Psychopharmacological Medication Use in Frontotemporal Dementia at the Time of Diagnosis: Comparison with Alzheimer’s Disease

INTRODUCTION

Frontotemporal dementia (FTD) is a neurodegenerative disease consisting of several clinical, genetic, and neuropathological subtypes. The most common clinical phenotype of FTD is the behavioral variant of frontotemporal dementia (bvFTD), followed by the language variants, non-fluent variant of primary progressive aphasia (nfvPPA) and semantic variant of primary progressive aphasia (svPPA), and by the FTD-plus movement disorders, corticobasal syndrome, and progressive supranuclear palsy[1, 2].

To date, there are no disease-modifying pharmacological treatments available for FTD, and recommendations regarding disease management emphasize the use of non-pharmacological treatment options, such as psychoeducation and physical-,speech-, or occupational therapy, as well as symptomatic pharmacological treatments [1, 3]. For Alzheimer’s disease (AD), an approved therapy, i.e., cholinesterase inhibitors, is available and recommended to be used already in the early disease stages [4]. Due to the diverse cognitive, neuropsychiatric, and behavioral symptoms, differential diagnostics especially between FTD and primary psychiatric disorders is challenging and, thus, the psychiatric symptoms are treated with different medications before the eventual diagnosis of FTD [5–7]. Furthermore, the most commonly used pharmacological treatments after the establishment of the FTD diagnosis rely on the use of psychopharmacological medications, including antipsychotics, antidepressants, mood stabilizers, and benzodiazepines [3, 8]. These pharmacological approaches aim to modify especially the burdensome behavioral or psychiatric symptoms, such as agitation, disinhibition, apathy, and psychotic symptoms [3]. However, there is limited knowledge on the actual risk/benefit ratios regarding specific psychopharmacological drug groups in the symptomatic treatment of FTD, and no generally accepted recommendations are available. Thus far, most of the psychopharmacological studies in FTD have been observational, retrospective, or case-series reports using small-sized cohorts [9–13]. Especially prospective comprehensive, randomized double-blind controlled trials are lacking [14]. Altogether, these factors may result in rather diverse practices regarding the off-label psychopharmacological treatments in FTD.

To our knowledge, limited real-world data (i.e., data reflecting actual clinical prescription preferences) are available on the use of psychopharmacological medications in the symptomatic management of FTD at the time of the diagnosis [15–17]. In the present study, we report the extent of psychopharmacological medication use in a large FTD cohort from Finland, which includes data from two separate tertiary centers for neurodegenerative disorders. Parallel data were collected from a cohort of patients with AD, which were used as a neurodegenerative disease reference group in thestudy.

MATERIALS AND METHODS

The study was performed according to the principles of the Declaration of Helsinki. The study protocol was approved by the research ethics committees of Northern Ostrobothnia and Northern Savo Hospital Districts. Written informed consent was obtained from the participants before enrollment.

The study cohort included altogether 222 FTD patients and 214 AD patients. All patients were diagnosed by a specialized neurologist at the dementia outpatient clinics of Kuopio and Oulu University Hospitals (41% of the FTD patients and 35% of the AD patients were from Oulu University Hospital). Of the FTD patients, 155 (69.8%) had probable and 67 (30.2%) definite FTD diagnosis according to the latest criteria (60 definite due to C9orf72 repeat expansion carriership and 7 due to neuropathological confirmation) [18–20]. The C9orf72 repeat expansion status (repeat expansion carrier with >60 repeats or non-carrier with <30 repeats) was available from 181/222 (81.5%) of the FTD patients. Other causal FTD mutations were not tested, as the prevalence of these mutations is extremely low or non-existing in Finnish patients [21–23]. The FTD group included 145 bvFTD, 44 nfvPPA, 9 svPPA, 1 progressive supranuclear palsy, and 23 FTD accompanied with motor neuron disease (FTD-MND) patients. All AD patients had at least probable AD diagnosis based on the criteria by McKhann and colleagues [24, 25]. Family history of both the FTD and AD patients was evaluated and scored according to the Goldman score [26]. The familial FTD group included patients with Goldman score ≤3 and/or FTD patients carrying the C9orf72 repeat expansion. The familial AD group included patients with Goldman score ≤3. Mini-Mental State Examination (MMSE) score was evaluated from the patients at the first diagnostic visit and used as a measure of baseline disease severity. MMSE was chosen as it was most comprehensively available for patients in bothcohorts.

Use of medication at the time of clinical assessment was reported for each patient at their diagnostic visit at the dementia outpatient clinic, and these data were collected retrospectively from the medical records. We included regularly used (daily use) medications at the time of the first visit (medications that had been prescribed by other doctors than the neurologist at the neurological outpatient clinic responsible for the dementia diagnosis) as well as medications that were prescribed during the diagnostic visit. Psychopharmacological medications were divided to subgroups based on drug classifications: 1) Antipsychotics (further subdivided to first-generation and second-generation or atypical antipsychotics); 2) Antidepressants (further subdivided to selective serotonin reuptake inhibitors (SSRI), serotonin and norepinephrine reuptake inhibitors (SNRI), and other/atypical antidepressants); 3) Benzodiazepines and Z-drugs; 4) Mood stabilizers (further subdivided to lithium, valproic acid/valproate, and other mood stabilizers); 5) Melatonin; 6) Cholinesterase inhibitors; 7) Memantine. Medication use was coded as dichotomous variables (yes or no) per each drug subgroup. Drug dosage was not separately assessed. Data for presence or absence of depressive symptoms, psychotic symptoms, and parkinsonism were collected as dichotomous variables. Depressive symptoms required persistent sadness or low mood most of the time. Psychotic symptoms required at least one evident psychotic symptom, i.e., either delusion or hallucination. Positive parkinsonism was defined when at least two symptoms of the following were observed: resting tremor, bradykinesia, rigidity, prominent hypomimia, postural instability, or loss of automatic movements. Prevalence of anxiety was not systematically evaluated in the whole cohort and thus not reported.

Statistical analyses were performed with IBM SPSS Statistics version 27. Dichotomous/categorical variables were compared with chi square test (or Fisher’s exact test when frequencies were less than five). Continuous variables were compared with independent sample t-test. Main categorical comparisons were further evaluated with a logistic regression model and adjusted for age, gender, and MMSE score (for disease severity). Odds ratios (OR) with p values are reported for those comparisons. Normality of the continuous variables was evaluated with visual inspection (histogram graph and q-q plot) before parametric tests. p values ≤0.05 were considered statistically significant.

RESULTS

The characteristics of the study cohort are shown in Table 1.

The most commonly used type of medication was antidepressants in both FTD and AD groups. However, compared to AD, FTD patients received more often antidepressants (26.1% versus 15.0%, p = 0.004 in chi-square, OR = 2.01 with p = 0.008 in adjusted logistic regression), antipsychotics (23.9% versus 9.3%, p < 0.001 in chi-square, OR = 3.15 with p < 0.001 in adjusted logistic regression), and/or mood stabilizers (6.3% versus 1.9%, p = 0.020 in chi-square, OR = 2.93 with p = 0.085 in adjusted logistic regression). Moreover, the use of benzodiazepines and Z-drugs was more common among the FTD compared to AD patients, although the difference was not statistically significant (17.6% in FTD versus 12.6% in AD, p = 0.149 in chi-square, OR = 1.65 with p = 0.087 in adjusted logistic regression). Only cholinesterase inhibitors were more commonly used among AD compared to FTD patients (11.2% in AD versus 1.4% in FTD, p < 0.001 in chi-square, OR = 0.14 with p = 0.002 in adjusted logistic regression). No significant difference was observed in melatonin use between the groups (4.2% in AD versus 3.2% in FTD, p = 0.559 in chi-square). None of the FTD patients and only three AD patients used memantine (0.0% in FTD versus 1.4% in AD, p = 0.117 in Fisher’s exact test) (Table 1). Statistical adjustments for cohort origin (Oulu University Hospital or Kuopio University Hospital) did not significantly affect any of the results.

The number of simultaneously used psychopharmacological medications per each subject ranged between 0–5. In the FTD group, 44.6% of the patients used at least one psychopharmacological medication at the time of the diagnosis compared to 36.0% of AD patients. Polypharmacy was more common in FTD patients: Percentages for the use of only one psychopharmacological drug were 19.8% in the FTD group and 22.9% in the AD group; two simultaneous drugs 14.9% in FTD and 7.9% in AD; three drugs 6.8% in FTD and 3.3% in AD; four drugs 2.3% in FTD and 1.4% in AD; and five drugs 0.9% in FTD and 0.5% in AD. In a regression model with the number of drugs used per each patient as the dependent variable, FTD diagnosis was a significant factor associating with the number of simultaneously used drugs per patient (p = 0.008, B = 0.281, 95% CI = 0.074–0.488), whereas age, sex, or MMSE score did not show significant associations.

Within the FTD group, the use of psychopharmacological medications trended as being more common in female FTD patients compared to male FTD patients (at least one drug was present in 50.0% of the female FTD patients and 39.5% of the male FTD patients, p = 0.115 in chi-square, not significant). When comparing separate drug groups, statistically significant difference was detected only in the use of benzodiazepines and Z-drugs. When compared between female and male FTD patients, use of antipsychotics was detected in 27.8% versus 20.2% (p = 0.184), antidepressants in 26.9% versus 25.4% (p = 0.811), benzodiazepine and Z-drugs in 23.1% versus 12.3% (p = 0.033), and mood stabilizers in 8.3% versus 4.4% (p = 0.227), respectively. When the medication use was examined in patients with different clinical FTD subtypes, bvFTD patients had the highest prevalence of medication use with at least one psychopharmacological drug used by 52.1% of the bvFTD patients. This percentage was 29.5% in the nfvPPA patients, 33.3% in the svPPA patients, and 34.8% in the FTD-MND patients.

No statistically significant differences were observed when comparing the medication use between the C9orf72 repeat expansion carrying and non-carrying FTD patients, although there was a trend of a more common use of antipsychotics in the expansion carriers (28.3% in carriers versus 19.5% in non-carriers, not significant). In contrast, C9orf72 repeat expansion non-carriers showed a trend towards more frequent use of antidepressants (21.7% in carriers versus 27.6% in non-carriers, not significant), mood stabilizers (5.0% in carriers versus 6.5% in non-carriers, not significant), and benzodiazepines (13.3% in carriers versus 18.7% in non-carriers, not significant). The most commonly used first-generation antipsychotic in the whole FTD group was perphenazine (4.5%) and the most commonly used second-generation antipsychotic was quetiapine (5.9%), risperidone (5.9%), or olanzapine (4.5%). The most commonly used antidepressants were escitalopram (4.1%), citalopram (4.1%), and sertraline (1.8%) (from the “SSRI” subgroup) and mirtazapine (9.5%) (from the “other antidepressants”subgroup).

In the FTD group, parkinsonism was more common in subjects receiving antipsychotics (32.1%) compared to subjects not using antipsychotics (10.2%). In a logistic regression analysis adjusted for age and sex, the use of antipsychotic medication independently associated with parkinsonism (OR = 3.01, 95% CI = 1.23–7.26, p = 0.014), antipsychotic symptoms (OR = 7.25, 95% CI = 3.44–15.28, p < 0.001), and depressive symptoms (OR = 2.50, 95% CI = 1.19–5.25, p = 0.015). Use of antidepressant medication associated with depressive symptoms (OR = 10.14, 95% CI = 4.72–21.79, p < 0.001), but not with psychotic symptoms or parkinsonism. Furthermore, the use of benzodiazepine or Z-drugs significantly associated only with depressive symptoms (OR = 2.19, 95% CI = 1.02–4.69, p = 0.044).

DISCUSSION

Prior scientific recommendations for the utilization of psychopharmacotherapy in FTD patients are thus far scarce [8, 27], and little is known about active use of these medications in FTD patients at different disease stages. Here, we report the use of psychopharmacological medications at the time of FTD diagnosis in a real-world data set from a large FTD cohort. Our data show that the use of psychopharmacological medication, including antipsychotics (23.9% versus 9.3%, OR = 3.15), antidepressants (26.1% versus 15.0%, OR = 2.01), mood stabilizers (6.3% versus 1.9%, OR = 2.93), and benzodiazepines (17.6% versus 12.6%, OR = 1.65), is more common in FTD as compared to AD patients. The female gender and bvFTD phenotype, alongside with psychotic or depressive symptoms, were the factors most strongly associating with the use of pharmacotherapy at the time of the diagnosis. Nearly half of the FTD patients (44.6%) were found to use at least one psychopharmacological medication at the time of FTD diagnosis. Polypharmacy, i.e., simultaneous use of more than one drug, was also more common in FTD as compared to AD patients.

To our knowledge, extremely limited amount of data are currently available on the actual (off-label) medication use in the FTD patients worldwide. In 2021, Leroy and colleagues reported similar findings in a French cohort to those in our present study. They found that especially the use of antidepressant (48.0% versus 27.0%), anxiolytic (33.2% versus 23.6%), and antipsychotic (17.5% versus 13.1%) medication was more common in FTD (690 FTD cases of whom 34 were genetic mutation carriers) compared to AD patients [16]. A study from 2010 in California (528 AD patients, 88 bvFTD patients) did not show significant differences in the overall use of psychiatric medication in FTD (48.9%) as compared to AD patients (41.3%), although especially antidepressants were significantly more commonly used in FTD (43.2%) compared to AD (29.9%) patients [15]. In our cohort, FTD diagnosis associated with a more common use of psychiatric medications (especially antipsychotics and antidepressants) compared to AD even after statistical adjustments. In contrast to the findings from California [15], and partly also to those from the French FTD cohort [16], the use of cholinesterase inhibitors (40.9% in California cohort and 12.0% in the French cohort) or memantine (29.5% in the California cohort and 5.7% in the French cohort) was extremely rare in the FTD patients in our present study, reflecting the current consensus, which does not support the use of these medications in the management of FTD [14, 28]. On the other hand, the use of antipsychotic pharmaceuticals was more common in our FTD cohort (23.9%) compared to the 2010 report from California (4.5%) [15], whereas the prevalence of the use of antipsychotics in our cohort was quite similar to that reported in 2021 in the French cohort (17.5%) [16]. These discrepancies could, for example, reflect differences in the prevalence of psychotic symptoms in separate FTD cohorts or in the disease stages of the FTD patients in these studies. Neuropsychiatric symptoms were common in the patients in our cohort and the cross-sectional cut-off for the medicine use was at the early stages of the disease in the present study. This together with the observed high prevalence of the use of psychopharmacological medications may reflect the fact that symptom-driven treatment approaches are often used (in other centers than tertiary memory center) before the eventual FTD diagnosis. Additionally, antipsychotics may have separate indications depending on the dosage. For example, low doses of quetiapine or olanzapine are used for insomnia without having actual antipsychotic effects. Also, the availability of psychopharmacological medications in different countries may vary. Thus, differences in prescription practices may affect the findings. Notably, our study did not include a control population without a neurodegenerative disease as a reference group. However, previous studies have reported approximately 11–20% prevalence for antidepressant use [29, 30] and 3–11% prevalence for antipsychotic use [30, 31] in the elderly populations in Finland. Even though these percentages may not be directly comparable due to different methodologies used in the studies, they may give at least an approximation on the average prescription practices regarding these medications in Finland in the elderly.

Out of the different psychopharmacological drug groups, the most congruent evidence for the improvement of psychiatric symptoms (most often evaluated as NPI score reduction) has been reported in FTD patients using serotonergic medications, mainly SSRI, but also these studies have lacked double-blind controlling [9–11, 14]. Meta-analysis in 2021 concluded that SSRI treatment in FTD patients appears to improve disinhibition, irritability, aggression, and aberrant motor activity across studies, whereas the results have not been as promising related to apathy or loss of empathy [12]. Interestingly, a tetracyclic antidepressant trazodone, in particular, has been suggested to most significantly reduce bvFTD symptoms [27]. However, this medication was not reported in the FTD patients of our present study, at least before or at the time of the diagnosis. Existing supporting research data are even more limited in the case of use of antipsychotics and benzodiazepines or Z-drugs, both drug groups commonly used in our cohort and in the previous ones [15, 16]. Efficacy of antipsychotics in the treatment of FTD patients has been evaluated mostly in case report or series studies, and there is a known risk for increased mortality and extrapyramidal side effects when using these medications [3]. In the present study, parkinsonism was more commonly detected in FTD patients receiving antipsychotics, which could at least partly indicate antipsychotic-induced parkinsonism, although also other reasons for this association are possible. However, the link between increased extrapyramidal symptoms and use of antipsychotics underlines the need for a careful risk/benefit assessment when prescribing these pharmaceuticals. Risk/benefit ratios of benzodiazepines have not been studied in FTD patients, but in general, benzodiazepine use in AD or dementia patients is known to associate with adverse effects [32]. Even though there is some clinical evidence from small-sized case-series studies suggesting that valproic acid [33] and lithium [13] may have beneficial effects in FTD patients, these medications were rather rarely used in our cohort (prevalence of 2.3% for both lithium or valproic acid use). In summary, the medication profile in our clinical FTD cohort in the diagnostic phase mainly follows the current, although limited, recommendations for treating FTD, as the antidepressants was the most commonly prescribed drug category and the use of cholinesterase inhibitors or memantine was nearly nonexistent in the FTD patients. On the other hand, the use of antipsychotics and anxiolytics was relatively common in our present cohort, despite the limited or discouraging efficacy studies. The high prevalence of the use of these medications could at least partially be explained by utilization of pre-diagnostic psychiatric treatment approaches in FTD patients and, thus, does not necessarily reflect prescription practices in the tertiary memory centers.

The strengths of the present study include the large and well-characterized FTD patient cohort containing systematically collected medication data at the early diagnostic phase. Data were also collected from two independent Finnish centers from different geographical areas, which reduces the risk for bias caused by specific prescription practices in one center. A limitation in the study is that no data are available regarding longitudinal use of these medications after the FTD diagnosis and, thus, the results merely reflect the use of these medications at the early or diagnostic phase. Moreover, no conclusions on the efficacy of these medications can be made based on our study due to its cross-sectional nature. We used MMSE score as a measure of baseline disease severity due to the fact that these scores were available from most of the FTD and AD patients throughout the cohort, although we acknowledge that MMSE is not the best measure of disease severity especially in FTD patients.

Several conclusions and future directions can be drawn based on the present study. First, the use of off-label psychopharmacological medications, especially antipsychotics, antidepressants, and mood-stabilizers, is more common in FTD compared to AD patients at the time of diagnosis. In particular, female gender and bvFTD phenotype, alongside with depressive and psychotic symptoms, are associated with the use of these medications among the FTD spectrum patients. Nearly half of the FTD patients used at least one psychopharmacological drug at the time of the diagnosis, and polypharmacy was also common. While our data describe the medication use only at the early diagnostic stages of the disease and may reflect the prevailing challenges in differential diagnostics and symptomatic treatment prior to the eventual and accurate diagnosis of FTD, it still highlights the considerably high prevalence of the use of psychopharmacological drugs among these individuals. Considering the thus far extremely limited research data on the assessment of actual risk/benefit ratios of these medications specifically in FTD patients, our findings from the early disease stage highlight the great need for prospective studies, which will evaluate the prevalence of the use and efficacies of symptomatic pharmaceutical treatments at different disease stages in patients of the heterogenous FTD spectrum.