Preoperative nodal staging of uterine cancer: is contrast-enhanced PET/CT more accurate than non-enhanced PET/CT or enhanced CT alone?

Objective: To determine whether contrast-enhanced PET/CT is more accurate than either non-enhanced PET/CT or enhanced CT alone for nodal staging of uterine cancer.
Methods: Forty patients with endometrial cancer and cervical cancer underwent conventional PET/CT scan with low-dose CT (ldCT), followed by full-dose CT with IV contrast (ceCT) before radical hysterectomy with pelvic and, when applicable, para-aortic lymphadenectomy. Three data sets of PET/ldCT, PET/ceCT, and enhanced CT images were interpreted separately by two readers. For region-specific comparisons, para-aortic and pelvic lymph nodes were divided into the bilateral para-aortic, common iliac, external iliac, internal iliac, and obturator areas. Based on histopathological findings as the gold standard, we compared the diagnostic accuracy between the three methods using McNemar test with Bonferroni's adjustment.
Results: Of the 40 patients, 21 underwent pelvic lymphadenectomy only. Region-based analysis showed that the sensitivity, specificity, and accuracy of PET/ceCT were 61.4% (27/44), 98.1% (308/314), and 93.6% (335/358), respectively, whereas those of PET/ldCT were 52.3% (23/44), 96.8% (304/314), and 91.3% (327/358), respectively, and those of enhanced CT were 40.9% (18/44), 97.8% (307/314), and 90.8% (325/358), respectively. Although PET/ceCT had the best sensitivity among the three imaging modalities, a significant difference was observed only between PET/ceCT and enhanced CT (p = 0.0027). Although PET/ceCT had better sensitivity and accuracy than PET/ldCT, the differences between the two imaging methods did not reach statistical significance (p = 0.046 and p = 0.047, respectively).
Conclusion: PET/ceCT is slightly but not significantly superior to PET/ldCT for nodal staging of uterine cancer. Nodal metastasis cannot be excluded even if PET/ceCT gives negative findings.