Bringing rare neurology drugs to the patients

Maximizing Study Power

EPoC Science is a neurology Efficacy Proof-of-Concept (phase 2) drug developer. Our strategy is to overcome the limitations of rating scale-based endpoints with implementation of digital measures and AI. Our innovative study design will integrate technologies for improving the statistical power.

Multiple system atrophy (MSA), a rare atypical Parkinsonism with no cure today, is our initial focus.

A path to phase 2 trial success

Q: Why neurology drug trials fail?

A: Underpowered endpoints.

Clinical rating scales like MDS-UPDRS* and UMSARS* are frequently used as clinical endpoints for drug development trials. For the purpose of drug development they are coarse and susceptible to noise. Small phase 2, proof-of-concept studies for therapeutic efficacy, are often underpowered**. Increasing the sensitivity and/or reducing variability (noise) should improve the statistical power.

* MDS-UPDRS: Movement Disorder Society Unified Parkinson's Disease Rating Scale, UMSARS: Unified Multiple System Atrophy Rating Scale
** Palma JA, Vernetti PM, Perez MA, Krismer F, Seppi K, Fanciulli A, Singer W, Low P, Biaggioni I, Norcliffe-Kaufmann L, Pellecchia MT, Martí MJ, Kim HJ, Merello M, Stankovic I, Poewe W, Betensky R, Wenning G, Kaufmann H. Limitations of the Unified Multiple System Atrophy Rating Scale as outcome measure for clinical trials and a roadmap for improvement. Clin Auton Res. 2021 Apr;31(2):157-164. doi: 10.1007/s10286-021-00782-w. Epub 2021 Feb 7. [Table 1]

Common results in PD and MSA clinical trials

Differential trajectories of the treatment and placebo arms indicate promising drug candidate.

But, the overlap in scores (i.e., confidence intervals) prevents clear separation of the two arms.

This is a weak and statistically non-significant result, requiring a larger study before regulatory approvals can be sought.

This may be a Type 2 error: Incorrect conclusion that no effect exist when actually one is present. Also called “false negative”.

Method 1: Increase endpoint sensitivity

By increasing the magnitude of change or by adopting more granular scales the arms would separate further, even if the score distributions remain unchanged.

Signal sensitivity can be increased by:

Applying digital sensor-based measurements to complement clinical rating scales
Improving objectivity and reducing inter-rater variability of functional assessments by using AI and digital
Focusing on the symptoms and functions that represent intended therapeutic effects; highlight specific assessment scores

Method 2: Reduce standard deviation

By reducing the standard deviation (i.e., noise), the confidence intervals could separate enough to show statistical significance.

Variability can be reduced by addressing the root causes of:

Interrater variability
Intrarater variability
Intrapatient variability

Higher power reduces study size

Higher powered study will require fewer patients to achieve the same probability of study success (PoSS). Smaller studies mean lower costs and shorter recruitment.

Mori, et al., simulated the impact in Parkinson's Disease studies showing potential ~50% reduction in sample size. (pubmed.ncbi.nlm.nih.gov/35949224/)

Page updated

Report abuse