ABSTRACT

This PhD dissertation is based on research performed at Rigshospitalet, Copenhagen, and describes the value of two novel methods for ovarian cancer diagnostics: the PET/CT scan and a seven biomarker panel analyzed by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS). The dissertation is based on the following prospective clinical trials:

In 210 patients with a pelvic mass (included in the Danish Pelvic Mass study), no history of cancer, and a risk-of-malignancy index (RMI) above 150, PET/CT was performed two weeks before surgery. Furthermore, blood samples were analyzed for seven serum biomarkers using SELDI-TOF MS. The biomarkers were combined into an ovarian-cancer-risk index (OvaRI) ranging from 0 (lowest risk) to 10 (highest risk). In ovarian cancer patients referred to chemotherapy, blood samples were drawn again before chemotherapy and the analysis of CA125 and OvaRI was repeated.

The sensitivity of PET/CT for diagnosing a malignant pelvic tumor was 95% (107/113) and the specificity was 91% (80/88). In 44% of 94 ovarian cancer patients, PET/CT demonstrated areas of abnormally increased metabolic activity indicating stage IV disease. On PET/CT, large bowel mesentery implants predicted incomplete cytoreduction (macroscopic residual tumor) with a sensitivity of 76% (31/41) and a specificity of 68% (17/25). In 177 patients the sensitivity of the OvaRI for diagnosing a malignant pelvic tumor was 62% and the specificity was 78% at a cut-off level of 6.61. At a cut-off level of 6.74 the OvaRI was found to predict incomplete cytoreduction with a sensitivity of 74% and a specificity of 62% in 98 stage I-IV patients. Both the CA125 levels and OvaRI values decreased after ovarian cancer surgery; the decrease was unrelated to the amount of residual tumor.

In 60 other patients clinically suspected of having recurrent ovarian cancer based on physical examinations, ultrasound findings, or increasing CA125, PET/CT was performed 68 times. The PET/CT results were compared with CT and ultrasound results. The sensitivity of PET/CT, CT, and ultrasound for diagnosing recurrence was 97% (56/58), 81% (47/58), and 66% (38/58), respectively. The specificity of PET/CT, CT, and ultrasound for diagnosing recurrence was 90% (1/10). PET/CT diagnosed recurrence in 50% (10/20) without recurrence according to CT and in 66% (19/29) without recurrence according to ultrasound. PET/CT found potentially inoperable recurrence in 42% (8/19) considered operable on CT and in 69% (27/39) considered operable by ultrasound. This dissertation shows that the use of PET/CT can alter the diagnosis and choice of treatment in ovarian cancer patients. However, additional studies including survival data are needed to further clarify the role of both PET/CT and proteomics for ovarian cancer diagnostics.