APRI Score as a Predictor of Significant Liver Fibrosis in Chronic Hepatitis B

• Published in: Open forum infectious diseases 2017-10, Vol.4 (suppl_1), p.S196-S196
• Main Authors: Ayed, Houda Ben, Koubaa, Makram, Yaich, Sourour, Mejdoub, Yosra, Smaoui, Fatma, Jemaa, Tarak Ben, Maaloul, Imed, Marrakchi, Chakib, Dammak, Jamel, Jemaa, Mounir Ben
• Format: Article
• Language: English
• Subjects: Abstracts
• ISSN: 2328-8957; 2328-8957
• DOI: 10.1093/ofid/ofx163.375

Abstract Background Chronic hepatitis B (CHB) is a global public health problem. Histologic staging of liver fibrosis is crucial to identify patients who need antiviral treatment. As an alternative to percutaneous liver biopsy (PLB), Aminotransferase Platelet Ratio index (APRI) score was recently validated by the WHO. We aimed to evaluate the performance of APRI score in predicting significant fibrosis (SF) in patients with CHB. Methods We conducted a retrospective study including 179 patients aged ≥ 15 years with documented CHB and who underwent a PLB during the period 2008-2016. A SF was defined according to Metavir score (≥ F2). ROC curves assessed the performance of APRI score in predicting SF. We estimated PLB cost of 60 Dollars in our institution. Results Mean age of patients was 37.6 ± 10 years and sex ratio was 1.48. There were 93 patients with SF (52%) who had a high level of aspartate aminotransferase (ASAT) (71.4 ± 38 vs. 34 ± 16 IU/L; P < 0.001) but a low level of platelet count (195 ± 53 vs. 208 ± 52 G/L; P = 0.04). APRI score was significantly higher in patients with SF (1.1 ± 0.7 vs. 0.48 ± 0.26; P < 0.001). Multivariate analysis using logistic regression showed that only APRI score was independently predictive of a SF (HR = 3.78, CI95% 1.23-11.6; P = 0.02). APRI predicted accurately SF with an Area Under the Receiving Operating Curve (AUROC) of 0.7 (CI 95% 0.62–0.77; P < 0.001). At a threshold of 0.5, APRI had a sensitivity of 62%, a specificity of 68%, a positive predictive value of 64.4% and a negative predictive value of 60.7%. The number of avoided PLB with APRI score was 112 PLB with a diagnostic accuracy of 62.5%. Subsequently, the saved cost was estimated to be around 6720 Dollars. Conclusion APRI score was well performing in predicting SF in patients with CHB. This could be of paramount importance particularly in developing countries given that this non-invasive score may help to assess liver fibrosis in CHB. Larger scale and analytic prospective studies are required in order to strengthen the accuracy of this score. Disclosures All authors: No reported disclosures.