Article Text

Download PDFPDF
PWE-110 Breathomics – distinguishing minimal from overt hepatic encephalopathy using volatile organic compound analysis: a pilot study
  1. J Covington1,
  2. M Mcfarlane2,
  3. L Kho1,
  4. S Wurie2,
  5. N O'Connell2,
  6. R Savage3,
  7. C Nwokolo2,
  8. KD Bardhan4,
  9. R Arasaradnam2,5
  1. 1School of Engineering, University of Warwick
  2. 2Gastroenterology, UHCW
  3. 3Systems Biology Centre, University of Warwick, Coventry
  4. 4Gastroenterology, Rotherham General Hospital, Rotherham
  5. 5CSRI, University of Warwick, Coventry, UK


Introduction Hepatic encephalopathy (HE) is a neurological manifestation of decompensated liver disease which develops in approximately 50% of patients with cirrhosis. The current diagnostic challenge presented by Hepatic encephalopathy is the detection of minimal HE, as opposed to the more clinically apparent overt HE. Rifaximin, licensed for overt HE, is an effective therapy for these patients, but earlier identification and treatment of HE could prevent liver disease progression and hospitalisation. We conducted a pilot study to analyse the breath samples of patients with different HE grades, alongside the breath of healthy controls, using a portable type electronic nose (uvFAIMS).

Method 42 patients were enrolled; 22 with HE and 20 healthy controls. Breath samples were captured at the bedside using a Warwick designed breath capture device. The samples were then analysed using an ultra violet FAIMS machine. This uses ultra violet light to energise electrons rather than ionising radiation in the traditional FAIMS devices. West Haven criteria were applied and MELD scores calculated. Data was analysed using a previously developed pipeline based on a 2D wavelet transform and threshold, removing background noise. Sensitivity, specificity, and Area Under Receiver Operator Curve (AUROC) were calculated in 10-fold cross validation and reported using sparse logistic regression.

Results 12 patients had minimal HE and 10 had covert/overt HE. Classification of HE vs controls showed a sensitivity of 0.88 (0.73 – 0.95) and specificity of 0.68 (0.51 – 0.81), AUROC 0.84 (0.75–0.93). Minimal HE was distinguished from covert/overt HE with a sensitivity of 0.79 (0.49 – 0.95) and specificity of 0.50 (0.37 – 0.63), AUROC 0.71 (0.57–0.84). There was no statistically significant differences between differing HE grades; AUROC 0.61 (0.43 – 0.79).

Conclusion This pilot study has confirmed the potential of detection and diagnosis of HE via breath analysis identification of VOCs signatures. Importantly this was performed utilising a non-invasive, portable bedside device and holds potential for future early diagnosis of minimal or covert HE.

Disclosure of interest None Declared.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.