OBJECTIVE: Accurate preoperative identification of the attachment site is the key to the successful surgical management of sinonasal inverted papillomas (IPs). This study investigated the value of preoperative imaging to identify the attachment sites of IPs. METHODS: We analyzed I 0 consecutive patients with pathologically proven IPs. Two radiologists predicted the attachment sites of IPs from computed tomography (CT), 3.0 Tesla magnetic resonance imaging (3.0T MRI), 1.5T MRI, and CT plus 3.0T MRI. The actual tumor attachment sites were confirmed via pathological examination of specimens and compared with the predicted sites. RESULTS: Computed tomography showed the highest sensitivity (P < .0001), although both MRI formats showed greater specificity (P < .0001). The sensitivity of MRI plus CT was equal to that of CT and better than that of MRI (P < .0001), whereas its specificity was better than that of CT (P < .0001) and comparable to that of MRI. Prediction using 3.0T MRI appeared slightly superior to that using 1.5T MRI in terms of sensitivity and specificity, although the differences were not significant. CONCLUSION: Computed tomography and MRI had different features for prediction of sinonasal IP attachment sites. Preoperative CT plus MRI provided more useful information than CT or MRI alone.
OBJECTIVE: Accurate preoperative identification of the attachment site is the key to the successful surgical management of sinonasal inverted papillomas (IPs). This study investigated the value of preoperative imaging to identify the attachment sites of IPs. METHODS: We analyzed I 0 consecutive patients with pathologically proven IPs. Two radiologists predicted the attachment sites of IPs from computed tomography (CT), 3.0 Tesla magnetic resonance imaging (3.0T MRI), 1.5T MRI, and CT plus 3.0T MRI. The actual tumor attachment sites were confirmed via pathological examination of specimens and compared with the predicted sites. RESULTS: Computed tomography showed the highest sensitivity (P < .0001), although both MRI formats showed greater specificity (P < .0001). The sensitivity of MRI plus CT was equal to that of CT and better than that of MRI (P < .0001), whereas its specificity was better than that of CT (P < .0001) and comparable to that of MRI. Prediction using 3.0T MRI appeared slightly superior to that using 1.5T MRI in terms of sensitivity and specificity, although the differences were not significant. CONCLUSION: Computed tomography and MRI had different features for prediction of sinonasal IP attachment sites. Preoperative CT plus MRI provided more useful information than CT or MRI alone.