Purpose: This study describes diagnostic accuracy of PET imaging in patients with ovarian tumours using histological diagnosis as gold standard.
Methods and results: Pet studies were performed in 19 patients who were scheduled to undergo exploratory surgery for a suspicious ovarian mass and in 5 patients with suspected recurrence. The PET data were analyzed visually and quantitatively and compared to the histologic findings. 6 patients had ovarian cancer, while in 13 patients a benign tumour was found including inflammatory processes in 4 cases. All malignant tumours showed an enhanced FDG uptake with the exception of one false-negative borderline carcinoma. 4 cases with inflammatory processes as well as endometrial and follicular cysts revealed a high FDG uptake. A successful localisation of a recurrent tumour was possible in 4 out of 5 cases. Disseminated peritoneal carcinomatosis in two patients could not be detected by PET.
Conclusions: Enhanced glucose metabolism of ovarian cancer enables PET diagnosis with a sensitivity of 83%. An intensive FDG uptake in inflammatory processes resulted in a specificity of only 54% in this study. The high sensitivity of PET for malignant tumours may be useful in the detection of recurrent ovarian cancer.