Positron Emission Tomography (PET) in Oncology

Free Full-Text 

1.8.1. Ovarian Cancer

There is mounting evidence that FDG-PET/CT has an increasing role in the management of ovarian cancer, with its main indication to detect tumor recurrence in presence of rising CA-125 serum values and negative conventional imaging studies [93]. The benefits of the use of FDG-PET/CT in these settings has been reported several times in the literature [94,95], with a sensitivity of more than 90% in detecting occult metastases. In the study of Zimny et al., FDG-PET/CT preceded the conventional diagnosis by a median of 6 months in patients judged clinically free of disease. Menzel et al. suggest that a PET indication is worthwhile at CA 125 levels of approximately 30 U/mL [96]. A more recent prospective multi-center, cohort study (90 patients) confirmed the impact of FDG-PET/CT in suspected recurrent ovarian cancer, which affected disease management decisions in 60% of the cases (in 49% with a high, in 11% with a medium clinical impact) with a much higher detection rate compared to conventional imaging [97].
For the characterization of asymptomatic adnexal findings, FDG-PET/CT has no place due to lack of sensitivity [98], and MRI remains the best imaging modality choice.
For the initial staging of ovarian cancer, FDG-PET/CT is not routinely used. Nevertheless, some publications noticed that it could be interesting in advanced epithelial ovarian cancer, in particular for the detection of supradiaphragmatic lymph node metastases like parasternal lymph nodes, with better accuracy than conventional CT (detection rate: 67% vs. 33%) [99]. However, increased mediastinal FDG uptake was not shown to play a significant prognostic role, while complete cytoreduction did [100]. For the initial preoperative staging of ovarian cancer, FDG-PET/CT may be superior compared to CT alone [101,102], but some publications also observed limits, as De Iaco et al., who reported a sensitivity and specificity of 78 and 68% respectively, with a high rate of false negative results in lesions <5 mm such as found in presence of peritoneal carcinomatosis [103].
However, conflicting results have been reported on the sensitivity of FDG-PET/CT scan in detecting peritoneal carcinomatosis; Turlakow, Suzuki and Kim reported higher diagnostic accuracy of FDG-PET/CT than CeCT in this settings, with a sensitivity and specificity for FDG-PET/CT of 67%–92.2% and 90%–94% respectively, as compared to 22%–88.5% and 65%–77% respectively for CeCT [104,105,106]. The sensitivity of FDG-PET/CT proved also similar to that of conventional MRI, and even better for detecting small peritoneal lesions (<2 cm) in patients with recurrent ovarian cancer [107]. However, FDG-PET/CT has limits, in particular for the detection of small peritoneal implants (<5 mm) because of the limited PET resolution, and surgical staging remains the gold standard [108]. The good performances of FDG-PET/CT in detecting peritoneal carcinomatosis lead to interesting information for optimizing patient selection for cytoreductive surgery in recurrent ovarian cancer; recently, Ebina et al. observed that FDG-PET/CT led to a change in management plan in 58.4% in that case, with a total number of patients in whom cytoreductive surgery was selected as the treatment of choice increased from 12 to 35 according to FDG-PET-CT results [109]. In the preoperative management, FDG-PET/CT is also able to detect distant metastases (25/95 patients upstaged from FIGO stage III to stage IV by FDG-PET/CT in a recent study [110]. However, upward stage migration did not worsen the prognosis of stage III patients, and in advanced ovarian cancer, the only prognostic factor that retained a significant prognostic value is the quality of response to cytoreductive therapy. Another study proposed FDG-PET/CT criteria such as FDG-PET/CT stage IV, pleural exudates, and PET-positive large bowel mesentery implants, which were statistically significant in the prognosis univariate analysis to guide the administration of neo-adjuvant chemotherapy in advanced ovarian cancer, but, once again, incomplete tumor debulking was the only statistically significant independent prognostic variable using multivariate analysis (p = 0.0001) [111]. Other prognostic factors like MTV or TGL may be interesting, but more data are needed at this time to confirm that [112].
- See more at: http://www.mdpi.com/2072-6694/6/4/1821/htm#sthash.MzNWUkeA.dpuf