S1578 Dynamic Limonene Breath Testing Maximizes Classification Performance for Subjects With Cirrhosis

Le résumé fourni par la source

Introduction: The current diagnostic pathway for liver cirrhosis results in up to 75% of the cases diagnosed at advanced stages with decompensation. Chronic liver diseases alter the systemic bioavailability of xenobiotics undergoing phase 1 metabolism. Limonene, one of these xenobiotics, is partially excreted unchanged in the breath, where it can be non-invasively measured to assess alterations associated with chronic liver disease. With an unprecedented study design, dynamic limonene breath analysis was evaluated for classification performance for cirrhosis, and correlations with risk of liver fibrosis and disease severity. Methods: A total of 29 subjects with cirrhosis [Child-Pugh: 22 A, 5 B, 2 NAs] and 29 controls [17 f, 12 m], were enrolled with random recruitment, and orally administered 100 mg limonene after an overnight fasting. Breath samples were collected before, and 20, 40, 60, 90, and 120 minutes after ingestion using the ReCIVA® breath sampler. Absolute limonene quantification was obtained by gas chromatography mass spectrometry. Blood tests were administered within 6 months from breath sampling. Logistic regression with cross-validation was used to assess classification performance. Correlation of breath limonene with risk of fibrosis and disease severity was evaluated using canonical correlation analysis (CCA). Results: All tested timepoints, showed higher levels of breath limonene in subjects with cirrhosis compared to controls (P < 0.01). At baseline limonene discriminated subjects with cirrhosis from controls with an AUC (SD) of 0.87 (0.11). Limonene administration induced a spike on breath of >100 folds with a peak (Cmax) after 20 or 40 minutes (Tmax) in more than 90% of the subjects. AUCs (sd) at post administration timepoints were respectively: 0.89 (0.10); 0.93 (0.06) (Figure 1A); 0.91 (0.08); 0.92 (0.07); 0.94 (0.06). In subjects with cirrhosis, breath limonene levels at Cmax showed a collective correlation with liver fibrosis risk estimated with FIB4 and APRI, and disease severity estimated with MELD score [R2 = 0.76, P < 0.001] (Figure 1B). No correlations were observed in the control group. Conclusion: The dynamic limonene breath analysis demonstrates excellent performance to discriminate cirrhosis. It may also serve to identify fibrosis and monitor disease regression after therapeutic intervention. This premise should be evaluated with/against Fibroscan in patients with various stages and etiologies.Figure 1.: Study results.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.