Eur Arch Otorhinolaryngol (2011) 268:17271734 DOI 10.

1007/s00405-011-1552-8

OTOLOGY

Computed tomography and magnetic resonance imaging ndings before and after treatment of patients with malignant external otitis
Khaled Al-Noury Alsaid Lotfy

Received: 17 November 2010 / Accepted: 14 February 2011 / Published online: 15 March 2011 Springer-Verlag 2011

Abstract The aim of our prospective study is to illustrate the role of computed tomography (CT) and magnetic resonance imaging (MRI) in the diagnosis and delineation of the extent of disease in malignant external otitis (MEO) at presentation and following successful treatment. The study was performed at Tertiary academic referral center. We included 18 patients (16 males and 2 females; age, 4979 years; mean, 65.3 years) with a clinical diagnosis of MEO in this study. CT and MRI were performed at presentation. Follow-up CT and MRI were performed after 6 and 12 months in 12 successfully treated patients. The main outcomes are CT scan and MRI ndings. Results show that all patients were known to have diabetes mellitus. We present the patients demographic and clinical data. All patients received local treatment as well as the intravenous 3rd generation antibiotic cephalosporin for at least 6 weeks. The detailed radiological ndings were reported. In conclusion, MEO is predominantly a disease of diabetic males and MEO can be managed by the control of diabetes, local treatment, and intravenous antibiotics for a long period. CT and MRI play complementary roles in the diagnosis and follow-up of patients with MEO. The extension of the disease before treatment and its disappearance from the involved areas are usually characteristic.

Keywords prognosis

Necrotizing otitis externa Diagnosis and

Introduction Necrotizing (malignant) external otitis is an invasive infection of the external auditory canal (EAC) that can lead to fatal infections of the skull base. Although elderly patients with diabetes remain the most commonly affected population, immuno-suppressed individuals are also susceptible [1]. Chandler rst coined the term malignant external otitis (MEO) in 1968 [2]. The word malignant was chosen to describe the high mortality associated with this disease prior to the availability of effective antibiotic therapy. As this is an infection and not a malignant process, more appropriate pathophysiological terms have been introduced, including necrotizing external otitis or the more inclusive description of skull base osteomyelitis [3]. Patients often experience exquisitely painful otalgia or deep pain that may be worse at night. Symptoms may begin after irrigation or trauma of the EAC [4]. Purulent otorrhea with a swollen and tender EAC are hallmarks of this disease, and granulation tissue is frequently seen in the oor of the canal at the bony-cartilaginous junction [5]. The most frequent pathogen is Pseudomonas aeruginosa [4, 5]. Diagnosis requires the culture of ear secretions and the examination of granulation tissue in culture. Cranial neuropathies are caused by the spread of the infection through the skull base, indicating a poor prognosis. Intracranial complications are the most frequent cause of death. Treatment of MEO includes the correction of immunosuppression (whenever possible), control of diabetes, local treatment of the auditory canal, long-term systemic

K. Al-Noury Department Otolaryngology, King Abdulaziz University, PO Box 80215, Jeddah 21589, Saudi Arabia A. Lotfy Chairman Radiology Department, International Medical Center, Jeddah, Saudi Arabia K. Al-Noury (&) PO Box 35135, Jeddah 21488, Saudi Arabia e-mail: Kalnoury@kau.edu.sa

123

1728

Eur Arch Otorhinolaryngol (2011) 268:17271734

antibiotic therapy, and surgery in selected cases [4]. The prolonged oral administration of quinolone as an antipseudomonal agent is the most frequently used treatment [1]. Early diagnosis can help in the control of this fatal osteomyelitis. The clinical features rarely permit an exact diagnosis of MEO to be made promptly, and it is difcult to distinguish from otitis externa. This explains the frequent delay in diagnosis with respect to the onset of symptoms. A physical examination almost always reveals the presence of non-specic granulation tissue in the oor of the bony-cartilaginous junction of the EAC, while the most common laboratory nding is a raised erythrocyte sedimentation rate (ESR) [4, 6]. Imaging in the form of computed tomography (CT) and/ or magnetic resonance imaging (MRI) has great diagnostic relevance. These techniques are very useful for spatial resolution, while radionuclide scanning and single photon emission CT are superior for detecting early osteitis and monitoring the response to therapy. The aim of this study was to illustrate the role of CT and MRI in the delineation of the extent of MEO at presentation and following successful treatment.

Subjects and methods Following approval from the hospital ethics committee, we enrolled 18 patients (16 males and 2 females; age, 4979 years; mean, 65.3 years) with a clinical diagnosis of MEO in this study. CT and MRI were performed on all patients at presentation. Follow-up CT and MRI were performed after 6 and 12 months on 14 patients who received the proper treatment. Medical data from the patients les were included. CT was performed using a high-resolution technique with 11.5 mm collimation and an intravenous (I.V.) contrast agent. An MRI examination was performed before and after the administration of the I.V. contrast agent.

Facial nerve palsy was present in four patients at presentation and a 5th patient developed it during the course of the disease. Only 2 patients had palsy of the 10th and 11th nerves. None of the patients were HIV positive. The ESR was elevated in 11 patients (61%), leukocytosis was observed in 7 patients (39%), and blood sugar levels were elevated in 14 patients (78%). Pseudomonas aeruginosa was isolated in 12 patients (67%) (Table 1). All patients were treated by the control of their blood sugar levels. Local treatment in the form of frequent local cleansing, insertion of an ear wick, gentamicin eardrops, and debridement of ear canal granulation were achieved. Systemic treatment with the 3rd generation antibiotic cephalosporin (Ceftazidime 2 g IV every 8 h) for at least 6 weeks was guided by the clinical symptoms, ear canal, ESR, and white blood cells. Antibiotic treatment was continued orally for 2 more weeks in four cases and had to be changed by the infectious disease team in three cases to antipseudomonal penicillin (Ticarcillin-clavulanate potassium 3 g IV every 4 h). These patients continued on oral ciprooxacin 500 mg orally every 12 h for 2 more weeks. Follow-up CT and MRI were performed in 12 patients because 4 patients were lost to follow-up (1 patient died from the disease and 1 patient died from another cause). A technetium-99m scan was performed in 6 patients and a gallium-67 scan was performed in 12 patients at
Table 1 Demographic and clinical data Patients No Gender Male Female Clinical details DM Type I Type II History of trauma Severe otalgia Ear discharge 5 13 6 17 16 11 14 13 14 11 7 12 5 2 2 27 73 33 94 88 61 78 72 78 61 39 67 27 11 11 16 2 89 11 %

Results All patients were known to have diabetes mellitus: type 1 was observed in 5 cases (27%) and type 2 in 13 cases (73%). MEO was bilateral in two cases, while recurrence occurred in the same ear in two cases and in the opposite ear in one case. Seventeen patients (94%) presented with severe otalgia and 16 (88%) with otorrhea, and 11 patients (61%) presented with different types of hearing loss. Six patients (33%) had a history of ear trauma prior to infection. Stenosis of the ear canal was present in 14 patients (78%), while granulation was present in 13 patients (72%).

Deafness Swollen canal Granulation High blood sugar ESR Leukocytosis Pseudomonas culture Cranial neuropathy VII CN X CN XI CN

123

Eur Arch Otorhinolaryngol (2011) 268:17271734

1729

presentation and at the follow-up examinations at 6 and 12 months (data not shown) (Tables 2, 3). Abnormalities of the EAC were in the form of soft tissue involvement and bone erosion (Fig. 1). Involvement of the infratemporal area and masticator space was seen in the majority of patients with EAC abnormalities; however, only one patient had infratemporal involvement without apparent abnormalities of the EAC on MRI or CT (Figs. 2, 3). Cortical bone abnormalities in the base of the skull were delineated by CT in eight patients, while MRI was not able to delineate these abnormalities. On the contrary, skull base medullary changes were elicited only by MRI in the form of hyper-intense foci in T2-weighted images (WI) and enhancement after I.V. contrast administration in 12 patients (Fig. 4). Temporomandibular joint (TMJ) involvement was detected in six patients by CT as bone erosion, widening of the joint space, or abnormal soft tissue and in eight patients by MRI as abnormal marrow intensity, abnormal soft

Table 2 Computed tomography (CT) and magnetic resonance imaging (MRI) ndings in 18 patients at presentation CT External auditory canals soft tissue External auditory canals bone erosion Involvement of infra-temporal region Skull base: cortex Skull base: medulla TMJ involvement Mastoid involvement Petrous apex Middle ear cavity Carotid canal Carotid artery occlusion Intracranial extension Nasopharyngeal extension 2 6 10 4 8 8 2 16 4 8 8 12 8 10 8 8 8 2 6 4 14 MRI 10

tissue, abnormal enhancement, and bone destruction (Fig. 5). Mastoid involvement was detected by CT and MRI; CT showed bone erosion and destruction, abnormal pacication of air cells, bony septae, and bone cortex abnormalities (Fig. 6). Abnormalities of mastoid air cells in MRI were detected as bright signals in T2-WI, abnormal soft tissue in T1- or T2-WI, and the presence of enhancements. Petrous apex abnormalities were detected in four patients by CT as an erosion of bones or soft tissue lesions, while for MRI we relied upon abnormal intensity in T2-WI, abnormal inltration of bone marrow, presence of soft tissue, and abnormal enhancements. All or some of these abnormalities were found in eight patients. Middle ear cavity abnormalities were found in eight patients in the form of abnormal opacity in CT and MRI (Fig. 7). The carotid canal areas were eroded in eight patients on CT, and the same patients showed enhanced soft tissues around the carotid canal on MRI (Fig. 8). Only one of these patients had an occlusion of the petrous carotid artery, the cavernous as well as supraclinoid portions, with no abnormal changes in the brain (Fig. 9). Intracranial extensions were best evaluated by MRI coronal post-contrast T1-WI, and they were found in six patients at presentation. These abnormalities were in the form of an extension of the inammatory process throughout the base of the skull or foramen ovale or in the form of meningeal enhancements. Nasopharyngeal abnormalities were seen in two patients by CT and in four patients with MRI (Fig. 10). Two of these four patients required a prolonged course of antibiotics and the other two patients required a change of antibiotic by the infectious disease team. Follow-up CT and MRI were only performed on 12 patients. We noticed that the rst abnormality to disappear was in the soft tissue of the EAC. Infratemporal fossa abnormalities usually persist for a long time and they were detected by MRI in six out of ten patients after 1 year, although the EAC showed no abnormal soft tissues. The erosion of the EAC was not obvious in two patients out of

Table 3 Computed tomography (CT) and magnetic resonance imaging (MRI) ndings in 12 patients at 6 and 12 months after treatment CT Present External auditory canals soft tissue External auditory canals bone erosion Involvement of infra-temporal region Skull base: cortical Skull base: medulla TMJ involvement Mastoid involvement Petrous apex 4 6 4 12 4 4 6 6 months 4 4 2 6 2 4 6 4 6 2 4 2 2 6 6 6 8 6 6 6 8 2 4 2 2 12 months 2 2 10 10 6 MRI Present 8 6 months 2 12 months

123

1730 Fig. 1 Soft tissue enhancement within the external auditory canal. a coronal CT, b axial CT, and c coronal post-contrast T1-Wl

Eur Arch Otorhinolaryngol (2011) 268:17271734

Fig. 2 Involvement of the infratemporal area and masticator space. a Coronal CT with fat plane obliteration (arrow). b Axial post-contrast T1-Wl of the infratemporal space (arrow)

Fig. 3 Coronal post-contrast T1-Wl. Extension of the process to the suprazygomatic masticator space (arrows)

four by CT after 1 year. The changes to the base of the skull bone on MRI (marrow abnormality) disappeared in four out of six patients after 1 year. The cortical changes in

the base of the skull persisted on CT in the six patients examined, but changes to the pattern of involvement also involved the medulla. This change of the pattern was in the

123

Eur Arch Otorhinolaryngol (2011) 268:17271734 Fig. 4 Axial post-contrast T1-Wl with skull base medullary involvement (arrows)

1731

Fig. 5 Temporomandibular joint involvement. a Post-contrast axial MRI T1-Wl with soft tissue enhancement of the right condyle (circle). b Normal left side MRI. c CT bone erosion and widening of the joint space

Fig. 6 Axial CT in two different patients. a Fluid in the mastoid cavity. b bone destruction

form of abnormal sclerosis of the skull base medullary areas. TMJ changes persisted in CT and in four out of six patients on MRI. Mastoid involvement and petrous apex showed improvement by CT and MRI.

Discussion The diagnosis of MEO is based on a combination of clinical ndings, an increased ESR, and radiological evidence

123

1732 Fig. 7 Middle ear opacity. a coronal CT and b coronal MRI T1-Wl

Eur Arch Otorhinolaryngol (2011) 268:17271734

Fig. 8 Axial CT erosion of the right carotid canal and postcontrast MRI T1-Wl soft tissue enhancement around the carotid

of soft tissue damage with or without bone erosion in the EAC and infratemporal fossa [1]. Skull base osteomyelitis secondary to MEO was rst described in 1959 [7]. The extension of the disease process posteriorly to the mastoid, anteriorly into the TMJ, and medially into the middle ear and the petrous apex occur less frequently. Actually, the tympanic membrane is resistant to the extension of the infection and middle ear involvement is considered by most investigators to be secondary to mastoid destruction [8, 9]. Facial nerve palsy is a grave prognostic sign and may occur as a result of mastoid destruction or more commonly as a result of the subtemporal spread of the infection to the stylomastoid foramen [10]. This disease typically extends inferiorly via tiny clefts in the cartilaginous portion of the EAC, the ssures of Santorini, allowing access to the soft tissues beneath the temporal bone around the stylomastoid foramen and the infra-temporal fossa [9]. However, we observed that the cranial nerve improved and was not an indicator of a worse prognosis [1114]. In our study, soft tissue lesions of the EAC were observed in most patients (n = 16/18) at presentation by CT and MRI; the erosion of the canal was best evaluated by CT, while infratemporal and masticator space involvement at presentation were clearly identied by MRI. In our study, the involvement of the skull base was detected by CT in the form of cortical bone erosion. MRI identied medullary

Fig. 9 Axial MRI T1-Wl with inltration of the bone marrow (arrow) compared with the normal (circle)

bony changes; however, it poorly elicited the erosion of the cortical bone. In our study, we observed two forms of mastoid involvement: uid with the preservation of bony septae or the destruction of the bony cortices; CT and MRI were equally able to delineate these abnormalities. We believe that the abnormalities in the middle ear were basically due to mastoid involvement rather than direct extension. However, the presence of effusion in the middle ear due to the distal inammatory process around the Eustachian tube

123

Eur Arch Otorhinolaryngol (2011) 268:17271734 Fig. 10 Axial CT and postcontrast MRI T1-Wl show soft tissue enhancement of the nasophyarynx

1733

may be another factor. The involvement of the petrous apex may be caused by the direct extension of the same process or indirectly through the invasion of the base of the skull. MRI was more sensitive than CT to detect petrous apex abnormalities (n = 8 in MRI and 4 in CT). The involvement of the carotid canal could occur by the same mechanism as the extension through the base of the skull. In our study, CT showed the erosion of the vertical and horizontal portions of the carotid canal in eight patients. In the same patients, MRI showed soft tissue encasement of the distal cervical and petrous portions of the internal auditory canal. MRI detected carotid artery occlusion; although CT suggested the diagnosis initially, MRI delineated the occlusion that involved the cervical, petrous, cavernous, and supraclinoid portions of the artery. However, these patients had no stroke at presentation and brain MRI did not detect any infarction even by uid attenuation inversion recovery and diffusion sequences. We could only detect intracranial extensions by MRI. These abnormalities consisted of an abnormal dural extension, direct extension of the process, or through the foramen ovale. No frank parenchymal lesions or abscess formation were observed. The nasopharyngeal extension of the lesions was detected by MRI, as suggested by Joshua et al. [12, 13] in patients with a signicantly prolonged treatment and reduced survival. In the follow-up evaluations with CT and MRI, we noticed that the rst abnormality to disappear was in the soft tissue of the EAC. The infratemporal fossa abnormalities usually persisted for a long time as determined with MRI after 1 year. The erosion of the EAC was not obvious on CT after 1 year. The changes to the base of skull bone detected by MRI (marrow abnormality) disappeared after 1 year. The cortical changes in the base of the skull persisted, but the pattern of involvement changed to also involve the abnormal sclerosis of the medullary areas in the skull base. These patterns of changes explain the persistent radiological ndings in spite of clinical improvement. They also explain the persistent positive

bone and gallium scans for long periods in some patients. Hence, clinical judgment of the cases is an important part of their management, especially during the period of antibiotic therapy.

Conclusions MEO is predominantly a male disease. The 3rd generation antibiotic cephalosporin is still active for the treatment of MEO, reserving other antibiotics for resistant cases, and there is no role for surgery. CT and MRI play complementary roles in the diagnosis and follow-up of MEO patients. The absence of EAC soft tissue involvement at presentation does not exclude the diagnosis. The extension of the disease before treatment and its disappearance from the involved areas are usually characteristic. Complete occlusion of the internal carotid artery can occur in an extensive infection. Nasopharyngeal involvement requires a prolonged treatment.

Conict of interest The authors declare that they have neither conict of interest nor nancial relationship with the organization that sponsored the research.

References
1. Grandis RJ, Branstetter BF, Yu VL (2004) The changing face of malignant (necrotizing) external otitis: clinical, radiological and anatomic correlation. Lancet Infect Dis 4:3439 2. Chandler J (1968) Malignant external otitis. Laryngoscope 78:1257 3. Hariga I, Mardassi A, Belhaj Younes F et al (2010) Necrotizing otitis externa: 19 cases report. Eur Arch Otorhinolaryngol 267:11931198 4. Handzel O, Halperin D (2003) Necrotizing (malignant) external otitis. Am Fam Physician 68:309312

123

1734 5. Carfrae MJ, Kesser BW (2008) Malignant otitis externa. Otolaryngol Clin North Am 41:537549 6. Sreepada GS, Kwartler JA (2003) Skull base osteomyelitis secondary to malignant otitis externa. Curr Opin Otolaryngol Head Neck Surg 11:316323 7. Arnold R, Baylin G (1959) Destructive lesions of the temporal bone. Laryngoscope 69(6):766788 8. Nadol JB (1980) Histopathology of pseudomonas osteomyelitis of the temporal bone starting as malignant external otitis. Am J Otolaryngol 1:359371 9. Kohut RI, Lindsay JR (1979) Necrotizing (malignant) external otitis: histopathologic processes. Ann Otol Rhinol Laryngol 88:714720 10. Amorosa L, Modugno GC, Pirodda A (1996) Malignant external otitis: review and personal experience. Acta Otolaryngol Suppl 521:316

Eur Arch Otorhinolaryngol (2011) 268:17271734 11. Soudry E, Joshua BZ, Sulkes J, Nageris BI (2007) Characteristics and prognosis of malignant external otitis with facial paralysis. Arch Otolaryngol Head Neck Surg 133(10):10021004 12. Joshua BZ, Sulkes J, Raveh E, Bishara J, Nageris BI (2008) Predicting outcome of malignant external otitis. Otol Neurotol 29:339343 13. Mani N, Sudhoff H, Rajagopal S, Moffat D, Axon PR (2007) Cranial nerve involvement in malignant external otitis: implications for clinical outcome. Laryngoscope 117(5):907910 14. Ali T, Meade K, Anari S, ElBadawey MR, Zammit-Maempel I (2010) Malignant otitis externa: case series. J Laryngol Otol 124:846851

123

Copyright of European Archives of Oto-Rhino-Laryngology is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.