Background & aims: Patients with hepatocellular carcinoma (HCC) in a palliative setting have a poor prognosis despite recent therapeutic progress. Several prognostic scores, such as the BCLC and the CLIP, have been shown to be useful in helping select treatment options ranging from transplantation to palliative care. However, the discriminatory ability of these scores is inadequate in palliative settings, which concern about 70% of HCC patients. In this paper, we propose and validate a new prognostic score for patients in the palliative setting.
Methods: The prognostic score was developed on a set of 416 patients from a negative randomized clinical trial conducted by the Fédération Francophone de Cancers Digestifs. It was then subsequently validated on a second set of 271 patients from another negative trial. Backward selection was used to identify independent baseline characteristics. Measures of discrimination and predictive values were computed to assess the quality of the developed score. Comparisons with the BCLC and the CLIP - with and without the WHO performance status (PS) score - were performed.
Results: Tumour morphology, portal vein obstruction, metastasis, ascites, jaundice, alpha-foetoprotein, and serum alkaline phosphatase were included in the final score. From the training dataset, three groups of increasing risk were defined, and these were associated with hazard ratios (HR) of 2.13 and HR = 5.72. Similar results were obtained on the validation dataset. This score provides a better discriminatory ability than BCLC and CLIP in this setting. Unfortunately, absolute performances for these scores remain poor.
Conclusions: The new prognostic score and CLIP + PS are recommended in palliative settings. However, new prognostic variables are necessary.
Copyright © 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.