HSC Section 8_April 2017

113 / 264

disease was likely with involvement of 2 or more of the fol-
lowing areas: temporomandibular joint, temporal bone, and
base of skull. The CT findings of our patients were stratified
into major and minor findings based on the studies by Peleg
et al13 and Soudry et al.4 Contrary to the observations by
Peleg et al,13 no correlation was seen between the presence
of major area involvement and poor outcome. This lack of
correlation was also noted by Sudhoff et al.14 Bone erosion
can only be seen once demineralization has occurred.
Demineralization becomes evident after weeks of inflamma-
tion,3 making early disease hard to detect. Moreover, once
demineralization has occurred, the bony changes persist
even after inflammation has settled.6 This loose association
between bone demineralization and disease activity could
account for the lack of reliability of CT scans in predicting
outcome. In cases where medial and intracranial extension
of disease had to be visualized, we used MRI to better
delineate soft tissue involvement.
Direct extension of disease medially from the petrous
temporal bone can progress to involve the clivus, a central
structure at the anterior-most portion of the basilar occipital
bone where it meets the sphenoid bone. We considered
patients with clival involvement to have central skull base
osteomyelitis (Figure 3). Central skull base osteomyelitis
can affect the surrounding soft tissues, compromising the
lower cranial nerves and brainstem.15 Extension of disease
through the petroclival synchondrosis can result in intracra-
nial involvement with the development of meningitis,
abscess, and venous sinus thrombosis.1 Clival involvement
was noted in 5 patients. All these patients had disease that
persisted after 6 weeks of antibiotics. In the group without
clival involvement, only 14.3% (n = 2) had persistent dis-
ease. Clival involvement was also seen in all the mortalities
that resulted from intracranial involvement. We suggest that
patients with clival erosion should be counseled appropri-
ately on prognosis and have more aggressive treatment.
Antibiotic Treatment
It has become increasingly difficult to isolate causative micro-
organisms from the EAC for culture-directed therapy because
of the use of otic antibiotics at primary care. Only 63.2% (n =
12) of ear swabs were positive. Increasing antibiotic resistance
in P aeruginosa represents another problem. This organism
has properties that make it inherently resistant to many drug
classes and able to acquire resistance through mutation.16,17
Taking into consideration all forms of infections caused by
P aeruginosa from 33 European countries, the European
Antimicrobial Resistance Surveillance System Annual
Report for 2006
reports/) reported that 18.0% of isolates were multidrug resis-
tant (ie, resistant to 3 or more antibiotics). Specific to P aeru-
ginosa in MOE, Berenholz et al18 raised the concern of
ciprofloxacin resistance when they reported that 33.0% of
isolates were resistant to ciprofloxacin. This was mirrored in
other studies in which the rates of ciprofloxacin resistance
have ranged from 31.0% to 37.5%.18-20 Similarly, 33.3% (n =
3) of P aeruginosa isolates in our series were ciprofloxacin
resistant. Levenson et al21 declared that ciprofloxacin mono-
therapy was the treatment of choice in MOE. However, cur-
rent resistance rates suggest that monotherapy with
ciprofloxacin is imprudent in almost one-third of cases.
Antibiotic choices in culture-negative patients therefore rep-
resent a therapeutic challenge in view of the high antibiotic
resistance rates of P aeruginosa. Our results suggest that the
empirical use of combination anti-pseudomonal therapy with
intravenous ceftazidime and oral ciprofloxacin in culture-
negative cases remains relevant despite increasing antibiotic
resistance in P aeruginosa . This is consistent with reports in
the literature.18,22 Rubin et al23 proposed the use of combina-
tion therapy with oral rifampacin and ciprofloxacin in treating
drug-resistant P aeruginosa, and they showed that their oral
and intravenous regimes were equally efficacious. However,
in another study by Korvick et al24 involving 121 patients
with positive blood cultures for P aeruginosa, no significant
benefit was demonstrated from the addition of rifampacin to
existing anti-pseudomonal therapy, suggesting that further
clinical studies on the anti-pseudomonal properties of rifam-
pacin are needed. In addition, due to the significant risk of
hepatotoxicity and drug interactions related to hepatic cyto-
chrome P450 upregulation, rifampacin was not considered for
our regime.
The role of surgery is limited. Only 3 patients in our
series underwent surgery, and the indications for surgery
were local debridement and to obtain cultures. The useful-
ness of surgery was limited in these cases as intraoperative
cultures did not yield additional information over ear swab
specimens, and disease persisted despite debridement.
Conclusion
Definite prognostic factors remain elusive. Our experience
has shown that age, severity of diabetes, and duration of
symptoms cannot be relied upon to predict prognosis. When
radionuclide scans are not readily available, anatomic
Figure 3. Clival erosion in central skull base osteomyelitis.
Loh and Loh
92