Sensitivity and specificity of computed tomography and magnetic resonance imaging in the diagnosis of isolated sphenoid sinus diseases.

OBJECTIVES/HYPOTHESIS: The sphenoid sinus is uncommonly affected by inflammatory or neoplastic lesions. Initial onset of isolated sphenoid sinus diseases (ISSD) is generally asymptomatic. The objectives of this study were to estimate the sensitivity and specificity of computed tomography (CT) and magnetic resonance imaging (MRI) in diagnosing ISSD and establish guidelines to declare hidden ISSD through correlation of radiological diagnosis to final pathological diagnosis.

STUDY DESIGN

A prospective cohort study.

METHODS

There were 66 patients with isolated sphenoid sinus lesions presenting to Ain-Shams University Hospitals, Cairo, Egypt. Provisional diagnosis of ISSD was done by CT and MRI, followed by histopathologic and immunohistochemical staining, and if needed microbiological examination of resected specimens to establish the final diagnosis.

RESULTS

Patients were classified into four groups according to the type of lesion: inflammatory, neoplastic, bony dysplastic disorders, and sphenoid sinus roof defect-related lesions. Radiological imaging provided the greatest diagnostic information and guided management. The sensitivity of CT and MRI in diagnosing inflammatory lesions was 95% versus 61%, whereas those of the neoplastic group were 72% and 100%, respectively. In the osseous group the sensitivity was 100% for both CT and MRI, whereas in sphenoid sinus roof defect the sensitivity was 50% and 100% for CT and MRI, respectively.

CONCLUSIONS

Because of CT's superiority in defining the bony margins and MRI's superior soft tissue resolution, CT and MRI should be used in a complementary manner in the evaluation of isolated sphenoid sinus disease in addition to mapping the lesion better and identifying intracranial and intraorbital extension. The use of one modality only should be restricted to straightforward lesions.