Corticosteroid Treatment Guided by Peripheral Blood Mononuclear Cell Counts in Cerebral Amyloid Angiopathy-related Inflammation: A Case Report


 Background: Cerebral amyloid angiopathy-related inflammation (CAA-I), a rare variant of cerebral amyloid angiopathy, is one of treatable causes of rapidly progressive dementia. CAA-I usually occurs in the elderly and corticosteroids are most often used for treatment. Thus it is more necessary to develop an individualized dosing regimen to maximize efficacy yet minimize adverse effects of corticosteroids. We report a CAA-I patient successfully treated by corticosteroid therapy guided by dynamic monitoring of peripheral blood mononuclear cell (PBMC) count. Case presentation: A 68-year-old female presented with subacute step-wise cognitive decline. Medical history and current examination revealed multiple risk factors for cerebral amyloid angiopathy, including apolipoprotein E ε4/ε4 genotype, hyperhomocysteinemia, hypertension, and chronic renal failure, while brain susceptibility-weighted magnetic resonance imaging revealed diffuse lobar microbleeds associated with extensive lesions disturbed throughout cerebral grey and white matter on T1- and T2-weighted MRI. The CD19+ B-cell fraction of PBMCs was markedly elevated, as were inflammatory markers such as erythrocyte sedimentation rate and C-reactive protein. Cerebrospinal fluid total protein and T-tau levels were also elevated, while Aβ42 was low and P-tau levels were normal. Corticosteroid treatment guided by dynamic CD19+ B-cell number monitoring reversed these inflammatory signs, imaging manifestations, and acute cognitive decline. Conclusions: Corticosteroid treatment individualized by dynamic monitoring of PBMC CD19+ B-cell fraction may provide for maximum therapeutic efficacy with minimal adverse effects in patients with CAA-I.


3
Cerebral amyloid angiopathy (CAA)-related inflammation (CAA-I) is a rare CAA subtype [1] characterized by reversible encephalopathy with acute or subacute cognitive decline, seizures, headache, and focal neurologic signs [2]. CAA-I can be divided into two subtypes, inflammatory CAA and amyloid-β (Aβ)-related angiitis, according to histopathology [2]. The majority of patients with CAA-I (~80%) respond to immunotherapy, but roughly 30% have symptomatic relapse [1]. Among the immunotherapies, corticosteroids are most often used for CAA-I treatment. Considering CAA-I usually occurs in the elderly, it is more necessary to develop an individualized dosing regimen to maximize efficacy yet minimize adverse effects of corticosteroid. Herein, we report a case of CAA-I successfully treated by corticosteroid therapy guided by peripheral blood mononuclear cell (PBMC) counts.

Case Presentation
A 68-year-old female presented with subacute step-wise cognitive decline and behavioral abnormalities for the previous eight months. She was diagnosed with chronic nephritis 30 years ago and with hypertension and chronic renal failure 1 year prior to the current presentation. The family history was unremarkable. Head computed tomography (CT) plain scan conducted at another hospital after current symptom onset revealed multiple punctate high-density lesions and extensive low-density lesions in both cerebral hemispheres. Accompanying magnetic resonance imaging (MRI) showed extensive hyperintense lesions on both T2-weighted images (T2WI) and fluid-attenuated inversion recovery (FLAIR) images without gadolinium enhancement on T1-weighted images (T1WI).
Brain susceptibility weighted imaging (SWI) was not conducted at that time. The patient is illiterate and scored 14/30 on the Mini-Mental State Examination (MMSE). The diagnosis was undetermined at that time and no specific therapies were given except for antihypertensive agents.

4
On admission at our institution, blood pressure was 143/102 mmHg. The patient was fully alert but disoriented and cognitive function tests revealed subacute cognitive decline, with MMSE score of 4/30, Montreal Cognitive Assessment (MoCA) score of 0/30, and Activities of Daily Living (ADL) scale score of 44/64. Results of cranial nerve and sensory examinations were negative, and motor examination revealed positive palmomental reflex bilaterally without weakness or dystonia of limbs.
In accord with prior neuroimaging results, brain CT plain scans showed multiple punctiform hyperdensities in the bilateral lobes ( Figure 1-A), and brain MRI scans showed extensive hypointense lesions on T1WI (Figure 1 Based on these findings, the patient was diagnosed with probable CAA-related inflammation [3] and received intravenous methylprednisolone (IVMP) starting at 500 mg/d, followed by oral prednisone with slow tapering guided by monitoring of PBMC counts presented with only one of the major symptoms (rapidly progressive dementia) without seizures, headaches, or focal neurologic deficits. However, brain CT scans at the time of dementia onset showed multiple lobar hemorrhages, a frequent manifestation of CAA in patients over the age of 55. Neuroinflammation due to CAA should be highly suspected in rapid onset dementia cases with patchy or confluent white matter hyperintensities on T2WI. Extensive lobar microbleeds on SWI acquired at our institution further supported probable CAA-I according to clinicoradiological criteria [3]. It is well known that both SWI and gradient-echo MRI are sensitive to hemorrhage, especially microhemorrhage and chronic-phase hemorrhage. Indeed, hypointensities on SWI are widely considered a biomarker for CAA reflecting underlying brain lesions [4].
We also performed a brain biopsy to determine the exact pathological subtype of CAA-I but this yielded no definitive results. This false-negative is not unusual, however, as amyloid-laden vessels and other signs may be sparse in certain regions depending on disease stage and so are easily missed. Indeed, only 57% of brain biopsies in patients with 8 dementia are diagnostic [5]. There is a general region-specific temporal pattern to vessels affected by Aβ accumulation in CAA. Initially, arteries of the leptomeninges show signs of pathology, followed shortly by penetrating arterioles in the neocortical grey matter.
Vessels to the posterior aspects of the brain are particularly impacted, most severely in the occipital lobe and parietal region, although in some cases the frontal cortex is the primary site of CAA pathology. In the next stage, vessels in the olfactory cortex, hippocampus, and cerebellum may be affected. Finally, deep grey and white matter are involved [6]. The biopsy site in this case was the deep white matter of the left occipitotemporal lobe, a region likely affected by inflammation rather than amyloid-laden vessels. Despite failing to obtain direct evidence for amyloid deposits, histopathology and the absence of gadolinium enhancement on T1WI ruled out brain tumor, viral and autoimmune encephalitis, and primary angiitis of the central nervous system.
The case patient demonstrated elevated ESR, CRP, and CSF protein with no signs of infection. It has been reported that about 33% of CAA-I cases exhibit elevated levels of such inflammatory markers in the peripheral blood, while 45% show pleocytosis in the CSF, and 71% show mildly elevated protein in the CSF [2]. It has also been reported that half of CAA patients meet the pathological diagnostic criteria for Alzheimer's disease (AD) [7]. However, this case demonstrated low CSF Aβ42, high T-tau, and normal P-tau (181p), further supporting CAA and excluding the possibility of AD, which is associated with high CSF P-tau. A previous study also documented lower CSF Aβ40 and Aβ42 concentrations in CAA compared to both controls and patients with AD [8]. Furthermore, the presence of microbleeds was reported to be associated with lower CSF Aβ42 in AD and vascular dementia, especially in apolipoprotein E ε4 carriers [9].
This case carried an APOE epsilon 4 (ε4) /epsilon 4 (ε4) genotype, which further supported a diagnosis of CAA as there is a dose-dependent association between APOE ε4 and 9 sporadic CAA [10]. Patients carrying the APOE epsilon 4 (ε4) or epsilon 2 (ε2) alleles are at greater risk for CAA-related hemorrhage than those with the common APOE epsilon 3 (ε3) allele [11]. Among the vessels affected by Aβ deposition, capillary involvement is strongly associated with the APOE ε4 allele [12]. Moreover, the APOE ε4/ε4 genotype is strongly associated with the occurrence of inflammation induced by CAA [1].
This case also exhibited hyperhomocysteinemia, which may have increased the disease load. Animal studies have demonstrated that high levels of homocysteine can lead to more severe CAA and parenchymal Aβ deposition [13]. Of note, this case also showed markedly elevated levels of TGAb and TPOAb, so Hashimoto encephalopathy (HE) should be considered in the differential diagnosis because this disease also causes subacute onset cognitive dysfunction and confusion with altered consciousness. However, MRI findings were inconsistent with the neuroimaging manifestations of HE. In HE, most MRI sequences are normal, but patients may demonstrate cerebral atrophy or nonspecific signal abnormalities in the subcortical white matter on T2WI [14].
This case also had long histories of hypertension and chronic renal failure, both of which can aggravate microbleeds. Amyloid-related pathology and hypertensive cerebral small vessel disease have synergistic effects on the progression of lobar microbleeds [15].
There is also evidence from animal studies for an age-independent effect of tubulointerstitial kidney damage on brain Aβ accumulation, which can be reinforced by the co-presence of cerebral microhemorrhages [16].
This case presented evidence of arteriosclerosis, but a statin rather than an antiplatelet agent was administered to prevent future ischemic cerebrovascular events due to the high risk for ICH in CAA, particularly in patients with severe microbleed burden or a history of lobar hemorrhage. A prospective cohort study found that aspirin induced an approximately fourfold increase in the risk for both ICH occurrence (when ≥ 5 microbleeds are present) and recurrence in CAA [17]. In addition, we administered several antihypertensive agents to control blood pressure because active antihypertensive treatments have been shown to reduce the risk of CAA-related and hypertension-related ICH (by 77% and 46%) [18].
Currently, there are insufficient data to recommend against statin agents so they were administered as preventative therapy [19].
We analyzed PBMC counts to assess immune status and the results showed an increased CD19+ B-cell fraction among PBMCs at baseline and a gradual return to normal over the course of treatment. It is generally believed that the CD19+ B-cell fraction reflects the activity of humoral immunity, which is strongly involved in CAA-I pathogenesis. Several studies have shown that anti-amyloid autoantibodies are increased in the CSF during the acute phase of inflammation and return to control levels during remission [20,21].
Unfortunately, this case was not tested for CSF anti-amyloid antibodies. However, it should be noted that the diagnostic value of CSF anti-amyloid antibodies has not been definitively demonstrated. Alternatively, we conducted dynamic monitoring of PBMC    renamed_71880.pdf