Recurrence of Lobar Hemorrhage: A Red Flag for Cerebral Amyloid Angiopathy-related Inflammation?

Abstract

Background. Recurrent lobar intracerebral hemorrhage is more commonly associated with cerebral amyloid angiopathy and less likely associated with hypertension. Cerebral amyloid angiopathy-related inflammation is a subgroup of cerebral amyloid angiopathy that can present with lobar intracerebral hemorrhage, encephalopathy, and seizures; wherein corticosteroids may facilitate favorable outcome. Whether recurrence of lobar intracerebral hemorrhage in cerebral amyloid angiopathy is related to cerebral amyloid angiopathy-related inflammation is unknown.

Case presentation. A 68-year-old woman presented with an acute onset of confusion. She was known to have a history of recurrent lobar intracerebral hemorrhage related to cerebral amyloid angiopathy. Brain imaging revealed previous sequelae of cerebral amyloid angiopathy and a new lobar intracerebral hemorrhage. An empirical diagnosis of cerebral amyloid angiopathy-related inflammation was made given the patent’s clinical course of recurrence. Utilizing current evidence of criteria used to diagnose cerebral amyloid angiopathy-related inflammation, corticosteroid therapy was initiated with significant improvement in clinical and imaging characteristics.

Discussion. Inflammatory pathways incited as a result of cerebrovascular amyloid deposition play a vital role in pathogenesis of cerebral amyloid angiopathy-related inflammation. We highlight the need to consider corticosteroid therapy in patients presenting with recurrent lobar intracerebral hemorrhage in the setting of cerebral amyloid angiopathy since inflammation may play a role in its pathophysiology. Evidence in the literature is sparse to suggest that cerebral amyloid angiopathy-related inflammation might be the root cause for the lobar intracerebral hemorrhage recurrence in cerebral amyloid angiopathy. Further studies are needed to identify mechanisms of recurrent hemorrhage, its correlations with cerebral amyloid angiopathy-related inflammation, and the potential role of corticosteroid therapy.

INTRODUCTION

Cerebral amyloid angiopathy (CAA) is a progressive age-related vasculopathy due to excess deposition of beta-amyloid (Aβ) in leptomeningeal and cortical arterioles. Mainly found in the elderly and demented patients, CAA presents with hemorrhagic stroke, ischemic stroke, cognitive impairment, and transient neurological symptoms.¹ Population-based autopsy studies associate CAA in 55 to 59 percent of demented elderly,² 90 percent of Alzheimer’s patients,^3,4 and 28 to 38 percent of nondemented elderly cohort.² CAA-related intracerebral hemorrhage (ICH) can manifest as lobar macrohemorrhages, cortical microhemorrhages, convexal subarachnoid hemorrhage, or superficial siderosis on imaging studies.¹ The lobar hemorrhages involve predominantly cortical and subcortical regions⁵ with a predilection to temporal and occipital lobes⁶ but can involve multiple sites.⁷

Recurrence in ICH has been reported up to 24 percent⁸ in cases of primary ICH cases and majority of them are lobar hemorrhages related to CAA.^9,10 SMASH-U criteria (structural lesion, medication, amyloid angiopathy, systemic/other disease, hypertension, undetermined), a pathogenetic classification proposed for the etiologic classification of ICH, also highlights the importance of CAA in the causation of ICH.¹¹ At present, there is no established treatment regimen to curb the recurrence of CAA-related lobar ICH.¹² CAA- related inflammation (CAA-RI) is a lesser known entity in which corticosteroid therapy has shown some promise.¹³ CAA-RI and its relation with recurrence of lobar ICH in CAA is largely unexplored. Here, we report a patient with recurrent CAA-related lobar ICH and inflammation who responded to corticosteroid therapy and developed improvement in higher cortical functioning and significant reduction in lesion severity on magnetic resonance imaging (MRI).

CASE PRESENTATION

A 68-year-old woman with a history of hypertension presented to the hospital with acute confusion in June 2013. Her exam was consistent with limb kinetic apraxia of the left arm and alien limb phenomenon of the left upper extremity and left arm and leg clumsiness. Brain imaging revealed a right frontal ICH with associated surrounding vasogenic edema and extension into the local subarachnoid space. There were innumerable microhemorrhages bilaterally, compatible with amyloid angiopathy. Her presentation was attributed to CAA and her home therapy was adjusted and she was discharged. A few months later, in April 2014, she was readmitted with acute onset confusion, speech difficulty and transient left side weakness. History was also suggestive of some transient seizurelike activity as per family history. Examination confirmed transcortical motor aphasia and mild left hemiparesis. Brain imaging revealed a new left frontal lobar ICH and was also evident was encephalomalacia in the right frontal lobe from previous ICH. Multifocal remote infarcts and scattered punctate foci of restricted diffusion in the inferior left cerebellum and the bilateral frontal and parietal lobes, corresponding to amyloid, were also noticed. She was treated conservatively with blood pressure control and an anticonvulsant regimen.

During her admission in May 2014, she was admitted again with acute confusion, headache, and vomiting and was hospitalized for workup of her acute encephalopathy. MRI brain showed a new right temporo-parietal ICH with surrounding vasogenic edema. On initial neurologic exam, the patient exhibited impaired orientation, comprehension, repetition, and recall. A mild left facial droop was also noted. No further abnormalities in mental status, cranial nerve function, strength, or sensation were detected. Biochemical findings were within normal limits except for elevated single-stranded DNA IgG antibody at 83 EU and C-reactive protein at 13mg/L. MRI images were obtained on admission (Figure 1).

FIGURE 1.

Axial fluid attenuated inversion recovery magnetic resonance imaging (MRI) brain scan on admission and after corticosteroid treatment. Temporo-parietal sulcal hyperdensity with surrounding vasogenic edema on admission (A) is noted to significantly improve after corticosteroid therapy (B). One month follow up MRI demonstrated resolution of the lesion (C).

Overall, the patient’s presentation on admission was highly suggestive of encephalopathy related to her CAA. The patient’s family declined further spinal tap and leptomeningeal biopsy. Given her multiple recurrence of lobar hemorrhage, we decided to initiate corticosteroid therapy entertaining the diagnosis of CAA-RI. A short course of corticosteroid pulse therapy was initiated with daily 1g of methylprednisolone sodium succinate administered intravenously for five days. Five consecutive days into steroid initiation, the patient completed repeat brain MRI and Montreal Cognitive Assessment (MoCA) examination. Her MoCA scores demonstrated significant improvement from 7 on Day 1 to 14 on Day 3 (Figure 2). There was accompanying radiological improvement noted on the brain MRI (Figure 1). The patient was discharged on 40mg daily oral prednisone and 75mg daily oral azathioprine. At one-month follow-up, MRI scans showed almost complete resolution of the lesion (Figure 1), and her MoCA score had improved to 19. She continues to show progress to date with the MoCA score reaching 27 at six months follow-up.

FIGURE 2.

Clinical course of steroid therapy. Abnormal imaging improved after steroid therapy and coincided with improvement in higher brain function. As MRI lesions decreased, Montreal Cognitive Assessment scores doubled. Drawings also became increasingly vivid and cohesive.

DISCUSSION

CAA commonly involves small cortical arteries and rarely veins.¹⁴ Aβ deposition results in thickened vasculature, smooth muscle destruction, endothelial dysfunction,¹⁵ cerebral dysautoregulation, ongoing inflammation, and disruption of the blood-brain barrier.¹⁶ The apolipoprotein E (Apo E, particularly the ε4 allele) gene located on chromosome 19 is considered the linkage between Alzheimer’s disease and CAA.¹⁷ CAA-RI is also termed cerebral amyloid angiitis, primary angiitis of CNS associated with CAA, and Aβ-related angiitis. CAA-RI can present with progressive encephalopathy, seizures, headache, and focal neurological deficits.¹⁸ CAA-RI in comparison to CAA has few distinctive features, including younger age at diagnosis, decreased prevalence of cognitive and neurologic deficits, leptomeningeal enhancement on contrast imaging, and better response to corticosteroid therapy.¹⁹ The relatively younger age, recurrence of ICH, presenting symptoms, and radiographic findings prompted the diagnosis of CAA-RI in our case.

CAA-RI pathophysiology presumably stems from the immune response to the cerebral Aβ deposits. The support for this theory comes from immunomodulation studies where vaccine administration in Alzheimer’s disease subjects led to the development of a clinical picture of acute to sub-acute meningoencephalitis, similar to CAA-RI and autoantibodies to Aβ42 amyloid form.²⁰ Also, increased levels of autoantibodies against Aβ40 and Aβ42 have been noted in the cerebrospinal fluid²¹ and against Aβ42 in the serum of patients with CAA-RI.²² Aβ deposits even result in partial activation of CD4 cells in close vicinity of major histocompatibility complex class II antigens that are expressed by macrophages after engulfing Aβ.²³ CAA-RI has two pathological subtypes that may or may not coexist: a perivasculitic form wherein there is visible perivascular multinuclear giant cells infiltration in the brain parenchyma and a vasculitic form (or transmural granulomatous vasculitis) where direct invasion of the vessel wall by inflammatory mediators are seen.²⁴

CAA-RI is a distinct entity that complicates CAA, and no data so far exist to suggest CAA-RI plays a role in CAA related ICH. In CAA-RI, there is Aβ deposition leading to various immunomodulation pathway activations, including the release of inflammatory mediator-triggering of complement pathway^25,26 and matrix metalloproteinase-9 activation.²⁷ These deposits can also incite ion channel formation, toxicity of the cell, disruption of the blood-brain barrier permeability,²⁵ and proliferation and activation of the monocyte/macrophage.²⁸ These inflammatory pathways ultimately cause disruption of the cerebrovasculature,²⁹ which sheds light on the key role of inflammation in CAA-related ICH. Various precipitating factors, including advanced age, hypertension, Alzheimer’s disease, use of antithrombotics, minor trauma, and anti-amyloid therapy such as bapineuzumab, play roles in the development of CAA-related ICH.³⁰Since Apo E gene ε4 allele is strongly associated with CAA-related ICH,¹⁷ CAA-RI,³ early onset CAA related ICH, and early recurrence of ICH,³² we postulate that the ε4 allele might be the epicenter with higher genetic predilection of immune response to Aβ in these individuals, linking all of these manifestations. From the above, we further postulate that CAA-RI may be involved in the CAA-related ICH development especially in cases of recurrence. Further research is required to decipher our theory of genetic linkage between CAA-related lobar ICH and CAA-RI.

Brain imaging and biochemical testing play central roles in the diagnosis of CAA-RI. Biochemical analysis involves estimation of cerebrospinal fluid concentrations of Aβ.³³ Danve et al¹³ noted that inflammatory markers, such as erythrocyte sedimentation rate and C-reactive protein, were elevated in as many as 30 percent of patients with CAA-RI¹³ and were also elevated in our patient. Given the patient’s reluctance for further spinal taps, we did not evaluate the cerebrospinal fluid concentrations of Aβ and autoantibodies against Aβ40 and Aβ42, these studies may be an important step in attaining the diagnosis of CAA-RI. Genetic testing for Apo E gene, especially ε2 and ε4 alleles, may be considered in patients with Alzheimer’s or of older age in order to ascertain their risks for CAA-related ICH and CAA-RI.

Key imaging characteristics on brain MRI that correlate to CAA include lobar cerebral microhemorrhages³⁴ with a predisposition to parietal lobes⁶ and leukoaraiosis³⁵ (deep cerebral white matter imaging changes due to axonal destruction), mild gliosis, and demyelination.³⁶ Other relevant MRI changes are localized convexity subarachnoid hemorrhage, cortical superficial siderosis (hemosiderin accumulation in superficial cerebral cortical layers),³⁷ and silent acute ischemic lesion (especially on diffusion weighed imaging).³⁸ When there is high clinical suspicion of CAA-RI but radiological findings are not definitive, brain biopsy may prove useful.³⁹

CAA-RI treatment as highlighted by multiple studies is primarily focused on corticosteroid therapy.^39,40 High dose (1g/day) of intravenous methylprednisolone for three days might be a reasonable choice at initiation. The remission can be maintained using a maintenance regimen of low-dose oral corticosteroids (approximately 1mg/kg/day), which can be gradually tapered off over 3 to 4 months (Table 1).^18,39 Studies have shown approximately 1 to 3 weeks are required from treatment initiation to observe the response to treatment.³¹

TABLE 1.

Management of cerebral amyloid angiopathy related inflammation (CAA-RI) using corticosteroids

INITIATION DOSAGE Pulse steroid therapy (1g/day) using intravenous methylprednisolone sodium succinate for 3 days

MAINTENANCE DOSAGE Low dose (1mg/kg/day) oral corticosteroids, which can be gradually tapered over 3–4 months after initiation

The exact mechanism of action of corticosteroids in CAA-RI is not known. An in-vitro study by Previti et al⁴¹ has demonstrated that corticosteroids can decrease the pro-inflammatory effects of Aβ deposition by increasing interleukin-6 and matrixmetalloproteinase-2 levels in the late inflammatory phase within vascular smooth muscle cells. Another plausible theory is the potential decrease in the surrounding area of vasogenic edema induced by the corticosteroids.⁴⁰ The mitigation of immune-related inflammation might be another mechanism for corticosteroid response as demonstrated by the normalization of the increased levels of autoantibodies against Aβ40 and Aβ42 at three months after initiation of corticosteroids.^21,22 The beneficial effects of immunosuppressants can also be explained by the tendency of CAA-RI to relapse in drug-free periods.¹⁸ Few studies have indicated that perivasculitic forms have the most optimistic prognosis⁴² and better response to corticosteroid treatment. This might be due to presence of more vasogenic edema in association with perivasculitic form as compared to the vasculitic form, which has more cytotoxic involvement.⁴⁰

Other immunosuppressive agents employed in the treatment of CAARI include cyclophosphamide, azathioprine, methotrexate, and mycophenolate mofetil, but further studies are needed to determine their true level of benefit.¹³ Cyclophosphamide may be an alternative of corticosteroid therapy when administered as 100mg/day for the initiation dose for nine months and 50mg/day as the maintenance dose.⁴³ Methotrexate 20mg/week, in addition to corticosteroids, can also benefit the patient.⁴⁴ Azathioprine 75mg/day is another effective alternative for the maintenance corticosteroid therapy.⁴⁵ We also noticed absence of recurrence on addition of azathioprine to corticosteroids as a part of the maintenance regimen in our patient.

CAA-related lobar ICH has a high tendency for recurrence.⁷ This recurrence can be attributed to the inflammatory variant of CAA. Hoshi et al⁴⁶ reported a case with recurrent CAA-related ICH that was treated with corticosteroids; they noted an increase in serum anti-nuclear antibody titers and erythrocyte sedimentation rate, which suggests an inflammatory process. From our clinical observation, we also believe that there might be an increase in recurrence associated with CAA-RI as compared to CAA alone; our patient presented with multiple recurrences of CAA-related ICH that terminated after corticosteroid initiation. CAA-RI treated with immunosuppressants is associated with a good prognosis; approximately 73 percent of patients treated with immunosuppressants have shown good outcomes, with Modified Rankin Scale scores of 3 or lower.¹⁹ Studies have observed complete remission in a majority of patients that have lasted up to five years.¹³

CONCLUSION

CAA-RI may be an underlying etiology for recurrent CAA-related lobar ICH. Lobar ICH is an integral component of Boston criteria^47,48 for CAA diagnosis, and inflammation might be underlying pathogenesis for its recurrence. Thus, we propose the addition of recurring lobar ICH to the criteria suggested by Chung et al¹⁸for probable CAA-RI (Table 2).

TABLE 2.

Diagnostic criteria for cerebral amyloid angiopathy related inflammation (CAA-RI)^*

CHARACTERISTICS	PROBABLE CAA-RI	DEFINITE CAA-RI
Clinical	Younger age group (≥40 years) Neurological symptoms with acute or sub-acute onset: altered mental status or behavior, focal neurological deficits, seizures, headache Exclusion of neoplastic, infectious, or other causes	Diagnostic criteria is same as that for probable CAA-RI in addition to the following: Evidence of perivascular, transmural, and/or intramural inflammation on biopsy Accumulation of amyloid in the vessel walls of the affected cortical area and leptomeninges
Radiological	Multiple cortical and subcortical hemorrhages or microhemorrhages and/or recurrent lobar hemorrhages evident on susceptibility weighted MRI sequences Presence of usually asymmetric patchy or confluent hyperintensity with or without mass effect on T2 or fluid attenuation inversion recovery. Leptomeningeal or parenchymal enhancement may or may not be seen
Pathological	None

^*This table is adapted from Chung et al.¹⁸ In addition to the criteria from Chung et al, we have included an additional criterion we believe supports the diagnosis of CAA-RI: “Recurrence of lobar hemorrhage.”

Corticosteroid therapy is the treatment of choice for CAA-RI; it might also be effective to prevent CAA-related lobar ICH recurrence and should be started as soon as the diagnosis of probable CAA-RI is made. In-depth studies are warranted in order to establish the genetic and pathophysiologic correlation between CAA-RI and CAA-related lobar ICH. Randomized controlled trials are necessary to evaluate the efficacy of immunosuppressive therapies, including corticosteroids, in CAA-RI as a form of prevention in recurring ICH among CAA patients.

AUTHOR CONTRIBUTIONS

VR and LLD drafted the manuscript. LLD administered the MOCA examinations. GA, SB, and AYK revised the paper critically for intellectual content. VSH conceived the study and participated in its design and coordination. All authors read and approved the final manuscript.

ACKNOWLEDGMENTS

We would like to extend our thanks to the patient and her family for allowing us to carry out this work and to all who participated in her interdisciplinary care.