NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis

Paolo Montuschi, Debora Paris, Dominique Melck, Vincenzina Lucidi, Giovanni Ciabattoni, Valeria Raia, Cecilia Calabrese, Andrew Bush, Peter J. Barnes, Andrea Motta

Research output: Contribution to journalArticle

96 Citations (Scopus)

Abstract

Background: Metabolomics could provide new insights into the pathophysiology of cystic fibrosis (CF) by identifying profiles of endogenous metabolites. Objectives: To investigate whether metabolomics of exhaled breath condensate could discriminate between patients with unstable CF, stable CF and healthy subjects, and whether selected metabolites were responsible for between-group differences. Methods: Twenty-nine patients with stable CF, 24 with unstable CF and 31 healthy subjects (age 9-24 years) participated in a cross-sectional study. Metabolomics was performed with high-resolution nuclear magnetic resonance spectroscopy. Partial least squares-discriminant analysis was used as classifier. The results were validated in a second independent study. Results: Intraclass correlation coefficients for betweenday and technical repeatability were 0.93 and 0.96, respectively. Bland-Altman analysis showed good within-day repeatability. Correct classification rate of CF (n=53) vs healthy subjects (n=31) was 96% (R 2=0.84; Q 2=0.79). Model validation with a testing sample set obtained from subjects not included in the primary analysis (23 CF and 25 healthy subjects) showed a sensitivity of 91% and a specificity of 96%. The classification rate of stable CF (n=29) vs unstable CF patients (n=24) was 95% (R 2=0.82; Q 2=0.78). Model external validation in 14 patients with stable CF and 16 with unstable CF showed a sensitivity of 86% and a specificity of 94%. Ethanol, acetate, 2-propanol and acetone were most discriminant between patients with CF and healthy subjects, whereas acetate, ethanol, 2-propanol and methanol were the most important metabolites for discriminating between patients with stable and unstable CF. Conclusions: Nuclear magnetic resonance spectroscopy of exhaled breath condensate is reproducible, discriminates patients with CF from healthy subjects and patients with unstable CF from those with stable CF, and identifies the metabolites responsible for between-group differences.

Original languageEnglish
Pages (from-to)222-228
Number of pages7
JournalThorax
Volume67
Issue number3
DOIs
Publication statusPublished - Mar 2012

Fingerprint

Metabolomics
Cystic Fibrosis
Magnetic Resonance Spectroscopy
Healthy Volunteers
2-Propanol
Acetates
Ethanol
Discriminant Analysis
Acetone
Least-Squares Analysis

ASJC Scopus subject areas

  • Pulmonary and Respiratory Medicine

Cite this

NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis. / Montuschi, Paolo; Paris, Debora; Melck, Dominique; Lucidi, Vincenzina; Ciabattoni, Giovanni; Raia, Valeria; Calabrese, Cecilia; Bush, Andrew; Barnes, Peter J.; Motta, Andrea.

In: Thorax, Vol. 67, No. 3, 03.2012, p. 222-228.

Research output: Contribution to journalArticle

Montuschi, P, Paris, D, Melck, D, Lucidi, V, Ciabattoni, G, Raia, V, Calabrese, C, Bush, A, Barnes, PJ & Motta, A 2012, 'NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis', Thorax, vol. 67, no. 3, pp. 222-228. https://doi.org/10.1136/thoraxjnl-2011-200072
Montuschi, Paolo ; Paris, Debora ; Melck, Dominique ; Lucidi, Vincenzina ; Ciabattoni, Giovanni ; Raia, Valeria ; Calabrese, Cecilia ; Bush, Andrew ; Barnes, Peter J. ; Motta, Andrea. / NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis. In: Thorax. 2012 ; Vol. 67, No. 3. pp. 222-228.
@article{2ff8dfbe806c41c68c49602268f65f7c,
title = "NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis",
abstract = "Background: Metabolomics could provide new insights into the pathophysiology of cystic fibrosis (CF) by identifying profiles of endogenous metabolites. Objectives: To investigate whether metabolomics of exhaled breath condensate could discriminate between patients with unstable CF, stable CF and healthy subjects, and whether selected metabolites were responsible for between-group differences. Methods: Twenty-nine patients with stable CF, 24 with unstable CF and 31 healthy subjects (age 9-24 years) participated in a cross-sectional study. Metabolomics was performed with high-resolution nuclear magnetic resonance spectroscopy. Partial least squares-discriminant analysis was used as classifier. The results were validated in a second independent study. Results: Intraclass correlation coefficients for betweenday and technical repeatability were 0.93 and 0.96, respectively. Bland-Altman analysis showed good within-day repeatability. Correct classification rate of CF (n=53) vs healthy subjects (n=31) was 96{\%} (R 2=0.84; Q 2=0.79). Model validation with a testing sample set obtained from subjects not included in the primary analysis (23 CF and 25 healthy subjects) showed a sensitivity of 91{\%} and a specificity of 96{\%}. The classification rate of stable CF (n=29) vs unstable CF patients (n=24) was 95{\%} (R 2=0.82; Q 2=0.78). Model external validation in 14 patients with stable CF and 16 with unstable CF showed a sensitivity of 86{\%} and a specificity of 94{\%}. Ethanol, acetate, 2-propanol and acetone were most discriminant between patients with CF and healthy subjects, whereas acetate, ethanol, 2-propanol and methanol were the most important metabolites for discriminating between patients with stable and unstable CF. Conclusions: Nuclear magnetic resonance spectroscopy of exhaled breath condensate is reproducible, discriminates patients with CF from healthy subjects and patients with unstable CF from those with stable CF, and identifies the metabolites responsible for between-group differences.",
author = "Paolo Montuschi and Debora Paris and Dominique Melck and Vincenzina Lucidi and Giovanni Ciabattoni and Valeria Raia and Cecilia Calabrese and Andrew Bush and Barnes, {Peter J.} and Andrea Motta",
year = "2012",
month = "3",
doi = "10.1136/thoraxjnl-2011-200072",
language = "English",
volume = "67",
pages = "222--228",
journal = "Thorax",
issn = "0040-6376",
publisher = "BMJ Publishing Group",
number = "3",

}

TY - JOUR

T1 - NMR spectroscopy metabolomic profiling of exhaled breath condensate in patients with stable and unstable cystic fibrosis

AU - Montuschi, Paolo

AU - Paris, Debora

AU - Melck, Dominique

AU - Lucidi, Vincenzina

AU - Ciabattoni, Giovanni

AU - Raia, Valeria

AU - Calabrese, Cecilia

AU - Bush, Andrew

AU - Barnes, Peter J.

AU - Motta, Andrea

PY - 2012/3

Y1 - 2012/3

N2 - Background: Metabolomics could provide new insights into the pathophysiology of cystic fibrosis (CF) by identifying profiles of endogenous metabolites. Objectives: To investigate whether metabolomics of exhaled breath condensate could discriminate between patients with unstable CF, stable CF and healthy subjects, and whether selected metabolites were responsible for between-group differences. Methods: Twenty-nine patients with stable CF, 24 with unstable CF and 31 healthy subjects (age 9-24 years) participated in a cross-sectional study. Metabolomics was performed with high-resolution nuclear magnetic resonance spectroscopy. Partial least squares-discriminant analysis was used as classifier. The results were validated in a second independent study. Results: Intraclass correlation coefficients for betweenday and technical repeatability were 0.93 and 0.96, respectively. Bland-Altman analysis showed good within-day repeatability. Correct classification rate of CF (n=53) vs healthy subjects (n=31) was 96% (R 2=0.84; Q 2=0.79). Model validation with a testing sample set obtained from subjects not included in the primary analysis (23 CF and 25 healthy subjects) showed a sensitivity of 91% and a specificity of 96%. The classification rate of stable CF (n=29) vs unstable CF patients (n=24) was 95% (R 2=0.82; Q 2=0.78). Model external validation in 14 patients with stable CF and 16 with unstable CF showed a sensitivity of 86% and a specificity of 94%. Ethanol, acetate, 2-propanol and acetone were most discriminant between patients with CF and healthy subjects, whereas acetate, ethanol, 2-propanol and methanol were the most important metabolites for discriminating between patients with stable and unstable CF. Conclusions: Nuclear magnetic resonance spectroscopy of exhaled breath condensate is reproducible, discriminates patients with CF from healthy subjects and patients with unstable CF from those with stable CF, and identifies the metabolites responsible for between-group differences.

AB - Background: Metabolomics could provide new insights into the pathophysiology of cystic fibrosis (CF) by identifying profiles of endogenous metabolites. Objectives: To investigate whether metabolomics of exhaled breath condensate could discriminate between patients with unstable CF, stable CF and healthy subjects, and whether selected metabolites were responsible for between-group differences. Methods: Twenty-nine patients with stable CF, 24 with unstable CF and 31 healthy subjects (age 9-24 years) participated in a cross-sectional study. Metabolomics was performed with high-resolution nuclear magnetic resonance spectroscopy. Partial least squares-discriminant analysis was used as classifier. The results were validated in a second independent study. Results: Intraclass correlation coefficients for betweenday and technical repeatability were 0.93 and 0.96, respectively. Bland-Altman analysis showed good within-day repeatability. Correct classification rate of CF (n=53) vs healthy subjects (n=31) was 96% (R 2=0.84; Q 2=0.79). Model validation with a testing sample set obtained from subjects not included in the primary analysis (23 CF and 25 healthy subjects) showed a sensitivity of 91% and a specificity of 96%. The classification rate of stable CF (n=29) vs unstable CF patients (n=24) was 95% (R 2=0.82; Q 2=0.78). Model external validation in 14 patients with stable CF and 16 with unstable CF showed a sensitivity of 86% and a specificity of 94%. Ethanol, acetate, 2-propanol and acetone were most discriminant between patients with CF and healthy subjects, whereas acetate, ethanol, 2-propanol and methanol were the most important metabolites for discriminating between patients with stable and unstable CF. Conclusions: Nuclear magnetic resonance spectroscopy of exhaled breath condensate is reproducible, discriminates patients with CF from healthy subjects and patients with unstable CF from those with stable CF, and identifies the metabolites responsible for between-group differences.

UR - http://www.scopus.com/inward/record.url?scp=84857440052&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857440052&partnerID=8YFLogxK

U2 - 10.1136/thoraxjnl-2011-200072

DO - 10.1136/thoraxjnl-2011-200072

M3 - Article

C2 - 22106016

AN - SCOPUS:84857440052

VL - 67

SP - 222

EP - 228

JO - Thorax

JF - Thorax

SN - 0040-6376

IS - 3

ER -