P1-139: Patterns of cognitive decline during long-term treatment of Alzheimer’s disease in relation to cognitive reserve

Abstract

Some patients treated for Alzheimer's disease (AD) do better than others but the factors that underlie different patterns of disease progression are poorly understood. Variability in cognitive reserve offers one explanation for why patients with similar treatments show variable disease expression. To identify patterns of disease progression in relation to initial conditions. Data come from long term follow–up of patients enrolled in European/Canadian trials of galantamine (Gal–Int 13). Of 240 patients with mild–moderate AD at baseline (age 74.5+/–7.8; 61% women) who were treated with galantamine for 48 months, 188 completed the study. The Alzheimer's Disease Assessment Scale–cognitive subscale (ADAS–cog) and Disability Assessment in Dementia (DAD), were assessed throughout. The Mini–Mental State Examination (MMSE) was available only at baseline. Patients were classified by iterative K–means clustering in three distinct groups according to their pattern of disease progression. The three groups of patients were identified as “least progression” (n=82), “intermediate” (n=82), and “persistent progression”(n=75) based on their ADAS–cog trajectories. The between–group differences in baseline characteristics were much higher than the within–group differences in the maximal rates of change. Of note, these differences do not diminish during the course of disease progression; rather they tend to increase: on average, over 48 months the ADAS–cog increased (representing decline) by 13 points in those with least progression, by 17 points in the intermediate group, and by 21 points for the persistent progression group (p<0.01). The maximal annual changes are 6, 8 and 10, respectively (p for trend <0.05). The baseline MMSE and ADAS–cog were significantly different between the three groups. For example, patients with the least progression had a baseline ADAS–cog of 15+/–8, those in the intermediate group had ADAS–cog 20+/–9 and those with persistent progression had ADAS–cog 30+/–10 (p<0.001). This is also reflected in the MMSE but not the DAD. Of note, only 10.8% of patients with the least progression dropped out, compared with 26.7% of persistent progressors (p<0.001). Differences in patterns of decline appear to arise from differences in the initial (baseline) state. These differences in initial conditions are consistent with differences in cognitive reserve.