A New Tool for UTI Detection
“When adults have a urinary tract infection [UTI], they can tell the doctor it burns when they pee,” says Nader Shaikh, MD/MPH associate professor of pediatrics at the University of Pittsburgh and UPMC Children’s Hospital of Pittsburgh. “You check their urine, and they’re usually right.”
With diaper-wearing infants and pre-verbal toddlers who present with a fever in the emergency department (ED), diagnosing UTIs is much more difficult. “You can’t ask babies to pee in a cup,” says Shaikh, “and they don’t talk, so we usually have to do a catheterization to get a clean sample.”
But catheterization is upsetting and painful. Children often need to be held down while the genital area is cleaned and the tube is inserted. Then, it’s fingers crossed that there even is any urine on board to be emitted, sampled, and sent to the lab. When children test positive, then they must also undergo an ultrasound.
The effort might be worth it if a UTI diagnosis is highly likely, but the probability of UTI is just 5 to 10 percent in a baby presenting with a fever in the ED, says Shaikh. “That means you have to test approximately 10 kids to get the child with the UTI.” So, which ones do you test, and which ones do you leave alone?
In spring 2018, Shaikh and his colleagues published in JAMA Pediatrics a study of UTICalc, a free, Web-based UTI probability calculator they developed at Children’s.
In a large data set of children who were 2 years old or younger (and another large data set to validate the first), the investigators demonstrated that UTICalc reduced UTI testing by 8 percent and decreased the number of UTI misdiagnoses to zero.
The first step of UTICalc assesses a child’s risk factors. It includes five yes/no questions on age, temperature, other source of fever, gender, and race. (White children are more likely to contract UTIs.) “Once you put in the child’s baseline information, you get the probability of UTI,” Shaikh says. Then the physician can decide whether to sample the child’s urine.
The second step focuses on interpreting the urine test results. When the labs come back, physicians enter specific data points (like white blood count). UTICalc then provides a more refined probability of UTI.
“If the probability goes down to 1 percent, you might say, I’m not going to treat; this is probably not a UTI,” says Shaikh. “If it goes up, then you might decide the other way.”
By improving diagnostic accuracy, the calculator can help reduce the number of patients who are being overtreated with antibiotics. It can also help kids who should be getting catheterized but aren’t.
“Both mistakes happen, but it’s more common that we overtreat,” says Shaikh. “And part of that problem is that the tests we have right now are good but not great. We can’t predict well who’s going to have a UTI or not, so we tend to be more careful and treat more children than need it.”
The UTICalc team put a lot of effort into building an accurate tool that keeps the number of required fields to a minimum, so that it’s user-friendly, Shaikh says. Ultimately, the end product’s variables were very similar to those of an American Academy of Pediatrics algorithm, he adds, but UTICalc also makes probability cutoffs (the percentage chance of UTI) more transparent to physicians.
In the first six months after the JAMA Pediatrics paper, analytic reports showed nearly 16,000 users and more than 45,000 entries. And several clinicians and researchers have contacted Shaikh about presenting UTICalc at a journal club or including it in future study designs.
Shaikh is a member of the UTI Center at Children’s, one of the few centers in the country devoted to addressing these elusive childhood infections.
“There are so many mysteries about UTI in babies,” Shaikh says. “It’s hard to understand why some get a UTI, and some don’t. They’re all sitting in diapers!”
And a Bright Spot for Parkinson's
Most cases of Parkinson’s disease seem to be caused by some murky combination of genetic and environmental influences. In about 1 out of 10 people with the condition, though, genetic glitches are solely at fault. The most common among these glitches is a mutation in a gene called LRRK2, which drives its encoded protein into a frenzy of activity.
Pharmaceutical companies are developing drugs that hobble LRRK2 protein activity to treat the disease in the small subset of patients who carry the mutation. But J. Timothy Greenamyre, professor of neurology at the University of Pittsburgh, thinks meddling with LRRK2 might have a bigger payoff. He and his colleagues report that the protein plays a central role in the disease, even when the gene is not mutated. “Rather than targeting something like 3 percent of Parkinson’s patients,” he says, drugs that tamp down LRRK2 activity “may help virtually everyone with Parkinson’s disease. And that’s good news.” The team published their findings last summer in Science Translational Medicine.
Greenamyre had a hunch that LRRK2 had deeper ties to the disorder than researchers had previously suspected. But studying the protein was a challenge because only a tiny amount of it is present in the cell. So as a first step, he and his colleagues set out to develop a sensitive method to visualize it in action. By creating a so-called molecular beacon, a marker designed to glow red when the protein is mid-task, they could see that LRRK2 tends to sit in a type of nerve cell that gradually atrophies in people with Parkinson’s disease.
When they deployed the molecular beacon in tissue taken from people who had died from nongenetic forms of the disease, they found that LRRK2 protein activity was surprisingly high. Further studies suggested that oxidative stress ratchets up LRRK2 activation, which could explain how environmental factors can kick the gene into disease-causing mode, even when it’s not mutated.
The researchers then turned to a rat model of Parkinson’s. In the model, animals are fed rotenone, a chemical used in pesticides, which causes a real mess—clumps of a protein called alpha-synuclein accumulating over time, much like they do in humans as Parkinson’s advances. Again, by deploying their molecular beacon, Greenamyre and his team found that LRRK2 glowed bright with activity within cells.
When the team gave the rats rotenone along with a compound that prevents LRRK2 activation, however, the alpha-synuclein clumps didn’t form. Alpha-synuclein is normally cleared away by a cellular waste disposal system called autophagy, but that process was disrupted in rats given rotenone. Their brain cells had an unusually low number of lysosomes—tiny membrane sacs that act as garbage trucks, ferrying detritus out of the cell.
“The basic story, then, is that in Parkinson’s disease, LRRK2 gets overactivated by oxidative stress,” explains Greenamyre. That interferes with a pathway that regulates autophagy, which in turn causes abnormal proteins such as alpha-synuclein to build up.
At least one company is already beginning to test LRRK2 inhibitors in Parkinson’s patients who have mutations in the gene.
In the meantime, Greenamyre’s lab wants to better understand the gene’s importance. He and his colleagues are working to further define how toxins and oxidative stress can affect autophagy via LRRK2 activity.
And the team has shown that, in rats, the new inhibitors can prevent abnormalities in cellular waste disposal caused by the disease. —by Alla Katznelson