The behavioral variants, which have a similar rate of atrophy to the nfvPPA variants (as illustrated on the previous slide), have more variability within the population assessed, leading to higher sample size estimates. Within 18 months from baseline, less than 50 subjects per arm are estimated to be required to observe a 40% slowing of the observed brain atrophy in the Thalamus.

In the nfvPPA population, by 18 months, less than 50 subjects per arm are required to observe a 40-80% slowing of atrophy in the Thalamus, Frontal Cortex, and Temporal Cortex regions, as well as expansion of the Lateral Ventricles.

In the svPPA population, as early as 6 months, fewer than 50 subjects per arm are required to observe a 40% slowing of atrophy in the Hippocampus and Temporal Cortex.

Comparing Cohen's d allows for the assessment of effect size across different platforms. For comparison purposes, we also evaluated the same dataset using FreeSurfer.

When looking at the effect size for the twelve-month data across all three FTD variants, a clear distinction emerges between the two analytical methods. In the PIANO™ analysis (on the left), the majority of brain regions in both svFTD and nfvPPA variants display Cohen's d values greater than 0.8 (above the blue line), which is considered indicative of a strong effect.

This data suggests that the PIANO™ approach is highly sensitive for detecting meaningful changes in atrophy over a relatively short period, allowing for smaller sample size requirements in clinical trials.

Conversely, the FreeSurfer analysis (shown on the right) yields substantially lower Cohen's d values for the same regions and time point. Lower effect sizes translates into a need for larger sample sizes to achieve statistically significant results.

This comparison underscores the advantage of the PIANO™ analysis pipeline in identifying robust therapeutic effects with fewer participants, streamlining the design of FTD clinical studies and potentially accelerating the evaluation of new treatments.