Necrotizing (malignant) external otitis

INTRODUCTION — Necrotizing (malignant) external otitis (NEO; also termed malignant otitis externa) is an invasive infection of the external auditory canal leading to skull base osteomyelitis. The infection typically occurs in older adult patients with diabetes mellitus.

The epidemiology, pathogenesis, clinical features, and management of necrotizing external otitis will be reviewed here. The spectrum of ear and eye infections due to P. aeruginosa and other clinical syndromes caused by this pathogen are discussed separately. (See "Pseudomonas aeruginosa infections of the eye, ear, urinary tract, gastrointestinal tract, and central nervous system", section on 'Ear infections' and "Pseudomonas aeruginosa infections of the eye, ear, urinary tract, gastrointestinal tract, and central nervous system", section on 'Eye infections' and "Pseudomonas aeruginosa skin and soft tissue infections" and "Pseudomonas aeruginosa bacteremia and endocarditis" and "Pseudomonas aeruginosa pneumonia".)

An overview of external otitis is also presented separately. (See "External otitis: Pathogenesis, clinical features, and diagnosis".)

NOMENCLATURE AND CLASSIFICATION — Necrotizing external otitis (NEO), previously referred to as malignant external otitis, is a rare invasive ear infection that can lead to skull base osteomyelitis if not promptly treated.

The extension of the infection is typically classified by radiographic findings. In early phases of NEO, only the soft tissues of the external auditory canal are involved. As the infection progresses, mastoid and then temporal bone involvement becomes evident, followed by involvement of the temporal-mandibular joint or the nasopharynx. If the infection is not promptly diagnosed and treated, the infection can progress to involve the skull base and then can extend to the contralateral side as well as the meninges and the brain parenchyma.

ANATOMY AND PATHOGENESIS — Necrotizing external otitis (NEO) is a complication of acute bacterial external otitis that occurs when the infection spreads beyond the skin to soft tissue, cartilage, and bone of the temporal region and the skull base (figure 1 and figure 2). The anatomy of the ear canal is discussed in detail elsewhere. (See "External otitis: Pathogenesis, clinical features, and diagnosis", section on 'Anatomy'.)

Ear exposure to water has been directly linked to simple external otitis [1-4] and Pseudomonas spp is commonly found in water sources. Forceful irrigation with tap water has been associated with progression to NEO. In one study of 13 patients with NEO, 62 percent had had tap water irrigation of the ears (for the purposes of removing cerumen) performed by a clinician prior to the onset of symptoms [5]. Other reports have also described this association between aural water exposure and necrotizing external otitis [6-11]. (See "Epidemiology, microbiology, and pathogenesis of Pseudomonas aeruginosa infection".)

EPIDEMIOLOGY

Patients with glucose intolerance — Older adult diabetic patients are overwhelmingly the population at risk for necrotizing external otitis (NEO). More than 90 percent of adults with this disease were found to have some form of glucose intolerance in one review [12]. A number of hypotheses have been proposed to explain this predisposition in diabetics including:

●Microangiopathy in the ear canal, which might also occur more commonly in older adult individuals [6,13]

●Increased pH in diabetic cerumen [14,15]

It is unclear whether patients with uncontrolled diabetes are at higher risk that those who have well-controlled diabetes mellitus [13], but many patients develop NEO despite good control of their diabetes.

Other patient populations — Necrotizing external otitis can also occur in other populations, such as those with an immunocompromising condition including untreated or advanced human immunodeficiency virus (HIV) [16-23]. Very rarely, NEO has occurred in patients with recent ear surgery [24]. Necrotizing external otitis is a rare disease in children with few cases reported in the literature. Affected children are more likely to be immunocompromised (eg, malignancy or malnutrition). Children tend to develop fever and leukocytosis in association with this infection and are more likely than adults to have a concomitant P. aeruginosa bacteremia [13].

MICROBIOLOGY — Necrotizing external otitis (NEO) is caused by P. aeruginosa in more than 95 percent of cases [12,25,26]. When more than one organism is recovered, isolates accompanying P. aeruginosa tend to be normal skin flora.

In our experience, the majority of the other 5 percent of cases tend to include Aspergillus species [23,27-35]. In one retrospective study of 11 patients with skull base osteomyelitis that did not respond to antibiotics, all were caused by Aspergillus spp [34].

Occasional reports of cases caused by other organisms have included Staphylococcus aureus [33,36], Proteus mirabilis [37], Klebsiella oxytoca [38], Burkholderia cepacia [39], and Candida parapsilosis [33,40]. Infection with these other organisms characteristically occurs in immunocompromised hosts, such as those with acquired immunodeficiency syndrome (AIDS) or cancer, although cases of Aspergillus fumigatus have been reported in patients with diabetes or without any apparent risk factors [41-43]. Fungal NEO often presents with prolonged duration of symptoms that persist despite multiple courses of antibiotics; cranial nerve palsies are frequently present at the time of diagnosis [41]. While most cases of fungal necrotizing external otitis are due to Aspergillus spp, other fungi have also been reported, including Scedosporium apiospermum, Pseudallescheria boydii, Candida ciferrii, Candida orthopsilosis [44], and Malassezia sympodialis [45,46].

CLINICAL MANIFESTATIONS — Necrotizing external otitis (NEO) is characterized by otalgia, otorrhea, and visualization of granulation tissue in the external auditory canal.

●Clinical symptoms – Patients with necrotizing external otitis classically present with exquisite otalgia and otorrhea, which are not responsive to topical measures used to treat simple external otitis. The pain is generally more severe than that found in simple external otitis, although the two may be difficult to distinguish in their early stages. The pain in necrotizing external otitis tends to be nocturnal and extend into the temporomandibular joint, resulting in pain with chewing. Hearing loss may or may not be present.

●Physical examination – On physical examination, the external auditory canal appears swollen and inflamed and has a purulent discharge. Visualization of granulation tissue in the inferior portion of the external auditory canal at the bone-cartilage junction (at the site of Santorini's fissures) is a classic finding. However, this finding may be absent in atypical patients (eg, immunocompromised individuals and children). Inflammation of surrounding structures, such as the pinna, periauricular and retromandibular areas, and the mastoid tip, may also be seen. Less commonly, clenching and spasm of the jaw (trismus) is present. In rare cases, the infection causes a nasopharyngeal abscess or phlegmon that may mimic a necrotic cancer.

Fungal ear infections have a typical appearance although such findings are often absent in NEO. Presence of black dots or a dark coating suggests an Aspergillus spp infection (picture 1) while presence of white, sebaceous-like material suggests a candidal infection.

●Laboratory markers – Laboratory parameters are generally normal in necrotizing external otitis, with the exception of an elevated erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP). Although nonspecific, a strikingly elevated ESR (eg, ESR >60mm/hr) is the most characteristic laboratory abnormality and is a useful way of monitoring disease activity [13].

●Complications – As the infection advances, osteomyelitis of the base of the skull and temporomandibular joint osteomyelitis can develop [13,15,47,48]. Progression of the osteomyelitis can be associated with cranial nerve palsies. In one series of 23 adult diabetic patients with necrotizing external otitis, 10 patients had cranial nerve involvement, six had only facial nerve involvement, and four had a combination of nerves involved (VI, VII, IX, X, XI, XII) [49]. Children with necrotizing external otitis have a higher incidence of facial palsy due to their relatively undeveloped mastoid process and the more medial location of the fissures of Santorini, which places the facial nerve in closer proximity to the ear canal [50]. The glossopharyngeal, vagal, and spinal accessory nerves can be afflicted at the jugular foramen, and the hypoglossal nerve can be affected as it exits the hypoglossal canal. The trigeminal and abducens nerves are rarely affected at the petrous apex, and there is a single report of optic nerve involvement in a patient with necrotizing external otitis [51]. The olfactory, oculomotor, and trochlear nerves appear to be unaffected by this disease.

Other central nervous system complications are rare but can be fatal when they occur. These include meningitis, brain abscess, dural sinus thrombosis, or pseudoaneurysm of the intrapetrous internal carotid artery [8,52]. (See "Clinical features and diagnosis of acute bacterial meningitis in adults" and "Pathogenesis, clinical manifestations, and diagnosis of brain abscess" and "Septic cavernous sinus thrombosis".)

DIAGNOSIS

When to suspect NEO — Necrotizing external otitis (NEO) should be suspected in a patient presenting with otorrhea (especially at night) and otalgia that has not responded to topical antibiotics [13,24].

Approach to diagnosis — For patients suspected to have NEO, we obtain blood work, swab collection of the ear canal drainage, and imaging.

Blood work — We obtain blood cultures and elevated erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) for patients in whom we suspect necrotizing external otitis. Although bacteremia is rare in necrotizing external otitis in adults, immunocompromised patients and children may be bacteremic without overt signs of toxicity. Most patients with NEO will have ESR levels greater than 60mm/hr and in many, the ESR is >100mm/hr. Serial ESR and CRP are also helpful for monitoring disease activity while on treatment. (See 'Duration and follow up' below.)

Ear canal drainage swab — We send swabs of ear canal drainage or granulations to microbiology for Gram stain, bacterial culture, fungal stain (eg, calcofluor), and fungal culture. Recovery of P. aeruginosa or a mold indicates the presence of a pathogen, since these organisms are not normal components of ear canal flora even in diabetic patients [3,53,54]. Often, cultures may be polymicrobial. We focus our attention on bacteria that are more likely to be pathogenic, such as methicillin-resistant Staphylococcus aureus (MRSA), Pseudomonas aeruginosa, and Enterobacterales spp. Absence of growth on cultures does not mean infection is not present and is more likely if the patient received topical antipseudomonal antibiotics as initial treatment.

Imaging — The goal of imaging is to determine the extent of infection and presence of bone erosion. If the patient is clinically stable to be managed as an outpatient, imaging should be performed within three to five days of initiating antibiotics for NEO to determine the extent of infection and obtain a baseline.

●CT imaging is preferred – Visualization of bone erosion is necessary to establish the diagnosis of advanced NEO. Computed tomography (CT) imaging with and without contrast of the temporal bone is the best imaging modality for assessing bone erosion. CT imaging also allows for anatomic localization and determining the extent of the infection. While some studies indicate that CT findings predict clinical outcome [55], others have reported that the CT findings do not correlate with the clinical course [56]. In one prospective study, the presence of bone erosion and/or soft tissue abnormalities in the subtemporal region was helpful in making the diagnosis of necrotizing external otitis [57]. Although bone did not remineralize, resolution of the soft tissue component did correlate with disease activity.

●MRI with contrast if CT imaging non-diagnostic/contraindicated – When a CT scan is non-diagnostic or contraindicated, magnetic resonance imaging (MRI) with contrast is the next best imaging modality. Although MRI is not as effective at demonstrating bone erosion, MRI is better in visualizing soft tissues, determining the extension of the infection, and monitoring response to therapy. In some cases, anterior extension into the retrocondylar fat on MRI may indicate the earliest sign of NEO [58]. In one prospective study of seven patients with NEO, MRI was slightly better at demonstrating medial skull base disease due to its ability to delineate changes in the fat content of the marrow [59]. In another study, extension of the infection in more than one direction (eg, anterior, medial, midline, intracranial, extracranial) on MRI was associated with worse prognosis [58].

●Other less common imaging modalities – Imaging modalities other than MRI and CT scan are rarely used but may be appropriate in circumstances where CT and MRI are either contraindicated or not available. These imaging modalities include gallium, bone, and white blood cell (WBC) scans, as well as positron emission tomography (PET)/CT. All of these imaging techniques lack specificity and are not helpful in monitoring response to therapy.

•Gallium scan – Gallium citrate (Ga-67) scanning appears to be more specific than bone scanning since the radioisotope is incorporated into granulocytes and bacteria. While several studies have reported that gallium scanning can be used to follow disease activity, others have noted that normal scans can be found in patients with recurrent disease [60-66]. The combination of Ga-67 with single photon emission computerized tomography (SPECT) scanning may be useful in the diagnosis and follow-up examination if CT and MRI are not available or contraindicated [24,64-66].

•Bone scan – Bone scanning with technetium (Tc 99m), where the radionuclide tracer accumulates at sites of osteoblastic activity, is very sensitive in making the diagnosis of osteomyelitis but is not specific, since there are reports of positive bone scans in simple external otitis [67]. Quantitative bone scanning, however, may be able to distinguish simple from necrotizing external otitis and demonstrate disease resolution [68,69]. Generally, bone scans are not suitable for following response to treatment since they do not normalize. However, use of the mouse monoclonal antibody linked to technetium (Tc 99m sulesomab) was reported in two cases of necrotizing external otitis where imaging results appeared to accurately reflect disease activity and response to treatment [70].

•Indium-111-labeled WBC scan – The WBC scan is similar to the other radionucleotide tracer scans (eg, Gallium, Tc 99m) in regards to sensitivity for diagnosing NEO [71]. Specificity is similar to a bone scan.

•PET/CT scan – Although PET/CT scan can detect bone erosion [72], it is non-specific and has difficulty distinguishing malignancy from infection. The findings can also take time to resolve which also limits its utility in monitoring response to therapy.

Establishing the diagnosis — In the absence of a single pathognomonic criterion, the diagnosis of necrotizing external otitis is based upon a constellation of clinical (otorrhea, otalgia, granulation tissue approximately midway down the external ear canal) and laboratory (elevated ESR and/or CRP) findings. The lack of growth on bacterial and/or fungal culture does not exclude an infection and does not form the basis of establishing a diagnosis.

Post-diagnostic evaluation — Once NEO is established, we perform further laboratory testing and imaging to evaluate for underlying risk factors (if not yet known) as well as determine extent of disease and baseline markers that can be followed over time.

●Determining extent of the disease – Radiographic findings (presence of bone erosion on CT or MRI) are helpful for determining the extent of disease. Many patients present with early disease that has not yet affected the bone; in such cases, imaging may not show any abnormalities. If bone involvement is present, we recommend obtaining consultation with an infectious disease specialist to help guide treatment, if available.

●Evaluation for underlying risk factors – In all patients, we check a hemoglobin A1C, CBC with differential, full chemistry panel (electrolytes, liver function tests), and obtain a chest radiograph. If the patient is not known to have diabetes mellitus or immunocompromise, we also check HIV and consider further imaging to look for an occult malignancy.

●Baseline ESR and CRP – We obtain an ESR and CRP at the time of diagnosis, even if this was recently checked because it is helpful to have a baseline at the time intravenous (IV) antibiotics are started. These lab values can be trended over time (eg, once weekly or less frequently) to determine the response to therapy.

DIFFERENTIAL DIAGNOSIS — Other conditions that can present with otorrhea (especially at night) and otalgia that has not responded to topical antibiotics include cholesteatoma and squamous cell carcinoma of the temporal bone.

●Squamous cell carcinoma – Squamous cell carcinoma of the temporal bone can also present as a painful draining ear canal. Since radiographic studies cannot differentiate tumor from necrotizing infection, biopsy is the only definitive method to distinguish between these two entities. A positive culture for P. aeruginosa and elevation of the erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP) are more commonly associated with infection. Symptoms that do not improve with the administration of broad-spectrum antibiotics suggest malignancy. There are only two reports of simultaneous presentation of temporal bone cancer and necrotizing external otitis [73,74].

●Cholesteatoma – Cholesteatoma is a keratinized collection of squamous epithelial cells in the middle ear or mastoid that can become secondarily infected. Cholesteatomas can present with otorrhea, hearing loss, and/or vertigo and is typically diagnosed by ENT via otoscopy. (See "Chronic otitis media and cholesteatoma in adults", section on 'Establishing the diagnosis'.)

Occasionally, a patient may be diagnosed with skull base osteomyelitis without a known cause. Aside from NEO, other causes of skull base osteomyelitis include infections or malignancies extending from the paranasal sinuses or the oral cavity. Magnetic resonance imaging (MRI) with contrast can be helpful in determining the underlying cause [75].

TREATMENT

Initial management and site of care — Most patients with early necrotizing external otitis (NEO; eg, short duration of symptoms and infection does not appear to have extended beyond the soft tissues of the external ear) and who are likely to have fluoroquinolone-susceptible Pseudomonas infection can be managed in the outpatient clinic with oral ciprofloxacin. Ciprofloxacin has the most anti-Pseudomonas activity of all the fluoroquinolones, and oral absorption is excellent so that serum levels produced by the oral antibiotic are usually equivalent to those produced by intravenous (IV) administration.

However, clinicians should have a low threshold for admission to the hospitalization for IV antibiotics, expedited imaging, and consultation with ENT and infectious diseases physicians, since the infection can spread quickly and inadequate treatment is associated with poorer outcomes. Hospital admission is warranted for the following:

●Patients with uncontrolled diabetes (HgbA1C not at goal)

●Immunocompromised individuals

●Evidence of bony involvement on imaging (if imaging is available)

●Presence of cranial nerve palsies or any extent of the infection beyond the soft tissues of the external ear

●Patients who have already received a course of oral fluoroquinolones and are continuing to have symptoms

●Patients within a community where Pseudomonas spp resistance to fluoroquinolones is high. This information can often be found in the local antibiogram. With the introduction and widespread use of both oral and topical fluoroquinolones, the emergence of ciprofloxacin resistance has been increasing [12,76-78].

Empiric antipseudomonal therapy — Antipseudomonal antimicrobials are the mainstay of therapy for NEO. Our approach to treatment depends on the severity of disease. The local rate of fluoroquinolone resistance in Pseudomonas spp may also be a consideration. (See "Epidemiology, microbiology, and pathogenesis of Pseudomonas aeruginosa infection", section on 'Prevalence of resistance'.)

Oral antibiotics for select immunocompetent outpatients — For immunocompetent patients with early and uncomplicated malignant otitis externa in settings with low rates of P. aeruginosa resistance to ciprofloxacin, we suggest oral ciprofloxacin (in adults: 750 mg orally every 12 hours; in children: 20 to 30 mg/kg per day orally divided every 12 hours, max 1500 mg/day) monotherapy. Patients taking ciprofloxacin along with medications that may interfere with absorption (eg, iron or calcium supplements) should space these appropriately to prevent interference with absorption.

Patients who cannot tolerate fluoroquinolones should receive an IV antipseudomonal beta-lactam, such as cefepime or piperacillin-tazobactam. (See 'IV antibiotics for all other patients' below.)

We expect improvement in symptoms within 48 to 72 hours. If symptoms do not improve, we admit the patient to the hospital for IV antibiotics and broaden the spectrum of activity of the antibiotic regimen. (See 'No clinical improvement after 48 to 72 hours' below.)

Most of the supportive evidence for antipseudomonal antibiotics for the treatment of NEO is observational [79-85]. Cases are rare, and thus randomized controlled comparisons of therapy are not feasible. Studies from 1988 to 1991 showed that prolonged therapy with ciprofloxacin alone was effective in treating most cases of early NEO [79-85], although, as above, resistance to fluoroquinolones has been increasing [12,76-78].

IV antibiotics for all other patients — For patients who are immunocompromised or who have poorly controlled diabetes, have complications of NEO (eg, evidence of bone involvement on imaging, cranial nerve abnormalities, skull base osteomyelitis, extension to the meninges, nasopharynx, or temporomandibular joint), or who reside in a setting with high fluoroquinolone-resistance among P. aeruginosa isolates, we suggest empiric combination treatment with oral ciprofloxacin and an anti-pseudomonal beta-lactam (eg, cefepime, ceftazidime, or piperacillin-tazobactam) to ensure adequate antibiotic activity against P. aeruginosa. Ciprofloxacin can be given IV in patients who have poor gastrointestinal absorption. Patients with non-severe NEO who cannot tolerate fluoroquinolones can be given aztreonam as the second anti-pseudomonal agent. We do not recommend combination therapy with an aminoglycoside since ototoxicity with vestibular disturbances has occurred despite assiduous monitoring of aminoglycoside serum concentrations.

Antibiotic dosing is provided in the table (table 1). Antibiotics should be tailored based on susceptibility results once these are available.

The duration of IV therapy depends on the severity of the infection and extent of disease on presentation. For example, patients who have extensive skull base osteomyelitis may require six to eight weeks of IV therapy or more (usually with monotherapy after susceptibility results are available), while those with early infection (eg, minimal bone involvement) and cultures that demonstrate a ciprofloxacin-susceptible Pseudomonas may be able to switch to oral ciprofloxacin once cultures and susceptibility results return and the patient is clinically improving (eg, resolution or abatement of symptoms, decreasing Erythrocyte sedimentation rate [ESR]). (See 'Duration and follow up' below.)

The objective of initial combination therapy is to increase the likelihood of using an active agent in a patient for whom ineffective antibiotic therapy could lead to progressive disease with high morbidity. However, no studies have compared initial empiric combination therapy versus monotherapy with ciprofloxacin in severe cases of necrotizing external otis. Overall, we believe the risk of choosing monotherapy with oral ciprofloxacin far outweighs any inconvenience with IV therapy. (See "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections", section on 'Role of combination antimicrobial therapy'.)

Tailoring therapy — We reassess patients with NEO 48 to 72 hours after the initiation of empiric therapy. Assessment consists of evaluating the patient's clinical response as well as reviewing microbiologic results. Occasionally, antibiotics can be tailored effectively and safely based on culture results.

Assessing response to therapy — Improvement of ear pain is often the first sign of response to therapy and typically occurs within one week of starting effective therapy, while complete resolution of ear pain may take much longer. Hearing loss, trismus, and neurologic signs (eg, cranial nerve palsies) may take longer to resolve as well.

No pathogen identified and clinically improving — For patients who are clinically improving and whose culture did not identify any bacteria, we continue empiric anti-pseudomonal therapy. If the patient initially was started on oral ciprofloxacin, we continue the ciprofloxacin. If the patient was initially on IV antipseudomonal combination therapy, we narrow to oral ciprofloxacin for the duration of therapy (see 'Duration and follow up' below). As described above, the decision of when to switch to oral ciprofloxacin alone must be made on a case-by-case basis depending on the severity of the findings on presentation as discussed above. Consultation with an infectious disease specialist is especially helpful in such cases. (See 'IV antibiotics for all other patients' above.)

Absence of growth on cultures does not mean infection is not present. Negative culture results are more likely if patient received topical antipseudomonal antibiotics as initial treatment.

Pathogen identified and clinically improving — For patients in whom a pathogen has been identified, tailoring therapy depends largely on the specific pathogen; we continue therapy against Pseudomonas spp even if Pseudomonas spp is not isolated on culture.

Pseudomonas aeruginosa — If cultures yield fluoroquinolone-sensitive P. aeruginosa, we continue ciprofloxacin. If cultures yield drug-resistant P. aeruginosa, we select an appropriate antibiotic based on susceptibility testing results, usually a beta-lactam (eg, piperacillin-tazobactam, cefepime, meropenem). We do not recommend combination therapy with an aminoglycoside since ototoxicity with vestibular disturbances has occurred despite assiduous monitoring of aminoglycoside serum concentrations. (See "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections".)

Other pathogens — Pathogens other than P. aeruginosa can occasionally cause NEO. These pathogens are typically Enterobacterales spp or Aspergillus spp (or rarely other molds). However, even if Pseudomonas spp does not grow on culture, as is often the case in patients who have received topical fluoroquinolones, we usually treat with antibiotics that will also treat Pseudomonas spp while including antibiotics that will treat another potentially but much less common pathogen that has grown on culture (eg, methicillin-resistant Staphylococcus aureus [MRSA] and Enterobacterales spp).

Normal skin flora isolated on culture are likely commensal organisms that are not the cause of NEO and thus generally do not need to be treated.

●Enterobacterales spp – Occasionally, NEO may be caused by Proteus spp or Klebsiella oxytoca. If these organisms are isolated, we administer antibiotic therapy with activity against the isolated pathogen based on susceptibility data as well as Pseudomonas spp.

●Normal skin flora – Normal skin flora, such as coagulase-negative staphylococci or Cutibacterium acnes are typically commensal organisms that do not cause NEO. We generally do not target or tailor antibiotic therapy against these organisms when they are isolated from culture. Instead, we continue empiric antipseudomonal therapy.

●Staphylococcus aureus – S. aureus can present as part of normal skin flora or (rarely) be the causative pathogen of NEO. Isolation of this pathogen requires the clinician to consider the risks and benefits of adding anti-staphylococcal therapy to the antibiotic regimen. If the patient is clinically improving without anti-staphylococcal therapy, we do not alter the antibiotic regimen. However, if the patient is improving slower than anticipated, clinical improvement has stalled, or clinically worsens, we add antistaphylococcal therapy (eg, cefazolin for methicillin-sensitive S. aureus [MSSA], vancomycin for MRSA) for the duration of therapy.

●Fungi – Fungi rarely cause NEO. When they do, Aspergillus spp are the most common. Isolation of Aspergillus or another mold almost always warrants treatment as molds are not part of normal skin and ear canal flora. Candida species may be isolated from ear canal swab cultures, particularly in ears that have been treated with topical antibiotics and corticosteroids; this does not necessarily indicate a candidal infection and most commonly is a commensal organism.

We choose antifungals based on culture results and susceptibility. If Aspergillus species is the causative organism, we treat with a prolonged course of voriconazole (eg, 12 weeks) [86,87]. If voriconazole is not an option, liposomal amphotericin B is an alternative and after initial improvement on this antibiotic, consideration of step-down therapy to isavuconazole or posaconazole can be considered. These two newer azoles may be noninferior and better tolerated than voriconazole [88], although treatment of NEO with these agents has not yet been described. Successful treatment of other head and neck invasive fungal infections (invasive fungal sinusitis) with isavuconazole has been described [89]. Success in treatment of NEO caused by Aspergillus flavus has been reported with a short course of amphotericin B and long-term oral itraconazole in a single case report [90]. (See "Treatment and prevention of invasive aspergillosis", section on 'Initial therapy'.)

No clinical improvement after 48 to 72 hours — In some patients, clinical improvement may take longer than 72 hours. If culture results are available and demonstrate a likely causative pathogen, it may be reasonable to continue appropriate therapy and monitor. However, if there is no improvement in pain or drainage and cultures are unrevealing, we broaden the antibiotic regimen to provide activity against multidrug resistant bacteria as well as other pathogens.

●Not improving on oral antibiotics – If the patient was treated at home with oral antibiotics, we admit them to the hospital to initiate antipseudomonal combination therapy with oral ciprofloxacin and an intravenous (IV) beta-lactam anti-pseudomonal agent. (See 'IV antibiotics for all other patients' above.)

●Not improving on IV antibiotics – If the patient was already started on IV anti-pseudomonal combination therapy (eg, ciprofloxacin plus cefepime) in the hospital, we broaden the antibiotic regimen to include activity against MRSA and multidrug resistant P. aeruginosa (eg, IV vancomycin plus meropenem).

●When to add antifungal therapy – If ear swab culture results are negative and the patient does not improve (eg, persistent symptoms or persistently elevated inflammatory markers) on broadened empiric therapy (eg, IV vancomycin and meropenem) after several more days, or if the patient is not improving despite appropriate antibiotic therapy based on ear swab culture results, we suggest initiating antifungal therapy with voriconazole and obtaining a biopsy to send tissue for culture, susceptibility testing, and pathology (to rule out squamous cell carcinoma). Although evidence surrounding the addition and timing of antifungals are scarce, lack of improvement on broad-spectrum antibiotics suggests either a fungal infection (most likely Aspergillus spp) or a non-infectious entity. (See 'Differential diagnosis' above.)

Duration and follow up — For bacterial NEO, treatment for six to eight weeks is generally recommended, as indicated for osteomyelitis. Fungal NEO generally requires a longer duration of >12 weeks. However, duration should be determined based on patient's pace of clinical improvement after initiation of treatment. Resolution of clinical symptoms (eg, ear pain, ear drainage) and normalization of the ESR and/or CRP are indicators of resolving infection.

Adjunctive options for refractory cases — In refractory cases, surgical debridement or hyperbaric oxygen may be considered, although the evidence that either is helpful is unclear.

●Surgical involvement – Rarely, cautious debridement may be necessary when the infection involves a fungal pathogen and/or when its refractory to systemic antibiotics in immunocompromised patients. However, more often, only a deeper surgical biopsy is indicated to obtain deeper cultures in order to confirm the diagnosis and exclude malignancy.

●Hyperbaric oxygen – Hyperbaric oxygen has been used on occasion with mixed results [47,91-94]. However, a systematic review found no randomized controlled trials comparing hyperbaric oxygen therapy to treatment with antibiotics and/or surgery and concluded that there was no clear evidence to demonstrate the efficacy of hyperbaric oxygen treatment [95] (see "Hyperbaric oxygen therapy").

Interventions with limited role — There is no role for topical antibiotics in necrotizing external otitis therapy. Instillation of antipseudomonal topical agents may only increase the difficulty in isolating the organism from the ear canal.

Surgical excision also plays no role in the current treatment of this disease, although this modality was used prior to the availability of systemic antipseudomonal antibiotics. Cautious debridement and/or biopsy to rule out cancer are the only indicated surgical procedures. (See 'Adjunctive options for refractory cases' above.)

OUTCOMES — Prior to the development of systemic agents, recurrences of necrotizing external otitis (NEO) were frequent, and mortality was approximately 50 percent [6]. The introduction of parenteral semisynthetic penicillins reduced the mortality to 20 percent [96], and a further reduction in mortality to 5 to 10 percent has occurred since the introduction of ciprofloxacin and antipseudomonal cephalosporins, such as ceftazidime and cefepime [12].

Since then, mortality rates continue to decrease due to earlier diagnosis of the infection [26,33]. In a systematic review of retrospective studies involving over 2200 patients with NEO, one-year mortality was 2 percent [26]. Risk factors for mortality include age above 80 years and presence of comorbidities (especially diabetes mellitus) [33]. A retrospective review of 73 cases at a tertiary care center suggested that patients who present with most of the classic clinical and radiographic criteria and/or who fail to respond to the initial course of intensive therapy have worse outcomes [97]. Individuals noted to have "severe" disease at presentation as manifest by cranial nerve palsy, positive fungal culture, relapse, and invasion on imaging also have a worse treatment outcome [98].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute otitis media, otitis media with effusion, and external otitis".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)

●Basics topic (see "Patient education: Outer ear infection (The Basics)")

●Beyond the Basics topic (see "Patient education: External otitis (including swimmer's ear) (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Definition – Necrotizing external otitis (NEO), previously referred to as malignant otitis externa, is a rare invasive ear infection that can lead to skull base osteomyelitis if not promptly treated. It typically occurs in older adults with diabetes mellitus. (See 'Introduction' above and 'Nomenclature and classification' above and 'Epidemiology' above.)

●Microbiology − NEO is caused by Pseudomonas aeruginosa in more than 95 percent of cases. Since P. aeruginosa is not a normal component of ear canal flora even in diabetic patients, its recovery indicates the presence of a pathogen. Other organisms that can cause NEO include Aspergillus species, Staphylococcus aureus, Proteus mirabilis, Klebsiella oxytoca, Burkholderia cepacia, and Candida parapsilosis. Infection with these other organisms characteristically occurs in immunocompromised hosts, such as those with HIV or cancer. (See 'Microbiology' above.)

●Clinical presentation − Patients with NEO classically present with deep ear pain and otorrhea, which are not responsive to topical measures used to treat simple external otitis. The pain is generally more severe than that found in simple external otitis, although the two may be difficult to distinguish in their early stages. On physical examination, the external auditory canal appears swollen and inflamed and has a purulent discharge. (See 'Clinical manifestations' above.)

●Diagnosis

•NEO should be suspected in a patient presenting with otorrhea (especially at night) and otalgia that has not responded to topical antibiotics. For patients suspected to have NEO, we obtain blood cultures and erythrocyte sedimentation rate (ESR) and/or C-reactive protein (CRP; along with other routine blood work), swab collection of the ear canal drainage, and computerized tomography (CT) imaging. (See 'Approach to diagnosis' above.)

•In the absence of a single pathognomonic criterion, the diagnosis of NEO is based upon a constellation of clinical (eg, otorrhea, otalgia, granulation tissue in the external ear canal) and laboratory (elevated ESR and/or CRP) findings. Radiographic findings (presence of bone erosion on CT or MRI) are helpful for determining the extent of disease. Many patients present with early disease involving only the soft tissues below the ear canal and that has not yet affected the bone; in such cases, imaging may not show any bony abnormalities (although MRI typically shows inflammation in the soft tissues). (See 'Establishing the diagnosis' above.)

•The lack of growth on bacterial and/or fungal culture does not exclude an infection and does not form the basis of establishing a diagnosis. (See 'Establishing the diagnosis' above.)

●Treatment

•Empiric therapy

-Select immunocompetent patients – For immunocompetent patients with very early, uncomplicated NEO (no skull base osteomyelitis, cranial nerve palsies, or meningitis) in settings with low rates of P. aeruginosa resistance to ciprofloxacin (<10-15 percent of isolates resistant to ciprofloxacin), we suggest oral ciprofloxacin (Grade 2C). (See 'Oral antibiotics for select immunocompetent outpatients' above.)

-All other patients – For immunocompromised patients, immunocompetent patients who do not meet the criteria for or are not improving on oral ciprofloxacin, or in settings with high rates of P. aeruginosa resistance to ciprofloxacin (>10-15 percent of isolates resistant to ciprofloxacin), we suggest IV combination therapy with ciprofloxacin and an antipseudomonal beta-lactam (eg, cefepime, ceftazidime, piperacillin-tazobactam) (table 1) (Grade 2C). (See 'IV antibiotics for all other patients' above.)

•Reassessment after 48 to 72 hours of empiric therapy – We reassess all patients with NEO 48 to 72 hours after the initiation of empiric therapy. Assessment consists of evaluating the patient's clinical response as well as reviewing microbiologic results. Those who are improving should continue the same antibiotic regimen unless microbiologic culture results allow for tailoring of therapy. (See 'Pathogen identified and clinically improving' above and 'No pathogen identified and clinically improving' above.)

For patients who do not improve on initial combination IV antibiotic therapy, we suggest broadening the antibiotics to include activity against methicillin-resistant Staphylococcus aureus (MRSA) and multi-drug resistant P. aeruginosa (eg, vancomycin and meropenem) (Grade 2C).

If ear swab culture results are negative and the patient does not improve on broadened empiric therapy (eg, persistent symptoms or persistently elevated inflammatory markers) or if the patient is not improving despite appropriate antibiotic therapy based on ear swab culture results, we suggest initiating antifungal therapy with voriconazole and obtaining a biopsy to send tissue for culture, susceptibility testing, and pathology (to rule out squamous cell carcinoma) (Grade 2C). (See 'No clinical improvement after 48 to 72 hours' above.)

•Duration of therapy − The duration of antibiotic treatment is generally six to eight weeks, as indicated for osteomyelitis. Fungal NEO generally requires a longer duration of >12 weeks. (See 'Duration and follow up' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Victor L Yu, MD, who contributed to earlier versions of this topic review.