RATIONALE AND OBJECTIVES: To identify magnetic resonance (MR) imaging characteristics allowing specific preoperative discrimination between colloid cysts (CCs) of the sellar region and third ventricle (CC3rdv) versus Rathke's cleft cysts (RCCs). MATERIALS AND METHODS: MR imaging data of 38 patients with histologically proven CCs/CC3rdv and RCC underwent retrospective analysis with respect to signal intensity and heterogeneity on T1- and T2-weighted images, presence of the dot sign, enhancement, size, location, and accompanying infundibular stalk abnormalities. RESULTS: Thirteen patients had CCs, 12 had CC3rdv, and 13 had RCCs. Signal intensity on T1-weighted images was partly or entirely hyperintense (n = 8), iso- or mixed iso/hypointense (n = 5) in CCs; hyperintense (n = 8), isointense, or mixed hypo/isointense (n = 3) in CC3rdv and hyperintense (n = 9); or mixed (n = 4) in RCCs. On T2-weighted images, signal intensity was hypointense (n = 12) or hyperintense (n = 1) in CCs, hypointense (n = 9) or hyperintense (n = 2) in CC3rdv, and hypointense (n = 5) or iso/hyperintense (n = 8) in RCCs. T2-weighted images were unavailable in two patients. Only one questionable enhancement was found in CCs, whereas an enhancing rim was consistently seen in RCCs. The dot sign was present in 7 CCs, 8 CC3rdv, and 4 RCCs. Mean cyst diameters were 12.6 mm for CCs and 14.5 mm for RCCs. RCCs showed more frequent and even solely suprasellar extent contrary to CCs. CONCLUSION: Cyst wall enhancement was found in all RCCs but in none of the CCs, making this feature a reliable discriminator between the two. Complementary, suprasellar extension was more frequent in RCCs, whereas signal hypointensity on T2w was more common in colloid cysts.
RATIONALE AND OBJECTIVES: To identify magnetic resonance (MR) imaging characteristics allowing specific preoperative discrimination between colloid cysts (CCs) of the sellar region and third ventricle (CC3rdv) versus Rathke's cleft cysts (RCCs). MATERIALS AND METHODS: MR imaging data of 38 patients with histologically proven CCs/CC3rdv and RCC underwent retrospective analysis with respect to signal intensity and heterogeneity on T1- and T2-weighted images, presence of the dot sign, enhancement, size, location, and accompanying infundibular stalk abnormalities. RESULTS: Thirteen patients had CCs, 12 had CC3rdv, and 13 had RCCs. Signal intensity on T1-weighted images was partly or entirely hyperintense (n = 8), iso- or mixed iso/hypointense (n = 5) in CCs; hyperintense (n = 8), isointense, or mixed hypo/isointense (n = 3) in CC3rdv and hyperintense (n = 9); or mixed (n = 4) in RCCs. On T2-weighted images, signal intensity was hypointense (n = 12) or hyperintense (n = 1) in CCs, hypointense (n = 9) or hyperintense (n = 2) in CC3rdv, and hypointense (n = 5) or iso/hyperintense (n = 8) in RCCs. T2-weighted images were unavailable in two patients. Only one questionable enhancement was found in CCs, whereas an enhancing rim was consistently seen in RCCs. The dot sign was present in 7 CCs, 8 CC3rdv, and 4 RCCs. Mean cyst diameters were 12.6 mm for CCs and 14.5 mm for RCCs. RCCs showed more frequent and even solely suprasellar extent contrary to CCs. CONCLUSION: Cyst wall enhancement was found in all RCCs but in none of the CCs, making this feature a reliable discriminator between the two. Complementary, suprasellar extension was more frequent in RCCs, whereas signal hypointensity on T2w was more common in colloid cysts.