Advanced Ovarian Cancer. A multimodal diagnostic approach to predict outcome

Sammanfattning: Primary debulking surgery (PDS) followed by platinum-based postoperative chemotherapy isthe standard of care for advanced ovarian cancer (AOC). Absence of macroscopic residualdisease after debulking surgery is the strongest prognostic factor achieved by surgery. Correctcharacterization of the tumor specimen and tumor spread in combination with patient’scharacteristics such as age, comorbidity, and personal wishes, can help to select more effectivetherapeutic approaches for each patient before initial intervention.The overall aim of this thesis were to evaluate diagnostic, from the preoperative to theintraoperative stage, to investigate how an accurate diagnosis could predict surgical outcomeand survival in patients with advanced ovarian cancer.Study I: A retrospective population-based review was conducted of 328 biopsies, in order toassess the adequacy, accuracy and safety of tru-cut biopsy in gynecological malignancies fromthe perspective of a daily clinical practice. The tru-cut biopsy was shown to be a reliable andsafe diagnostic method, with adequacy of 86.3%, accuracy of 97.5% and a complication rate of1.3%. The adequacy of tru-cut biopsy depends on the site of the tissue sample, indications forthe biopsy and the experience of the operator.Study II: A single-center, retrospective population-based study was conducted on 358 patients,to evaluate the reliability of intraoperative FS diagnosis for planning the treatment of patientswith suspected ovarian cancer (OC), from a multidisciplinary perspective. Prevalence,sensitivity, specificity, positive predictive value and negative predictive value for invasivemalignancies on FS were 54.0%, 88.1%, 98.8%, 98.9% and 87.6% respectively. Malignancywas observed to be underestimated, but overestimation in malignancy grading was rare.Borderline-related tumors were more likely to be incorrectly graded by FS, as were rare tumortypes. Despite diagnostic difficulties, in some of the cases, the oral communication duringdeliverance of frozen section diagnosis resulted in adequate treatment decisions, whichminimalized the risk for reoperation or delay of chemotherapy treatment.Studies III and IV: A single-center, retrospective population-based study was conducted on 118patients with AOC, to determine whether the PCI and the quantity of ascites visualized bycomputed tomography (CT) could assess the extent of the tumor (S-PCI) and residual disease(RD) for AOC patients treated with PDS. Furthermore, in study IV, we examine the impact ofthe tumor extent on survival. CT-PCI correlated well with S-PCI and the risk of RD, with a cutoffof 21 for CT-PCI (0.715, p = 0.000). The risk of RD was 3.5 times higher when the quantityof ascites on CT (CTascites) was estimated to be above 1000ml. Regardless of the completenessof cytoreductive surgery or the complication rate, the extent of the tumor at the beginning ofsurgery seemed to affect OS in patients with AOC. PCI above 18.5 doubled the risk of dyingof the disease. CT-PCI seemed to play a prognostic role for PFS, but its prognostic role for OSis still to be investigated.Conclusions: The existing methods of preoperative material retrieval and histopathologicaldiagnosis of ovarian cancer are reliable, when performed by highly trained specialists. Thepreoperative CT is accessible and can be used by an experienced radiologist as a singletechnique to select patients as candidates for PDS. The complete removal of the tumor is a veryimportant prognostic factor for prognosis in AOC, but patient’s status and tumor biology arealso important factors in the decisionmaking on a treatment plan. This thesis maintain the ideathat centralization of cancer care to tertiary centres results in highly specialized pathology,radiology, oncology and surgical departments, and that the multidisciplinary diagnostic andtherapeutic efforts improve health care, and possibly the outcome, for patients with AOC.