WHICH CORRELATION AMONG HISTOLOGICAL ASSESSMENT OF FIBROSIS, NON-INVASIVE MARKERS AND PORTAL HYPERTENSION IN HCV TRANSPLANTED PATIENTS?

Introduction HVPG is an established marker of portal hypertension, while transient elastography (TE) has been proposed as a non-invasive marker of portal hypertension/disease progression. Collagen proportionate area (CPA),1 a histological quantitation of liver collagen, was recently shown to be a better correlate with HVPG than Ishak stage. The aim of this study is to evaluate relationships between liver collagen, TE, APRI score, HVPG and Ishak stage. Methods Consecutive HCV transplanted patients had transjugular liver biopsies combined with HVPG, TE and CPA. Linear regression analysis was used to assess correlation: CPA and TE, and CPA, TE and HVPG. ROC curves to assess sensitivities/specificities of each factor. Results 58 HCV transplanted patients: 47 males, mean age 53 years, mean time 50±39 (9–169) months from LT. Median liver stiffness (LSM) 11.3 kPa (3.1–50.3), mean HVPG 5 mm Hg (0–32), mean CPA 7.9% (1–45), mean APRI score 0.06 (0.01–0.7). Correlation between LSM and CPA r2=0.58, (p<0.000), between LSM and HVPG r2=0.16; (p=0.002) and CPA and HVPG r2=0.28(p<0.000). Univariately CPA, Ishak stage, TE (all p<0.0001), APRI score (p=0.04) were associated with HVPG≥6 mm Hg; multivariately CPA was the only independent factor (OR 1.336, 95% CI 1.136 to 1.571, p=0.000). Univariately CPA, Ishak stage, TE and APRI score were also associated with HVPG≥10 mm Hg but multivariately only CPA (OR 1.1, 95% CI 1.02 to 1.2, p=0.020). For HVPG≥6 mm Hg, AUROC for CPA was 0.91(95% CI 0.83 to 0.99). The best cut-off was 8.7% (78% sensitivity and 87% specificity); AUROC for stiffness was 0.83 (95% CI 0.70 to 0.96), cut-off of 10.9 kPa had 78% sensitivity and 87% specificity. For CPA >8%, the AUROC for stiffness was 0.98(95% CI 0.96 to 1.011). A cut-off of 9.6 kpa could predict CPA>8% (93% sensitivity and 99% specificity). For stiffness>11 kPa, AUROC for CPA was 0.96(95% CI 0.89 to 1.018) with a cut-off of 8.4% (87% sensitivity and 99% specificity). Conclusion CPA as a quantitative measure of liver collagen had a better correlation with TE than HVPG. Thus it may represent a better histological index for TE rather than stages which are descriptive categories and not quantitative assessments. CPA was better than Ishak stage or TE in predicting portal hypertension. CPA should be evaluated to subclassify cirrhosis.