A case of cerebral amyloid angiopathy-type hereditary ATTR amyloidosis with Y69H (p.Y89H) variant displaying transient focal neurological episodes as the main symptom

Abstract

The patient was a 43-year-old Japanese man. At age 41, he underwent carpal tunnel release surgeries for bilateral carpal tunnel syndrome. At age 42, he began experiencing recurrent episodes of nausea and dizziness followed by diminished consciousness and momentary lightheadedness approximately once a month. Nine months later, he suffered from a bout of syncope and was referred to our hospital. General and neurological examinations upon admission revealed no abnormal findings. Routine laboratory tests were all normal. Cerebrospinal fluid (CSF) analysis showed slightly elevated cell count and protein. Electrocardiography, electroencephalography, and nerve conduction study displayed no remarkable results. Gadolinium-enhanced magnetic resonance imaging showed a contrast effect along the meninges of the brain and spinal cord (Figure 1(A–C)). Open brain biopsy from the arachnoid mater and cerebral parenchyma of the frontal lobe was performed for further evaluation. The biopsied tissue showed arachnoid membrane hypertrophy and subarachnoid vessel dilations with massive deposition of amyloid (Figure 1(D)). Amyloid deposition was also observed on the cerebral surface (Figure 1(E)). All amyloid deposits were specifically immunolabelled with the anti-TTR antibody (Figure 1(F, G)). On the cerebral surface, ATTR amyloid deposited mainly on the pia mater and subpial area (Figure 1(H)). The cerebral parenchyma itself was substantially preserved. However, positron emission tomography using Pittsburgh compound B (PiB-PET) [1] displayed no abnormal retention in the brain (Figure 1(I)). Transthyretin (TTR) gene analysis identified heterozygosity for the c.265T>C, p.Y89H (Y69H) variant. Matrixassisted laser desorption ionization/time-of-flight mass spectrometry (MALDI/TOF MS) analysis of serum TTR detected an ion peak of molecular mass (m/z) 13,762 originating from wild-type TTR, whereas that of CSF TTR detected a doublet-shaped peak originating from both wild-type and variant TTR (Supplementary Figure S1). After the evaluation at our hospital, the patient began experiencing transient episodes of sensory disturbance and weakness in the left side of the body lasting for approximately 30min. He was diagnosed as having transient focal neurological episodes (TFNEs) attributed to ATTR-type cerebral amyloid angiopathy (CAA) and treated with carbamazepine, which reduced TFNEs. To date, three hereditary ATTR amyloidosis families with the Y69H variant have been reported [2–4], and all patients with this variant displayed central nervous system (CNS) symptoms similar to those in our patient. The most likely pathomechanism underlying the TFNEs in our patient was cortical spreading depression or focal seizures, which was supported by the clinical effectiveness of carbamazepine. Concerning the pathogenesis of CNS-type hereditary ATTR amyloidosis, in vitro and in vivo studies have demonstrated that CNS-type variants are highly unstable and exhibit very low secretion efficiency from hepatocytes into the serum, while the choroidal cell line is more permissive in its ability to secrete unstable variants into the CSF. Indeed, MALDI/TOF MS analysis of our patient revealed undetectable serum variant TTR but a considerable amount of the Y69H variant in the CSF. The most intriguing aspect of this case was the PiB-PET findings. An earlier PiB-PET study of a large series of posttransplant hereditary ATTR amyloidosis patients demonstrated that PiB-PET could detect ATTR amyloid deposition in the CNS approximately 10 years before the onset of TFNEs [1]. On the other hand, PiB-PET results were negative in our patient despite recurrent episodes of TFNEs and considerable amyloid deposition in brain biopsy specimens. Recently, Pilebro et al. reported that the PiB-retention index of type B fibrils (composed of full-length TTR) was significantly higher than that of type A fibrils (mixture of fulllength TTR and TTR fragments), suggesting that C-PiB bound much more strongly to type B than to type A [5]. Thus, it was possible that the amyloid fibril deposits in our patient were type A to produce a false-negative PiBPET result.