenytoin was the most commonly used AED (51.5%; 95%CI 42.7-60.15). Lamotrigine
was the second most frequently prescribed AED in a monotherapy regimen (33.6%; 95%CI
25.8-42.3), followed by levetiracetam (29.1%; 95%CI 21.7-37.7).
Figure 1 Temporal trend in the proportion of inpatients on appropriate AEDs and suboptimal
AEDs between Y1 (2009-2010) and Y2 (2015-2019).
Figure 2 AED regimens including monotherapy and polytherapy among 134 older adult inpatients.AED:
antiepileptic drugs.
When comparing prescribing patterns of AEDs in Y2 (2015-2019) with the results of
the previous study in Y1 (2009-2010), a significant change was observed in prescribing
patterns of AED regimens over time, with the proportion of patients prescribed a suboptimal
regimen in Y1 (73.3%) proving higher than that in Y2 (51.5%) (p<0.001). This reduction
in the use of suboptimal AEDs was due to a considerable decline in phenobarbital prescription
(from 9.2 to 0.75%). In addition, a modest decrease in the use of phenytoin was found
(from 64 to 51.5%). Although carbamazepine was not listed as suboptimal, its prescription
dropped from 16.5% (Y1) to 9.7% (Y2) in this study.
[Figure 3 ] shows changes in the proportion of prescribed AEDs between Y1 and Y2, including
phenytoin, phenobarbital, valproic acid, lamotrigine, oxcarbazepine, carbamazepine,
clobazam, gabapentin, and levetiracetam. A trend toward the prescription of lamotrigine
(33.6 versus 5.5%) and levetiracetam (29.1 versus 0%) was observed in Y2 compared with Y1, whereas valproate, oxcarbazepine, carbamazepine,
and gabapentin were less often prescribed over time.
Figure 3 Temporal trend in the proportion of inpatients treated with the most commonly used
AEDs.PHT: phenytoin; PB: phenobarbital; VP: valproic acid; LTG: lamotrigine; OXCBZ:
oxcarbazepine; CBZ: carbamazepine; CLB: clobazam; GB: gabapentin; LEV: levetiracetam.
Multivariate analysis showed patients who did not present with convulsive SE were
seven times more likely to be in the appropriate regimen group than those patients
with convulsive SE [1⁄(e-2.0141 )=7.4940] (p<0.001). Patients who did not have remote or progressive etiologies associated
with acute symptomatic seizures were four times more likely to be in the appropriate
regimen group than those patients who did [1⁄(e-1.42784 )=4.1697] (p=0.02).
Eight patients who required drug discontinuation due to severe adverse reactions were
identified and, in these cases, the culprit drug was replaced by another AED; among
patients taking phenytoin, four had idiosyncratic reactions and one of them developed
Stevens-Johnson syndrome. One patient had erythema multiforme in the first few days
of levetiracetam use; two patients exhibited a rash during lamotrigine dose escalation,
and another patient presented with valproic acid-induced hepatotoxicity.
DISCUSSION
There are limited data evaluating prescribing patterns of AEDs in the older adult
population, despite published guidelines and expert opinion recommendations regarding
the treatment of epilepsy in this population[24 ]. The present study investigated changes in prescribing patterns of AEDs that occurred
from 2009 to 2019 in older adult inpatients. The authors hypothesized there would
be a downward trend in the use of phenytoin and phenobarbital, formerly deemed inappropriate
in the previous retrospective study on prescribing patterns of AEDs in older adult
inpatients[12 ], and consequently an increase in the use of newer AEDs over time, typifying an improvement
in the quality of treatment. The decision to categorize an AED regimen as appropriate
or inappropriate (suboptimal) considered the evidence-based clinical recommendations
and expert consensus opinions. These guidelines included phenobarbital and phenytoin,
both enzyme inducers, in the group of inappropriate drugs, because of their unsuitable
safety profile for older adult patients[6 ],[9 ]. Although carbamazepine is also an enzyme inducer, it is well-tolerated by older
adult patients and so it was not included in the inappropriate group[9 ],[11 ],[25 ].
Significant changes were found in prescribing patterns of AEDs for older adult inpatients
with late-onset epilepsy throughout the ten-year study period. Although often utilized,
a significant decrease in the prescription of suboptimal AEDs was observed, which
was due to a considerable decrease in phenobarbital prescription. This consistent
change in the quality of AED prescription coincided with the increasing dissemination
of newer AEDs into clinical practice in Brazil. The modest drop in phenytoin prescription
and the fact it remains the most commonly prescribed AED in this study may be explained
by the characteristics of these hospitalized older adult patients. One third of the
included patients presented with SE, thus requiring treatment with intravenous (IV)
AEDs according to the guidelines on the management of SE. Despite the increased availability
of other IV AEDs in several countries, IV phenytoin remains the drug of choice for
most patients in emergency settings. In the center where this study was conducted
(HSR), phenytoin, phenobarbital, valproic acid, and lacosamide are currently available
in IV formulation. Levetiracetam is not available in IV formulation in Brazil. Phenytoin
was prescribed to 69 patients in initial care, with 27.5% of them requiring replacement
for another AED such as lamotrigine or levetiracetam. The present results are in line
with a previous study on older adult patients, which found phenytoin to be the most
commonly prescribed AED; conversely, a significant reduction in the use of phenobarbital
was observed in the present research[11 ]. Since 1985, data supporting the status of phenobarbital as a suboptimal AED have
been widely cited[26 ]; a subsequent meta-analysis confirmed this notion and recommended against the use
of phenobarbital due to adverse drug effects[27 ]. Despite scientific evidence favoring newer AEDs over older ones, changes in clinical
practice face some obstacles related to prescriber profile, higher cost of new drugs,
and the circumstances in which the AED is used.
A randomized trial of gabapentin, lamotrigine, and carbamazepine in new-onset geriatric
epilepsy has demonstrated these AEDs to have similar efficacy, but has found the discontinuation
rate to be the lowest with lamotrigine and the highest with carbamazepine[25 ]. When compared with lamotrigine and carbamazepine, levetiracetam had the lowest
discontinuation rate among older adults with new-onset epilepsy[28 ]. Moreover, despite similar effectiveness between drugs, levetiracetam had the lowest
discontinuation rate when compared with valproate and carbamazepine among older adult
patients[29 ]. A recent systematic review with meta-analysis concluded lamotrigine was better
tolerated than carbamazepine among older adults with epilepsy; levetiracetam was seemingly
associated with a higher probability of seizure freedom when compared with lamotrigine,
although no difference was found in long-term efficacy and tolerability[5 ]. Levetiracetam and lamotrigine may cause fewer drug-drug interactions than enzyme
inducers, such as phenytoin, carbamazepine, and phenobarbital, and therefore the former
tend to be prescribed more often with advancing age. A revision of guidelines on the
management of epilepsy in England recommends two first-line AEDs (carbamazepine and
lamotrigine) and three alternative AEDs in monotherapy (levetiracetam, oxcarbazepine,
and valproate); however, notwithstanding the substantial increase in the use of newer
AEDs over time, the authors observed persistent widespread use of suboptimal therapy
with phenytoin, primidone, and phenobarbital[30 ].
Multiple comorbidities are common among older adult inpatients, and polypharmacy tends
to be the rule[31 ], which may affect AED choice. Potential drug interactions of statins, antihypertensive
agents, warfarin, and chemotherapeutic agents with phenytoin, carbamazepine, phenobarbital,
and valproate are reasons for concern[32 ],[33 ]. Enzyme inducers may interact with many other drugs commonly prescribed to older
people such as tricyclic antidepressants, selective serotonin reuptake inhibitors,
antipsychotics, and calcium channel blockers[33 ].
Post-stroke epilepsy usually occurs after 5-6 months, and no specific AED is recommended,
although there has been an attempt to use levetiracetam in this setting[34 ].
Limitations and strengths
This study was conducted on a relatively small sample of a heterogenous population
of older adult patients from a single center who had epilepsy, which is also a heterogenous
disease. The patients received specialized care from a neurology team, which could
affect the generalizability of the reported findings to some extent. Additionally,
the present results may not reflect some categories of individuals such as those residing
in nursing homes or other long-term care facilities.
Key strengths of this study include: its prospective design, which allowed the authors
to identify acute symptomatic seizures as well as epileptic and nonepileptic events
and to determine seizure type and etiology according to the ILAE recommendations;
and its focus on the prescription of AEDs to patients who had been diagnosed with
epilepsy.
In conclusion, this study showed a downward trend in suboptimal therapy for older
adult inpatients over a period of time, notably fewer phenobarbital prescriptions.
The use of newer AEDs substantially increased, although phenytoin remains the most
commonly prescribed drug to older adult inpatients. Convulsive SE and some acute symptomatic
seizures were found to be independent risk factors for suboptimal prescription among
this population.
These results suggest ongoing commitment to reducing the prescription of suboptimal
AEDs to older adults, particularly phenytoin, in Brazilian emergency rooms.
