Background: The Cancer of the Liver Italian Program (CLIP) staging system for hepatocellular carcinoma (HCC) was subdivided into 36 subgroups. We aimed to validate the prognostic value of CLIP scoring.
Methods: This study included 3868 HCC cases treated between 1986 and 2002. Survival and prognostic impact of all subgroups were analysed.
Results: In primary CLIP, comparisons of each score showed a significant difference (P<0.001) and exhibited a linear trend (P<0.001). A CLIP score of 0 was used as control group. Portal vein thrombosis, Child–Pugh B, α-fetoprotein (AFP) ≥400 ng/ml and multinodular with tumour extension ≤50% of the four subgroups with a CLIP score of 1 exhibited decreasing univariate hazard ratios and 95% confidence intervals, with values of 2.99 (2.05–4.37), 2.39 (2.00–2.86), 1.66 (1.40–1.96) and 1.39 (1.18–1.63) respectively. Homogeneity in the same score was evaluated by comparing subgroup survival curves. For scores 1–5, 83.3% (5/6), 57.1% (16/28), 24.4% (11/45), 3.6% (1/28) and 16.7% (1/6) pairs of survival curves significantly differed, respectively, with decreasing linear trend (P<0.001).
Conclusion: Different prognostic weighting of four predictive factors caused intrascore heterogeneity. Lower CLIP scores were associated with increased differences in intrascore. In conclusion, the CLIP staging scoring system is a reasonable ordinal scale, but the clinician must be aware of the heterogeneity of mortality risk within a given score.