Predictors of Progressive Supranuclear Palsy (PSP) disease progression using “A Study to Assess Efficacy, Safety, Tolerability, and Pharmacokinetics of ABBV-8E12 in Subjects With Progressive Supranuclear Palsy (PSP)”

Lead Investigator: Anne-Marie Wills, Massachusetts General Hospital
Title of Proposal Research: Predictors of Progressive Supranuclear Palsy (PSP) disease progression using “A Study to Assess Efficacy, Safety, Tolerability, and Pharmacokinetics of ABBV-8E12 in Subjects With Progressive Supranuclear Palsy (PSP)” (ClinicalTrials.gov ID NCT02985879)
Vivli Data Request: 9077
Funding Source: None
Potential Conflicts of Interest: None

Summary of the Proposed Research:

Progressive Supranuclear Palsy (PSP) is a rare disorder related to Parkinson’s disease and Frontotemporal Dementia affecting approximately 2-3/100,000 persons in the US. It is a neurological disease that causes impaired eye movements, speech, balance, and falls, and rapidly progresses to disability and death. There are currently no treatments for PSP and very few studies using patient-reported outcomes measures in PSP. The PSP-Quality of Life scale (PSP-QoL) was first published in 2006; however, there have been no studies looking at how the scale changes over time. Importantly, the PSP-QoL has not been compared to the validated PSP Rating Scale except in a small single-center cross-sectional cohort study. Validating the PSP-QoL as a patient-reported outcome measure would be extremely important for developing new treatments for this disease. Common metabolic tests, including blood and urine panels, have not been well studied as risk or prognostic factors in progressive supranuclear palsy (PSP).

Previous cross-sectional, case-control studies have suggested a relationship between low urate levels and PSP, with oxidative damage as a hypothetical explanation. We aim to perform longitudinal analyses of the AbbVie trial to investigate the prognostic utility of laboratory data in PSP.

Statistical Analysis Plan:

The Abbvie ABBV-8E12 study was a large, international, multi-center clinical trial of PSP and enrolled 378 patients with this rare disease. We hope to access this data to perform the following secondary analyses:

Aim 1: We will assess the repeated measures association between change in PSP-QoL and change in PSPRS over time. Correlational assessments will be first performed at baseline using Spearman’s correlation. The repeated measures correlation will be calculated using an atypical application of ANCOVA via the rmcorr package in the R programming language. The PSP-QoL will also be compared to the EQ-5D using the same methodologies. We hypothesize that the PSP-QoL and PSPRS will correlate with a rho value of greater than or equal to 0.7 at baseline and 0.5 over time while the PSP-QoL and EQ-5D will correlate with a rho value of equal to or less than 0.4 at baseline as in the original Schrag et al. paper and 0.2 over time. Missing data will be handled using multiple imputation.

Aim 2: To analyze change from baseline over time using the PSPRS, a continuous variable, as our outcome of interest, we use a linear mixed effects modeling approach. In this approach, we first test each covariate one at a time in a series of models including a covariate by time interaction term as well as a main covariate effect and time effect. Covariates as part of interaction effects with a p-value < 0.1 will be included in the multivariable model. We will consider the following covariates: Baseline Demographic variables: age, age*age, gender, disease duration Baseline and Time-Varying (change over first 3 months) Disease severity: PSPRS total score Medication use: medications will be classified by WHO ATC classification. Medication classes used by greater than 10% of participants will be included in the model, including but not limited to: cholinesterase inhibitors (donepezil, rivastigmine), anticholinergic medications (including diphenhydramine, trihexyphenidyl, antispasmodics), SSRIs, SNRIs, SARIs, TCAs, benzodiazepines, stimulants including modafinil, dopaminergic medications (levodopa, dopamine agonists). Medications will be treated as time-varying covariates. Dosage, start and stop dates and indications for each medication class used by >10% of the participants.
Past Medical history (experienced by greater than 10% of participants at baseline) including (but not limited to) hypothyroidism, CAD, arthritis, stroke, anxiety, depression, urinary frequency, constipation, sleep apnea, insomnia.

We will perform linear mixed-effects modeling to assess the potential association between baseline laboratory data and PSP Rating Scale (PSPRS) progression. Further, we will control for relevant confounding variables including medical history and baseline disease severity. To investigate how the relationship between lab measurements and disease progression changes with time, we plan to perform latent growth modeling.

After our selection procedure, we will include k covariates in our multivariable model. We will then use a Bonferroni correction for multiple comparisons to control the familywise error rate. Thus, we will test the statistical significance of each individual fixed effect parameter in the multivariable model at alpha = 0.05/k.

Requested Studies:

An Extension Study of ABBV-8E12 in Progressive Supranuclear Palsy (PSP)
Data Contributor: AbbVie
Study ID: NCT03391765
Sponsor ID: M15-563

A Randomized, Double-Blind, Placebo-Controlled Multiple Dose Study to Assess Efficacy, Safety, Tolerability, and Pharmacokinetics of ABBV-8E12 in Progressive Supranuclear Palsy
Data Contributor: AbbVie
Study ID: NCT02985879
Sponsor ID: M15-562

Summary of Results:

Common metabolic tests, including blood and urine panels, in progressive supranuclear palsy (PSP) progression have not been well characterized. Previous studies have suggested a relationship between low urate levels and PSP. We sought to perform an unbiased analysis of baseline laboratory data to validate previous findings and to identify novel predictors of PSP Rating Scale (PSPRS) progression.

We thus first analyzed baseline laboratory and clinical data from 122 participants in the extension study of ABBV-8E12 (tilavonemab) of the AbbVie trial over 52 weeks. We observed that higher baseline calcium levels were significantly associated with more rapid worsening on the PSPRS after correction (p < 0.0014), while higher phosphate levels were associated with slower worsening (p < 0.0014). These associations remained significant after adjusting for potential confounding comorbidities such as history of fracture, renal insufficiency, Vitamin D deficiency, and baseline disease severity. However, the observed associations did not hold when we gained access to the full study cohort and replicated our analyses. We thus concluded that initial associations were likely due to a smaller sample size. Though we did not complete our SAP, this study still replicated previously published findings, that baseline urate levels were not significantly associated with PSP progression.