Abstract
SUMMARY: Skull base osteomyelitis is a relatively rare condition, generally occurring as a complication of advanced otologic or sinus infection in immunocompromised patients. Skull base osteomyelitis is generally divided into 2 broad categories: typical and atypical. Typical skull base osteomyelitis occurs secondary to uncontrolled infection of the temporal bone region, most often from necrotizing external otitis caused by Pseudomonas aeruginosa in a patient with diabetes. Atypical skull base osteomyelitis occurs in the absence of obvious temporal bone infection or external auditory canal infection. It may be secondary to advanced sinusitis or deep face infection or might occur in the absence of a known local source of infection. Atypical skull base osteomyelitis preferentially affects the central skull base and can be caused by bacterial or fungal infections. Clinically, typical skull base osteomyelitis presents with signs and symptoms of otitis externa or other temporal bone infection. Both typical and atypical forms can produce nonspecific symptoms including headache and fever, and progress to cranial neuropathies and meningitis. Early diagnosis can be difficult both clinically and radiologically, and the diagnosis is often delayed. Radiologic evaluation plays a critical role in the diagnosis of skull base osteomyelitis, with CT and MR imaging serving complementary roles. CT best demonstrates cortical and trabecular destruction of bone. MR imaging is best for determining the overall extent of disease and best demonstrates involvement of marrow space and extraosseous soft tissue. Nuclear medicine studies can also be contributory to diagnosis and follow-up. The goal of this article was to review the basic pathophysiology, clinical findings, and key radiologic features of skull base osteomyelitis.
ABBREVIATIONS:
- ASBO
- atypical skull base osteomyelitis
- EAC
- external auditory canal
- Ga-67
- gallium-67 citrate
- IgG4
- immunoglobulin G4
- Tc99m MDP
- technetium Tc99m methylene diphosphonate
- NEO
- necrotizing external otitis
- SBO
- skull base osteomyelitis
- TSBO
- typical skull base osteomyelitis
Skull base osteomyelitis (SBO) is a rare, potentially life-threatening infection that can present a diagnostic challenge clinically and radiologically.1⇓⇓-4 While reports differ in terminology, there are generally 2 categories of SBO: typical and atypical. Typical SBO (TSBO) is the most common and classically occurs in elderly patients with diabetes as a result of necrotizing external otitis (NEO) caused by Pseudomonas species. (Fig 1).1,3 Atypical SBO (ASBO), also called central SBO, predominantly involves the basisphenoid and basiocciput and occurs without preceding otologic infection (Fig 2).2,5 Recognition of SBO is increasing, and it is clear that radiologic evaluation plays a critical role in diagnosis and management. The goal of this article was to review the pathophysiology, clinical presentation, and detailed radiologic findings using multiple modalities including CT, MR imaging, and nuclear medicine.
TSBO
TSOB is considered a part of the NEO spectrum, and these terms are often used interchangeably.1,2,6 TSBO can also occur secondary to other otologic infections, including complicated otomastoiditis or petrous apicitis.4,7,8 It can also occur secondary to trauma or as a surgical complication.7⇓⇓-10
Disease may begin as localized otitis externa (Fig 3). Progressive, deeper infection leads to NEO, in which microorganisms invade local cartilage and bone, spreading through natural gaps in the cartilaginous framework of the external auditory canal (EAC) (fissures of Santorini).1 Infection can extend from the EAC through the foramen of Huschke into the temporomandibular joint (Fig 4). Progression of NEO ultimately causes TSOB, with local osseous destruction and localized marrow infiltration. From the EAC, infection most often spreads anteromedially to the infratemporal and preclival soft tissues, petrous apex, and clivus (Fig 5).1
Patients with TSBO have otorrhea and severe otalgia, with pain often out of proportion to the physical findings. Local adenopathy may be present, but fever and leukocytosis are often absent.1 The erythrocyte sedimentation rate is generally increased and can be used to monitor treatment. Trismus may occur with involvement of the masticator space. Infiltration of the EAC or nasopharyngeal soft tissues may produce mass effect and suggests underlying malignancy. Eustachian tube obstruction can cause further fluid accumulation and phlegmon in the middle ear. Cranial neuropathies can occur and may indicate a poor prognosis. Facial nerve palsy is seen in 25% of patients as the disease spreads medially to involve the stylomastoid foramen.1,11 Extension to the petrous apex may result in Gradenigo syndrome (facial pain, cranial nerve VI palsy, and persistent otorrhea).12
Inferomedial spread of infection to the jugular foramen and carotid space can result in multiple lower cranial neuropathies (cranial nerves IX–XII).1,2 Involvement of the sympathetic plexus along the internal carotid artery can also produce Horner syndrome. Rare cases of Villaret syndrome (neuropathies of cranial nerves IX–XII plus Horner syndrome) have been reported when SBO affects the jugular foramen.13 Secondary thrombophlebitis of the jugular bulb and sigmoid sinus can also occur.11 The internal carotid artery can be affected anywhere from the neck to the cavernous sinus, producing infectious arteritis, thrombosis, pseudoaneurysm, and stroke.
Intracranial spread can result in meningitis, epidural abscess, and cavernous sinus thrombosis. If the cavernous sinus is affected, multiple upper cranial nerves can be involved.11 Cranial nerve involvement is typically unilateral but can be bilateral in advanced SBO crossing the midline.
Patients with diabetes are particularly prone to NEO/TSBO due to a combination of immune dysfunction and microvascular angiopathy. Pseudomonas aeruginosa is responsible for 98% of cases.14 The virulence of this Gram-negative bacterium is related to angioinvasion and small-vessel thrombosis.1 Rarely, other bacteria including Staphylococcus species are reported. Otologic fungal infections, especially Aspergillus species, are increasingly reported in immunocompromised patients and can lead to fungal SBO.4,15
TSBO has been traditionally associated with high morbidity and mortality despite intensive antibiotic therapy, with reported survival rates of around 50%.2,16 Although prognosis has improved, recent studies still report mortality of up to 30%, and it is still difficult to cure.17⇓-19
ASBO
ASBO, or central SBO, for has a predilection for the clivus and occurs without precipitating otologic infection. It can be idiopathic or secondary to regional infections of the sinus, deep face, or oral cavity.4,8,10,16,19 The distinction between typical and atypical is not always clear because patients may have occult or partially treated infection before diagnosis of SBO. Furthermore, infection of the temporal bone can spread medially to the central skull base and vice versa, making the true origin uncertain in some cases.6,20
The clinical features of ASBO are nonspecific. Patients are generally middle-aged to elderly with underlying diabetes or other immunocompromised states (HIV, chronic steroid use, and so forth).2 Ridder et al18 reported that 70% of patients with ASBO had a predisposing factor affecting bone vascularization, including diabetes (45%). Additional predisposing factors include previous radiation therapy, anemia, malnutrition, chronic cardiopulmonary disease, Paget disease, and osteoporosis. Rare patients have ASBO with no relevant pre-existing illness.
The most common symptoms of ASBO are headache and cranial neuropathies, with sinonasal symptoms reported in 25%.2 Fever is uncommon, found in <20%.2,3,5 The erythrocyte sedimentation rate can be elevated, but leukocytosis is absent more often than not.3,5 ASBO can begin with persistent sinus or other local infections, with spread from pneumatized space or soft tissues to the osseous skull base. Involvement of preclival or nasopharyngeal soft tissues raises concern for nasopharyngeal neoplasm. Intracranial extension can lead to meningitis, multiple cranial neuropathies, and cavernous sinus thrombosis.2,5,16
Gram-positive bacteria, including Staphylococcus species, are more common than Pseudomonas species.2,3 ASBO can be caused by fungal organisms, especially mucoraceal family in ketoacidosis and Aspergillus species in patients with neutropenia (Fig 6).3,15,21 Nontuberculous Mycobacteria species are being increasingly recognized in the immunocompromised population. Some cases are polymicrobial.
In ASBO, a 90.5% survival rate has been reported with aggressive management at 18-month follow-up, though up to one-third of the patients experienced residual neurologic sequelae.2
A summary of the comparison of clinical characteristics of TSBO and ASBO is shown in the Table.
Pathology
Biopsy is often necessary in the clinical course to exclude neoplasm, examine features of non-neoplastic tissue, and obtain direct microbiologic specimens for Gram stain, culture, and antibiotic-sensitivity assessment. In cases of SBO, pathologic specimens show inflammatory changes that vary from edema to purulence, with varying degrees of tissue necrosis. Histology may not reveal microbes directly, and cultures are important for definitive diagnosis and specific treatment.5 However, there have been emerging reports of culture-negative SBO, typically in patients with previous incomplete/partial treatment with topical or oral antibiotics.22,23 There is a wide range of reported culture-negative cases in the literature.24⇓-26 In a systematic review by Mahdyoun et al,26 culture-negative cases of NEO in the literature ranged from 0% to 36%. Djalilian et al,22 in 2006, reported 8 cases of SBO, all of which were culture-negative. Some authors have reported 100% positive cultures in their studies of patients with SBO,27 whereas some have reported up to 70% of patients with negative cultures, ultimately requiring empiric treatment.28 These data are limited by the absence of prospective studies and small sample sizes; a systematic analysis of these data to assess differences between typical and atypical SBO in this regard is beyond the scope of this article.
Imaging.
Radiologic evaluation is critical for prompt diagnosis of SBO. The imaging approach depends on presenting symptoms. However, in general, a combination of complementary studies using high-resolution bone CT and gadolinium-enhanced MR imaging is often necessary. In challenging cases, molecular imaging studies can provide functional and metabolic information.
CT.
Unenhanced CT is often first-line for suspected head and neck infections and is adequate for identifying opacification, mucosal thickening, and air-fluid levels in the temporal bones and sinuses. High-resolution submillimeter-section CT using a bone algorithm can be reformatted in multiple planes and is the study of choice for identifying cortical bone erosion or trabecular demineralization that accompanies osteomyelitis.
In suspected TSBO, it is critical to assess cortical bone loss, which can be subtle. In areas without prominent marrow and trabecular bone, this may be the only clue to SBO. Several key areas should be evaluated, including the bony EAC, mastoid tip, temporomandibular joint, petrous apex, petro-occipital fissure, foramen lacerum, jugular foramen, and clivus. In cases of ASBO, there is often evidence of invasive sinusitis with cortical erosions of the paranasal sinuses, particularly the sphenoid or ethmoid sinuses and possibly along the anterior clivus and foramina of the central skull base.5
Contrast-enhanced CT with soft-tissue windows can also be useful. In TSBO, the anteromedial spread of cellulitis and phlegmon in the infratemporal soft tissues manifests as poorly defined enhancement and soft-tissue infiltration. Occasionally, a frank abscess can be identified in the preclival soft tissues. Asymmetric soft-tissue fullness of the nasopharynx is common and can mimic an infiltrating neoplasm, especially nasopharyngeal carcinoma (Fig 7A). The soft-tissue invasion and neurovascular complications can occur before or without frank bone destruction, especially in early or aggressive diseases such as fungal SBO. CTA or CTV can be of additional benefit for evaluation of vascular complications, including cavernous sinus thrombosis or stroke.
MR Imaging.
MR imaging of the skull base is complementary to CT and is superior for evaluating soft-tissue extent, marrow involvement, and intracranial complications related to SBO.3 A combination of MR images is necessary to fully evaluate the skull base and surrounding structures, including T1, T2, STIR, DWI, and T1-weighted fat-saturated contrast-enhanced images. Early TSBO may demonstrate abnormal signal and enhancement of the EAC, with marked edema and inflammation of the auricular soft tissues. With progression, there is anteromedial spread as described above.
On T1 images, the ill-defined soft-tissue process demonstrates hypo- or isointensity to muscle and causes obliteration of normal fat planes: retromandibular fat, parapharyngeal fat, and retropharyngeal fat in the preclival region. The infiltration consists of inflammation, edema, and phlegmon and produces T2 hyperintensity and heterogeneous enhancement on T1-weighted fat-saturated contrast-enhanced sequences. As noted, the soft-tissue abnormality in the nasopharynx may be the dominant feature and can be indistinguishable from an infiltrative neoplasm (Fig 7B, -C).
When osteomyelitis affects the bone marrow, there is loss of normal fat signal in the marrow space, causing T1 hypointensity and STIR hyperintensity. The affected marrow demonstrates heterogeneous gadolinium enhancement.4,7,21 With advanced infection, bone marrow may become necrotic and evolve into an abscess, producing a region of peripherally enhancing tissue. Mucormycosis infection, in particular, can produce a combination of abnormal enhancement and nonenhancing areas of devitalized soft tissue and bone.15,21
Early in TSBO, the volume of marrow space abnormality may be small and attention to detail and use of fat-suppression are necessary to identify subtle involvement of mastoid bone, petrous apex, or occipital bone. Progression of disease can lead to more diffuse involvement of the marrow space and can include the clivus. While most cases will present as unilateral abnormalities, disease can progress to bilateral skull base involvement.
With ASBO, the MR signal abnormalities of the affected soft tissues and the bone marrow will be similar to those in TSBO. The primary difference is that the epicenter of disease will be the sphenoid bone. Paranasal sinus opacification may be conspicuous. The primary marrow signal abnormality will be in the clivus but may also involve the lesser and greater wings of the sphenoid or petrous apices. The soft-tissue abnormalities will also involve the preclival soft tissues but may be more symmetric, producing diffuse fullness in the nasopharynx. Soft-tissue infiltration of the pterygopalatine fossa with obliteration of normal fat in these regions is typical of invasive sinusitis associated with SBO.
DWI, especially non-EPI DWI, may be of benefit in evaluating SBO. Diffusion restriction in nonenhancing fluid collections can help confirm an abscess. Additionally, ADC values may allow distinction between SBO and neoplasm, with bacterial SBO values shown to be higher than those in nasopharyngeal carcinoma or lymphoma.17 On postcontrast imaging, focal abscesses could show a peripheral rim of enhancement, whereas a neoplasm would generally demonstrate enhancement within the diffusion-restricting tissue.29
CT or MR imaging is not necessarily helpful in long-term monitoring of the disease because radiologic findings lag behind clinical improvement.1,11,30 Overall, improvement in soft-tissue findings is the best radiologic indicator of early improvement,11,31 but abnormalities of bone may persist for weeks to months despite a clinical response to treatment.1,3,5,31
Nuclear Medicine.
Before the advent of CT and MR imaging, nuclear imaging served as a cornerstone for evaluation of SBO.32 The various radionuclide studies provide functional and metabolic information that can help confirm and localize infection of the skull base and can be complementary to clinical findings and anatomic imaging to monitor treatment response.
Technetium Tc99m methylene diphosphonate (Tc99m MDP) can demonstrate increased osteoblastic bone activity that occurs in response to infection. A 3-phase Tc99m MDP bone scan is more sensitive than CT for early detection of SBO, with sensitivity approaching 100%, including SPECT Tc99m MDP scans, which are reported to be more sensitive and a better prognosticator for patients with malignant external otitis.33 It typically shows abnormal increased tracer uptake in bone on all 3 phases (ie, immediate blood flow, blood pool [5–10 minutes], and delayed phase [3–4 hours]), whereas isolated soft-tissue infection will be differentiated by a normal delayed phase. If available, delayed-phase SPECT improves anatomic localization. A bone scan, however, lacks specificity for infection because it can demonstrate abnormal bone activity in malignancies, trauma, recent surgery, or noninfectious inflammation. Furthermore, in the setting of osteomyelitis, a bone scan can remain abnormal even after satisfactory treatment due to bone healing and remodeling (Fig 8).1,2,34,35
A gallium-67 citrate (Ga-67) scan targets acute-phase reactants like lactoferrin and bacterial siderophores and can bind to white blood cells engaged in the immune response to infection. This feature provides high specificity for infection and is complementary to the bone scan (Fig 9). A normal Ga-67 scan, even with an abnormal bone scan, reliably excludes SBO, and increased uptake on a Ga-67 scan confirms infection. A Ga-67 scan plays an important role in monitoring of treatment response, converting to normal findings after successful treatment; persistent increased uptake suggests residual infection. The scan can be repeated to monitor antibiotic response until findings become normal.6,34 This repetition can be reassuring for the consulting physician and the patient, especially in complicated cases in which the diagnosis was delayed or in doubt. The major limitation of a Ga-67 scan is the long scan time requiring delayed images up to 48–72 hours.
A technetium-labeled white blood cell scan is less commonly used but like Ga-67, it has a high specificity for SBO in the initial diagnosis. A tagged white blood cell study can be used for confirming healing at the end of antibiotic therapy.1,11,36
Overall, the literature has variable data regarding the overall diagnostic value of nuclear medicine studies.37 A recent review of malignant otitis externa literature revealed pooled sensitivities for technetium-99 and gallium-67 of 85.1% and 71.2%, respectively, with poor specificity; however, the data were deemed insufficient for a meta-analysis. The authors, therefore, advised against the routine use of these studies in SBO management in patients with a known diagnosis on conventional imaging. However, these examinations were considered to be reasonably sensitive tests in patients with an unclear diagnosis despite an otomicroscopic examination or other imaging studies.37 The authors also concluded that there were insufficient data to determine the usefulness of these modalities during follow-up and that larger prospective studies would be necessary.
[18F] FDG- PET detects increased glucose metabolism. FDG is nonspecific and accumulates at sites of high glucose demand, including active infection, but also in postoperative, inflammatory, or neoplastic tissue. The advantages of FDG-PET/CT over other nuclear studies are wider clinical availability, shorter imaging time, and higher spatial resolution. It can be complementary to determine the extent of infection in confirmed cases of SBO and for evaluation of treatment response. In a recent study comparing the diagnostic performance of [18F] FDG-PET/CT with MR imaging, both modalities had comparable sensitivities (87.5 versus 81.25%, respectively), but PET-CT had better specificity (71.0% versus 28.5%, respectively) in identifying infection. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of FDG-PET/CT in detecting SBO were 96.7%, 93.3%, 98.3%, 87.5%, and 96.1%, respectively.38 With wider availability of hybrid PET-MR imaging scanners with superior soft-tissue detail and metabolic information in a single imaging session, PET-MR can be used to follow patients with SBO.6,38,39 PET-MR imaging with gadolinium in combination with high-resolution CT is an excellent approach for suspected new SBO, and a combination of FDG-PET with either MR imaging or CT may provide sufficient follow-up (Fig 10).6
Differential Considerations.
The primary diagnostic dilemma for SBO arises from neoplastic processes because they can also infiltrate the skull base and the adjacent soft tissues. Carcinoma of the EAC can have similar clinical and radiologic features to those of TSBO and, therefore, needs to be excluded. Other neoplasms to consider are metastatic disease, nasopharyngeal carcinoma, sinonasal neoplasm, and lymphoma.3
Neoplasms that affect the skull base may arise locally from various sources that are amenable to clinical evaluation. For example, a primary skin lesion such as EAC squamous cell carcinoma or a primary mucosal lesion such as nasopharyngeal carcinoma may have a dominant, clinically obvious primary soft-tissue mass that heavily influences image interpretation and preliminary management decisions. Other neoplasms of the skull base such as lymphoma, myeloma, or metastatic disease may be accompanied by pertinent clinical history and other extracranial findings to indicate underlying systemic metastatic or multifocal disease and would typically present with more bulky lesions than ill-defined skull base pathologies.
Most primary forms of local invasive neoplasms that affect the skull base have a dominant soft-tissue mass or nodule that secondarily invades the bone. While the osseous skull base may have significant derangement due to invasion and destruction, the primary lesion presents on cross-sectional imaging as nodular or masslike enhancement in the extraosseous soft tissues. As opposed to SBO, malignant neoplasms typically displace or replace normal anatomy without preservation of tissue planes. Enhancement and fullness without destruction of fascial planes may support a diagnosis of SBO over tumor.29,40 Additionally, at presentation, MR imaging has been reported to show greater disease involvement compared with CT in central SBO, whereas malignancy generally shows equal involvement; and combining the CT and MR imaging information can help differentiate between central SBO and malignancy.40 With high-resolution imaging, especially T1-weighted contrast-enhanced fat-saturated MR imaging, the margins of the tumor can be reasonably mapped and measured. If a lesion involving the skull base has a dominant, solidly enhancing soft-tissue component that correlates with a clinically obvious mass involving the skin or mucosa, neoplasm is strongly favored over infection. In these cases, biopsy of the apparent lesion is indicated as initial management. In situations in which a primary lesion is not clinically apparent, imaging can help direct a potential biopsy of suspected viable tumor by identifying localized extraosseous soft-tissue enhancement.29
The dilemma is made more difficult in the setting of a small or occult primary tumor, deep subcutaneous or submucosal invasion, or tumor with significant ulceration or necrosis. In these cases, a dominant soft-tissue nodule or mass may be lacking both clinically and radiologically. This situation occasionally occurs in the setting of nasopharyngeal carcinoma or adenoid cystic carcinoma that affects the central skull base. In such circumstances, the radiologic picture may be dominated by osseous demineralization on CT and abnormal marrow space signal and enhancement on MR imaging. In these cases, infiltrating tumor can be difficult to distinguish from SBO. In addition, focal soft-tissue infection associated with SBO can produce phlegmon, occasionally taking on masslike qualities of soft-tissue fullness, mass effect, and enhancement.
The decision to biopsy tissue is based on multidisciplinary consultation. Surface mucosal or submucosal disease can be biopsied through endoscopic approaches by ear, nose and throat surgeons, whereas more deep-seated disease may require image-guided biopsies. Radiologic findings can provide specific clues in cases in which there have been repeat biopsies negative for malignancy. Alternate biopsy targets can be suggested on the basis of the imaging appearance to plan surgical approaches to the pterygopalatine fossa or orbital apex for biopsy and tissue analysis.23 Often during the course of the disease, these cases will need a multidisciplinary approach, with consultations among the referring clinician, surgeons, infectious disease specialists, and radiologists at different steps.23
Neoplasms that are intrinsic to the skull base such as chordoma or chondrosarcoma can be considered invasive or infiltrative, but they tend to be slow-growing, focally expansile, and relatively well-circumscribed. Tumor grows beyond the margins of the bone into adjacent soft tissues, but there is not typically an inflammatory response.
Rare differential considerations would include non-neoplastic diseases, including granulomatosis with polyangiitis and other granulomatous diseases (eg, tuberculosis, sarcoidosis).3 Idiopathic skull base inflammation (inflammatory pseudotumor), an idiopathic noninfectious inflammatory condition, may primarily involve the skull base or extend from the orbit and can appear identical to SBO.41 Immunoglobulin G4 (IgG4)-related disease can affect almost any organ, most commonly the submandibular, lacrimal, or parotid glands, but it can also involve the skull base. IgG4-related disease typically shows increased IgG-4-positive plasma cells on tissue sampling, and elevated serum IgG4 concentrations are also seen.42 An elevated IgG4/IgG ratio of >0.4 was detected in 40% of cases in a study of inflammatory pseudotumor and helped to distinguish them from SBO in some instances because none of the SBO cases had a ratio of >0.4.43 Ultimately, radiologic findings alone are insufficient to differentiate these inflammatory entities from SBO and malignancy. These entities often are suspected in the absence of a mass or signs of infection, but endoscopic biopsy/tissue sampling will be needed for diagnosis.23,43
Primary bone conditions of the skull base, including fibrous dysplasia and Paget disease, can be in the differential for SBO on MR imaging; however, CT would show their typical appearances with bony expansion and no associated soft-tissue abnormality. Ground glass opacification with variable lytic foci would be seen in fibrous dysplasia and osseous expansion with a lytic lesion (osteoporosis circumscripta) or mixed lytic-sclerotic foci having a cotton wool appearance as seen in Paget disease.3
Management/Treatment.
TSBO often has a classic presentation and is not difficult to diagnose, whereas ASBO is often a diagnostic dilemma due to the nonspecific initial presentation. For any infiltrative/destructive process of the central skull base, neoplastic processes including nasopharyngeal carcinoma, lymphoma, or leukemia need to be ruled out first with other aforementioned inflammatory or noninflammatory conditions also considered. Skull base or nasopharyngeal biopsies need to be performed in a timely manner to rule out these differential possibilities as well as to obtain tissue samples due to the potential for rapid progression of SBO.11,18 Tissue samples should undergo microbial analysis with culture and flow cytometry for lymphoma in addition to pathologic analysis, especially if the clinical suspicion is high and no obvious soft-tissue lesions are seen.1,5 Pathogen-specific antibiotic therapy including IV antibiotics followed by long-term oral antibiotics would be the mainstay of treatment, currently recommended for 6–20 weeks,2,18,44 with wide variations in the duration of treatment observed in a survey-based study of otolaryngologists in the United Kingdom.45 Initially however, broad-spectrum antimicrobials, including coverage for P aeruginosa and methicillin-resistant Staphylococcus aureus particularly for non-otologic causes, are recommended to cover the possibility of polymicrobial infection in ASBO before culture and sensitivity information is available.18 Antipseudomonal antibiotics, such as carbapenems and third-generation cephalosporins, with ciprofloxacin in the long term, are considered an alternative to single initial therapy with ciprofloxacin in view of growing ciprofloxacin resistance in the intensive care setting, including in culture-negative cases, in which the antibiotic choice can be difficult.1,18,24 Surgical debridement of necrotic bone and soft tissue, especially for fungal disease, may be required in advanced cases with drainage of involved air cells or sinuses and of abscesses to also help improve antimicrobial penetration. However, an early and aggressive surgical approach has also been found to be beneficial and is recommended by some authors, especially in patients with prolonged ear infections and at the first signs of cranial neuropathy.18 Hyperbaric oxygen therapy has also been suggested as an ancillary treatment but has not shown an impact on survival.1,2
Summary
Diagnosis of SBO, clinically and radiologically, requires a high index of suspicion, and a delay in diagnosis is common. It should be considered in the differential consideration for any infiltrative skull base process, particularly if biopsies are negative for malignancy. Thin section, high-resolution bone CT of the skull base would be necessary to identify early cortical erosion followed by multiplanar pre- and postcontrast MR imaging to identify marrow space involvement. Nuclear medicine imaging studies can play an important role in difficult-to-diagnose cases and in follow-up. Long-term antibiotics with surgical debridement in advanced cases are the mainstay of management.
Footnotes
Disclosures: Philip R. Chapman—UNRELATED: Employment: University of Alabama Birmingham, Comments: I am an Associate Professor at University of Alabama; Payment for Lectures Including Service on Speakers Bureaus: Los Angeles Radiological Society, Comments: I received an honorarium for a total of 5 lectures at a recent annual meeting in Los Angeles, California, January 2020; Royalties: Elsevier, Comments: royalties for textbooks: 1) Chapman PR, Harnsberger HR, Vattoth S. Imaging Anatomy: Head & Neck. 1st ed. Elsevier, 2018 (September): ISBN: 978-0323568722; Shaaban AM, ed, Diagnostic Imaging, Oncology, 2nd ed, November 2019, ISBN: 9780323661126. Gagandeep Choudhary—UNRELATED: Employment: University of Alabama at Birmingham.
Indicates open access to non-subscribers at www.ajnr.org
References
- Received May 8, 2020.
- Accepted after revision August 21, 2020.
- © 2021 by American Journal of Neuroradiology