Does computed tomography scan predict suboptimal primary cytoreduktion in patients with advanced ovarian cancer?

Einleitung/Falldarstellung:

In advanced ovarian cancer, optimal primary cytoreduction is the most important prognostic factor. Patients with a low likelihood to achieve optimal cytoreduction should be scheduled for neoadjuvant chemotherapy. The aim of this study is to identify features on preoperative computed tomography (CT) scans for prediction of suboptimal cytoreduction in primary surgery of ovarian cancer.

Methodik/Material:

We performed a retrospective review of preoperative CT scans of patients with primary surgery of ovarian cancer in the university hospital of Rostock between 2010 and 2014 (n = 75). Thirteen radiological criteria were assessed by an experienced radiologist blinded to surgical results.

Multivariate stepwise logistic regressions models were applied to determinate sites of disease which were associated with suboptimal surgical outcome..

Resultate/Ergebnisse:

Optimal cytoreduction with no residual tumor was achieved in 37.3% of patients, in 34.7% residual tumor was ≤1 cm, in 21.3% > 1 cm, and 6.7% of patients were inoperable. Univariate logistic regression revealed that diagphragm disease, large bowel mesentery implants, omental cake, diffuse peritoneal thickening, ascites and peritoneal implants > 2 cm were significantly associated with suboptimal surgical outcome (p < 0.05). In multivariate logistic regression analysis, only diffuse peritoneal thickening was identified to predict incomplete cytoreduction (p = 0.045). Using this predictor, CT scan had a sensitivity of 78.7% (95%CI 64.3%-89.3%), and specificity of 64.3% (95%CI 44.1%-81.4%).

Schlussfolgerung/Zusammenfassung:

Although radiological features could be identified as predictors of suboptimal surgery, results should be used with caution due the moderate value for specificity. A prospective study with higher number of cases should be done for validation of results.