Comparing recist and Choi’s criteria to evaluate radiological response to chemotherapy in patients with advanced pancreatic cancer.

Abstract

e15069 Background: Assessment of response after chemotherapy (CTH) for pancreatic cancer (PC) is currently based on RECIST criteria. In 2007 Choi et al. published a new classification system.The purpose of this study was to evaluate the accuracy of these classification systems for radiological response to CTH in patients with advanced PC. Methods: From 2006 to 2012, 66 untreated patients with advanced PC underwent palliative CTH. Fourty (60 %) had a locally advanced PC and 26 (40%) a metastatic disease. All patients were treated with a GEM-based CTH or FOLFIRINOX. We assessed radiological response after three months of first-line therapy applying both RECIST criteria, which evaluate differences in CT size, and Choi’s criteria, which consider changes both in size and in density at CT. We evaluated the accuracy in restaging, comparing the class of response with overall survival (OS). OS was calculated with Kaplan-Meier method. The accuracy in restaging was assessed through log rank test and multivariate analysis with Cox Regression. Results: At restaging, using RECIST criteria, we registered 7 (10.6 %) patients with partial response (PR), 32 (48.5 %) with stable disease (SD), and 27 (40.9 %) with disease progression (PD). Instead Choi’s criteria assessed 19 PR (28.8 %), 12 SD (18.2%) and 35 PD (53.0%). Comparing each classification with OS, we observed that patients with different prognosis were better stratified with Choi’s criteria. Using RECIST criteria we found a borderline significant difference in OS between patients with PR (13.47 months), SD (13.67 months) and PD (9.97 months) (p=0.05). Instead we found a significant statistical difference in OS using Choi’s criteria between patient with PR (14 months), SD (16.37 months), PD (9.7 months; p=0.004). Multivariate analysis showed a statistically significant difference in OS between Disease Control Rate (DCR, PR+SD) and PD patients (14.47 vs. 9.67 months, p=0.02), only using Choi’s criteria. Conclusions: In our experience, Choi’s criteria seem to better assess radiological response of CTH in PC patients than RECIST criteria. Due to the small number of patients, larger prospective studies are needed.