Abstract

Our objective was to identify, for diagnostic purposes, potential volatile biomarkers of human microbial pathogens. We analysed the head space of cultures of medically important bacterial and fungal respiratory pathogens for 2-Pentylfuran (2PF) production through the use of Solid Phase Micro Extraction (SPME) and Gas Chromatography/Mass Spectroscopy (GC/MS). Our results confirm that 2PF is consistently produced by Aspergillus fumigatus, Fusarium spp., Aspergillus terreus, Aspergillus flavus and to a lesser extent by Aspergillus niger. 2-Pentylfuran was not detected from most of the bacterial strains except for Streptococcus pneumoniae. In human studies, four litre breath samples were collected from patients with cystic fibrosis (CF), with and without colonisation by A. fumigatus and other pathogens, as well as healthy volunteers. 2-Pentylfuran was detected in breath samples collected from 4/4patients with CF and A. fumigatus colonization, 3/7patients with CF and no microbiological evidence of A. fumigatus and 0/10healthy control individuals. These results suggest that 2PF may be a biomarker for lung colonization/infection by fungal pathogens. To our knowledge, this is the first report describing the detection in breath samples of a volatile biomarker of a pathogen resident in the lungs. Breath analysis has the potential of being a non-invasive diagnostic method of detecting respiratory infection including invasive aspergillosis

Introduction

Aspergillus and other filamentous fungi are feared infectious agents with high mortality in severely immunocompromised and neutropenic patients 1 despite recent advances in the prophylaxis and treatment 1–3. Aspergillus fumigatus is the most important member of the genus as it causes more than 90% of Aspergillus infections in immunocompromised patients 4. Aspergillus flavus, A. niger, A. terreus and other Aspergillus species have occasionally been reported to cause invasive infections. While other filamentous fungi, such as Fusarium spp., Mucor spp., Scedosporium apiospermum and zygomycetes may be rare etiologic agents of human disease, they have been identified more frequently in recent trials of antifungal prophylaxis.

The diagnosis of A. fumigatus and other fungal infection remains difficult. Sputum culture and direct microscopy are insensitive and biopsies are often hazardous in severely immunocompromised patients. PCR and galactomannan (GM) testing of bronchoalveolar lavage (BAL) fluid were initially reported to be highly sensitive for specific Aspergillus infections by some investigators but subsequent studies have reported variable results 5–11.

The detection of microbial volatile organic compounds (MVOCs) may offer an alternative approach to the diagnosis of fungal infections. All microorganisms produce volatile compounds as a result of their normal metabolism. The detection of differences in metabolism among pathogens is a fundamental tool in laboratory microbiology that has been exploited primarily through liquid and agar based techniques. New highly sensitive technologies offer the possibility of a new approach to diagnosis by detection of MVOCs.

One strategy is to use profiles of MVOCs that have been found to be specific for some organisms 12. An alternative would be to identify unique specific markers of pathogenic organisms. A potential candidate marker for A. fumigatus is 2-Pentylfuran (2PF) 11. It is a small molecule with a molecular weight of 138 g/mol and a distinctive mass fragmentation (Fig. 1). It is volatile in nature (vapour pressure is estimated to 160 Pa at 25°C) and poorly water-soluble (42mg/l at 25 °C). MVOC production by A. fumigatus has been studied in agriculture, food technology and the building industry 13, 14 and 2PF has been found to be produced on different substrates such as malt extract and gypsum board 14. Human pathogens such as Scedosproium spp., Mucor spp., Fusarium spp., Pneumocystis jiroveci and Cryptococcus spp. have not been systematically studied. Given that 2PF is produced in detectable quantities in vitro, it is feasible that it may be produced in vivo and detected in breath samples from individuals whose airways are colonized or infected by fungi.

Fig. 1

Structure and mass fragmentation of 2-Pentylfuran from NIST library.

Fig. 1

Structure and mass fragmentation of 2-Pentylfuran from NIST library.

As a first step in assessing the diagnostic potential of 2PF we needed to establish whether 2PF production was specific for members of the genus Aspergillus as compared with other filamentous fungi and medically important lower respiratory bacterial pathogens under biological rather than environmental growth conditions. The next step was to conduct a pilot study to determine if 2PF could indeed be detected on the breath of patients chronically colonized with A.fumigatus.

Materials and methods

Strains and culture media

Clinical isolates of Aspergillus flavus, A. fumigatus, Aspergillus niger, Aspergillus terreus, Candida albicans, Fusarium spp., Scedosporium apiospermum, Rhizopus microsporus and Staphylococcus aureus, Pseudomonas aeruginosa, Pseudomonas fluorescens, Haemophilus influenzae, Streptococcus pneumoniae, Moraxella catarrhalis, Escherichia coli, Legionella pneumophila and Burkholderia cepacia were used in these experiments (Table 1). All bacterial stains and most of the fungal strains were obtained from the Microbiology Department of Canterbury Health Laboratories, Christchurch, New Zealand, while other fungal isolates were obtained from the Department of Microbiology of Lab Plus, Auckland. Organisms were cultured on Sabouraud's dextrose agar (SDA; Oxoid NZ, Auckland, New Zealand) and their identities checked by colonial and microscopic morphology before being tested 15.

Table 1

Source of fungal isolates tested.

Species (no. of strains) Site of isolation (no. of strains) 
A. fumigatus (15) Cerebellar abscess1 (1), BAL2 (1), sputum3 (13) 
A. flavus (9) Right upper lobe biopsy (1), sputum (2), ear swabs (6) 
A. niger (5) BAL (1), sputum (2), ear swab (2) 
A. terreus (6) Sputum (1), hip aspirate (1), arm swab (1), ear swab (3) 
Fusarium spp. (10) Corneal scraping (1), finger and toe nails (9) 
S. apiospermum (6) Peritonsillar biopsy (1), sputum (3), shin swab (1), sclera tissue (1) 
R. microsporus (4) Lung abscess4 (1), ethmoid biopsy (1), appendix abscess pus (1), pessary (1) 
C. albicans (7) Blood isolates (2), sputum (5) 
Species (no. of strains) Site of isolation (no. of strains) 
A. fumigatus (15) Cerebellar abscess1 (1), BAL2 (1), sputum3 (13) 
A. flavus (9) Right upper lobe biopsy (1), sputum (2), ear swabs (6) 
A. niger (5) BAL (1), sputum (2), ear swab (2) 
A. terreus (6) Sputum (1), hip aspirate (1), arm swab (1), ear swab (3) 
Fusarium spp. (10) Corneal scraping (1), finger and toe nails (9) 
S. apiospermum (6) Peritonsillar biopsy (1), sputum (3), shin swab (1), sclera tissue (1) 
R. microsporus (4) Lung abscess4 (1), ethmoid biopsy (1), appendix abscess pus (1), pessary (1) 
C. albicans (7) Blood isolates (2), sputum (5) 

1Aspergillus endocarditis with multiple cerebral emboli.

2Invasive aspergillosis post lung transplant.

3Bronchiectasis and cystic fibrosis patients.

4Post renal transplant for diabetic nephropathy.

All organisms for Gas Chromatography/Mass Spectroscopy (GC/MS) testing, except for H. influenzae and Legionella spp. were cultured on Colombia sheep blood agar and nutrient agar (Fort Richard Laboratories, Auckland, New Zealand). H. influenzae was cultured on chocolate agar (Oxoid NZ, Auckland, New Zealand) and Legionella spp. on buffered charcoal yeast extract agar (BCYE, Oxoid NZ, Auckland, New Zealand). In addition, A. fumigatus and Fusarium spp. were grown on a BCYE and D-MEM enriched with calf serum without antibiotics (Invitrogen, Carlsbad, California, USA).

Cultures

All test organisms were grown on 10 ml of the appropriate culture media within 100 ml sterile glass vials sealed airtight with screw caps incorporating a teflon-coated silicone rubber septum. Inocula were prepared aseptically by streaking the growth from stock cultures onto blood agar plates and incubating the latter for 48 h. Bacterial and yeast cells were transferred with sterile loops from the latter plates into 5ml of sterile physiological saline to prepare 0.5 McFarland standard suspensions. Spores of filamentous fungi were harvested from the blood agar plates with sterile water containing 0.05% Tween to develop 0.5 McFarland standard suspensions. Five hundred microlitres of this suspension was introduced to the sealed culture vial by injecting through the septum onto the culture medium. Cultures were maintained at 37°C for 2 days, with the vials being flushed for 1 min with purified air after 24 h incubation to remove all volatile compounds. The head space gas was sampled after a further 24 h of incubation. Uninoculated media served as negative controls and were processed in the same way as the culture samples.

Solid–Phase Micro Extraction (SPME) Gas Chromatography/Mass Spectrometry (GC/MS)

Sample preparation

Preconcentration was performed using Solid Phase Micro Extraction (Supelco, Bellefonte, USA). The fiber (DVB/Carboxen/PDMS) was conditioned at 250°C for 15 min in the hot injector. It was exposed into culture vials for 3 h then directly desorbed in the injection port for 15 min.

GC/MS parameters

A Saturn 2200 system (Varian, Palo Alto, USA) was used to perform the GC/MS analysis. The column was a Zebron 30 m×0.25 mm×1.4 µm ZB-624 (Phenomenex, Auckland, New Zealand).The temperatures of the injector, ion trap, manifold and transfer line were 250°C, 200°C, 60°C and 250°C, respectively. The oven program commenced at 60°C for 2 min and was raised to 250°C at a rate of 10°C/min, at which temperature it was maintained for a further 2 min. Helium flow was set at a constant rate of 1.2 ml/min. The split vent was opened to a ratio of 1:50 after 1 min. Fragmentation was performed in the EI-mode as full scan.

Calibration and semi-quantification

2-Pentylfuran as 98% pure standard (Sigma-Aldrich, St. Louis, Missouri, USA) was used for calibration and semi-quantification. Ten microlitres of 2PF solutions diluted in methanol were deposited into 100 ml headspace vials containing 10 ml of buffer. The final calibration was made up as total amounts in the headspace vials containing 0.1, 12.5, 5, 20 and 50 nmol/mol. The fibre was exposed into the headspace vials and desorbed the same way as described above. The calibration curve was linear for 2PF (Fig. 2). Because of differences in volatiles in the headspace above the culture samples as compared with conditions used to produce the standard curve, 2PF levels from cultures can only be regarded as semi-quantitative. Results are therefore reported as 03 = trace (≤3 nmol/mol), 1+ (410 nmol/mol), 2+ (1120 nmol/mol), 3+ (>20 nmol/mol).

Fig. 2

Calibration curve of 2-Pentylfuran.

Fig. 2

Calibration curve of 2-Pentylfuran.

Human subjects

Ethical approval for the study was obtained from the Canterbury Ethics Committee, and participants gave their informed consent to take part in the study.

Patients with cystic fibrosis

All subjects were under the care of specialist respiratory services at Christchurch Hospital and had a diagnosis of cystic fibrosis confirmed by sweat test. Criteria for chronic colonization were records of at least three positive cultures of A. fumigatus from sputum, broncholveolar lavage or cough swabs within the last 12 months, and one within the last 1 month. Patients were excluded from the study if they were currently undergoing treatment with itraconazole. Critieria of the absence of Aspergillus colonization were negative results in at least 3 sputum cultures over the last 12 months of which at least one having been done in the last 1 month. The presence of bacteria in sputum samples was determined from the patients' microbiological reports.

Healthy controls

The health status of prospective control patients was determined by questionnaire and any person with a history of acute or chronic illness including respiratory disease or recent antibiotic use were excluded.

Breath sampling

Breath samples were collected in 4 l tedlar bags (SKC Inc., Eighty four PA USA), each one of which incorporate a valve, disposable mouthpiece, and septum that could be pierced for sampling. Samples were collected by asking participants to exhale through the mouthpiece into the bag until full. The valve in the bag was then closed and samples transported immediately to the laboratory for testing. Sampling was done at times convenient to the subjects between 9:00 am and 5:00 pm.

Analysis of breath samples by GC/MS-MS

Breath samples were analysed by GC/MS-MS for presence of 2PF. The conditioned SPME fiber was exposed into the collection bags for 48 h and then desorbed directly into the injection port for 5 min. Ion preparation for MS-MS analysis was EI mode; the selected parent ion was m/z 81 with an isolation window of m/z 3; excitation storage level was 35; excitation amplitude was 35; the resulting MS-MS spectra featured two main peaks at m/z 53 (100) and m/z 81 (82). Because of the semi-quantitative nature of the assay in vivo results are reported as detected (D) and not detected (nd).

Results

Quantification of 2-Pentylfuran by GC/MS, calibration and standardization

Semi-calibration curves of headspace gas analysis of serial aqueous dilutions of 2PF were plotted. The resulting calibration curve proved to be linear in the range 150 nmol/mol (Fig. 1). However quantification via SPME is considered semi-quantitative only, especially with manual injection.

Detection of 2-Pentylfuran from control media

All media except BCYE produced some 2PF after incubation in the absence of inoculum, presumably as decomposition products. Flushing the vials with purified air after 12 h of incubation reduced the 2PF to very low levels as shown for controls (Table 2).

Table 2

2-Pentylfuran production of species on blood agar.

Species (no. of strains) 2-Pentylfuran 
A. fumigatus (15) 2 + 
A. flavus (9) 1 + 
A. niger (5) 1 + 
A. terreus (6) 1 + 
Fusarium spp. (10) 3 + 
S. apiospermum (6) 1 + 
R. microsporus (4) nd 
C. albicans (7) nd 
Ps. aeruginosa (5) nd 
Ps. fluorescens (5) nd 
B. cepacia (5) nd 
S. aureus (7) nd 
S. pneumoniae (7) 3 + 
M. catarrhalis (6) nd 
E.coli (6) nd 
L. pneumophila (6) nd* 
H. influenzae (6) nd ** 
  
Negative controls 
Blood agar (22) Trace 
BCYE (4) nd 
Chocolate media (5) Trace 
Species (no. of strains) 2-Pentylfuran 
A. fumigatus (15) 2 + 
A. flavus (9) 1 + 
A. niger (5) 1 + 
A. terreus (6) 1 + 
Fusarium spp. (10) 3 + 
S. apiospermum (6) 1 + 
R. microsporus (4) nd 
C. albicans (7) nd 
Ps. aeruginosa (5) nd 
Ps. fluorescens (5) nd 
B. cepacia (5) nd 
S. aureus (7) nd 
S. pneumoniae (7) 3 + 
M. catarrhalis (6) nd 
E.coli (6) nd 
L. pneumophila (6) nd* 
H. influenzae (6) nd ** 
  
Negative controls 
Blood agar (22) Trace 
BCYE (4) nd 
Chocolate media (5) Trace 

Nd, not detected.

*BCYE media.

**Chocolate media.

Detection of 2-Pentylfuran from laboratory cultures

When cultured on blood agar, all strains of A. fumigatus, A. flavus, Fusarium spp., and S. apiospermum produced detectable levels of 2PF that were clearly greater than found with the control media (Table 1). Detectable levels of 2PF were distinguishable from controls with all but one strain of A. niger and one A. terreus isolate.A. fumigatus strain 1 was tested five times on different occasions, with the measured 2PF levels of 2+ on all tests.

When cultured on nutrient agar, 2PF was detected from the headspace of cultures of 2 strains of A.fumigatus (2 + ), A. flavus (1 + ), and Fusarium spp. (3 + ), but was not detected in cultures of S. apiospermum, A. terreus, C. albicans and any of the bacteria tested including S. pneumoniae.

When cultured on BCYE increased 2PF levels were also detected in cultures of A. fumigatus (2 + ) and Fusarium spp. (2 + ).

With the exception of S. pneumoniae, increased levels of 2PF were not detected in the head space of any bacterial cultures as shown in Table 1 or in six strains of H. influenzae cultured on chocolate agar. However, some preliminary additional experiments with S. pneumonia revealed a set of other volatile metabolites which were emitted by this bacterium and not found in the head space of fungal cultures.

Detection of 2-Pentylfuran from breath samples

Relevant demographic data and 2PF detection are shown in Table 3. 2-pentylfuran was detected in breath samples from 4 of 4 subjects with CF and known Aspergillus colonization, but not detected in any of the healthy controls.

Table 3

Demographic and microbiological data of patients enrolled in the study, and results of testing breath by SPME/GC-MS for presence and quantity of 2-Pentylfuran. ‘Aspergillus colonization’ refers to colonization with Aspergillus fumigatus only.

 ID Age Sex Underlying disease S. aureus S. pneumoniae M. catarrhalis H. influenzae Ps. aeruginosa A. fumigatus 2-Pentylfuran 
CF + aspergillus 12 20 CF    
 16 20 CF    
 CF     
 30 12 CF     
            
CF, no aspergillus 10 CF    
 11 CF    nd 
 17 19 CF     
 19 CF      nd 
 20 CF     nd 
 21 CF    
 22 CF     nd 
            
Normal controls 18 30 none       nd 
 13 25 none       nd 
 14 33 none       nd 
 15 31 none       nd 
 24 35 none       nd 
 25 36 none       nd 
 26 36 none       nd 
 27 44 none       nd 
 28 37 none       nd 
 29 57 none       nd 
 ID Age Sex Underlying disease S. aureus S. pneumoniae M. catarrhalis H. influenzae Ps. aeruginosa A. fumigatus 2-Pentylfuran 
CF + aspergillus 12 20 CF    
 16 20 CF    
 CF     
 30 12 CF     
            
CF, no aspergillus 10 CF    
 11 CF    nd 
 17 19 CF     
 19 CF      nd 
 20 CF     nd 
 21 CF    
 22 CF     nd 
            
Normal controls 18 30 none       nd 
 13 25 none       nd 
 14 33 none       nd 
 15 31 none       nd 
 24 35 none       nd 
 25 36 none       nd 
 26 36 none       nd 
 27 44 none       nd 
 28 37 none       nd 
 29 57 none       nd 

D, detected. Nd, not detected.

Discussion

The aim of this study was to determine if 2PF, a volatile metabolic product of A. fumigatus, is indeed unique and if it could be a possible volatile biomarker for the diagnosis of aspergillosis in breath. We have demonstrated that 2PF was produced by both A. fumigatus and Fusarium spp. on all four media tested suggesting that production is a function of the micro-organism rather than a function of the nutritional base. Previous studies have shown that 2-pentylfuran is also produced by A. fumigatus when the fungus is cultured on malt extract and plasterboard supporting the contention that this occurs under a wide variety of conditions 14. In contrast 2PF production by S. apiospermum, A. terreus, and A. niger was not detected when the fungi were cultured on nutrient agar and it may not be as reliably produced as with A. fumigatus.

2-pentylfuran production appears to be relatively specific but not unique for filamentous fungi in vitro. It was not detected in cultures of Candida albicans or any of the bacterial strains except S. pneumonia. Additional volatile compounds produced by S. pneumonia but not by fungi have been detected suggesting that it may be possible to distinguish between these pathogens using GC/MS.

None of the healthy individuals showed detectable levels of 2PF on the breath which is in keeping with one previous study which identified more than 3000 compounds in human breath but 2PF was not among the fifty most frequently occurring VOC's 16. In contrast 2PF was detected in the breath samples of all four patients colonized by A. fumigatus and of 3 of 7 patients without laboratory evidence of Aspergillus colonization. The positive results returned by some of the individuals not previously proven to have Aspergillus colonization may be explained in a number of ways. First, these patients may in fact represent cases of unrecognized aspergillus colonization, a not unlikely possibility, given the difficulties of diagnosing Aspergillus colonization (and hence ABPA) in the paediatric CF population. Second, detection of 2PF may be a non-specific product of inflammation in the lung associated with bacterial or fungal colonization or other metabolic processes associated with cystic fibrosis.

The important point to emerge from this data set is that healthy, normal individuals appear not to produce 2PF while it may be detected with many patients with pulmonary colonization by organisms that produce 2PF. These experiments are preliminary in nature and describe a very limited patient population that fail to take into account several potentially confounding factors. Despite these concerns, detection of volatile organic compounds produced by pathogens resident in the lungs may explain these findings and opens the possibility of non-invasive diagnosis of pulmonary and possibly disseminated infections. The significance of these findings will depend on carefully controlled clinical studies of defined patient groups. The main target population will be the immunosuppressed host in which invasive aspergillosis is a particular diagnostic and treatment problem. In most instances bacterial infection should not be a confounding problem in this patient population as most are heavily pre-treated with antibiotic therapy.

Conclusion

In conclusion, this in vitro study has confirmed that 2PF is consistently produced by A. fumigatus but also by other fungi such as Fusarium spp., S. apiospermum, A. terreus, A. flavus and A. niger which represent a broad range of fungal pathogens in the immunocompromised host. As 2PF has not been reported to occur in the breath of healthy humans we conclude that 2PF might indeed be a possible volatile biomarker in breath for fungal infections of the lung. Proof will depend on further well conducted clinical trials. This study was funded in part by the Health Kids New Zealand and Lotteries Health, New Zealand.

Acknowledgements

We wish to thank Ros Podmore of the Microbiology Laboratory of Canterbury Health Laboratories and Karen Rogers, from the mycology service of Lab Plus Microbiology Laboratory, Auckland, for their help in identifying and supplying the fungal strains used in this study. This study was funded in part by the Health Research Council of New Zealand, Cure Kids, New Zealand and Lotteries Health, New Zealand.

References

1.
Patterson
TF
Denning
DW
Mandell
Bennett
Dolin
Aspergillus species
Principles and Practice of Infectious Diseases
 , 
2005
6th edn
Edinburgh
Churchill Livingstone
(pg. 
2958
-
2973
In
 
Ch. 256
2.
Ullmann
AJ
Lipton
JH
Vesole
DH
, et al.  . 
Posaconazole or fluconazole for prophylaxis in severe graft-versus-host disease
N Engl J Med
 , 
2007
, vol. 
356
 (pg. 
335
-
347
)
3.
Herbrecht
R
Denning
DW
Patterson
TF
Bennett
JE
, et al.  . 
Invasive Fungal Infections Group of the European Organisation for Research and Treatment of Cancer and the Global Aspergillus Study Group. Voriconazole versus amphotericin B for primary therapy of invasive aspergillosis
N Engl J Med
 , 
2002
, vol. 
347
 (pg. 
408
-
415
)
4.
Lin
SJ
Schranz
J
Teutsch
SM
Aspergillosis case-fatality rate: systematic review of the literature
Clin Infect Dis
 , 
2001
, vol. 
32
 (pg. 
358
-
366
)
5.
Levy
H
Horak
DA
Tegtmeier
BR
Yokota
SB
Forman
SJ
The value of bronchoalveolar lavage and bronchial washings in the diagnosis of invasive pulmonary aspergillosis
Respir Med
 , 
1992
, vol. 
86
 (pg. 
243
-
248
)
6.
Scotter
JM
Campbell
P
Anderson
TP
, et al.  . 
Comparison of PCR-ELISA and galactomannan detection for the diagnosis of invasive aspergillosis
Pathology
 , 
2005
, vol. 
37
 (pg. 
246
-
253
)
7.
Maertens
J
Verhaegen
J
Demuynck
H
, et al.  . 
Autopsy-controlled prospective evaluation of serial screening for circulating galactomannan by a sandwich enzyme-linked immunosorbent assay for hematological patients at risk for invasive aspergillosis
J Clin Microbiol
 , 
1999
, vol. 
37
 (pg. 
3223
-
3228
)
8.
Sulahian
A
Boutboul
F
Ribaud
P
, et al.  . 
Value of antigen detection using an enzyme immunoassay in the diagnosis and prediction of invasive aspergillosis in two adult and pediatric hematology units during a 4-year prospective study
Cancer
 , 
2001
, vol. 
91
 (pg. 
311
-
318
)
9.
Verweij
PE
Rijs
AJ
De Pauw
BE
, et al.  . 
Clinical evaluation and reproducibility of the Pastorex Aspergillus antigen latex agglutination test for diagnosing invasive aspergillosis
J Clin Pathol
 , 
1995
, vol. 
48
 (pg. 
474
-
476
)
10.
Costa
C
Costa
JM
Desterke
C
, et al.  . 
Real-time PCR coupled with automated DNA extraction and detection of galactomannan antigen in serum by enzyme-linked immunosorbent assay for diagnosis of invasive aspergillosis
J Clin Microbiol
 , 
2002
, vol. 
40
 (pg. 
2224
-
2227
)
11.
Chamilos
G
Kontoyiannis
DP
Defining the diagnosis of invasive aspergillosis
Med Mycol
 , 
1995
, vol. 
44
 (pg. 
163
-
172
)
12.
Magan
N
Evans
P
Volatiles as an indicator of fungal activity and differentiation between species, and the potential use of electronic nose technology for early detection of grain spoilage
J Stored Prod Res
 , 
2000
, vol. 
36
 (pg. 
319
-
340
)
13.
Schnurer
J
Olsson
J
Borjesson
T
Fungal volatiles as indicators of food and feeds spoilage
Fungal Genet Biol
 , 
1999
, vol. 
27
 (pg. 
209
-
217
)
14.
Gao
P
Korley
F
Martin
J
Chen
BT
Determination of unique microbial volatile organic compounds produced by five Aspergillus species commonly found in problem buildings
AIHA J (Fairfax Va)
 , 
2002
, vol. 
63
 (pg. 
135
-
140
)
15.
Larone
DH
Medically Important Fungi. A Guide to Identification
 , 
2002
4th edn
Washington, DC
ASM Press
16.
Phillips
M
Herrera
J
Krishnan
S
, et al.  . 
Variation in volatile organic compounds in the breath of normal humans
J Chromatogr B Biomed Sci Appl
 , 
1999
, vol. 
729
 (pg. 
75
-
88
)