Experts slug it out: What is 'latency'?

A huge subject about which little is known

(Editor's note: This is the first of two articles on the issue of TB latency. The second article will appear in the July issue. )

When TB Monitor recently asked a handful of experts for their views on the subject of latency, responses ran the gamut. While everyone agrees that precious little is known for certain about the subject, amicable disagreement abounded as to what latent TB infection is - or, indeed, if such a thing even exists.

Everyone agreed, however, that getting a handle on latency is a singularly important challenge facing TB researchers.

"Latency is an absolutely huge issue," says Bess Miller, MD, MSc, associate director for science at the Centers for Disease Control and Prevention's (CDC) Division for Tuberculosis Elimination. "Having one to two billion people on the globe who are latently infected makes TB control difficult - and it makes eradication and elimination extremely difficult."

The phenomenon also ensures that TB will be hanging around for quite a long time, says Barry Bloom, PhD, investigator at the Howard Hughes Medical Research Institute of the Albert Einstein College of Medicine in the Bronx, NY.

"Typically, an epidemic cycle lasts a couple of years - perhaps a dozen years at most," says Bloom. "But because TB can persist in people for a lifetime, most models say the cycle for TB lasts between 100 to 300 years."

And what about the approximate third of the world's population infected with latent disease? Scientists still have very little idea (except when it comes to external forces, such as HIV infection or malnutrition) about what factors drive people into reactivation, or what spares them.

A T-cell trying to get in

So far, the search for agents that could prevent someone who is latently infected from reactivating - in other words, for something that would act as a post-exposure vaccine - has turned up empty-handed, says Ian Orme, PhD, professor of microbiology at Colorado State University in Fort Collins, and head of the program for vaccine screening in animals for the National Institutes of Health.

Of all the post-exposure vaccine candidates Orme's shop has tested so far, nothing has come close to working, he says. Small wonder, he adds, "What you've got is a bug inside a macrophage, which is inside a granuloma; and about 10 miles outside that granuloma is a T-cell, trying to get in."

Even if something promising did turn up, experts add, how would we know? "Think how long a trial [in humans] would have to last - literally for decades," says Clifton Barry III, PhD, head of the mycobacterial research unit of the National Institutes for Allergy and Infectious Diseases in Hamilton, MT. "You try telling that to [officials in] a drug company. They just roll their eyes."

For some, of course, there's isoniazid prophylaxis, with efficacy typically rated at about 70%, says Miller. (If subjects are more compliant, a recent study shows efficacy rises to a figure more on the order of 90%, she adds.) Here, experts such as Donald Enarson, MD, director of scientific activity for the International Union Against Tuberculosis and Lung Disease, argue that in most cases, INH prophylaxis is a waste of time. Even in developed countries, many doctors dislike prescribing it; almost universally, patients dislike taking it.

That leaves scientists trying to find the clue to the latency riddle. What, exactly, is going on in a state of latency? "We don't know if the TB bug is sitting there doing nothing, or growing at a really slow rate," says Bloom. "Is it asleep? Just puttering along and slowly dividing? Or hidden away in some sort of cryptic form?"

One major obstacle is lack of a proper model, most experts agree. "There's no evidence whatsoever that what happens in vitro happens in a mouse or a human being," says Bloom. At the same time, available animal models have plenty of shortcomings, too.

In the Cornell mouse model, mice are infected with TB and given INH until there's no evidence any TB bugs remain. Test agents are administered; finally, the mice are dosed with immunosuppressants. Whatever grows back is said to be "latent."

"It's a very, very artificial model," says Barry. "It's very hard to draw any conclusions from it."

Guinea pigs aren't much better, says Bloom. Unlike mice, they do form a granulomatous response; but like mice, without benefit of chemotherapy, they bypass anything resembling a self-contained state of latency and simply fall sick and die.

Who knows, Bloom adds, perhaps rabbits hold out some promise. But so far no one's done much in the way of finding out, he says.

That leaves humans. Though inarguably the best available model, they make terrible subjects for study. Along with all the obvious shortcomings, in people infected with TB, "most of the time you can't find the bugs," says Bloom.

Inside chunks of lungs, finding viable bacilli

Not that it hasn't been done. Consider Lawrence Wayne, PhD, head of the TB research laboratory at the Department of Veterans Affairs Medical Center in Long Beach, CA. In the 1950s, Wayne and some of his colleagues began looking at resected chunks of lung, often taken from patients who'd been sputum-negative for months, perhaps even years, but who still had a lesion surgeons decided to take out, says Wayne. Inside some of those lesions, no matter how blocked and airtight it appeared to be, Wayne would occasionally find viable tubercle organisms.

He noticed something else that seemed odd, too. In an unshaken culture of tuberculosis mycobacteria, bugs grew at a linear rate - much more slowly than the log rates one would logically expect from a creature that multiplies by doubling.

Eventually, Wayne developed an in-vitro model to explain the phenomenon. What was happening, he concluded, was that the bugs were adapting to an oxygen gradient that had established itself in the unshaken test tube. Some had entered a non-replicating phase Wayne described as microaerophilic; others further down the oxygen gradient had shifted into an even lower gear and had entered a truly anaerobic phase.

According to Barry, who took Wayne's model and fine-tuned it, the key to the bug's descent into anaerobic dormancy lies in some key protein, or set of proteins, that act essentially as preservatives, allowing the dormant bug to continue carrying on essential activities.

Far from being an obligate aerobe, "This is a bug with big-scale metabolic capabilities," Barry says. Both he and his colleague Wayne believe the key to unraveling the puzzle of latency lies hidden somewhere in the microaerophilic stage - a drug that could keep the TB bug from making its "preservative" proteins would prevent it from developing the tools it needed to survive in an anaerobic stage.

"Rubbish," says Orme. "There is zero evidence that a microaerophilic stage occurs in a living, breathing animal or human being. You don't catch TB in a test tube, and you don't drive the oxygen tension to zero in your lungs." Orme goes a step further. He doesn't even believe the TB bug ever enters anything resembling dormancy.

Dynamic or dormant?

"The TB bug is always in a dynamic state, ready to kill you any chance you give it," he asserts. "What's happening is you're preventing it from growing - but it's not dormant."

Here's the proof, he goes on. "If you take the bugs out of an animal when they're in a chronic state and put them into a mouse with no immune system, they behave exactly the same as bugs that are actively dividing and start growing at exactly the same rate. If they were dormant, they'd need some time to wake up. But there's no lag time; they grow straightaway at the same rate."

Further proof is found in how the host, instead of taking a break during periods of supposed latency, maintains a continuous immune response. he adds. "There's an interferon `message' in the lungs the whole time. Obviously, you're making it, and the surveillance system is activated because the bugs are doing something."

The reason bacterial loads appear to flatten out in latency reflects what Orme hypothesizes is actually a wave that rises and falls but in amounts too small to be detectable.

Perhaps, he speculates, what's going on is that there may be two kinds of chronic disease. In one kind, some key factors are present - perhaps genetically determined in the host - that make some subjects prone to reactivate and others able to resist reactivation.

Whatever it is, "We think something happens almost immediately after you're first infected, in those first two to three weeks," Orme says. "So we've gone back to that long course of granuloma formation and breakdown, to look at what comes in, what goes out, what cytokines and chemokines are made, and so on." Indeed, he adds, when one examines the granuloma formed by resistant mice and compares them to granulomas from susceptible mice, the cellular populations differ strikingly.

Other researchers, Bloom among them, doubt that the key to latency is to be found either wholly in the host or completely in the bug. Bloom favors tweaking various genes, to see how doing so will affect latency; other researchers are looking at other bacilli that enter dormancy.

Next month, TB Monitor takes a look at research under way in some of those arenas.