HSC Section 8_April 2017

Otolaryngology–Head and Neck Surgery 148(6)

normalization of inflammatory markers and without relapse within 12 weeks. Those in whom disease resolved after a 6-week course of intravenous antibiotics were deemed as having a good out- come. Those who had persistent symptoms despite 6 weeks of therapy were considered to have a poor outcome. The clinical characteristics and laboratory and imaging findings between the 2 groups were compared to identify possible prognostic factors. Data analysis was performed with SPSS 16.0 software (SPSS, Inc, an IBM Company, Chicago, Illinois). Statistically significant differences in categorical variables were cross-tabulated and analyzed with the Fisher exact test. The Mann-Whitney U test and t test were used to analyze continuous variables as appropriate. P values .05 were considered statistically significant. Results Demographics and Outcome Twenty-two cases of MOE were identified. Three were excluded from analysis: 2 foreign patients chose to continue treatment in their own country, and 1 died of pneumonia shortly after initiating treatment. The remaining 19 patients consisted of 16 men and 3 women. Mean age was 69.1 years (range, 51-86 years). Disease resolved in 63.2% (n = 12) after 6 weeks of antibio- tics. In 36.8% (n = 7), disease was persistent after 6 weeks, and additional antibiotics were required. Four in this group ultimately died. The overall mortality rate in our series was 21.1% (n = 4). Three patients died from MOE-associated intracranial complica- tions. The remaining patient died from intracranial hemorrhage secondary to ceftazidime-induced thrombocytopenia. There was a tendency for those 65 years and older to have disease that persisted after 6 weeks of antibiotics (45.0% of those 65 years and older vs 25.0% of those younger than 65 years), but this did not reach statistical sig- nificance ( P = .633). Comorbidities All patients were immunocompromised. Diabetes was present in 94.7% (n = 18). The only patient (n = 1) without diabetes was on long-term steroids for rheumatoid arthritis. The macrovascular and microangiopathic effects of diabetes were present in a significant proportion: 63.2% (n = 12) had ischemic heart disease, 15.8% (n = 3) had peripheral vascular disease, and 26.3% (n = 5) had end-stage renal failure. Diabetes severity was assessed based on glycated hemoglobin (HbA1c) levels at diagnosis. Of those with diabetes, 55.6% (n = 10) had HbA1c in excess of 7.0%, reflecting poor glu- cose control prior to diagnosis of MOE. There was no associ- ation between HbA1c levels and outcome ( P = 1.00). Clinical Presentation The most common complaint, present in 73.7% (n = 14), was concomitant otalgia and otorrhea; 10.5% (n = 2) had parotid swelling. Facial nerve palsy was the main symptom in 21.0% (n = 4). The mean duration of symptoms before MOE was diagnosed was 6.79 weeks (range, 1-12 weeks).

There was no statistically significant difference in the dura- tion of symptoms prior to diagnosis between the group in which disease resolved after 6 weeks of intravenous treat- ment and the group in which disease persisted. Mean dura- tion of symptoms was 6.67 weeks in the former group compared with 7.0 weeks in the latter group ( P = .859). Although patients with cranial nerve involvement tended to have more persistent disease, this did not reach statistical significance (50.0% of patients with cranial nerve palsy had poor outcome vs 33.3% of those without; P = .603). Inflammatory Markers Inflammatory markers were evaluated at diagnosis and seri- ally as treatment progressed. Total white cell count (TWC; normal range, 3.40-9.60 3 10 9 /L), C-reactive protein (CRP; normal range, 0-10 mg/L), and erythrocyte sedimentation rate (ESR; normal range, 5-15 mm/h) levels were measured. Leukocytosis was less prominent compared with derange- ments in CRP and ESR levels. At diagnosis, only 26.3% (n = 5) had abnormal TWC levels compared with 72.2% (n = 13) with raised CRP and 84.2% (n = 16) with raised ESR. The mean (SD) values of these inflammatory markers were as fol- lows: TWC, 9.98 (4.12) 3 10 9 /L; CRP, 42.56 (51.89) mg/L; and ESR, 66.82 (34.73) mm/h. Inflammatory marker levels were compared between the 2 outcome groups at the time of diagnosis and at the 2- and 6-week point after initiating antibiotics. The results are pre- sented in Table 1 . No significant difference in mean CRP and ESR levels was detected between the 2 outcome groups at these 3 intervals. The group in which disease progressed had a statistically significant higher TWC after 6 weeks of antibiotics compared with the group in which disease resolved, but overall, leukocytosis was not prominent and remained at the upper limit of normal. Although the abso- lute levels of the inflammatory markers appeared to be of limited usefulness in predicting outcome, a more prominent downward trend of levels was seen as treatment progressed in the group in which disease resolved, as compared with the group in which it persisted ( Figure 1 ). Imaging Computed tomography (CT) was the imaging modality of choice at our institution and was performed in 89.5% (n = 17) at diagnosis. Computed tomography findings were divided into minor and major findings based on a study by Soudry et al 4 that showed a correlation between major CT findings and persistent disease. Minor findings were defined as EAC tissue swelling, EAC bony erosion, and mastoid involvement. Major findings were defined as infratemporal fossa involvement, temporomandibular joint involvement, parapharyngeal involvement, and nasopharyngeal involve- ment. The CT findings are presented in Table 2 . The presence of major findings on initial CT scans was not seen to be predictive of outcome (33.3% in the poor out- come group and 36.4% in the good outcome group had major findings; P = 1.00).

89