Computed tomography scans and cancer risk: The latest findings


  • Carrie Printz

Illustration 1.

Two new studies add evidence to the risks associated with scanning

Editor's note: This is the first installment of a 2-part series.

The potentially harmful effects of ionizing radiation exposure associated with computed tomography (CT) scans, particularly among children, continue to be a subject of concern among both the general public and the medical community. At the same time, there is debate as to how large the risk really is, and what should be done about it.

Although some experts believe the potential cancer risks associated with CT scans have been exaggerated in the mainstream news media, it is clear that the rate of such scanning has increased significantly. According to a special report published in Radiology in 2011, the collective dose to the US population resulting from medical imaging has increased 6-fold in the past 25 years.1 In addition, the authors note, the number of CT scans has increased more than 10% per year over the past 15 years, and the per capita rate of CT usage is among the highest in the world (second only to Japan).

Nevertheless, radiologists and others say that one reason CT scanning is on the rise is because it offers many benefits, including the rapid diagnosis of life-threatening vascular conditions, shorter hospital stays, and more efficient treatment after injury, to name just a few. “As long as an exam is appropriately ordered, the benefit to that individual patient far outweighs any risk,” says Richard L. Morin, PhD, chair of the American College of Radiology Dose Index Registry and chief medical physicist at the Mayo Clinic in Jacksonville, Florida. “There's no question that you can see disease sooner than ever before and that thousands of lives have been saved.”

However, he adds, the radiology community and other researchers have not been able to develop an effective way to measure “benefit.” And now, 2 recently published studies have added more evidence to the risk side of the equation.

Direct Evidence of Risk

In one report, published in June 2012, children and young adults who were scanned multiple times by CT were found to have a small increased risk of leukemia and brain tumors in the decade after their first scan.2 The findings were from a study of more than 175,000 children and young adults led by researchers from the National Cancer Institute (NCI) and the Institute of Health and Society at Newcastle University in Newcastle-Upon- Tyne in the United Kingdom.

The research is the first cohort study of pediatric CT scans and subsequent cancer risk, according to senior investigator Amy Berrington de González, PhD, of the NCI's division of cancer epidemiology and genetics, who puts the findings into context: “It is the first direct evidence that CT scans in childhood may increase the risk of developing leukemia or brain tumors. The magnitude of the excess risks are, however, very small.” Investigators estimate that for every 10,000 CT scans of the head performed in children aged 10 years or younger, 1 case of leukemia and 1 brain tumor would occur in the decade after the first CT beyond what would have been expected had no CT scans been performed.

Dr. Berrington de González and her colleagues obtained CT examination records from radiology departments across Britain and linked them to data regarding cancer diagnosis and deaths. Participants underwent CT scans at British National Health Service hospitals from birth to age 22 years between 1985 and 2002. Data concerning cancer incidence and mortality from 1985 to 2008 were gathered from the National Health Service Central Register database.

Despite such evidence of the risk associated with CT scans, Dr. Berrington de González notes, “As long as the scan is clinically justified and performed using appropriate scanner settings, then the benefits should far outweigh the small radiation risks.”

Some in the radiology community question the study's findings because it was not a prospective randomized trial that included both children who were irradiated and those who were not and who were then followed for the next 20 years or so. These individuals also state that newer CT scan technology involves significantly lower doses of ionizing radiation than that used during the study period.

However, Dr. Berrington de González maintains that the historical cohort is as valid a design as a prospective cohort. Furthermore, she notes that investigators' risk estimates are based on the estimated cumulative radiation dose to each patient rather than the number of CT scans. They then converted the dose estimates back to the number of scans, assuming newer technology and current settings.

Illustration 2.

CT scans in childhood may increase the risk of developing leukemia or brain tumors. The magnitude of the excess risks are, however, very small. –Amy Berrington de González, PhD

Still more evidence concerning the impact of CT scans on potential pediatric cancer risk is being gathered. Dr. Berrington de González notes that 2 other major cohort studies are currently ongoing, one in Australia and one in Canada. In addition, a multicenter European study will conclude around 2015. Meanwhile, her group is following their cohort and will conduct more in-depth studies, with reports likely to be released within the next 2 years.

Scanning Overused in Hodgkin Lymphoma

The second study, also published in June 2012, found that CT scans are overused for the routine surveillance of patients with intermediate and advanced stage pediatric Hodgkin lymphoma.3 The Children's Oncology Group study found that the majority of children with the disease develop disease recurrence within the first year after therapy or had disease detected based on changes in clinical symptoms. Of the 216 patients treated in the trial, only 4 patients had their recurrence detected exclusively by surveillance imaging after the first year.

Investigators undertook the study because pediatric cancer patients are increasingly surviving, particularly those with Hodgkin lymphoma, who now have a 95% survival rate. Radiologists at Boston Children's Hospital began to consider whether physicians could reduce the amount of surveillance of these patients while still maintaining their survival, says lead author Stephan Voss, MD, PhD, director of nuclear medicine and molecular imaging at Boston Children's Hospital in Massachusetts.

Dr. Voss and his colleagues were somewhat surprised by the results, which demonstrated that the patients who fared the worst developed early disease recurrence. When recurrences were found in asymptomatic patients 4 years later by routine CT scan, researchers noted no impact on their long-term survival. Dr. Voss speculates that cancers that are chemosensitive tend to be less aggressive. “During that first year, aggressive monitoring is important, but there's not a need for that 4-year scan,” he says. “We just don't want to scan 100% of the patients to find the 1% who do relapse. The clinical exam findings should guide the kind of imaging you do.”

Although Dr. Voss and his colleagues did calculate the “attributable risk” of developing cancer directly due to the CT scans, it is not an easy figure to arrive at because these patients also had undergone more chemotherapy and radiotherapy than the standard population. “One argument is that the additional CT scanning doesn't really matter because these children had exposure to such high doses of radiation,” he says. “The alternative argument is that it does matter for that very reason because these patients are already at higher risk of developing cancer.”

Their calculations, based on those that other groups have done, indicated that in a population of approximately 1700 patients, about 8 new cancers would develop due to the use of CT imaging for surveillance. “It's a small number, but it's not 0,” says Dr. Voss. “The general principle is that if it's not going to help, don't do it.” While physicians were not immediately willing to embrace this idea, they are now beginning to do so because their patients with Hodgkin lymphoma are doing so well, he adds.

For this study, Dr. Voss and his colleagues examined patients who were at intermediate and high risk and are planning to eventually analyze results for low-risk patients. Other research groups are beginning to consider whether surveillance imaging can be reduced in patients with other cancers, particularly those that are less aggressive. Dr. Voss and his colleagues also are studying when it is appropriate to reduce radiation doses for scanning purposes.

Many of these issues still need to be decided because some facilities may be reducing radiation doses across the board. “My feeling is we should tailor the technique to the clinical indications,” he says. Part 2 of this article, which will appear in the March 1 issue, will review what the radiology community and others are doing to ensure the safest, most effective dose of ionizing radiation in CT scans.