Abstract

Background:

Multicapillary column ion-mobility spectrometry (MCC-IMS) may identify volatile components in exhaled gas. The authors therefore used MCC-IMS to evaluate exhaled gas in a rat model of sepsis, inflammation, and hemorrhagic shock.

Methods:

Male Sprague–Dawley rats were anesthetized and ventilated via tracheostomy for 10 h or until death. Sepsis was induced by cecal ligation and incision in 10 rats; a sham operation was performed in 10 others. In 10 other rats, endotoxemia was induced by intravenous administration of 10 mg/kg lipopolysaccharide. In a final 10 rats, hemorrhagic shock was induced to a mean arterial pressure of 35 ± 5 mmHg. Exhaled gas was analyzed with MCC-IMS, and volatile compounds were identified using the BS-MCC/IMS-analytes database (Version 1209; B&S Analytik, Dortmund, Germany).

Results:

All sham animals survived the observation period, whereas mean survival time was 7.9 h in the septic animals, 9.1 h in endotoxemic animals, and 2.5 h in hemorrhagic shock. Volatile compounds showed statistically significant differences in septic and endotoxemic rats compared with sham rats for 3-pentanone and acetone. Endotoxic rats differed significantly from sham for 1-propanol, butanal, acetophenone, 1,2-butandiol, and 2-hexanone. Statistically significant differences were observed between septic and endotoxemic rats for butanal, 3-pentanone, and 2-hexanone. 2-Hexanone differed from all other groups in the rats with shock.

Conclusions:

Breath analysis of expired organic compounds differed significantly in septic, inflammation, and sham rats. MCC-IMS of exhaled breath deserves additional study as a noninvasive approach for distinguishing sepsis from inflammation.

What We Already Know about This Topic
  • Septic patients are not always readily identified by caregivers, and there is a short window for optimum antibiotic treatment

What This Article Tells Us That Is New
  • Exhaled gas from rats given endotoxin compared with the gas from rats with bacterial sepsis was found to be significantly different and different from rats who were in hemorrhagic shock

  • Breath analysis appears to be able to distinguish inflammation from infection

THOUSANDS of volatile organic compounds (VOCs) are expelled with each breath. The composition of these VOCs can change with the state of health.1–10  Breath analysis has consequently become a promising field of research. One suitable tool for breath analysis is multicapillary column ion-mobility spectrometry (MCC-IMS). The MCC-IMS is a noninvasive and rapid method for detection of VOCs in exhaled breath. The principles of MCC-IMS have been described previously.11,12  MCC-IMS is already used extensively for detection of drugs, chemicals, and explosives.13,14  But this technique has already been used for evaluation of pulmonary diseases, that is, chronic obstructive pulmonary disease,1–3  acute respiratory distress syndrome,4  lung cancer, and airway infections.1,6,15  Furthermore, it also appears possible to detect volatile metabolites arising from various bacterial species.7–10 

Based on these results using MCC-IMS for diagnosis of pulmonary diseases and inflammation, the question rises if systemic inflammation and sepsis are also appropriated targets for this novel noninvasive device. To our knowledge, only one study from Guamán et al.15  used gas chromatography–mass spectrometry to evaluate the systemic inflammatory response to intraperitoneal lipopolysaccharide in rats.

But whether MCC-IMS can be used on exhaled breath as a measure of sepsis and systemic inflammation remains unknown. We therefore compared patterns of exhaled organic compounds in septic, endotoxemic shock (ES), and sham-control rats during an observation period of 10 h. A further group of pressure-controlled hypotension was used to distinguish between infection, inflammation, and noninfectious shock in exhaled metabolites. Noninfectious shock was included to distinguish changes in VOCs resulting from infection/inflammation versus hypotension alone.

Material and Methods

Animals

All experiments were conducted with approval from our Animal Care and Use Committee (Landesamt für Soziales, Gesundheit und Verbraucherschutz; Saarbrücken; Germany) and in accordance with the German Animal Welfare Act. Male Sprague–Dawley rats (200 to 300 g body weight) were obtained from Charles River (Sulzfeld, Germany) and kept in the institutional animal facility under controlled conditions (temperature 20° ± 2°C and 50 ± 5% humidity). Animals had free access to water; standard pellet food was withheld for 12 h before the experiment.

Chemicals

Pure lipopolysaccharide of Escherichia coli serotype O26:B6 was obtained from Sigma-Aldrich (Munich, Germany).

Surgical Procedures

Rats were anesthetized with pentobarbital 60 mg/kg intraperitoneally and positioned on a warming plate; an open tracheotomy was performed to facilitate breathing. The right external jugular vein was catheterized to allow for infusion, and the left carotid artery was catheterized for continuous measurement of mean arterial pressure (MAP) and heart rate (Philips IntelliVue MP20 Junior, Boeblingen, Germany).

Afterward the rats were connected to the respirator (KTR-5 small animal ventilator; Hugo Sachs Elektronik-Harvard Apparatus, March-Hugstetten, Germany) and ventilated with highly purified synthetic air (Air Liquid, Ludwigshafen, Germany). The MCC-IMS (type Breath discovery; B&S Analytik, Dortmund, Germany) was connected to the same tank of synthetic air.

Tidal volume was adapted to a physiological lift of the thorax wall and respiratory rate set to 65 min−1 at an inspiration ratio of 45% and a positive end-expiratory pressure of 1 to 2 cm H2O. General anesthesia was maintained throughout the study with pentobarbital intravenously as needed. The rats were given intravenous isotonic solution at a rate of 10 ml kg−1 h−1 (Sterofundin ISO; B. Braun, Melsungen, Germany).

Experimental Protocol

We evaluated four groups of 10 rats each, randomly assigned to sepsis (CLI), ES, hemorrhagic shock (HS), and sham control:

  • CLI: Sepsis was induced using a modified model of cecal ligation and incision as previously described.16,17  The cecum was exteriorized through a midline laparotomy, and the cecum and paracecal blood vessels were ligated below the ileocecal valve. The ligated cecum was incised 1.5 cm on the antimesenteric side and replaced in the abdomen, and the abdominal wall was closed.

  • Sham: Sham-operated rats were treated likewise but without ligation and incision.

  • ES: ES was induced by administration of 10 mg/kg lipopolysaccharide E. coli serotype O26:B6 intravenously.

  • HS: A modified pressure-regulated shock model was used as described preciously,18,19  but without resuscitation and retransfusion. HS was induced by rapid arterial blood withdrawal by way of the carotid artery with a MAP of 35 ± 5 mmHg.

Induction of sepsis or sham operation was carried out by the same investigator and subsequently blinded during the observation period. ES and HS groups were also carried out by the same investigator, but blinding was not realizable. When measurements were finished, all surviving rats were killed with an overdose of pentobarbital.

Arterial blood (0.2 ml) was sampled every 2.5 h for analysis of pH, base excess, lactate, hemoglobin, glucose, partial pressures of oxygen, and carbon dioxide (Radiometer ABL 800 Basic, Willich, Germany). After insertion of the arterial catheter, and 5 h after induction of CLI, blood samples (0.5 ml) were taken for enzyme-linked immunosorbent assay (ELISA) analysis of pro- and antiinflammatory cytokines.

MCC-IMS Measurements

We used a Bioscout 2011 MCC-IMS based on a 550 MBq 63Ni ionization source with an electrical field strength of 300 V/cm and equipped with MCC using a general-purpose, slightly polar standard capillary column phase OV-5 (Multichrom, Novosibirsk, Russia).

Before connecting the rats to the MCC-IMS, we measured background concentrations of organic volatile compounds from the respirator for 1 h. Throughout the experiment, the rats were ventilated with synthetic air as described previously.20  To avoid any extraneous signals, polytetrafluoroethylene (Bohlender, Grünsfeld, Germany) was used for all external connections, and the MCC-IMS sampling tube was directly connected to the exhalatory line of the respirator. Breath humidity and cluster reactions between ions and water lead to significant changes of the IMS spectra; therefore we used a MCC coupled with an ion-mobility spectrometer to avoid this error.

Exhaled gas (10 ml per measurement) was collected in the sample loop before entering the multicapillary column of the ion-mobility spectrometer. This arrangement allowed three measurements per hour. The retention time in relation to the preseparation of the MCC, drift time of the ions within the IMS, and the intensity of the analytes were delivered by the MCC-IMS and subsequently identified using the BS-MCC/IMS-analytes database (Version 1209; B&S Analytik).

Peaks representing various VOCs were identified using Visual Now 3.1 (B&S Analytik). All signals above the threshold of 1 mV to 5 V (peak range: lower limit is defined as five times over the background noise; the upper limit depends on the saturation of the drift tube) were defined as a peak and characterized by their position with drift time, retention time, and concentration related to the peak intensity.

Cytokine Assay

Cytokine concentrations were measured by ELISA. Positive controls of each cytokine were measured routinely with each assay (ELISA Antibodies BD OptEIA; BD Biosciences Pharmingen, San Diego, CA).

Statistical Analysis and Signal Processing

Mathematical and statistical evaluation was carried out using SigmaPlot (version 12.5; Systat Software, Erkrath, Germany).

Based on our previous experience, we confirmed a number of 10 rats per group as sufficient to see clear differences. Animals were randomized in four groups using a random number generator (Excel 2010; Microsoft, Redmond, WA).

Survival time was analyzed using a log-rank model according to Kaplan–Meier, followed by an all-pairwise comparison using Tukey test.

Data were tested for distribution normality (Kolmogorov–Smirnov test) and analyzed using repeated-measures ANOVA, followed by post hoc multiple comparisons with Holm–Sidak method. When appropriate, repeated-measures ANOVA on ranks were used. Based on the survival rate, statistical analysis for each volatile compound was performed from 0 to 6 h in the sham, sepsis, and endotoxemic group.

Hemorrhagic shock group was merely analyzed after 2 h using one-way ANOVA followed by post hoc multiple comparisons with the Holm–Sidak method. A two-tailed P value of less than 0.05 was considered statistically significant.

Results

Survival Time, Blood Gas Analysis, and Cytokines

All sham rats survived the observation period, whereas others died earlier: HS: mean, 2.5 ± 0.3 h; 95% CI, 1.9 to 3.1 h; CLI: mean, 7.9 ± 0.4 h; CI, 7.2 to 8.7 h; and, ES: mean, 9.1 ± 0.4 h; and CI, 8.3 to 9.9 h. All P values less than 0.001 versus sham (fig. 1).

Fig. 1.

Survival rate. After hemorrhagic shock (HS), endotoxemic shock (ES), and sepsis (CLI), mean survival time was significantly decreased compared with sham control. *P < 0.001 for sham versus HS, ES, and CLI.

Fig. 1.

Survival rate. After hemorrhagic shock (HS), endotoxemic shock (ES), and sepsis (CLI), mean survival time was significantly decreased compared with sham control. *P < 0.001 for sham versus HS, ES, and CLI.

All blood gas values remained stable and normal in sham-operated rats. In contrast, blood lactate concentrations increased significantly over time in CLI, ES, and HS rats. Arterial base excess and pH were significantly lower in the CLI and ES groups than sham rats (table 1). Respiratory parameters (partial pressure of oxygen, partial pressure of carbon dioxide, and oxygenation index), hemoglobin, and blood glucose remained within normal ranges throughout the experiment in both groups.

Table 1.

Arterial Blood Gas Analysis

Arterial Blood Gas Analysis
Arterial Blood Gas Analysis

Baseline concentrations of tumor necrosis factor-α, interleukin-6, and interleukin-10 were similar at the beginning of the experiment. However, concentrations of all three increased significantly in the septic and endotoxemic rat (table 2).

Table 2.

Cytokine Response

Cytokine Response
Cytokine Response

Hemodynamic Parameters

Baseline MAP was similar in all groups and remained stable throughout in the sham-operated rats (fig. 2). Induction of HS resulted in a significant decline of MAP. Injection of lipopolysaccharide resulted in an early reduction of MAP from 1 to 2 h and further at 5 h and thereafter. MAP gradually declined in the CLI rats, becoming significantly lower than in the sham group after 5 h. The heart rate remained within normal ranges during the entire observation period, without significant differences between the groups.

Fig. 2.

Mean arterial pressure was significantly lower in the hemorrhagic shock (HS) than in any of the other groups. In endotoxemic shock (ES) and sepsis (CLI) groups, mean arterial pressure decreased over time throughout the experiment compared with sham control. Statistical analysis for sham, CLI, and ES was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Statistical analysis for HS was performed at 2 h: &P < 0.05 for HS versus all other groups. Data are given as means ± SEMs (n = 10 per group at the beginning of the experiment).

Fig. 2.

Mean arterial pressure was significantly lower in the hemorrhagic shock (HS) than in any of the other groups. In endotoxemic shock (ES) and sepsis (CLI) groups, mean arterial pressure decreased over time throughout the experiment compared with sham control. Statistical analysis for sham, CLI, and ES was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Statistical analysis for HS was performed at 2 h: &P < 0.05 for HS versus all other groups. Data are given as means ± SEMs (n = 10 per group at the beginning of the experiment).

Multicapillary Column Ion-mobility Spectrometry

During the experiment, about a hundred different signals between a threshold of 1 mV and 5 V (peak definition) were detected in exhaled air. Seven signals presented with significant differences between groups during the first 6 h of the observation period. Those seven were considered potentially useful VOCs for distinguishing among treatments. Figure 3 illustrates the typical IMS chromatogram of 3-pentanone and a three-dimensional plot for this setting.

Fig. 3.

Ion-mobility spectrometry chromatogram of 3-pentanone. (A) Shows a typical ion-mobility spectrometer chromatogram for, respectively, individual rats in the sham, sepsis (CLI), endotoxemic shock (ES), and hemorrhagic shock (HS) groups after 4 h. (B) A three-dimensional plot of the same rats. Signals above a threshold of 1 mV defined a peak. The peaks were characterized by their position in the ion-mobility spectrometry chromatogram, using the inverse reduced mobility (axis of abscissa; 1/K0-value—which is proportional to the drift time of the ions) and the retention time (axis of ordinates) as well as the peak expansion. Subsequently, peaks were identified according to database BS-multicapillary column/ion-mobility spectrometry-analytes (Version 1209; B&S Analytik, Dortmund, Germany). For visualization, 3-pentanone is labeled as representative peak in this setting. Comparing the 3-pentanone peak, the drift time of 0.551 Vs/cm2, and a retention time of 8.2 s are equal in all chromatograms. The peak intensity after 4 h was 685 mV for sham, 370 mV for CLI, 60 mV for ES, and 585 mV for HS for this peak and is given as difference in color in A and size in B.

Fig. 3.

Ion-mobility spectrometry chromatogram of 3-pentanone. (A) Shows a typical ion-mobility spectrometer chromatogram for, respectively, individual rats in the sham, sepsis (CLI), endotoxemic shock (ES), and hemorrhagic shock (HS) groups after 4 h. (B) A three-dimensional plot of the same rats. Signals above a threshold of 1 mV defined a peak. The peaks were characterized by their position in the ion-mobility spectrometry chromatogram, using the inverse reduced mobility (axis of abscissa; 1/K0-value—which is proportional to the drift time of the ions) and the retention time (axis of ordinates) as well as the peak expansion. Subsequently, peaks were identified according to database BS-multicapillary column/ion-mobility spectrometry-analytes (Version 1209; B&S Analytik, Dortmund, Germany). For visualization, 3-pentanone is labeled as representative peak in this setting. Comparing the 3-pentanone peak, the drift time of 0.551 Vs/cm2, and a retention time of 8.2 s are equal in all chromatograms. The peak intensity after 4 h was 685 mV for sham, 370 mV for CLI, 60 mV for ES, and 585 mV for HS for this peak and is given as difference in color in A and size in B.

With the BS-MCC/IMS-analytes database (Version 1209), all seven potential compounds could be identified according to their retention time and drift time. All these VOCs are shown in table 3.

Table 3.

Identification and Changes in MCC-IMS Peaks

Identification and Changes in MCC-IMS Peaks
Identification and Changes in MCC-IMS Peaks

As a result of the measured survival time (fig. 1), the statistical analysis of each VOC was performed at 0, 2, 4, and 6 h in the sham, CLI, and ES (each n = 10 per group at these time points). Sham-operated rats presented with approximately stable concentrations of all VOCs throughout the observation period. During the experiment, seven VOCs decreased significantly in lipopolysaccharide-treated rats, when compared with sham. At 2 h and thereafter 1-propanol (P < 0.04), butanal (P < 0.001), acetophenone (P < 0.022), 1,2-butandiol (P < 0.001), 3-pentanone (P < 0.001), acetone (P < 0.001), and 2-hexanone (P < 0.001), all decreased (table 3 and figs. 4–6).

Fig. 4.

Peak course of 3-pentanone. For 3-pentanone, sham and hemorrhagic shock (HS) rats exhibited stable signal strength during the experiment. In endotoxemic shock (ES) and sepsis (CLI), 3-pentanone concentrations declined during the observation period. Statistical analysis was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. §P < 0.05 for CLI versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Data are given as means ± SEMs.

Fig. 4.

Peak course of 3-pentanone. For 3-pentanone, sham and hemorrhagic shock (HS) rats exhibited stable signal strength during the experiment. In endotoxemic shock (ES) and sepsis (CLI), 3-pentanone concentrations declined during the observation period. Statistical analysis was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. §P < 0.05 for CLI versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Data are given as means ± SEMs.

Fig. 5.

Peak course of acetone. Examples of an acetone peak declining during hemorrhagic shock (HS), endotoxemic shock (ES), and sepsis (CLI). Statistical analysis was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Data are given as means ± SEMs.

Fig. 5.

Peak course of acetone. Examples of an acetone peak declining during hemorrhagic shock (HS), endotoxemic shock (ES), and sepsis (CLI). Statistical analysis was performed from 0 to 6 h: *P < 0.05 for sham versus CLI. **P < 0.05 for sham versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Data are given as means ± SEMs.

Fig. 6.

Peak course of 2-hexanone. 2-Hexanone increased significantly more after induction of hemorrhagic shock (HS) than other treatments. 2-Hexanone decreased in endotoxemic shock (ES) compared with sham control and sepsis (CLI). Statistical analysis was performed from 0 to 6 h: **P < 0.05 for sham versus ES. §P < 0.05 for CLI versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Statistical analysis for HS was performed at 2 h: &P < 0.05 for HS versus all other groups. Data are given as means ± SEMs.

Fig. 6.

Peak course of 2-hexanone. 2-Hexanone increased significantly more after induction of hemorrhagic shock (HS) than other treatments. 2-Hexanone decreased in endotoxemic shock (ES) compared with sham control and sepsis (CLI). Statistical analysis was performed from 0 to 6 h: **P < 0.05 for sham versus ES. §P < 0.05 for CLI versus ES. #P < 0.05 for CLI versus corresponding baseline. ##P < 0.05 for ES versus corresponding baseline. Statistical analysis for HS was performed at 2 h: &P < 0.05 for HS versus all other groups. Data are given as means ± SEMs.

In the septic rats, though, only two compounds decreased significantly. 3-Pentanone (P < 0.001) and acetone (P = 0.02) decreased at 6 h (table 3 and figs. 4 and 5).

Overall, there were significant differences between ES and CLI in the concentrations of three VOCs: butanal (P < 0.001 at 2 h and thereafter), 3-pentanone (P < 0.001 at 2 h and thereafter), and 2-hexanone (P < 0.001 at 2 h and thereafter (table 3 and figs. 4 and 6).

Noninfectious shock was included to distinguish changes in VOCs resulting from infection or inflammation versus hypotension alone. Due to the high mortality in this group, a single statistical evaluation was performed at 2 h after induction of HS:

There was significant increase in 2-hexanone (P < 0.001) when compared with all other groups, and for butanal (P < 0.006), 3-pentanone (P < 0.001), and acetone (P < 0.001) when compared with endotoxemia. However, none of the VOCs decreased significantly compared with sham (table 3).

Table 4 gives an overview of changes in volatile compounds with significant differences after 6 h.

Table 4.

Overview of Changes in Volatile Compound Profiles

Overview of Changes in Volatile Compound Profiles
Overview of Changes in Volatile Compound Profiles

Discussion

Currently, diagnosis of sepsis and septic shock relies on nonspecific physiological signs, and early appropriate treatment remains of prime importance. Common diagnostic tools, such as blood culture techniques, procalcitonin, ELISA, or polymerase chain reaction, harbor a series of drawbacks and limitations.21–26  MCC-IMS may prove to be an innovative tool to complement conventional diagnostic techniques.15  Major advantages of the technique lie in its noninvasive character and near-real-time bedside analysis.8 

We are aware of only a single previous study in which expired gas was used to detect a systemic inflammatory response.15  In that study, IMS and gas chromatography–mass spectrometry were used to evaluate the systemic inflammatory response to intraperitoneal lipopolysaccharide in rats. Our use of MCC-IMS technology for VOC assessment of sepsis appears to be novel.

We observed a total of 100 signals in exhaled air, between a threshold of 1 mV and 5 V, which was defined as peak range. Seven VOCs presented with significant difference between the investigated groups during the first 6 h of the experiment and have to be chosen as potential VOCs to distinguish between research groups.

All detected compounds could be identified; the most were ketones, hydrocarbons, alcohols, and ethers. Previous investigators, using different methodologies, found similar compounds in the breath of humans,27,28  rats,15  and mice29  and also in the headspace samples of bacteria cultures.7,8,10,30  Our focus, though, was not on peak identity, but in differences between sepsis, inflammation, and shock.

Our experimental model of fulminant four-quadrant peritonitis presumably better represents clinical sepsis than lipopolysaccharide endotoxemia, but also poses several problems. During CLI sepsis, rats presented with three clinical entities: infection, inflammation, and shock. To distinguish among them, we used two further models: one with a systemic inflammatory response induced by lipopolysaccharide and another with a pressure-controlled HS.

In comparison with untreated animals, most volatile compounds detected in sepsis and inflammation animals declined over time, as did hemodynamic compensation. However, none of these compounds similarly declined during HS; in contrast, HS was the only condition associated with an increase in VOCs. Also surprisingly, the “sterile” inflammation after injection of lipopolysaccharide seemed to have more altered the most VOCs than the complex model of polymicrobial sepsis. Thus, the observed alterations in volatile compounds during sepsis are unlikely to be the result of hemodynamic decompensation, but seem more likely due to factors specific to sepsis, inflammation, or consequent pulmonary impairments. Furthermore, all animals presented with a normal oxygenation index during the experiment, so the distinction between groups is unlikely influenced by the degree of lung injury.

A detailed peak examination revealed an early and strongly reduced release of ketones, acetone, and 3-pentanone during sepsis and inflammation. Our findings are consistent with previous reports showing that blood ketone bodies and the ketogenic capacity of the liver are inhibited during sepsis in rats.31,32  These findings seem to be independent of nutritional conditions because fasting before and during sepsis did not result in hypoglycemia.32  Instead, increased tumor necrosis factor-α and interleukin-6 plasma concentrations may mediate ketogenic impairment.33 

Volatile acetone decreased over time in sepsis, inflammation, and HS, suggesting a hemodynamic dependence. But interestingly, other studies reported an increased acetone in exhaled breath of patients with heart failure, associated with an increase in blood brain natriatic peptide concentration.34,35  It thus seems unlikely that reduced acetone concentrations result from cardiac decompensation.

Guamán et al.15  selected acetone and a further 14 volatile compounds as possible compounds that can be potentially used to separate healthy rats from endotoxemic rats 24 h after lipopolysaccharide treatment using solid phase microextraction gas chromatography–mass spectrometry Schubert et al.35  reported that n-pentane concentrations are higher and that acetone concentrations are lower in the blood of mechanically ventilated septic patients than in healthy controls. However, GC/MS revealed no differences in the concentration of either substance in the breath of the patient groups. In contrast, previous studies showed increased concentrations of acetone, dimethyl sulfide, 2-butanone, and 2-pentanone in the breath of patients with liver cirrhosis and in the headspace samples of different bacteria.8,29,36  The extent to which any of these VOCs might serve as useful diagnostic criteria for sepsis thus remains unclear.

Our study is limited in so far as rodent models only partially correlate with human conditions.37  However, the sepsis and the inflammation model we chose are standard ones that have been used in many previous studies. A more important limitation is that we do not know whether expired breath markers of inflammation and sepsis can detect the conditions in humans before they are otherwise clinically apparent.

In summary, MCC-IMS analysis of expired gas may help identify sepsis or inflammation. Future studies should evaluate plausible mechanisms for changes in VOC quantities and which breath profiles best indicate a septic or inflammatory state. Thereafter, breath analysis for sepsis and inflammation needs to be confirmed in humans, and its sensitivity and specificity determined.

Acknowledgments

The authors thank Jan Baumbach, Ph.D. (Department of Mathematics and Computer Science, University of Southern Denmark, Odense, Denmark), for his encouragement and Stefan Wagenpfeil, Ph.D. (Department of Biometry, Epidemiology, and Medical Informatics, University Medical Center, Homburg (Saar), Germany), for his statistical support. This study contains data taken from the thesis presented by Nathalie Heim (Department of Anesthesiology, Intensive Care, and Pain Therapy, University Medical Center) as part of the requirements for the obtention of the degree “Doctor of Medicine” at Saarland University Medical Center and Saarland University Faculty of Medicine.

The data analysis work performed in this article (Dr. Baumbach) was supported by Deutsche Forschungsgemeinschaft (Bonn, Germany) within the Collaborative Research Center (Sonderforschungsbereich) SFB 876 “Providing Information by Resource-Constrained Analysis,” project TB1 “Resource-Constrained Analysis of Spectrometry Data.” All other funding was provided from department sources.

Competing Interests

The authors declare no competing interests.

References

References
1.
Westhoff
M
,
Litterst
P
,
Maddula
S
,
Bödeker
B
,
Rahmann
S
,
Davies
AN
,
Baumbach
JI
:
Differentiation of chronic obstructive pulmonary disease (COPD) including lung cancer from healthy control group by breath analysis using ion mobility spectrometry.
Int J Ion Mobil Spec
2010
;
13
:
131
9
2.
Westhoff
M
,
Litterst
P
,
Madulla
S
,
Bödeker
B
,
Baumbach
JI
:
Statistical and bioinformatical methods to differentiate chronic obstructive pulmonary disease (COPD) including lung cancer from healthy control by breath analysis using ion mobility spectrometry.
Int J Ion Mobil Spec
2011
;
14
:
139
49
3.
Bessa
V
,
Darwiche
K
,
Teschler
H
,
Sommerwerck
U
,
Rabis
T
,
Baumbach
JI
,
Freitag
L
:
Detection of volatile organic compounds (VOCs) in exhaled breath of patients with chronic obstructive pulmonary disease (COPD) by ion mobility spectrometry.
In J Mobil Spec
2011
;
14
:
7
13
4.
Schubert
JK
,
Müller
WP
,
Benzing
A
,
Geiger
K
:
Application of a new method for analysis of exhaled gas in critically ill patients.
Intensive Care Med
1998
;
24
:
415
21
5.
Baumbach
JI
,
Westhoff
M
:
Ion mobility spectrometry to detect lung cancer and airway infection.
Spectroscopy Europe
2006
;
18
:
22
7
6.
Westhoff
M
,
Litterst
P
,
Freitag
L
,
Urfer
W
,
Bader
S
,
Baumbach
JI
:
Ion mobility spectrometry for the detection of volatile organic compounds in exhaled breath of patients with lung cancer: Results of a pilot study.
Thorax
2009
;
64
:
744
8
7.
Dwivedi
P
,
Puzon
G
,
Tam
M
,
Langlais
D
,
Jackson
S
,
Kaplan
K
,
Siems
WF
,
Schultz
AJ
,
Xun
L
,
Woods
A
,
Hill
HH
Jr
:
Metabolic profiling of Escherichia coli by ion mobility-mass spectrometry with MALDI ion source.
J Mass Spectrom
2010
;
45
:
1383
93
8.
Maddula
S
,
Blank
LM
,
Schmid
A
,
Baumbach
JI
:
Detection of volatile metabolites of Escherichia coli by multi capillary column coupled ion mobility spectrometry.
Anal Bioanal Chem
2009
;
394
:
791
800
9.
Purkhart
R
,
Köhler
H
,
Liebler-Tenorio
E
,
Meyer
M
,
Becher
G
,
Kikowatz
A
,
Reinhold
P
:
Chronic intestinal Mycobacteria infection: Discrimination via VOC analysis in exhaled breath and headspace of feces using differential ion mobility spectrometry.
J Breath Res
2011
;
5
:
027103
10.
Jünger
M
,
Vautz
W
,
Kuhns
M
,
Hofmann
L
,
Ulbricht
S
,
Baumbach
JI
,
Quintel
M
,
Perl
T
:
Ion mobility spectrometry for microbial volatile organic compounds: A new identification tool for human pathogenic bacteria.
Appl Microbiol Biotechnol
2012
;
93
:
2603
14
11.
Baumbach
JI
:
Ion mobility spectrometry coupled with multi-capillary columns for metabolic profiling of human breath.
J Breath Res
2009
;
3
:
034001
12.
Baumbach
JI
,
Eiceman
GA
:
Ion mobility spectrometry: Arriving on site and moving beyond a low profile.
Appl Spectrosc
1999
;
53
:
338A
55A
13.
Borsdorf
H
,
Eiceman
GA
:
Ion mobility spectrometry principles and applications.
Appl Spectrosc Rev
2006
;
41
:
323
75
14.
Eiceman
GA
,
Stone
JA
:
Ion mobility spectrometers in national defence.
Anal Chem
2004
;
76
:
390A
7A
15.
Guamán
AV
,
Carreras
A
,
Calvo
D
,
Agudo
I
,
Navajas
D
,
Pardo
A
,
Marco
S
,
Farré
R
:
Rapid detection of sepsis in rats through volatile organic compounds in breath.
J Chromatogr B Analyt Technol Biomed Life Sci
2012
;
881–882
:
76
82
16.
Fink
T
,
Heymann
P
,
Taha-Melitz
S
,
Taha
A
,
Wolf
B
,
Rensing
H
,
Volk
T
,
Mathes
AM
:
Dobutamine pretreatment improves survival, liver function, and hepatic microcirculation after polymicrobial sepsis in rat.
Shock
2013
;
40
:
129
35
17.
Scheiermann
P
,
Hoegl
S
,
Revermann
M
,
Ahluwalia
D
,
Zander
J
,
Boost
KA
,
Nguyen
T
,
Zwissler
B
,
Muhl
H
,
Hofstetter
C
:
Cecal ligation and incision: An acute onset model of severe sepsis in rats.
J Surg Res
2009
;
151
:
132
7
18.
Rensing
H
,
Jaeschke
H
,
Bauer
I
,
Pätau
C
,
Datene
V
,
Pannen
BH
,
Bauer
M
:
Differential activation pattern of redox-sensitive transcription factors and stress-inducible dilator systems heme oxygenase-1 and inducible nitric oxide synthase in hemorrhagic and endotoxic shock.
Crit Care Med
2001
;
29
:
1962
71
19.
Jellestad
L
,
Fink
T
,
Pradarutti
S
,
Kubulus
D
,
Wolf
B
,
Bauer
I
,
Thiemermann
C
,
Rensing
H
:
Inhibition of glycogen synthase kinase (GSK)-3-β improves liver microcirculation and hepatocellular function after hemorrhagic shock.
Eur J Pharmacol
2014
;
724
:
175
84
20.
Wolf
A
,
Baumbach
JI
,
Kleber
A
,
Maurer
F
,
Maddula
S
,
Favrod
P
,
Jang
M
,
Fink
T
,
Volk
T
,
Kreuer
S
:
Multi-capillary column-ion mobility spectrometer (MCC-IMS) breath analysis in ventilated rats: A model with the feasibility of long-term measurements.
J Breath Res
2014
;
8
:
016006
21.
Mylotte
JM
,
Tayara
A
:
Blood cultures: Clinical aspects and controversies.
Eur J Clin Microbiol Infect Dis
2000
;
19
:
157
63
22.
Sankar
V
,
Webster
NR
:
Clinical application of sepsis biomarkers.
J Anesth
2013
;
27
:
269
83
23.
Schaub
N
,
Frei
R
,
Müller
C
:
Addressing unmet clinical needs in the early diagnosis of sepsis.
Swiss Med Wkly
2011
;
141
:
w13244
24.
Vincent
JL
,
Donadello
K
,
Schmit
X
:
Biomarkers in the critically ill patient: C-reactive protein.
Crit Care Clin
2011
;
27
:
241
51
25.
Shafazand
S
,
Weinacker
AB
:
Blood cultures in the critical care unit: Improving utilization and yield.
Chest
2002
;
122
:
1727
36
26.
Casserly
B
,
Read
R
,
Levy
MM
:
Multimarker panels in sepsis.
Crit Care Clin
2011
;
27
:
391
405
27.
Ligor
T
,
Ligor
M
,
Amann
A
,
Ager
C
,
Bachler
M
,
Dzien
A
,
Buszewski
B
:
The analysis of healthy volunteers’ exhaled breath by the use of solid-phase microextraction and GC-MS.
J Breath Res
2008
;
2
:
046006
28.
Phillips
M
,
Herrera
J
,
Krishnan
S
,
Zain
M
,
Greenberg
J
,
Cataneo
RN
:
Variation in volatile organic compounds in the breath of normal humans.
J Chromatogr B Biomed Sci Appl
1999
;
729
:
75
88
29.
Vautz
W
,
Nolte
J
,
Bufe
A
,
Baumbach
JI
,
Peters
M
:
Analyses of mouse breath with ion mobility spectrometry: A feasibility study.
J Appl Physiol (1985)
2010
;
108
:
697
704
30.
Filipiak
W
,
Sponring
A
,
Baur
MM
,
Filipiak
A
,
Ager
C
,
Wiesenhofer
H
,
Nagl
M
,
Troppmair
J
,
Amann
A
:
Molecular analysis of volatile metabolites released specifically by Staphylococcus aureus and Pseudomonas aeruginosa.
BMC Microbiol
2012
;
12
:
113
31.
Lanza-Jacoby
S
,
Rosato
E
,
Braccia
G
,
Tabares
A
:
Altered ketone body metabolism during gram-negative sepsis in the rat.
Metabolism
1990
;
39
:
1151
7
32.
Ohtoshi
M
,
Jikko
A
,
Asano
M
,
Uchida
K
,
Ozawa
K
,
Tobe
T
:
Ketogenesis during sepsis in relation to hepatic energy metabolism.
Res Exp Med (Berl)
1984
;
184
:
209
19
33.
Pailla
K
,
El-Mir
MY
,
Cynober
L
,
Blonde-Cynober
F
:
Cytokine-mediated inhibition of ketogenesis is unrelated to nitric oxide or protein synthesis.
Clin Nutr
2001
;
20
:
313
7
34.
Marcondes-Braga
FG
,
Gutz
IG
,
Batista
GL
,
Saldiva
PH
,
Ayub-Ferreira
SM
,
Issa
VS
,
Mangini
S
,
Bocchi
EA
,
Bacal
F
:
Exhaled acetone as a new biomaker of heart failure severity.
Chest
2012
;
142
:
457
66
35.
Schubert
JK
,
Miekisch
W
,
Birken
T
,
Geiger
K
,
Nöldge-Schomburg
GF
:
Impact of inspired substance concentrations on the results of breath analysis in mechanically ventilated patients.
Biomarkers
2005
;
10
:
138
52
36.
Van den Velde
S
,
Nevens
F
,
Van Hee
P
,
van Steenberghe
D
,
Quirynen
M
:
GC-MS analysis of breath odor compounds in liver patients.
J Chromatogr B Analyt Technol Biomed Life Sci
2008
;
875
:
344
8
37.
Seok
J
,
Warren
HS
,
Cuenca
AG
,
Mindrinos
MN
,
Baker
HV
,
Xu
W
,
Richards
DR
,
McDonald-Smith
GP
,
Gao
H
,
Hennessy
L
,
Finnerty
CC
,
López
CM
,
Honari
S
,
Moore
EE
,
Minei
JP
,
Cuschieri
J
,
Bankey
PE
,
Johnson
JL
,
Sperry
J
,
Nathens
AB
,
Billiar
TR
,
West
MA
,
Jeschke
MG
,
Klein
MB
,
Gamelli
RL
,
Gibran
NS
,
Brownstein
BH
,
Miller-Graziano
C
,
Calvano
SE
,
Mason
PH
,
Cobb
JP
,
Rahme
LG
,
Lowry
SF
,
Maier
RV
,
Moldawer
LL
,
Herndon
DN
,
Davis
RW
,
Xiao
W
,
Tompkins
RG
;
Inflammation and Host Response to Injury, Large Scale Collaborative Research Program
:
Genomic responses in mouse models poorly mimic human inflammatory diseases.
Proc Natl Acad Sci U S A
2013
;
110
:
3507
12