Giant rats detect tuberculosis

Low-income countries struggling to keep tuberculosis under control might get a boost from an unlikely source — giant African rats.

The big rodents spotted hundreds of TB-positive sputum samples that a standard microscope test missed on first pass, researchers report in the December American Journal of Tropical Medicine and Hygiene.

The TB bacterium currently infects one in three people worldwide, the World Health Organization estimates, with the highest rates in Africa.

Giant African rats, also called Gambian pouched rats (Cricetomys gambianus), are native to much of Africa and have been used before to sniff out land mines. Training captive-bred rats to detect TB takes about five or six months, says study coauthor Alan Poling, a psychologist and animal-learning expert at Western Michigan University in Kalamazoo. Attempts to train captured giant rats didn’t work because the animals proved unmanageable, he says. “In the wild, they’re really nasty.”

The rats are exposed to sputum samples through holes in the floor of a cage, and if they correctly pause for five seconds to smell a TB sample, they are rewarded with a mouthful of banana. Lingering over non-TB samples gets no reward. Eventually, the rats can check a string of holes moving “about as fast as they can walk,” Poling says.

In the largest analysis to date, Poling and a research team in Tanzania collected sputum samples from more than 10,000 people and tested them using a standard microscopic analysis for TB. The researchers found that about 1,400 people had the disease. The TB samples had been rendered noninfectious using heat and pressure, a safety measure designed to prevent transmission of live TB from dropped or spilled samples, Poling says.

When the rats sniffed the same sputum samples, the animals correctly identified more than 90 percent of those found as positive in the lab. But the rats also tagged more than 1,400 additional people as positive. When reanalyzed more closely under the microscope, those samples turned out to include 620 people who indeed had TB. Thus, while the standard test found that 13.3 percent of people were TB-positive, the rats bumped that figure up to more than 19 percent.

It may seem that the rats turned in a lot of false positives, sensing samples as TB-positive that a second microscope test couldn’t confirm. But many might not be false, Poling says. “We’re thinking that in at least some of these people, there are bacteria present that our people [using microscopes] didn’t see. There might be low concentrations.”

The gold standard for TB testing is to culture the bacteria in the sputum, but that can take weeks. “There’s a need to deploy a new generation of technologies to get new diagnostics out there against TB,” says Peter Hotez, a physician and microbiologist at George Washington University in Washington, D.C. “[Using rats] is definitely a low-tech approach,” he says.

But he says the rats’ accuracy rate would need to improve for this to become a standard screening strategy. A new lab test that uses a polymerase chain reaction is 98 percent accurate in detecting TB, researchers reported in the Sept. 9 New England Journal of Medicine. That technique, which copies key bacterial DNA, appears to be the TB test of the future, Hotez says. It is being tested in the field and may become commercially available soon.

Poling and his group are using the DNA test to verify results from the standard microscopy and the rat tests. But the DNA test currently costs about $20 per sample, he says, which is far too high for developing countries. It’s unclear how much the rats would cost per sample, he says, since the method has not been scaled up commercially. But once the rats are trained, they can whip through a dozen samples in less than a minute and be used again and again.