Doctors may soon be able to check your health by analysing your breath. Nick Houtman reports
Doctors may soon be able to check your health by analysing your breath. Nick Houtman reports
Imagine sitting in your doctor’s examining room having your annual medical. In addition to the usual poking and prodding, the doctor asks you to blow into a small device about the size of a mobile phone. S/he studies data displayed on a small LCD screen, makes a few adjustments and asks for another breath sample. The results, s/he says, suggest that your ulcer has cleared up nicely but that you might come down with a cold in a day or two.
No such device exists, of course, but chemists and engineers in Europe, the US, Japan and Israel are busy laying the foundations. Their goal is to produce technology that can rapidly, reliably and accurately measure any of more than 400 chemical compounds found in air exhaled from the respiratory tract. Although devices to collect and analyse breath samples have been around for years, none of them fit the toolkit, not to mention the budget, of most doctors’ surgeries, A&E departments or clinics. Yet rapid breath analysis could be a lifesaver in monitoring health and responding to emergencies.
Technologies to identify the components of complex gaseous mixtures are not new. For over 50 years, police have monitored alcohol levels with breathalysers based on wet chemistry, and now they have smaller devices that use infrared spectroscopy and fuel cells. Electronic noses designed to detect odorous compounds in ambient air have been in development for more than 10 years, and some devices are now on the market. These systems use a variety of approaches: chemiresistive metal oxides, mid-IR spectroscopy and GC-mass spectrometry. Uses include quality control in perfume production, sniffing out illegal drug labs and alerting military inspectors to the presence of chemical weapons.
Not surprisingly, human breath research is driven primarily by medical needs. Researchers have discovered that breath constituents can provide clues about liver, pancreas and kidney function, ulcer-causing bacterial infection, lung and prostate cancer, asthma and schizophrenia. Blood chemistry affected by these conditions provides the tell-tale signs. It’s the thin membrane between blood and air in the lung that allows diffusion of volatile organics into breath, and vice versa. It may be years before taking a person’s breath print will be part of a standard medical check-up or A&E procedure, but scientists have announced significant advances in recent years.
Innovision of Denmark has developed a device that uses infrared photoacoustics to determine levels of the six different compounds in a single breath sample. The European Space Agency has delivered this technology to Nasa for monitoring the health of astronauts during space missions. Researchers at Keele University have worked with doctors at North Staffordshire Hospital to adapt the selected ion flow tube coupled with a mass spectrometer, or Sift-MS, for breath analysis in patients with kidney disease. Other breath analysis systems are being developed in Israel, Germany, Japan and the US.
Researchers at the University of Maine in the US have focused on two primary tasks: accurately characterising the range and variability of breath constituents using a Fourier transform-ion cyclotron resonance mass spectrometer (FT-ICR MS), equipped with a pre-concentrator and a GC; and developing a portable nitric oxide sensor based on chemiresistive thin film platforms. A chemiresistive sensor works on the principle that chemical reactions between adsorbed gases and the sensor surface induce resistance changes in the underlying sensor film.
According to UMaine chemist Touradj Solouki, FT-ICR MS has the highest mass resolution of any mass spectrometry technique, differentiating between molecules separated by fewer than one in 100,000 mass units. Sensitivity is comparable to GC interfaced with quadrupole MS, at the parts per billion level. Thus, it is Solouki’s hope that FT-ICR technology will set a gold standard against which other portable sensor technologies can be measured.
FT-ICR works by measuring the unique cyclotron frequency of ions as they move through a homogeneous magnetic field. Moreover, by storing ions in the magnetic field, researchers can lengthen the data collection time, improving both sensitivity and resolution. In tests on human breath over the past year, Solouki and his colleagues used a 7T IonSpec FT-ICR MS coupled to a GC and two different breath pre-concentrators.
Pre-concentration of the analytes is a crucial step. Many breath compounds occur at 10-9 and 10-12 g l-1 levels, says Solouki, and concentrating the sample increases the chances that they will be detected. Just as important is removal of the dominant constituents. Trace gases can be overwhelmed by O2, N2, H2O and CO2, which comprise over 95 p er cent of a typical breath sample. Solouki’s team uses GC separation to introduce the pre-concentrated analytes into the FT-ICR MS for accurate analysis.
In addition, Solouki and his colleague David Labrecque have written a software program called Mass Identification Smart Tool (Mist), to analyse data and identify compounds. By calculating possible combinations of atoms and comparing them with actual data, Mist rapidly generates a unique picture of the possible types and amounts of compounds that might be present in an individual’s breath sample.
Adopting FT-ICR and Mist to create a medically useful database of breath profiles will require collaboration among clinicians and researchers. ’We have to examine statistical variations of the experimental data from smokers and non-smokers, females and males, young and old’, says Solouki. ’Once we have acquired that information, we need to identify biomarkers that will tell us something about the health status of the donor. We can use the presence or absence of the identified biomarkers in a breath sample, and we can develop sensors to detect or monitor identified biomarkers that we think are important.’ With recent funding from the US National Science Foundation (NSF), Soloukis’s team is working to improve the sensitivity of PC/GC FT-ICR MS analysis.
UMaine efforts in breath analysis grew out of ongoing work on thin metal oxide films and a project funded by the US National Institutes of Health (NIH) at Sensor Research and Development (SRD) in Orono, Maine. NIH’s goal was to develop a less expensive and more reliable way of monitoring nitric oxide (NO) in breath. NO relaxes blood vessels, acts as a neurotransmitter and mediates many bodily functions such as our immune response to pathogens and other toxic agents (Chem. Br., July 2000, p30). Being able to detect it in breath has become a desirable clinical objective.
A hot molecule
SRD worked closely with the Laboratory for Surface Science and Technology (LASST) at UMaine to secure additional support for NO research through the US Defense Advanced Research Projects Agency (Darpa). Darpa’s interest stems from concerns about the potential exposure of military personnel to biological weapons, says Carl Freeman, president of SRD. ’They are interested in the possible diagnosis of exposure and the respiratory distress associated with it. The symptoms are typically flu-like in the early stages.’ Elevated NO levels in breath, it turns out, may provide an early warning of exposure prior to the appearance of headache, nausea or other symptoms.
’Nitric oxide has been a hot molecule in the medical community for a number of years now’, says Robert Lad, director of LASST. ’You can buy a detector for about $30,000 [ca ?20,000]. It weighs 80lbs [36kg] and does the job. Every major hospital probably has one or two of these instruments to look at NO from a medical standpoint’.
In addition to monitoring disease, doctors may one day look at NO detection to calibrate drug delivery, Lad says. Since NO levels in the breath rise with the strength of an infection or other health problem, changes in NO can indicate the effectiveness of a given medication. ’Real time diagnostics of NO levels can be done to watch how a drug interacts with the body’ he adds.
However, existing detectors are too heavy and expensive to be routinely used in the field by emergency personnel. Thus there is a need for a new reliable detection method based on technology that can be mass produced. ’We feel that thin film technology, playing with the microstructure, getting the film structure just right, gives this technology stability’, Lad says.
To date, researchers at LASST have succeeded in demonstrating a stable and sensitive system for detecting nitric oxide. At the heart of the system is a microelectronic chip on a sapphire crystal about the size of a 10p coin (see Fig). Equipped with platinum electrodes and a heating element and layered with a thin film (typically 50nm) of tungsten oxide (WO3), the chip operates like a miniature hot plate. Operating at 300 to 500?C, it burns gases on the WO3 sensor surface and indicates the presence of a specific target gas by monitoring a change in the film’s electrical properties.
Chemiresistive gas sensor for NO detection
Precise control over the thickness and composition of the WO3 film is the key to developing an effective sensor platform. Lad and his colleagues have experimented with procedures to achieve a reproducible film microstructure. Among the necessary steps is an annealing process that is carried out after the film is deposited. To be effective, the annealing is done in a specific oxygen-argon atmosphere. The result is a microstructure that is stable through repeated heating and cooling cycles.
Subsequent work focused on exploring phase fields and structure as a function of film deposition parameters. LASST researchers have investigated a variety of thin film recipes using pure WO3 with oxides of titanium, copper, tin, gallium, silicon and indium as doping agents.
’In terms of NO in human breath, we’re very close to having something that works quite well’, says Lad. ’Now if you want to couple it with another sensor that provides additional health signatures in breath, such as a sensor for isoprene or ketones, to create an even more useful diagnostic, then that adds complexity.’
The selectivity hurdle
One of the remaining hurdles, and possibly the highest, is the issue of selectivity. ’If you can make these little metal oxide sensors selective, you’ve got a home run. They’re inexpensive, and they’re highly sensitive. The problem is that they’re sensitive to everything’, says Freeman.
LASST researchers are addressing that problem by continuing to modify the sensor surface and filtering the breath sample before it gets to the sensor. Compounds that might confuse the sensor are removed, and since NO does not react well with the WO3 film, the gas is changed to another gas, nitrous oxide (NO2), before it hits the sensor element. Ideally, the whole system could be miniaturised and placed directly on the chip, an idea that LASST researchers are currently pursuing.
For a first generation of devices based on thin film technology, Lad estimates that a commercial product is still several years away. LASST focuses on the underlying science, he emphasises, and not product development.
’Eventually you could think of this product as a wafer that has thousands of individual sensors. The military is talking about having multi-sensor modules in which you can power up the ones that you need. You could mass produce them on a single sensor chip - a temperature sensor, magnetic field sensor, a chemical sensor, a biosensor. If you can fabricate it in a clean-room with micro-fabrication techniques, you can make wafers cheaply. You have the flexibility to power only the ones you want for a specific application. You can have a device that can be customised’, Lad says.
In addition to the standard tools of MS and thin films, researchers elsewhere are using chemiluminescence, in vitro cell systems and other techniques to answer a variety of questions about chemicals in breath. They have found, for example, that most NO in human breath tends to come from the nose rather than from deep in the respiratory tract. Acetone and isoprene, two volatile organic compounds, have been found in high enough concentrations in breath to pose a concern for indoor air quality under densely crowded conditions. Based on proton transfer MS, breath concentrations of isoprene have been found to correlate with cholesterol and LDL in blood, thus providing a non-invasive monitoring technique for lipid lowering therapies.
Ultimately, responding effectively to Darpa’s need for a breath sensor that can effectively prioritise medical needs will require consideration of a host of complexities. ’Baseline NO levels are very reproducible for a given person, but variations in the NO level between different people make it difficult to diagnose abnormalities conclusively with a single measurement. Our measurements have shown that either a history of a person’s normal NO level or additional markers must be detected’, says Lad.
The thin film metal oxide technology also has promise for other military applications. ’If you go into a more complicated business such as chemical warfare agent detection, you have an environment that may have very small amounts of things like nerve agents, Sarin and the like, but you also have other things like diesel fuel exhaust, burning tyres, cigarette smoke. For that, you need a much bigger sensor system that has many more sophisticated arrays.’
Rapid response may be the greatest benefit of the developing sensor technology. ’Many of the applications for which there are existing technologies are not real time’, adds Freeman. ’You go out, take samples and process them. It’s wet chemistry. It’s multiple processes, and three hours later or three days later, you get the highly accurate result of what happened three days ago. That’s hardly acceptable. They [the funding agencies] are looking for accurate real time sensors. It’s a pipe dream in many cases, but there is real promise with this technology’.
Source: Chemistry in Britain
- C. M. Harris, Anal. Chem., 2001, 73, 658A.
- J. E. Szulejko and T. Solouki, Anal. Chem., 2002, 74, 3434.
- J. E. Szulejko et al in Proceedings of the 50th ASMS conference on mass spectrometry and allied topics. Clinical Chemistry, 233, 2002.
- T. Solouki and J. E. Szulejko in Proceedings of the 49th ASMS conference on mass spectrometry and allied topics. WPM267, 2001.
- T. Solouki and J. E. Szulejko in Proceedings of the 50th ASMS conference on mass spectrometry and allied topics. Bioinformatics, 37, 2002.
- J. E. Szulejko and T. Solouki in Proceedings of the 49th ASMS conference on mass spectrometry and allied topics. WPQ331, 2001.
- S. C. Moulzolf, S. A. Ding and R. J. Lad, Sensors and Actuators B, 2001, 77, 375.
- B. Fruhberger et al, Sensors and Actuators B, 2001, 77, 226.
Contact and Further Information
Senior News Writer
University of Maine department of public affairs