Haris, a cheerful 7-year-old, wasn’t himself on returning from a family trip. His bounce was missing, said his father, and he’d lost weight he’d put on the previous summer spent in Central India, sampling the home cooking of a doting grand-mother.
Unable to diagnose Haris’ malaise after several office visits, his pediatrician reached wider, based on the boy’s stay where the TB burden was high. A tuberculin skin test came up negative. So did the sputum sample and its six weeks’ culture for TB mycobacteria. The child had no fever, no markers for inflammation. His only symptoms were fatigue and weight loss. Just to be sure, a second culture was ordered.
This time, six weeks later, results were positive. By now, diagnosis had stretched more than three months, plus the nine months since Haris’ exposure to TB. His CTs now showed bronchial blockage and dead lung tissue. He had a cough and early pneumonia. Once in the hospital, he was started on heavy IV antibiotics, a necessary gamble since lab work to uncover drug resistance would cost more time.
A small admission here: Haris’ case happens to be hypothetical. “But our frustration is real,” says Sanjay Jain, the pediatric infectious disease specialist who laid the basis for it, “because the scenario’s so realistic.” Recently, Jain worked with a team in Central India, where sampling is easier, to update medicine’s take on pediatric TB diagnosis. “It’s maddening that accurate diagnosis can be so elusive in children,” he says.
What makes it difficult? Waiting to culture slow-growing bacteria is a prime reason. But Jain’s work with kids five and younger highlights others. Unlike adults, children often show extrapulmonary TB, with disease heaviest in brain, lymph nodes or swaths of non-lung tissues. Yet positive sputum tests rely upon mucus from infected airways. That means children with bad TB elsewhere can test negative, as in the India study: The gold-standard lab culture flagged only 15 percent of them.
“Even with positive tests,” Jain says, “nothing convincing exists to tell us where TB has spread.”
A sea change is underway. The past five years, Jain and colleagues with Hopkins’ Center for Infection and Inflammation Imaging Research (CI3R), which he heads, have crossed hurdles in visualizing live bacteria wherever they exist in the body.
CI3R researchers borrowed and greatly adapted cancer-imaging techniques from oncology that rely on sensing areas of high or unusual metabolism in the body. Last fall, they published news of prototype imaging that’s exquisitely sensitive to particular bacteria—benign, pathogenic, whatever—“so we can’t mistake what we see for inflammation or tumors,” Jain says. The technique uses mildly radioactive tracers and PET or related scanners to detect them.
Four years ago, he explains, his lab singled out molecules taken up by bacteria but not human cells. One hit was sorbitol—a favorite breath mint sweetener. The team bonded sorbitol to a radioactive isotope tracked by PET scanners. Bacteria took to the radioprobe 1,000 times more readily than human or animal cells.
Next, as a test case, mice were injected with the fast-growing gut bacteria, E.coli. The rodents had earlier eaten chow laced with sorbitol radioprobes. Within hours, their PET scans glowed like homing beacons in thigh regions where infection had settled (see photo). Most exciting was the lab’s trial of E.coli “superbugs” resistant to many antibiotics. Jain’s group monitored fadeout of the PET signal over a day, as the antibi-otic ceftriaxone killed bacteria.
“This proof of principle drives our parallel work for TB with other, specific tracers. Plans are underway for human trials soon,” says Jain. “I remain very optimistic.”