Neonatal infection
test may save lives, reduce hospitalizations
BY WILLIAM A. WELLS
A new test devised at
Stanford may allow rapid and reliable diagnosis of
infections in newborns. This could reduce
hospitalizations and antibiotic usage for uninfected
babies, while saving the lives of babies whose infections
might otherwise go undetected.
Infants' immune systems
are not fully developed at birth, so infections
contracted before or during birth are extremely
dangerous. But these infections are also hard to
diagnose.
"Newborns can die
within six hours if they are infected, but they don't
have the usual signs of infection," said Leonore
Herzenberg, professor of genetics and senior author of a
research report on the new test in the March Journal of
Pediatrics. "So we collaborated with our colleagues
in pediatrics to devise a blood test that would identify
infected newborns, using only a single drop of
blood."
Knowing which babies to
treat could save millions of dollars annually in
unnecessary health care expenses. Every year, up to
300,000 newborns in the United States receive treatment
for suspected infections, at a cost of more than $800
million even though only one in 17 has a confirmed
infection, the researchers note in their paper.
Erica Weirich, a
final-year medical student and the lead author,
coordinated the study and conducted most of the lab work.
In addition to Weirich and Herzenberg, the other authors
are Dr. Ronald Rabin, now at the National Institute of
Allergy and Infectious Disease; Dr. Yvonne Maldonado and
Dr. William Benitz, associate professors of pediatrics;
Siv Modler, pediatrics study coordinator; and Leonard
Herzenberg, professor of genetics.
"In neonates, it's
hard to define what is an infection and what is not. The
babies may act irritable, they may not feed well, or they
may have a whole host of very minimal symptoms in the
first day or so of life," Maldonado said. "You
end up treating a lot of children for suspected bacterial
infections when in fact they have either a benign viral
infection or just variations in normal behavior. It's
worth the cost if it saves lives, but you want to
minimize a newborn's time in the hospital and away from
parents."
The most direct way to
diagnose a bacterial infection is by trying to grow the
offending bug in a laboratory culture, starting with a
sample of urine, blood or spinal fluid. Another way is to
test for C-reactive protein (CRP), produced by the immune
system in response to infection.
Both methods can take days
to yield results. Bacterial cultures need at least 24 to
48 hours to grow, and CRP testing requires three
consecutive days of negative results to rule out an
infection, because the body can be slow to begin
producing the protein. Meanwhile, antibiotic treatment
and hospitalization must continue. Cultures can also miss
bacteria that fail to grow, and with the CRP test, stress
from a difficult birth or other disease conditions can
lead to false-positive results.
The new test measures a
different protein, called CD11b. This protein rapidly
appears on the outside of white cells called neutrophils,
as part of the immune system's response to infection. In
test tube experiments, the amount of CD11b on the cell
can increase within five minutes of exposure to bacteria.
With such a speedy
response, the Stanford researchers were able to identify
infections using blood samples collected from infants
right after birth.
The researchers studied
106 infants who were being monitored for bacterial
infection in the neonatal intensive care unit of Lucile
Packard Children's Hospital. Five had bacterial
infections that were later confirmed by cultures. All
five of these infants showed high levels of CD11b when
they were first admitted to intensive care, but did not
show high CRP levels until later generally, one to two
days after admission.
Two babies had confirmed
viral infections. Both of them had high CD11b levels at
admission, and neither had high CRP levels.
In 17 babies, doctors
suspected bacterial infection because of outward
symptoms, but cultures did not confirm the infection.
Sixteen of these infants had high CD11b levels at
admission. Fifteen later showed high CRP levels.
Of the remaining 82 babies
with no evidence of infection, all had low CD11b levels
at admission, although two later showed high CRP levels.
Thus, within six hours of
birth, the CD11b test correctly predicted the diagnosis
of all the babies with confirmed infections and all but
one of those with suspected infections. It also correctly
excluded all babies who were infection-free.
This predictive value of
105/106 is extremely high, said Leonore Herzenberg. To
confirm it, she and her colleagues propose a multicenter
trial looking at up to 3,000 babies.
"A test like this has
value in giving doctors more information about the
infants they treat," she said. "The second
value may be in cutting down on intensive care admissions
if we get the predictive values high enough. And the
third issue is looking at all babies to catch infections
that might be missed. At Packard Children's Hospital we
tend to err on the side of safety, treating all babies
who show any signs of infection, but that isn't possible
in all places."
The new test uses a flow
cytometer, or FACS machine, to measure the amount of
CD11b protein on the outside of each cell. The first flow
cytometer was developed at Stanford in the early 1970s by
a team led by Leonard Herzenberg. Although this is a
complex technology, it is now widely available, used for
everything from detecting cancer cells to measuring the
levels of immune cells in AIDS patients.
To collect a single drop
of blood as soon as possible after each newborn in the
study was admitted to neonatal intensive care, medical
student Erica Weirich stayed on call for many, many
hours. "What she did required exceptional diligence
and planning," said Leonore Herzenberg.
"There was a lot of
running between the neonatal intensive care unit and the
laboratory," Weirich recalled.
As part of a standard
hospital protocol, however, the test should be simple to
use, Herzenberg added.
"This may be a new
and faster method to discriminate between infected and
non-infected neonates," commented Maldonado.
"It seems like a really nice way to keep children
out of the hospital who don't need to be there."
The study was supported by
a National Cancer Institute Outstanding Investigator
grant to Leonard Herzenberg and by Health and Human
Services grant M01-RR00070 (General Clinical Research
Centers). The flow cytometry was done at the shared FACS
facility in the Beckman Center. SR
|
