Abstract Content

With respect to the requirements of non-invasiveness and bedside applicability, breath represents a promising alternative to serum chemistry and other invasive diagnostic procedures such as bronchoscopy or biopsy. For more than 20 years researchers have spent much effort in searching volatile marker substances which are specifically linked to a diseased state. Many substances have been presented as biomarker candidates. Many of these results could never be reproduced, none of them made it into clinical practice.In most cases, correlations between biomarkers and corresponding diseases are reasonably good only as long as disease prevalence is high in the population under investigation. When the prevalence is low, as it is usually when a normal population has to be screened, false negatives and false positives can induce crucial problems. Biomarkers have not only to be sensitive and specific but prognosis of the underlying disease has also to be influenced significantly by early diagnosis. If more than one parameter is used to identify disease specific patterns correlations found in this way may be accidental or may be produced by the fact that too many parameters and too few distinct measurements were used. So, is it hopeless to try to use breath analysis in clinical practice?

YES, if we continue publishing non-randomized, non-prospective studies with low patient numbers advertising the umpteenth version of some fancy biomarkers.

NO, if we realize the strong points of breath analysis and if we efficaciously use the knowledge that has been accumulated during the last 20 years:

• Preconcentration techniques were improved requiring now not more than a few cc of exhaled air to do the analyses (SPME, NTME), GC technology progressed in the way that enhanced and miniaturized devices yield reliable substance separation within a few minutes (fast GC, GC/GC, GC on the chip).

• Modern mass spectrometry (ion trap, MS/MS, TOF-MS) enables detection and substance

Abstract Content

Background: A large number of volatile organic compounds (VOCs) are emitted by bacteria and cells. Production and consumption of these VOCs fundamentally depend on metabolic status of the cultures and on composition of culture media. In addition, media themselves may emit volatile compounds. In order to exploit the potential of headspace analysis of biological cultures, this study was intended to systematically assess how different cell types and proliferation status impact onto VOC profiles.

Methods: Two fast proliferating murine stem cell lines and one slow proliferating murine fibroblast cell line were monitored over 3 or 7 days. VOC pre-concentration was done from a headspace volume of 20 ml by means of polymer needle trap devices (NTDs) every 24h. Pure media samples were analyzed in parallel. Susbtances were thermally desorbed from the NTDs, separated and identified by means of GC-MS (Agilent 7890A gas chromatograph coupled to an Agilent 5975C inert XL MSD with triple axis detector).

Results: Both murine stem cell lines emitted increasing concentrations of thiirane and methyl-methoxyhydroxymethyl- amine (MMHA) over a period of three days. Murine fibroblasts did not emit thiirane or MMHA

at any time point of headspace analysis. Concentrations of aldehydes –especially benzaldehydes–were higher in pure media samples than in all cell cultures. The production of thiirane and MMHA correlated with increasing numbers of cells during the linear growth phase.

Conclusions: Thiirane and MMHA specifically indicated stem cell proliferation and may be used to distinguish stem cells and non-stem cells from each other. These results strongly suggest that VOC headspace analysis could be applied for destruction-free monitoring of stem cell proliferation.

Abstract Content

Background: Infections are one leading cause of death worldwide. Influenza is one of the most frequently occurring virus disease. In the last decades, breath analysis raised increasing interest as a non-invasive means for detection of infectious diseases. Since Influenza A infection takes place in the respiratory tract, VOC profiles exhaled from the lung can be expected to change during infection. Influenza A causes infection in pigs as in humans and pigs show sufficiently high similarity to human physiology and genetics. Hence, pigs are well suited as a large animal infection model. This study was intended to characterize breath profiles during an Influenza A infection in spontaneously breathing awake pigs.

Methods: Breath samples were taken from 10 pigs before infection with Influenza A and on day 2, 4, 7 and 14 after infection. In parallel three healthy pigs were analyzed as controls. Pigs’ snout was placed into a breathing mask and breath samples were taken by means of a 50 ml glass syringe. Breath gas was withdrawn under capnometer control through a virus proof filter connected to the mask. VOCs were preconcentrated from 20 ml aliquots onto polymer needle traps. Separation and identification of substances were done by means of gas chromatography-mass spectrometry.

Results: On the day when swaps from pigs’ nasopharynx became positive for Influenza A specific compounds appeared in pigs’ breath or reached higher concentrations than in the control group. After attenuation of the infection these compounds disappeared from breath in the infected group or reached similar concentrations as in the control group.

Conclusions: Volatile compounds in pigs’ breath mirrored onset and attenuation of an Influenza A infection. Hence, VOC analysis may enable a non-invasive disease detection and may be used for infection monitoring.