Six Striking Findings from Major TB Conference

Mycobacterium tuberculosis drug susceptibility test. Photo by CDC on Unsplash

News & Features

2nd December 2025 | Elri Voigt

From studies of new medicines and a mask used to diagnose TB, there was no shortage of interesting findings presented at the recent Union World Conference on Lung Health, held in Copenhagen, Denmark. Spotlight rounds up six studies that stood out.

1. People do better if we dispense all TB prevention pills at once

One of the most important questions in TB is how to best prevent people from getting ill if they’ve been exposed to the bug. While effective treatments to prevent TB disease in people who have been exposed exist, uptake has generally been poor.

Now, researchers have found that, dispensing all the pills in a three-month course of TB preventive Therapy (TPT) at once, instead of asking people to collect pills at the clinic every few weeks, led to many more people completing the treatment course.

The study, called ThiPhiSA, was conducted in four clinics and communities in KwaZulu-Natal. 268 households who qualified to receive TPT were enrolled in the trial. 301 participants from these households were randomised to one of two arms, explained Dr Adrienne Shapiro, Assistant Professor of Global Health and Infectious Diseases at the University of Washington.

In the first arm, 159 people were given a two-week supply of the standard of care 3HP (consisting of the drugs isoniazid and rifapentine, taken once a week for 3 months). They then had to go to clinic to receive their refills as per the clinic schedule. They received weekly sms reminders to take their pills and were visited by researchers at month one and two and at the end of the study.

In the second arm, 142 people were given all the pills for the full three-month course of 3HP at once. While no clinic visits were required, they were remotely registered at their clinics just in case they had to visit the clinic. This group also got weekly sms reminders to take their pills and were visited at month one and two and at the end of the study.

Whether people had taken their pills was assessed through self-reporting, as well as a calendar dosing diary, pill count and assessment of urine colour change if the visit was on a day when the participant had recently taken a dose of 3HP.

For those who had to go to the clinic at regular intervals to collect their pills, only 28% completed their treatment course. By contrast, 86% of those who had the full course dispensed at once completed their treatment.

Much of the difference was due to the fact that some people in the prior group simply did not go to the clinic every time to collect pills. There was also a drug stockout at one of the clinics – which somewhat skewed the results in favour of the latter group, but not enough so to change the fact that people were more likely to complete treatment if they got all the pills at once.

Dispensing a full course of TPT at once was safe, according to Shapiro, with no serious adverse events seen in the study.

“Multi-month delivery of TPT is safe, and person-friendly approaches improving the convenience of TPT should be adopted to decompress health facilities and improve TPT coverage to meet TB prevention goals,” she concluded.

2. A new medicine might help shorten TB treatment

Much TB research in recent years have focussed on reducing the duration of TB treatment – it typically takes six months. In addition, researchers have also been looking for medicines that have fewer side effects. Growing resistance to some existing TB drugs is also a concern.

One of the big talking points at this year’s conference was data on an experimental new drug called sorfequiline. It is thought that sorfequiline could be a replacement for bedaquiline, arguably the most important TB drug developed in recent decades. This is because sorfequiline appears to be more potent than bedaquiline and because of worries over TB strains that are resistant to bedaquiline.

The new data is from a phase 2 trial of sorfequiline used in combination with two other TB drugs – pretomanid and linezolid – to treat drug susceptible TB. The regimen is called SpaL for short. 309 participants with newly diagnosed TB were either given the standard of care first-line drugs isoniazid, rifampicin, pyrazinamide, and ethambutol (HRZE) for 26 weeks, the BPaL regimen consisting of the drugs bedaquiline, pretomanid and linezolid for 26 weeks (not all drugs in these first two arms are taken for the entire period), or one of three different doses – 25mg, 50mg or 100mg – of sorfequiline along with linezolid and pretomanid for 8 weeks.

Once those in the sorfequiline arms completed the initial 8-week course, they had to take the drugs isoniazid and rifampicin (HR) for another 7 weeks and were then tested for TB again. Meaning at this point they had gotten treatment for 15 weeks (or around 3 and a half months). If they tested negative and had no TB symptoms, they could stop treatment. But if they tested positive for TB and had symptoms then they’d have to continue taking isoniazid and rifampicin.

Among the participants who got sorfequiline, 64% in the 100mg arm were able to stop treatment after 15 weeks, compared to 46% in the 50mg arm, and 28% in the 25mg arm.

Study participants gave regular sputum samples that were tested for the presence of TB bacteria. The researchers then estimated the probability of a “stable sputum culture conversion at week 8”. In simple terms, this means the researchers wanted to find out what the probability is that all the TB bacteria had been killed by the different regimens after 8 weeks of treatment.

For the 25mg arm, there was 31% stable culture conversion, in the 50mg arm 48%, and in the 100mg arm 59%. For both HRZE and BPaL it was 45%. In other words, 100mg of sorfequiline plus pretomanid and linezolid showed better stable sputum culture conversion after eight weeks than HRZE, the regimen currently used to treat drug-susceptible TB in South Africa and most other countries.

“[W]e believe that SPaL is a promising four-month regimen,” said Dr Morounfolu Olugbosi, the medical lead for the study, which is being conducted by the non-profit TB Alliance.

The regimen was well tolerated at all dose levels, according to Olugbosi, with no difference in safety signals in the sorfequiline arm compared to the other treatment arms. “So that lack of observable difference is what we consider positive news,” he said.

While this trial looked at people with drug susceptible TB, the research team will be investigating how well it works for drug resistant TB in an upcoming phase 3 study, said Dr Maria Beumont, the Chief Medical Officer at TB Alliance, during a press conference.

Indeed, while these phase 2 results are promising, the real test for this drug will be in the larger phase 3 study to come.

3. Co-morbidities are really important when people have TB

A prospective cohort study in South Africa looked at the burden of co-morbidities and the impact it has on TB mortality. The researchers followed around 2 000 adults with pulmonary TB, diagnosed with Gene Xpert (a molecular TB test), and looked at mortality rates for 1 896 of those people after 15 months. Of those, 272 people (14.3%) had passed away during the study duration.

According to the study presenter, Dr Greta Wood, a Clinical Research Fellow in Infectious Diseases at the University of Liverpool, the prevalence of TB multimorbidity among the whole study group was 86%. Meaning most had TB as well as another illness or risk factor.

The researchers looked at five key co-morbidities identified by the World Health Organization (WHO) – HIV, smoking, undernutrition, diabetes and alcohol use. “These five key comorbidities alone explained over half of the mortality that we saw in this cohort,” Wood said.

The researchers found that the more co-morbidities a person had, the higher their risk of dying was when they got ill with TB. The risk of dying for someone with TB was 19% if they had three or more co-morbidities compared to 16% if they had two, and 11% if they had no co-morbidities.

The key conditions driving mortality in this group of people with TB is HIV and undernutrition. Undernutrition in particular was flagged in the study, as in this setting it was responsible for around one in five TB deaths in people under the age of 40, according to Wood.

“[I]n this cohort, we didn’t find an association between diabetes, smoking, alcohol and mortality, but that has been demonstrated in other settings,” she added.

This data should lead to urgent action, concluded Wood, saying that “to reduce the risk of death, we need to urgently start operationalising screening for these key five TB comorbidities and linking people into treatment”.

4. Point-of-care testing leads to people starting treatment faster

As shown in several studies at this year’s conference, the details of how TB services are delivered can make a significant difference to TB outcomes.

One such study, led by researchers from the University of Cape Town (UCT), explored whether it made a difference if someone had a TB test done at a mobile van, or had their sputum sample collected and sent off to a lab. In other words, the study was indirectly testing whether it makes a difference if someone gets a test result right after testing, versus having to wait a day or two to be contacted with a result.

The study, of over 7 000 people, was conducted in South Africa, Zambia, Zimbabwe, and Mozambique. Around half of those screened in the study were at high risk for TB and randomised to either receive point of care testing on a GeneXpert machine in a mobile van or centralised GeneXpert testing at a laboratory.

In the point of care arm, results were available for 1 641 people, and of those 55 (3.3%) had microbiologically confirmed TB. While 67 (4.1%) of the 1 632 tested in the centralised testing arm had microbiologically confirmed TB. Overall, across both arms, 93 of the people diagnosed with TB (76%) were successfully linked to care.

“When compared to those who had their Xpert performed at a central laboratory, those who had their Xpert done at point of care had a 43% lower probability of treatment initiation failure and initiated treatment twice as fast,” said Tahlia Perumal, a researcher at UCT, who presented the results. Participants in the point of care arm on average started treatment four days sooner than those whose TB tests were done in a centralised laboratory.

“[T]here is an argument to be made about the clinical significance of a four-day reduction. We are in the process of doing transmission modelling to be able to provide more granular details about the difference this may make in active case finding models in larger population sizes,” she said.

5. Promising signs for a portable TB test

In the above study, point-of-care testing was done in a relatively large machine in a mobile van. We may however be able to go much smaller.

Tessa Mochizuki, a Research Scientist at University of California in San Francisco, presented results from a multi-country study evaluating how accurate a portable, battery-operated testing device, called MiniDock MTB, was at diagnosing TB from sputum swabs and tongue swabs.

“The test is run on a small device about the size of my hand, and results are available in under 30 minutes, often even faster for positive results,” Mochizuki said.

1 380 people, aged 12 years and older with presumptive TB were enrolled across seven countries – South Africa, India, Nigeria, the Philippines, Uganda, Vietnam, and Zambia. Sputum samples from all participants were tested using two MIDGIT cultures (a test in which the bug will grow if present), smear microscopy (where the bug is looked for under a microscope), and GeneXpert Ultra (a molecular test).

Results from these tests were then compared with results from the MiniDock MTB machine.

For the tongue swab test, a healthcare worker runs a swab over the participants tongue for 30 seconds, then it is put into a buffer liquid, mounted on a testcard which is run through the portable machine. For the sputum swab, a swab is dipped into a test tube that contains sputum for about 15 seconds, put into a buffer liquid and mounted onto a testcard and run through the portable machine.

When comparing sputum swabs results to the Xpert Ultra results there was not a statistically significant difference, according to Mochizuki, as sputum swabs showed 87% sensitivity compared to GeneXpert Ultra’s 89%. Sensitivity is a measure of how likely the test is to detect a bug if the bug is present in the sample.

Tongue swabs performed a bit worse with 81% sensitivity. This was however much better than the 62% with microscopy. Microscopy is rarely used for TB diagnosis in South Africa, but this finding could be important in other countries where health systems haven’t switched as fully over to molecular testing as we’ve done here.

All sample types and tests achieved 98% specificity. Specificity is an indication of how likely a test is to give a negative result if the bug being tested for is not present in the sample.

These findings meet the WHO target product profile requirements – a minimum of 85% sensitivity and 98% specificity for sputum tests and a minimum of 75% sensitivity and 98% specificity for non-sputum tests.

“We submitted this data to the WHO guideline development group that convened last week, and we look forward to news on any official recommendations in the coming year,” Mochizuki said. “These results show that we can achieve high accuracy with a low-complexity platform, bringing molecular testing closer to people seeking care without sacrificing performance.”

6. A mask that can help diagnose TB

An even more interesting idea that some researchers have been working on is to use a diagnostic mask to diagnose TB.

At this year’s conference, Dr Rouxjeane Venter, a researcher based at the Clinical Mycobacteriology and Epidemiology (CLIME) research group at Stellenbosch University, presented a proof-of-concept study testing whether a mask, called the Avelo Mask, can be used to diagnose whether TB bacteria is present in the air a person breathes out.

58 adults, across four clinics in Cape Town, who had TB symptoms and tested positive for TB on a molecular test were given the mask to wear for 45 minutes. The filter in the mask is able to trap tiny particles from .3 micrometers and above – meaning it can trap viruses and bacteria. This filter is then pushed into a buffer tube using a sample stick – where it can be stored or tested directly. The mask as well as the stick and buffer tube are part of the Avelo mask kit developed by Avelo Diagnostics. For this study, the researchers used a qPCR test – a rapid test that looks for TB DNA – to detect TB bacteria.

When the mask filters were tested, 34 people were found to be negative for TB bacteria and 24 were positive. When compared to their Xpert Ultra sputum results, it was found that there were two false positives.

Overall, according to Venter, the mask had a sensitivity of 71% when compared to GeneXpert Ultra and 65% when compared to the Microbiological Reference Standard and a specificity of about 92%.

People with higher bacillary loads – meaning lots of bacteria – in their sputum were more likely to be positive, but there was still a large percentage of participants with low or very low bacillary loads that were picked up by the mask.

These numbers aren’t nearly as good as those for the MiniDock MTB, but it is positive that masks like these are showing promise. A long-standing problem in TB diagnosis is that not everyone can produce sputum samples. The more alternatives we have, be it tongue swabs or masks, the better.

Republished from Spotlight under a Creative Commons licence.

Read the original article