A study involving thousands of British children provides the first direct evidence that low-dose radiation used in diagnostic imaging produces a small but real increase in a child's risk of developing cancer within 10 to 15 years, researchers say. The study, published today in Lancet, found that two to three head CT scans can triple a child's risk of getting a brain tumor. Five to 10 scans that deliver radiation to the bone marrow triple the risk of leukemia, the research showed.
In absolute terms, however, the estimate translates into one excess brain tumor or case of leukemia per 10,000 young patients, the study says. This finding is critical, researchers say, because it indicates that an individual's odds of getting brain cancer or leukemia from CT radiation exposure is low.
"If you need a CT scan, get one," says health physicist Owen Hoffman, of the consulting firm SENES Oak Ridge, in Oak Ridge, Tenn. "This doesn't mean that you're going to come down with cancer. The added information from a CT scan far outweighs the cancer risk. But the risk is there, and it's such that you should ask your doctor, 'Do I need the scan?' The doctor should have the information to reassure a patient that the scan is really necessary."
Still, the findings are cause for concern, researchers say, because the number of CT scans done in the United States is soaring. The most recent data suggests that U.S. doctors perform at least 70 million scans each year, 5 percent to 10 percent of them in children. A CT scan—a computer-enhanced series of X-rays routinely used to identify brain trauma, cancer and other conditions that might be missed or take longer to diagnose by other means—delivers at least 10 times the radiation of a mammogram and up to 600 times the radiological punch of a single X-ray. Studies suggest that a third or more of CT scans may be unnecessary, says health physicist David Brenner, director of the Center for Radiological Research at Columbia University Medical Center.
"If only a few people had CT scans, the risk wouldn't have any significance at all," Brenner says. "When you have a small risk but big numbers of people, there's the potential for more people to end up with cancer."
Brenner, who wasn't involved in the British study, was lead author of a landmark and highly controversial paper published a decade ago predicting that childhood diagnostic CT scans could cause cancer in later years. That study provoked a nationwide debate about exposing children to medical radiation. Many makers of CT scanners and children's hospitals took steps to recalibrate CT scanners to reduce young patients' exposures to harmful X-rays.
[See our 2012-13 rankings of Best Children's Hospitals, released yesterday.]
Children are especially susceptible to radiation-induced cancers, because their cells multiply rapidly. Radiation can damage the cellular controls that keep multiplication in check, over time transforming healthy tissue into tumors.
Brenner's paper provoked a furious backlash among radiologists and other health physicists who objected that data from the Japanese atomic bomb survivors couldn't be used to evaluate the risks of medical radiation exposures, such as CT scans. The new study, however, strengthens his case. It demonstrates that the cancer risks from low-dose CT scans are statistically on par with those seen in atomic bomb survivors, says Andrew Einstein, a cardiologist and radiologist at New York-Presbyterian University Hospital of Columbia and Cornell.
What makes the latest research different from earlier studies, including Brenner's, is that it doesn't rely on risk assessments derived from post-World War II studies of atomic bomb survivors, says Einstein, author of a commentary on the new study in Lancet. "Here we're looking directly" at young people exposed to CT scans, he says.
By necessity, it's a huge group. Radiation is a weak carcinogen, Einstein says, so researchers must study tens of thousands of people to detect a relatively small number of cancers.
Mark Pearce of the Institute of Health and Society at Newcastle University in Newcastle upon Tyne in England, and his team studied more than 175,000 people younger than 22 without previous cancer diagnoses who had CT scans in Great Britain between 1985 and 2002. The study participants were followed for at least a decade. Researchers found a statistically significant link between the radiation dose delivered by the CT scan and cancer in children younger 15. Seventy-four children developed leukemia, and 135 children developed brain cancer. The analysis indicates there's just a 5-percent likelihood that the cancer-CT link could have occurred by chance.
Despite the previous lack of evidence from a medical population, Dr. Einstein says, the controversy had a positive impact, provoking a dialogue that changed radiology. The American College of Radiology and other groups began developing new guidelines for CT scans that use lower doses of radiation. Manufacturers of CT scanners began designing their machines with options to reduce radiation exposure without diminishing the quality of the image. In 2008, the Society for Pediatric Radiology joined with the American College of Radiology, the American Society for Radiologic Technologists, and the American Association of Physicists in Medicine to launch the "Image Gently" campaign to raise awareness about radiation risks and to provide radiologists with low-dose imaging protocols and other safety information.
Stephen Swensen, a professor of radiology and radiation-safety advocate at the Mayo Clinic in Rochester, Minn., says refusing CT scans when they're called for carries risks too. "What we don't want is a vast frenzy where people don't want exams that will save their lives," he says. "We want public understanding of their risks and benefits."
To avoid inappropriate scans—such as those for chronic headache, minor trauma, and low back pain—Swensen advises every person offered a CT scan—and every parent advised that their child might benefit from one—to ask the doctor:
• How could the test result change my (or my child's) care, if at all?
• Can you recommend an alternative, such as an ultrasound or MRI, that doesn't involve radiation?
• Can a dose at the low end of the scale be used?
• What is the accuracy of the test at the recommended center? Some operators are better at getting clear images than others, Swensen says.
• Do you have a financial interest in the scanner? Doctors that do, he says, have a greater incentive to refer patients for tests.
Pearce's study was funded in part by the U.S. National Cancer Institute but was carried out in Great Britain. That's because the U.S., with its fragmented health system, doesn't have the uniform national respository of medical records available through the National Health Service. The National Cancer Institute's SEER database, for Surveillance Epidemiology and End Results, provides information on cancer's geographic variations but doesn't record whether patients have had a CT scan, Hoffman says. The lack of a national cancer registry makes it much harder to study radiation epidemiology at home—and forced the NCI to seek partners abroad.
Brenner says he spent years trying to get a similar study off the ground, but finally abandoned the effort. "I went gray trying to do a study here," he says.