FDG-PET/CT in advanced ovarian cancer staging: value and pitfalls in detecting lesions in different abdominal and pelvic quadrants compared with laparoscopy.

INTRODUCTION AND AIM Ovarian carcinoma (OC) is a common cancer in the Western Countries, and an important cause of death in patients suffering with gynaecologic malignancies. The majority of patients present with advanced disease at the time of diagnosis. Treatment with debulking surgery followed by chemotherapy is the standard approach while chemotherapy is contemplated when surgery is not possible. A correct pre-operative staging is important to ensure a most appropriate management. Laparoscopy (LPS) is the standard diagnostic tool for the assessment of intraperitoneal infiltration but is invasive and requires general anaesthesia. FDG-PET/CT is increasingly used for staging different types of cancer, and the aim of this study is to assess the value of FDG-PET/CT in staging advanced OC and its sensitivity to detect lesions in different quadrants of the abdominal-pelvic area compared to laparoscopy. MATERIALS AND METHODS From September 2004 till April 2008, 40 patients with high suspicion of OC were referred to our hospital for diagnostic LPS to explore the possibility of optimal debulking surgery. Those who were not suitable for surgery were referred for chemotherapy. Before chemotherapy, the patients underwent an FDG-PET/CT scan. The findings in 9 quadrants of abdominal-pelvic area (total 360 quadrants) for PET/CT and LPS were recorded and compared. RESULTS In 14/360 areas (3.8%), surgical evaluation was not possible because of presence of adhesions, thus the number of areas explored by laparoscopy was 346. Tumour was found in 308 quadrants (38 quadrants free of disease). PET/CT was positive in all 40 patients with true negative results in 26/346 quadrants (7.5%), and true positives results in 243/346 quadrants (70.2%). False positive and negative PET/CT results were found in 12/346 and 65/346 quadrants, respectively. False positive PET/CT findings were evenly present in all quadrants. False negative PET/CT findings were present in 31/109 (28.4%) upper abdominal quadrants (epigastrium and diaphragmatic areas). Final analysis showed a sensitivity and specificity for PET/TC of 78.9 and 68.4% respectively with a positive predictive value of 95.3%. A significant difference was noted between mean SUVmax associated with lesions smaller or larger than 0.5 cm (p=0.006). CONCLUSION Our results suggest that PET/CT may prove a useful tool for pre-surgical staging of ovarian cancer with a sensitivity and specificity of 78 and 68%, respectively. However, it may be used in combination with laparoscopy for better results. PET/CT showed an adequate correlation between SUVmax values and laparoscopy findings of lesions>5mm, but a high rate of false negative results in lesions<5mm such as in carcinomatosis. PET/CT should be used carefully in early stage disease, with low risk of peritoneal infiltration, because of high rate of false positive results, to avoid unnecessary therapy procedures.