OUP user menu

Mycotic aneurysms: a case report, clinical review and novel imaging strategy

M. Fisk, L.F. Peck, K. Miyagi, M.J. Steward, S.F. Lee, M.B. Macrae, S. Morris-Jones, A.I. Zumla, D.J.B. Marks
DOI: http://dx.doi.org/10.1093/qjmed/hcq240 181-188 First published online: 8 January 2011

Case report

A 62-year-old man presented to a district general hospital with a 4-week history of fever, drenching sweats, lethargy and intermittent thoracic back pain. Two weeks prior to onset, a pacemaker had been inserted for recurrent syncope secondary to carotid sinus hypersensitivity. His other past medical history included a permanent tracheostomy following resection of a laryngeal carcinoma 8 years previously, and ischaemic heart disease. The latter culminated in an acute coronary syndrome 7 months before the current presentation, for which primary angioplasty was undertaken with insertion of six drug-eluting coronary artery stents requiring dual antiplatelet therapy for at least 1 year.

On admission, he was febrile at 38.5°C. There were no peripheral stigmata of infective endocarditis, auscultation of the pre-cordium identified no murmurs and the pacemaker site was not inflamed. Systemic examination was otherwise unremarkable, as were chest radiography and urinalysis. Transthoracic echocardiography demonstrated good systolic function, with no valvular regurgitation or pacemaker lead vegetations. In the absence of a clear septic focus, he was commenced on piperacillin–tazobactam. Blood tests showed haemoglobin of 11.6 g/dl, white cell count of 5.7 × 109 cells/l and platelet count of 52 × 109 cells/l. C-reactive protein (CRP) was elevated at 272 mg/l. Coagulation screen was normal. Multiple blood cultures were drawn, which grew methicillin-sensitive Staphylococcus aureus. Intravenous flucloxacillin and rifampicin were commenced.

He was consequently transferred to the London Heart Hospital for further management. The pacemaker was explanted and there was marked clinical improvement. Nonetheless, despite antibiotic therapy he continued to have high-grade pyrexias, without diurnal variation, and elevated serum CRP concentrations. An extensive septic screen, including urine and blood cultures, chest radiographs and repeated transthoracic and transoesophageal echocardiograms revealed no abnormalities.

Computed tomography (CT) scan of the thorax, abdomen and pelvis to identify any occult septic focus revealed a large infrarenal aortic aneurysm, as well as aneurysms of the right common iliac (RCI) and femoral (RCF) and left popliteal (LP) arteries. Fluorodeoxyglucose positron emission tomography (FDG-PET) showed significantly increased uptake in the RCI, right common femoral and LP artery aneurysms (Figure 1), as well as in the intervertebral discs at C6/7 and L3/4. A diagnosis of multiple mycotic aneurysms and intervertebral discitis secondary to metastatic endovascular S. aureus infection was made.

Figure 1

Fused transaxial CT-PET images for the diagnosis and surveillance of multiple mycotic aneurysms, demonstrating a sterile abdominal aortic aneurysm and inflammatory aneurysms of the RCI, RCF and LP arteries. An interval scan (February 2010) after a prolonged course of antibiotics demonstrated inflammatory suppression at all sites. This was maintained following endovascular repair of the aortic aneurysm (August 2010), although peri-stent activity was observed at this time. Maximum standard uptake values (SUVmax) are shown.

Intravenous linezolid was added to flucloxacillin and rifampicin, and continued for 1 month. The fevers remitted, repeated blood cultures remained sterile and CRP returned to normal within 2 weeks. The vascular surgeons opted not to intervene at this time, in view of the response to antibiotics, high anaesthetic risk and mortality associated with early discontinuation of antiplatelet therapy. He was discharged from hospital and completed a further 10 weeks course of oral rifampicin and flucloxacillin. He remained well with no evidence of progression of the mycotic aneurysms on follow-up CT angiography. Surveillance CT-PET 4 months after completion of antibiotics further demonstrated minimal signal in all aneurysms, consistent with microbiological suppression. Elective endovascular repair of his aortic aneurysm was consequently conducted without immediate complications. Unfortunately, a further surveillance CT-PET 6 months later demonstrated peri-stent inflammation in the abdominal aorta. He subsequently deteriorated and died secondary to septicaemia.

Review of mycotic aneurysms

Mycotic (or infective) aneurysms are localized, irreversible vascular dilatations caused by weakening and destruction of the vessel wall by an invasive organism establishing an infective arteritis. They are now rare in clinical practice (due to effective and prompt antibiotic therapy) constituting only 1–3% of all arterial aneurysms,1 but potentially life threatening. The term ‘mycotic’ derives from the mushroom-like appearance of the aneurysms originally described,2 and not their underlying microbiological aetiology.3 Endarterial infection may arise through haematogenous seeding from distant septic foci (endocardial vegetations, infected thrombi or intravascular devices) either direct to the arterial intima or to deeper mural layers via the vasa vasorum; lymphatic spread (particularly tuberculosis),4 contiguous extension (purulent pericarditis or osteomyelitis)5 or direct inoculation (iatrogenic during angiography or through intravenous drug misuse).6

Clinical presentation

Initial symptoms are often non-specific, the most common presentation being a febrile illness with insidious onset, general malaise and weight loss, gradually deteriorating into uncontrolled sepsis.7 Those presenting late may manifest with profound septicaemia or with consequences of rapid aneurysm expansion or rupture. There is typically an elevated peripheral blood leucocyte count (neutrophilia in 65–83%), erythrocyte sedimentation rate and CRP.8 The initial search for an infective source is usually negative, and thus mycotic aneurysms are an important differential in patients with pyrexia of unknown origin (PUO) or unexplained persistent bacteraemia. Mycotic aneurysms usually affect major arteries, classically at branch points. The femoral artery is most frequently involved, followed by the abdominal, thoracoabdominal and thoracic aortas.1 Aneurysms of peripheral limb and intracranial vessels are uncommon, and visceral arteries are extremely rare.9 The majority are solitary, but occasionally multiple synchronous lesions develop.10

The natural history of untreated mycotic aneurysms is of fatality from either massive haemorrhage or fulminant sepsis.7,11 Manifestations of aneurysm expansion correlate with anatomical location. Although occurring as late as complications in terms of the natural history, up to 45% rupture, with fistula formation in 18%.7 Thoracic aneurysms may present with tearing chest or interscapular pain, and abdominal aneurysms with abdominal pain and a pulsatile expansile mass,7,11,12 and both may produce radiofemoral delay. Dissections of the ascending thoracic aorta may cause sudden severe aortic regurgitation, or myocardial ischaemia or infarction if the coronary ostia are involved or cardiac tamponade; those in the descending or abdominal aorta may cause renal failure if the renal arteries are occluded. Rupture results in massive internal haemorrhage and hypovolaemic shock.

Peripheral arterial aneurysms present with localized pain, a pulsatile mass with palpable thrill and local inflammatory changes.6,13 They are associated with lower mortality of 0–15% due to their relatively superficial location, leading to earlier presentation.6,13,14 Most are proximal, and only rarely are vessels below the popliteal involved. There may be distal vascular compromise or embolization and compressive neuropathic phenomena. Intracerebral aneurysms may cause headaches, seizures or focal neurological signs, but in the majority are asymptomatic prior to rupture.15 Sudden onset of cerebrovascular events with neutrophils and bacteria in the cerebrospinal fluid should alert clinicians to the possibility of a ruptured cerebral mycotic aneurysm. The anterior circulation is most frequently affected, and 20–33% arise on the circle of Willis and must be differentiated from berry aneurysms.15,16 Most visceral mycotic aneurysms involve the superior mesenteric artery and manifest as mesenteric ischaemia or bowel infarction,17 and there are also cases reported affecting the hepatic,18 splenic,17 renal,19 pulmonary20 and coronary arteries.21

Risk factors

In the pre-antibiotic era, the most common pre-disposing factor was bacterial endocarditis.22 This is now only present in the minority, with the exception of intracranial lesions that are almost exclusively associated with intracardiac sources.23 The principal current risk factor is atherosclerosis8 (and its determinants male sex,22 increasing age24 and cigarette smoking): damage to the intimal lining increases susceptibility to microbial colonization and secondary degeneration.24 Vascular anomalies including pre-existing aneurysms, aortic coarctation or patent ductus arteriosus also elevate risk and easily rupture.

Immune competence is another important determinant with increased frequency of mycotic aneurysms observed with malignancy, diabetes mellitus, alcohol misuse, use of immunosuppressive medication and HIV infection. Unusual pathogens, such as fungi, cause disease in those with profound immunocompromise.25 Intravenous drug users comprise a distinct and important subgroup. The behaviour is particularly associated with peripheral artery lesions, principally the brachial artery followed by subclavian, radial, femoral and popliteal.6 Importantly in this cohort, fever may be absent in up to 70%, although clinical signs consistent with a peripheral aneurysm are usually apparent.26 The danger is in misrecognition of a vascular lesion as an abscess, and attempting to treat this by incision and drainage results in arterial rupture or fistula formation.


It is important to obtain a microbiological diagnosis given the need for protracted courses of antimicrobial therapy. Blood cultures are positive in 50–75% of patients,7,22,27 with reduced rates in those already on antibiotics or with anaerobic infection. In such individuals, it is typically possible to obtain positive cultures from surgically resected aneurysm tissue, achieving microbiological evaluation in nearly all patients.22 Polymicrobial cultures are uncommon,28 but when found most frequently occur in intravenous drug users.14 Initial empirical treatment is often required and typically targets the common aetiological agents according to the individual epidemiological risks. It is possible that nucleic acid amplification techniques may further improve microbiological diagnosis.29

The vast majority of mycotic aneurysms are bacterial. Gram-positive cocci predominate, with S. aureus accounting for around 45% of cases and streptococci for 10%.22,30 Combined, these rise to 90% for intracranial lesions, due to the association with infective endocarditis.23 Approximately 30–40% are caused by enteric-derived organisms, of which half are non-typhi Salmonella species;22,31 these are particularly prevalent in Asian populations32 and older age groups (since they exhibit predilection for atherosclerotic arteries),24 and carry high incidence for early rupture.24

Treponemal aneurysms, once considered commonplace, are now exceedingly rare and in fact the least common manifestation of cardiovascular syphilis.33 When they occur, it is due to lodging of treponemes within vasa vasorum of the aortic adventitia, causing obliterative endarteritis and necrosis. The proximal aorta appears particularly susceptible and eggshell calcification may be visible on plain radiography; extrathoracic disease is exceptional. Microbiological diagnosis is usually made upon positive serology using a combination of less specific non-treponemal investigations [such as the venereal disease research laboratory (VDRL) test], followed by confirmation with specific treponemal assays. Mycobacterial aneurysms are also extremely rare but may occur in areas of high prevalence.4 In the majority, there is a contiguous source of infection, usually tuberculous lymphadenitis in anatomical association with the vascular tree, and thus the thoracoabdominal aorta is principally affected.

Fungal aneurysms constitute ∼1% of mycotic aneurysms,34 and are seen in immunocompromised individuals. They are most commonly due to candidal species35 or Aspergillus,25 although other fungi including Penicillum25 and Histoplasma22 have been reported. In post-transplant patients, infection is usually transmitted from the donor and shows particular pre-dilection for graft vessel anastomoses.35 Mortality is high.


In all patients in whom the diagnosis of mycotic aneurysm is considered, transthoracic followed by transoesophageal echocardiogram should be performed to exclude endocarditis. These modalities may also reveal aneurysms of the thoracic aorta. If negative, this should prompt the search for another source of endovascular infection.

Multi-detector CT angiography remains the modality of choice for aortic mycotic aneurysms, with sensitivity of 92–96% and specificity of 93–100%.36–40 Additional advantages include detection of synchronous or source lesions and three-dimensional reconstruction for planning of intervention. Suggestive features include new aneurysm formation; rapid expansion or morphological change of known aneurysms; synchronous lesions; and intramural or perivascular gas, oedema, soft tissue mass or stranding.37,41 Ring enhancement may been seen.9 Disruption or disappearance of aortic calcification is a late sign, and may herald imminent rupture;24in contrast, extravasation indicates this has already occurred. Risk of rupture correlates with aneurysm diameter and rate of enlargement; there are no other imaging features that accurately predict outcome.9

Magnetic resonance angiography (MRA) is an alternative modality but currently restricted by longer examination times, increased susceptibility to motion artefact and small volume coverage.42,43 It is nonetheless particularly useful for intracranial lesions, with sensitivity of 95–100% and specificity of 82–96%.44,45 Invasive aortography is reserved for patients in whom mycotic aneurysm cannot be excluded by non-invasive tests, as it is limited by its capacity to image only the vessel lumen and not extravascular changes, and carries risks of distal embolization and rupture of the already fragile inflamed arterial wall.46

Peripheral arteries are best imaged initially by ultrasound. Aneurysms appear as circumscribed, hypoechoic lesions adjacent to the main arterial lumen with turbulent flow on Doppler.47 Differentiation of infective from sterile lesions requires correlation with the clinical presentation, but suspicious features include a lobulated vascular mass; an indistinct irregular arterial wall; and peri-aneursymal oedema or soft tissue mass.9 Conventional angiography remains the gold standard, but suffers the same limitations as direct aortography.37

There has been interest in using radionuclide-based imaging to identify and monitor inflammatory lesions. Leucocyte scintigraphy has been employed, but lacks the sensitivity and specificity of CT.37,48 Our case adds to recent reports on the utility of PET, of proven value in investigating PUO in which relevant foci were identified in almost 50% of cases.49,50 In terms of vascular imaging, it has high sensitivity for medium and large vessel vasculitis and aortic graft infection,51 with the caveat that mild uptake is seen in atherosclerotic lesions.52 PET in isolation has an ∼15% false-positive rate;49 this can be overcome by performing synchronous CT. There has been, to our knowledge, only one previous report of its use in the detection of a mycotic abdominal aortic aneurysm.53 Our case further highlights not just its diagnostic utility, but also detection of synchronous lesions, differentiation of infective from sterile aneurysms and monitoring response to antibiotic therapy.

Antibiotic therapy

There are no randomized controlled trials to inform on optimal treatment of mycotic aneurysms, but expert opinion recommends the combination of intensive antibiotic therapy alongside surgical repair. The goals of the latter are to confirm the diagnosis, achieve microbiological source control, prevent or contain rupture and haemorrhage and reconstruct the arterial tree. Intensive broad-spectrum antibiotic therapy should be instituted promptly and subsequently refined based on microbiological results. Empirical initial treatment should cover streptococcal and staphylococcal species as well as gram-negatives; treponemal, mycobacterial and fungal aneurysms are treated according to the standard protocols.25 Poor prognostic factors include advanced age, delayed diagnosis, gram-negative aetiology, immunocompromise, thoracic location, non-surgical management, rupture, embolic events and septic shock.46

There is no consensus on overall duration of treatment. Emergency operations to control impeding rupture or severe recalcitrant sepsis should not be delayed, but in other individuals pre-operative antibiotics for 2–4 weeks are associated with better outcomes.46 Unruptured aneurysms should be monitored with serial imaging: those which are not resolving or enlarging despite antibiotics necessitate urgent surgical intervention. The optimum post-operative duration is also unclear, but should continue for a minimum of 6 weeks. Most sources recommend a minimum of 3–6 months, and discontinuation only when there is no longer clinical evidence of ongoing sepsis, inflammatory markers have normalized and blood cultures are sterile.22,34,46 A longer course should be administered in patients with immunocompromise. A case has been made for lifelong antibiotics in patients without source control,34 Salmonella infection24 or who have undergone prosthetic vascular bypass.54 The strongest predictor of poor long-term outcome is persistent infection, with a 1-year survival of only 39%.55,56

Surgical intervention

Medical therapy alone is associated with poor outcome, with in-hospital mortality of 50% and event-free 1-year survival of only 32%.57 Complete surgical excision is curative, whereas although antibiotics may control sepsis they do not reduce risk of rupture of the weakened vessel wall.58 Operative technique depends on aneurysm site, local expertise and primary source of endovascular infection (taking into account management of contiguous or distant septic foci).13

Surgical management of infrarenal aortic lesions can be accomplished by either combined proximal and distal arterial ligation with extra-anatomic bypass (usually axillary-bifemoral or two axillary-unifemoral anastomoses) or in-line prosthetic graft, both accompanied by complete excision of the infected field.41 While in-line grafting is appropriate for low-grade infections, there is early mortality of 25%, similar incidence of septic complications and 20% require subsequent formal reconstruction.59 Extra-anatomic bypass is therefore indicated for high-grade lesions. Vascular homografts have also been trialled successfully, although arterial grafts risk future aneurysmal dilatation.60 Venous grafts, particularly the superficial femoral vein, while less prone to infection or aneurysmal dilatation, require longer operating times and have associated complications at harvest sites.61 Aneurysms of the thoracic and suprarenal aorta appear at lower risk of persistent or recurrent infection, and are therefore usually most suitably managed with prosthetic grafts,7 as extra-anatomic bypasses are more prone to thrombosis and need for future reconstruction.62

Peripheral artery aneurysms are managed by ligation or reconstruction (most frequently ileofemoral bypass).63 Ligation and excision of the infected field with delayed revascularization has lower morbidity than immediate reconstruction: it is less challenging, faster, establishes definitive haemostasis and is unlikely to lead to major tissue loss (although there are well-documented risks of critical limb ischaemia requiring amputation).64 Intracranial and visceral aneurysms may be treated surgically with clipping or excision,16,18,65 or endovascularly by embolization with non-selective cyanoacrylate occlusion or coiling.36 Success rates are 80–100%, with low procedural mortality.66

Endovascular repair is well-established for non-infective aneurysms. There have been concerns about adopting a similar approach for mycotic aneurysms, principally whether residual infection due to lack of excision and debridement of the infected field could be overcome by antibiotic therapy alone, and whether placement of the graft as a foreign body would allow long-term colonization. Persistent infection is a strong predictor of long-term mortality.56 Nonetheless, endovascular intervention has been successfully applied: most series suggest 30-day mortality of ∼10%, endoleak rate of 20%, persistent infection in 22.9%56 and a 2-year survival of 75%.54,56,67,68 Results can be improved by delaying the procedure if possible until antibiotic suppression therapy has achieved negative blood cultures, soaking or coating stents in antibiotic, performing adjunctive debridement and continuing long-term post-operative antimicrobials.


Mycotic aneurysms are rare but universally fatal without appropriate management. Diagnosis and surveillance remain clinical challenges, and CT angiography is currently considered the modality of choice. The combined use of PET is a valuable addition: as illustrated by our case, it not only augments sensitivity and specificity, but also distinguishes inflammatory from non-inflammatory aneurysms, highlight synchronous lesions and monitors remission/recurrence. This is particularly pertinent if a long-term antibiotic strategy is adopted in patients deemed high surgical or anaesthetic risk. Surgical or endovascular intervention, alongside intensive antibiotic therapy, remains definitive treatments, although patients must to be assessed on a case-by-case basis. The optimal duration of antibiotic cover is a matter for expert opinion, although prospective use of CT-PET to determine inflammatory activity could prove valuable to guide such recommendations.


The authors thank Dr Jamshed Bomanji for assistance in interpretation of the nuclear medicine images. S.M.-J., A.I.Z. and the Department of Nuclear Medicine receive financial support from the UK Department of Health NIHR Comprehensive Biomedical Research Centres funding scheme.

Conflict of interest: None declared.


  • *These authors contribute equally to this work.


View Abstract