Measuring time saved in Alzheimer's disease: What is a meaningful slowing of progression?

Abstract

INTRODUCTION

Minimal clinically important differences (MCIDs) for Alzheimer's disease (AD) have previously been estimated using clinician-based anchors. However, MCIDs have been criticized for not reflecting the preferences of people living with AD (PLWAD). Furthermore, interpretations of clinical trial results have been criticized for conflating within-person meaningfulness thresholds and between-group differences. Here, we simulate scenarios of disease slowing and compare those to published MCIDs.

METHODS

Scenarios of 5%–95% disease slowing were simulated using Alzheimer's Disease Neuroimaging Initiative (ADNI) data. Time saved and point differences on the Clinical Dementia Rating scale—Sum of Boxes (CDR-SB) were estimated for these scenarios and compared to published MCIDs.

RESULTS

Scenario analyses resulted in estimates of time saved at ∼3 weeks–17 months and mean changes at 0.08–1.5 CDR-SB points over 18 months. The often referenced MCID for mild cognitive impairment (0.98) thereby corresponded to 11 months slowing, whereas the MCID for mild dementia (1.63) corresponded to >17 months slowing.

DISCUSSION

Translating trial endpoints to estimates of time saved supports that often-referenced MCIDs may not be aligned with realistic and meaningful slowing of clinical progression.

Highlights

• AD slowing of clinical progression by 5%–95% resulted in 0.74–17 months saved and 0.08–1.5 CDR-SB points change at 18 months.
• Slowing of at least 60% or 11 months of time saved over 18 months met an often-cited MCID threshold of 0.98 points for mild cognitive impairment.
• For mild AD dementia, an MCID of 1.63 meant that even an 18-month delay over 18 months would be considered only borderline meaningful—a face invalid and unrealistic proposition.