LAUSR

We read with interest the article from Grötsch et al on the modified version of the Progressive Supranuclear Palsy Rating Scale (mPSPRS). Here, we applied the mPSPRS to our singlecenter PSP cohort and investigated its sensitivity to change in patients with Richardson syndrome (PSP-RS) and the other variant syndromes of PSP (vPSP). The mPSPRS was computed based on its original version for 36 patients (29 PSP-RS and 7 vPSP) (Supplementary data) assessed at least twice (mean standard deviation follow-up: 15.33 9.78 months). PSP diagnosis and phenotype attribution were determined according to the Movement Disorder Society criteria. Table 1 shows the sample size required to detect 20% and 50% change from baseline in PSP-RS and vPSP for the mPSPRS and its original version. Supplementary data show sensitivity to change for the whole PSP cohort for single items and subscores of the scales. In line with Grötsch et al, our data confirm that for the whole PSP cohort the mPSPRS has a slight lower sensitivity compared to its original version to detect a 50% change over follow-up. Sample sizes for power calculations in our cohort are, in general, smaller than those reported by Grötsch et al, possibly due to the longer follow-up (15 vs. 12 months). When considering PSP phenotypes, the mPSPRS presented higher sensitivity for PSP-RS than vPSP. In keeping with the slower disease progression in vPSP, the effect sizes are smaller; thus, larger samples would be needed to detect significant changes over time. Indeed, the effect size for mentation, history, and bulbar subscores of the mPSPRS for vPSP is so small that 100 patients would be required to demonstrate a 50% change (Table 1). This is likely due to the small sample size of our cohort of vPSP. Complementarily, such result could be explained by the elimination of sensitive items for vPSP (ie, irritability for PSP with frontal presentation). On the contrary, mPSPRS possesses good clinimetric properties for both PSP-RS and vPSP in regard to ocular and gait subscores (Table 1). The mentation subscore from the mPSPRS had the smallest effect size in all settings (whole cohort, PSP-RS, and vPSP). Grötsch et al left only the emotional incontinence item and eliminated grasping in such subscore. However, in our itemper-item analysis, grasping was significantly more sensitive to change, with no apparent relation to emotional incontinence (P = 0.797) (Supplementary data). Considering also that grasping and imitative behavior can precede emotional incontinence in the disease course, it could be beneficial to consider both items within the mentation subscore. In conclusion, our data from an independent cohort of PSP patients further support the use of the proposed mPSPRS in clinical practice when all the disease phenotypes are considered. However, we also highlight that more work needs to be performed to improve sensitivity to change of rating scales and sample size calculations for vPSP. Given the heterogeneous forms of diseases included in the vPSP category, a revision rather than a simple compression of the original PSPRS would be advisable.