Sniffing out tuberculosis in human breath

In the remote areas of Yogyakarta, on the island of Java, tuberculosis (TB) runs rampant. Access to healthcare is scarce, and even when a clinic is available, doctors may be absent. As a result, individuals with the disease are often unnoticed, leading to a high transmission and mortality rate.

TB is an infectious disease caused by Mycobacterium tuberculosis bacteria. It spreads from one person to another through the air. Once airborne, the pathogens can linger for hours. The bacteria infect the lungs, but they can also spread to other organs, such as the kidneys (1). The World Health Organization recommends that health providers in high-risk areas screen for TB through chest X-rays, rapid molecular tests, or symptom analyses. However, these options are not always accessible to these populations.

“It's very difficult for those who live in the remote areas to reach the health center,” said Antonia Saktiawati, a clinical scientist at Gadjah Mada University in Indonesia. People often visit the clinic only when they are in the late stage of the disease, where their lung tissues have become damaged, and the bacteria have accumulated in the large airways (2). At that point, recovery is usually difficult, and chances are, they’ve passed on the disease to others around them. What is critically needed, “is active case finding, not passive,” she said. Instead of relying on patients to seek care, Saktiawati urges healthcare providers to proactively identify individuals at risk of TB. And for that to be feasible, they need a portable instrument to diagnose them, she said.

Tuberculosis in exhaled breath

The World Health Organization estimates that about a quarter of the global population has been affected by TB. Five to 10 percent of this population are active TB patients, meaning they acquire symptoms and develop the disease. The rest of the infected population harbors what is considered inactive or latent TB, in which the pathogens live inside their bodies without causing any symptoms. These asymptomatic individuals are usually not contagious. However, if left untreated, inactive or latent TB infection can become active.

In its most severe cases, TB can be life-threatening, but the condition is preventable. “TB is a curable disease if we can diagnose it quickly,” said Jane Hill, a chemical and biological engineer at the University of British Columbia.

Hill’s work focuses on studying volatile molecules in breath as they relate to various diseases. She leads an ongoing initiative called the Human Breath Atlas that maps molecules in human breath to determine health status.

A research technician in pink personal protective equipment and mask is collecting breath samples of a person using a plastic collection bag outside a building. — A researcher collects a breath sample from a local resident in Yogyakarta, Indonesia.
Credit: C. Maharani / University of Gadjah Mada

“When you're breathing in and out and your blood is pumping, you're sampling the full body,” Hill said.

A person’s breath can reveal the interaction between the Mycobacteria and their microbiome, which reflects the person’s immune response, Hill explained (3). “What we're looking at in breaths for a bacterial infection is a combination of these metabolites, [which is] a direct conduit to the chemistries that are going on [inside the body],” she said.

Compared to blood, urine, or sputum, breath is relatively easy to sample, making it an ideal option for those who have difficulty providing these more invasive kinds of samples, Hill said.

Renato Zenobi, an analytical chemist at ETH Zurich, agreed with Hill. Zenobi has dedicated his career to developing mass spectrometer methods for breath analyses.

So, if a person’s breath can reveal so much information about their health, why haven’t TB clinics taken advantage of it?

“The technology is really young,” Zenobi said. “[The] diagnostic industry is hanging on to dear life to their existing technologies.” But with the rise of new diagnostic tools, Zenobi, Hill and many scientists are turning to breath as an avenue for routine disease diagnosis (4).

The birth of eNose-TB

For Saktiawati, the urgency of the TB crisis in her hometown motivated her to develop a portable screening device that healthcare workers can easily deploy in remote areas.

In the early 2010s, Saktiawati began exploring the use of an electronic nose (eNose), a device with sensors that can pick up breath signals in the form of volatile organic compounds (VOCs).

An eNose, as the name suggests, works like a biological nose, but in place of an olfactory receptor, it contains an array of chemical sensors that distinguish various volatile compounds in the air (5). The average eNose is about the size of a shoebox. “It's portable. It’s cheap, and you can use it for door-to-door screening,” Saktiawati said.

From 2013 to 2015, Saktiawati, who obtained her PhD at the University of Groningen, Netherlands, pilot-tested an eNose device built by a Netherlands-based company, to screen for TB in Yogyakarta (6).

Saktiawati and her colleagues recruited 360 patients with suspected TB who underwent clinical tests including a sputum smear, chest X-ray, and symptoms screen. To collect the breath samples, the patients breathed into a mouthpiece connected to the eNose device. The breath signals detected by the device sensors generated data that was analyzed using an artificial neural network specifically trained to recognize a VOC pattern produced by M. tuberculosis or the patient’s metabolism.

The researchers found that the sensitivity of the device at 85 percent was comparable to the gold standard sensitivity of TB examination, but its specificity fell short of the 86 percent benchmark (7).

Determined to improve the system, in 2018 Saktiawati and her team built an in-house eNose device in Yogyakarta, dubbed the eNose-TB. She estimated that the tests would cost less than $2 USD per sample. The device can accommodate offline sampling to allow the patients to breathe into a disposable bag rather than directly into the instrument.

A portrait of Antonia Saktiawati, a female clinician, in a white blouse and black blazer. — Antonia Saktiawati is a TB expert at the University of Gadjah Mada who developed a portable eNose device to detect TB in the remote areas of Yogyakarta, Indonesia.
Credit: Antonia Saktiawati / University of Gadjah Mada

This feature came in handy in 2020, when the COVID-19 outbreak began. During the pandemic, the scientists repurposed the eNose device for COVID-19 testing (8). They distributed as many as 4,820 devices across Indonesia within months, Saktiawati said. “Our TB research was very delayed for two years, but it's still going.”

In 2021, Saktiawati published the initial results of the eNose-TB testing involving 27 TB patients and 24 healthy controls (9). To her delight, the device reached 95 percent sensitivity and 82 percent specificity. While the findings were promising, the sample size was small, and there were plenty of logistical issues to improve, Saktiawati said. She found that individual differences from people’s diet, sex, race, and other health conditions could vary the results.

Because the method depends on a breath pattern instead of a specific disease marker, the researchers will need a large dataset to train the machine learning algorithm to pick up those variations. Using more sophisticated techniques such as gas chromatography-mass spectrometry (GCMS), however, can address this limitation, Saktiawati said. This approach combines two instruments designed to separate volatile molecules in a sample and measure their masses.

If Saktiawati and her team can identify specific TB biomarkers, they can make the eNose smarter, she said. In 2022, the team began to recruit more patients and expand the study to other provinces in Indonesia. They plan to launch a biomarker study using GCMS this year.

A different approach

Like Saktiawati, George Chimowa, a nanotechnologist at Botswana International University of Science and Technology, was also motivated by the TB crisis in his country. Tucked in the southern African desert, Botswana is one of the countries with the highest global tuberculosis burden.

“In southern Africa, we have informal settlements, such that the houses are so close together. And TB, being a very highly infectious disease, quickly spreads,” he said.

Instead of collecting breath samples and analyzing their pattern, Chimowa used a bottom-up approach to pinpoint the molecular signals unique to TB.

In a recent study, Chimowa and his team identified eight potential markers specific to patients with tuberculosis using GCMS (10). The researchers derived these compounds from the breath of 41 TB patients across Botswana. Among the eight unique markers they identified, six were previously reported in the literature as being associated with TB. They identified o-cymene as a novel biomarker of TB, and they found that benzyl 1-methyl 4-(1-propynyl) could distinguish between susceptible and multidrug resistant TB cases. The researchers found that 4-methyloctane was also present in individuals with other pulmonary diseases, such as lung cancer and chronic obstructive pulmonary diseases.

Lung diseases tend to have two or three overlapping biomarkers, Chimowa said. “So, the idea is to have a profile of many VOCs to eliminate the chances of false positives here.”

While GCMS is a very accurate technique to determine VOCs, Chimowa said that he hopes to stop using it. The instrument can cost up to $100,000 USD and is about the size of a dishwasher. “Our desire is really to come up with a low cost and very fast technique for diagnosing diseases,” Chimowa said. His team is now developing VOC sensors for an eNose based on the biomarkers they discovered.

Beyond developing nations

The effects of TB reach beyond Indonesia and Africa. “TB is not just a disease of the developing world, it's affecting the US and other highly developed nations,” said Michael McLoughlin, the Vice President of Zeteo Tech, a medical device company in Maryland that develops a cost-effective mass spectrometry instrument to analyze breath for infectious disease and biodefense purposes.

TB is not just a disease of the developing world, it's affecting the US and other highly developed nations.
- Michael McLoughlin, Zeteo Tech

As of March 2025, 67 individuals have become ill from the outbreak that began in Kansas last year. “For TB, you really want to be able to find the spreaders to cut down the transmission,” McLoughlin said.

Scientists and engineers at Zeteo Tech are developing a portable Matrix-Assisted Laser Desorption/Ionization (MALDI) mass spectrometer, called BreathBiomics. The instrument is about the size of a briefcase, which is much smaller than a conventional MALDI instrument found in a research lab, which is typically about the size of a large fridge. McLoughlin and his team demonstrated their instruments’ ability to detect up to 80 proteases linked to lower respiratory tract infections (11). Among these proteases, three were specific to TB.

Beyond VOCs, the company is exploring other components of exhaled breath, such as droplets or aerosols that contain neutrophil markers specific to the immune responses of people with TB and other lung diseases (12). They are hoping to secure more funding to conduct further tests.

Breath analyses for TB and beyond

For Zenobi, the key to diagnosing and controlling a treatable infection like TB is in developing an affordable and accessible diagnostic instrument. “Everybody can give a breath sample. It is rather universal, instantaneous, and it's painless,” he said.

Beyond the clinic, Zenobi sees the potential of breath tests as a strategy to expand disease screening into the public spaces. “One could place an instrument at an airport, at a shopping mall, at a train station, and make sure that people are not infectious,” he said. The holy grail would be to breathe into an instrument and find out in a matter of seconds what a person’s health condition is. “That's my big fantasy there,” he said.

References