OBJECTIVE: To review the treatment and outcomes of lateral and central skull base osteomyelitis (SBO) patients at a single academic tertiary center.
STUDY DESIGN: Retrospective case series.
PATIENTS: Twenty-three adult patients diagnosed with SBO via radiographic imaging and biopsy/culture.
METHODS: Electronic medical records were reviewed for ICD9 and ICD10 codes for osteomyelitis between 2005-2018. Demographic information including age, gender, and medical co-morbidities was collected. Disease specific data including length of symptoms, presenting symptoms, number and route of biopsies, surgical intervention, radiographic and nuclear imaging findings, pathology, culture results, total length of antibiotic treatment, and number of antibiotic courses was also collected. Patients were subdivided into groups depending on the portion of the skull base involved. Disease limited to the temporal bone was considered lateral SBO (LSBO), whereas disease limited to the clivus was considered central SBO (CSBO). SBO that involved both was considered lateral and central SBO (LCSBO).
RESULTS: Twenty-three patients were included (18 males, 5 females). The average age at presentation was 68.8 years. The most common presenting symptom was otalgia (83%) followed by headache (48%) and cranial nerve (CN) palsy (26%). The average length of each antibiotic course (IV and PO) was 9 weeks, the total length of antibiotic treatment was 20 weeks and number of courses was 2.4. The most common isolated bacteria were Staphylococcus species (39%) followed by Pseudomonas aeruginosa (30%). On subgroup analysis, 8 patients had LSBO, 6 CSBO and 9 LCSBO. Patients with CSBO had symptoms for an average of 5.4 months prior to diagnosis compared to 2.62 months for the LSBO group (p=0.03, 95% CI 0.248 to 5.336). Compared to CSBO, LCSBO was significantly more likely to present with a CN palsy (p=0.04). Seven out of nine (78%) of patients with LCSBO grew out a staphylococcus species on culture, which was higher than the LSBO group (33%, p=0.05) and significantly greater than the CSBO group (0%, p=0.007) The CSBO group underwent more biopsies than the LSBO group (mean 2.5 vs 1.25 respectively, p=0.04 [95% CI 0.03 to 2.47]). CSBO was read as a “mass” on initial imaging in 83% of cases compared to LSBO where osteomyelitis was read 100% of the time.
CONCLUSIONS: SBO continues to present a diagnostic and treatment challenge. In our single institutional experience, Staphylococcus species are the most commonly isolated bacteria in SBO. Many sources site a 12-week course of antibiotics as the recommended treatment, however, in the current study, the average course of antibiotics was only 9 weeks which may be why patients required an average of 2.4 courses of treatment. This suggests that SBO may be undertreated initially in many instances. The overall mortality rate of SBO in this study was 9%.