The aim of this study was to determine the prognostic significance of PET/CT findings in women with cervical cancer and describe the normalization of lymph node SUVmax (nSUVmax).Materials and Methods
A retrospective review was performed of 113 patients with cervical cancer who underwent a PET/CT before receiving definitive therapy. SUVmax measurements were normalized to the SUV of the pelvic blood pool. Patient, tumor, and PET/CT data were correlated to extracervical recurrence-free survival (ecRFS) and lymph node pathology.Results
Of 113 patients, there were 23 (20%) extracervical recurrences. On univariate analysis, stage, histology, nSUVmax, and radiographic size of the primary tumor, and nSUVmax of the most hypermetabolic lymph node were significantly associated with ecRFS. On multivariable analysis, nSUVmax and radiographic size of the primary tumor remained associated with ecRFS (both P < 0.001). Sixty-six patients underwent pelvic, common iliac, and/or para-aortic nodal sampling. The sensitivity, specificity, false-negative, and false-positive rates of PET/CT for lymph node metastases were 53%, 75%, 6%, and 82%, respectively. On univariate analysis, nSUVmax, and radiographic size of the primary tumor, and nSUVmax of the most hypermetabolic lymph node, and radiographic size of the largest lymph node, were associated with the presence of at least one pathologically positive lymph node. On multivariable analysis, only the radiographic size of the largest lymph node remained significantly associated with lymph node metastases (P < 0.001).Conclusions
The size and nSUVmax of the primary tumor were associated with ecRFS. PET/CT has a low false-negative rate but high false-positive rate for lymph node metastases.