Preoperative detecting metastases of cervical cancer in pelvic and para-aortic lymph nodes：comparison of integrated ~（18）F-FDG PET/CT with or without contrast enhancement

Purpose： Compared the performance of contrast-enhanced PET/CT and non-enhanced PET/CT for preoperatively detecting pelvic and para-aortic lymph node metastases in patients with cervical cancer. Methods： This prospective study included 72 patients with clinically M0 cervical cancer. They underwent surgery within two weeks of PET/CT imaging. Imaging consisted of a whole-body PET/CT protocol without intravenous contrast, followed by abdominal and pelvic PET/CT protocol including contrast-enhanced CT. We compared the diagnostic efficiency between the methods on per-patient and per-lesion basis. Results： Patient-based analysis showed that the sensitivity, specificity, and accuracy of contrast-enhanced PET/CT were 63.6% （14/22）, 94.0% （47/50）, and 84.7%（61/72）, respectively, whereas those of non-enhanced PET/CT were 54.5% （12/22）, 88.0% （44/50）, and 77.8% （56/72）, respectively, and those of enhanced CT alone were 36.4% （8/22）, 80.0% （40/50）, and 66.7% （48/72）, respectively. Lesion-based analysis showed that the sensitivity, specificity, and accuracy of contrast-enhanced PET/CT were 77.7% （87/112）, 98.7%（938/950）, and 96.5% （1025/1062）, respectively, whereas those of non-enhanced PET/CT were 69.6% （78/112）, 97.5% （926/950）, and 94.5% （1004/1062）, respectively, and those of enhanced CT were 54.4% （61/112）, 96.1% （913/950）, and 91.7% （974/1062）, respectively. Contrast-enhanced PET/CT had the best sensitivity, specificity and accuracy. Although patient-based analysis showed no significant difference between contrast-enhanced PET/CT and non-enhanced PET/CT （p =0.540, 0.295 and 0.286）, the specificity and accuracy of these two methods were significantly different on lesion-based analysis （p =0.043 and 0.027）.