Tuberculosis Through the Looking Glass

A prismatic view of novel research to combat an ages-old disease

Published on December 9, 2015by Emily Augustini and Prativa Baral

White plague. Phthisis. Consumption. Depicted in early Egyptian art and referenced in the Bible, tuberculosis (TB) has been called many different names since it began infecting humans long before recorded history. For centuries, its cause was elusive. Ever since the 1882 discovery of Mycobacterium tuberculosis, the infecting agent, we have been searching for new ways to track the bacterium and minimize its effects. In the mid-twentieth century, antibiotics seemed to promise an end to the disease, but the microbe has become stubbornly drug resistant in the twenty-first century.

One third of the world’s population is infected with TB. In 2014 alone, 9.6 million people fell ill with the disease. Across the globe, TB kills over 1.5 million people each year. Cases of multi-drug resistant (MDR) and extreme drug resistant (XDR) TB are rising at an alarming rate. Moreover, TB and HIV have developed into a co-epidemic, with 1.2 million people infected by both the TB bacterium and HIV virus worldwide. Of those living with both diseases, one in three people are never notified that they have TB and do not receive treatment for it.

Now, the scientific community is more focused than ever on finding innovative methods to combat this chronic infectious disease.

Together, a passionate, dedicated core of faculty in Columbia’s Department of Epidemiology are imagining new approaches to TB that span the spectrum of the challenges it presents. Their innovative research ranges from tracking the molecular messages that TB sends to better detect the disease, and telling the tale of transmission dynamics through the tubercule itself, to a more humane way of delivering what historically has been one of the more punitive population-level treatment protocols.


The majority of people who are infected with TB do not have the disease—the bacterium remains dormant in 90-95 percent of people, kept in check by their immune systems.

Latent TB can turn into active disease years after a person is infected, which obscures transmission. From a clinical perspective, the “when,” “where” and “how” don’t necessarily matter: regardless of the means by which an individual becomes infected, treatment is much the same. Those seeking to control the spread of this infectious disease, however, must concern themselves with such details.

“People lie but bugs don’t”

From an epidemiological perspective, an epidemic demands a systemic—and systematic— approach: households, individuals and their respective contacts all need to be investigated, which requires tracing and verification. But memories are unreliable at times—especially given that infection may have occurred years ago and doesn’t even require direct physical contact (the bacterium that causes TB is airborne). What’s more, individuals may not want to share information because of stigma or social constraints. All of these factors hinder TB control because they make it difficult to paint a clear picture of how the bacterium is being spread using traditional epidemiological methods.

A pathogen-centric approach can lead to a much deeper understanding of TB transmission because while, with respect to their exposure, people might forget, or may even be compelled to lie, bugs simply don’t.

From microorganism to macro movements

Historically, the pathogen responsible for TB, the Mycobacterium tuberculosis bacterium, has been observed superficially: we started out by recognizing its shape, texture, and growth characteristics. In other words, basic microbiology relied on the phenotype of the organism for classification purposes. Modern methods are more granular, relying on genotyping, a laboratory-based method that analyzes and recognizes the genetic configuration of the organism, to detect outbreaks earlier and recognize unsuspected relationships.

Dr. Barun Mathema is an assistant professor of epidemiology at Columbia University’s Mailman School of Public Health. He uses a combination of molecular genotyping and epidemiological methods, such as surveillance, spatial analysis, and network analysis in order to understand the spread of TB epidemics. He says TB is fascinating because it has endured years of insults by the host’s immune system and constant attacks by a slew of antibiotics.

Understanding the pathogen through molecular genetics and mechanisms of characterization is essential to painting a bigger picture of TB transmission and consequentially, global epidemics.

“The bug innately wants to survive—it’s a very well-adapted human pathogen, a product of hundreds of thousands of years of evolution, and fine-tuning. And in that process, it adapts, spreads and mutates,” says Dr. Mathema.

Dr. Mathema uses data to understand this evolution and to observe the natural changes that occur over time. “Over the years, the organism has gone through a Darwinian process of selection and adaptation such that the pathogen has learned to co-adapt and survive with us as a host,” he says.

Tracking the evolution of the bug is one way to control the epidemic because the specific changes and mutations that occurred during evolution can indicate possible targets that can be used to develop more effective drugs. Furthermore, analysis of the strains can demonstrate similarities and differences, and allow for early detection of mutations that may result in multi-drug-resistance.

All of this is vital to public health’s ongoing efforts to control the epidemic: the pathogen-centric lens leads to a more nuanced understanding of transmission. The pathogen doesn’t just tell us who gave TB to whom. It provides insight into specific mutations that affect the ease of transmissibility from one person to the next and drug resistance. Such examination is the key to identifying chinks on the microbe’s armor, which helps clinicians select appropriate treatment protocols and researchers to avoid wasting time on pharmaceutical dead-ends.

It should be noted that simply recognizing the pathogen is alone inadequate because transmission is complex and at times, perplexing. Distinctive strains may be observed within the same household, which suggests a non-linear method of transmission. A father, for example, may contract TB from someone at work, while his children pick up a different variant of the bacterium at school.

Tracking this non-linear transmission requires spatial and social network analysis. Understanding the pathogen through molecular genetics and mechanisms of characterization is essential to painting a bigger picture of TB transmission and consequentially, global epidemics.

Social medium

Indeed, poverty and TB are closely associated, where the poorest and most vulnerable communities are most at risk of developing the disease. In an ideal world, eradicating poverty and the discrepancies in the wider societal structures would eliminate TB. But until then, Dr. Mathema says, hotspot mapping, a visualization of high-density occurrences, is a valuable public health tool that can advance our understanding of the disproportionate prevalence and the population dynamics of TB.

South African gold miners, for instance, have an extremely high TB prevalence. We now know that this is a consequence of their exposure to silica dust in the mines. But it was through spatial analysis and genotype analysis of the strains prevalent in that population that led public health scientists to conclude that TB emerged from the gold mines. Recognizing the strain-specific catalysts of TB transmission, mapping the highly prevalent locations and finally, prioritizing interventions tailored towards these highly dense areas is key for TB control.

Merging the fields of public health and biomedicine can lead to a better understanding of the complexities of TB transmission for the prevention and control of the disease. This way, a more targeted implementation of public health strategies can occur, kick-started by a thorough understanding of the pathogen and its transmission.

According to Dr. Mathema, “Transmission is a real Achilles heel of TB—we really need to knock down the incidence. And the best way for us to do that is to understand the pathogen and then concentrate on hotspots.”

The emphasis now is on a worldwide hunt for biomarkers of TB that can be used to precisely determine disease status, and therefore improve the diagnosis and prognosis of TB patients.


For a disease as ancient as TB, we have shockingly few tools with which to fight it. Most people who have attended college have had the tuberculin skin test, which measures a patient’s antibody levels to determine if they’ve been exposed to TB bacteria. While relatively cheap and easy, the skin test has a high false positive rate. The Interferon-Gamma Release Assay, or IGRA, is more reliable but far too expensive for resource-poor settings. These are the two most common diagnostic tests for TB.

The tests available for assessing treatment outcomes are even more inadequate. The “gold standard” for drug trials is merely determining whether subjects are still infected with TB after two years.

The emphasis now is on a worldwide hunt for biomarkers of TB that can be used to precisely determine disease status, and therefore improve the diagnosis and prognosis of TB patients. Dr. Max O’Donnell, an assistant professor in the Division of Pulmonary, Allergy, and Critical Care Medicine and the Department of Epidemiology at Columbia, described his search for new methods of assessing and optimizing treatment outcomes through the use of biomarkers.

Disease signals

The term “biomarker” is reminiscent of other buzzwords like “synergy” and “empowerment”—trendy, catchall terms with many meanings. This is because “biomarkers” aren’t just one thing; they comprise a vast array of molecular signals. When a person is infected with TB, these molecular messages sound off.

According to Dr. O’Donnell, “A biomarker is a surrogate that’s close to the disease process and reflects disease activity.” These surrogates are biochemical signals that are given off by either the bacterium itself or the human host, and can be quantified in order to determine disease status and progression.

Host-derived biomarkers are often antibodies that the immune system creates to fight the infection. These antibodies are an attractive target because of our knowledge of the human immune system, but may provide an imperfect measure of disease activity. Instead of monitoring the bacteria, the results are filtered through the immune response of the host.

Dr. O’Donnell’s research focuses on biomarkers derived from the TB bacterium itself. His lab has created a virus that feeds on TB. It releases a fluorescent green protein when it finds a snack.

Researchers collect a sputum sample from the patient and add the modified virus. If the sample glows green under a fluorescent microscope, TB bacteria are present. If the sample glows in a well when antibiotics are added, the TB is drug resistant.

The virus created by Dr. O’Donnell’s team only infects TB bacteria, making it more accurate than the traditional tests, and it can detect extremely small concentrations of bacteria. When the procedure was tested in a clinic in South Africa, it was found to be more sensitive but less specific for detection of TB—meaning it was better at detecting true cases of TB than it was weeding out the non-cases—than an expensive, sophisticated PCR machine. It did not make a difference if the patient also had HIV, which is critical in the developing world, where co-infection is extremely common.

Predicting resistance

The uses for this modified virus are virtually limitless. One of Dr. O’Donnell’s priorities is examining whether treatments are likely to generate drug-resistant strains of TB.

“Almost all of our data on rates of mutation come from a petri dish. How frequently the bug mutates is based on in vitro data, and it may be that some of the second-line drugs we’re giving have mutagenic potential. Because they interfere with…enzymes that are involved in DNA replication, it’s totally plausible that the rates of mutation are quite different. We’re trying to use deep sequencing and this phage, this modified virus, to detect subpopulations that are drug resistant as they emerge,” says Dr. O’Donnell.

TB is something of a perfect storm for paternalistic public health interventions.


For individuals, a diagnosis of TB is unlike any other. Quarantine is common. Treatment is laced with mistrust, precisely because of concerns regarding transmission and treatment resistance. Often, health care workers must witness patients taking their medications, because treatment non-adherence is expected. And patients not completing their six-month regimens are labeled defaulters. Never mind the toxic side effects.

The treatment for TB is arduous, lengthy, and unaided by its impersonal, rather punitive nature.

A perfect storm for paternalism

TB is something of a perfect storm for paternalistic public health interventions. The disease is airborne, so it is spread simply by breathing; there’s no deliberate action required on the part of the exposed. There aren’t any corporate stakeholders to push back against TB control, like tobacco companies do with smoking. Those infected are usually of lower socioeconomic status and, in the United States, are often foreign-born, limiting their ability to access services and complete treatment.

The standards for TB treatments were developed originally with sound reasoning and no intent to demean patients. Directly observed therapy, or DOT, for TB treatment is not an inherently bad idea, just an outdated one. As medicine and public health become more technologically advanced and the biology of disease is better understood, the importance of humanity—human factors, humane engagement—is starting to emerge.

From punitive to patient-centered

Dr. Yael Hirsch-Moverman, an assistant professor of epidemiology at Columbia, focuses on a holistic, empowering approach to TB treatment known as patient-centered care, which is gaining traction worldwide.

According to Dr. Hirsch-Moverman, patient-centered care is “Basically the way it sounds, which is putting the patient at the center of the treatment.” In patient-centered care, providers consider the barriers that prevent patients from seeking and completing treatment, including social, educational, psychological, and structural factors.

This approach isn’t new, though. It has long roots is patient advocacy—activists for diseases like HIV/AIDS have been championing patients’ rights for years. However, patient-centered care only recently made its way into TB treatment. The WHO just adopted patient-centered care as one of the three pillars of the End TB strategy for 2015-2030.

This shifting mentality is due in part to the relative failure of current TB control methods. Though TB deaths have decreased markedly since 1990, there were still 1.5 million TB deaths worldwide in 2014. Dr. Hirsch-Moverman says, “That’s unacceptable. It’s a preventable, curable disease. We’re not doing well, and we need to fix it.”

There are myriad reasons for these gaps in TB management, including absence of funding, lack of visibility, and the latent nature of the disease. Dr. Hirsch-Moverman and many others hope that patient-centered care for TB will not only empower patients but also improve their ultimate outcomes, all while being cost effective.

Scaling up these culturally and locally specific interventions presents a grand challenge.

Packaged for patients

In a trial that Dr. Hirsch-Moverman is working on in Lesotho with the Columbia University organization ICAP, a patient-centered approach means combining multiple interventions that have been shown to be effective into a comprehensive package. This package includes reimbursement for transportation to the clinic, educational programs to improve patients’ understanding of their treatment, job aides for nurses, and an SMS reminder system for appointments.

Each of the interventions is extremely inexpensive, and together they stand to improve treatment completion and survival rates. The team is doing a similar study in Ethiopia, but with the goal of enabling HIV positive patients to complete preventive treatment regimens for TB. The packages are tailored to the region in which they’re used. For example, in Ethiopia literacy rates are lower than in Lesotho, so the SMS text reminders are replaced by automated voice reminders.

Scaling up these culturally and locally specific interventions presents a grand challenge. However, Dr. Hirsch-Moverman is confident that it can be done in a cost effective manner. “We designed these interventions to be sustainable. Not too expensive, nothing super fancy, because at the end of the day we’re going to have to depend on the local ministries of health to deliver these things,” she says.

The high level of acceptance and buy-in achieved in the communities will usher the process along. Although the research team won’t know the final results for a few months, the trials have already had a significant effect on the patients that participated.

“What’s been really interesting is to see them feeling like somebody cares about them,” Dr. Hirsch-Moverman says. “Through the messages, they feel cared for. They feel that someone in the clinic is thinking about them. It’s automated, but to them it’s not necessarily.” Emphasizing the humanity of this approach, Dr. Hirsch-Moverman adds, “It’s beautiful.”

Precision medicine

Indeed, the human touch is a focal point for patient-centered care. However, with his work on biomarkers, Dr. O’Donnell is interested in yet another dimension of individualized treatment for TB. He’s striving to foster a shift towards the use of precision medicine for the disease. The current drug regimens for TB are highly standardized, and often don’t detect when patients have drug-resistant strains until several treatment programs have failed. Dr. O’Donnell’s work with the TB-detecting virus his lab developed, as well as biomarker research in general, has the potential to change those outcomes.

“It’s a way of detecting way early if [the patient] has two populations of bacteria, one’s antibiotic resistant and one’s susceptible. As you get antibiotics that kill down the drug-susceptible, the drug resistant will emerge. As it’s emerging, we want to detect it at say, one part in 100,000 rather than waiting until it’s 50-50.”


If we have half a chance of fighting this disease that causes so much morbidity and mortality globally, it’s going to take the complimentary, coordinated efforts of researchers represented by this Columbia thought collective. This group’s unique set of perspectives, which encompasses the micro and the macro, the local and the global, offers fresh hope for one day adequately addressing this ages-old killer.

Edited by Dana March. Image by Dana March.