Mycotic aneurysm presenting in a patient with a ventriculoatrial shunt with positive CSF cultures for Cutibacterium acnes: illustrative case

Abstract

BACKGROUND

Mycotic aneurysms (MAs) typically present with intracranial hemorrhage but are a rare cause of subdural hematoma (SDH). Although infective endocarditis is the most common etiology, MAs may also result from bacteremia associated with intravenous drug use, meningitis, or poor dentition. Patients with ventriculoatrial shunts (VASs) are at increased risk for shunt-related bacteremia compared to those with ventriculoperitoneal shunts, potentially predisposing them to MA formation.

OBSERVATIONS

This is the case of a 61-year-old female with a fusiform distal left middle cerebral artery MA presenting as an acute-on-chronic SDH without endocarditis, but with positive blood cultures and CSF cultures for Cutibacterium acnes from a long-term VAS.

LESSONS

MAs can rarely present as SDHs and should be considered in the differential diagnosis for patients with a long-standing VAS due to the risk of subacute bloodstream infections.

https//thejns.org/doi/10.3171/CASE25625

Mycotic aneurysms (MAs) account for < 5% of all intracerebral aneurysms and usually present as subarachnoid, intraparenchymal, or combination hemorrhages, with only 4 reports of cases presenting as pure acute subdural hematomas.¹ Poor dental hygiene, intravenous drug use, or direct extension from meningitis or cranial abscesses can precipitate MA formation. However, infective endocarditis is present in 65% of cases.² The most recent literature has shown that ventriculoatrial shunts (VASs) have a similar risk of infection to ventriculoperitoneal shunts (VPSs). However, VASs carry an increased risk of bacteremia with septicemia, endocarditis, nephritis, and thromboembolic or cardiac complications.³ There have been no reports about VAS infection leading to MA formation.

Illustrative Case

Presentation

A 61-year-old female with a significant past medical history of congenital hydrocephalus treated with a VPS at age 3 years and a pituitary adenoma resected at age 31 years presented to the emergency department after an unwitnessed fall 5 days earlier. The patient complained of nausea and vomiting but was intact on a neurological examination. She denied having the worst headache of her life or symptoms of infection such as fever or chills, but reported amnesia from the fall. She was found to have left greater than right acute-on-chronic subdural hematomas (SDHs). The patient was admitted for observation and had stable repeat brain imaging. She was discharged home with plans for return in 1 week for outpatient left middle meningeal artery (MMA) embolization (Fig. 1). During a technically successful left MMA embolization with N-butyl cyanoacrylate, the patient was found to have a distal left middle cerebral artery (MCA) M4 segment fusiform aneurysm that was adjacent to the known acute-on-chronic subdural on both conventional and CT angiographic imaging (Figs. 2 and 3). Given the distal location and fusiform appearance of the discovered aneurysm, an infectious etiology was suspected. On further investigation, it was discovered that the patient underwent multiple shunt revisions due to infection, with the last one at age 18 years. The patient’s VPS was distally revised to the right atrium at age 17. The history regarding her multiple shunt revisions and infections is not available in our records, and the patient and her family do not recall further details about specific pathogens and treatments. The patient was admitted for further infectious workup including blood cultures, urinalysis, a respiratory viral panel, and a complete blood count. One of two blood cultures drawn at admission for MA workup grew Cutibacterium acnes after 5 days, initially thought to represent probable contamination. The remainder of the infectious workup was negative. The patient was febrile to 102.6°F 2 days after MMA embolization; repeat blood cultures at this time grew methicillin-sensitive Staphylococcus aureus (MSSA), which was thought to represent iatrogenic bacteremia secondary to ultrasound-confirmed left upper extremity thrombophlebitis from the peripheral intravenous line, for which she was started on vancomycin and ceftriaxone.

FIG. 1.

Axial (left) and coronal (right) noncontrast CT images showing a left holohemispheric mixed-density SDH.

FIG. 2.

Anteroposterior (A), lateral (B), and 3D-reconstructed (C) angiograms demonstrating a 3.4 × 3.6–mm aneurysm arising from the middle M4 segment of the angular artery, originating from the inferior division of the left MCA. The aneurysm localizes to the subdural space at the parieto-occipital border. Delayed contrast opacification and internal stasis are consistent with a pseudoaneurysm or MA.

FIG. 3.

Axial (left) and sagittal (right) CT angiograms revealing an aneurysm (arrows) on the left cortical surface surrounded by a mixed-density SDH.

On hospital day 4, the patient developed acute onset of aphasia and mild right-sided weakness. Noncontrast head CT and CT angiography studies were negative for acute infarct, new vascular abnormality, or subdural enlargement. The symptoms resolved spontaneously within 24 hours and were attributed to seizures, for which the patient was medically managed with Keppra. Long-term monitoring EEG showed no signs of seizures, but did show left frontal and bilateral temporal sharp waves. A left parieto-occipital craniotomy for aneurysm repair was planned for treating the distal fusiform aneurysm. Medical management alone was not pursued due to the high clinical suspicion of aneurysm rupture, which conferred a significant risk of rerupture.

Operative Report

The patient underwent a left parieto-occipital craniotomy on admission day 10. Neuronavigation was used to confirm the location of the aneurysm throughout the case. After dural opening, a thick organized SDH was encountered, which was adherent to the dura and a pulsatile aneurysm on the cortical surface. The aneurysm dome appeared partially adherent to the dura and surrounding subdural clot (Fig. 4A). After subdural clot evacuation, the aneurysm was visualized along with the proximal and distal MCA vessel. The inflow MCA vessel was visualized as originating deeper from the underlying sulcus, while the outflow MCA vessel ran across the cortical surface. During dissection of the aneurysm neck and inflow and outflow vessels from the surrounding blood clot, the aneurysm ruptured. A 4-mm curved Yaşargil mini clip was placed across the aneurysm, resulting in immediate cessation of the bleeding (Fig. 4B). Because of the fusiform morphology of the aneurysm, a portion of it remained outside of the clip blades, which was confirmed by injecting indocyanine green (ICG) dye. To address this, two additional 3-mm straight Yaşargil mini clips were placed proximal and distal to the aneurysm on the inflow and outflow MCA branches to trap the aneurysm (Fig. 4C). A repeat ICG dye injection study confirmed no residual flow in the aneurysm, continued flow in the proximal inflow MCA branch, and retrograde filling of the distal outflow MCA branch (Fig. 4D). Intraoperative cultures were not obtained as the yield was low in the setting of ongoing broad-spectrum antibiotic coverage for 8 days preoperatively. A subgaleal drain was left in place at the conclusion of the surgery. The patient tolerated the procedure well with no complications and had no new neurological deficits.

FIG. 4.

A: Subdural membrane adherent to a pulsatile aneurysm (x) with an underlying mixed-density subdural collection. B: Interval evacuation of subdural collection with subdural membrane adherent to the aneurysm dome, and a Yaşargil mini clip placed across the aneurysm dome following intraoperative rupture. C: Inflow MCA branch emerging from the sulcus (x). D: A 3-mm Yaşargil mini clip placed on the inflow MCA branch. E: A 3-mm Yaşargil mini clip placed on the outflow MCA branch. F: ICG dye injection showing no residual flow in the aneurysm, continued flow in the proximal inflow MCA branch (white arrow), and retrograde filling of the distal outflow MCA branch (black arrow).

Postoperative Course

Postoperative CT showed the expected postoperative changes; the subgaleal drain was removed 3 days postoperatively. Transthoracic and subsequent transesophageal echocardiogram studies did not reveal infection-related growths characteristic of endocarditis. Given the unclear source of infection, the patient’s shunt reservoir was tapped for a CSF sample on postoperative day (POD) 2, which showed many gram-positive rods (GPRs) and eventually grew C. acnes, suggesting an indolent shunt infection as a possible source of bacteremia and MA development. Shunt removal was discussed, but due to the corkscrew and flanged configuration of the proximal shunt catheter, the occipital approach, and the chronicity of the shunt (being in place for more than 40 years), with extensive scarring without evidence of shunt malfunction, the surgical risks of complete removal were considered high, and the decision was made to trial medical treatment with long-term antibiotic therapy and delayed resampling of CSF (Fig. 5). The patient was discharged on POD 6 on ceftriaxone therapy without shunt removal. She was seen 20 days after surgery as an outpatient and had no neurological deficits but reported intermittent headaches. She had an additional 4 weeks of ceftriaxone therapy followed by a 2-week hold. Repeat CSF sampling revealed GPRs on Gram stain but no growth with cultures held for 2 weeks. The patient was then started on 2 months of doxycycline suppression therapy. At the 5-month follow-up, the patient underwent a repeat shunt tap after a 2-week hold of immunosuppressive doxycycline. CSF cultures showed no growth after 2 weeks. The patient has been doing well with no new neurological symptoms throughout her follow-up visits and continues suppressive therapy. Given the demonstrated CSF clearance, the infectious disease team recommends no additional shunt taps unless new clinical findings emerge. The neurology team is managing her persistent nonpositional headaches as migraines, and she is scheduled to receive Botox injections.

FIG. 5.

Axial (left) and coronal (right) noncontrast CT images showing a right occipital approach flanged proximal shunt catheter located near the midline in the collapsed right lateral ventricle.

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

This is a case of a 61-year-old woman with a history of childhood hydrocephalus and multiple shunt infections and revisions, culminating with transition to a VAS at age 17, who presented after a fall with the finding of a chronic-appearing SDH. During planned MMA embolization for the chronic SDH, a fusiform distal M4 aneurysm was seen, which was located at the cortical surface within the SDH on imaging, subsequently raising concern for a ruptured MA. During microsurgical aneurysm clipping, the aneurysm was confirmed to be located on the cortical surface and surrounded by a thick adherent clot, highly suggestive of a prior rupture. Infectious workup revealed a C. acnes shunt infection and bacteremia suggestive of a chronic VAS infection as a possible etiology for MA.

This case highlights the diagnostic complexity of MAs, especially when presenting atypically. Our patient presented with headaches and a SDH after a fall, with no infectious symptoms, which is not the typical clinical and radiographic presentation for an MA ictus. In the literature, the most common presenting symptoms for MAs are headaches and fevers, but some can remain clinically silent until death.² Although the patient denied suddenly having the worst headache of her life, her amnesia from the fall raised suspicion for a hemorrhagic event that preceded the traumatic episode. In addition, angiographic imaging showed a distal left M4 aneurysm on the cortical surface, with an aneurysm dome located within the organized SDH, suggesting aneurysm rupture as the source of her SDH. Further supporting the unusual nature of our case, a literature review showed that 72% of MAs present with rupture, and 28% of these cases present as an intraparenchymal hemorrhage (ICH).² There are 14 case reports of MAs presenting as acute SDHs.⁴ Of those, only 4 had an SDH with no associated subarachnoid hemorrhage (SAH) or ICH, as in our patient.¹ One case report described an MA presenting as an empyema.⁵

The most common etiology of MAs (65%) is bacterial endocarditis from degenerative heart valve disease, prosthetic valve replacement, and intravenous drug use. However, other etiologies, such as bacterial meningitis, periorbital cellulitis, poor dental hygiene, and intracranial venous sinus thrombosis, are possible.¹ Our patient did not have any insidious infectious symptoms to suggest the presence of an MA or a known history of endocarditis. However, the distal location of her aneurysm pointed toward an infectious etiology.¹ Furthermore, the patient did have a long-standing VAS for congenital hydrocephalus. VASs were the first-line treatment for shunt diversion until the 1970s, as they were deemed to mimic more physiological CSF absorption into the bloodstream. Since then, VASs have been used less, as the risks of abdominal complications from VPSs have decreased with the introduction of new shunt materials and surgical techniques.³ Nevertheless, it is important to remember that VASs require access to the intravascular space, which can lead to bloodstream infections and precipitate bacteremia, endocarditis, septic emboli, and nephritis.⁵ There have been 6 reported cases of VASs complicated by infective endocarditis.^6–12 Our patient had negative transthoracic and transesophageal echocardiogram studies, with the latter having a sensitivity of 100% for endocarditis on native valves.¹³

Turning to the microbiological findings in this case, further complexity arises in evaluating the significance of the isolated pathogens. The patient had iatrogenic MSSA bacteremia from thrombophlebitis, which followed the diagnosis of her aneurysm. The only infectious result that was a possible source of the MA was her positive shunt CSF culture growing C. acnes, with one of her blood cultures also positive for C. acnes. However, this organism did not grow on any other cultures, initially suggesting contamination. To date, there have been no reported cases of VAS infection leading to MA. Overall, 79% of cases with C. acnes endocarditis occur in the setting of prosthetic cardiac devices.¹⁴ Even though our patient had a negative echocardiogram and no cardiac device, the distal shunt catheter tip in the left ventricle could have acted as a surface area for C. acnes biofilm to seed into the bloodstream. In addition, C. acnes has been shown to cause insidious infections that are difficult to diagnose due to its long incubation period and inability to grow in most culture media, both of which may have additionally contributed to our patient’s single positive blood culture.^15,16 Case reports of infective endocarditis with VASs recommend complete removal of the shunt for source control and prevention of cardiac complications such as future emboli or outflow obstruction.^6,9,10,16 Even though shunt removal, temporary CSF diversion, and antibiotic therapy are the standard of care, some studies have shown 85%–95% cure rates of shunt infection with a combination of intrathecal and parenteral antibiotic therapy without shunt removal.^17,18

Given these considerations, our patient’s management required careful weighing of infectious risk against procedural morbidity. We opted for preservation of the shunt and only intravenous antibiotic therapy with repeat delayed CSF cultures, concluding that the risk of hemorrhage from shunt removal was higher due to the occipital location, collapsed ventricle, and flanged proximal catheter. After proximal catheter removal, intraventricular hemorrhage (IVH) rates of approximately 25%–30% have been reported, with only 2% of those being greater than 5 mL, and Calayag et al. found an IVH rate of 36% for occipital catheter revision compared to 0% for frontal shunt revision, contributing to our decision to avoid shunt revision.^19,20 This difference in IVH rate for occipital revisions is likely due to the proximity to the course of the choroid plexus. Of the 2 available case series exploring IVH after shunt revision, one found that IVH led to shorter time to re-revision, while the other found the opposite. This is likely due to differences in IVH volume between the 2 case series. Based on these results, as well as the development of newer techniques such as endoscopy to visualize choroid plexus adhesions and the use of diathermy for adhesion removal, clinicians now either revise long-standing shunts or place completely new catheters while abandoning old ones. The latter is not feasible in cases o infection such as ours.

Finally, the possibility remains that this aneurysm may not have been mycotic in origin at all. It is hard to establish whether the MA was a result of the patient’s current subacute shunt infection or her history of previous shunt infections with possible meningitis (direct extension of intracranial infection), remaining silent until rupture. The aneurysm can also be nonmycotic in origin. Studies have shown that 0.5%–7.9% of non-MAs present as SAHs with acute SDHs. Aladawi et al. found that in 31 of 108 patients, nonmycotic MCA aneurysms presented as isolated SDHs.²¹

Lessons

This case highlights the diagnostic challenges of MAs, particularly in patients with atypical hemorrhage patterns such as an isolated SDH and long-standing VAS. While C. acnes shunt infection was the suspected etiology, the absence of definitive microbiological or intraoperative confirmation limits the certainty of causation, emphasizing the need for high clinical suspicion and individualized risk-benefit assessment in managing infected shunts with associated vascular complications.

Acknowledgments

We used ChatGPT for reference formatting. We carefully reviewed all the references and made the pertinent modifications for this submission.

Disclosures

Dr. Torabi reported being a consultant for Boston Scientific outside the submitted work. Dr. Wolman reported consulting fees from Johnson & Johnson.

Author Contributions

Conception and design: Amaral-Nieves, Moldovan, Santos-Fontanez, Shao, Torabi, Shaaya, Wolman. Acquisition of data: Amaral-Nieves, Moldovan, Shao, Shaaya, Wolman. Analysis and interpretation of data: Amaral-Nieves, Shao, Klinge, Wolman. Drafting the article: Amaral-Nieves, Acevedo-Jetter, Santos-Fontanez, Chuck, Torabi. Critically revising the article: Amaral-Nieves, Moldovan, Xu, Acevedo-Jetter, Santos-Fontanez, Doberstein, Chuck, Klinge, Torabi, Shaaya, Wolman. Reviewed submitted version of manuscript: Amaral-Nieves, Moldovan, Santos-Fontanez, Doberstein, Shao, Chuck, Klinge, Torabi, Shaaya, Wolman. Approved the final version of the manuscript on behalf of all authors: Amaral-Nieves. Administrative/technical/material support: Amaral-Nieves, Shao, Torabi, Wolman. Study supervision: Amaral-Nieves, Torabi, Wolman.

Correspondence

Natalie Amaral-Nieves: Brown University, Providence, RI. natalie_amaral-nieves@brown.edu.