Breath analyser shows promise for sniffing out high-risk MASLD

Exhaled breath analysis using the so-called eNose technology not only accurately identifies individuals with metabolic dysfunction-associated steatotic liver disease (MASLD) but also pinpoints those at high risk of accelerated disease progression.

In a proof-of-concept study, the eNose distinguished between healthy individuals and patients with MASLD with 100-percent sensitivity (p<0.001, cross-validation verification of 96 percent), independent of age and sex. [Aliment Pharmacol Ther 2025;doi:10.1111/apt.70176]

Unbiased clustering of breath profiles led to the identification of a subgroup of MASLD patients with a significantly worse prognosis. Over a 5-year follow-up, a higher number of patients in cluster 2 vs clusters 1 and 3 had disease progression (42 percent vs 17 percent and 0 percent, respectively; p=0.03), experienced liver-related events (17 percent vs 4.3 percent and 0 percent; p=0.06), and died from any cause (12.5 percent vs 0 percent and 0 percent).

Disease progression was defined as either a transition from noncirrhotic MASLD to overt cirrhosis or the development of a new liver decompensation event in patients with compensated cirrhosis.

Baseline clinical and demographic characteristics were similar across the three clusters, except for a significantly higher serum hyaluronic acid levels (median 140 vs 82 and 56 μg/L; p=0.003) and greater percentage of patients showing endoscopic signs of portal hypertension (67 percent vs 30 percent and 0 percent; p<0.001) in cluster 2 vs clusters 1 and 3.

The analysis included 30 healthy volunteers (median age 31 years, 70 percent female, median BMI 22 kg/m²), 30 patients with Child-Turcotte-Pugh class A MASLD cirrhosis (median age 70 years, 30 percent female, median BMI 34 kg/m²), and 30 patients with noncirrhotic MASLD (median age 62 years, 33 percent female, median BMI 35 kg/m²).

“This is the first study to establish clusters of patients with MASLD based on unbiased exhaled breath volatile organic compound (VOC) analysis by eNose technology,” said lead researcher Dr Rohit Sinha from the University of Edinburgh, Edinburgh, UK, and colleagues.

The VOC composition in the exhaled breath reflects the disturbed metabolism in MASLD not captured through traditional approaches, and an unbiased clustering approach allows for the grouping of patients based solely on molecular mechanisms and pathophysiology rather than traditional fibrosis scores or clinical comorbidities, Sinha and colleagues explained.

The results of the proof-of-concept study are “encouraging,” highlighting the potential of eNose with unbiased clustering approach in routine clinical practice, they said. This method could be used to noninvasively screen for MASLD in primary care settings, as well as for risk stratification or prognostication of patients with known MASLD. “This could allow targeting of at-risk patients for earlier intervention and enhanced monitoring.”

In an accompanying editorial, Dr Takefumi Kimura from Shinshu University School of Medicine, Matsumoto, Japan, emphasized that the study by Sinha and colleagues is a crucial first step in bringing “breathomics” to the field of hepatology, where existing noninvasive biomarkers often lack precision. [Aliment Pharmacol Ther 2025;doi:10.1111/apt.70185]

“Breathomics offers a promising approach of risk detection that goes beyond simple anthropometric or biochemical assessments. [And the findings of the present study] demonstrate that exhaled breath profiling, combined with unbiased clustering, can uncover high-risk MASLD phenotypes invisible to traditional clinical evaluation,” Kimura noted. “With further validation, this noninvasive approach could reshape early risk stratification and support more personalised management in MASLD.”