Cerebral Amyloid Angiopathy and Atrial Fibrillation: An up to Date Case Report

Abstract

Concurrent Cerebral Amyloid Angiopathy (CAA) and Atrial Fibrillation are becoming an increasingly common dilemma in clinical practice due to the aging population and the comorbidities associated with it. In such patients, the physician must appreciate and strike the difficult balance between the risk of ischemic strokes from atrial fibrillation on one hand, and that of intracerebral hemorrhage from coexisting CAA on the other. Anticoagulation is necessary for the former but potentially deleterious for the latter. In this case report, we present the case of a 67-year-old woman with a long history of atrial fibrillation on rivaroxaban who recently began to experience recurrent transient neurological deficits that were later diagnosed as amyloid spells related to concomitant CAA. While there is no clear-cut consensus in published literature on how to best manage these patients regarding the use of anticoagulation, it is recommended to involve a multidisciplinary team for optimal management of these patients.

Introduction

Cerebral amyloid angiopathy (CAA) is characterized by progressive deposits of amyloid beta-peptide in the media of the small and medium vasculature of the cerebral cortex and leptomeninges. Cerebral amyloid angiopathy is one of the leading causes of spontaneous intracerebral hemorrhage. The hemorrhages occur in varying sizes and locations which ultimately determine the patient’s clinical presentation; however, the vast majority of patients tend to be asymptomatic. ¹ With increase in amyloid deposition and number of intracerebral hemorrhages (ICH), a patient’s presentation can progress from being asymptomatic or having mild deficit in cognition such as memory loss, to hemiparesis, hemiplegia, and depressed consciousness due to involvement of the pyramidal motor tracts and compression of the reticular system, respectively. ²

A characteristic clinical presentation associated with CAA is termed as “amyloid spells” which are transient focal neurological episodes (TFNEs) of positive symptoms such as recurrent transient paresthesias and negative symptoms such as sudden onset limb weakness lasting for varying short durations. ³ Transient focal neurological episodes are a predictive factor of high early risk of symptomatic ICH. ⁴ Despite the presence of many potential predisposing risk factors, the only evident risk factor for the development of CAA is its relationship with increased amyloid deposits found in elderly patients over the age of 60. The 2010 modified Boston criteria, which encompass the imaging, clinical, and pathological parameters of CAA, were proposed in order to standardize the diagnosis. Moreover, the inclusion of superficial siderosis in the modified criteria enhanced the sensitivity for CAA-related hemorrhage, without loss of specificity. ⁵

There are currently no effective treatments available to cure or slow down the progression of CAA. Nevertheless, awareness of the diagnosis is critical because of the increasing number of cases and the challenge or dilemma posed by concurrent diseases such as Atrial fibrillation (A-Fib) that tend to require lifelong anticoagulation. One must then walk the delicate balance between managing the risk of cerebral thromboembolism associated with A-Fib and the risk of cerebral hemorrhage associated with CAA. ⁶

In this report, we present the case of a 67-year-old woman with A-Fib and onset of amyloid spells.

Case Presentation

A 67-year-old Emirati woman presented with a 12-month history of episodes of transient weakness of her right arm which occurred suddenly, were unrelated to activity, and resolved spontaneously within 15 min. The weakness was not associated with pain or numbness in the arm and there was no facial drooping or speech difficulties during these episodes. The episodes had recently increased in frequency to 1 or 2 episodes per day within several weeks from presentation.

Her past medical history was notable for A-Fib, hypertension, dyslipidemia, osteoporosis, and hypothyroidism which were all well-controlled. Her routine medications included Rivaroxaban (20 mg/OD), Valsartan (160 mg/OD), Aspirin (75 mg/OD), Amlodipine (5 mg/OD), Atorvastatin (80 mg/OD), Levothyroxine (100mcg/OD), Amitriptyline (25 mg/TID), and Alendronate Sodium. Her family history was insignificant. She has never smoked, consumed alcoholic beverages, nor used recreational drugs. Review of systems was notable for long-standing neck pain attributed to cervical spondylosis.

On neurological examination, there was no cranial nerve deficit, sensory loss, nor pronator drift in either arm. Inverted supinator jerk was observed bilaterally during tendon reflex assessment, but plantar response was downgoing bilaterally and Hoffman’s sign was absent. The patient was well oriented to time, place, and person with no evidence of cognitive impairment. However, a comprehensive mini-mental status exam was not performed.

The patient underwent plain MRI of the brain which revealed multiple, variable sized, bilateral foci of dark signal during T2 star-weighted angiography (SWAN) sequences distributed along the cortex and subcortex of both cerebral hemispheres with additional involvement of the cerebellar hemispheres (Figure 1A). Intracranial and extracranial MR angiography did not show any hemodynamically significant stenosis, standard scalp EEG was normal, and transthoracic as well as transesophageal echocardiogram did not show any intramural cardiac thrombi.

In light of her clinical presentation (of recurrent sudden arm weakness which fits the description of “Amyloid spells”) and the findings and distribution of multiple cerebral and cerebellar microbleeds on MRI, a diagnosis of CAA was considered most likely. According to the 2010 Modified Boston Criteria, the patient falls under the category of “Probable CAA,” without supporting pathology as she did not undergo a cortical biopsy. ⁵

Moreover, due to the fact that she was on Rivaroxaban for A-Fib, which may increase her risk of intracranial bleeding in the setting of coexisting CAA, she was referred back to her Cardiologist for a discussion about possible alternative treatment options other than oral anticoagulation. The patient was offered the option of ablation of the left atrial appendage by the cardiologist as a substitute to oral anticoagulation. While this discussion continued, the patient was considered for anti-epileptic treatment which has been shown by some authors to have some efficacy against amyloid spells. ⁷ During the last multi-disciplinary discussion with the patient and family, they expressed their wish to not pursue any interventions. The increased risk of significant intracerebral bleed was discussed, and the family accepted to continue with anticoagulation alone and Aspirin was discontinued. Fortunately, the amyloid spells stopped spontaneously before anti-epileptics were started. Patient continues to be followed closely at the Cardiology out-patient clinic and has not had any complications in the last 6 months.

Discussion

Concurrent CAA and Atrial Fibrillation (A-Fib) pose a clinical dilemma in terms of managing the risk of thromboembolism and hemorrhage simultaneously. However, despite extensive literature in past years promoting hesitancy in the use of oral anticoagulation in patients with CAA, a more up to date approach suggests careful weighing of A-Fib risks and benefits of anticoagulation in the given clinical context.

The clinical presentation of patients with CAA is variable, with only a small fraction of people developing CAA-related clinical events, including symptomatic ICH. ¹ A recently recognized presentation of CAA is TFNEs; which may hinder reaching a definitive diagnosis, as the presentation may overlap with Transient Ischemic Attacks (TIA), epileptic seizures, or migraine aura. This heterogeneous presentation may lead to unregulated use of anticoagulant therapy, increasing the risk of ICH. ²

The patient described in this report presented with transient, self-resolving episodes of upper extremity weakness, which, with the aid of appropriate imaging, proved to be due to amyloid spells and unlikely to be related to TIA, and this prompted consideration of modification of her existing treatment for A-Fib. The clinical differentiation between CAA and TIA is of clinical significance, as there is a risk of inappropriate use of anticoagulation and antiplatelet therapy. Despite this conundrum, this therapeutic approach can be considered in the setting of comorbid Atrial Fibrillation in such patients. Stanton et al. identified several distinct clinical features of CAA-related TFNEs compared to TIA, demonstrating that the former more commonly manifest with migratory spread, affect sensation, and reoccur in a stereotyped manner. ⁸ In the absence of concurrent A-Fib, the knowledge of these symptoms can guide the management strategy.

Cerebral microbleeds (CMBs), which may be seen on imaging studies of patients with CAA (eg, our patient), may be harbingers of both ICH and ischemic stroke. This was shown by a meta-analysis where in patients with lobar microbleeds, the pooled absolute event rate for future ischemic stroke was 9.3%, compared to 3.6% for ICH. ⁹ Puy et al ¹⁰ described that in patients with ischemic Stroke/TIA and A-Fib, the presence of CMBs is an indicator of greater benefit from anticoagulation, due to their higher absolute rates of ischemic stroke. Furthermore, CMBs are markers of both ischemic and hemorrhagic events and mortality in embolic stroke of undetermined source ¹¹ ; however, the absolute rates of ischemic events are higher in non-ICH populations.

The feared consequence of CAA is a spontaneous lobar ICH, ¹ which has a high risk of recurrence, ⁴ about 5-fold greater risk than hypertensive ICH. ¹² The risk of ICH in CAA is even further increased by additional use of anticoagulation and the presence of disseminated cortical superficial siderosis, cortical subarachnoid hemorrhage, and CMBs on imaging.^7,13 This risk has generally constituted a major impediment to prescribing anticoagulants in such patients. The review by Cannistraro et al ¹³ concluded that the risk of ICH is associated with the dose of anticoagulation, depending on the number of CMBs. In particular, patients with 2 or more CMBs are at an increased risk of ICH, and would benefit from a multidisciplinary approach discussing the case-specific risks and benefits. Puy et al ¹⁰ agreed that CMBs have a prognostic value in terms of ICH risk; however, the authors concluded that the presence of CMBs does not modify the clinical outcomes of antithrombotic therapy, hence may not be a reliable guide for the choice of therapy.

With regards to management, some authors have suggested that newer oral anticoagulants (OAC) such as Non-Vitamin K antagonist OAC may be safely used in patients with ICH from CAA, given their absolute rates of ICH being similar to Aspirin monotherapy. ¹⁴ Kang-Ho et al ¹⁵ compared the use of OAC and antiplatelet therapy and concluded that the benefit of OACs can outweigh the risks compared to antiplatelets in the secondary prevention of stroke in patients with A-Fib and multiple or lobar CMBs. More specifically, the patients in their analysis who were on OACs had significantly lower rates of recurrent ischemic strokes, major adverse cardiovascular events, and all-cause death as opposed to the antiplatelet group; whereas the rates of ICH were similar. A similar paper by Hilt et al. concluded that even in the setting of left ventricular thrombus, evidence for treatment with OACs in patients with CAA is still deficient; however, it is reasonable based on its efficacy and safety profile. ¹⁶ Consequently, oral anticoagulation might actually be of benefit in this patient population who have A-Fib coexisting with CAA.

Finally, a conceivable approach to the issue of anticoagulation in these patients is to consider Left Atrial Appendage Closure (LAAC), which allows for avoidance of long-term oral anticoagulation. A cohort study evaluating LAAC in patients with CAA has proven this method to be safe and effective in reducing stroke risk in patients with CAA; thus, allowing avoidance of oral anticoagulation and its associated risks. ⁶

PROTECT AF and PREVAIL are two randomized trials which compared LAAC to warfarin therapy for stroke prevention in patients with A-Fib. ¹⁷ However, this data for ischemic stroke prevention via LAAC is difficult to interpret. While LAAC can be considered for selected patients (those with a high bleeding risk), the periprocedural complications and caveats of the available data need to also be taken into account.

Conclusion

Managing patients with CAA who have a history of A-Fib presents a challenge to clinicians. This is largely due to the complex risk-benefit analysis of anticoagulation that is required. Aside from anticoagulation treatment, another viable option which can reduce the risk of stroke is LAAC. A multidisciplinary team approach involving neurologists, cardiologists, and radiologists is recommended to effectively manage these patients from all perspectives. Patients should be well informed of the situation and the various treatment options including their pros and cons before a decision is reached. Furthermore, each case should be individualized as the indication for anticoagulation will usually need to be assessed and risk stratified. This ensures the need for anticoagulation is justified and management plan tailor-made.

Appendix.

Figure 1A.

(A) Red box indicates a dark signal identifying a microbleed along the frontoparietal region during a star-weighted angiography sequence; multiple small foci may also be seen scattered around the right cortex. (B) Mild subcortical arteriolar sclerotic leukoencephalopathy can be appreciated in the parieto-occipital region. Multiple small foci of dark signals are also observed.

ORCID iDs

Umberto Battistin https://orcid.org/0000-0002-5003-0254

Noora Al Qassim https://orcid.org/0000-0002-0581-6629