Twenty-five diagnostic accuracy studies (n greater than 1,044) were included. Eight studies were prospective, 12 studies were retrospective; the design of the remaining 5 studies was not clear.
Newly diagnosed cervical cancer.
The pooled sensitivity and specificity of FDG-PET for detecting aortic node metastasis (4 prospective studies) were 0.84 (95% CI: 0.68, 0.94) and 0.95 (95% CI: 0.89, 0.98), respectively. The pooled sensitivity and specificity for the detection of pelvic node metastasis were, respectively, 0.79 (95% CI: 0.65, 0.90) and 0.99 (95% CI: 0.96, 0.99) with FDG-PET (2 prospective and 2 retrospective studies), and 0.72 (95% CI: 0.53, 0.87) and 0.96 (95% CI: 0.92, 0.98) with MRI (1 prospective and 1 retrospective study). The pooled sensitivity for CT was 0.47 (95% CI: 0.21, 0.73) (1 prospective, 1 retrospective study). Two retrospective studies investigating progression-free survival provided some limited but supportive evidence for the prognostic significance of pre-treatment lymph node staging using FDG-PET.
Recurrent cervical cancer.
The pooled sensitivity and specificity of FDG-PET were 0.96 (95% CI: 0.87, 0.99) and 0.81 (95% CI: 0.58, 0.94), respectively, for detecting recurrent cervical cancer with clinical suspicion (3 retrospective studies), and 0.92 (95% CI: 0.77, 0.98) and 0.75 (95% CI: 0.69, 0.80) without clinical suspicion for recurrence (2 retrospective studies). The results of one retrospective study investigating progression-free survival showed that post-treatment surveillance FDG-PET is predictive of progression-free survival in individuals with a history of cervical cancer. One retrospective and one prospective study suggested that the use of FDG-PET influenced treatment choices.
Ovarian cancer.
The sensitivity and specificity for detecting recurrence with clinical suspicion were, respectively, 0.90 (95% CI: 0.82, 0.95) and 0.86 (0.67, 0.96) with FDG-PET, 0.68 (95% CI: 0.49, 0.83) and 0.58 (95% CI: 0.33, 0.80) with conventional imaging (2 studies), and 0.81 (95% CI: 0.62, 0.92) and 0.83 (95% CI: 0.58, 0.96) with CA-125 (2 studies). The sensitivity and specificity for detecting recurrence without clinical suspicion were 0.54 (95% CI: 0.39, 0.69) and 0.73 (95% CI: 0.56, 0.87) when conventional imaging and CA-125 were negative (3 studies). In 2 studies of patients without clinical evidence of recurrence, FDG-PET was not sensitive for the detection of microscopic residual disease. When CA-125 was rising and conventional imaging was negative, the pooled sensitivity and specificity of FDG-PET were 0.96 (95% CI: 0.88, 0.99) and 0.80 (95% CI: 0.44, 0.97), respectively (3 studies).
No publication bias was detected.