PP-011 Collagen proportionate area: a continuous quantitative of histological collagen has the best correlation with transient elastography

IntroductionLiver stiffness (LS) measured by transient elastography (TE) is related to liver fibrosis, which histologically is categorised into stages which do not have quantitatively relationships. Collagen proportionate area (CPA) measured by computer-assisted digital image analysis (DIA) on biops...

Full description

Saved in:
Bibliographic Details
Published inGut Vol. 59; no. Suppl 1; p. A44
Main Authors Isgro, G, Luong, T, Andreana, L, Garcovich, M, Maimone, S, Manousou, P, Calvaruso, V, Tsochatzis, E, Patch, D, Thornburn, D, Dhillon, A, Burroughs, A
Format Journal Article
LanguageEnglish
Published London BMJ Publishing Group LTD 01.04.2010
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:IntroductionLiver stiffness (LS) measured by transient elastography (TE) is related to liver fibrosis, which histologically is categorised into stages which do not have quantitatively relationships. Collagen proportionate area (CPA) measured by computer-assisted digital image analysis (DIA) on biopsies stained with Sirius red is an histological method for quantifying collagen in liver biopsies. The aim of this study is to correlate LS to Ishak score and CPA in patients with chronic viral hepatitis.MethodsLS was measured by TE in 100 patients with chronic viral hepatitis (HCV, HBV).ResultsWe evaluated 100 patients: 58 male, mean age 46.4 (±11.5).HCV (n=51), mean age 48.8, 28 males, mean CPA (6.4±6.6), mean LS 12.6 (±15.1). Univariately CPA (r2 0.57 p<0.0001), METAVIR (r2 0.37 p< 0.0001) and Ishak (r2 0.37 p< 0.0001), were correlated with LS but multivariately only CPA was independently associated (p<0.0001, 95% CI 1.139 to 2.789). Ishak and METAVIR were both correlated with CPA univariately, but multivariately only Ishak (p<0.0001, 95% CI 1.544 to 4.135). HBV (n=49), mean age 43.9, 29 males, mean CPA 4.09 (±4.5), mean LS (8.1 ±6.5). Univariately, LS correlated with CPA (r2 0.59, p<0.000), Ishak (r2 0.49, p<0.0001) and METAVIR (r2 0.55, p< 0.000), mutivariately with CPA (p=0.002, 95% CI 0.305 to 1.286) and METAVIR (p=0.03, 95% CI 0.441 to 9.446). METAVIR (r20.69, p< 0.0001) and Ishak (r20.70, p<0.0001) correlated with CPA univariately, but not multivariately. In HCV patients with Ishak stage >4 the AUROC for CPA was 0.96 (95% CI 0.90 to 1.03) and the best cut off was 11.7 (90%sensitivity, 100%specificity). For METAVIR F4 AUROC was 0.98 (95% CI 0.96 to 1.01) and the best cut off for CPA was 12.4 (87%sensitivity, 99%specificity). In HBV patients, the AUROC for CPA was absolute (ie, 1, 95% CI 1.00 to 1.00) in both patients with Ishak >4 and patients with METAVIR F4. In the first case the best cut off was 8.9 (100%sensitivity, 100%specificity), in the latter the best cut off for CPA was 10.3 (100%sensitivity, 100%specificity).ConclusionIn chronic viral hepatitis, the measurement of CPA which is a continuous quantitative measurement of liver collagen was correlated well with current stage systems. CPA had a much better correlation with TE than METAVIR or Ishak stages. CPA should replace traditional histological staging systems for quantifying fibrosis in comparison with TE.
ISSN:0017-5749
1468-3288
DOI:10.1136/gut.2009.209015c