Management of Intracranial Aneurysms Caused by Infection

Published on 13/03/2015 by admin

Last modified 13/03/2015

Print this page

This article have been viewed 2257 times

Chapter 150 Management of Intracranial Aneurysms Caused by Infection

Infectious intracranial aneurysms are a heterogeneous group of rare lesions that develop from inflammation-induced weakening of arterial walls. Their first description was provided by Church, who in 1869 described a 13-year-old male with endocarditis who died from rupture of a middle cerebral artery (MCA) aneurysm.¹ Osler then coined the term mycotic aneurysm to describe an aortic aneurysm that arose in the setting of endocarditis. This term has since been employed as a misnomer to describe all aneurysms of infectious origin.² By 1923, 34 cases of intracranial bacterial aneurysms were reported.³ Combined with 16 cases reported in 1954,⁴ only 4 patients survived. Fortunately, as time has progressed, the prevalence of infectious intracranial aneurysms has been on an overall decline as rheumatic heart disease and endocarditis wane (Table 150-1).⁵^–¹¹ Across series reported in recent years, 0.7% to 6% (mean 2.3%, 182 of 7999 patients across all series) of aneurysms are infectious.⁷^–¹¹ However, the true incidence of infectious intracranial aneurysms remains unknown given their ability to grow and regress while remaining clinically silent. We comprehensively review their natural history, pathogenesis, diagnosis, treatment modalities, and management algorithms.

Table 150-1 Prevalence of Mycotic Aneurysms as Reported in Several Large Series

Natural History and Presentation

We reviewed 22 original series with 245 patients harboring 316 infectious aneurysms (see the Appendix, summarized in Table 150-A).^7-,11–29 The overall male-to-female ratio was 1.2:1; 59% of aneurysms were in the MCA territory, 16% were in the posterior cerebral artery (PCA) territory, 9% were in the anterior cerebral artery (ACA) territory, and 7% were along the cavernous internal cerebral artery (ICA) (Table 150-2). Among aneurysms in the MCA territory, 55% were specified as being distal MCA while only 8% were specified as being proximal MCA (the remainder were unspecified). This is reinforced by an earlier review of 81 patients with infectious aneurysms—69% were in the distal MCA territory, while only 9% were in the proximal cerebral vasculature.³⁰

Table 150-B Review of Etiologies and Offending Organisms Where Specified Among Modern Series of Infectious Aneurysms

Table 150-2 Summary Table for Data Across Our Reviewed Series

Presenting symptoms include seizures or those referable to either ischemic or hemorrhagic strokes, such as hemiparesis, dysphasia, visual field deficits, and change in level of consciousness. Patients with infectious intracranial aneurysms of the cavernous carotid segment (most frequently secondary to cavernous sinus thrombophlebitis) often present with persistent ophthalmoplegia and often have accompanying meningitis.^14,¹⁷ Given that infectious aneurysms may form from septic emboli, it is not surprising that patients with these lesions can present with ischemic symptoms that may precede aneurysm formation. However, Salgado et al. reviewed 68 patients with infective endocarditis and infectious aneurysms and found that only 43% of patients had neurologic prodromes prior to their diagnosis (median 17 days prior).³¹ In comparing this group of patients to those with endocarditis without infectious aneurysms, there were no particular neurologic symptoms distinguishing this particular group. Across three of the larger series, 40 of 59 patients (68%) presented with hemorrhage while 5 of 59 patients (8%) presented with only ischemic infarction; 7 of 59 patients (12%) presented with seizure.^7,^21,²² Excluding patients with cavernous aneurysms, across 17 series with 202 patients, as shown in Table 150-2, 153 patients (76%) had ruptures of at least one infectious aneurysm. The bleeding pattern may involve any combination of subarachnoid, intraparenchymal, and intraventricular hemorrhage. Epidural hemorrhage from middle meningeal mycotic aneurysm rupture has also been reported,²¹ as has subdural hemorrhage.^27,^32,³³

Table 150-3 summarizes outcomes across our 22 reviewed series. Overall mortality was 20% (50 of 245 patients). This is an underestimate, because patients with catastrophic/lethal hemorrhage may not even come to neurosurgical attention, may not obtain an actual diagnosis of infectious aneurysm because they do not undergo angiograms, or are explicitly excluded from certain series.²⁶ Long-term neurologic morbidity, where reported, was seen in 36% of patients, and 52% of patients (67 of 129 patients) were neurologically intact at follow-up. These data are promising, because more dated, oft-cited reviews quote significantly more pessimistic outcome figures. A detailed literature review of 85 cases in 1978 cited an overall mortality rate of 45% and a mortality rate of 42% from initial rupture (8 of 19 patients).¹⁵ It also found that 80% of patients with aneurysmal subarachnoid hemorrhage during hospitalization for endocarditis died (a frequently cited figure based on 15 patients), while 3 of 10 patients (30%) with unruptured mycotic aneurysms treated for endocarditis died. In another literature review of 45 patients from 1977, excluding patients who were “inevitably” going to die, 29% of patients died of aneurysmal rupture (20% overall).³⁴

Table 150-3 Overall Outcome from Series of Infectious Intracranial Aneurysms/Total in Study (%)

Pathogenesis

Infectious intracranial aneurysms can develop in the setting of (1) septic emboli (e.g., endocarditis) or (2) contiguous spread (e.g., meningitis or cavernous sinus thrombophlebitis).^14,^35,³⁶ Alternative sources for contiguous spread described in the literature include cerebral abscesses,^9,^37,³⁸ osteomyelitis,³⁹ sinusitis,⁴⁰ and otitis media.^21,⁴¹ Across 21 series with 225 patients specifying infectious aneurysm etiology (Table 150-B), 180 cases (80%) were due to endocarditis, 18 cases (8%) were from meningitis, and 10 cases (4%) from cavernous sinus thrombophlebitis. The pattern of aneurysm formation is defined by the mechanism by which it arises—those from septic emboli are more likely to occur in multiples and tend to occur distally at branch points in the MCA territory, while those from contiguous spread tend to develop along the proximal circulation.¹⁹ A recent case series suggested a relatively better prognosis for patients with infectious aneurysms due to endocarditis compared to for those with infectious aneurysms from “other” causes—10% of patients with endocarditis and infectious aneurysms died, compared to 47% with infectious aneurysms from other etiologies.⁹

Table 150-4 Prevalence of Infectious Aneurysms in Series of Patients with Endocarditis

Endocarditis

Despite medical advances, the persistence of endocarditis in populations is likely attributable to continued intravenous drug abuse and rheumatic heart disease in underdeveloped countries. Neurologic sequelae are seen in 20% to 40% (average 30%) of cases ⁴² (Table 150-4). Several series have noted that neurologic complications are more likely to occur and occur more quickly¹⁸ with Staphylococcus aureus infection than with streptococcal. In Kanter and Hart’s series, 67% of patients with S. aureus endocarditis had neurologic complications compared to 22% of patients with streptococcal infections.⁴³ Tunkel and Kaye reviewed 1014 patients with endocarditis and found central nervous system involvement in 53% to 71% of cases of S. aureus endocarditis.⁴²

The most common neurologic complication of endocarditis is embolic infarction, afflicting as many as 31% of patients with this disease.⁴² As reinforced by several case series,^43,⁴⁴ Hart et al.’s literature review of 2119 patients with endocarditis found an overall 5% rate of intracerebral hemorrhage.⁴⁵ Interestingly, in an autopsy study of 16 patients with intracranial hemorrhage and endocarditis, 9 cases (56%) were hemorrhagic transformations of infarcts, 4 cases (25%) were arterial ruptures from pyogenic arteritis, and only 3 cases (19%) were from infectious aneurysm rupture.⁴⁶ In Hart et al.’s series of 209 patients with endocarditis, 17 patients (8%) had intracerebral hemorrhages, 4 patients (24%) attributed endocarditis to hemorrhagic infarcts, 4 patients (24%) attributed it to pyogenic arteritis, and only 2 patients (12%) attributed it to infectious aneurysm rupture.⁴⁵ In their larger review of 2119 patients with endocarditis, only 1.7% had infectious aneurysms.⁴⁵ Table 150-4 summarizes the prevalence of infectious aneurysms in several large series of patients with endocarditis, underscoring that infectious aneurysms are an overall rare complication of this disease.^8,^12,^43,^44,⁴⁷

Molinari et al. developed an elegant model that continues to be cited as the likely mechanism of infectious aneurysm formation in the setting of septic emboli.⁴⁸ They demonstrated that septic particles embolize into the vasa vasorum, leading to inflammation of the adventitia first, followed by centripetal progression into the muscularis and then the internal elastic lamina. Early-developing aneurysms were first noted at the proximal or distal end of an occluded segment as quickly as 24 hours after embolization. Enlargement then occurred via pulsations against the necrotic wall in some cases. This model also demonstrated the efficacy of antibiotic treatment, because partially treated aneurysms, as opposed to untreated aneurysms, had indurated, intact fibrotic walls.

Contiguous Spread: Meningitis and Cavernous Sinus Thrombophlebitis

Patients with bacterial meningitis likely develop infectious aneurysms via invasion of the arterial wall from the adventitia spreading inward. An infective arteritis weakens the wall, leading to aneurysm formation. Given the extravascular etiology in this case, it is not surprising that most aneurysms in association with meningitis develop along the proximal intracranial vasculature.

Cavernous carotid infectious aneurysms are special cases that most often develop from cavernous sinus thrombophlebitis. These cases have the overall best prognosis of infectious intracranial aneurysms. A recent case series of 8 patients with 9 cavernous carotid infectious aneurysms reported that all were “cured” at follow-up.⁹ In a comprehensive literature review of 36 cases of cavernous carotid infectious aneurysms (7 bilateral, primarily from case reports), 24 patients (67%) were improved at follow-up.⁴⁹ Although 6 patients (17%) died, 2 patients had fungal aneurysms and 1 patient never received antibiotic treatment. The pathogen was unknown in 17 cases (47%) but was staphylococcal in 12 (33%), streptococcal in 3 (8%), and fungal in 3 (8%).

Microbiology

Although most commonly due to bacteria, infectious intracranial aneurysms have been attributed to fungal (true mycotic),⁵⁰ spirochetal,⁵¹ amebic,⁵² and viral pathogens, including the human immunodeficiency virus (HIV).^53,⁵⁴ A recent review of the literature of HIV-associated mycotic aneurysms found 17 cases in children and 6 cases in adults.⁵⁴ The aneurysms had a tendency to occur in younger patients and in the anterior circulation, and some resolved following antiretroviral therapy.^55,⁵⁶ Across reviewed series specifying microbiology in 195 cases (see the Appendix, Table 150-A), 32% were due to streptococcus (8% specified as Viridans streptococcus), 24% were from staphylococcus (19% specified as S. aureus), 4% were fungal, and 3% were due to enterococci. In 30% of these cases, cultures were negative. This is similar to the results of a review of 81 cases in the literature that found 44% due to streptococcal species, 19% due to S. aureus, no growth from cultures in 12% of cases.³⁰

Fungal cases deserve particular attention given their exceedingly poor prognosis. Among the first 16 cases reported in the literature, none survived.^19,⁵⁷ A recent review of the literature consisting primarily of case reports identified 35 cases of fungal (“true mycotic”) aneurysms.⁵⁸ Although 2 of 35 cases (6%) had favorable outcomes, the vast majority died; 18 cases were seen in immunocompetent patients, most commonly after surgery (11 cases) but also following sinusitis in 3 cases and trauma in 2 cases. The most common organism seen, potentially from its angioinvasive nature, given its ability to produce elastase, is Aspergillus. Aneurysms from aspergilli typically arise from local intracranial disease. The second tier of common fungal organisms includes Candida (typically from endocarditis or systemic infection) and Phycomycetes (associated with diabetes).

Diagnosis

Patients with endocarditis and new neurologic symptoms should undergo computed tomography (CT) angiography followed by digital subtraction angiography (Figs. 150-1 and 150-2). Van der Meulen et al., however, demonstrated that surveillance angiography in neurologically asymptomatic patients with endocarditis should not be performed because it does not confer a survival benefit overall.⁵⁹ Patients with meningitis, cavernous sinus thrombophlebitis, or other potential contiguous infectious processes with persistent or new neurologic symptoms should also undergo CT angiography followed by digital subtraction angiography. The diagnosis of an infectious aneurysm is then rendered in the event of a positive angiogram. Proximal saccular aneurysms discovered in the setting of endocarditis should be viewed with scrutiny as potential true hemodynamic berry aneurysms, though the treatment algorithm (Fig. 150-3) does not need modification—ruptured aneurysms should be treated, and unruptured, stable aneurysms should be observed. Kannoth et al. reported a sensitivity of 96% and specificity of 100% for a diagnosis of infectious intracranial aneurysms in patients with a cerebral aneurysm and a diagnosis of infective endocarditis, meningitis, orbital cellulitis, or cavernous sinus thrombophlebitis.⁶⁰ Further specific, supportive features from their study include distal location of the aneurysm, multiplicity, fusiform shape, change in size or appearance, patient age less than 45 years, recent fever, recent lumbar puncture, and intraparenchymal hemorrhage on imaging. However, several studies have demonstrated that lumbar puncture is not particularly correlated with the presence of infectious aneurysms in patients with endocarditis.^26,^31,⁴⁴