Prediction of non-resectability in tubo-ovarian cancer patients using Peritoneal Cancer Index – A prospective multicentric study using imaging (ISAAC study)

Abstract

Background

The aim was to evaluate the performance of the Peritoneal Cancer Index (PCI) using imaging (ultrasound, contrast-enhanced computed tomography (CT), and whole-body diffusion-weighted magnetic resonance imaging (WB-DWI/MRI) in assessing peritoneal carcinomatosis and predicting non-resectability in tubo-ovarian carcinoma patients.

Methods

This was a prospective multicenter observational study. We considered all patients with suspected primary ovarian/tubal/peritoneal cancer who underwent preoperative ultrasound, CT, and WB-DWI/MRI (if available). The optimal cut off value for assessing the performance of the methods in predicting non-resectability was identified at the point at which the sensitivity and specificity were most similar. The reference standard to predict non-resectability was surgical outcome in terms of residual disease >1 cm or surgery not feasible. Agreement between imaging methods and surgical exploration in assessing sites included in the PCI score was evaluated using the Intraclass Correlation Coefficient (ICC).

Results

242 patients were included from January 2020 until November 2022. The optimal PCI cut-off for predicting non-resectability for surgical exploration was >12, which achieved the best AUC of 0.87, followed by ultrasound with a cut-off of >10 and AUC of 0.81, WB-DWI/MRI with a cut-off of >12 and AUC of 0.81, and CT with a cut-off of >11 and AUC of 0.74. Using ICC, ultrasound had very high agreement (0.94) with surgical PCI, while CT and WB-DWI/MRI had high agreement (0.86 and 0.87, respectively).

Conclusion

Ultrasound performed by an expert operator had the best agreement with surgical findings compared to WB-DWI/MRI and CT in assessing radiological PCI. In predicting non-resectability, ultrasound was non-inferior to CT, while its non-inferiority to WB-DWI/MRI was not demonstrated.