Comparison of computed tomography histogram analysis and chemical-shift magnetic resonance imaging for adrenal mass characterization

Abstract

Background: Computed tomography (CT) histogram analysis and chemical-shift magnetic resonance imaging (MRI) are currently used modalities for adrenal mass characterization. However, it is not yet clear which modality can be regarded as most sensitive in terms of adrenal mass characterization.

Purpose: To prospectively compare CT histogram analysis and chemical-shift MRI in the characterization of adrenal masses.

Material and methods: Between May 2007 and November 2008, 93 patients (45 males, 48 females; mean age 56.7 years, range 22-85 years) with 109 adrenal masses prospectively underwent both unenhanced CT and chemical-shift MRI examinations. These masses consisted of 67 adenomas and 42 metastases. Histogram analysis was applied with a circular region of interest (ROI) that recorded mean attenuation, total number of pixels, number of negative pixels, and the percentage of negative pixels on unenhanced CT images for each adrenal mass. In the CT histogram analysis, a 10% negative pixel threshold for unenhanced CT was calculated. In chemical-shift MRI, signal intensity drop between in-phase and opposed-phase images was quantitatively calculated so that adrenal-to-spleen chemical-shift ratios and adrenal signal intensity indexes were determined for each of the adrenal masses. A mass was regarded as an adenoma if it contained more than 10% negative pixels by CT histogram analysis, showed an adrenal-to-spleen chemical-shift ratio of less than 0.71, and had an adrenal signal intensity index of more than 16.5% by chemical-shift MRI. The results were compared to reveal which method was most sensitive in the diagnosis of adrenal masses and whether or not a correlation exists between these two modalities. Final diagnoses were based on imaging follow-up of minimum 6 months, biopsy, surgery, and adrenal washout study.

Results: On unenhanced CT examinations, all of the 67 adenomas and 21 out of 42 metastases exhibited negative pixels. None of the metastases showed more than 10% negative pixels on CT histogram analysis. An increase in the percentage of negative pixels correlated well with a decrease in the mean CT attenuation. CT histogram analysis using a 10% negative pixel threshold gave a 91% sensitivity and 100% specificity for the diagnosis of an adenoma. On chemical-shift MRI, for an adrenal-to-spleen chemical-shift ratio of less than 0.71, a 97% sensitivity and 100% specificity were achieved, while a 97% sensitivity and 93% specificity were obtained for an adrenal signal intensity index of more than 16.5% for adenoma diagnosis.

Conclusion: CT histogram analysis method using a 10% negative pixel threshold on unenhanced CT had a good sensitivity and perfect specificity for the differentiation of adrenal adenomas from non-adenomas. In spite of the good results obtained with the CT histogram analysis method, chemical-shift MRI using adrenal-to-spleen chemical-shift ratio and adrenal signal intensity index formulas had a higher sensitivity and could help in the characterization of adrenal masses appearing indeterminate by CT histogram analysis.