[ Fall 2008 ]
Extensively drug-resistant tuberculosis is research focus
Covering 200 square kilometers of arid scrubland in South Africa’s KwaZulu-Natal Province, the rural district of Tugela Ferry seems to swallow its inhabitants, its parched mountain ridges shutting out the world beyond. Within the region, rural and poor, many villages have neither running water nor electrical service. A lone main road cuts like a jagged scratch through the rocky landscape. Men work the gold mines up north, or in the cities, returning to their families every few weeks or months and bringing HIV home from girlfriends and prostitutes.
Few outsiders had ever heard of Tugela Ferry before a frightening outbreak began unfolding there in 2005: At the Church of Scotland hospital, a group of patients suddenly began dying—succumbing, as it turned out, to a tuberculosis “super-bug.”
TB is rife in the developing world, especially wherever HIV, the AIDS virus, cripples the immune systems of huge numbers of people. In Tegula Ferry, state-of-the-art treatment regimens had begun offering hope to patients infected with both TB and HIV. But in several whose HIV loads were virtually undetectable, thanks to this treatment, TB proved lethal. Against cultures from these patients, every class of the powerful antibiotics required to vanquish multidrug-resistant disease (MDR-TB) proved powerless.
Similar scattered cases had been recorded by the U.S. Centers for Disease Control and Prevention (CDC) as far back as 2000, but Tegula Ferry’s outbreak had no precedent. Of 53 people definitively identified as carrying this strain, most of them HIV positive, 52 died. Alarmed, the CDC sped up one multinational TB survey and released the results in March of 2006. Officials christened the deadly disease “extensively” drug-resistant tuberculosis, or XDR-TB. News reports encircling the globe called it “virtually untreatable.”
No one knows precisely how virulent XDR-TB might be among healthy people whose immune systems aren’t already enfeebled. But with $14 million in new funding from the National Institutes of Health, researchers at the Harvard School of Public Health (HSPH) aim to find out. Driving the School’s effort is Principal Investigator Megan Murray, an assistant professor at Harvard Medical School and an associate professor of epidemiology at HSPH.
When the World Health Organization (WHO) first began tracking drug-resistant cases of TB, in 1994, the problem was thought to be confined to populations whose immune systems were compromised—people with HIV/AIDS or cancer, as well as the elderly. But today, drug-resistant strains are more prevalent than scientists ever thought possible. A WHO global survey published in early 2008, the first in four years, found MDR-TB in 72 countries and XDR-TB in 49, including the United States.
According to WHO, TB affects more than 9 million people a year and kills nearly 2 million annually by destroying the lungs. The bacterium, Mycobacterium tuberculosis, infects about one in three people on Earth; it lies harmlessly dormant in most but becomes reactivated in one in 10. Garden-variety, drug-sensitive TB is readily curable, as long as doctors prescribe the right medications and patients stick with them.
And there’s the rub. Patients must take four different antibiotics over a six-month period, a level of compliance that is hard to maintain. “Many people stop taking their pills once they’re feeling better,” HSPH’s Murray explains. As the bacteria multiply, a few by sheer chance develop genetic mutations that enable them to survive the drugs. These in turn replicate, threatening the patient; some of these strains may also be “fit” enough to jump to other people. Today, five percent of all new cases worldwide are MDR-TB, defined as being resistant to at least two crucial “first-line” antibiotics, rifampin and isoniazid.
Wiping out MDR-TB is an even longer, more grueling ordeal. Patients must take a complex regimen of six or more “second-line” antibiotic pills and injections for two years, at a cost 100 to 1,000 times higher than the first-line drugs. Side effects can include hearing loss, acute nausea, depression, and psychosis.
When health workers supervise TB patients daily to ensure they take their medications—a strategy backed by WHO since 1993 called directly-observed therapy, or “DOTS”—cure rates are high. Even for MDR-TB, and even in poor countries, cure rates of 85 percent can be achieved, as demonstrated in 1998 in Peru by the cofounders of the nongovernmental organization Partners in Health: HSPH’s Jim Yong Kim, now director of the François-Xavier Bagnoud Center for Health and Human Rights at HSPH, and Paul Farmer, a Harvard Medical School professor. In response, WHO changed its policies, adding to the DOTS strategy drugs to combat MDR-TB.
In Peru, the duo worked with second-line drugs and with the government to achieve universal access to MDR-TB care. But elsewhere, compliance problems, misdiagnoses, and the inappropriate use of antibiotics paved the way for XDR-TB on an epidemic scale. Particularly in South Africa, Russia, Indonesia, China, and India, HIV and TB are ravaging populations in a worst-case combination.
THE NEED FOR SPEED
Megan Murray grew up caring for TB patients. Her father and mother, an internist and a microbiologist, took the family from their Minnesota home to impoverished Niger on charitable medical missions. Before earning an MD at Harvard Medical School and a ScD at HSPH, Murray worked on TB screening in Thailand. Today she specializes in the molecular epidemiology of the bacterium, tracking strains by their telltale DNA footprints as they migrate.
“We want to understand the genetic changes and mutations that confer resistance when you apply drug pressure on TB,” Murray says, referring to how drugs influence the bacterium’s DNA makeup by killing off all but mutated strains capable of withstanding those drugs. What Murray learns could one day help companies develop new pharmaceuticals. But for now, she says, “Our overarching goal is to create a low-cost technology that can be used to detect resistance to TB drugs immediately, in the field.”
Cheap, rapid diagnostics are most urgently needed in developing countries, especially where both HIV and TB thrive. Until recently, diagnosis invariably took at least three weeks, as a fleet of antibiotics were tested, one by one, against cultures grown from sputum. In the United States, where active TB is rare, hospitals routinely screen suspected cases this way, but low-income nations lack the resources. In June, however, WHO and other entities endorsed new, low-cost tests developed by teams in Germany and South Africa that detect MDR-TB by flagging rifampin and isoniazid resistance in just one day.
“That’s great news,” Murray says, “but we need to expand the tool,” whose limitations include reliance on pure DNA and the availability of sophisticated laboratory equipment. “What we want to see next is a test that immediately picks up resistance to second-line drugs in the field.” With colleagues at the Broad Institute of Harvard and MIT, where she is an associate of the Broad’s Infectious Disease Initiative, Murray is working toward this goal, which she predicts is at least several years away.
SHATTERING OLD THEORIES
If drug pressure were the only explanation for proliferating MDR-TB strains, there would be less cause for worry. But there’s a simpler way to become infected: through close contact with carriers. Bacteria can travel by air, in the droplets of a cough. Strangers on a train can transmit TB, as can friends, just by breathing. This route is known as transmission.
It was HSPH Dean Barry R. Bloom, an internationally known immunologist and TB expert, and his then postdoctoral fellow David Alland, now a professor at New Jersey Medical School, who helped shape prevailing ideas about TB transmission. Previously, scientists thought most cases arose from the reactivation of latent infections acquired years, even decades, earlier. But in 1994, using genetic fingerprinting, a then-new technique of molecular epidemiology, Alland and Bloom showed that at least one-third of active TB cases in New York City resulted from recent person-to-person transmission.
The researchers found that certain clusters of patients were infected by identical TB strains. If the illnesses had been caused by the reactivation of latent TB, the strains would have been more genetically diverse, reflecting variations in time and geography at the point patients were first infected, the scientists reasoned. Similar results were obtained in a San Francisco-based study by another team.
These findings unleashed new worries that if drug-sensitive TB were transmissible, drug-resistant strains might be, too. That notion flew in the face of WHO’s treatment policies, however. At that time, WHO officials assumed that genetic mutations conferring drug-resistance also imposed “fitness costs” on TB bacteria, rendering them too weak to achieve human-to-human transmission. Drug-resistant TB “might kill you, but since it didn’t spread well, it wasn’t a public health priority,” explains HSPH’s Megan Murray.
But in 2004, Murray and Ted Cohen, an HSPH alumnus and then postdoctoral fellow who is now an assistant professor at Harvard Medical School, published evidence to the contrary. By analyzing published TB case data, they concluded that at least some drug-resistant varieties were indeed transmissible. In a paper for Nature Medicine, they used a mathematical model to show that if even small numbers of MDR-TB strains weren’t appropriately detected and treated, they could eventually overwhelm drug-sensitive TB, creating an even bigger threat.
Today, Murray and colleagues want to find out just how transmissible MDR and XDR-TB really are. With the Broad Institute, they are sequencing strains from around the world in a bid to catalogue as many resistance mutations as possible. They want to learn which mutations impose fitness costs that weaken a bacteria’s transmissibility. Conversely, they hope to learn how bacteria compensate for their genetic shortcomings to gain vigor.
MDR and XDR-TB can vary by just a few dozen points of difference within the microbes’ 4-million-letter DNA code. Murray and collaborators in South Africa made this discovery in 2007 after sequencing XDR-TB for the first time, specifically the strain that caused KwaZulu-Natal’s deadly outbreak.
James Galagan, the associate director of the Broad’s microbial sequencing center, calls a few of these mutations “spooky.” For example, one simple DNA spelling error (“single nucleotide polymorphism”) called gidB confers resistance to streptomycin, which in 1946 became the first TB cure.
“What’s kind of scary is that gidB disposes bacteria toward sudden leaps from low to high levels of resistance,” Galagan says. “That means we have to be careful how we treat these organisms. They seem to be poised for resistance at the slightest challenge.”
Documenting movements of strains with the gidB variant—or any TB strain, for that matter—has proved devilishly hard. That’s in part because people can remain symptom-free for decades. Scientists tracking TB’s travels at the genetic level must wait for new cases to be diagnosed. Among large numbers of HIV-positive people whose immune systems are in tatters, this can happen swiftly, as it did in Tugela Ferry and, since, much of Africa.
But in healthy populations, disease tracking takes time. That’s a challenge HSPH alumna Mercedes Becerra, an assistant professor of social medicine at Harvard Medical School, is undertaking now in Lima, Peru, where TB prevalence remains high. With funding from Murray’s NIH grant—and working with Socios En Salud, the Peruvian affiliate of Partners In Health—Becerra heads the most comprehensive effort yet to follow new drug-resistant cases in households where a family member is already infected.
“We’re looking at 7,200 households. We’ll capture as many cases as we can and follow their family members over time to see if any get sick,” Becerra says. “We’ll also be looking at various factors that affect people’s susceptibility, such as nutritional status.”
In a closely watched display of molecular epidemiology in action, individual TB strains will be genetically analyzed and archived. Ultimately, the project should discover which drug-resistant strains become contagious, and why.
Meanwhile, HSPH’s Sarah Fortune, an assistant professor of immunology and infectious disease and a collaborator with Murray’s research team, is asking another, related question: Why are some TB strains able to develop so many drug-resistance mutations when doing so requires mutation of the DNA, which is normally deleterious to the organism? One theory, Fortune says, is that highly drug-resistant strains are “hypermutable”—that is, they have acquired the ability to turn off pathways that normally make DNA replication so accurate. This may be a successful survival strategy in clinical cases, where the bacterium must withstand an onslaught of multiple antibiotics.
Ultimately, some of the biggest questions about XDR-TB come down to virulence, the speed with which it kills. One look at the Tugela Ferry experience reveals how quickly patients die in the presence of HIV; nearly all patients in the initial outbreak perished within a few weeks. But whether the disease will behave similarly in people with intact immune systems isn’t known.
“As far as we know, based on TB biology, XDR-TB won’t spread from human to human any faster than other forms do. It will just be more difficult to treat,” observes Murray. In other words, it may kill your friend or neighbor or countryman, but it won’t necessarily affect you. Of course, there is as yet no way to know for sure, Murray concedes. “There may be things about XDR that we just don’t know yet.”
A new study of HIV-negative drug-resistant TB patients in Lima, published in the New England Journal of Medicine in August of 2008, raises both hope and concern. On one hand, the study, led by an HSPH alumna turned Harvard Medical School instructor and Partners In Health research associate, Carole Mitnick, showed that with proper drugs, care, and psychosocial support, about 60 percent of XDR-TB patients could be cured—despite having undergone several failed drug regimens. Moreover, patients recovered at home, outside the confining walls of a hospital, using a model of care suitable for low-income countries.
On the other hand, “Only 60 percent can be cured when they’ve already been subjected to multiple inappropriate drug regimens,” says Mitnick, who calls Murray a mentor and whose research team included Becerra and six other HSPH alumni. She adds, “Changing global policy to permit early screening for, and proper treatment of, drug-resistant TB would lead to more promising prognoses.”
Overcoming hurdles to diagnosing multi-drug-resistance faster is a top priority. So says Christine Sizemore, the microbiologist who heads the TB, Leprosy, and Other Microbial Diseases Section at the National Institute of Allergy and Infectious Diseases, which funds Murray’s project team.
Time won’t wait. The longer patients get no drugs, or drugs that don’t work, Sizemore warns, the harder MDR and XDR-TB will be to contain.
Charlie Schmidt has written for Discover magazine, the Washington Post, Environmental Health Perspectives, Science, and Nature Medicine. Karin Kiewra is editor of the Review.
Photos: Megan Murray and TB researchers, Kent Dayton; TB image, Photoresearchers, Inc.
Originally published in Fall 2008