AM ER IC AN JOURNAL OF OT OLARYNGOLOGYH E A D A N D NE CK M E D IC IN E A ND S U RGE RY 3 7 (2 0 1 6) 4 254 3 0

Available online at www.sciencedirect.com

ScienceDirect
www.elsevier.com/locate/amjoto

Malignant external otitis: Factors predicting patient

outcomes,,

Sagit Stern Shavit, MD a, b,, Ethan Soudry, MD a, b ,

Yaniv Hamzany, MD a, b , Ben Nageris, MD b, c,
a
Department of Otolaryngology-Head and Neck Surgery, Rabin Medical Center, Petach Tikva, Israel
b
Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
c
Department of Otolaryngology-Head and Neck Surgery, Meir Medical Center, Kfar Saba, Israel

ARTI CLE I NFO A BS TRACT

Article history: Objective: Malignant external otitis (MEO) is an aggressive infection, primarily affecting
Received 23 February 2016 elderly diabetic patients. It begins in the external ear canal and spreads to adjacent
structures. This study investigated the clinical characteristics of patients diagnosed and
treated for MEO and analyzed factors affecting patient outcomes.
Study design: Historical cohort.
Setting: Tertiary medical center.
Methods: Medical records of all patients diagnosed and treated for MEO from 1990 to 2013, were
retrospectively reviewed. Clinical features, laboratory, imaging and outcomes were analyzed.
Results: 88 patients were included, mean age was 73 11.5 years, 61 (69%) were male. Of these, 75%
had diabetes. Mean follow-up was 60 months. The most common presenting symptoms were otalgia
(89%), external ear canal edema (86%) and otorrhea (84%). Pseudomonas aeruginosa was isolated in 61% of
ear cultures. All patients were treated with antibiotics, 22% had surgery and 8% hyperbaric oxygen.
Overall survival rate was 38% in 5 years, with disease specific mortality 14%. DM, facial nerve palsy,
positive CT scan and age above 70 were found to correlate and predict disease-specific mortality.
Conclusions: MEO carries a grave prognosis. The presence of two or more of the following
features, DM, facial nerve palsy, positive CT scan and age above 70, predicts poor outcome,
and highlights the need for prolonged, vigorous treatment.
2016 Elsevier Inc. All rights reserved.

1. Introduction 1968 due to the high, disease-specific mortality rate of over

50% prior to the introduction of appropriate antibiotic
Malignant (necrotizing) external otitis (MEO) is an aggressive treatment [1]. MEO predominantly affects elderly patients
infection. It begins in the external ear canal and spreads to and those with diabetes, although it has been described
adjacent structures. It was termed malignant by Chandler in among non-diabetics as well [2]. Pseudomonas aeruginosa, the

Author Disclosures.

Sponsorship: None.

Funding source: None.
Correspondence to: S.S. Shavit, Rabin Medical Center, 39 Jabotinsky St., Petach Tikva, Israel 49100. Tel.: +972 3 9376458; fax: +972 3 9376467.
Correspondence to: B. Nageris, Dept. of Otolaryngology-Head and Neck Surgery, Meir Medical Center, 89 Tshernichovsky St., Kfar
Saba, Israel 4428164. Tel.: +972 9 7472553; fax: +972 9 7471291.
E-mail addresses: [email protected] (S.S. Shavit), [email protected] (B. Nageris).

http://dx.doi.org/10.1016/j.amjoto.2016.04.005
0196-0709/ 2016 Elsevier Inc. All rights reserved.
426 AM ER IC AN JOURNAL OF OT OLA RYNGOLOGYH E A D A N D NE CK M E D ICI N E AN D S U RGE RY 3 7 (2 0 1 6) 4 25 4 3 0

most common pathogen known to cause MEO, is responsible All patients who met the diagnostic criteria for MEO (n =
for over 90% of cases. Staphylococcus aureus, Proteus mirabilis 31) were added to the previous cohort of patients diagnosed
and fungi, mostly aspergillus and candida species have also with MEO in our department from 1990 through 2008 (n = 57)
been described [35]. [9]. Diagnosis and treatment for all patients was according to
In 1987, Cohen and Friedman described diagnostic criteria the departmental protocol, which was not altered from 1990
to stratify disease severity. Major criteria include otalgia, till 2013. The following information was collected from the
otorrhea, edema, granulation tissue, positive 99mTc scan, medical records: age, sex, comorbidities, physical examina-
failure to response to treatment after more than a week and P. tion features, treatment, biochemical results, culture and
aeruginosa isolation. Minor criteria are positive radiograph, biopsy results, imaging, follow-up data and prognosis for all
diabetes mellitus, old age and cranial nerve involvement [6]. 88 patients.
MEO is diagnosed if all major criteria are present. If some are
absent, then a 13 week trial of intensive local and systemic 2.1. Statistical analysis
treatment is recommended; failure to respond clarifies the
diagnosis [7]. Treatment includes systemic anti-pseudomonas All statistical analyses were carried out using SAS 9.2
antibiotics for at least 6 weeks and topical therapy. If insufficient software. Continuous variables are presented as mean and
response, modalities such as surgery and hyperbaric oxygen standard deviation. Normally distributed variables were
treatment are considered, as well as additional antibiotic or compared using t-test and categorical data were analyzed
antifungal treatment. using chi-square or Fisher exact test, as appropriate. Kaplan
Due to the difficulty of diagnosis and its low incidence, Meier with log Rank test was performed for univariate
evidence-based knowledge is derived from small case series analysis. Multi-variable and hazard ratios were calculated
or historical cohorts, several of which are based on our using Cox regression and the Fine and Gray model for
institutional experience [2,4,79]. The goals of the current competing risks. Logistic regression was performed to design
study were to evaluate and identify characteristics and a prognostic model. Statistical significance was inferred at
factors predicting outcomes and to design a prognostic tool P < 0.05.
to predict mortality.

3. Results
2. Materials and methods
3.1. Ethical considerations
The computerized database of the Department of
Otolaryngology-Head and Neck Surgery of a university- The study protocol was approved by the local Institutional
affiliated medical center was retrospectively reviewed for all Review Board. Informed consent was not required.
adult patients admitted for MEO between January 2009 and From 2009 through 2013, 31 patients were diagnosed with
December 2013. The medical center is a major referral site for MEO and added to the previous cohort of patients with MEO
MEO. Patients were diagnosed with MEO if presenting with all diagnosed in our department from 1990 through 2008. Overall,
of the following criteria: (1) external otitis with compatible mean follow up was 60.6 54 (range 1268) months. The
physical examination severe otalgia, external ear canal mean age was 73 11.5 years and 61 (69%) were male. Most
edema, exudate and granulations, (2) failure to respond to patients (72%) had 2 or more comorbidities and 9 (10%) had
antibiotic treatment for at least one week, (3) positive finding more than 4. The most common were diabetes mellitus (DM)
on bone scan (technetium with or without gallium) or (66/88, 75%), hypertension (HTN) (64%) and ischemic heart
evidence of bone destruction on CT scan, (4) histological disease (IHD) (34%).
findings compatible with inflammation (patients with
diagnoses such as squamous cell carcinoma of the external 3.2. Physical examination and imaging
ear were excluded), and (5) microabscess in specimens
obtained at surgery, if performed. If one criterion was lacking Among the 88 patients, all but 13 had unilateral MEO, only 3
and disease failed to improve after 1 week of intravenous presented with simultaneous bilateral ear disease. Mean
antibiotics and aggressive local treatment, MEO was diagnosed, duration of otalgia prior to hospitalization was 41 38 days.
as well. All patients failed to improve after at least one week of topical
According to departmental protocol, all patients undergo and oral antibiotics prior to diagnosis. Physical examinations
technetium scan for diagnosis of MEO. Those with a positive signs at presentation included: otalgia in 79 patients (90%),
scan undergo gallium scan to strengthen the diagnosis. CT is edema 76 (86%), otorrhea 74 (84%), granulations 66 (75%),
routinely performed to identify disease spread. All patients facial nerve palsy 15 (17%), bone erosions 4 (4.5%) and trismus
undergo biopsy of the external ear canal for histological 1 (1%).
confirmation of inflammation and to exclude squamous cell Technetium scan was performed on 85/88 patients. Three
carcinoma. Antibiotics are prescribed for at least 6 weeks. patients with aggressive disease died before having the scan;
Follow up examinations, CRP and ESR measurements, and all 3 had a positive CT scan. Almost all technetium scans
gallium scans are repeated as needed until disease resolution. (95%) demonstrated positive findings in the petrous (32%),
Cure from disease is defined as a complete resolution of mastoid (25%), other temporal bone location (28%) or multiple
clinical symptoms, normal sedimentation rate and negative locations. Fifty patients underwent gallium scan to strengthen
gallium scan. the diagnosis, of which 64% had positive gallium absorption.
 AM ER IC AN JOURNAL OF OT OLARYNGOLOGYH E A D A N D NE CK M E D IC IN E A ND S U RGE RY 3 7 (2 0 1 6) 4 254 3 0 427

Localization in the technetium and gallium scans agreed in all Disease recurred in 12 patients (14%) after a mean of 7
but 2 patients. A total of 61 patients underwent CT scan as well, 3.5 months. These patients were treated with a second, 6-week
with 75% demonstrating features compatible with external ear course of antibiotics, most with a combination of ceftazidime
infection and 84% indicating disease spread to adjacent structures: and voriconazole. Additional treatment modalities were used
temporomandibular joint (TMJ) (41%), infratemporal fossa (20%), in 42%, including surgery (20%), hyperbaric oxygen (40%), or
nasopharynx (13%) and skull base (11%). both (40%).
Gallium scan was repeated during patient follow-up in Univariate analysis was performed to identify which
54%. The second scan was negative or demonstrated a significant criteria affect overall patient mortality risk. Granulation during
improvement in 59%. physical examination (P = 0.023), positive CT scan (P = 0.0046),
facial nerve palsy (P = 0.024) and older age (P = 0.014) were
3.3. Bacteriology and treatment significant. (Fig. 1).
Other variables, including sex, other physical examination
Bacterial cultures were taken from all admitted patients; 20/ findings, bacterial cultures, and positive technetium or
88 cultures were negative. There were a total of 78 positive gallium scans were not found to affect the prognosis. No
bacterial or fungal isolates. Among these, 68 patients had a significant relationship was found between persistent disease,
single organism and 10 had 2 or more. Bacterial and fungal recurrence and treatment modalities and mortality. Compari-
species are elaborated in Table 1. Fungal infection was found son of prognoses between the populations followed from 2009
in 12, of which 7 also had bacterial infection and one had both through 2013 and the previous cohort from 1990 through 2008
aspergillus and candida growth. demonstrated no difference in disease specific mortality
All patients received antibiotic treatment, 73% intrave- between the two cohorts.
nously, mostly with ceftazidime. High-dose oral quinolones Diabetes was a risk factor for overall mortality. However, it
were given to 27%. Five patients with fungal isolates only in was an extremely strong risk factor for disease-specific
repeat cultures received voriconazole in addition to antibiotic mortality (P < 0.0001). Disease distribution according to CT
therapy. Thirteen patients received 2 and 2 were treated by a scan (anterior invasion to TMJ, nasopharynx, infratemporal
3-drug cocktail. Mean treatment duration was 47 27 days, fossa, posterior invasion to mastoid or invasion to skull base)
maximum 180 days. was not related to prognosis.
Surgery was a second treatment modality for 20 (23%) A multivariate analysis for disease-specific mortality
patients, of whom 12 had external canal debridement only. demonstrated that DM, positive CT scan, facial nerve palsy
Mastoidectomy was performed in 8 patients, of whom 4 had and older age increased the hazard ratio (Table 2). A regression
canal wall-down procedure and 4 canal wall-up. Hyperbaric model using these four variables was built to predict mortality
oxygen was administered as an additional modality to 7 patients, within a year of diagnosis in order to create a prognostic scale.
4 after surgery and 3 while receiving antibiotics. Each variable received one point. Three or more points predicted
an 8-fold increased risk of mortality (range 227) with 75%
3.4. Prognosis and risk factors sensitivity and 72% specificity.

The survival rate after a mean follow-up of 61 months was

38%. There was 14% disease-specific mortality and 48%
mortality from other causes. Mean time-to-death from MEO 4. Discussion
diagnosis was 9.5 13 months, while time to death from
other causes was 56.5 46 months (P = 0.001). Malignant external otitis is a rare disease affecting elderly
Disease persisted beyond the initial 6 week treatment diabetic patients and caries a poor prognosis. Since it was
period in 24 patients (27%), who required additional antibiotic described by Chandler in 1968, data regarding disease
treatment for an average of 63 80 days. A third of these were prevalence, diagnosis and outcomes have been derived from
treated with multiple modalities, including surgery (37.5%), small case series and historical cohorts. Some of the largest
hyperbaric oxygen (25%) or both (37.5%). series were described among Israeli patients [2,610]. This
finding is attributed to the warm, humid climate.
Many studies include the definition and diagnosis of MEO,
Table 1 Bacterial and fungal species isolated from ear as well as bacterial and imaging findings. Since the numbers
cultures. are small, only a few were able to identify factors affecting
patient outcomes. Stevens et al. published a protocol that
Bacteria/Fungus Positive cultures N (%)
stratified MEO into severe and non-severe based on a
Pseudomonas aeruginosa 39 (50) literature review. Severe MEO was defined by clinical features
Enterobacteriace 18 (23) such as seven nerve palsy, fungal culture, disease relapse and
Candida 10 (13)
requirement of surgerysurgical intervention, and imaging
Staphylococcus aureus 7 (9)
features such as TMJ, ITF, Tegmen, NP and intracranial
Aspergillus 2 (3)
Streptococcus 1 (1.3) involvement. In a series of 28 patients, those with two or
Coagulase negative 1 (1.3) more positive criteria, had a worse prognosis [11].
Normal flora/Negative 20 The current study aimed to identify factors predicting patient
Based on a total of 78 positive isolates among 88 patients. outcomes in the largest cohort in the literature and includes our
total institutional experience since 1990, describing 88 patients.
428 AM ER IC AN JOURNAL OF OT OLA RYNGOLOGYH E A D A N D NE CK M E D ICI N E AN D S U RGE RY 3 7 (2 0 1 6) 4 25 4 3 0

Fig. 1 KaplanMeier curves reporting follow up (FU) in days, from diagnosis, demonstrate increased mortality related to (A)
granulation tissue upon examination, (B) positive CT scan, (C) CN7 nerve palsy and (D) age above 70 at presentation.

They were diagnosed with Type 1 MEO according to Cohen and Multivariate analysis was conducted to identify which
Friedman criteria [6]. All patients were elderly, most with DM diagnostic criteria affect patient risk for disease-specific
(75%) and positive ear culture for P. aeruginosa (50%). These mortality. Age above 70 years, DM, facial nerve palsy and
findings are similar to reports from other countries, including positive findings on CT scan at diagnosis were significant
USA, UK, Australia, and Singapore among others [1215]. factors. Persistent or recurrent disease was not found to
By the 5-year follow-up, 14% of the 88 patients had died increase mortality, compatible with a previous report from
from the disease within a year and an additional 48% died our institution [9].
from unrelated causes within 4.5 years of diagnosis. Initial
reports of MEO mortality prior to the advent of anti-pseudomonas 4.1. Older age
antibiotics were 50% [1]. Despite a clear decline in disease-specific
mortality, MEO has high mortality for an infectious disease. The Patients older than 70 had up to 11-fold higher risk for
overall mortality of 62% is not surprising taking into account mortality; comparable with previous findings [9]. Older
patients' ages and comorbidities. patients are at higher risk for all cause mortality and for
disease-specific mortality. In Israel, MEO in patients over
70 years of age increases the mortality rate by a factor of 6
(5.47% vs. 32.88%) [16]. Franco-Vidal et al. in an analysis of 46
patients and Loh and colleagues in an analysis of 19, did not
Table 2 Relation of prognostic factors to disease specific
find that age affected patient outcomes [14,17].
mortality.
Odds 95% Wald p 4.2. Diabetes mellitus
ratio value
confidence limits
Early publications suggested that MEO occurs exclusively in
Positive CT scan 4.5 1.513.6 <.0001
patients with DM [1]. Since then, several studies have reported
Diabetes mellitus 1.6 0.475.4 0.009
MEO in non-diabetic patients, as well [2,1114,18,19]. Contem-
Facial nerve palsy 3.3 110.6 <.0001
Old age 11 1.590 0.0021 porary studies diagnosed MEO in as many as 49% of non-
diabetic patients [13] and in as few as 5% [14]. Similar to other
 AM ER IC AN JOURNAL OF OT OLARYNGOLOGYH E A D A N D NE CK M E D IC IN E A ND S U RGE RY 3 7 (2 0 1 6) 4 254 3 0 429

reports [17,20], 25% of the patients in our series did not have 4.4. Positive CT scan
diabetes.
The classic form of MEO is thought to be due to Many imaging methods have been suggested for diagnosing
P. aeruginosa infection in diabetic patients. In a series of 20 and following MEO patients. CT scan has been proven as a
patients, Candace et al. demonstrated that fewer MEO useful diagnostic tool. However, its role in routine follow-up
patients with other bacteria (MRSA for example) have and prognosis is debatable. CT-positive features such as bone
diabetes (55% vs. 100%) [15]. Chen et al. described similar erosion, which occurs after bone demineralization, may take
findings [18]. Our results differed in that 75% of patients in months to develop and are irreversible [23].
each group (those with and without positive P. aeruginosa We found that positive CT scan findings in disease
culture) had diabetes. Diabetes predisposes patients to MEO diagnosis, increase mortality within the first year 4.5-fold.
by causing microangiopathy and impaired blood circulation However, no specific area of involvement TMJ, NP, IFT,
in the external auditory canal, hence, interfering with the mastoid or skull base and no anterior or posterior involvement
ability of immune cells to respond to P. aeruginosa and other was a significant cause of increased mortality. Peleg et al.
bacterial infections [1920]. Chin et al. followed 24 patients suggested that MEO patients with severe disease will
with MEO, of which 54% were diabetic, with an average HA1C have involvement of two or more of the following areas in
of 9% (range 7.4%13.4%). They suggested that poor glycemic CT scan TMJ, temporal bone and skull base [10].
control contributes to disease development [12]. However, Chin et al. designed an imaging grading system, where grade I
others failed to demonstrate a relationship between glycemic represents soft tissue involvement, II mastoid, III temporal bone and
control and disease outcomes [14,20]. IV anterior or posterior skull base. They found a positive correlation
Diabetes was found to be a significant risk factor for between hospitalization days and the grading system [12].
disease-specific mortality, since all patients who died from Loh et al. correlated prognosis to minor (EAC tissue
causes related to MEO, within a year had diabetes and it swelling, mastoid involvement) and major (IFT, TMJ and NP
increased mortality by 3-fold. These results can be attributed involvement) CT findings, but found no significant correlation.
to other mortality risk factors, such as decreased renal However, among 5 patients in which imaging demonstrated
function and vasculopathies that are attributed to diabetes clivus involvement, a positive correlation to mortality was seen
and influence patient response and tolerance to treatment. [14]. Sudhoff et al. found no correlation with prognosis [23].
Stevens et al. included involvement of 2 sites on CT scan as a
4.3. Facial nerve palsy designation of severe MEO [11].
We advocate CT scan for all patients when MEO is
Early publications about MEO reported cranial nerve paralysis suspected. Positive CT findings will not only help confirm
at disease presentation, of which the facial nerve is most the diagnosis, but also predict a more severe outcome. In
commonly involved [1,21]. This factor was reported to contrast to others who evaluated CT as prognostic tool, we did
increase mortality by 50% [21]. A previous report from our not find specific areas of involvement correlated with
institution, compared 7 patients with facial paralysis at prognosis and we believe that when any site involve with
disease presentation to 41 patients without paralysis and disease in a CT scan is sufficient to predict worse outcomes.
found no significant difference in survival. More severe Other imaging techniques commonly used in our center are Tc
manifestation was seen on CT in the facial paralysis group and gallium scans. Technetium scan is highly sensitive for diagnosing
[8]. Similarly, Mani et al. found no difference in mortality for 6 MEO, but less specific, since it is also positive for malignancies and
patients with facial nerve paralysis among 23 with MEO. Their trauma. The scan will remain positive until cessation of osteoblastic
results demonstrated that patients with facial nerve palsy did activity, thus its limited role in follow up [6,13]. Gallium scan is the
not regain complete recovery despite medical treatment [22]. imaging of choice for follow-up, since it returns to normal once
Franco-Vidal et al. published their experience of 46 patients inflammation subsides. However, both methods do not provide
and showed that facial paralysis presented in 20% and was a adequate anatomical reference for disease spread and neither was
significant risk factor for mortality [17]. Loh et al. showed a found to correlate with disease severity in our results. MRI was shown
similar trend in 19 patients 4 of whom presented with palsy to be useful for soft tissue involvement; [10,24] but it is less accessible
[14]. According to the classification of Stevens et al., facial than other imaging modalities in our center.
nerve palsy defines a severe form of MEO and was correlated Recently, other advanced nuclear imaging techniques have
to increased mortality [11]. been suggested [25]. These methods combine physiological as well
Among the 88 patients with MEO in our center, 15 as anatomical views of the infection and will hopefully provide a
presented with facial nerve paralysis, of which 5 died from single, superior method for the diagnosis and follow up of MEO.
the disease (42% disease-specific mortality). Facial nerve This study is limited by its retrospective design, relatively
paralysis was related to a 3-fold increased risk of mortality small sample, and restriction to single-center experience.
in the first year compared to patients with no paralysis.
The facial nerve is in close proximity to the external
auditory canal. Hence, it is the first cranial nerve to be 5. Conclusion
affected by the disease. It is logical to assume that the
aggressive inflammatory features of MEO can penetrate to 5.1. Proposed prognostic scale
the temporal bone and other adjacent nerves. Small sample
sizes might have prevented previous series from finding a We propose the following model as a prognostic scale for
significant correlation. disease-specific mortality. It is a simple scale based on clinical
430 AM ER IC AN JOURNAL OF OT OLA RYNGOLOGYH E A D A N D NE CK M E D ICI N E AN D S U RGE RY 3 7 (2 0 1 6) 4 25 4 3 0

and imaging information available from patients diagnosis to [9] Soudry E, Hamzany Y, Preis M, et al. Malignant external otitis:
evaluate prognosis directly from patients presentation and Analysis of severe cases. Otolaryngol Head Neck Surg 2011;
144:75862.
may help in determining treatment aggressiveness and
[10] Peleg U, Perez R, Raveh D, et al. Stratification for malignant
patient and family education. Each of the following variables
external otitis. Otolaryngol Head Neck Surg 2007;137:3015.
has a disease-specific mortality risk and will receive one [11] Stevens SM, Lambert PR. Malignant otitis externa: A novel
point: age above 70 years, DM, facial nerve palsy, and positive stratification protocol for predicting treatment outcomes.
CT scan findings. Our results demonstrate that having more Otol Neurotol 2015;36(9):14928.
than 2 points increases mortality in the first year, 8-fold [12] Chin R, Roche P, Sigston E, et al. Malignant otitis externa:
compared to having 2 points or fewer. Larger, prospective An Australian case series. Surgeon 2010;10:2737.
[13] Ali T, Meade K, Anari S, et al. Malignant otitis externa: Case
studies are needed to improve understanding of this devas-
series. J Laryngol Otol 2010;124:84651.
tating disease. [14] Loh S, Loh WS. Malignant otitis externa: An Asian perspective
on treatment outcomes and prognostic factors. Otolaryngol
Head Neck Surg 2013;148:9916.
Author contributions [15] Hobson CE, Moy JD, Byers KE, et al. Malignant otitis externa:
Evolving pathogens and implications for diagnosis and
treatment. Otolaryngol Head Neck Surg 2014;151:1126.
Sagit Stern - conception and design, acquisition and analysis
[16] Central bureau of statistics. Statistical Abstract of Israel;
of data drafted the article. 2014http://www.cbs.gov.il.
Yaniv Hamzany - conception and design, acquisition of [17] Franco-Vidal V, Blanchet H, Bebear C, et al. Necrotizing
data, drafted the article. external otitis: A report of 46 cases. Otol Neurotol 2007;28:
Ethan Soudry - conception and design, acquisition of data, 7713.
drafted the article, [18] Chen JC, Yeh CF, Shiao AS, et al. Temporal bone
osteomyelitis: The relationship with malignant otitis
Ben Nageris - conception and design, acquisition and analysis
externa, the diagnostic dilemma, and changing trends.
of data, drafted the article. ScientificWorldJournal 2014;2014:591714. http://dx.doi.org/
All authors gave final approval of version to be published. 10.1155/2014/591714.
[19] Geerlings SE, Hoepelman AI. Immune dysfunction in patients
with diabetes mellitus (DM). FEMS Immunol Med Microbiol
REFERENCES 1999;26:25965.
[20] Rubin J, Yu VL. Malignant external otitis: Insights into
pathogenesis, clinical manifestations, diagnosis and therapy.
[1] Chandler JR. Malignant otitis externa. Laryngoscope 1968;78: Am J Med 1988;85:3918.
125794. [21] Chandler JR. Pathogenesis and treatment of facial paralysis
[2] Shpitzer T, Stern Y, Cohen O, et al. Malignant otitis externa in due to malignant external otitis. Ann Otol Rhinol Laryngol
nondiabetic patients. Ann Otol Rhinol Laryngol 1993;102: 1972;81:64858.
8702. [22] Mani N, Sudhoff H, Rajagopal S, et al. Cranial nerve
[3] Bayardelle P, Jolivet-Granger M, Larochelle D. Staphylococcal involvement in malignant external otitis: Implications for
malignant otitis externa. Can Med Assoc J 1982;126:1556. clinical outcome. Laryngoscope 2007;117:90710.
[4] Hamzany Y, Soudry E, Preis M, et al. Fungal malignant [23] Sudhoff H, Rajagopal S, Mani N, et al. Usefulness of CT scans
external otitis. J Infect 2011;62:22631. in malignant external otitis: Effective tool for the diagnosis,
[5] Rubin Grandis J, Branstetter 4th BF, Yu VL. The changing face but of limited value in predicting outcome. Eur Arch
of malignant (necrotising) external otitis: Clinical, radiological, Otorhinolaryngol 2008;265:536.
and anatomic correlations. Lancet Infect Dis 2004;4:349. [24] Ismail H, Hellier WP, Batty V. Use of magnetic resonance
[6] Cohen D, Friedman P. The diagnostic criteria of malignant imaging as the primary imaging modality in the diagnosis
external otitis. J Laryngol Otol 1987;101:21621. and follow-up of malignant external otitis. J Laryngol Otol
[7] Joshua BZ, Sulkes J, Raveh E, et al. Predicting outcome of 2004;118:5769.
malignant external otitis. Otol Neurotol 2008;29:33943. [25] Chen YH, Hsieh HJ. Single photon emission computed
[8] Soudry E, Joshua BZ, Sulkes J, et al. Characteristics and tomography/computed tomography for malignant otitis
prognosis of malignant external otitis with facial paralysis. externa: Lesion not shown on planar image. Am J Otolaryngol
Arch Otolaryngol Head Neck Surg 2007;133:10024. 2013;34:16971.