Fabrizia D'Antonio (1,2), Julie Bertrand (3,4), Yucheng Sheng (5), Robert Howard (1), Suzanne Reeves (1,2)
(1)Division of Psychiatry - University College London - London (2) Department of Old Age Psychiatry - King's College London - London (3) 4 UMR 1137, Inserm - University Paris Diderot - Paris (4) University College London Genetic Institute (5) Department of Pharmaceutics - UCL School of Pharmacy - London
Objectives: The psychosis phenotype in Alzheimer’s disease (AD) may be comprised of two subtypes: paranoid (persecutory delusions), and misidentification (misidentification phenomena ± visual and auditory hallucinations).Psychosis is associated with a faster speed of cognitive decline, but whether this effect is subtype dependent has not been explored. This study aim to model how psychotic subtypes impact cognitive trajectories, evaluated by AD Assessment Scale (Adas-cog), in the longitudinal dataset ADNI2.
Methods: ADNI2 participants categorised as late mild cognitive impairment or AD over a 4 years observation period were included in the analysis. Trajectories were fitted with simplified versions of the Richard’s function (Samtani et al. 2012, Conrado et al. 2014). These models may allow to capture an inflection point beyond which the rate of decline reaches a plateau. Adas-cog being bounded between 0 and 70, we explored both a constant error model on the log-transformed observations and an extended logit error model. The psychosis effect on the progression rate parameter was explored as follows: “psychotic symptoms y/n”, “delusion y/n”, “hallucination y/n” and finally psychotic subtypes. Potential confounding covariates were also explored. Parameter estimation was performed using the SAEM algorithm implemented in the Monolix software (v4.3.3) and model selection was based on likelihood ratio test and visual predictive checks.
Results: Among the 528 patients in this analysis, there were 96 psychotic subjects, of whom 38 were pure paranoid subtype, 29 pure misidentifications subtype and 28 mixed subtype. Samtani’s function with a constant error model fitted the data better. A correlation between Adas-cog at baseline and rate of progression was required to ensure model stability. The inflection point for the Adas-cog score trajectory was at 46.8 units, consistent with previous literature. The presence of psychotic symptoms significantly increased the rate of disease progression from 1.3 to 3.0% in ADAS-cog score per year (P<0.001). Misidentification subtype alone increased the rate from 1.3 to 2.8% (P =0.021), paranoid subtype was not significantly associated, but the presence of both psychotic subtypes (mixed) increased the rate from 1.3 to 3.9% (P =0.0016).
Conclusion: We quantified how presence of psychosis yield faster cognitive decline. Moreover, we showed how this effect is essentially associated with mixed and misidentification subtype.
References:
[1] Cook SE, Miyahara S, Bacanu SA, Perez-Madriñan G, Lopez OL, Kaufer DI, Nimgaonkar VL, Wisniewski SR, DeKosky ST, Sweet RA.. Psychotic symptoms in Alzheimer disease: evidence for subtypes. Am J Geriatr Psychiatry. 2003 Jul-Aug;11(4):406-13
[2] Reeves SJ, Gould RL, Powell JF, Howard RJ. Origins of delusions in Alzheimer’s disease. Neurosci Biobehav Rev. 2012 Nov;36(10):2274-87
[3]Samtani MN, Farnum M, Lobanov V, Yang E, Raghavan N, Dibernardo A, Narayan V; Alzheimer’s Disease Neuroimaging Initiative. An improved model for disease progression in patients from the Alzheimer’s disease neuroimaging initiative. J Clin Pharmacol. 2012 May;52(5):629-44
[4]Conrado DJ, Denney WS, Chen D, Ito K. An updated Alzheimer’s disease progression model: incorporating non-linearity, beta regression, and a third-level random effect in NONMEM. J Pharmacokinet Pharmacodyn. 2014 Dec;41(6):581-98.
Reference: PAGE 25 (2016) Abstr 5947 [www.page-meeting.org/?abstract=5947]
Poster: Drug/Disease modeling - CNS