Abstract
Homotaurine supplementation may have a positive effect on early Alzheimer’s disease. Here, we investigated its potential neuroprotective effect on the hippocampus structure and episodic memory performances in amnestic mild cognitive impairment (aMCI). Neuropsychological, clinical, and neuroimaging assessment in 11 treated and 22 untreated patients were performed at baseline and after 1 year. Magnetic resonance data were analyzed using voxel-based morphometry to explore significant differences (Family Wise Error corrected) between the two groups over time. Patients treated with homotaurine showed decreased volume loss in the left and right hippocampal tail, left and right fusiform gyrus, and right inferior temporal cortex which was associated with improved short-term episodic memory performance as measured by the recency effect of the Rey 15-word list learning test immediate recall. Thus, homotaurine supplementation in individuals with aMCI has a positive effect on hippocampus atrophy and episodic memory loss. Future studies should further clarify the mechanisms of its effects on brain morphometry.
Keywords
INTRODUCTION
Alzheimer’s disease (AD) is characterized by progressive cognitive and functional impairment and neural degeneration in specific brain areas, such as the hippocampus and entorhinal cortex (for typical presentation), and frontal, parietal, or occipital cortices (for atypical presentations). Classically, hippocampal atrophy has been considered a reliable imaging marker of typical AD and included in research diagnostic criteria [1]. More recently, however, updated criteria have been proposed (the International Working Group -IWG- 2 criteria) [2]; namely, it has been suggested that in vivo evidence of decreased amyloid-β (Aβ) together with increased Total-tau or Phosphorilated-tau in cerebrospinal fluid (CSF), increased tracer retention on amyloid positron emission tomography (PET), and AD autosomal dominant mutation present in PSEN1, PSEN2, or AβPP, should be considered the core pathological features of typical AD. Within this scenario, downstream topographical biomarkers of the disease, such as volumetric magnetic resonance imaging (MRI) and fluorodeoxyglucose PET, might better serve in the measurement and monitoring of the disease course and in characterizing clinical phenotypes, being less useful for diagnostic purposes as previously thought. As for typical presentation, specific clinical phenotype should also be considered, such as the presence of an early and significant episodic memory impairment including: (1) a gradual and progressive change in memory function reported by patient or informant over more than 6 months, and (2) an objective evidence of an amnestic syndrome of the hippocampal type, based on significantly impaired performance on an episodic memory test. Numerous studies have shown that a deficit in episodic memory is also the definitive hallmark of amnestic mild cognitive impairment (aMCI) [3] which, in turn, is considered a putative early clinical state of AD.
In the past decade, several compounds have been proposed as possible disease modifiers for AD [4]. Previous data suggested that homotaurine, a small aminosulfonate compound which is present in different species of marine red algae, might provide protection against progressive memory impairment [5] by slowing neurodegeneration; this is substantiated by its effect on hippocampus atrophy [6]. Despite positive results in AD patients, no data are available on the potential effect of homotaurine on brain morphology and cognitive performance in subjects at higher risk of developing the typical presentation of AD, i.e., aMCI patients. Thus, the primary aim of this study is to investigate its effect on hippocampal volume loss and episodic memory impairment in aMCI. A secondary aim is to explore the neuroprotective effect of the compound in brain areas other than the hippocampus. We predicted decreased volume loss and improved episodic memory performance in supplemented patients in comparison with untreatedpatients.
MATERIALS AND METHODS
Subjects
In the present study, we included a convenience sample of 33 patients who underwent the first diagnostic assessment for memory problems in the Santa Lucia Foundation outpatient memory clinic in Rome. All patients meet the criteria for aMCI [3, 7] and had a Clinical Dementia Rating (CDR) score of 0.5 [8]. No patient had taken antidementia drugs lifetime, or psychotropic drugs (i.e., antidepressants, benzodiazepines and antipsychotics) in the previous 12 months. Detailed inclusion criteria for aMCI patients are described in a previous paper [9]. In particular, impaired performance on at least one memory test in association or not with impaired performance in at least one additional cognitive domain (i.e., praxis, attention, language, and executive functions) in the absence of functional impairment was required for a diagnosis of aMCI. We referred to the normative data and cutoff scores for the Italian population on neuropsychological tests (see Neuropsychological Assessment section) [10]. Exclusion criteria were: 1) major medical illnesses and comorbidity of primary psychiatric or neurological disorders (to identify a homogenous group of aMCI patients and to ensure the reliability and short-term stability of diagnosis, subjects with mood disorders were excluded, given that a clinically significant improvement in cognition may be observed after depression is improved. Clinical examination was used to exclude patients with cognitive deficits secondary to somatic disorders such as unbalanced diabetes, heart disease, vitamin B12 and folate deficiency, or other major medical illnesses causing secondary cognitive impairment); 2) drug/alcohol dependence or abuse; 3) lack of a reliable informant, defined as someone who was able to ensure patient’s compliance with assessment procedures and who contacted the patient at least twice weekly, with one contact being a personal visit; 4) any potential brain abnormality at T1 and T2 weighted and DWI scans and microvascular lesion apparent on conventional FLAIR-scans; in particular, the presence, severity and location of vascular lesions was computed according to the semi-automated method recently published by our group [11]; and 5) a Hachinski score higher than 4 [12].
Patients were naturalistically (not randomly) in-cluded in the study based on the decision of the clinician who, however, was not aware of the aims of the study. Thus, 11 of the 33 aMCI patients included were treated with homotaurine (aMCI-T) tablets, 50 mg, QD for 2 weeks and BID for the next year, and 22 were untreated (aMCI-NT). At baseline, the two groups did not differ for age, educational level, gender, or global cognitive level as measured by the Mini-Mental State Examination (MMSE) [13] (see Table 1). Patients’ outcome was assessed by raters who were blind of the treatment they had received. The risk of ascertainment bias was minimized by basing efficacy outcomes on objective variables.
The study was part of the MCI-DA protocol approved by the local ethics committee of the IRCCS Santa Lucia Foundation. Written informed consent was obtained after presentation of the nature and purposes of the study to both patients and their informants.
Neuropsychological, behavioral, and functional assessment
A trained senior research psychiatrist (GS) made the diagnosis of aMCI and two trained post-doc neuropsychologists made the cognitive assessment. The acceptable inter-rater reliability level for the present study was set at k > 0.80 for the diagnostic procedure and the neuropsychological-neuropsychiatric assessment. The MMSE was administered to obtain a global index of cognitive impairment at baseline and the Mental Deterioration Battery (MDB) [10] to measure performance in specific cognitive domains at baseline and the1-year follow-up. In particular, the MDB was administered to make a more thorough cognitive examination at the diagnostic level, assessing logical reasoning (Raven’s Colored Progressive Matrices, RCPM), language (Phonological Verbal Fluency, PVF), information processing speed (Stroop test-word reading, ST-WR), resistance to interference (Stroop test color-word interference, ST-CWI), concept formation and set-shifting (Modified Wisconsin Card Sorting Test-Perseverative errors, MWCST-PE), constructional praxis and visuospatial abilities (Rey-Osterreith Complex Figure Test immediate copy, ROCFT-IC) [14]. The above-mentioned tests were administered to diagnose the aMCI disorder and to exclude the overt diagnosis of dementia. Neuropsychological test scores (means and standard deviations) are shown in Table 2.
In order to measure episodic memory performance, which is crucial for the main aim of the present study, Rey’s 15-word list learning test (RWLLT) wasadministered. In this task, participants are given a list of 15 unrelated words that are repeated in five different trials and asked to recall them in any order immediately afterwards (immediate recall (I-RWLLT); score range: 0–75). After a 15-minute interval, during which non-verbal tasks are given, the patient is asked to recall, without list repetition, as many words as possible in any order (delayed recall (D-RWLLT); score range: 0–15). The task allows characterizing patterns of performance relative to the position of the items in the word list, because recall accuracy varies as a function of the item position in the study list, i.e., it is greater for words at the beginning (primacy effect) and the end (recency effect) of the list compared to the mid-list (intermediate) items [15]. Specific scoring of word-list recall data for serial position has been suggested to improve discrimination of normal aging fromdementia. In particular, the number of correctly recalled words in the early list positions (from word 1 to word 5), the intermediate list positions (from word 6 to word 10), and the number of correctly recalled recency items (from word 11 to word 15) was calculated. The latter effect is the most efficient measure of the short-term component of episodic memory, as recency items are stored in a short-term phonological buffer and probably still present in working memory when recall is solicited [16].
The Neuropsychiatric Inventory [17] was used to measure the frequency and severity of neuropsychiatric symptoms. Functional impairment was evaluated by assessing instrumental activities of daily living (IADL) [18]. The IADL includes abilities that allow a person to live independently (e.g., food preparation, housekeeping and laundry, managing financial matters, shopping, and using a telephone). When the ability is completely lost a score of 0 is assigned.
MRI acquisition and processing
Both aMCI-T and aMCI-NT patients underwent the same imaging protocol, which included 3D T1-weighted, T2-weighted, and FLAIR sequences, using a 3T Allegra MR imager (Siemens, Erlangen, Germany) with a standard quadrature head coil. Whole-brain T1-weighted images were obtained in the sagittal plane using a modified driven equilibrium Fourier transform sequence (TE/TR = 2.4/7.92 ms, flip angle 15°, voxel size 1 × 1 × 1 mm3) (MDEFT). Participants underwent the entire MRI protocol when they were included in the study (baseline, T0) and at the 1-year follow-up (T1).
Data preprocessing and analysis was performed with the VBM8 toolbox (http://dbm.neuro.uni-jena.de/vbm), which is incorporated in the SPM8 software (http://www.fil.ion.ucl.ac.uk/spm/) running on MATLAB® (Mathworks). Preprocessing of the imaging data prior to the statistical analysis was carried out using a specific batch for longitudinal data as implemented in VBM8 (no modulation was used because it is not necessary in longitudinal designs where the focus is on relative differences between the same objects).
Individual images were first aligned to a T1 template in MNI-space (Montreal Neurological Institute) to bring them into a common reference frame with respect to translation and rotation. A mean image was calculated from these realigned images and a first realignment of raw data followed, enclosing this mean image as a reference. At this stage individual images were bias-corrected to account for signal inhomogeneities. The resulting mean image was segmented into grey matter (GM), white matter (WM), and CSF. The segmentation procedure was further refined by accounting for partial volume effects [19] with adaptive maximum a posteriori estimations and by applying a hidden Markov random field model [20]. The segmented GM images were nonlinearly normalized using Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) [21]. The resulting normalization parameters were applied to the realigned and bias-corrected single images. Afterwards, these images were segmented and a second realignment followed. Smoothing of GM segments was done using an 8 mm FWHM Gaussian kernel (suitable for relatively small samples sizes [22]), and data were passed to the statistical analysis.
Statistical analysis: Clinical data
Demographic data were compared using Student’s t-test for age and educational level and the chi-square test for gender. Unpaired t-tests were also performed to evaluate potential differences at baseline in cognitive functioning.
The effect of treatment on patient performance in the 15-word list learning test (and more specifically on the recency effect) was assessed by means of repeated measures Analyses of Variance (ANOVAs) using time (baseline versus 1-year follow-up) as within-subjects factor and group (aMCI-T versus aMCI-NT) as between-subjects factor, with two levels. Indeed, our main aim was to look at the interaction between the “time” and “group” factors in order to test the hypothesis of a possible improvement in performance only in aMCI-T individuals. As to emphasize the clinical significance of possible treatment-related changes in neuropsychological performance, differential scores (T1-T0) were calculated for cognitive tests and delta scores submitted to non-parametric tests of groupdifferences (Mann-Whitney U-test).
Statistical analysis: Neuroimaging data
We used a repeated measure ANOVA with a flexible factorial design with “subject” and “time” as within factors and “group” (aMCI-T versus aMCI-NT) as between factor. To infer a significant difference between groups over time (i.e., more volume reduction between time points in the aMCI-NT group with respect to the aMCI-T group), a t-contrast was formulated comparing GM segments of both groups at time point 1 against those of the baseline (i.e., T0 versus T1). Analyses were performed using two different approaches. First, we tested the a priori hypothesis that aMCI-NT patients would show a greater decrease in hippocampus volume between time points compared to aMCI-T patients. To this aim, analyses were performed using left and right hippocampi derived from a standard template [23] as Regions of Interest (ROI). Second, we performed a whole-brain analysis using an absolute threshold masking; a threshold of 0.2 was applied to avoid possible edge effects between different tissue types. The statistical threshold for these contrasts (i.e., ROI and whole brain) was set to the level of p < 0.05, family-wise error (FWE) corrected to account for multiple comparisons and exclude a false positive Type I error.
Finally, we extracted mean volumetric values of the areas where significant effects were found for eachsubject using an in-house shell script. These values were subsequently used to create bar plots and to investigate the relationship between time point volume loss and clinical data by means of univariate correlation analyses.
RESULTS
Clinical data
As shown in Table 2, the two groups of treated and untreated patients did not differ in terms of cognitive performance at baseline. The only significant effect was found on the recency effect at 1-year follow up.
Table 3 shows longitudinal values of the three subcomponents of the I-RWLLT in aMCI-T and aMCI-NT individuals. There was a statistically significant time by group interaction for the recency effect. In particular, aMCI-T individuals’ performance had improved as compared to aMCI-NT subjects, who conversely showed decreased performance. Accordingly, the analysis performed on delta scores revealed a significant effect of homotaurine only on the number of correctly recalled recency items. The effect magnitude (as expressed by the Cohen’s d metric) was 1.33 (95% Confidence Interval 0.54–2.12; variance 0.16) suggestive of a large to very large difference in the immediate memory performance (recency effect) after homotaurine supplementation. On the contrary, no significant results were found for the Primacy and Intermediate effect of the I-RWLLT.
Neuroimaging data
Table 4 summarizes the neuroimaging results found using the whole brain analysis approach.
First, ROI analysis showed a higher volume loss between time points in the aMCI-NT sample with respect to the aMCI-T group in the bilateral hippocampus. This increased volume loss of aMCI-NT individuals was located in the hippocampus tail (Fig. 1, Table 4).
Second, whole-brain analyses revealed three clusters of GM where aMCI-NT patients showed significantly increased volume loss with respect to aMCI-T patients. The first two clusters were located in the fusiform gyrus bilaterally and, in agreement with the ROI analysis, extended to the tail of the hippocampus; the third cluster was located in the right inferior temporal cortex (Fig. 2, Table 4).
DISCUSSION
The present study aimed at investigating the neuroprotective effect of homotaurine supplementation on hippocampal volume changes and episodic memory impairment in a sample of patients diagnosed with aMCI. The main results of this study support our hypothesis that individuals with aMCI who are treated with homotaurine for 1 year may have a beneficial improvement in the short-term component of episodic memory as compared to untreated subjects. Moreover, this increased performance in memory functioning is mirrored by reduced volume loss in the bilateral hippocampus tail and in specific brain areas other than the hippocampus, such as the bilateral fusiform gyrus and right inferiortemporal cortex.
With regard to memory functioning, several studies suggest that deficits in delayed recall and lack of benefit from cues in cued recall and recognition tasks are the most important indices of memory loss in MCI and a risk factor for developing AD [24, 25]. When the pattern of performance relative to the position of items in the word list was explored, data showed that subjects with MCI scored worse than normal controls but better than AD patients on the end-list items [25]. A tentative explanation of the observed phenomenon is that, as a result of repeated learning trials, normal subjects are able to increase their recall of words at the end of the list by storing them in long-term memory. In contrast, MCI patients, like AD patients, can recall the last words on the list only by using short-term memory.
Apart from data coming from delayed free-recall tasks, in which no words are recalled from immediate memory after a delay, some investigators [26] showed that at the first trial, mild and very mild AD groups lacked a significant primacy effect but had a robust recency effect in the immediate recall of a 16-word list. Similarly [27], MCI patients showed diminished primacy and intact recency effects compared to normal controls in the CERAD Word-List task, which consists of a list of 10 unrelated words read aloud by subjects in each of three trials with immediate recall after each trial. The words are in a different order in each trial, which allows comparing serialposition recall dissociated from learning in previous trials. Also asymptomatic persons at risk for AD by virtue of family history, exhibit the same serial position pattern observed in mild AD, but less severe, suggesting greater reliance on immediate as opposed to episodic memory [28]. As normal aging does not affect the basic shape of the serial position curve, the observed pattern (an exaggerated recency compared with primacy effect) might be a marker of cognitive vulnerability to AD. Indeed, since a left hippocampal lesion preferentially produces the same pattern [29], presumably by adversely affecting episodic memory and increasing reliance on working memory, the distinctive serial position effect observed in persons at risk for AD and along the continuum from MCI to AD, likely reflects the decline in hippocampal function that characterizes the disorder [30]. In this sense, our finding of a beneficial effect of homotaurine on the short term component of episodic memory probably indicates its positive action on the pathological hippocampal volume reduction described in patients diagnosed with aMCIor AD.
Indeed, results of several studies report hippocampal atrophy as an important marker for AD, because incident AD cases present abnormalities in medio-temporal brain areas up to 5 years before diagnosis and this neurodegenerative process seems to progressively reach the temporo-parietal cortices in the AD preclinical phase [30]. Moreover, Chételat and colleagues found that hippocampus abnormalities are specifically associated with rapid conversion from MCI to AD [31]. Actually, previous longitudinal studies demonstrated that higher rates of volume loss in MCI patients who later converted to AD, as compared to non-converters, were concerned with the hippocampal structures [32] and that hippocampal volume atrophy was correlated with the annual decline rates in memory test scores of individuals with MCI [33].
Concerning our second result, the effect of homotaurine supplementation on hippocampal atrophy has already been described in patients with an overt diagnosis of AD. In the study ALPHASE, a large scale phase III randomized clinical trial in patients with mild to moderate AD [4], post-hoc analysis of volumetric MRI data showed a significant dose-dependent reduction in hippocampus volume loss in patients treated with homotaurine as compared to a placebo group. Moreover, an exploratory analysis of ALPHASE data indicated a positive effect of homotaurine on memory, language, and praxis skills; furthermore, this pro-cognitive effect was more evident in APOE4 carriers [5], who have a well-known risk for AD.
The above-mentioned result in AD patients included in the ALPHASE trial, similar to that observed in our sample of aMCI individuals, could have been due to the action of homotaurine on the pathophysiological mechanisms of AD, specifically the presence of amyloid deposits and tau aggregates. First, homotaurine may act by inhibiting amyloid aggregation, thus preventing the deposition of neurotoxic amyloid into plaques in brain regions known to be burdened by plaque deposition [34], such as the hippocampus and entorhinal cortex [32]; accordingly, our patients treated with homotaurine showed decreased volume loss in the caudal portion of the hippocampal formation, which has been suggested to be involved in retrieval processes of episodic memory [35]. Second, in animal models homotaurine promotes tau polymerization into fibrillar aggregates that are not toxic for neurons, thus inhibiting the production of tau hyperphosphorilated intracellular tangles [36]. Third, homotaurine is a strong GABA-receptor agonist, and GABA-mediated mechanisms are known to play a role in learning and memory processes [37]. Finally, homotaurine has been shown to exert cytoprotective action through its effect on oxidative damage to DNA caused by free radicals [38].
Intriguingly, in our study individuals supplemented with homotaurine showed less atrophy in brain areas other than the hippocampus. Indeed, whole-brain analyses found clusters of GM where non-supplemented MCI subjects showed significantly increased volume loss compared to supplemented ones. In particular, two clusters were located bilaterally in the fusiform gyrus extending to the tail of the hippocampus, and one cluster was located in the right inferior temporal cortex. These results are coherent with findings in untreated MCI subjects that indicate attenuated hippocampal and fusiform gyrus activation during functional neuroimaging retrieval experiments [39] and reduced functional response in the temporal cortex [40]. Conversely, some pieces of evidence [41] suggest that increased activity during working memory tasks in MCI patients, in areas including the fusiform gyrus, might be related to the involvement of alternative processing techniques and brain regions during performance of working memory tasks.
Thus, supplementation with homotaurine may result in reduced neurodegeneration in hippocampal and extra-hippocampal structures that are involved in specific memory networks and are well described as already impaired in untreated aMCI.
Before conclusions, some limitations of this study should be acknowledged. This is an open-labelobservational study, and patients were not randomized for allocation in the untreated or treated group. Thus, aMCI group taking homotaurine assumed tablets for 1 year, with the expectancy of some therapeutic effect, while the other aMCI group assumed no therapy and, as such, placebo effects cannot be excluded. There is a great amount of evidence claiming that placebo modulates functions within the brain, including analgesic responses and dopaminergic release [42], as well as cognitive performance [43]. While placebo modulation of the memory improvement observed in our study cannot be ruled out, it must be noted that a putative placebo effect should have been generalized and extended to all cognitive functions, rather than to a specific aspect of episodic memory. Further, studies investigating placebo effect on brain functions are mainly based on functional data, such as PET and functional MRI, while, at the best of our knowledge, no evidence exists of a possible effect on brain volume. Therefore, interpreting the reduced hippocampal atrophy observed in our study in terms of placebo, seems highly unlikely. Although double-blind randomized studies are the gold standard of treatment efficacy, open-label trials are less complex and can be conducted at lower costs; therefore, these kind of studies are important to open new avenues of treatment.
The study can also be criticized because of the small sample size of the two groups (22 non treated aMCI, 11 treated aMCI), which limits the generalizability of the findings. Nevertheless, although these results cannot be considered definitive because of these limitations, this is the first study that has analyzed the effect of 1 year of homotaurine supplementation on cognition and brain atrophy in subjects with aMCI, and results are statistically significant (even after appropriate correction for multiple comparisons) and clinically relevant, considering the magnitude of the effect and the concomitant behavioral change. Both the statistical significance, which is beyond the alpha error (i.e., false positive findings) and the clinical relevance, make the concept of protection against neurodegeneration more reliable and solid. Finally, it should be noticed that, as evident from our clinical data, patients included in the study were at risk of the amnesic presentation of AD but not of all the AD phenotypes and this might further limit the generalizability of our results, although it is estimated that non-amnesic atypical AD presentation accounts for only 6–14% of all cases [44].
In conclusion, results of this study suggest that supplementation with homotaurine in subjects with aMCI improves episodic short-term memory and protects against hippocampal volume loss. Unfortunately, there are no treatments specifically approved for MCI. Further, drugs currently used for AD treatment are not recommended for routine treatment of MCI because they do not appear to provide lasting benefit. Thus, double-blind studies with larger samples are justified to confirm definitively these results.
DISCLOSURE STATEMENT
Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/15-0484r1).