Analysis of misdiagnosis of primary skull base chordomas: A case-control study

Abstract

Objective To investigate the clinical and imaging features of misdiagnosed cases of skull base chordomas and to improve the rate of correct diagnosis. Methods A retrospective analysis was conducted on 155 cases of skull base chordomas that were consecutively admitted to Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University from February 2005 to December 2014. We measured the signal intensity ratio in T2WI and enhanced T1WI sequences (RT2 and REN). The demographics, radiological and pathological features were compared between the misdiagnosed group (n=50) and control group (n=105). Further inter-group comparison and analysis were conducted about the tumor location and common features of misdiagnosed cases. Results Among the 50 cases in misdiagnosed group, 19 cases were misdiagnosed as schwannomas, 12 as epidermoid cysts, 8 as meningiomas, 5 as other osseous tumors, 3 as pituitary adenomas and 3 as cavernous malformations. The mean RT2 and REN in misdiagnosed group were 2.49 (0.79) and 0.99 (0.77), respectively, which were not statistical different from those in control group (P>0.05). However, there were significant differences when the tumor location (P<0.001) and extent of bone invasion (P<0.001) were concerned. Multivariate logistic regression showed that tumors in spheno-petrous (SP) (P=0.001) and petrous-occipital (PO) (P=0.008) regions were easily misdiagnosed compared with spheno-clival (SC) type. Exophytic chordoma was significantly easier to be misdiagnosed compared with endophytic type (P<0.001). Conclusions Skull base chordomas seem to demonstrate varying imaging features and those located in the lateral skull base and presented in the exophytic status are more prone to be misdiagnosed. The most common mimicking tumors are schwannomas in the lateral skull base and epidermoid cysts in the midline area. Signal intensity ratio on MRI could provide quantitative reference, which needs further research on its value in the diagnosis and differential diagnosis of chordomas. Key words: Chordoma; Skull base neoplasms; Diagnostic errors; Magnetic resonance imaging; Signal intensity ratio