Abstract

Fluid signal in the mastoid can be an incidental finding on T2-weighted magnetic resonance imaging and often is interpreted as mastoiditis by radiologists. This study examines 28 consecutive cases of such erroneously diagnosed "mastoiditis" and documents the presence or absence

of otologic symptoms and clinical signs. We found a very low prevalence of otologic symptoms orpathology and no cases of mastoiditis in these patients, and we determined that magnetic resonance imaging is not an effective screening modality for mastoiditis. We also reviewed the literature on current diagnostic criteria for mastoiditis and propose alternative terminology to replace the use of excessively alarming terms such as "mastoiditis" unless they apply to that specific clinical diagnosis.

Introduction

Use of magnetic resonance imaging (MRI),as with many evolving technologies, is increasing in the United States and shows no sign of plateauing. Data from more than 30 million Medicare enrollees showed that MRI use increased 16.6% from 1998 to 2001 in the ambulatory setting alone. (1) Another study of Medicare data found that from 1993 to 1998, head and neck MRI increased 43.6%. (2) Concomitant with the overall increase in MRI use is an increase in the number of incidental findings.

Increased fluid signal in the mastoid air cells on T2-weighted MRI is sometimes interpreted as mastoiditis, and otherwise asymptomatic patients are referred to an otolaryngologist or otologist/neurotologist. Many times, patients have been told they have mastoiditis but have received little additional counseling. Imaging can be important in diagnosing middle ear and mastoid disease and becomes more critical as the disease becomes more complex. Imaging is warranted when a complicated case of mastoiditis is suspected, (3) but it is unnecessary in the diagnosis of mastoiditis, which can be established through clinical evaluation.

Computed tomography (CT) is the study of choice in diagnosing mastoiditis and also may be necessary in planning surgery for complicated middle ear infection. CT's sensitivity for mastoiditis is approximately 97%. (4) The hallmark finding is osseous destruction of the mastoid air cells. Opacification of the mastoid air cells is nonspecific and not indicative of acute mastoiditis but, rather, is often seen with acute otitis media (AOM), along with preservation of the ossicular chain, trabeculae, and cortical bone. CT's imaging of the bony components involved in mastoiditis is superior to MRI's, and CT is less expensive, faster, and more widely available. MRI might be necessary when CT is inconclusive for intracranial complications but is, in general, rarely indicated for evaluating possible mastoiditis. (5)

This study investigates clinical scenarios in which the normal sequence is inverted: when radiologic study takes place first and diagnosis second, in the absence of history taking or physical examination. We delineate the significance of bright signal found in the mastoid on T2-weighted MRI in consecutive patients referred for otologist/neurotologist evaluation at an academic medical center.

Patients and methods

After approval by the Institutional Review Board, we reviewed MRIs, radiographic reports, and medical records of 28 consecutive patients seen by the senior author from March 2004 to December 2006 at the Department of Otolaryngology, University of South Florida College of Medicine. Sixteen of the patients had been referred for evaluation of mastoiditis or fluid seen on MRI, and 12 were either established patients or had undergone MRI at our request for other noninfectious indications.

All MRIs were of the brain, with and without gadolinium contrast, and some studies focused on the internal auditory canals or pituitary gland. Indications for MRI in the 28 patients were: hearing loss in 8; headache in 5; vertigo/dizziness in 5; follow-up evaluation of a brain lesion in 3; otalgia in 3; and vision loss, cerebral palsy, jaw pain, and cerebrovascular accident in 1 each.

MRIs had been performed at 10 different facilities and read by a total of 12 radiologists. One facility accounted for 15 (53.6%) of the reports reviewed. Reports did not always specify which imaging sequence was used to generate a diagnosis, but we confirmed the presence of bright signal as reported by the interpreting radiologists in all cases.

As recorded in the medical records, patients presenting with an MRI diagnosis of mastoiditis were asked about symptoms consistent with mastoiditis surrounding the time MRI was performed. Ear-specific medical history was noted, and results of thorough head and neck examination, performed on each patient, were recorded.

Established patients had a documented history and had received head and neck examination prior to MRI. In patients who were seen by interpreting radiologists on the day of their MRI, findings of bright signal were noted by the senior author. Patients were enrolled in our study only if the interpreting radiologist recorded fluid signal in the mastoids in the subsequent note. These radiologists were blinded to the results of the same-day examination. Age, sex, interval between MRI and examination, examination findings, otology-related diagnoses, and treatments were recorded for each patient (table).

Results

The age range of the cohort of 28 patients was 23 to 80 years (mean, 57.4 years). The female-to-male ratio was 15:13 (54% to 46%). Anamnesis surrounding the time of the MRI revealed that 4 patients (14.3%) had symptoms with a clear connection to the MRI findings. Only 2 of them had physical examination findings consistent with fluid in the middle ear.

The range between MRI and patient examination was 0 to 161 days (average, 51 days). A diagnosis that could be consistent with MRI fluid signal in mastoid air cells was found in 9 patients (32%). In 7 of them, the diagnosis was eustachian tube dysfunction (ETD), which was the most common diagnosis that we related to the finding of bright signal in the mastoid on T2-weighted MRI. Mastoiditis was not diagnosed in any patients.

We also evaluated the terminology the radiologists used in their reports. In 26 (92.9%) of the reports, the term "mastoiditis" or "inflammation" was used when describing the MRI findings. The other 2 reports used the term "fluid" and "increased signal." Eleven (39.3%) reports gave a diagnosis of right mastoiditis, 7 (25%) of left mastoiditis, and 7 (25%) of bilateral mastoiditis. Three (10.7%) reports did not specify a side.

Two patients (7%) had a clearlyvisible, ongoing pathologic process underlying findings that were identifiable on MRI. One of these patients required myringotomy for chronic otitis media with effusion, and the other had acute serous otitis, which was managed expectantly and resolved.

Discussion

MRI is highly sensitive to varying phases, making it an ideal modality for imaging soft tissues. Areas of the body composed of tissues with a high hydrogen content, such as that of water, tend to give off more signals. Pathologic conditions--such as neoplasia, inflammation, and edema--tend to produce more signals because of the increase in free water compared with surrounding tissue. (6)

Because bright signal with increased water content is revealed on T2-weighted images, it can theoretically be a sensitive indicator for the presence of edematous mucosa. Air appears black on these images, so an aerated and well-pneumatized temporal bone should appear mostly black. The presence of edematous mucosa within a well-pneumatized temporal bone, therefore, becomes starkly prominent against a dark background.

The prevalence of abnormal temporal bone findings on MRI in asymptomatic patients is low. In compiling a database of normal MRI results, Katzman et al found only 4 temporal bone abnormalities in 1,000 asymptomatic volunteers/Similarly, in a series of 2,700 consecutive temporomandibular joint MRIs, Orhan et al found radiographic evidence of otomastoiditis in only 10 patients, 7 of whom (prevalence, 0.39%) were found on subsequent otolaryngologic examination to have inflammatory disease of the ear. (8) In a series of 644 screening MRIs to rule out retrocochlear hearing loss, Mirza et al reported a 5% incidence of middle ear/ mastoid disease. (9)

Findings in study patients. Our 28 patients were clearly outside the pediatric age group, in which AOM and mastoiditis are typically found. Most had no physical findings suggestive of mastoiditis. Two patients had otitis media. In one, acute serous otitis media, which resolved spontaneously without intervention, involved the left side, but the MRI report did not specify the side of what was termed "mastoiditis." In the other patient, right chronic otitis media with effusion was found, which the MRI report described as "very extensive right mastoiditis, less extensive left mastoiditis" (figure 1).

The increased signal was explainable in 12 patients: 9 had either ETD or effusion, 1 had metastatic breast cancer, and 1 had a history of recent endolymphatic sac surgery. In the 12th patient, a bright signal in the right mastoid was noted as "right mastoiditis" on T1-weighted, T2-weighted FLAIR (fluid-attenuated inversion recovery), and FIESTA (fast imaging employing steady-state acquisition) MRI images (figure 2,A). The signal's source, as revealed on follow-up CT, was bone marrow from an unpneumatized mastoid (figure 2, B).

The vast majority of patients in our series had incidental, nonspecific MRI findings that did not require intervention. Our conjecture is that ETD was the culprit in many of these cases, such as the patient in figure 3, who was seen on the day of his MRI. Seven of our patients (25%) had symptoms consistent with ETD.

Potential conclusions. Our findings raise the question of whether fluctuating middle ear pressure in patients with ETD is capable of inducing changes on MRI. A study by Swarts et al provides excellent evidence of the sensitivity of MRI in detecting middle ear underpressures. (10) To study MRI responses to middle ear pressure changes, the researchers introduced carbon dioxide through the eustachian tube into the middle ear space in cynomolgus monkeys. Since carbon dioxide is a rapidly diffusing gas, the result was underpressure within the middle ear space. After 1 hour of gas stabilization, bright signal on T2 images emerged in the mastoid air cells. This experimentally observed phenomenon underlies our theory that in some patients, the otherwise false-positive T2 MRI signal is attributable to negative middle ear pressure.

A noteworthy finding is the normal physical examination results in the five patients seen immediately after MRI. One of them, examined on the same day as his MRI, had undergone labyrinthectomy on the side ipsilateral to reported findings of mastoiditis. The remaining four had no recent or ongoing history consistent with ear infection or eustachian tube dysfunction.

Our findings reinforce our belief that MRI is undesirable as a screening study for mastoiditis. Not only is it expensive and of limited accessibility, other methods, including history taking and physical examination, are superior. Incidental MRI findings should be interpreted in the context of patient presentation. Evaluation by an otolaryngologist should be reserved for signs of mastoiditis, including indications of AOM with postauricular swelling and erythema.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Another undesirable feature sometimes found on MRI reports is the use of such terms as "inflammation of the mastoid" or "mastoiditis" to describe increased fluid signal within the mastoid. These terms are alarming to many primary care physicians and patients and can lead to unnecessary medical treatment with broad-spectrum antibiotics. Patient interests are better served by appropriate and timely specialist referral.

[FIGURE 3 OMITTED]

References

(1.) Bhargavan M, Sunshine JH. Utilization of radiology services in the United States: Levels and trends in modalities, regions, and populations. Radiology 2005;234(3):824-32.

(2.) Rao VM, Parker L, Levin DC, et al. Use trends and geographic variation in neuroimaging: Nationwide Medicare data for 1993 and 1998. Am J Neuroradiol 2001;22(9):1643-9.

(3.) Maroldi R, Farina D, Palvarini L, et al. Computed tomography and magnetic resonance imaging of pathologic conditions of the middle ear. Eur J Radiol 2001;40(2):78-93.

(4.) Migirov L. Computed tomographic versus surgical findings in complicated acute otomastoiditis. Ann Otol Rhinol Laryngol 2003; 112(8):675-7.

(5.) Vazquez E, Castellote A, Piqueras J, et al. Imaging of complications of acute mastoiditis in children. Radiographics 2003;23(2):359-72.

(6.) Aygun N, Oliverio PJ, Zinreich SJ. Overview of diagnostic imaging of the head and neck. In: Cummings CW, Flint PW, Haughey BH, et al, eds. Otolaryngology Head and Neck Surgery, 4th ed. St. Louis: Mosby; 2005:29-35.

(7.) Katzman G, Dagher A, Patronas NJ. Incidental findings on brain magnetic resonance imaging from 1000 asymptomatic volunteers. JAMA 1999;282(1):36-9.

(8.) Orhan K, Nishiyama H, Tadashi S, et al. MR of 2270 TMJs: Prevalence of radiographic presence of otomastoiditis in temporomandibular joint disorders. Eur J Radiol2005;55(1):102-7.

(9.) Mirza S, Malik T, Ahmed A, et al. Incidental findings on magnetic resonance imaging screening for cerebellopontine angle tumours. J Laryngol Otol 2000;114(10):750-4.

(10.) Swarts JD, Alper CM, Chan KH, et al. In vivo observation with magnetic resonance imaging of middle ear effusion in response to experimental underpressures. Ann Otol Rhinol Laryngol 1995; 104(7):522-8.

James R. Meredith, MD; K. Paul Boyev, MD

From the Division of Otology/Neurotology, Department of Otolaryngology-Head and Neck Surgery, University of South Florida College of Medicine, Tampa.

Corresponding author: K. Paul Boyev, MD, Division of Otology/Neurotology, Department of Otolaryngology-Head and Neck Surgery, University of South Florida College of Medicine, MDC 73, 12901 Bruce B. Downs Blvd., Tampa, FL 33612. Phone: (813) 974-6573; e-mail: pboyev@health.usf.edu

Table. Patient characteristics in consecutive cases of
MRI-diagnosed "mastoiditis"

                                  Time
                                between           Ear and
        Age                     MRI and          otoscopic
No.     (yr)        Sex         exam (d)           exam

1        44          M            132             Normal
2        62          M             86             Normal
3        23          F            161             Normal
4        51          M             78             Normal
5        35          F             55             Normal
6        74          M             26             Normal
7        71          F             20             Normal
8        65          M             76             Normal
9        70          M             33             Normal
10       80          M             39            Abnormal
11       54          M             89             Normal
12       54          F            105             Normal
13       53          M             64             Normal
14       72          F             45             Normal
15       40          M             49             Normal
16       58          F             23             Normal
17       69          F            100             Normal
18       53          M              0             Normal
19       76          F              7            Abnormal
20       66          F             16             Normal
21       55          M              0             Normal
22       58          F             87             Normal
23       47          F             13             Normal
24       60          F             75             Normal
25       63          M             45             Normal
26       50          F              0             Normal
27       48          F              0             Normal
28       56          F              0             Normal

                 Pertinent
        Age         ear         Otologic      Comment/
No.     (yr)     history?      treatment?     diagnosis

1        44         No            No          ETD
2        62         No            No          ETD
3        23         No            No
4        51         No            No
5        35         No            No
6        74         No            No          ETD
7        71         No            No
8        65         No            No
9        70         No            No
10       80         Yes           No          Acute SOM
11       54         No            No
12       54         No            No          ETD
13       53         No            No
14       72         No            No
15       40         No            No
16       58         No            No          ETD
17       69         No            No
18       53         Yes           No          History of
                                              labyrinthectomy
19       76         Yes           Yes         Chronic OME
20       66         No            No
21       55         No            No          ETD
22       58         No            No          ETD
23       47         No            No
24       60         No            No
25       63         No            No
26       50         No            No
27       48         No            No
28       56         No            No

Key: ETD = eustachian tube dysfunction; SOM = serous otitis
media; OME = otitis media with effusion.