Predictors of survival in a series of clinically diagnosed progressive supranuclear palsy patients

Background: Investigations into prognostic factors in progressive supranuclear palsy have shown conflicting results. We performed a retrospective study in order to identify clinical predictors of survival in clinically diagnosed progressive supranuclear palsy patients referred to our centre. Methods: Data on medical history, survival and five clinical disability milestones (inability to walk unassisted, unintelligible speech, severe dysphagia, dementia and institutionalization) were collected from outpatients' medical records and by a telephone interview to caregivers. Patients were subdivided into Richardson's syndrome and PSP-Parkinsonism according to symptoms during the first 2 years of disease. Survival was analyzed by the Kaplan-Meier method and Cox regression analysis. Results: Forty-three consecutive patients were enrolled (86% Richardson's syndrome). Motor disturbances were the most frequent symptoms of onset. During the follow-up, 60.5% of patients died after a median survival of 7.1 years (2.2-18). Older age at onset (>63) (HR 2.8; 95% CI: 1.3-5.7; p = 0.007), early dysphagia (HR 2.3; 95% CI: 1-5.3; p = 0.05) and early cognitive deficits (HR 3.6; 95% CI: 1.6-8.2; p = 0.002) were predictors of shorter survival. Compared to PSP-Parkinsonism patients, Richardson's syndrome patients had shorter survival and higher mortality risk although not statistically significant (HR 3 95% CI: 0.9-9.9; p = 0.07). Seventy-seven percent of patients developed severe disability during follow-up: shorter time to the first clinical disability milestone predicted shorter survival (HR 7.8; 95% CI: 2.3-26; p = 0.0008). Conclusions: early dysphagia, cognitive impairment, older age at onset, and time to disability were predictors of shorter survival; Richardson's syndrome had a less favorable course than PSP-Parkinsonism. Clinical milestones should be considered as possible endpoints in future clinical trials.