| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Anesthesiology and Intensive Care Medicine, University Hospital of Rostock, Schillingallee 35, 18057 Rostock, Germany.
2
Department of Anesthesiology and Intensive Care
Medicine, University Hospital of Freiburg, Hugstetter Strasse 55, 79106
Freiburg, Germany.
aAuthor for correspondence. Fax 49-381-4946402; e-mail wolfram.miekisch{at}medizin.uni-rostock.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Arterial and venous blood samples were taken from mechanically ventilated patients. Additional blood samples were taken from selected vascular compartments of 19 mechanically ventilated pigs. Volatile substances were concentrated by means of solid-phase microextraction (SPME), separated by gas chromatography, and identified by mass spectrometry.
Results: Detection limits were 0.02–0.10 nmol/L. Venous concentrations in pigs were 0.2–1.3 nmol/L for isoprene, 0–0.3 nmol/L for pentane, and 1.2–15.1 nmol/L for dimethyl sulfide. In pigs, substances were not equally distributed among vascular compartments. In humans, median arteriovenous concentration differences were 3.58 nmol/L for isoprene and 1.56 nmol/L for pentane. These values were comparable to pulmonary excretion rates reported in the literature. Acute respiratory distress syndrome (ARDS) patients had lower isoprene concentration differences than patients without ARDS.
Conclusions: The SPME method can detect volatile substances in very low concentrations in the blood of humans and animals. Analysis of volatile substances in vascular compartments will enlarge the diagnostic potential of breath analysis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The diagnostic potential of breath analysis has been obscured by a lack of knowledge of the origin, distribution, and physiological meaning of the volatile substances. For example, isoprene, one of the main components of human breath, has not only been related to the mevalonic pathway of cholesterol synthesis (14), but also to the activation of neutrophils (9), the presence of lung cancer (5), and even the uptake of environmental pollutants (10). Additional studies require that volatile substances be determined in selected vascular compartments, preferably within the controllable setting of an animal model, but the concentrations of volatile substances such as isoprene, pentane, or dimethyl sulfide in animal blood are below the detection limits of available analytical methods (15).
Analysis of volatile compounds by solid-phase microextraction (SPME)1 has been established for many environmental (16)(17) and forensic (18)(19)(20) applications. In most of these cases, SPME is more sensitive, more precise, and more reliable than commonly used headspace techniques. We therefore developed an analytical method using SPME coupled to ion-trap detection to assess trace amounts of the hydrocarbons isoprene and pentane, the ketone acetone, the volatile anesthetic isoflurane, and the thioether dimethyl sulfide in the blood of humans and laboratory animals.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Materials.
SPME devices together with their fiber assemblies
were purchased from Supelco Inc.
SPME procedure
Before the SPME procedure, the Carboxen/polydimethylsiloxane
(PDMS)-coated fiber was pretreated in the injection port of a gas
chromatograph at 285 °C for 30 min.
Whole blood (5 mL) was drawn into a heparin-containing syringe and immediately transferred into a sealed headspace vial containing 4.9 mL of an aqueous solution of NaCl (30 g/L), 0.1 mL of 500 g/L phosphate buffer (pH 7.0), and a magnetic stirring bar. Internal standard (5 µL) was added through the septum by means of a microsyringe.
The vial was put into an aluminum block heater (Reacti Therm Heating-Stirring Module; Pierce) for stirring and heating at 40 °C. After the vial was heated for 5 min, the syringe needle of the SPME device was inserted into the headspace of the vial for 5 min. The needle containing the SPME fiber was withdrawn and introduced into the port of a capillary gas chromatograph. Port temperature was 270 °C. The SPME device was held in the port for 2 min to allow complete desorption of the components.
calibration solutions
Calibrators were prepared using methanol as a solvent. The
methanolic stock solution containing 5 µL of isoprene, 3 µL of
pentane, 3 µL of dimethyl sulfide, 5 µL of isoflurane, and 10 µL
of acetone in a 22-mL gastight headspace vial was diluted with methanol
(1:110). The solution was stirred by a magnetic microstirring device
for 10 min at 20 °C. Using a microsyringe, we added aliquots of the
calibration solution (0.2–300 µL) to the samples containing 5 mL of
fresh frozen plasma, 4.9 mL of an aqueous solution of 30 g/L NaCl, and
100 µL of phosphate buffer.
analytical instrument settings
Gas chromatographic–mass spectrometric analyses were carried out
on a CX 3400 (Varian) gas chromatograph coupled to a Varian Saturn 2000
ion-trap mass spectrometer. Data were acquired using the Varian Saturn
Software.
Gas chromatography conditions.
The column was a 30-m (0.32 mm
i.d.) Poraplot Q column (Chrompack). The carrier gas was helium at a
head pressure of 12 psi using a split/splitless injector with a
splitless time of 60 s followed by a split ratio of 1:20 for the
remaining run time. The oven temperature was initially held at 50 °C
for 2 min, and then was increased by 10 °C/min to 180 °C,
increased by 8 °C/min to 220 °C, and held at the final
temperature for 2 min.
Mass spectrometry conditions.
Total ion current was monitored
for all samples using electron-impact ionization (70 eV). A scan rate
of 1 scan/s was applied. Quantification was performed using
characteristic masses.
The peak at m/z 58 [M (molecular ion)] was used for the quantification of acetone. For pentane, both m/z 43 (base peak) and m/z 71 [M - 1] can be used for quantification. Although the m/z 71 fragment is more significant for pentane than the common m/z 43 fragment, the latter was used for quantification because of the higher intensity of this peak. For the quantification of isoprene, the peak at m/z 67 [M - 1] was used. For the quantification of dimethyl sulfide, we used the molecular ion at m/z 62; for isoflurane, we used the m/z 117 fragment [M - Cl].
analytical procedures
Linear range.
The linear range of the method was investigated
by determining calibration curves in the concentration ranges of
interest. Aliquots of a methanolic calibration solution (described
above) containing acetone, dimethyl sulfide, pentane, isoflurane, and
isoprene at concentrations of 0.01–200 nmol/L were added to a vial
containing 5 mL of fresh frozen plasma and 5 mL of an aqueous solution
of 30 g/L NaCl. The line of best fit for the relationship between peak
areas (obtained by integrating the selected m/z
chromatograms) and concentrations of the analytes in the sample was
determined by linear regression.
Precision.
Eight replicate measurements of the calibration
solutions in human plasma containing identical concentrations (3 ng/5
mL) of acetone, dimethyl sulfide, isoflurane, isoprene, and pentane
were performed. Precision was assessed by calculating the mean,
standard deviation, and relative standard deviation (% RSD) of the
observed values.
Limits of detection.
The limit of detection (LOD) for each
calibrator was estimated based on the signal-to-noise ratio obtained
for calibration solutions containing the compounds of interest at low
concentrations. The LOD was defined as that concentration of an analyte
that produced a signal 3 SD above the mean of a blank sample using
ion-trap detection and single-ion traces. The average signal-to-noise
ratio of four measurements using a blank was used for calculation.
Accuracy.
The accuracy of the method was investigated by
creating a five-point calibration curve for dimethyl sulfide [0–30
nmol/L; y = 2.24 x
10-4(Area); r2
= 0.98], isoflurane [0–26 nmol/L; y = 1.70 x
10-4(Area); r2
= 0.99], isoprene [0–33 nmol/L; y = 3.82 x
10-5(Area); r2
= 0.99]; and pentane [0–29 nmol/L; y = 1.38 x
10-5(Area); r2
= 0.99] in the range of interest. A solution containing 2.4 ng (3.9
nmol/L) of dimethyl sulfide, 6.8 ng (4.6 nmol/L) of isoflurane, 4.2 ng
(6.2 nmol/L) of isoprene, and 2.3 ng (3.2 nmol/L) of pentane in 5 mL of
human plasma was prepared. Triplicate analyses were performed.
Recovery.
The recovery was investigated by adding methanolic
calibration solutions (containing 2.4 ng of dimethyl sulfide, 6.8 ng of
isoflurane, 4.2 ng of isoprene, and 2.3 ng of pentane) to human blood
samples containing known amounts of volatile substances. The resulting
concentrations were 3.9 nmol/L dimethyl sulfide, 4.6 nmol/L isoflurane,
10.5 nmol/L isoprene, and 3.5 nmol/L pentane. Triplicate measurements
were performed.
extraction time
The extraction time profiles were established by plotting detector
responses against the extraction time (10, 30, 60, 120, 240, 300, and
500 s) using plasma samples to which calibrators had been added.
To demonstrate that the extraction time profiles were similar in whole
blood, we repeated an extraction profile measurement limited to five
points (30, 120, 240, 300, and 500 s) using whole blood to which
calibrators had been added. Because acetone, isoprene, pentane, and
dimethyl sulfide are a priori present in whole blood, the resulting
concentrations in plasma and blood samples were different.
Equilibration was reached when an additional increase of the extraction
time did not produce a significant increase of the detector response.
effect of temperature
The effect of temperature on the SPME process has already been
investigated (21). We therefore chose only five temperatures
in the desired range (20, 30, 40, 50, and 70 °C) to optimize the
amount extracted.
patients and animals
After approval by the local ethics committee and after having
obtained informed consent from the patient or the patient’s nearest
relative, we examined 33 mechanically ventilated patients. From August
1998 to July 1999, all patients with an indwelling pulmonary artery
catheter were included in the study. Because of the mechanical
ventilation, all patients received sedating and pain-relieving
medication. Therefore, a normal diurnal rhythm that could have
influenced isoprene excretion (22) did not exist in these
patients. Nevertheless, we took all measurements at approximately the
same time (0900–1400).
Nine patients suffered from an acute respiratory distress syndrome
(ARDS). The criteria for ARDS were a ratio of the arterial
O2 pressure to the fraction of inspired oxygen
(PaO2/FIO2)
200 mmHg regardless of positive end expiratory pressure, bilateral
infiltrates on anterior-posterior chest radiographs, and pulmonary
artery wedge pressure 18 mmHg (23). Ten patients
were at risk for developing ARDS. Patients with multiple injury or
(sterile) pancreatitis, and patients following major abdominal or
cardiovascular surgery were regarded as being at risk to develop ARDS.
Three of these patients underwent orthotopic liver transplantation.
Fourteen patients had sepsis. Criteria for sepsis were temperature
>38.0 °C, leukocytes >12 000/µL or <3600/µL, and confirmed
bacterial or mycotic infection (24).
Blood samples from 19 mechanically ventilated pigs were examined. The animals had been acutely instrumented to study the effects of epidural application of local anesthetics on splanchnic perfusion (25). Mixed venous and arterial blood samples were taken from all 19 animals; additional samples were taken from the portal and hepatic venous compartments of 4 animals. In addition, venous blood samples from four rabbits that were used for an isolated lung model of ventilation and perfusion were analyzed.
All blood samples were analyzed within 3 h after sampling. After >3 h, isoprene (12% after 5 h) and pentane (8% after 5 h) concentrations were significantly decreased.
statistical analysis
Because substance concentrations were not normally distributed,
the Mann–Whitney U-test, Wilcoxon rank-sum test, or the
Kruskal–Wallis test was used when appropriate. Results are reported as
medians and 25th to 75th percentiles. P <0.05 was
considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. For nonpolar substances, the Carboxen/PDMS fiber reached
equilibrium after 
4–5 min in plasma and in whole blood samples.
Only for polar substances, such as acetone, was a further increase in
the extracted amount observed. Therefore, an extraction time of 5 min
was used for the measurements.
|
effect of extraction temperature
The relative peak areas of a calibration mixture at different
temperatures are shown in Table 1
. With the exception of acetone, the highest amounts were
extracted at temperatures between 40 and 50 °C. Because protein
denaturation takes place at temperatures >43 °C, a temperature of
40 °C was used for the extractions.
|
linear range
The linear ranges, precision data, and detection limits are shown
in Table 2
. RSDs for the target compounds were <6% (n = 8). The LOD
ranged from 0.02 nmol/L for isoflurane to 0.10 nmol/L for pentane.
|
accuracy
Using the calibration curves, we found concentrations of 3.9
± 0.2 nmol/L dimethyl sulfide, 4.6 ± 0.1 nmol/L isoflurane,
6.1 ± 0.3 nmol/L isoprene, and 3.2 ± 0.2 nmol/L pentane in
the supplemented plasma samples containing 3.9 nmol/L dimethyl sulfide,
4.6 nmol/L isoflurane, 6.2 nmol/L isoprene, and 3.2 nmol/L pentane.
recovery
Recoveries were 93% ± 3% for isoprene, 96% ± 5% for pentane,
92% ± 3% for dimethyl sulfide, and 102% ± 2% for isoflurane.
blood analysis
Analysis of blood from mechanically ventilated patients.
Fig. 2
shows a chromatogram of a venous blood sample from a
mechanically ventilated patient compared with a plasma blank. Fig. 3
shows a selected-ion chromatogram of a mixed venous and an
arterial blood sample from the same patient.
|
|
The median isoprene concentration was 9.08 nmol/L in venous and 5.73
nmol/L in arterial blood (range, 0.52–24.4 nmol/L in venous blood and
0 (LOD) to 18.0 nmol/L in arterial blood). Table 3
shows venous and arterial isoprene concentrations for all
patients included in the study and the resulting concentration
differences. (Cvenous -
Carterial)/Cvenous
represents a virtual pulmonary substance extraction. Values were
5–96%.
|
Differences between venous and arterial isoprene concentrations were lower in patients with ARDS and at risk to develop ARDS than in septic patients [2.20 nmol/L (range, 0.5–3.93 nmol/L) vs 4.09 nmol/L (3.53–5.17 nmol/L); P = 0.038].
The median pentane concentration was 11.8 nmol/L in venous and 5.8 nmol/L in arterial blood [range, 0–58.0 nmol/L in venous and 0 (LOD) to 48.1 nmol/L in arterial blood]. Differences between venous and arterial pentane concentrations (median difference, 1.56 nmol/L) were lower in patients with ARDS and at risk to develop ARDS than in septic patients {0.25 nmol/L [range, 0.0 (LOD) to 5.82 nmol/L] vs 3.21 nmol/L (0.39–8.76 nmol/L); P = 0.041}. The median dimethyl sulfide concentration was 0.41 nmol/L in venous and 0.27 nmol/L in arterial blood [range, 0 (LOD) to 1.72 nmol/L in venous and 0–0.74 nmol/L in arterial blood]. Median isoflurane concentrations were 59.6 nmol/L in venous and 33.1 nmol/L in arterial blood.
Acetone concentrations were always higher than the linear range of the method; therefore, exact quantification of this compound was not possible.
Analysis of animal blood.
In pigs, isoprene concentrations
were 0.2–1.3 nmol/L in venous blood and 0 (below LOD) to 0.8 nmol/L in
arterial blood. In rabbits, isoprene concentrations were 0.3–0.7
nmol/L in venous blood. In pigs, the pentane concentrations were 0
(below LOD) to 0.3 nmol/L in venous blood and 0 (below LOD) to 0.2
nmol/L in arterial blood; pentane was not detected in rabbits. In pigs,
dimethyl sulfide concentrations were 1.2–15.1 nmol/L in venous and
0.8–12.7 nmol/L in arterial blood; dimethyl sulfide was not detected
in rabbits.
Isoprene concentrations measured in different vascular regions in
acutely instrumented pigs are shown in Fig. 4
. In three pigs, the highest concentrations were found in portal
and in mixed venous blood. In one pig, the hepatic venous concentration
was highest. Concentrations in arterial blood were always lower than in
portal, mixed venous, and hepatic venous blood.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The accuracy of acetone determinations suffered from its high concentration in human blood. The upper limit of the linear range in our method for acetone was 60 nmol/L, which is far below the concentrations in patients’ blood. Nevertheless, we thought that it could be of interest to determine this substance together with the other relevant breath markers in a single analytical step because acetone is a well-known volatile metabolic marker. However, analytical settings (pH, temperature, extraction time, and dilution) would have to be changed considerably to assess acetone concentrations with sufficient precision. Under these changed conditions, nonpolar substances such as isoprene, pentane, and dimethyl sulfide could no longer be determined quantitatively. However, acetone concentrations in blood may easily be assessed by several other analytical methods (26).
The stability of whole blood samples represents another analytical problem. Because whole blood samples are not stable for more than 180 min at room temperature, there is not enough time to carry out the number of analyses needed for statistical relevance of any validating study. In addition, substances under investigation, such as isoprene, acetone, and pentane, are frequently present in blood samples. Hence, plasma was used for calibration and for validating measurements. Recovery studies and limited five-point extraction time measurements in whole blood showed that results obtained in validation studies using plasma can be transferred to whole blood measurements.
One of the main problems in the analysis of volatile substances in blood and breath is the distinction between environmental contaminants and endogenous substances. In fact, many volatile substances are found in human blood and breath that are of environmental origin. Furthermore, it is difficult to identify the origin of lipophilic substances such as pentane because long-term storage of exogenous fractions of these substances cannot be excluded. Springfield and Levitt (27) reported a significant decrease in breath pentane after washout in rats. They concluded that the main part of breath pentane represents a contaminant that had been taken up from environmental sources and had been stored in fat tissue. In humans, however, there is strong experimental evidence that the substances used in this study (isoprene, pentane, acetone, and dimethyl sulfide) represent endogenous markers (4)(5)(6)(7)(9)(10)(11)(12)(28)(29)(30)(31)(32)(33)(34). The correlation between clinical conditions and substance concentrations in expired air strongly suggests that long-term storage is of less importance than production of endogenous substances.
Obviously, concentrations of halogenated hydrocarbons depended on previous exposure, e.g., during anesthesia.
Concentrations of isoprene or pentane in animal blood are up to 10-fold lower than in human blood. Previously, isoprene, for example, could not be determined in animal blood because its concentration was lower than the detection limits of the available analytical methods (15). Now, however, the low detection limits of the SPME method allow assessment of trace amounts of volatile substances, such as isoprene, in the blood of humans as well as in the blood of laboratory animals. As a consequence, an experimental model can be established to assess volatile substances in blood and exhaled gas under controllable physiological and pathological conditions. Hence, detailed information on the origin, distribution, and physiological meaning of the volatile substances found in exhaled air can be obtained.
Initial measurements in acutely instrumented pigs demonstrated that volatile substances are, in fact, not equally distributed among the different vascular compartments. In pigs, isoprene concentrations tended to be higher in portal and mixed venous blood and lower in hepatic venous and in arterial blood. This distribution of concentrations suggests that isoprene may be generated by intestinal bacteria or mucosal cells. Decreasing substance concentrations in the hepatic venous blood suggest that isoprene is metabolized in the liver. High mixed venous concentrations suggest that there is a peripheral origin of isoprene, e.g., from muscle cells.
Interesting aspects arose from simultaneous measurements of substance concentrations in mixed venous and arterial blood of patients with an indwelling Swan-Ganz catheter. The differences between mixed venous and arterial concentrations represent the amounts of volatile substances that are exhaled through the lungs. Pulmonary extraction of hydrophilic volatile substances, such as acetone, depends on the extent of dead-space ventilation in the lung. Extraction of lipophilic substances, such as isoprene, depends on shunt perfusion of the lung. Furthermore, pulmonary extraction may be influenced by the ratio of cardiac output and minute ventilation, which can be quite different from 1 in mechanically ventilated patients. Therefore, pulmonary extraction of substances varies over a wide range, depending on solubility, ventilation/perfusion ratio, and ventilator settings. For the same reason, arteriovenous concentration differences or pulmonary extraction do not correlate with venous concentrations. Thus, it is not possible to assess pulmonary extraction of volatile metabolites by means of a single (venous) blood measurement.
Differences between venous and arterial isoprene concentrations were comparable to the pulmonary excretion rates reported in several other studies (6)(33)(35). In accordance with previously reported results (6)(34), differences between venous and arterial isoprene concentrations were lower in patients suffering from or at risk for developing an ARDS. This may be attributable to an effect on membrane repair mechanisms caused by an impaired cholesterol metabolism (14).
Differences in pentane concentrations were highest in septic patients. Because infection induces significant (per)oxidative and radical activity, high pentane concentrations in patients with sepsis are in agreement with pentane being a marker of lipid peroxidation (4)(28)(29). In several studies investigating patients with high inflammatory activity, breath pentane concentrations comparable to the arteriovenous concentration differences in the septic group of our study were reported (9)(30)(31)(32).
The median pentane concentration differences that we found in the ARDS group were in the same range as breath pentane concentrations reported by Kohlmüller and Kochen (8), Euler et al. (33), and Mendis et al. (2) for volunteers, and were comparable to breath pentane concentrations reported previously for mechanically ventilated ARDS patients (6)(34).
The extremely low pentane concentrations that we found in pigs were possibly related to the fact that we studied animals without any sign of infection.
Dimethyl sulfide was not regularly detected in all patients. Venous concentrations were highest in patients after liver transplantation. Thus, dimethyl sulfide may serve as a marker of impaired liver function in humans (12)(13).
In conclusion, the SPME procedure is well suited for the assessment of volatile substances in the blood of animals and humans. The SPME method may provide more detailed information on the origin, distribution, and physiological meaning of volatile substances in the body. Hence, our understanding of breath analysis could be enlarged, and its potential use as a diagnostic tool could be facilitated.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, 2,3-dimethylbutadiene (internal standard);
, dimethyl sulfide; 
, isoprene; X, pentane; *, isoflurane;
•, acetone.
