Quantitative Bone and 67Ga Scintigraphy in the Differentiation of Necrotizing External Otitis From Severe External Otitis Nechama
Uri, MD; Sara Gips, MD; Avi Front, MSc; Shlomo Walter Meyer, MD; Ruth Hardoff, MD, DSc
\s=b\ Qualitative technetium Tc 99m bone
scintigraphy using phosphate compounds and gallium 67 scintigraphy were described as a helpful means in diagnosing necrotizing external otitis (NEO). They were, however, claimed to be nonspecific. Quantitative Tc 99m methylene diphosphonate bone scintigraphy and gallium 67 scintigraphy were performed in eight patients with NEO and in 20 patients with severe external otitis, in order to prove usefulness of scintigraphy in the diagnosis of NEO. Ratios of lesion to nonlesion for bone scintigraphy were 1.67 \m=+-\0.16 in patients with NEO and 1.08 \m=+-\0.09 in patients with severe external otitis, and for gallium 67 scintigraphy they were 1.35 \m=+-\0.24 in NEO patients and 1.05 \m=+-\0.03 in patients with severe external otitis. There was no difference in uptake between diabetic patients with severe external otitis and nondiabetic patients. The scintigraphic studies were also evaluated using a qualitative scoring method (scores 0 to +4), according to the intensity of the radiopharmaceutical uptake. This method was found to be inferior in the diagnosis of NEO compared with the quantitative method. We conclude that lesion-to-nonlesion ratios greater than 1.5 and 1.3 on bone and gallium 67
are indicative of NEO. Quantitative bone scintigraphy, which is quicker to perform, may be used as a single imaging modality for the diagnosis of NEO. (Arch Otolaryngol Head Neck Surg.
external otitis (NEO) described by Meltzer and Keleman.' This disease has been characterized by a stubborn Pseudo¬ monas external otitis in elderly dia¬ betic patients, with a lesion of fleshygranulation tissue on the floor of the external ear canal at the osseous-car¬ tilaginous junction appearing in the majority of patients. This entity has been known for the last 30 years; how¬ ever, changes in the diagnostic, thera¬ peutic, and prognostic aspects of this illness have been recently described.2 Among these changes is the observa¬ tion that a group of elderly diabetic patients who present with all the above cliical signs may have a benign course of the disease, since they do not suffer from temporal bone involve¬ ment and improve after a limited top¬ ical or oral antimicrobial therapy. This group of patients can be classified as patients with severe external otitis was
(SEO). Qualitative bone scintigraphy has
Accepted for publication September 4, 1990. From the Departments of Otolaryngology (Drs Uri and Meyer) and Nuclear Medicine (Drs Gips, Front, and Hardoff), Lady Davis Carmel Hospital, Haifa, Israel. Reprint requests to Department of Nuclear Medicine, Lady Davis Carmel Hospital, 7 Michal St, Haifa 34362, Israel (Dr Hardoff).
been used as a common tool for detect¬ ing temporal bone involvement. How¬ a
study,3 using a qualita¬
tive scoring method, has questioned its utility by showing that an increased uptake in the temporal bone can also be found in SEO. The objective of this study was to find out whether a quantitative ap¬ proach using technetium (Tc) 99m méthylène diphosphonate (MDP) and gallium 67 (67Ga) and calculating the ratio of the affected and nonaffected temporal bone (lesion-to-nonlesion ratio)4 might be used as additional ob¬ jective parameter to the clinical diag¬ nosis of NEO. MATERIALS AND METHODS Patient Population Between the years 1986 and 1988, a pro¬ was undertaken on 28 pa¬ tients with severe external ear infection resistant to conventional therapy and ex¬ amined in a primary outpatient clinic. Be¬ fore entry to the study, the patients were treated with ear drops and oral antibiotics (not including the quinolones). Since the ear infection could not be controlled, the patients were referred to the hospital. Those with previous ear surgery, previous NEO or SEO, bilateral NEO, and chronic otitis media were excluded from the study. The patients were examined, cultures were obtained from the external ear canal, and a
Downloaded From: http://archotol.jamanetwork.com/ by a Monash University Library User on 06/20/2015
Table 1.—Management and
Diagnosis of Patient Population* No. of Patients
patients Improvement within 1 wk Diabetic patients Improvement within 1 wk No improvement within 1
Diabetic and nondiabetic groups consisted of 14 external otitis; and SEO, severe external otitis.
Fig 1.—Technetium 99m bone scintigraphy demonstrating four skull views: anterior (ANT), posterior (POST), right lateral (RT LAT), and left lateral (LT LAT). An increased uptake is demonstrated in the right temporal bone, best appreciated on the posterior view (arrows on the posterior and right lateral views).
After the completion of bone scintigra¬ phy, patients were injected with 5 mCi (185 MBq) of Ga 67. Forty-eight-hour images were obtained using a medium energy collimator and 15% window on the three photopeaks: 94,185, and 296 KeV. The same positions and the same routine of calcula¬ tion were used. To avoid any bias of patient
management or scintigraphic interpreta¬ tion, analysis was performed at the end of the study. Thus, physicians taking care of the patient were not aware of the results of the analysis before the completion of the study, and the physicians who analyzed the results were not informed about the clinical diagnosis and management.
Fig 2.—A rectangular region of interest drawn on the posterior view lining the area of in¬ creased uptake. The same region was then moved to the unaffected side. The lesionto-nonlesion ratio was 1.6.
quantitative bone scintigraphy, followed by a 48-hour Ga 67 scintigraphy, was performed. Radionuclide Studies All patients underwent bone scintigra¬ phy and 24 of the 28 patients had Ga 67 scintigraphy performed. Prior to the bone scintigraphy, the patients were injected with 20 mCi (740 MBq) of Tc 99m MDP. Four hours later, the patients were posi¬ tioned under a large field of view gamma camera
(Apex 415, Elscint Ltd, Haifa, Israel) and fitted with an all-purpose parallel-hole collimator. A 15% window was used centered the 140-KeV photopeak. The anterior, right lateral, left lateral, and posterior skull views were obtained collecting 500 X 103 counts per view. The images were stored on a 256 X 256-matrix size for quan¬ titative analysis. on
Patient Management Nondiabetic patients with severe edema and closure of the ear canal or granulation tissue, fever, and severe pain were initially treated with intravenous antibiotics ac¬ cording to culture's sensitivity, as well as surgical débridement and ear drops. Non¬ diabetic patients with moderate edema re¬ ceived oral antibiotics, surgical dé¬ bridement, or local treatment only. All nondiabetic patients responded to the above treatment within a week and, there¬ fore, were classified as SEO patients. Diabetic patients were treated iden¬ tically. Patients who responded to treat¬ ment within the first week, by demonstrat¬ ing improvement of edema, pain, and neg¬ ative ear cultures, were classified as SEO patients and, therefore, were treated with oral antibiotics with or without local treat¬ ment. Diabetic patients who demonstrated persistent ear edema and had persistent positive Pseudomonas cultures were classi¬ fied as NEO patients and they were admin¬ istered prolonged antibiotic treatment.5 From 1986 to 1987, the treatment consisted of intravenous therapy with an aminoglycoside, mezlocillin or piperacillin sodium for a period of 6 weeks.5 Since 1987, when the quinolones became available, treatment has been started with 2 weeks of intrave¬ nous aminoglycosides, followed by 4 weeks of oral treatment using ofloxacin or ciprofloxacin.6 Thus, the hospitalization pe-
each. IV indicates intravenous; NEO,
riod could be shortened. Nondiabetic patients were followed up in the outpatient clinic for a period of 2 months, while diabetic patients were fol¬ lowed up for 1 to 3 years. Radionuclide Data Analysis
Data processing was carried out by two experienced observers (S.G. and R.H.) at the end of the study, without a previous knowledge of the affected side or the patient diagnosis. After observing the four views on the computer screen, a rectangular re¬ gion of interest, measuring 10 X 10 pixels, was drawn within the area of the mastoid bone of the affected side in the posterior view (Fig 1). The same region of interest was moved to the nonaffected side, and the counts of both regions of interest were re¬ corded. The measurements were repeated twice on two different occasions by two ob¬ servers independently, and an average was used. A lesion-to-nonlesion ratio was then calculated. The intraobserver and interobserver variabilities were calculated using linear regression analysis. The intraobserver and interobserver repeatability coefficients were calculated using the method by Bland and Altman.' The images were also read by the two ob¬ servers using the scoring method by Levin et al1 (0 to +4), by consensus. Sensitivity and specificity for the diagnosis of NEO were calculated.
RESULTS Clinical Results
Table 1 shows the clinical findings and patient management in nondia¬ betic and diabetic patients. There were 14 nondiabetic patients.
Three patients had severe edema, and treatment was started with intrave¬ nous antibiotics. Eight patients had granulation tissue and underwent sur¬ gical débridement. All 14 patients im¬ proved within 1 week and were classi¬ fied as SEO patients. The diabetic group consisted of 14
Downloaded From: http://archotol.jamanetwork.com/ by a Monash University Library User on 06/20/2015
Table 2.—Clinical and Laboratory Data of Patient Population* NEO
Variables Mean age
Tc 99m MDP
Pseudomonas infection Granulation tissue
NEO indicates necrotizing external otitis; SEO external otitis.
No. of Patients
Ratio, Mean ±SD (Range)
Male-to-female ratio Mean time of No. of
Table 5.—Patients' Radionuclide Scores for Tc 99m MDP and 67Ga
Table 4.—Tc 99m MDP and 67Ga in Diabetic and Nondiabetic SEO Patients*
(1.5-2.8) 1.35 ±0.24