Malignant Otitis Externa

Malignant external otitis (MEO) is an infection that affects the external auditory canal and temporal bone. The causative organism is usually Pseudomonas aeruginosa, and the disease commonly manifests in elderly patients with diabetes. The infection begins as an external otitis that progresses into an osteomyelitis of the temporal bone. Spread of the disease outside the external auditory canal occurs through the fissures of Santorini and the osseocartilaginous junction.[1]

Frequency

United States

Malignant external otitis (MEO) is more common in humid and warm climates than in other climates.

Mortality/Morbidity

Cranial neuropathy

Cranial nerves can be affected by inflammation along the skull base or by a neurotoxin produced by Pseudomonas species. The facial nerve (VII) is affected most commonly, usually at the stylomastoid foramen. As the disease progresses, cranial nerves IX, X, and XI can be affected at the jugular foramen, followed by XII at the hypoglossal canal. Cranial nerves V and VI can be affected if the disease extends to the petrous apex.

In 1977, Chandler reported a 32% incidence of facial nerve paralysis.[6] The incidence of facial nerve paralysis appears to have decreased with the development of more effective medical therapy as shown by Franco-Vidal et al who reported a 20% incidence of facial nerve paralysis in 46 treated patients.[7] The other cranial nerves are affected less frequently than the seventh cranial nerve. The development of cranial neuropathy generally was thought to reflect advanced-stage disease associated with a worse prognosis. More recently, Corey et al, Soudry et al, and Mani et al suggested that the presence of facial nerve paralysis does not worsen the prognosis.[8, 9] Recovery of facial nerve function is poor and unpredictable, and should not be used as an indicator of successful treatment. Other cranial nerves that are affected have a higher rate of recovery.

Intracranial complications

These complications rarely occur in the absence of cranial nerve palsies. Meningitis, brain abscess, and dural sinus thrombosis may ensue. Cranial neuropathies related to the jugular foramen should raise concern for sigmoid sinus thrombosis. Cavernous sinus thrombosis should be considered if cranial nerves V or VI are affected. Intracranial complications reflect severe disease and are commonly fatal.

Comorbid conditions

Patients with malignant external otitis (MEO) almost always have diabetes, often with other multiple medical problems.[10] During the course of therapy, Chandler found some deaths related to pneumonia, uremia, myocardial infarction, strokes, and liver failure. Franco-Vidal showed that patients with systemic immunodeficiencies had a worse prognosis.[7]

A nationwide, population-based case-control study from Taiwan, by Yang et al, supported the association between diabetes and malignant external otitis (MEO). The investigators reported that 54.8% of patients with MEO had previously diagnosed diabetes, compared with 13.9% of controls. The adjusted odds ratio for previously diagnosed diabetes in the presence of MEO was 10.07.[11]

A study by Sylvester et al analyzing 8300 cases of inpatient malignant external otitis from the National (Nationwide) Inpatient Sample database (2002-2013) found that diabetes mellitus in adult and elderly patients was associated with a greater number of comorbidities, longer hospitalizations (5.5 days vs 4.0 days in patients without diabetes), and higher hospital charges ($25,118 vs $17,039 in patients without diabetes). However, in-hospital mortality rates were not significantly greater in patients with diabetes than in those without it (0.6% vs 0.5%, respectively).[12]

A retrospective study by Peled et al of patients with malignant external otitis (MEO), over 90% of whom had diabetes mellitus, looked at the effects of glycemic control in MEO. The investigators found that higher hemoglobin A1c (HbA1c) levels were associated with longer hospital stays, the concentration being 7.6% in patients who where hospitalized for less than 20 days, and 8.7% in those who were hospitalized for 20 days or more. Moreover, while P aeruginosa was found in 26.7% of patients with a mean blood glucose level of 140 mg/dL or lower, the bacterial species existed in 51.0% of those with a higher mean concentration; however, disease outcome was not affected by the glucose level.[13]

Sex

Malignant external otitis (MEO) is more common in males than in females.

Age

Malignant external otitis (MEO) has been reported in all age groups but is most common in patients who are elderly (age, >60 y).

See the list below:

• Diabetes (90%) or immunosuppression (illness or treatment related)

• Severe, unrelenting, deep-seated otalgia

• Temporal headaches

• Purulent otorrhea

• Possibly dysphagia, hoarseness, and/or facial nerve dysfunction

See the list below:

• Inflammatory changes are observed in the external auditory canal and the periauricular soft tissue.

• The pain is out of proportion to the physical examination findings.

  • Marked tenderness is present in the soft tissue between the mandibular ramus and mastoid tip.

  • Granulation tissue is present at the floor of the osseocartilaginous junction. This finding is virtually pathognomonic of malignant external otitis (MEO). Otoscopic examination may also reveal exposed bone.

• The cranial nerves (V-XII) should be examined.

• Mental status examination should be performed. Deterioration of the mental status may indicate intracranial complication.

• The tympanic membrane is usually intact.

• Fever is uncommon.

See the list below:

• Diabetes (90% of patients)

  • Diabetes is the most significant risk factor for developing malignant external otitis (MEO).

  • Small-vessel vasculopathy and immune dysfunction associated with diabetes are primarily responsible for this predisposition.

  • The cerumen of patients with diabetes has a higher pH and reduced concentration of lysozyme, which may impair local antibacterial activity.

  • No difference in predisposition is found between diabetes types I and II.

  • The predisposition is not necessarily related to the severity of glucose intolerance or periods of hyperglycemia.

• Immunodeficiencies, such as lymphoproliferative disorders or medication-related immunosuppression

• AIDS

  • Malignant external otitis (MEO) associated with AIDS may have a different pathophysiology than classic malignant external otitis (MEO).

  • Patients present with similar symptoms but are generally younger and do not have diabetes.

  • Granulation tissue may be absent in the external auditory canal.

  • Pseudomonas is not necessarily the dominant causative organism.

  • Patients with AIDS generally have a poorer outcome than patients with diabetes.

• Aural irrigation: As many as 50% of cases of malignant external otitis (MEO) have been reported to be preceded by traumatic aural irrigation in patients with diabetes.

Leukocyte count

The leukocyte count in malignant external otitis (MEO) is usually normal or mildly elevated. A left shift is not commonly found.

Erythrocyte sedimentation rate

The erythrocyte sedimentation rate (ESR) is invariably elevated, with an average of 87 mm/h. It begins to decrease within 2 weeks of initiating therapy but takes many months to return to normal. ESR can be used to support the clinical diagnosis since acute external otitis or ear canal malignancy usually does not cause a rate elevation in this lab test.

Serum chemistry

Patients with known diabetes need an evaluation of the serum chemistry to determine if the infection is affecting their baseline glucose intolerance. Patients without a history of diabetes should be tested for glucose intolerance.

Culture and sensitivities from the external auditory canal

Culture from the ear drainage should be performed ideally before antimicrobial therapy is initiated. The most common causative organism is P aeruginosa (95%). This organism is an aerobic, gram-negative rod. Pseudomonas species have a mucoid coating that deters phagocytosis. Exotoxins (ie, exotoxin A, collagenase, elastase) can cause tissue necrosis, and some strains produce a neurotoxin that may be partially responsible for cranial neuropathies.

Less common organisms identified include Aspergillus and Proteus species, Candida species, Staphylococcus aureus, and S epidermidis.

A study by Hobson et al concluded that vigilance for nonpseudomonal organisms is necessary in treating patients with malignant external otitis, especially those without diabetes, since methicillin-resistant S aureus (MRSA) is increasingly being diagnosed as the causative organism, and not always in the presence of diabetes. Among 20 patients with malignant external otitis, the investigators found 9 patients with P aeruginosa and 3 with MRSA. They also determined that although diabetes was present in all patients with P aeruginosa, it existed in only 1 of the patients with MRSA and in only 55% of all patients with nonpseudomonal infections.[14]

A study by Gruber et al indicated that in patients with refractory, culture-negative malignant external otitis, a fungal infection should be suspected. The study involved a cohort of 19 patients with malignant external otitis, three of whom had refractory, culture-negative disease. Polymerase chain reaction (PCR) assay revealed a fungal pathogen in all three, with empirical targeted therapy being used to successfully treat their condition.[15]

These are important adjuncts for determining the presence of osteomyelitis, the extent of disease, and response to therapy.

Technetium Tc 99 methylene diphosphate bone scanning

Technetium Tc 99 methylene diphosphonate bone scanning is based on binding to osteoblasts. This scan depicts as little as a 10% increase in osteoblastic activity. However, this test is not specific since tumors or bony dysplasias, in addition to osteomyelitis, can cause osteoblastosis.

It is useful in the initial evaluation because a positive finding in the correct clinical context can lead to confirmation of the diagnosis.

The test is not useful for assessing the response to therapy since results remain persistently positive long after clinical improvement because of continuous bone remodeling and reformation.

This test may also have limited usefulness for patients with a prior history of mastoiditis or otologic surgery.

The application of single-photon emission computed tomography (SPECT) technology has improved the poor spatial resolution traditionally associated with this test.

Gallium citrate Ga 67 scan

Gallium citrate Ga 67 scan is very sensitive but is not specific because gallium binds to actively dividing cells, including inflammatory cells, tumor cells, and osteoblasts.

Uncertainty is possible regarding whether a positive test result represents an inflammatory condition, soft tissue, or bone disease.

This test is most helpful when used as a monitor of successful treatment. Improvement of a positive test result correlates with therapeutic response.

A baseline test is usually obtained at the initial diagnosis for comparison with follow-up studies during treatment.

A quantitative comparison of the lesion to the nonlesion side may improve the interpretation of these studies for distinguishing acute external otitis from malignant external otitis (MEO) and for determining the efficacy of therapy.

The application of SPECT technology has improved the poor spatial resolution traditionally associated with this test.

Indium In 111–labeled leukocyte scan

Indium In 111–labeled leukocyte scan attempts to provide the same sensitivity as a gallium citrate Ga 67 scan but is more specific to an inflammatory process.

It does not appear to provide an improvement in scintigraphic technique for helping to establish the diagnosis. It may be better than gallium citrate Ga 67 scans for assisting in establishing the correct timing of disease resolution.

This test can be unreliable for imaging chronic osteomyelitis in other areas of the body. Thus, the accuracy of this application needs further study.

CT scanning and MRI

CT scanning and MRI are both useful for evaluating the anatomic extent of soft tissue inflammation, abscess formation, and intracranial complications.

CT scanning fails to diagnose early osteomyelitis because 30-50% of bone destruction is required to detect osteomyelitis by CT scanning.

MRI provides poor bone resolution. The soft tissue manifestations regress on CT scanning and MRI with response to therapy.

Bone changes remain persistently abnormal on CT scans for at least one year and are not well demonstrated by MRI studies. Thus, neither of the tests can be used to determine osteomyelitis resolution.

A retrospective study by Goh et al suggested that MRI findings in skull base osteomyelitis secondary to malignant external otitis (MEO) can help to differentiate the osteomyelitis from advanced nasopharyngeal carcinoma by the presence of a combination of lateral extension, increased T2 signal in adjacent soft tissues, deficiency of architectural distortion, and enhancement equaling or surpassing that of mucosa.[16]

Most authors advocate obtaining a CT scan with the initial evaluation for all patients, whereas Benecke advocates obtaining this test selectively for patients with cranial neuropathy, extensive bone changes on technetium scan, or poor clinical response to treatment. Grandis et al and Okpala et al support obtaining a CT scan early in the diagnostic/treatment algorithm. Peleg et al showed that there is a correlation between clinical course and the extent of anatomical areas involved as measured on initial CT scan findings.[17]

MRI and CT scanning are equally sensitive in detecting the soft tissue extent of the disease, but MRI is more sensitive for detecting intracranial complications.

See the list below:

• Obtain a biopsy of the external auditory canal to exclude carcinoma or other etiologies.

Nadol described the histopathology of 2 temporal bones affected by malignant external otitis (MEO). The infection did not spread through the pneumatized air tracts of the temporal bone. Rather, it spread along the vascular and fascial planes on exiting the temporal bone through the external auditory canal osseocartilaginous junction or fissures of Santorini. The otic capsule appeared to be resistant to the disease process. Linthicum described histopathologic findings in 5 temporal bones. Extensive destruction in the wall of the bony external auditory canal and osteomyelitic destruction of the wall of the fallopian canal in the descending portion of the facial nerve was seen. The infection spread beneath the otic capsule to erode the wall of the carotid canal and then extended into the central skull base.

See the list below:

• Levenson et al, Corey et al, Benecke, and Davis et al have proposed staging systems for malignant external otitis (MEO).[18, 19, 20]

  • These staging systems are generally based on extent of soft tissue/bony involvement or development of neurologic complications.

  • None of these staging systems has been widely adopted.

Treatment includes meticulous glucose control, aural toilet, systemic and ototopic antimicrobial therapy, and hyperbaric oxygen therapy.[4, 5]

Systemic antibiotic choice

Until the development of third-generation antipseudomonal cephalosporins, long-term intravenous antibiotics using an antipseudomonal penicillin and aminoglycoside were the mainstay of medical treatment.

Several authors have demonstrated the effectiveness of intravenous ceftazidime monotherapy in the treatment of malignant external otitis (MEO).

Fluoroquinolones that attain high soft tissue and bone levels with oral doses were then developed. Subsequently, several authors have demonstrated the efficacy of oral ciprofloxacin monotherapy.

Although no established treatment guidelines are available, case series and anecdotal experience suggest that initial outpatient therapy with oral ciprofloxacin is efficacious for patients without a fluoroquinolone allergy, cranial neuropathy, or intracranial complication and who do not require hospital admission for diabetes or pain management.

The widespread use of fluoroquinolones for upper respiratory infections and simpler ear infections is beginning to confound the typical clinical spectrum of malignant external otitis (MEO). Ciprofloxacin-resistant P aeruginosa has been increasingly isolated in patients with malignant external otitis (MEO), accounting for as many as 33% of isolates in patients who failed outpatient management in a study by Berenholz et al.[21] Most notably in this patient population, 63% of isolates from 1998-2001 were resistant to ciprofloxacin, whereas only 15% of isolates were found to be resistant in the 10 years before this 3-year period. No increased morbidity or mortality was found in patients with ciprofloxacin-resistant Pseudomonas. Patients with resistant P aeruginosa require parenteral antibiotics with antipseudomonal beta-lactam antibiotics with or without an aminoglycoside.

A retrospective study by Carlton et al indicated that as rising antimicrobial resistance complicates treatment for malignant external otitis (MEO), multidrug and long-term parenteral antibiotic therapy with extended inhospital treatment are needed in select cases.[22]

Duration of therapy

Symptoms and examination findings improve with appropriate therapy, but these changes do not correlate with the length of needed therapy. Despite symptom relief, prolonged antimicrobial treatment as indicated for osteomyelitis is still indicated.

Imaging studies are helpful in determining the adequate length of treatment for each patient.

Treatment response should be evaluated with a gallium citrate Ga 67 scan every 4-6 weeks during treatment. Benecke recommended ending treatment 1 week after the gallium citrate Ga 67 scan findings return to normal and confirming this with a repeat scan 1 month after the treatment is stopped. Using this protocol for 13 patients, the average duration of treatment was 8.8 weeks with a range of 4-17 weeks.

Hyperbaric oxygen therapy

This should be used only as an adjunct to antimicrobial therapy; it should not be used alone. Hyperbaric oxygen therapy may be helpful for patients with complications, experiencing a poor response to therapy, or with recurrent cases.

See the list below:

• Chandler advocated surgery in his original report when appropriate antimicrobials were not available; he had very poor results, with a 50% mortality rate.

• Surgical removal of the lesion requires resection of large portions of the temporal bone.

• Because of the histopathology of malignant external otitis (MEO), removal of contiguous areas of bone may not be sufficient because of the spread of infection through vascular and fascial planes.

• Surgery is now reserved for local debridement, removal of bony sequestrum, and abscess drainage.

• Facial nerve decompression is not indicated for patients with facial paralysis.

See the list below:

• Consultation with internal medicine specialists is required for the management of diabetes and other comorbidities.

• Infectious-disease consultants may help with the choice of antibiotics in complicated cases.

The goals of pharmacotherapy are to reduce morbidity and prevent complications.

Class Summary

The choice of antimicrobial therapy and whether treatment can be delivered PO versus IV, as monotherapy, or on an outpatient basis need to be carefully individualized to each patient. Refer to Medical Care for further discussion of these issues.

Ciprofloxacin (Cipro)

A DNA gyrase inhibitor that prevents DNA replication.

Ceftazidime (Fortaz, Tazicef, Tazidime)

Third-generation cephalosporin that inhibits bacterial cell wall synthesis by adhering to penicillin-binding proteins. Can be either bactericidal or bacteriostatic depending on the organism, tissue penetration, and drug dosage.

Ticarcillin/clavulanate (Timentin)

Extended-spectrum penicillin/beta-lactamase inhibitor. Ticarcillin binds to penicillin-binding proteins, thus inhibiting bacterial cell wall synthesis. Clavulanate binds irreversibly to beta-lactamases, thus preventing inactivation of ticarcillin.

Gentamicin (Garamycin, Jenamicin)

Binds to 30S ribosomal subunit, thus inhibiting bacterial protein synthesis. Aminoglycosides have bactericidal activity against P aeruginosa. Bacterial strains resistant to one aminoglycoside still may be sensitive to other antibiotics within this category.

A study by Lee et al indicated that prognostic factors in malignant external otitis include the duration of diabetes mellitus, the C-reactive protein level and the erythrocyte sedimentation rate (inflammatory markers), the extent of the malignant external otitis as assessed through imaging studies, and cranial nerve involvement.[23]

In a retrospective study, Stevens et al stratified malignant external otitis into severe and nonsevere categories, with severe cases characterized by at least two of the following characteristics: cranial nerve VII palsy, a positive fungal culture, relapse, surgical treatment, and major radiographic findings. Compared with cases in the nonsevere category (16 patients), those classified as severe (12 patients) were associated with longer courses of antibiotics and with more admissions and relapses related to the condition. Moreover, cure was achieved in only four of the severe patients, compared with 14 of the nonsevere cases.[24]

A literature review by Mion et al indicated that in patients with fungal malignant external otitis, treatment tends to be effective in cases in which surgical debridement has not been employed, facial palsy is absent, the condition was caused by an Aspergillus species, and imaging findings at diagnosis and follow-up are absent. However, the investigators cautioned that surgical debridement may have been associated with less effective treatment because, possibly, it tended to be used in patients with more aggressive, advanced disease.[25]

A retrospective study by Schwam et al indicated that in patients with malignant external otitis (MEO), a prolonged length of hospital stay (8 days or longer) was independently associated with readmission within 30 days. Performance of a surgical procedure and the occurrence of central nervous system (CNS) complications were independent risk factors for prolonged length of stay.[26]

Disease recurrence

Disease recurrence is reported in 9-27% of patients. It usually is related to inadequate length of therapy and manifests as recurrent headaches and otalgia, not as otorrhea. The ESR begins increasing again.

Malignant external otitis (MEO) can recur as long as one year after treatment is completed; thus, a patient should not be considered cured until that time.

Mortality

Chandler reported a mortality rate of 50% in the original series. The mortality rate has decreased to 20% with the introduction of appropriate antibiotics, improved imaging modalities, and increased awareness of the disease.

Most current studies report a mortality rate of less than 10%, but mortality remains high for patients with cranial neuropathies (other than VII), intracranial complications, or with irreversible systemic immunosuppression.