Chromatographia

, Volume 77, Issue 5, pp 501–509

Volatile Organic Compounds Analysis in Breath Air in Healthy Volunteers and Patients Suffering Epidermoid Laryngeal Carcinomas

Authors

    • Department of Chemical and Environmental TechnologyRey Juan Carlos University
  • Victoria Morales
    • Department of Chemical and Environmental TechnologyRey Juan Carlos University
  • Sergio Martín
    • Department of Chemical and Environmental TechnologyRey Juan Carlos University
  • Estela Vilches
    • Department of Chemical and Environmental TechnologyRey Juan Carlos University
  • Adolfo Toledano
    • Department of OtorhinolaryngologyFoundation Hospital Alcorcón
Original

DOI: 10.1007/s10337-013-2611-7

Cite this article as:
García, R.A., Morales, V., Martín, S. et al. Chromatographia (2014) 77: 501. doi:10.1007/s10337-013-2611-7

Abstract

Exhaled breath contains thousands of gaseous volatile organic compounds (VOCs) that may be used as non-invasive markers of head and neck epidermoid cancer. We hypothesized that solid phase micro-extraction coupled to gas chromatography–mass spectrometry can discriminate patients with epidermoid head and neck cancer from healthy controls by analyzing the gaseous volatile organic compounds, VOC-profile, in exhaled breath, thus identifying some non-invasive biomarkers to be used in early detection. Twenty healthy subjects participated in a cross-sectional study plus 11 patients with epidermoid supraglottic laryngeal cancer. VOCs from T3 supraglottic cancer were clustered distinctly from those of T1 and healthy subjects. Up to seven VOCs were detected differently from healthy volunteers, mainly 2-butanone and ethanol. Thus VOC-patterns of exhaled breath may discriminate patients with epidermoid head and neck cancer from healthy controls.

Keywords

GC–MSSPMEVOCsLaryngeal carcinomaBiomarkers

Introduction

Neoplasms of the head and neck are particularly important because surgical treatment may cause extensive aesthetic deformities and impair important functions such as ingestion and speaking. Overall, head and neck carcinomas represent 5–10 % of malignant tumors diagnosed annually in Spain, and cause about 5 % of cancer deaths [1]. About 500,000 new cases are diagnosed each year, representing the sixth most common cancer worldwide [2]. Over 90 % of tumors in this region are epidermoid laryngeal carcinomas. The most common head and neck cancer is in the larynx, even though other points are the oral cavity, oropharynx, hypopharynx, and nasopharynx [3, 4]. The incidence of this group of tumors has been increasing over the past 30 years, with the most notable increase in tumors of the oral cavity and pharynx. Some 90 % of epidermoid carcinomas of head and neck are related to lifestyle, especially with the consumption of cigarettes and alcohol. The incidence of larynx cancer estimated in Spain in 2000 was 0.33 cases per 100,000 inhabitants for women, and 19.91 cases per 100,000 inhabitants for men, representing, in the case of males, the highest among European Union countries [2].

There are currently no effective programs for early detection of head and neck cancer. Close monitoring is recommended for people with known risk factors, such as heavy smokers and drinkers [5]. On the other hand, no tumor marker has yet been able to identify epidermoid carcinomas of the head and neck to help early diagnosis. There are some early molecular events in carcinogenesis of head and neck, but their utility as a tumor marker has yet to be confirmed. Furthermore, the detection of some oncogenes, tumor suppressor genes, and DNA repair genes help us to provide prognostic information for some head and neck tumors, but do not allow early detection programs [6]. Since the prognosis of these patients depends on the detection of cancer in early stages, it seems that it might of interest to develop a method for early diagnosis of disease in those patients with risk factors.

There exists strong evidence suggesting that some types of cancers may be detected through the analysis of exhaled air [710]. In 1971, Pauling et al. [11] identified over 200 volatile organic compounds (VOCs) present in exhaled air. This work has provided a starting point for research, allowing the association of these compounds with certain diseases [810]. Since 1985, further studies have been conducted in order to clarify the spectra of exhaled air, since a spectra of over 200 compounds may be very complicated to interpret. From statistical studies between healthy and cancer-diseased lungs, Gordon et al. and O’Neill et al. [10, 12] identified some potentially significant compounds for diagnosis. The search for organic compounds is not just limited to the information collected through exhaled air. This information is complementary to that obtainable from other samples: sweat, urine, and blood, which have already provided positive results in other applications such as drug analysis [13, 14].

The technique of solid phase micro-extraction (SPME) may be useful for the analysis of exhaled air. It is one of the methods used to concentrate volatile organic compounds emitted by different sources [15, 16]. The equipment is not expensive,is easy to use, and its small size makes it an easily transportable device [17, 18]. It was developed in the late 1980s by Arthur and Pawliszyn [19], and the method has been applied successfully in a wide variety of fields: environment, toxicology, pharmacology, and food and drug analysis. The coupling of gas chromatography with mass spectrometry allows the separation of the compounds to be carried into the chromatograph, while the identification is performed through the mass spectrometer [2024].

This work is intended to investigate the viability for detecting some VOCs in the air breathed out by patients, with the potential for them to be used as biomarkers in the early detection of epidermoid laryngeal carcinomas. The SPME for analysis of exhaled air combined with gas chromatography–mass spectrometry (SPME/GC–MS) was used as our analytical technique.

Experimental

Chemical and Materials

Analytical standards used were: indoor air standard 50 component, 1,000 μg mL−1 each component in methanol:water (97:3) (Aldrich); C1–C6 n-paraffins, 15 ppm each component in nitrogen (Fluka); and BTEX mix in nitrogen, 10 ppm each component in nitrogen (Fluka). All SPME fibers assemblies, fiber holders for manual and automatic sampling, 0.75 mm liners for SPME, and 5-L Tedlar Bags were from Supelco.

Sample Preparation

Samples from 10 healthy non-smokers, 10 healthy smokers, and 11 people with larynx cancer at different stages, summarized in Table 1, were collected in 5-L Tedlar sampling bags. For cleaning, the bags were initially flushed three times with nitrogen. Samples have been taken after at least 8 h of fasting in order to minimize the number of VOCs from the ingestion of food or drink that could alter the analysis results obtained. Air samples were stored under room temperature and atmospheric pressure, proceeding to their injection into the gas chromatograph–mass spectrometer, if possible, on the day of collection. The study was approved by the ethics committee of the Alcorcón Hospital. The research was conducted in compliance with international recommendations on clinical research in the Declaration of Helsinki.
Table 1

Description of the subject studied

Subject

Age (years)

Sex

T-stage

Description

1

51

Female

Healthy non-smoker

2

54

Female

Healthy non-smoker

3

60

Male

Healthy non-smoker

4

59

Male

Healthy smoker

5

54

Male

Healthy smoker

6

58

Male

Healthy smoker

7

27

Female

Healthy non-smoker

8

29

Female

Healthy non-smoker

9

29

Male

Healthy smoker

10

28

Male

Healthy smoker

11

53

Male

T3N0M0

Supraglottic

Laryngeal carcinoma

11′

58

Male

T3N0M0

Supraglottic

Laryngeal carcinoma

12

48

Male

T3N0M0 glottic

Laryngeal carcinoma

12′

50

Male

T3N0M0 glottic

Laryngeal carcinoma

13

82

Male

T1N0M0 glottic

Laryngeal carcinoma

13′

68

Male

T1N0M0 glottic

Laryngeal carcinoma

14

53

Female

T1N0M0 glottic

Laryngeal carcinoma

14′

64

Male

T1N0M0 glottic

Laryngeal carcinoma

15

75

Male

T1N0M0

Laryngeal carcinoma

15′

62

Male

T1N0M0

Laryngeal carcinoma

16

58

Male

T3N0M0

Supraglottic extirpated 1 year ago

Laryngeal carcinoma

SPME Procedure

Once the patient air breath has been transferred to the Tedlar Bags, SPME sorption was performed using several fibers: CAR/PDMS, DVB/PDMS, PDMS, and CAR/PDMS/DVB. The extraction was performed at 25 °C within 15 and up to 120 min in order to optimize the adsorption time. The fiber was introduced into the GC injector for analysis immediately after sampling. Two bags of each sample were analyzed by this procedure to ensure reproducibility. The fibers were initially conditioned according to the instructions of the manufacturer in order to remove contaminants and to stabilize the solid phase. Conditioning was carried out in an extra split/splitless port with helium carrier gas prior to each adsorption.

Instrumentation

To obtain the experimental data, a mass spectrometer Saturn 2000 ion trap GC/MS (Varian) was employed. The chromatographic column used was of type VF-5 ms capillary FOUR factor (Varian). The column length was 60 m, internal diameter 0.25 mm, and the thickness of the stationary phase was 0.25 μm. A splitless injection mode was selected with a valve off-time of 5 min; during this time, the SPME fiber was inserted into the gas chromatograph injector at 175 °C for 5 min, since the compounds adsorbed by the fiber are extracted with increasing temperature. The initial temperature of the column was held at 35° for 5 min, and then programmed to rise to 115° at 5 °C min−1 and to 250 °C at 20 °C min−1. Helium was used as the carrier gas at a flow rate of 1 mL min−1. The mass spectrometer was operated in electron impact ionization mode at 70 eV. The electron-impact (EI) mass spectra of the analytes were recorded in scan mode (scan range 35–280 m/z) to determine retention times and characteristic mass fragments.

Results and Discussion

To conduct this study, ten healthy non-smokers and ten healthy smokers were voluntarily chosen, in order to perform a preliminary study of the VOCs present in the exhaled air of healthy people, and, therefore, to confirm that the air exhaled by all subjects analyzed follows a certain pattern. The comparison was made by age, and exhaled air bags from people with an age range between 20 and 60 years were analyzed. After studying the characteristics, in terms of the composition of the exhaled air from a population of healthy smokers and non-smokers at different ages, the exhaled air from people with larynx cancer at different stages was analyzed: four patients with supraglottic epidermoid larynx cancer T3, six patients with supraglottic epidermoid larynx cancer T1, and one patient with supraglottic epidermoid larynx cancer T3 who had been operated on and the tumor removed.

The technique of SPME is based on the adsorption of compounds from the sample of exhaled air to the adsorbent fiber, placed on the SPME equipment, leaving a sufficient contact time for the analyte concentration to either reach or get close to adsorption equilibrium. Thus, to test real samples, it is necessary to optimize the method of compound adsorption/concentration with the aim of identifying the greatest number of VOCs present in the air. Thus, two parameters that exert significant influence on the SPME technique with gas samples were optimized: the nature of the adsorbent phase SPME fiber and the time of adsorption.

Study of the Influence of Adsorbent Fiber Type

For the analysis of VOCs in exhaled air, the behaviors of four different adsorbent fiber types were evaluated to select a suitable fiber for detecting volatile compounds. Samples of exhaled air, under the same conditions of temperature and pressure, were studied in order to choose the adsorbent type that is able to retain the maximum amount of VOCs. For this purpose, four SPME fiber types that differ in the type of adsorbent filler were tested: polydimethylsiloxane (PDMS), characterized by having an apolar behavior, Carboxen and polydimethylsiloxane (CAR/PDMS), polydimethylsiloxane divinylbenzene (DVB/PDMS) and Carboxen, and finally polydimethylsiloxane and divinylbenzene (CAR/PDMS/DVB), all of them with hybrid polar/apolar nature. Table 2 summarizes the VOCs detected in the full chromatograms for the fibers used in the study. The adsorption efficiencies on the SPME fibers were evaluated by comparing the number of compounds detected and the peak areas. The highest efficiency among the four fibers was obtained using polydimethylsiloxane/divinylbenzene, since the highest number of VOCs were obtained, as displayed in Table 2.
Table 2

Identified compounds in terms of the SPME fiber type in use

Extraction

No. VOCs

Compounds

PDMS

11

Acetone, toluene, α-pinene, β-pinene, n-decane, limonene,n-undecane, n-tridecane, n-tetradecane, n-pentadecane, n-hexadecane

CAR/PDMS

15

Acetone, 2-butanone, n-hexane, bromodicloromethane, isoctane, n-heptane, toluene, ethylbencene, m+p-xylene, o-xylene, α-pinene, β-pinene, 1,2,4-trimethylbencene, n-decane, limonene

DVB/PDMS

25

Acetone, isopropanol, ethyl acetate, n-heptane, toluene, butyl acetate, n-octane, ethylbencene, m+p-xylene, estirene, o-xylene, n-nonane, α-pinene, 3-ethyltoluene+4-ehyltoluene, 1,3,5-trimethylbencene, 2-ethyltoluene, β-pinene, 1,2,4- trimethylbencene, n-decane, 1,4-dichlorobencene, limonene, n-undecane, n-DODECANE, n-tridecane, n-tetradecane

CAR/PDMS/DVB

24

Acetone, isopropanol, 2-butanone, bromodicloromethane, isoctane, n-heptane, toluene, butyl acetate, n-octane, ethylbencene, m+p-xylene, o-xylene, n-nonane, α-pinene, 2-ethyltoluene, β-pinene, 1,2,4-trimethylbencene, n-decane, 1,4-diclorobencene, limonene, n-undecane, n-dodecane, n-tridecane, n-tetradecane

Adsorption Time Optimization

Once having established the type of fiber to be used, the adsorption time featuring each pair of analytes/fibers was the next parameter to be considered. To determine the optimal, four different times were chosen from 15 to 120 min, using the fiber DVB/PDMS chosen as optimal in the previous section.

In the analyzed chromatograms (not shown), the influence of adsorption time is much more pronounced in the compounds shown in the third part of the chromatogram, in which a significant increase of compounds, from 15 to 30 min, is noticed. After examining the results, in general, the chromatographic peak area of the compounds augments as the adsorption time increases. The longer the contact time between the fiber and the sample, the greater the concentration retained. However, this trend is fulfilled up to near the equilibrium time, when the analyte concentration remains constant. A maximum area was achieved by 60 min, and therefore this was established as optimal for the further studies.

Determination of the Limit of Detection (LOD) in the SPME/GC–MS Process

The Limit of Detection (LOD) is the lowest concentration of analyte in a sample that can be detected. The instrument LOD is an important factor to be determined because the lack of certain compounds may be due either to their absence in the exhaled air, or because their concentration is below the LOD. In both cases, the analytes would be undetectable by the analytical equipment used.

n-Hexane was used as reference compound, by diluting a standard mixture at different concentrations (500, 100, 50, and 10 ng mL−1), and thereafter concentrated with the SPME technique, and analyzed by GC–MS. In the proposed method, the LOD was 40 ng mL−1, which may be sufficient for the analysis of VOCs, taking into account that the concentration of the exhaled air is estimated in the literature to be in the range 10−9–10−12 M [7, 26].

Feasibility Study of Tedlar Bags Storage

As noted above, the exhaled air is stored in Tedlar bags, which are sealed and easily transportable. However, during the analysis of the samples, it was observed that the chemical composition of the bag interfered with the results obtained, due to adsorption of volatile compounds from the bag into the SPME fiber.

The chromatographic analysis showed that, during the first 20 min, no presence of compounds was detected, and therefore the identified peaks in real samples in this time interval are inherent to the composition of exhaled air, and no subtraction of the blank is needed. After 20 min, two main compounds were detected: dimethyl acetamide and phenol, with a relatively high intensity. In addition, during the study of the real samples, these two compounds have the highest intensities in the chromatographic analysis. As a consequence, the different compounds released by the sample bag were subtracted from the real samples, since these compounds came from the decomposition of plastic bags [2729].

Analysis of Real Samples of Healthy Non-Smokers and Smokers

Figures 1 and 2 display the air exhaled chromatograms by five non-smokers and five smokers segregated by age as representative examples of the volunteers. The chromatograms were divided into three time regions to facilitate analysis and identification of compounds (10–20, 21–24, and 25–28 min). After thorough analysis of the chromatograms, seven chromatographic peaks have been excluded as being identical in intensity, retention time, and ion spectra of compounds from the Tedlar bags. Identification of the compounds was accomplished by using the mass spectra and retention times of authentic standards. Chromatographic peaks were also identified with the help of the NIST mass spectrum library, using similarity indexes higher than 75 % and with authentic standards, i.e., indoor air standard 50 component, 1,000 μg/mL each component in methanol:water (97:3), C1–C6 n-paraffins 15 ppm each component in nitrogen, and BTEX mix in nitrogen, 10 ppm each component in nitrogen.
https://static-content.springer.com/image/art%3A10.1007%2Fs10337-013-2611-7/MediaObjects/10337_2013_2611_Fig1_HTML.gif
Fig. 1

GC chromatogram of exhaled breath air in healthy non-smoker subjects

https://static-content.springer.com/image/art%3A10.1007%2Fs10337-013-2611-7/MediaObjects/10337_2013_2611_Fig2_HTML.gif
Fig. 2

GC chromatogram of exhaled breath air in healthy smoker subjects

From the study of the chromatograms, it can be concluded that the air exhaled by healthy people, both smokers and non-smokers, is characterized by exhibiting similar chromatograms, making possible the comparison between samples, and allowing to the creation of a typical pattern for the analysis of air exhaled by healthy people. An average of 45 VOCs present in exhaled air was detected for each subject. These results are consistent with those obtained by Phillips [26] for the analysis of VOCs present in air exhaled by humans, these authors identifying up to 100 compounds using other systems for sample pre-concentration.

Studying the patient ages, small differences can be found, as shown in the chromatograms of Fig. 1 for non-smoker subjects. Subjects over 50 (1, 2, and 3) display slightly greater numbers of VOCs than subjects over 20–30, during the first 20 min of analysis. At least, two clear compounds are distinguished, ethyl acetate and acetic acid. Additionally, a 15 % higher intensity in the peaks areas from the older subjects can be clearly seen.

In the chromatograms obtained when analyzing smokers patients, no significant differences comparing the patient’s age are noticed, indicating that the age factor for this population is not significant. In this category, the exhaled air is conditioned by the consumption of cigarettes, obtaining higher intensities in the subjects over 50–60 years. However, performing a comparison between smoker and non-smoker subjects aged between 20 and 30, there are clear differences in the peaks of the chromatograms, not only regarding the different compounds but also in the intensity of some common analytes to both families of subjects.

By comparing both age families, a greater number of compounds are detected in the chromatograms of subjects aged between 50 and 60, regardless of their status as a smoker or not. Focusing on the subjects between 50 and 60, certain differences between smokers and non-smokers are found, similar to the subjects in their 20–30s. For some common compounds, there are differences in intensities. Concentrations of certain compounds, in the case of smokers, are indeed relevant; for example, n-nonane with an increase of the peak area around 50 %.

Additionally, four possible potential compounds that may be indicative of potential marker for smokers have been detected. These compounds were only identified in these samples, while undetected in the breathed out air from non-smokers [8, 25]. The compounds are benzene (diagnostic ions at m/z 78 and 51), furaldehyde (diagnostic ions at m/z 96, 67, and 39), 4-isobutyl-1-(1-hydroxyethyl)-benzene diagnostic ions at m/z 122, 91, and 65) and 2,3,5-trimethylhexane (diagnostic ions at m/z 85, 57, and 43).

Analysis of Exhaled Air by People with Larynx Cancer

Figure 3 shows the total ion chromatograms obtained after analyzing exhaled air samples during the first and second time intervals (up to 20 min of analysis). Patients with supraglottic epidermoid larynx cancer T3 present high concentrations of certain compounds with 6.45 and 10.71 min retention time, which are unnoticed in samples of healthy people. Compound identification in real samples was made by comparison of experimental retention times and background corrected mass spectra with those of pure compounds and matching with NIST mass spectral Search Program for the NIST/EPA/NIH Mass Spectral Library Version 2.0. For an accurate identification, Fig. 4 displays the chromatogram at m/z 45 and mass chromatogram corresponding to ethanol and 2-butanone, respectively. The mass spectrum was also featured by intense ions at m/z 45, 31, 29, and 15 in the case of ethanol, and m/z 57, 43, 29, and 15 for 2-butanone. These ions are representative of the co-eluting aliphatic hydrocarbons, which underlines the importance of likely interferences caused by the abundant compounds in the exhaled air sample and the need to use a selective detector (i.e., MS). In addition, five compounds were identified which appear only in the case of patients with supraglottic epidermoid larynx cancer T3, shown in Fig. 5, which illustrates the total ion chromatograms for the patients in the study, with five representative peaks in T3 patients assigned to the analytes 2,3-butanediol, 9-tetradecen-1-ol, octane derivative compound, cycloheptane derivative compound, and cyclo-nonane derivative compound (diagnostic ions included in Table 3 for compound numbers 3 up to 7), respectively, and according to the NIST/EPA/NIH Mass Spectral Library.
https://static-content.springer.com/image/art%3A10.1007%2Fs10337-013-2611-7/MediaObjects/10337_2013_2611_Fig3_HTML.gif
Fig. 3

GC chromatogram of 5–20 min retention time of exhaled breath air in people with T1 and T3 laryngeal carcinoma (subjects 14 and 11, respectively), supraglottic carcinoma extirpated 1 year ago (subject 16), and smoker and non-smoker healthy subjects (subjects 4 and 1, respectively)

https://static-content.springer.com/image/art%3A10.1007%2Fs10337-013-2611-7/MediaObjects/10337_2013_2611_Fig4_HTML.gif
Fig. 4

GC–MS chromatogram of ethanol and 2-butanone biomarker compounds compared in different patients

https://static-content.springer.com/image/art%3A10.1007%2Fs10337-013-2611-7/MediaObjects/10337_2013_2611_Fig5_HTML.gif
Fig. 5

GC chromatogram of 20–30 min retention time of exhaled breath air in people with T3 laryngeal carcinoma (subject 11), supraglottic carcinoma extirpated 1 year ago (subject 16), non-smoker healthy subject (subject 1), and smoker healthy subject (subject 4)

Table 3

Potentially markers compounds of laryngeal carcinoma

 

Compound

Retention time (min)

m/za

m/zb

1

Ethanol

6.450

46

45, 31, 29

2

2-Butanone

10.710

72

57, 43, 29

3

2,3-Butanediol

19.510

90

57, 45, 43, 29

4

9-Tetradecen-1-ol

24.870

212

109, 82, 55

5

Octene derivative

25.290

126

83, 70, 55

6

Cycloheptane derivative

25.565

108

93, 80, 39

7

Cyclononane derivative

25.879

136

101, 94, 91, 79

aMolecular ion m/z

bIon fragmentation m/z used for identification

Meaningfully, no presence of significant concentrations of these compounds in the case of supraglottic epidermoid larynx cancer T1 was found, showing a chromatographic pattern very similar to healthy subjects. Table 3 shows the potential markers that have been found from the analysis of subjects with supraglottic epidermoid larynx cancer T3, after comparison with the compounds found in patients with supraglottic epidermoid larynx cancer T1, and healthy patients, both smokers and non-smokers.

The study of the exhaled air of patients with laryngeal cancer at different stages determines that the analysis by GC–MS is a useful technique for identifying compounds that are potential biomarkers of epidermoid carcinomas of head and neck. Up to seven different compounds have been identified, regarding patients with the same disease in a less advanced stage, and healthy subjects, the most significant being ethanol and 2-butanone. Analyzing the results, we conclude that VOCs appear only in locally advanced tumors. Further investigations improving the analytical technique and the concentrations to be detected will be necessary to extend the results to small tumors and to the histological characteristics of the tumor.

Indeed, although these results are preliminary, an interesting path is open to explore the development of devices (electronic noses) to detect volatile substances that are potential markers of respiratory diseases, in this case of epidermoid larynx carcinomas of the head and neck.

Conclusions

In the light of the results, we can conclude that the optimized method of coupling the techniques of solid phase micro-extraction, SPME, and GC–MS can identify volatile organic compounds in exhaled air, and also discriminate between subjects of different ages, and smokers and non-smokers. The method is feasible as a starting point for a study of patients with laryngeal cancer, since some compounds that may have potential utility as biomarkers in these diseases have been identified.

The sample should always be collected in the same room where the ambient air has been previously analyzed. The patient must be fasting 8 h before blowing, and no alcohol or cigarettes must be consumed. The technique of solid phase microextraction, SPME, concentrated the compounds present in exhaled air samples between 100 and 500 times. Additionally, the variables that most influence the effectiveness of the process were optimized, such as the type of adsorbent phase, concluding that the fiber DVB/PDMS provides the best results, and the extraction time, which is set at 60 min.

A total of 31 common VOCs in non-smokers have been determined, whereas smokers reach up to 45 VOCs, thus differentiating the two types of populations. The comparative analysis between healthy smokers and non-smokers reveals four marker compounds related to cigarette consumption, since they have only been detected in samples from smokers. These compounds are: benzene, furaldehyde, 4-isobutyl-1-(1-hydroxyethyl)-benzene, and 2,3,5-trimethylhexane.

The study of the exhaled air of patients with laryngeal cancer at different stages determines that the analysis by GC–MS is a useful technique for identifying compounds that are potential biomarkers of epidermoid carcinomas of the head and neck. Up to seven different compounds have been identified, regarding patients with the same disease in a less advanced stage, and healthy subjects, the most significant being ethanol and 2-butanone.

Acknowledgments

We would like to thank all the participants in our study, especially those who are suffering from cancer. The financial support of Fundación Mútua Madrileña and the Spanish government (CTQ2008-05909/PPQ and CTQ2011-22707 projects) is gratefully acknowledged.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013