Presenilin‐1 mutation position influences amyloidosis, small vessel disease, and dementia with disease stage

Abstract INTRODUCTION Amyloidosis, including cerebral amyloid angiopathy, and markers of small vessel disease (SVD) vary across dominantly inherited Alzheimer's disease (DIAD) presenilin‐1 (PSEN1) mutation carriers. We investigated how mutation position relative to codon 200 (pre‐/postcodon 200) influences these pathologic features and dementia at different stages. METHODS Individuals from families with known PSEN1 mutations (n = 393) underwent neuroimaging and clinical assessments. We cross‐sectionally evaluated regional Pittsburgh compound B‐positron emission tomography uptake, magnetic resonance imaging markers of SVD (diffusion tensor imaging‐based white matter injury, white matter hyperintensity volumes, and microhemorrhages), and cognition. RESULTS Postcodon 200 carriers had lower amyloid burden in all regions but worse markers of SVD and worse Clinical Dementia Rating® scores compared to precodon 200 carriers as a function of estimated years to symptom onset. Markers of SVD partially mediated the mutation position effects on clinical measures. DISCUSSION We demonstrated the genotypic variability behind spatiotemporal amyloidosis, SVD, and clinical presentation in DIAD, which may inform patient prognosis and clinical trials. Highlights Mutation position influences Aβ burden, SVD, and dementia. PSEN1 pre‐200 group had stronger associations between Aβ burden and disease stage. PSEN1 post‐200 group had stronger associations between SVD markers and disease stage. PSEN1 post‐200 group had worse dementia score than pre‐200 in late disease stage. Diffusion tensor imaging‐based SVD markers mediated mutation position effects on dementia in the late stage.


Statistical analysis models and equations
2.1.Linear Mixed Effect Models, Sample Size, and EYO range grouping All linear mixed effects models were ran using SAS software.We run the analysis with EYO as continuous and as categorical measures per range of 5years or 10 years.

Regional PiB-uptake and EYO
To evaluate the effects of mutation position on regional amyloid burden as a function of estimated years to symptom onset (EYO), we used the model (1) and model (2), accounting or not for age of onset, respectively.Both models included APOE-ɛ4 status (presence or not of at least one APOE-ɛ4 allele), sex at birth, and education in years as fixed effects and family cluster as random effect.The linear mixed effect model below was applied and run individually for each FreeSurfer region of interest (ROI, n=40) and results of the model were corrected for multiple comparisons using the Benjamin-Hochberg method (Benjamini and Hochberg 1995).

Regional WMH volumes and EYO
To evaluate the effects of mutation position on total and regional white matter hyperintensity volumes as a function of EYO, we used the model (3) and model (4), accounting or not for age, respectively.Both models included APOE-ɛ4 status, sex at birth, mean arterial pressure (MAP), and education in years as fixed effects and family cluster as random effect.The linear mixed effect model below was applied and run individually for each WMH volume (Vol, n=4).Due to the small number of regions and the exploratory nature of the analysis, the results were not adjusted for multiple comparisons.

Peak width of Skeletonized Mean Diffusivity and EYO
To evaluate the effects of mutation position on global white matter injury (as measured by peak width of skeletonized mean diffusivity (PSMD)) as a function of EYO, we used the model ( 5) and model ( 6), accounting or not for age, respectively.Both models included APOE-ɛ4 status, sex at birth, MAP, and education in years as fixed effects and family cluster as random effect. ( These 2 models were used with EYO as a categorical variable to estimate   ±   for each group (NC, Pre-200, and Post-200 MC) and compare the groups at different disease stage as ranged per 10years (3 EYO categories: beyond -10, -10 to 0, above 0).The total sample size with non-missing PSMD data and other variables included in model ( 5) and (6) = 190 and was repartitioned as follow for the 3 groups and 3 EYO categories.When categorized by EYO range of 5, some group size were n<4.

Cognition and EYO
Because our clinical measures do not follow a linear progression with EYO, we investigate the effects of mutation position on cognition and clinical measures with disease stage by categorizing EYO by range of 5years.We used the model (7) and model (8), accounting or not for age of onset, respectively.Both models included APOE-ɛ4 status, sex at birth, and education in years as fixed effects and family cluster as random effect.The linear mixed effect model below was applied and run individually for each clinical and cognitive measures (n=3).Due to the small number of comparisons and the exploratory nature of the analysis, the results were not adjusted for multiple comparisons.
(7) / ~0 +  We fit the mixed effect model the relationship between mH and mutation position.The distribution of count is modeled as the zero-inflated negative binomial distribution to account for its dispersion and mass on zero.The effect of mutation position is adjusted for EYO, its interaction with EYO, APOE-ɛ4 status, education, sex, age and MAP.The intra-correlation within families is modeled by the random effect.In the second model of each mH outcome, we remove age from the model.

Mediation Analysis
To evaluate potential indirect effect of mutation position on clinical measures, via markers of small vessel disease.We used the lme4 and Mediation packages in R: cran.r-project.org/web/packages/mediation/vignettes/mediation.pdfEach PiB SUVR regions (n=40) or markers of small vessel disease that showed a significant effect in LME analyses were further evaluated in individual mediation analyses.For the approach, the mutation position corresponded to the independent variable, the clinical measure the dependent variable, and the regional PiB SUVR or marker of small vessel disease the mediator.The models included APOE-ɛ4 status, sex at birth, and education in years as fixed effects and family cluster as random effect.We tested the significance of this indirect effect using bootstrapping procedures.Unstandardized indirect effects were computed for each of 1000 bootstrapped samples, and the 95% confidence interval was computed by determining the indirect effects at the 2.5th and 97.5th percentiles.For both evaluations (regional amyloid or SVD as mediator) EYO is treated as continuous for conditional mediation at EYO=-15, -10, -5, 0, 0.5, and 1. Mediation analyses were corrected for multiple test when evaluating the 40 PiB SUVR regions.Tables S1 -Baseline characteristics of participants per mutation status and sub cohort study    0.12 0.00 0.00 0.58 0.16 0.01 0.00 0.66 0.15 0.00 0.00 0.63 0.14 0.00 0.00 0.84 0.16 0.00 0.00 0.79 0.18 0.00 0 0.13 0.00 0.00 0.42 0.16 0.03 0.02 0.48 0.15 0.00 0.01 0.66 0.14 0.00 0.00 0.89 0.16 0.00 0.00 1.14 0.19 0.00 0.00 0.11 0.00 0.00 0.45 0.13 0.01 0.00 0.43 0.13 0.00 0.00 0.56 0.12 0.00 0.00 0.74 0.13 0.00 0.00 0.96 0.16 0.00 0.00

PiB SUVR Region Pre-200 slope vs. NC slope Post-200 slope vs. NC slope Pre-200 slope vs. Post-200 slope Estimate SE P ROI- adjusted Estimate SE P ROI- adjusted Estimate SE P ROI- adjusted
Annotations: SE -Standard error; ROI = Region of interest; PSMD= peak width of skeletonized mean diffusivity; EYO= estimated years to symptom onset; CI= confidence interval; NC= non-carrier.Cell colors: Yellow highlights for significant p-values, green highlights for the largest cortical and subcortical regional effects in each comparison (pre vs nc, post vs nc, and pre vs post).

Table S3a -
LME Estimates of mean difference SUVR per EYO range in all 40 regions for Pre-200 MC versus NC Annotations: SE -Standard error; ROI = Region of interest; PSMD= peak width of skeletonized mean diffusivity; EYO= estimated years to symptom onset; CI= confidence interval.P-value <0.05 are highlighted in yellow.

Table S3b -
LME Estimates of mean difference SUVR per EYO range in all 40 regions for Post-200 MC versus NC Annotations: SE -Standard error; ROI = Region of interest; PSMD= peak width of skeletonized mean diffusivity; EYO= estimated years to symptom onset; CI= confidence interval.P-value <0.05 are highlighted in yellow.

Table S3c -
LME Estimates of mean difference SUVR per EYO range in all 40 regions for pre-200 MC versus post-200 MC Annotations: SE -Standard error; ROI = Region of interest; PSMD= peak width of skeletonized mean diffusivity; EYO= estimated years to symptom onset; CI= confidence interval.P-value <0.05 are highlighted in yellow.

Table S4 -
LME results mutation by EYO effect on SVD measures (Global PSMD, total, PV, Ant, and Post WMH volumes) P-value <0.05 are highlighted in yellow.In Bold, tests close to significance (P-value <.10).

Table S5a -
LME Estimates of mean difference in Global SVD measures per EYO range for pre-200 MC versus NC, post-200 MC versus NC, and pre-200 versus post-200 MC

Table S5b -
LME Estimates of mean difference in WMH Regions per EYO range for pre-200 MC versus NC, post-200 MC versus NC, and pre-200 versus post-200 MC

Table S6a -
Results of Negative binomial mixed effect models evaluating the effect of mutation position on Deep

Table S6b -
Results of Negative binomial mixed effect models evaluating the effect of mutation position on Deep

WMH volumes per EYO range
Annotations in TablesS6a and S6b: EYO= estimated years to symptom onset; SE -Standard error; WMH = white matter hyperintensity.

Table S7 -
Microhemorrhages prevalence, count, and location per mutation group

Table S8 -
Mutation position effect conditional of EYO on microhemorrhage count per region Significant pairwise comparisons are shaded in yellow, and trends are indicated in bold.Comparisons without enough power or with too large effects are shaded in grey.Annotations: EYO= estimated years to symptom onset; CI= confidence interval.P-value thresholds based on unadjusted comparison: * <0.05; ** <0.005.

Table S9 -
LME Estimates of mean difference in Clinical measures per EYO range for pre-200 MC versus NC, post-200 MC versus NC, and pre-200 versus post-200 MC Significant pairwise comparisons are shaded in yellow, and trends are indicated in bold.Annotations: EYO= estimated years to symptom onset; SE= standard error; CDR-SB= clinical dementia rating sum of boxes; MMSE= mini-mental state examination; CogComp= cognitive composite.P-value thresholds based on unadjusted comparison: * <0.05; ** <0.005.

Table S10 -
Mediation Analyses conditional per EYO with individual PiB regions as mediator Cell with adjusted P-value <0.05 are highlighted in yellow and trend in bold.Annotations: EYO= estimated years to symptom onset; CI= confidence interval.ACME= average causal mediation effect; ADE = average direct effect; TE = total effect.