Rapidly Sequential and Fatal Hemorrhaging in a Case of Cerebral Amyloid Angiopathy

Background

Cerebral amyloid angiopathy (CAA) is an increasingly recognized cause of lobar intracerebral hemorrhage (ICH) and cognitive impairment in the aging population [1]. CAA accounts for 15% of non-traumatic ICHs in the older population (greater than 60 years of age) [2]. Definitive diagnosis of CAA requires histopathological analysis, which is of minimal practical use. Thus, magnetic resonance imaging (MRI) is the most reliable option for clinical diagnosis of suspected CAA and can be prognostic for recurrent hemorrhaging [3,4].

CAA is a vasculopathy resulting from deposition of β-amyloid protein in the media and adventitia of small- and mediumsized arteries as well as capillaries in the leptomeninges and cerebral cortex [1,4]. CAA-related bleeds have a posterior predilection with most bleeds occurring in the occipital lobe [5]. The pathophysiology of microbleeds and ICH in CAA is not well understood, but it is thought to be the result of vessel weakening and rupture secondary to amyloid deposition [1].

Little evidence has been established pertaining to the time course of recurrent CAA-related microbleeds or larger hemorrhages. Although several risk factors have been associated with an increased risk of ICH in CAA, there are no current treatment guidelines specific to recurrent hemorrhaging in CAA [6].

Here we present a rare case of sequential and fatal lobar hemorrhaging in the setting of probable CAA in a 63-year-old female patient over a one-month period.

Case Report

A 63-year-old female with a history of multiple sclerosis, chronic right upper extremity weakness, heavy tobacco use, depression, migraines, hypertension, and dyslipidemia, who was taking 81 mg aspirin daily, was found unresponsive by her family.

The patient was emergently seen in a Chicago-area hospital where computed tomography (CT) of the head demonstrated left frontal lobe hemorrhage with intraventricular extension. The patient was admitted to the intensive care unit where she was hemodynamically stable, but exhibited worsening mental status. She was aphasic with right hemiplegia, and was unable to follow commands. A repeat CT head demonstrated a 0.8 cm midline shift, associated edema and mass effect, and acute intra-axial hematoma, measuring 7.5×4.2 cm axially, centered in the left frontal lobe with intraventricular extension and notable left subarachnoid hemorrhage (Figure 1A). Emergent craniotomy was performed without complication. An intraoperative diagnosis of arteriovenous malformation was made, which pathology defined as a cavernoma.

During the following days the patient exhibited improving mental status, however, she remained aphasic with right hemiplegia. One week later she was found to be more lethargic and a follow-up head CT demonstrated a new right frontal lobe hemorrhage (Figure 1B). A brain MRI demonstrated numerous microbleeds on susceptibility weighted imaging (SWI) suggestive of CAA (Figure 1D–1F).

After this acute event, the patient continued to regain strength bilaterally. Two weeks later the patient was started on subcutaneous heparin (5,000 units every eight hours) for deep venous thrombosis prophylaxis.

Three days later the patient was found to be more confused with increased drowsiness and left facial droop. A head CT demonstrated new right parietal cortical and subcortical hemorrhages. Simvastatin and subcutaneous heparin were immediately discontinued. Coagulation studies were within normal limits.

The next day the patient became more lethargic and a head CT revealed three new right frontal bleeds (the largest measuring 4.7 cm) with edema and midline shift (Figure 1C). A decision was made with the family to pursue comfort care only. The patient died three days later.

Discussion

Cerebral amyloid angiopathy (CAA) should be suspected in patients over the age of 55 years old with spontaneous lobar microbleeds or a larger hemorrhage in the setting of normotension [5]. Prognosis in a patient with CAA is correlated with the burden of microbleeds at initial presentation [3]. Several risk factors are associated with CAA-related bleeds including age, Alzheimer disease, apolipoprotein E ε2 and ε4 alleles, head trauma and, of controversy, medications including thrombolytics, anticoagulants, anti-platelets, and cholesterol-lowering statins [1,5–7].

When the diagnosis of CAA with microbleeds is probable, strict medical management is imperative [6]. Current treatment of symptomatic CAA-related microbleeds and hemorrhages follows the guidelines for the management of all spontaneous intra-cerebral hemorrhage [5,8]. Because the bleeding in CAA is often sporadic, effective prevention strategies are imperative [1].

The PROGRESS trial demonstrated that blood pressure-lowering treatment provides protection against all types of intracerebral hemorrhage including CAA-related hemorrhage (77% risk reduction), which may represent a practical therapeutic approach [9]. Our patient’s blood pressure ranged from 172/90 mm Hg to 96/42 mm Hg and elevations were not correlated with ICH (Figure 2). The spatial pattern of the microbleeds was strictly posterior and lobar, inconsistent with the typical non-lobar, deep location of hypertensive intracerebral hemorrhages [10,11].

A pathologic diagnosis of cavernoma was made during the initial craniotomy, which may have marginally contributed to the initial hemorrhage. However, the size of the initial hemorrhage was magnitudes larger than the mean size of the hemorrhage (13.4±7.3 mm) described by Jeon et al. [12] in 410 patients with cavernomas. The sequential hemorrhaging was also inconsistent with an etiology of cavernoma; and multiplicity of cavernomas has not been shown to be a risk for subsequent hemorrhage [13].

Additionally, the literature is replete with reports of medication-associated CAA-related bleeds, explicitly associated with thrombolytic, anticoagulant, antiplatelet, and statin therapy [2,7,9,14–21]. However, most controversially is the role of statin therapy in cerebral microbleeds and ICH. Reports have demonstrated that statin therapy can increase the risk of cerebral microbleeds [19,20], and Westover et al. [21] recommended the avoidance of statins in patients with a history of ICH. Alternatively, a meta-analysis of 31 randomized controlled trials of statin therapy did not find an association between statin therapy and increased risk for ICH [22]. The exact contributory role of statin therapy in our case was unclear. It seemed clear that the rate of hemorrhaging was increased by concomitant use of subcutaneous heparin, but ongoing statin therapy may have been an additional risk factor.

According to the guidelines for the management of spontaneous intracerebral hemorrhages adopted by the American Heart Association and American Stroke Association [8]: “After documentation of cessation of bleeding, low dose subcutaneous low-molecular-weight heparin or unfractionated heparin may be considered for prevention of venous thromboembolism in patients with lack of mobility after 1 to 4 days from onset.” In our case, deep vein thrombosis prophylaxis with subcutaneous heparin (5,000 units every eight hours) was initiated after the first two hemorrhages had stabilized and the patient was clinically improving. Despite the cautious timeframe, three days after initiation of subcutaneous heparin, multiple new hemorrhages developed resulting in fatal herniation.

The recurrence of hemorrhaging in this case was extremely rapid and fatal. In a large Japanese nationwide survey, the average interval between CAA-related lobar ICH was 11.3 months with recurrent hemorrhages in 31.7% of patients over the 35.3 month study period [23]. In a recent study, Biffi et al. reviewed 505 cases of recurrent lobar ICH and found zero recurrences within the first three months and only one recurrence within the first six months [24]. Greenberg et al. found recurrent lobar hemorrhages in 27 of 94 patients (28%) with a mean secondary hemorrhage at 20.8±16.7 months [3]. While a timeline was not specifically indicated for tertiary bleeds and onward, our patient comparatively demonstrated an accelerated time-line of sequential hemorrhages.

Conclusions

The combination of sequential hemorrhages compounded by the use of subcutaneous heparin broadens our understanding of the disease course and risk factors associated with CAA. What has been always thought of as a benign anticoagulant, subcutaneous heparin must now be examined, like other anticoagulants, as having the potential to increase the risk of hemorrhaging in patients with CAA.