Eur Radiol . 2018 Oct;28(10):4265-4273. doi: 10.1007/s00330-018-5410-6. Epub 2018 Apr 20.
Giovanni Marchegiani, Valentina Todaro, Enrico Boninsegna, Riccardo Negrelli, Binit Sureka, Debora Bonamini, Roberto Salvia, Riccardo Manfredi, Roberto Pozzi Mucelli, Claudio Bassi
Objectives: To assess factors associated with radical resection (R0) of pancreatic ductal adenocarcinoma (PDAC) after induction treatment with FOLFIRINOX.
Methods: Patients with either locally advanced (LA) and borderline resectable (BR) PDAC undergoing surgical exploration after FOLFIRINOX were retrospectively enrolled. Two pancreatic radiologists reviewed the CT blinded to the final outcome and assessed chemotherapy response and resectability. Patients were then divided into R0 resected (group A) and not resected/R1 resected (group B), which were compared.
Results: Of 59 patients included, 19 were defined as unresectable (32%), 33 borderline resectable (56%) and 7 resectable (12%) during the blind radiological evaluation after FOLFIRINOX. Once in a surgical setting, 27% were non-resectable, whereas 73% received surgical resection with a 70% R0 rate. Consequent sensitivity and specificity were 86% and 29%. At imaging review, significant decreases in longest tumour dimension were observed in both groups: from 32 mm (95% CI 15-55) to 21 (10-44) in group A and from 34 (18-70) to 26 (7-60) in group B, p < 0.05. However, a significant increase in tumour attenuation in all phases was only observed for R0 resected, from 52 HU (26-75) to 65 (35-92) in arterial phase (p < 0.001) and from 62 (36-96) to 78 (40-120) in the venous (p = 0.001).
Conclusion: After neoadjuvant FOLFIRINOX, CT predicted resectability with acceptable sensitivity but low specificity. The observation of increased tumour attenuation at CT scan after FOLFIRINOX treatment might represent a reliable predictor of R0 resection.
Read Full Article Here: https://doi.org/10.1007/s00330-018-5410-6