Phenotypic Spectrum of Progressive Supranuclear Palsy: Clinical Study and Apolipoprotein E Effect

Article information

J Mov Disord. 2024;17(2):158-170

Publication date (electronic) : 2024 January 30

doi : https://doi.org/10.14802/jmd.23178

Amina Nasri^1,2,3, Ikram Sghaier^1,3, Anis Neji¹, Alya Gharbi^1,2,3, Youssef Abida^1,2,3, Saloua Mrabet^1,2,3, Amina Gargouri^1,2,3, Mouna Ben Djebara^1,2,3, Imen Kacem^1,2,3, Riadh Gouider^1,2,3,

¹Department of Neurology, LR18SP03, Razi University Hospital, Tunis, Tunisia

²Faculty of Medicine of Tunis, University of Tunis El Manar, Tunis, Tunisia

³Clinical Investigation Center (CIC) “Neurosciences and Mental Health”, Razi University Hospital, Tunis, Tunisia

Corresponding author: Riadh Gouider, MD Department of Neurology, Razi University Hospital, 1 rue des orangers Manouba, Tunis 2010, Tunisia / Tel: +216-70-162-243 / Fax: +216-71-601-300 / E-mail: riadh.gouider@gnet.tn

Received 2023 September 9; Revised 2023 December 8; Accepted 2024 January 30.

Abstract

INTRODUCTION

Progressive supranuclear palsy (PSP) is a rare neurodegenerative disorder defined as a neuropathological entity linked to abnormal tau protein accumulation. Classically, PSP is characterized by clinical manifestations, including early postural instability, vertical supranuclear gaze palsy, cognitive decline, pseudobulbar palsy, levodopa-resistant bradykinesia and axial rigidity [1]. Despite the emphasis of earlier reports on motor signs, cognitive decline is henceforth considered a major manifestation of PSP, with a fronto-executive deficit being a core feature [2]. In fact, the classical presentation of PSP, which was initially reported and is now termed PSP-Richardson’s syndrome (PSP-RS), appears to be one of many clinical expressions on a large spectrum. Indeed, in 2017, the Movement Disorder Society (MDS) acknowledged several atypical phenotypes beyond PSP-RS according to the predominant clinical features, including various motor and cognitive deficits [3].

Unfortunately, the factors underlying the clinical expressions and heterogeneity of PSP remain unclear. Several studies have shown that these phenotypic differences are primarily due to the heterogeneity of the distribution of 4-repeat tau pathologies [4]. Genetic factors seem to be implicated in PSP etiopathogenesis considering the reported cases of familial PSP and the role of the microtubule-associated protein tau (MAPT) H1 haplotype in PSP [5]. Furthermore, the role of the apolipoprotein E (APOE) gene as a risk factor for other late-life tauopathies, such as Alzheimer’s disease [6], supports its probable implication as a genetic contributor to PSP phenotypic variability [7].

To date, few cohort studies have attempted to describe the distribution of PSP phenotypes according to the updated MDS classification [7] and to characterize the motor and neuropsychological aspects across these phenotypes. Moreover, to the best of our knowledge, no studies have assessed the impact of APOE on these different phenotypes. Hence, we investigated motor and cognitive features across PSP phenotypes through our Tunisian cohort and evaluated the influence of APOE gene variants on the phenotypic expression of PSP.

MATERIALS & METHODS

RESULTS

Phenotype distribution and demographic findings

We included 112 patients who met the criteria for 42 probable (37.50%), 34 possible (30.35%) or 36 suggestive of PSP (32.14%). We included only patients with suggestive PSP who ultimately (beyond the first 3 years) developed probable or possible PSP according to the 2017 diagnostic criteria. After applying the MAX criteria, all patients were determined to have only one predominant phenotype, which was considered for the study. Forty-eight patients met the criteria for the PSP-RS subgroup (42.85%), 34 patients (30.35%) met the criteria for the PSP-cortical subgroup, and 30 (26.78%) met the criteria for the PSP-subcortical subgroup. The cortical phenotypes included 17 PSP-CBSs (15.2%), 10 PSP-Fs (8.9%) and 7 PSP-SLs (6.2%). The PSP subcortical phenotypes included 13 PSP-Ps (11.6%), 9 PSP-PIs (8%), 3 PSPOMs (2.7%), 2 PSP-PGFs (1.8%), 2 PSP-C (1.8%) and 1 PSP-PLS (0.9%).

A male predominance was noted in our total cohort, with a sex ratio equal to 1.38, without a significant difference between PSP subgroups. Additionally, the PSP subgroups were similar in terms of age, education and disease duration. The mean age at disease onset was 60.82 years, and the age at disease onset was lower for the subcortical subtype and higher for PSP-RS (p= 0.004) (Table 1).

Table 1.

Demographic and motor characteristics of total study cohort and across PSP subgroups

Motor and neuropsychological findings across PSP subgroups

The detailed motor and neuropsychological characteristics of the total study population and across subgroups are detailed in Tables 1 and 2, respectively.

Table 2.

Neuropsychological characteristics of total study cohort and across PSP subgroups

PSP-RS patients had more akinetic-rigid and levodopa-resistant parkinsonism (p = 0.006), while PSP-cortical patients had more tremors and asymmetric and/or levodopa-responsive parkinsonism (p = 0.025).

CI was found in 98 PSP patients (87.5%). Overall, the PSPsubcortical subgroup seemed to be significantly less cognitively affected (Table 2 and Figure 1A).

Figure 1.

Neuropsychological and clinical profiles of PSP according to subtype (A, B) and APOE carrier status (C). Radar charts comparing the percentages of PSP-RS, PSP-cortical, and PSP-subcortical patients with neuropsychological findings (A) and Movement Disorder Society PSP diagnostic criteria clinical features (B) and APOEε4 carrier vs. noncarrier Movement Disorder Society PSP diagnostic criteria clinical features (C). PSP-RS, PSP-Richardson’s syndrome; PSP, progressive supranuclear palsy.

The clinical profiles of the three PSP subgroups (RS, cortical, and subcortical) and comparisons of clinical features between the patients with the typical and atypical PSP phenotypes are depicted in Figure 1B and 2.

Figure 2.

Clinical profile of PSP patients with typical and atypical phenotypes. Radar charts comparing the percentages of PSP-RS (typical phenotype) and PSP atypical phenotype patients according to the clinical features of the Movement Disorder Society PSP diagnostic criteria. PSP-RS, PSP-Richardson’s syndrome; PSP, progressive supranuclear palsy.

The evaluation of correlations between clinical variables in each subgroup revealed various associations (Figure 3). Concerning the subcortical strata, we found a significant inverse correlation between memory impairment and the occurrence of freezing of gait (p = 0.048, r = -0.45) as well as oculomotor dysfunction and memory impairment (p = 0.013, r = -0.46).

Figure 3.

Correlation between clinical variables among PSP subgroups. A: Total cohort. B: PSP-RS. C: PSP-cortical. D: PSP-subcortical. PSP, progressive supranuclear palsy; PSP-RS, progressive supranuclear palsy-Richardson’s syndrome; PSP-cortical, progressive supranuclear palsy-cortical; PSP-subcortical, progressive supranuclear palsy-subcortical.

For the cortical subgroup, a correlation was detected between freezing and the initial FAB score (p = 0.048, r = 0.42), while an association was detected between frequent falls and oculomotor dysfunction (p = 0.047, r = 0.34). Regarding the PSP-RS phenotype, we found a correlation between oculomotor dysfunction and dystonia (p = 0.05, r = 0.29) and between oculomotor dysfunction and dysarthria (p = 0.047, r = 0.29). Moreover, we noted another correlation between freezing and memory impairment (p = 0.026, r = -0.10) (Figure 3).

APOE genotype and its impact on the clinical profile across PSP subgroups

The APOE genotype had different frequencies across PSP subtypes. Approximately 6.31% of PSP patients were carriers of the homozygous form of APOE ε4/ε4, and 40.0% of PSP patients carried one copy of the APOE ε4 allele. Accordingly, significant differences were noted in the genotype distributions across the PSP subgroups (p = 0.025) (Table 3).

Table 3.

APOE allele and genotype distribution across PSP subgroups

Analysis of the associations of APOE ε4 with the whole cohort and the other variables revealed an association between APOE ε4 and the age of parkinsonism onset (p = 0.019). Indeed, all PSP-RS patients in the subcortical and cortical subgroups with APOE ε4 developed parkinsonian syndrome at an earlier age than the noncarriers (63.61 years and 62 years vs. 66.64 years and 66 years).

A stratified investigation of the impact of APOE ε4 on motor profiles across PSP subgroups revealed earlier onset parkinsonism in ε4 mutation carriers (p = 0.011), as did tremor and tremor at onset (p = 0.027 and p=0.022, respectively). Earlier repeated falls were associated with APOE ε4 carriage status (p = 0.038). Freezing was significantly associated with ε4 carriage (p = 0.047). Additionally, a marginal association was noted between oculomotor dysfunction and APOE ε4 (p = 0.052) (Table 1).

Our results adjusted for APOE ε4 carriage among PSP subtypes, according to the cognitive profile, revealed new associations in terms of memory impairment (p = 0.025), with a higher frequency of cases with memory impairment at onset (p = 0.019) and earlier stages (p = 0.008). Additionally, APOE ε4 carriers had a hippocampal profile of memory impairment (p = 0.049). Interestingly, APOE adjustment revealed new associations with early language impairment (p = 0.003). We also reported more frequent mood and behavior disorders (p = 0.038) in the PSPcortical subgroup with the APOE ε4 allele (67.64%) (Table 2).

The analysis of correlations among different motor and nonmotor parameters in each PSP subgroup according to APOE ε4 carriage status revealed an inverse correlation between the ε4 allele and freezing in the PSP-RS subgroup (p = 0.032) (Figure 3A). In addition, the APOE ε4 allele was associated with tremor (p = 0.047) in the PSP-cortical subgroup (Figure 3C) and with a significant correlation between APOE and dysarthria (p = 0.035) and freezing (p = 0.05) in the PSP-subcortical subgroup (Figure 3D).

Analysis of the clinical profiles of PSP patients according to APOE ε4 carriage revealed differences in graph shape (Figure 1C). In fact, compared with that of noncarriers, carriers of the APOE ε4 allele presented the most frequent predominance of cognitive dysfunction (Figure 1C).

DISCUSSION

In the present study, we provided one of the first and largest cohort studies comparing clinical motor and neuropsychological characteristics across PSP phenotypes and subtypes diagnosed according to the MDS criteria with the exploration of the impact of APOE on expression [18,19]. Only a few studies have been conducted in Africa to explore the clinical features of PSP in African patients [7], and even fewer have explored the genetic aspects of the disease [20]. The present study is the first to explore the phenotypic spectrum of PSP among Tunisian and North-African patients and one of the few studies that has examined PSP phenotypes across populations, which is shown in Table 4. Notably, we based our phenotyping process on disease presentation during the first three years of the disease, when diagnosis might be the most challenging. Indeed, PSP subtypes tend to change over time in an individual patient, and most patients eventually fulfill the criteria for PSP-RS [17].

Table 4.

Distribution of PSP phenotypes and main findings for previous studies using the MDS 2017 criteria

Our cohort included a large variety of phenotypes. The typical PSP-RS was the most common PSP phenotype (42.85%), which was in line with the findings of previous studies (24% to 69%), followed by PSP-CBS, PSP-P, PSP-F, PSP-PI, and PSP-SL. The rarest ones were PSP-C and PSP-PLS. Unlike in previous studies, the PSP-CBS phenotypes were more frequent than the PSP-P phenotypes among our cohort. The reasons for this discrepancy might be the lack of pathologically confirmed diagnoses in our series and the different levels of diagnostic certainty across studies. In fact, the highest incidence of the PSP-CBS phenotype was found in studies without neuropathological evidence, and the lowest was found in pathologically confirmed diagnoses [3,19]. Due to the lack of studies comparing cortical and subcortical presentations, we calculated the rates of presentation from available reports. In our cohort, the PSP-cortical and subcortical strata were more closely related (30.35% vs. 26.78%), in contrast to previous reports in which PSP-cortical subgroups tended to be less common (22% vs. 44%). Our findings may be due to the higher PSP-CBS and PSP-SL rates and the lower PSP-P rate compared to those reported in the literature [3,19].

Consistent with the findings of previous reports, no significant differences were noted among the three PSP subgroups according to sex (p = 0.321) [21,22] or age of parkinsonism onset (p = 0.522). Interestingly, the mean age of disease onset was younger in the PSP-subcortical subtype (58.51 years vs. 62.23 years in PSP-RS). Consequently, the age of diagnosis was significantly higher in the PSP-RS subgroup than in the other subgroups (p = 0.017). Our findings contradict those of other studies that outlined the later age of onset in the subcortical PSP. Those studies explained this feature by the more benign clinical trajectory with less CI delaying the perception of disease onset and the age at first consultation [17,23]. Our results may be explained by the fact that parkinsonian motor features that are prominent in PSP-P and PSP-PGF phenotypes (consisting of the subcortical subgroup of PSP) may be more obvious and earlier detected than more subtle cognitive features.

As expected, akinetic-rigid, predominantly axial, and levodopa-resistant parkinsonism was found in two-thirds of the PSP patients and was significantly more frequent in the RS vs. atypical phenotypes, contrary to tremor-associated and/or asymmetric and/or levodopa-responsive parkinsonism (p = 0.025), which was a major characteristic of the PSP-P22 and PSP-CBS phenotypes. Even if tremors were present, almost no PSP patients had a TD form of parkinsonism according to the UPDRS-III assessment, and the PIGD phenotype prevailed across all PSP subgroups (Table 2) [24].

In addition to motor symptoms, patients with PSP commonly exhibit cognitive and behavioral disorders [2,3]. Compared to the former cohorts, which included individuals with CI and dementia, our patients had higher rates of both items (57% vs. 87.5% with CI and 33% vs. 45.5% with dementia) [25] and lower mean MMSE scores (21.63 vs. 26) [14]. The cognitive pattern according to altered domains in all subgroups confirmed the prevailing fronto-executive deficits as a core feature of the PSP cognitive profile. Notably, it was observed that 35.7% of patients in our series were PSP patients with a hippocampal profile of memory disorders. However, memory impairment was not a predominant feature to be considered an exclusion criterion according to the 2017 MDS criteria [3], and all patients with a hippocampal profile developed features of PSP, notably vertical gaze palsy. In addition, recent observations revealed either an Alzheimer disease copathology in PSP patients or a distinct subregional distribution of tau pathology in the hippocampus in PSP patients, suggesting that there may be an AD-independent pathogenic process involved in the hippocampus [26-28].

However, when comparing the subgroups in our cohort, all cognitive domains were significantly less impaired in the subcortical subgroup, particularly for executive functions (p = 0.001). In fact, these findings suggest that subcortical phenotypes are mainly motor and cortical phenotypes that are predominantly cognitive. Although few cohort studies have compared PSP phenotypes in the literature, few have clinically compared PSP-RS, PSP-cortical and PSP-subcortical subgroups, revealing either more CI in the cortical subgroup [29] or no significant differences in cognitive scores [30]. Additionally, a recent study revealed cognitive and behavioral differences in PSP phenotypes, with more semantic verbal fluency in PSP-RS and PSP-CBS patients and more ideomotor apraxia in PSP-CBS patients [19].

Next, we investigated the correlations of different clinical parameters in each subgroup independently (Figure 3). Among the subcortical strata, we found mutual associations between oculomotor dysfunction, memory impairment and freezing of gait at disease onset. This result may be explained by considering the prevailing role of the frontal lobe in the PSP-subcortical subgroup at early stages of the disease in terms of oculomotoricity, gait control and the frontal component of memory functioning. The role of the frontal lobe in the cortical PSP subgroup may emerge later in the disease course, as we also found significant associations between the FAB score and freezing during the disease course [31].

Despite the well-established pathological heterogeneity of PSP, its correlation with PSP phenotypes remains elusive. In fact, the regional variations in the type of tau lesions or differences in tau load may explain the clinical variety of PSP [4,27]. Pathologically, APOE ε4 may lead to increased Aβ aggregation and decreased clearance of Aβ from the brain [32]. However, the influence of APOE isoforms on tau pathology is unclear. Although tau is a cytoplasmic protein, it has been reported that tau protein is also found in the extracellular space [33]. In fact, intracellular tau aggregates are in equilibrium with extracellular tau, whose aggregation might mediate the spread of pathological species of tau between cells. Since APOE may be present in the cytoplasm, it could form a complex with tau and modulate its metabolism in the brain [34]. Thus, APOE can also modulate tau aggregation and related pathologies [34]. Furthermore, it is hypothesized that the potentially different complex-forming properties of tau and APOE isoforms (ε3, ε2, and ε4) might affect tau metabolism to variable degrees. Overall, these findings may lead us to speculate that one of the possible mechanisms by which APOE could have an impact on PSP phenotypes is through the modulation of Tau metabolism, aggregation and distribution. Additionally, genome-wide association studies have shown a strong correlation between APOE and cerebrospinal fluid tau and phosphorylated tau (p-tau) [35].

The impact of APOE on the phenotypic diversity of PSP has not yet been adequately investigated, apart from one study that confirmed similar frequencies of APOE genotypes in PSP-RS and PSP-P [22]. Accordingly, we aimed to evaluate the effect of APOE ε4 on PSP phenotypes. Overall, a significant association was detected between the APOE genotype distribution and PSP subgroup (p = 0.025). Furthermore, significant differences in demographic, motor and nonmotor variables were detected between APOE ε4 carriers and noncarriers. In fact, earlier age of parkinsonism onset was significantly associated with APOE ε4 carriage (p = 0.019). Indeed, it has been reported that patients with PD and carriers of APOE ε4 have a significantly earlier onset of parkinsonism [36,37]. Early tremor at onset was less common among patients who were carriers of the APOE risk allele (p = 0.031).

Concerning the cognitive profile, the study of APOE ε4 compared to non-APOE ε4 revealed significant differences, with APOE ε4 carriers presenting more marked cognitive dysfunction. Indeed, carriers of the risk allele developed earlier memory impairment (p = 0.049). We noted a significant impact of APOE ε4 on attention, executive functions, and apraxia but not agnosia among the PSP subgroups. In fact, the associations found previously continued to be significant regardless of APOE ε4 carrier status adjustment.

Nonetheless, we reported the emergence of a new significant association between memory deficits and disease onset by adjusting for APOE ε4 status (p = 0.019). These findings agree with previous results in which APOE ε4 was associated with a faster rate of memory decline [38]. At the cellular level, this could be due to a possible APOE ε4 association with enhanced tau deposition in medial-temporal regions [31]. Additionally, an association between APOE status and early language impairment was reported (p = 0.003). In fact, noncarriers of the ε4 allele developed language impairment earlier than allele carriers did. These findings might be explained by previous observations that revealed that patients with the APOE risk allele exhibited more brain atrophy in the medial temporal lobe than noncarriers, while the nonε4 patients exhibited more brain atrophy in the frontal and parietal lobes than the other patients [39,40].

The initial NPI was associated with APOE ε4 carriage (p = 0.038). This finding was in line with the findings of other studies suggesting a link between the mutant APOE allele ε4 and the risk of depressive symptoms in patients with dementia, probably due to reduced amygdala volume [41].

A correlation analysis of the total PSP cohort revealed a positive association between APOE and attention disorders (p = 0.032, r = 0.24). A major possibility for the APOE-attention link is the alteration of cholinergic transmission associated with the cortical areas that are important for attentional operations [42]. However, the correlation between APOE and PSP severity changed with stratification.

The findings of our study have to be considered in light of several limitations. Due to its nature, our retrospective study relies on the use of charts; occasionally, the charts were not designed to collect adequate information for research. This fact could inevitably lead to missing data but without substantially impacting the PSP subtyping process. Additionally, despite the large cohort of PSP patients included in the present study, the different subgroups were restricted. Finally, the limited APOE genotype data and the lack of neuropathological confirmation could be major limitations.

This study demonstrated the wide phenotypic spectrum of PSP among Tunisians. Disease onset and akinetic-rigid and levodopa-resistant parkinsonism were the hallmarks of the PSPRS phenotype, while milder CI was characteristic of the PSPsubcortical subgroup. The APOEε4 allele seemed to play a role in defining a more altered cognitive profile in PSP patients. This effect mainly concerned executive function, memory, attention and apraxia. Our study may enrich the scientific literature on PSP North African patients. Further studies could also help to elucidate the interplay between APOE genotyping and other potential ethnic and genetic singularities of PSP in these regions.

Article information Continued