Association of Progressive Supranuclear Palsy Rating Scale with Progressive Supranuclear Palsy Quality of Life Scale

Progressive supranuclear palsy (PSP) is a degenerative atypical parkinsonian disorder that shortens lifespan and severely impacts quality of life (QoL) [1, 2]. Like other atypical parkinsonian disorders, PSP causes a set of challenges distinct from idiopathic Parkinson’s disease (PD) that greatly impact health-related quality of life (Hr-QoL), including more rapid progression and poor/no response to dopaminergic medications or other treatment modalities [3-5]. At a time when patient-reported outcomes are increasingly being recognized as important – and even primary – end points in clinical trials for parkinsonian disorders, it is crucial to have a validated disease-specific tool that assesses aspects of PSP that most impact patients’ QoL [4].

While several PD-specific tools are available for the measurement of Hr-QoL [6], few disease-specific QoL scales exist for atypical parkinsonian disorders. Despite the growing recognition of PSP as a uniquely disabling condition with a likely underestimated incidence and prevalence (due to misdiagnosis), the PSP quality of life scale (PSP-QoL) remains the only published and validated tool for Hr-QoL assessment of this disease [3, 7, 8]. Developed by Schrag and colleagues in 2006 following detailed interviews with clinically diagnosed PSP patients, the PSP-QoL questionnaire was validated in a population of nearly 200 Richardson syndrome PSP (RS PSP) patients in the UK and found to have good psychometric properties [7]. However, in its most recent 2015 guidelines, the Movement Disorder Society (MDS) Guideline Task Force was only able to encourage, but not fully recommend, the use of this instrument due to its limited use in the clinical research setting [8]. As the MDS task force implied, the application of the scale within other PSP patient groups is necessary to establish the instrument as a clinically recommended tool.

In addition to applying the PSP-QoL to different populations, it is necessary to examine other aspects of the scale that were not investigated during its initial development and validation. First, since the publication of the PSP-QoL, the clinical characterization of PSP has undergone expansion, with a growing recognition of PSP phenotypes other than the classic RS [9-16]. Respondek and colleagues [17] found that of 100 pathologically confirmed PSP cases, the majority presented with atypical constellations of symptoms within the first 2 years of disease and that these different phenotypes had different rates of symptom progression [14]. The PSP diagnostic criteria underwent a major revision in 2017 [18] to include additional PSP variants described since the prior criteria were published in 1996 [19]. Therefore, it is important to evaluate whether the PSP-QoL is applicable to a clinical population that reflects the heterogeneity of PSP presentations beyond RS. Second, the decade since the PSP-QoLs publication has seen the emergence of the PSP Rating Scale as a broadly validated and highly recommended tool for physician-assessed disease severity and progression [8, 17, 18, 20]. Given the PSP Rating Scale’s widespread application and use as an end point in clinical trials, it would be valuable to explore its interscale correlations with PSP-QoL total and subscale scores. This correlation analysis could ensure not only that the 2 scales yield congruent information regarding disease status but also provide insight into the physician-scored aspects of disease that best associate with patient-reported outcomes and overall life satisfaction.

The goal of our study was to compare the PSP rating scale and the PSP-QoL in a heterogeneous PSP patient group to determine what aspects of the PSP Rating Scale correlate with PSP-QoL total score and subscale items. We also compared patients with a diagnosis of PSP-RS with non-RS PSP (the latter group included patients with nearly all currently recognized PSP phenotypes based on the MDS PSP diagnostic criteria [18]), in order to determine whether there are differences in QoL determinants between RS PSP and non-RS PSP variants.

This study includes 60 PSP patients who were assessed in the Johns Hopkins Atypical Parkinsonism Multidisciplinary Clinic between 2015 and 2018. This study received Johns Hopkins Institutional Review Board approval, and written informed consent was obtained for all patients. Inclusion in these analyses required 2 main elements. First, the clinical diagnosis of PSP needed to meet the 2017 Hoglinger et al. [18] MDS PSP diagnostic criteria (suggestive of, possible, or probable PSP). Note that these criteria were applied retrospectively to all PSP cases, who were seen prior to the publication of the criteria in 2017. Overall, 56 of the 60 patients met either clinically possible or probable PSP criteria, while the remaining 4 had the “suggestive of” PSP-Postural Instability variant [18]. Second, patients included in the analyses were required to have a PSP Rating Scale and PSP-QoL completed within 1 calendar month of each other. PSP-QoL questionnaires were administered during the same clinic visit or mailed to the patients shortly afterward to fill out and return.

Based on a consensus review of the cases by 2 trained neurologists (A.P, or J.B.) patients were placed into a phenotypic category using the 2017 Hoglinger et al. [18] diagnostic criteria for PSP, which separate patients based on predominant symptomatology. We refer to these phenotypic classifications as PSP variants. Two PSP variant subgroups were included in the analyses: (1) patients with classic RS PSP featuring recurrent falls and supranuclear gaze palsy early in disease and (2) those without this early clinical presentation (non-RS PSP). The non-RS phenotypes included individuals who presented with a predominance of parkinsonism resembling PD (PSP-P), postural instability (PSP-PI), corticobasal syndrome (PSP-CBS), speech or language disorders (PSP-SL), frontal lobe cognitive or behavioral presentations (PSP-F), and progressive gait freezing (PSP-PGF). These non-RS variants could not be analyzed separately because of small patient numbers. Data on variants and diagnostic certainty are presented in online suppl. Figure 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000514519).

PSP-QoL is a 45-item patient-rated questionnaire that assesses the difficulty of mental and physical tasks within the previous 4 weeks and provides an overall score for life satisfaction. The first 22 items on the PSP-QoL scale provide a measure of physical function (Physical subscale), and the last 23 items assess mental function (mental subscale) (see online suppl. Table 1). The raw scores for each item (0–4) for the 2 subscales are summed and transformed to a range of 0–100 {100 × [(observed score minus minimum possible score)/(maximum possible score minus minimum possible score)]}, with greater scores indicating greater impairment [7]. The PSP-QoL also contains an associated visual analog scale requesting the patient provide a life satisfaction rating within a range of 0–100, with higher scores indicating greater life satisfaction. Mean PSP-QoL total and subscale scores and life satisfaction rating scores were determined for all patients, as well as for RS and non-RS subgroups separately.

The PSP Rating Scale also contains a Physical subscale and a Mentation subscale [20]. Mean PSP Rating Scale total and subscale (i.e., history, mentation, physical, bulbar, ocular, limb, and gait) scores were calculated for the complete cohort, as well as separately for the RS and non-RS subgroups. Scores range from 0 to 100, with greater scores indicating greater impairment. Of note, the physical subscale score of the PSP Rating Scale represents a composite of the bulbar, ocular, limb, and gait scores. Each PSP Rating Scale was scored by 1 of 2 trained neurologists (A.P. or J.B.).

Two sample t tests were used for bivariate comparisons of continuous variables and Pearson’s χ² test or Fisher’s exact test was used for categorical variables. The Wilcoxon rank-sum test was used to examine disease duration, due to the skewness of the measure. Pearson’s correlations between the QoL scores and PSP ratings with Sidak-adjusted [21] significance levels (to adjust for multiple comparisons) were calculated to examine the correlation between the 2 sets of measures. Linear regression models using PSP ratings as the dependent variables and the QoL scores as predictors, adjusting for age, gender and disease duration, were used to examine the relationship between the 2 scales. Interaction terms between QoL scores and PSP phenotype (RS PSP vs. non-RS PSP) were examined to determine whether the relationship changes depending on PSP phenotype. Considering the sample size, the confounders were not included in the models with the interaction terms. Scatter plots were generated to visualize interscale relationships for total and subscale scores. Correlation coefficients with p values <0.05 were considered statistically significant. All statistical analyses were performed using STATA version 14.0 (StataCorp, 2015).

Sixty patients met the study inclusion criteria for clinical PSP diagnosis and had a PSP-QoL) and a PSP Rating Scale completed within 1 month of each other. The demographics of these participants are presented in Table 1. Of the 60 patients in the cohort, 43 (71%) fulfilled possible or probable PSP-RS criteria, while the remaining 17 patients had non-RS PSP variants. To date, definite PSP diagnosis has been confirmed on autopsy in 3/3 patients in this cohort. Table 1 also summarizes PSP-QoL and PSP Rating Scale scores, including average total, mentation, and physical subscale values for each instrument.

Overall, there were no significant differences in demographic features between the PSP-RS and the non-RS PSP patients (Table 1). There was also no significant difference between the mean satisfaction rating scores of PSP-RS and non-RS patients (p = 0.32). To examine the relationship between disease duration and participant characteristics, we divided the subjects into those with disease duration of ≤4 years and those with disease duration >4 years. Among those with disease duration ≤4 years, patients with PSP-RS had a significantly worse ocular subscale scores on the PSP Rating Scale than patients with non-RS variants (p = 0.02) Among those with disease duration >4 years, patients with non-RS variants had a longer average disease duration than patients with RS (p = 0.04) and also tended to have worse limb subscale scores (p = 0.05) (see online suppl. Table 2). Online suppl. Figure 1 provides additional information about PSP variants in our cohort. Online suppl. Figure 2 provides details of the breakdown of the subjects based on meeting criteria for probable versus possible PSP.

Among the entire cohort, there were significant interscale correlations between the total, physical, and mental/mentation scores of the 2 instruments (r = 0.65, p = 0.001; r = 0.62, p = 0.001; r = 0.48, p = 0.006), as well as significant correlations between many of the other subscales. For the physical examination items from the PSP Rating Scale (i.e., bulbar, ocular, limb, and gait), only gait scores had a significant correlation with QoL (total scores [r = 0.58, p < 0.001], physical scores [r = 0.67, p < 0.001]), but not mentation (r = 0.40; p = 0.09), and there were no significant correlations between PSP Rating Scale Total or subscale scores and patient-reported Life Satisfaction ratings. Table 2 shows these results.

As the next step in our analysis, we considered the associations between the PSP Rating Scale and the PSP-QoL in a robust linear regression model, adjusted for age, gender, disease duration, and PSP phenotype (RS vs. non-RS variants). In this analysis, PSP Rating Scale total scores tended to be negatively associated with life satisfaction (adjusted p = 0.08) (Table 3). Gait scores of the PSP Rating Scale demonstrated significant negative associations with life satisfaction ratings (adjusted p = 0.028), while other subscale scores did not correlate with life satisfaction significantly (Table 3).

Table 4 shows the results of the comparisons between the 2 scales based on PSP subtype. PSP-QoL Total scores correlated with PSP Rating Scale Gait subscale scores more strongly for the non-RS PSP variants than for PSP-RS patients (p = 0.026). The same held true for Gait scores and PSP-QoL Physical subscale scores (p = 0.008). Also, for non-RS PSP patients, life satisfaction ratings correlated more strongly with PSP Rating Scale Total scores (p = 0.02), mentation scores (p = 0.02), and gait scores (p = 0.002) than for PSP-RS patients. All other interscale correlations did not appear to be differentially associated with PSP diagnostic status (PSP-RS vs. non-RS PSP).

This study found strong positive interscale correlations between the total and subscale scores of the patient-rated PSP-QoL 45-item questionnaire and those of the physician-assessed PSP Rating Scale. These correlations were largely maintained among patients with and without classic RS, suggesting that the PSP-QoL could be a useful patient-related outcome in the assessment of Hr-QoL in clinical research, including clinical trials.

We also explored the potential impact of PSP phenotype on patients’ QoL years into the course of illness. This study found that most of the associations between the PSP Rating Scale and the PSP-QoL were comparable for the 2 groups of patients (PSP-RS and non-RS PSP). On the other hand, PSP-QoL total and physical scores correlated with PSP Rating Scale gait subscale scores more strongly for the non-RS PSP variants than for PSP-RS. Additionally, life satisfaction ratings correlated more strongly with PSP Rating Scale total, mentation, and gait scores for non-PSP variants than the PSP-RS patients. It should be noted that the correlations presented in Table 2 included the total study sample and did not adjust for any confounders, while the results in Table 3 were adjusted for age, gender, duration, and PSP phenotype (RS vs. non-RS). The difference between these results indicates that for the overall sample, there is no significant linear correlation between the QoL physical score and the PSP Rating Scale bulbar, ocular, or limb scores on average. However, for patients with comparable age, gender, duration and PSP phenotype, the QoL physical score is positively associated with the PSP Rating Scale bulbar, ocular, and limb scores.

There were no baseline demographic or PSP Rating Scale/PSP-QoL score differences among PSP-RS and non-RS variants; age and disease duration were also similar in both subgroups (and were controlled for in the correlation analyses). Thus, the reason for the apparent divergence in determinants of life satisfaction between PSP-RS and non-RS patients cannot be explained based on these characteristics. We suggest that it may be due to important differences in the rates of progression of physical and cognitive disability in these subtypes [15, 17, 22-25] and plan to address this question through ongoing longitudinal patient follow-up. Focusing on specific PSP-related impairments and correlating with longitudinal change in QoL, such as apathy, will also be important. Pekmezović et al. [26] used the Short Form Health Survey (SF-36, Serbian version) assessment to longitudinally evaluate QoL in a sample of 28 patients with PSP at baseline and 1-year follow up. In their final models, only apathy (as assessed by the Apathy Evaluation Scale) had a statistically significant negative predictive value for worsening on the Physical and Mental Composite Scores of the SF-36.

It is noteworthy that the analysis of the overall study sample did not yield significant associations between patients’ Life Satisfaction ratings and the total or subscale scores of the PSP Rating Scale. The Life Satisfaction rating is a global patient-reported outcome; the PSP Rating Scale was validated in PSP- RS and includes a History section (7 items), for which responses are obtained from patients and their care partners and 6 Examination sections (21 items) based on neurological assessment by a clinician. Our results suggest that for a mixed sample of RS and non-RS PSP, total and subscale scores on the PSP Rating Scale may not be consistent predictors of overall patient-reported life satisfaction. Moreover, care partner input for the History section of the PSP Rating scale (obtained for every assessment) confounds its comparison with the patient-reported Life Satisfaction rating; as has previously been demonstrated in community-residing persons with dementia of different etiologies, QoL correlates depend heavily on whether the rating is done by the patient or care partner [27].

Non-RS PSP variants were common among our patient population. While at first glance our findings appear to diverge from the original descriptions of PSP variants of Respondek et al. [17] (in which only 24% of pathology-confirmed PSP cases presented with all the cardinal features of disease within the first 2 years), most of our cohort consisted of patients with >2 years of disease duration, which likely skewed the diagnostic classification toward the classic PSP-RS phenotype. As others have remarked, such findings underscore the difficulty of finding a sensitive and specific clinical examination marker for early PSP disease [16, 28] and the importance of inclusive diagnostic guidelines. At the same time, as in Respondek’s cohort, our study was comprised of tertiary referral center cases. Therefore, referral bias may have skewed inclusion toward atypical (non-RS variant) patients.

Outside of possible referral bias, other limitations of our study include the relatively small sample size among the non-RS group, which prevented a more detailed subgroup analysis. A larger study population across multiple centers would have allowed for further examination of this heterogeneous population by individual subtype. Additionally, our study lacked data on how much help patients received with completing the PSP-QoL. Therefore, it is unclear what proportion of our results are influenced by caregiver perspective. This important aspect of PSP care can be evaluated with separate assessment tools such as the Zarit Caregiver Burden interview, which is now being utilized at our center [29]. Another future plan involves longitudinally evaluating discrepancies between answers given by patients and care partners on the History portion of the PSP Rating Scale; whether these discrepancies are a harbinger of cognitive or overall decline in PSP remains to be determined. Our study was also limited to cross-sectional observations, and longitudinal prospective analysis of the PSP-QoL is required to determine its overall long-term applicability and relationship to changes in the PSP Rating Scale. Longitudinal studies may provide an opportunity for examining these scales in relation to autopsy confirmation of subjects’ PSP diagnosis.

In conclusion, our study supports the utility of the PSP-QoL in clinical research settings and suggests that it may also be useful as a patient-reported outcome in clinical trials. Not only does it demonstrate good congruence with the widely used PSP Rating Scale, but it also appears capable of yielding valuable adjunct information on QoL determinants among different PSP subtypes. Further application of the scale to larger multicenter cohorts is needed to validate these findings. In addition, longitudinal assessment of these patients, which is ongoing at our center, will enable important analysis of the rate of change in both the PSP Rating Scale and PSP-QoL, which may vary for different PSP phenotypes. Future avenues of study could also include analysis of PSP-QoL correlations with other disease assessments, such as MR imaging, Positron Emission Tomography tau tracer studies, and cerebrospinal fluid, and blood biomarker levels.

We would like to thank the patients and their family members who took their time to fill out questionnaires for this study. We would also like to thank Carrie Berlett and Kecia Garrett, who coordinated the clinic visit schedule and sent out questionnaires, and Lavinia Rizvi, who helped with data entry.

This research complies with the guidelines for human studies and should include evidence that the research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. All participants gave their written informed consent, and the study protocol was approved by our institute’s committee on human research (Johns Hopkins University School of Medicine Institutional Review Board-2, Approval Reference Number IRB00062534).

The authors have no conflicts of interest to declare.

This work (protected time to carry out the research and prepare the manuscript) was supported by the National Institutes of Health (NIA K23AG059891, Pantelyat).

Alexander Pantelyat: project conception, data collection, data analysis, and manuscript preparation. Lenora Higginbotham: data analysis and manuscript preparation. Liana Rosenthal: data analysis and manuscript preparation. Diane Lanham: data collection and manuscript preparation. Vanessa Nesspor: data collection and manuscript preparation. Mina AlSalihi: data analysis and manuscript preparation. Jee Bang: data collection and manuscript preparation. Jiangxia Wang: data analysis and manuscript preparation. Marilyn Albert: project conception, data analysis, and manuscript preparation.

Alexander Pantelyat and Lenora Higginbotham contributed equally.