TB treatment: faithful PET?

6 March 2015



Medical imaging techniques could be used to predict the effectiveness of antibiotic drug regimens being tested to treat tuberculosis patients, according to researchers at the National Institute of Allergy and Infectious Diseases. Dr Clifton Barry, chief of the Tuberculosis Research Section, explains to Ross Davies how PET and CT scans could play a vital role in treating the growing number of multidrug-resistant cases of the disease.


The public health threat posed by tuberculosis (TB) remains undisputed. After HIV/AIDS, it is the biggest killer worldwide. Of the nine million people that fell ill with the disease in 2013, 1.5 million of them died - according to the World Health Organization's (WHO) 2014 global tuberculosis report.

It has been more than 21 years since WHO declared TB to be an international health emergency. Speaking in April 1993, Arata Kochi, the now former manager of the organisation's TB programme, decreed the disease to be "humanity's greatest killer".

Of course, the past two decades have been marked by considerable progress in the treatment and control of TB; the death rate has dropped by 45% since 1990, and initiatives such as World TB Day, held annually on 24 March, have also done much to raise awareness.

The statistics contained in the 2014 report still make for grave reading, however. In an editorial published in January, the peer-reviewed general medical journal The Lancet condemned the current death rate from TB as "unacceptable in the 21st century".

Another cause for concern is the leap in cases of multidrug-resistant TB (MDR-TB), which totalled 480,000 worldwide in 2013. Caused by improper administration of treatment regimens, MDR-TB has become particularly problematic in TB hotspots such as sub-Saharan Africa, Asia and eastern Europe, where drug-resistance testing and treatment services are generally found wanting.

Worse still, 9% of people with MDR-TB are said to have extensively drug-resistant tuberculosis (XDR-TB); in other words, there are roughly 50,000 people in the world with a form of TB that remains completely untreatable.

The effectiveness of antibiotic drug regimens in the treatment of TB is a focal area of research at the National Institute of Allergy and Infectious Diseases (NIAID), which operates under the umbrella of the US National Institutes of Health. In December, researchers from NIAID revealed research suggesting that medical imaging techniques could be used to reliably predict treatment outcomes for TB patients.

Portrait of a killer

The claim is based on the idea of using a combination of positron emission tomography (PET) and computed tomography (CT) - more commonly used in the detection of cancer - to create new biomarkers of the disease, which are essential in ascertaining whether a particular drug regimen is effective or not.

In light of the prevalence of MDR-TB and XDR-TB, NIAID's findings could herald a considerable stride in care and control of the disease. Across two separate studies - on TB-infected monkeys and TB-infected humans respectively - it was found that early changes in PET/CT scans carried out two months into the course of treatment could accurately predict patient outcomes at 30 months.

According to Dr Clifton Barry, chief of NIAID's Tuberculosis Research Section (TRS), the institute's interest in the potential of PET/CT technology dates back approximately a decade and was prompted by the successful adoption of such imaging techniques in the field of oncology.

"PET/CT first came to our attention in 2004," he explains. "I had watched with interest how PET scanning in particular had changed oncology drug development and trials, and I remember thinking that, in terms of using imaging technology, we were really behind the times with infectious diseases. So it caught my eye as being something quantitative that could be assessed independently in many different areas, using defined criteria."

The TRS's first study into PET/CT scanning soon followed, in which researchers monitored macaques treated with two different antibiotics and observed changes in scans, as well as determining the amount of bacteria remaining in the animals' lungs post-treatment, the standard method for assessing TB therapy effectiveness.

The same approach was then applied to TB-infected people taking one of the antibiotics, with results indicating that changes in human PET/CT scans during treatment were comparable with those seen in the monkeys.

"In terms of using imaging technology, we were really behind the times with infectious diseases. So it caught my eye as being something quantitative that could be assessed independently."

"Contrary to popular belief, it was actually much easier to do PET/CT scanning on people as opposed to the macaques because the machines already existed," says Barry, looking back on NIAID's first group of tests.

"We worked with the likes of Siemens to engineer a PET/CT system able to accommodate animals and non-human primates. Investing in that hardware development meant it took several years before we could actually do any scanning. It took a lot of engineering."

Cultural change

Up until now, sputum cultures have been the most common means of predicting treatment response in TB patients. However, it is widely accepted that sputum studies are limited in their ability to foretell drug efficacy as well as disease relapse. Barry is unequivocal when it comes to the advantages of PET/CT technology over traditional microbiological methods - even if it is not exactly cheap.

"At present, our end points in clinical trials for TB are all based on microbiology," he says. "What we are counting on quantifying is essentially what the patient coughs up in their sputum. That's a pretty variable measure to follow and doesn't really correlate well with how the patient is responding to chemotherapy.

"The criticism people often voice about PET/CT scanning is that it is expensive. True, it is much more expensive than microbiology, but it's also much more powerful," explains Barry. "The quantitative aspects of it really allow us to talk about sample sizes that are, in order of magnitude, lower than the sample sizes you need with microbiology."

On its website, the TRS is described as being "highly interactive worldwide". Would it be fair to assume that the promise of the imaging techniques used by Barry and his team is already common knowledge among the wider community involved in researching TB?

"It is now, for sure," he says. "We are already working with groups in South Africa, China and Brazil, and also in other parts of the world, to implement our findings in the context of phase-II and phase-III studies.

"We started talking about this before much had been published, so as to try to generate interest. Despite the first reaction being, 'Oh, that's too expensive,' the response has been very positive, and people have come around to seeing the value of having quantitative radiographic end points in trials."

TB continued

Just as cancer drug treatment provided the initial touchstone for NIAID's experimentation with PET/CT, the ongoing fight against HIV - which promulgates a message of treatment as prevention - also serves as a source of inspiration to Barry as he looks ahead to further stages of study. As it happens, the risk of contracting TB is estimated to be 26-31 times higher for people living with HIV, according to WHO.

"I've watched the advances in HIV drug development with a lot of interest," says Barry. "The HIV clinical trials field has been well and truly transformed through the ability to take the quantity of viral load assays, and measure the host and the amount of pathogen.

"Up until now, TB has really had neither of those two things. I'm hopeful that with these bacteria-specific probes, we will have complementary assays that pair with viral loads, and we will be in a position to be able to emulate the HIV model for finding the best combinations and best durations of treatment, and really move things forward."

Undoubtedly, much more legwork will need to take place if we are to see the use of PET/CT scans in trials of investigational drugs to predict the effectiveness of a treatment regimen in its early stages.

While NIAID findings suggest that the use of such technology could save resources and shorten the duration of clinical trials, questions over cost are yet to fully dissipate, as hinted at by Barry.

The Global TB Strategy, as laid out by WHO, aims to reduce TB deaths by 95% between 2015 and 2035. It's an ambitious target, especially when taking into account the three million people diagnosed with TB each year - classified as the 'missing three million' - that continue to miss out on care.

The development of guaranteed drugs and treatment regimens will play a considerable role in the fight against TB, which Barry pointedly defines as "the world's most neglected disease".

If PET/CT imaging is proven to correlate with treatment outcomes in patients, and particularly those with MDR-TB, WHO's goal to effectively rid the world of TB-related deaths over the course of the next 20 years might not seem so lofty after all.

Preliminary research evidence has found that changes in CT scans taken over the course of treatment are correlated with treatment outcomes.


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.