December 1, 2008
The predictive model can improve the possibility of correctly identifying the optimally cytoreducible patients from 85% as assessed by CT scan, to the clinically relevant alternative of 100% as assessed by laparoscopy.
Laparoscopy followed by laparotomy with scores calculated from assessments made during each surgical approach.
9 parameters used to forecast optimal cytoreduction surgery: ovarian masses, omental cake or nodules, peritoneal carcinomatosis, diaphragmatic carcinomatosis, mesenteric retraction, bowel infiltration, stomach infiltration, liver metastases and lymph nodes
Unresectable Score = 2
Resectable Score = 0
Total predictive index (PIV), followed by sensitivity, specificity, PPV, NPV and accuracy with respect to residual disease after laparotomy
At least 2 of the following:
Ascites >500 mL
Elevated CA 125 >500 UI/mL
CT evidences of metastatic disease equivalent to stage III/IV
Clinically early stage
Very large abdominal mass
Disease progression during NACT or achieved an optimal response during NACT
With PIV >8, the probability of optimally resecting disease at laparotomy is 0 and the rate of unnecessary laparotomy is 40.5% (PPV= 100%; NPV 59.5%).
Mean PIV 5.
Overall accuracy of laparoscopic procedure 77.3%-100%
Metric: Assessable / NPV / PPV / Accurancy
Peritoneal carcinosis: 100% / 100% / 100% / 100%
Omental cake: 96.5 / 91.4 / 98.6 / 96.3
Diaphragmatic carcinosis: 91.1 / 86.8 / 100 / 95.1
Mesenteral retraction: 75.2 / 93.7 / 100 / 95.3
Bowel infiltration: 85.8 / 71.6 / 86.4 / 77.3
Stomach infiltration: 84.9 / 97.8 / 60.0 / 95.8
Superficial liver metastasis: 93.8 / 88.5 / 89.4 / 88.7
The proposed laparoscopic model appears to be a reliable and flexible tool to predict optimal cytoreduction in advanced ovarian cancer