Introduction

Apathy is a common behavioral syndrome in neuropsychiatric disorders with prefrontal cortex (PFC) and basal ganglia (BG) pathology, such as Huntington’s disease (HD) [1, 2]. It is broadly defined as the primary absence of motivation, lack of initiative and drive, as well as emotional indifference [3]. Apathy can be divided into three major syndrome domains—deficient emotional-affective function, cognitive function, or auto-activation [2].

In HD, apathy is the most common neuropsychiatric syndrome that correlates directly with disease progression [4–6]. Loss of dopamine (DA) receptor expression in fronto-striatal circuits was proposed as a key pathophysiological mechanism of apathy in HD [7, 8]. Neurodegeneration begins in the striatum as early as 15 years prior to motor onset, and then extends to frontal and PFC cortex regions [9–11]. Pathological changes in the orbital and medial PFC and the projections to limbic brain regions, predominantly the ventral striatum (VS), have been associated with the development of apathy in HD [2]. Magnetic resonance imaging (MRI) is capable of measuring atrophy [12–14] as well as alterations in mesolimbic DA processes [15, 16], which are linked to anticipation and processing of reward or punishment. In premanifest HD patients, an aberrant ventral striatal response during a monetary incentive delay task has been observed [17].

Despite of the high prevalence and disease burden of apathy in HD, research on therapeutic options for apathy is rare, and no effective treatment is at hand [18, 19]. This is the first controlled trial (CT) on the treatment of apathy in HD. It was the aim of this trial to evaluate the efficacy and safety of bupropion in the treatment of apathy in HD. We chose the antidepressant bupropion for its mode of action of blocking norepinephrine and DA reuptake, thereby potentially increasing DA neurotransmission in areas relevant for apathy. In addition, several single case reports and results of small series suggested the effectiveness of bupropion for the treatment of apathy in HD and other neurodegenerative diseases [20–22]. In addition, we investigated the effect of bupropion on DA-associated reward processing in an established gambling task using fMRI [23, 24].

Materials and methods

ACTION-HD (Apathy cure through Bupropion in Huntington’s disease) is a multi-center, randomized, double-blind, placebo-controlled, 2x2 crossover phase 2b investigator-initiated trial (IIT) that was conducted at four sites in Germany between May 2012 (recruitment of first patient) and May 2014 (last patient leaving the trial). The ACTION-HD trial was registered at the EudraCT clinical trial register (EudraCT number 2009-013698-16) on 24^th March 2011 prior to inclusion of the first patient. We later registered the trial at clinicaltrials.gov. The authors confirm that all ongoing and related trials for this drug/intervention are registered. The protocol for this trial is available as supporting information; see S1 Clinical trial protocol.

Results

Between May 2012 and October 2013, 40 apathetic HD patients were randomized into two treatment groups (20/20) Fig 1. The subject characteristics are summarized in Table 1. Eighteen study participants consented to participate in the adjunct MRI study. The last study participant finished the study in April 2014. The study database was locked on May 30, 2014. Other than duration of disease (group 1: 8.4 ± 3.6 years vs. group 2: 5.5 ± 2.4 years; p<0.05), there were no statistically significant treatment group differences for any demographic parameter or baseline values of primary or secondary outcome variables Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT flow diagram.

https://doi.org/10.1371/journal.pone.0173872.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic data and neuropsychiatric profiles (full analysis set).

https://doi.org/10.1371/journal.pone.0173872.t001

A first important observation of the study was that the study participants rated their symptoms of apathy (AES-S) as significantly less severe than the caregivers (AES-I; t (df = 38) = -6.0504, p<0.001) or the clinical investigators (AES-C; t (df = 38) = -5.6101, p<0.001) Table 1.

Due to major protocol violations, two study participants had to be excluded from the per protocol analysis, because one did not follow the washout and crossover medication sequence between visit 4 and visit 4a, and the other took less than 80% of the study medication during the treatment period. Four participants discontinued the study; two for unwillingness to resume study visits, and one because of increased irritability while on bupropion. A fourth participant discontinued the study due to mood changes (cf. severe adverse events (SAEs), participant five).

As for the safety and tolerability of bupropion, 21 of 40 study participants had at least one of 56 AEs. Five AEs were rated as severe and 11 AEs were rated as moderate. The five SAEs resulted in hospitalizations: two after a fall, in one case with a shoulder fracture and an ensuing depression, a third for an elective varicose vein stripping, all on placebo. The fourth participant with a SAE was hospitalized after a short episode of lack of responsiveness three days after the ending of bupropion treatment (visit 4; S1 Fig). A fifth participant was hospitalized due to an increased instability of mood, irritability and aggression which evolved one day prior to the tapering of bupropion and while on concomitant medication with venlafaxine. The 11 moderate AEs included a mild to moderate increase in chorea and gait disorder (7 on bupropion/4 on placebo), an increase in irritability (1 on bupropion/6 on placebo), concentration difficulties (3 on placebo), a sleeping disorder (2 on bupropion), agitation and restlessness (1 on bupropion/1 on placebo), and an increase of perseverative thinking (1 on placebo). One study participant developed disinhibition, another hyperphagia with an increase of 20 kg of body weight while on bupropion. No participant developed active suicidality (answer “yes” to questions 4 and 5 of the “suicidal ideation” section of the C-SSRS), or indicated suicidal ideation upon questioning by the clinical investigators.

Bupropion does not change apathy in HD as quantified by the AES-I

Comparing the primary outcome variable between the treatment groups using the linear mixed model adjusting for baseline, there was a slight but statistically non-significant change of the AES-I score due to bupropion treatment, which amounted to -0.81 (95% CI = -3.12, 1.52) in the full analysis set, and to -0.09 (95% CI = -2.56, 2.39) in the per protocol set Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. No change of apathy, irritability, cognition or function after 10 weeks of treatment with bupropion.

https://doi.org/10.1371/journal.pone.0173872.t002

In any case, this level of change had no clinical relevance (defined as a change of 10 points or more on the AES, cf. Materials & methods). The results did not change when using multiple imputations to take missing values into account. Including random effects for the site produced a corresponding variance of random effects equal to zero, showing that there was no relevant heterogeneity between sites. When we grouped the AES questions according to the three major apathy syndrome domains (cognitive, behavioral, emotional) and analyzed them separately, we still failed to observe a significant treatment effect on any of the three apathy domains. However, when looking at each AES question separately, a significant treatment effect was observed for question 2 “he/she get things done during the day” (-0.22 [95% CI = -0.44, 0.01; p<0.05]).

Trial participation/placebo effects alleviate symptoms of apathy

Irrespective of the intervention (bupropion or placebo), informants and clinical investigators observed a significant improvement of the symptoms of apathy (AES-I, AES-C) or the apathy syndrome (NPI-Apathy, UHDRS-Apathy) predominantly during the first treatment period Table 3. Notably, this was not observed in the AES-S scores.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Change of apathy after the first treatment period suggests activation through trial participation.

https://doi.org/10.1371/journal.pone.0173872.t003

Study participants also rated their symptoms of apathy as significantly less severe than the caregivers or the clinical investigators Table 1. There was also a tendency of a period effect when grading irritability by the UHDRS-Behavior, the SIS or the NPI. However, the observations were not consistent between instruments, and the changes in irritability scores were too low to have any clinical relevance (data not shown).

Discussion

ACTION-HD is the first randomised controlled clinical trial on the treatment of apathy in HD. We did not find a clinically significant effect of a 10-week treatment course with bupropion (at a dose of 300 mg per day) on the severity of apathy in HD based on assessments by informant, study participant and clinical investigator using the AES. Two further informant-based assessment tools (NPI-Apathy and UHDRS-Apathy) also failed to reveal a treatment effect of bupropion.

A concern when designing this trial was the lack of robust longitudinal data on the severity of apathy in HD, since no ‘gold standard tool’ for the assessment of apathy exists in general [30], and for HD in particular. We chose the SCIA-D as a screening instrument for the diagnosis of apathy, and the AES as a baseline and follow-up tool to measure the severity of apathy. The SCIAD-D is based on the criteria of Starkstein et al. [25], which were recently incorporated into the consensus criteria for the diagnosis of apathy proposed by an international task force [31]. The criteria refer to the prior functional status of the study participant, which we judged to be more appropriate than an arbitrarily determined baseline cutoff score. The AES is the most widely used rating scale for the symptoms of apathy. Clinical studies have also relied on single item measures like the UHDRS-Behavior or the NPI. Only recently, the AES has been validated in HD [32]. Importantly, confounding factors like lack of awareness [33], comorbid depression, motor symptoms, or side effects of concomitant medication have to be considered when evaluating apathy in HD.

The reasons for choosing the AES as the primary rating scale in the ACTION-HD trial included i. the availability of a validated German version [34], ii. the possibility of separate ratings by participant, informant and clinician, iii. the good inter-rater reliability for the AES-C [35], iv. the separate evaluation of all three domains of apathy [2, 26], v. the provision of performance guidelines by the author, allowing for further standardization [27], vi. the suitability of the AES for statistical evaluation due to the 4-point Likert scale, vii. the reasonable length and complexity, allowing for informative answers by cognitively impaired study participants. Based on previous reliability studies in AD, the test-retest reliability is 0.94 for the AES-I, 0.88–0.89 for the AES-C, and 0.76 for the AES-S [27, 36]. Power calculations were performed using longitudinal AES data from a recently published HD cohort [32]. AES threshold scores have been suggested for AD with a range from ≥ 30 to ≥ 41.5 [37], but also for conditions like traumatic brain injury [38]. Baseline scores of some of our patients were below these values Table 1, but there was no doubt about the presence of apathy in these study participants based on the SCIA-D as well as on the clinical assessments. Moreover, we found a mean NPI apathy score of 8.6 at baseline in our study. NPI apathy scores above 4 have recently been suggested to define clinically significant apathy in dementia [39].

We chose the informant version of the AES, the AES-I score, as the primary outcome variable since we assumed that HD patients rate their symptoms of apathy less severe than their caregivers or physicians [40, 41]. In HD, like in other neuropsychiatric diseases with frontal pathology, lack of insight into symptoms and functional capacities is common [42, 43]. Marin originally evaluated the AES in several diseases, including AD [26, 27], where he observed that patients graded the symptoms of apathy as less severe than their caregivers or clinicians. An association with frontal dysfunction has been suggested also for AD [44]. On the other hand, one has to consider that caregivers may be biased by the burden that the patient’s apathy bestows on them [45]. As expected, participants of Action-HD rated their symptoms of apathy as significantly less severe than their caregivers or the clinical investigator. In addition, a significant period effect was observed for AES-I and AES-C but not for AES-S scores Tables 1 and 3. In contrast to these findings, Mason and Barker at the University of Cambridge (UK) did not find a significant difference between AES-I and AES-S sores during an observation period of 19 months on average [32]. An explanation for this discrepancy could be that we physically separated patients and informants when performing the AES, as we previously observed that patients may adapt their judgements to the ratings of the caregivers when assessed in the same setting. In addition, the relative number of participants with a TFC score between 7 and 10 was small in the Cambridge study, whereas the mean TFC scores in our study were 7.4 and 8.8 in the intervention groups, respectively. Of the 40 patients rated as apathetic by their informants, 21 were lost to follow up in the Cambridge study, which creates a bias. Moreover, we excluded patients with dopamine-depleting drugs, and patients with comorbid depression in our study in order to reduce confounding effects. This has also recently been recommended by Cummings and coworkers [39]. However, we do acknowledge that HD patients with both depression and apathy may constitute a subgroup that responds better to bupropion than HD patients with apathy alone.

To quantify the affective-emotional components of apathy we additionally used an established gambling task to measure reward perception as one of the major preconditions for motivation [24, 29, 46]. DA is thought to be involved in reward and motivation, possibly through modification of the mesolimbic and mesocortical pathways, as well as the orbitomedial PFC-striatal projections. The loss of DA receptor 1 and 2 expression in fronto-striatal circuits was proposed as a key pathophysiological mechanism of apathy in HD [7, 8]. Concurrent with the clinical results, we did not detect significant effects of bupropion treatment on brain responses and functional connectivity during reward processing in apathetic HD patients in our exploratory fMRI analysis. However, it is difficult to draw any conclusion from the small sample size.

When considering the study duration, we feel that 8 weeks on the full dose of bupropion should have been sufficient for the detection of a potential treatment effect. The duration of most randomized CTs on apathy in neurodegenerative diseases is between 6 and 12 weeks. In fact, Cummings et al. recommended a study duration of 8 to 12 weeks when designing clinical trials for apathy in neurodegenerative diseases since the worsening of apathy in both the verum and placebo arms may result in an underestimation of an initial treatment effect in the intervention arm if the intervention period would be extended beyond 12 weeks [39].

It is possible that bupropion did not reach sufficient concentrations within the central nervous system at a daily dose of 300mg. However, we felt that daily doses of 450 mg could be problematic in advanced stages of HD as the risk of seizures, suicidality and irritability may increase.

HD patients in more advanced disease stages often require symptomatic treatment [19]. We cannot exclude the possibility that concomitant medication in some of the study participants may have aggravated apathy or mitigated the effect of bupropion. This is particularly true for tiapride, a dopamine D2 and D3 receptor antagonist, which is commonly used to treat chorea in Germany. A type 2 error may also result from low sensitivity of the available instruments for grading apathy. In the case of the AES, the poor differentiation between the answers 2 (“slightly characteristic”) and 3 (“somewhat characteristic”) on the four point Likert scale may reduce sensitivity. It is noteworthy that the validation of the AES in HD is still incomplete [41]. Our trial adds important data for this process.

Many recent drug trials in HD, e.g. the MermaiHD study [47] and the deutetrabenazine trial [48], have detected significant placebo effects. Interestingly, Cubo et al. observed a significant placebo effect on behavior, but not on motor symptoms or cognition in a large multicenter drug trial in HD [49]. This effect persisted over 3 years and was prominent if depression was absent. In Action-HD, we observed an improvement of the primary outcome measure AES-I for all study participants predominantly prior to crossover. This was replicated for the clinician-rated AES-C score. Study participants were informed about the crossover design. Since we compared period 1 and 2 data with the original baseline, the results suggest, that study participation/placebo effects have a positive impact on the severity of apathy. (cf. Table 3). Notably, the study participants themselves did not rate any improvement on the AES-S. These important findings indicate that our study was not underpowered to detect significant changes of apathy using the AES and that, when studying apathy in HD, activation by study participation should be considered when interpreting trial results.

Conclusion

The ACTION-HD trial suggests that bupropion does not influence the severity of apathy in non-depressed HD patients as quantified by the AES and other behavioral assessment tools. We observed a significant period effect irrespective of the intervention (bupropion or placebo). These findings highlight the general need for controlled symptomatic treatment trials in HD.

Supporting information

Acknowledgments

The authors thank Sarah L. Mason and Roger A. Barker (Centre for Brain Repair, University of Cambridge, UK) for providing preliminary AES data for the power analysis. We also thank David Craufurd (University of Manchester, UK) for sharing his experience with assessments in HD. We are indebted to Katrin Barth, Carolin Geitner, Sonja Trautmann, Barabara Kaminski, Christiane Stamm and Ralf Reilmann from the Huntington Centres in Ulm, Bochum and Münster for their assistance with the trial. We thank Andrea Hassenpflug, Yvonne Kamm and Stefan Hetzer (BCAN, Berlin) for their expert assistance with neuroimaging. Last but not least, we thank the many participants and caregivers who generously agreed to participate in this trial. The authors also thank those at HSC/HSG and EHDN, who did not lose confidence in the finalization of this first IIT on the symptomatic treatment of neuropsychiatric symptoms in HD.