Iatrogenic Creutzfeldt-Jakob disease with Amyloid-β pathology: an international study

Abstract

The presence of pathology related to the deposition of amyloid-β (Aβ) has been recently reported in iatrogenic Creutzfeldt-Jakob disease (iCJD) acquired from inoculation of growth hormone (GH) extracted from human cadaveric pituitary gland or use of cadaveric dura mater (DM) grafts.To investigate this phenomenon further, a cohort of 27 iCJD cases - 21 with adequate number of histopathological sections - originating from Australia, France, Italy, and the Unites States, were examined by immunohistochemistry, amyloid staining, and Western blot analysis of the scrapie prion protein (PrP^Sc), and compared with age-group matched cases of sporadic CJD (sCJD), Alzheimer disease (AD) or free of neurodegenerative diseases (non-ND).Cases of iCJD and sCJD shared similar profiles of proteinase K-resistant PrP^Sc with the exception of iCJD harboring the "MMi" phenotype. Cerebral amyloid angiopathy (CAA), either associated with, or free of, Thioflavin S-positive amyloid core plaques (CP), was observed in 52% of 21 cases of iCJD, which comprised 37.5% and 61.5% of the cases of GH- and DM-iCJD, respectively. If only cases younger than 54 years were considered, Aβ pathology affected 41%, 2% and 0% of iCJD, sCJD and non-ND, respectively. Despite the patients' younger age CAA was more severe in iCJD than sCJD, while Aβ diffuse plaques, in absence of Aβ CP, populated one third of sCJD. Aβ pathology was by far most severe in AD. Tau pathology was scanty in iCJD and sCJD.In conclusion, (i) despite the divergences in the use of cadaveric GH and DM products, our cases combined with previous studies showed remarkably similar iCJD and Aβ phenotypes indicating that the occurrence of Aβ pathology in iCJD is a widespread phenomenon, (ii) CAA emerges as the hallmark of the Aβ phenotype in iCJD since it is observed in nearly 90% of all iCJD with Aβ pathology reported to date including ours, and it is shared by GH- and DM-iCJD, (iii) although the contributions to Aβ pathology of other factors, including GH deficiency, cannot be discounted, our findings increase the mounting evidence that this pathology is acquired by a mechanism resembling that of prion diseases.

Keywords: Amyloid-β; Cerebral amyloid angiopathy; Pathology; Thioflavin S; iCJD.

Fig. 1

Ring doughnut charts visualizing distributions of codon 129 genotypes, resPrP^Sc types and disease phenotypes among iCJD and sCJD cases examined. a whole iCJD cohort; inner circle: percent distribution of codon 129 genotype (MM, MV, VV); intermediate circle: distribution of resPrP^Sc types (T1, T2, Ti, Ti ± 2, T1 + 2); outer circle: distribution of histopathological phenotype. b GH-iCJD, legend as a; c DM-iCJD, legend as a; d sCJD. T1: type 1; Ti: type intermediate; T2: type 2; Ti + 2: type i + 2; T1 + 2: type 1 + 2; n: number of cases; na: not available; Atyp.: atypical; Undet.: undetermined. Percentages refer to the distribution of codon 129 genotype

Fig. 2

WB profile of resPrP^Sc from iCJD and sCJD controls using high resolution gel electrophoresis. BHs from the frontal cortex (lanes 8 and 9) and cerebellum (lanes 1–7) treated with 100 U/ml PK (~2000 μg/ml) were probed with 3F4. PrP bands were resolved in a 15% Tris–HCl, 20 cm-long gel, and visualized with the near-infrared LI-COR system. The resPrP^Sc profiles from iCJD (lanes 1, 3, 5, 7 and 8) and matching sCJD subtypes (lanes 2, 4 and 6) are similar as shown by the co-migration of the unglycosylated resPrP^Sc type 2 (19) (lanes 1–4) and type 1 (21) (lanes 6–8). The different thickness of the type 2 (19) band is within the variability range for this subtype. The unglycosylated resPrP^Sc band of ~20 kDa, also identified as type i (i), is visible in GH-iCJDMVi + 2 (lane 3), sCJDMVi + 2 (lane 4) and GH-iCJDMMi case (lane 5); a thin band of ~20 kDa (arrowhead) is also present in CJDMM1. An ~18 kDa fragment (18) is also present in CJDVV2 and CJDMVi + 2

Fig. 3

Aβ and tau pathology of iCJD, sCJD and AD. a and b typical Aβ CP (arrows) from the frontal cortex of iCJD (a) and AD (b); inset a: Aβ CP at higher magnification immunoprobed for Aβ (top) or stained with Thioflavin S (bottom); inset b: Thioflavin S-positive Aβ CP. c diffuse plaques from the frontal cortex of a case of sCJD; inset: Thioflavin S-negative staining of plaques (dashed rectangle). d co-localization of Aβ (cyan dye) and PrP (brown dye) immunoreactivity in the same plaque from cerebral (I and III) and cerebellar (II) cortices of iCJD. I: plaque exhibiting a PrP-positive core surrounded by an Aβ immunoreactive crown; II-III: plaques showing either the immunostaining opposite to that of I, with Aβ-positive core and a PrP immunoreactive crown (II), or random distribution of PrP- and Aβ-reactive aggregates (III). e-g Aβ deposits (e) and Thioflavin S-positive staining (f and g) of vessels in frontal cortex (e), hippocampus (f) and cerebellum (g) in iCJD; f: DG, Dentate gyrus; g: asterisk, Aβ CAA in the subarachnoid space between two cerebellar folia. h Aβ CP predominantly affecting the granular layer (Grl. L) and Purkinje layer of the cerebellum in iCJD (case 9, Table 1); arrow indicates Aβ CP; large and small insets: enlargements of Aβ CP identified by the dashed rectangle in the main figure and a Thioflavin S-positive Aβ CP, respectively. i-j hippocampus free of tau pathology in iCJD (i) but severely affected in AD (j). k Tau-positive DN associated with an Aβ CP in frontal cortex of iCJD (case 8, Table 1) (dashed square); top inset: enlargement of tau reactive DN identified in dashed square; bottom inset: DN associated with a PrP kuru plaque in iCJD (occipital cortex, case 9, Table 1). l globose NFT (arrow) and neuropil threads (arrowhead) in the nucleus basalis of Meynert of iCJD (case 6, Table 1); inset: flame-shaped NFT in the temporal neocortex of the same iCJD case. Abs: 3F4 (d), 4G8 (a-e, h) and AT8 (i-l) to PrP, Aβ and phosphorylated tau, respectively; Thioflavin S to Aβ (a-c, f-h)

Fig. 4

Aβ phenotype and age distribution in Aβ-positive iCJD and control cases. a: iCJD (N = 11); b: sCJD (N = 9); c: non-ND (N = 2). Scores for the Aβ phenotype is relative to the presence of either Aβ CP or CAA (score 1), or combinations of both (score 2). Brackets below dotted line in the X-axis underline cases of the same age; GH: growth hormone; DM: dura mater

Fig. 5

Graphic representation of demographic data, NFT pathology, codon 129 genotype and resPrP^Sc type in iCJD with (+) and without (−) Aβ pathology. a-c age, disease duration, incubation period, and NFT pathology relative to All iCJD (a), GH-iCJD (b) and DM-iCJD (c); yrs.: years; mo: months. d and e ring doughnut charts visualizing genotype frequency at codon 129 (d) and resPrP^Sc type distribution (e) in iCJD with (iCJD-Aβ⁺) and without (iCJD-Aβ⁻) Aβ pathology; M: methionine; V: valine; T1: type 1; T2: type 2; Ti: type intermediate; Ti + 2: types intermediate +2; T1 + 2: types 1 + 2

Fig. 6

Type 1 or capillary CAA in one iCJD (a and b) and one AD (c and d). a-d Aβ deposits along the basement membrane of capillaries in the parahippocampal gyrus (a), molecular layer of the cerebellum (b) and hippocampus (c and d). The mean diameter of capillaries in longitudinal sections resulted from three (a and b) and six (d) measurements along the capillary whereas the diameter of the capillary in cross-section in c is the mean diameter calculated by the software (Image-Pro Plus)