Decreased Cerebrospinal Fluid Amyloid β 38, 40, 42, and 43 Levels in Sporadic and Hereditary Cerebral Amyloid Angiopathy

Abstract

Objective: Vascular amyloid β (Aβ) accumulation is the hallmark of cerebral amyloid angiopathy (CAA). The composition of cerebrospinal fluid (CSF) of CAA patients may serve as a diagnostic biomarker of CAA. We studied the diagnostic potential of the peptides Aβ38, Aβ40, Aβ42, and Aβ43 in patients with sporadic CAA (sCAA), hereditary Dutch-type CAA (D-CAA), and Alzheimer disease (AD).

Methods: Aβ peptides were quantified by immunoassays in a discovery group (26 patients with sCAA and 40 controls), a validation group (40 patients with sCAA, 40 patients with AD, and 37 controls), and a group of 22 patients with D-CAA and 54 controls. To determine the diagnostic accuracy, the area under the curve (AUC) was calculated using a receiver operating characteristic curve with 95% confidence interval (CI).

Results: We found decreased levels of all Aβ peptides in sCAA patients and D-CAA patients compared to controls. The difference was most prominent for Aβ42 (AUC of sCAA vs controls for discovery: 0.90, 95% CI = 0.82-0.99; for validation: 0.94, 95% CI = 0.89-0.99) and Aβ43 (AUC of sCAA vs controls for discovery: 0.95, 95% CI = 0.88-1.00; for validation: 0.91, 95% CI = 0.83-1.0). All Aβ peptides except Aβ43 were also decreased in sCAA compared to AD (CSF Aβ38: AUC = 0.82, 95% CI = 0.71-0.93; CSF Aβ40: AUC = 0.88, 95% CI = 0.80-0.96; CSF Aβ42: AUC = 0.79, 95% CI = 0.66-0.92).

Interpretation: A combined biomarker panel of CSF Aβ38, Aβ40, Aβ42, and Aβ43 has potential to differentiate sCAA from AD and controls, and D-CAA from controls. ANN NEUROL 2023;93:1173-1186.

Figure 1:

CSF Aβ38, Aβ40, Aβ42 and Aβ43 levels in sCAA patients and controls for discovery. (A) Scatterplots in all panels (depicting median and interquartile range). P-values adjusted for age and sex are depicted. Panel A: CSF Aβ38 levels were decreased in sCAA patients (p=0.015). Panel B: CSF Aβ40 levels were decreased in sCAA patients (p=0.015). Panel C: CSF Aβ42 levels were decreased in sCAA patients (p<0.0001). Panel D: CSF Aβ43 levels were decreased in sCAA patients (p<0.0001). Panel E: ROC analysis showed moderately high to high accuracy levels for discrimination of probable CAA from controls in the discovery group (CSF Aβ38: AUC of 0.65 (95% confidence interval (CI): 0.51–0.79); CSF Aβ40: AUC of 0.63 (95% CI: 0.49–0.75); CSF Aβ42: AUC of 0.90 (95% CI: 0.82–0.99); CSF Aβ43: AUC of 0.95 (95% CI: 0.88–1.00). Abbreviations: sCAA= sporadic cerebral amyloid angiopathy, CSF= cerebrospinal fluid, ROC: receiver operator curve. * = p<0.05, *** = p<0.001.

Figure 2:

CSF Aβ38, Aβ40, Aβ42 and Aβ43 levels in sCAA patients, AD patients and controls for validation (A). Scatterplots in all panels (depicting median and interquartile range). p-values adjusted for age and sex are depicted. Panel A: CSF Aβ38 levels were decreased in sCAA patients compared to controls (p<0.0001), and compared to AD patients (p<0.0001). Panel B: CSF Aβ40 levels were decreased in sCAA patients compared to controls (p<0.001), and compared to AD patients (p<0.0001). Panel C: CSF Aβ42 levels were decreased in sCAA patients compared to controls (p<0.0001) and compared to AD patients (p=0.012), and decreased in AD patients compared to controls (p<0.0001). Panel D: CSF Aβ43 levels were significantly decreased in sCAA patients compared to controls (p<0.001) and in AD patients compared to controls (p< 0.001), but similar between AD and sCAA patients (p=0.86). Panel E: ROC analysis showed moderately high to high accuracy levels for discrimination of probable sCAA from controls in the validation group (CSF Aβ38: AUC of 0.81 (95% CI: 0.71–0.91); CSF Aβ40: AUC of 0.61 (95% CI: 0.46–0.75); CSF Aβ42: AUC of 0.94 (95% CI: 0.89–0.99) CSF Aβ43: AUC of 0.91 (95% CI: 0.83–1.0). Panel F: ROC analysis showed moderately high to high accuracy levels for discrimination of probable sCAA from AD patients (CSF Aβ38: 0.82 (95% CI: 0.71–0.93), CSF Aβ40: AUC of 0.88 (95% CI: 0.80–0.96), CSF Aβ42: AUC of 0.79 (95% CI: 0.66–0.92), CSF Aβ43: AUC of 0.68 (95% CI: 0.53–0.82). Abbreviations: Abbreviations: AD= Alzheimer patients with a CSF biomarker profile indicative of Alzheimer’s disease. AUC= area under the curve, sCAA= sporadic cerebral amyloid angiopathy, CSF= cerebrospinal fluid, *p<0.05, *** = p<0.001.

Figure 3:

Effect of center on CSF Aβ38, Aβ40, Aβ42 and Aβ43 levels. CSF Aβ38, Aβ40, Aβ42 and Aβ43 levels in sCAA patients from the validation cohort: CAA patients from RUMC, LUMC and MGH. In all panels scatterplots are shown, depicting median and interquartile range. The mean age of the sCAA patients from RUMC (n=12) was 71±7 years and 33% was male, the mean age from the sCAA patients from LUMC (n=9) was 74±9 years and 30% was male, and the mean age from the sCAA patients from MGH (n=19) was 63±8 years and 63% was male. Panel A: CSF Aβ38 levels in patients with sCAA from LUMC were decreased compared to the patients with sCAA from RUMC (p=0.046). Panel B: CSF Aβ40 levels were decreased in patients with sCAA from LUMC compared to patients with sCAA from RUMC (p<0.0001). Panel C: CSF Aβ42 levels were decreased in LUMC patients compared to the RUMC patients (p=0.013). Panel D: CSF Aβ43 levels were similar in patients with sCAA from LUMC, RUMC and MGH. p-values are adjusted for age and sex. Abbreviations: CAA= cerebral amyloid angiopathy, CSF= cerebrospinal fluid, LUMC= Leiden University Medical Centre, Leiden, the Netherlands, MGH= Harvard Medical School, Boston, USA, RUMC= Radboud University Medical Center, Nijmegen. *=p<0.05, *** = p<0.001.

Figure 4:

(A) Correlation of CSF Aβ38, Aβ40, Aβ42 and Aβ43 with age, MoCA score, and amongst the Aβ peptides (patients with sCAA and controls for discovery (n=67)). MoCa score (n=12) was only available for a subset of sCAA patients. Spearman correlation coefficients are stated. (B): Correlation of CSF Aβ38, Aβ40, Aβ42 and Aβ43 with age, MoCa score, SVD burden score, and amongst the Aβ peptides (patients with sCAA and controls for validation (n=78). Spearman correlation coefficients are stated. MoCA and SVD burden scores were only available for a subset of sCAA patients (both n=21). (C) Correlation of CSF Aβ38, Aβ40, Aβ42 and Aβ43 with age, MoCA score, SVD burden score, and amongst the Aβ peptides in patients with D-CAA and controls. Spearman correlation coefficients are stated. MoCA (n=21) and SVD burden scores (n=19) were only available for (a subset of) D-CAA patients. The correlation of SVD burden score with the Aβ peptides was adjusted for age. The correlation between CSF Aβ40 and SVD burden score showed a trend (p=0.075). * indicates a significant correlation (p<0.05). Abbreviations: CSF= cerebrospinal fluid; Aβ= amyloid beta, MoCA= Montreal Cognitive Assessment, SVD = small vessel disease.

Figure 5:

CSF Aβ38, Aβ40, Aβ42 and Aβ43 levels in D-CAA patients and controls. (A) Scatterplots in all panels (depicting median and interquartile range). p-values adjusted for age and sex are depicted. Panel A: CSF Aβ38 levels were significantly decreased in presymptomatic and symptomatic D-CAA patients versus their respective controls (both p<0.0001). Panel B: CSF Aβ40 levels were significantly decreased in presymptomatic and presymptomatic D-CAA patients versus their respective controls (both p<0.001), and in symptomatic versus presymptomatic patients (p=0.047). Panel C: CSF Aβ42 levels were significantly decreased in presymptomatic and symptomatic D-CAA patients versus their respective controls (both<0.001), in symptomatic D-CAA patients versus controls (p<0.001), and in symptomatic versus presymptomatic patients (p=0.026). Panel D were significantly decreased in presymptomatic and presymptomatic D-CAA patients versus their respective controls (both p<0.0001). Abbreviations: D-CAA= Dutch-type cerebral amyloid angiopathy, CSF= cerebrospinal fluid. Presymp D-CAA= presymptomatic D-CAA patients, Symp D-CAA=symptomatic D-CAA patients. *=p<0.05, *** = p<0.001.