David Brenner

Two researchers from CUMC’s Center for Radiological Research made waves in November by sounding the alarm about radiation exposure from CT scans. David J. Brenner, Ph.D., professor of radiation oncology and of public health, and Eric J. Hall, D.Phil., director of the Center for Radiological Research, suggested in the Nov. 29, 2007, issue of the New England Journal of Medicine that CT scans may cause tens of thousands of cancers in the future. This resulted in a storm of media coverage and, as with any finding that challenges accepted dogma and practice, the scientists’ findings generated controversy and debate within the medical community. InVivo spoke with Dr. Brenner about the debate and its significance.

Why have you and Dr. Hall chosen to speak out about CT scans now?
The number of CT scans being done each year in this country has skyrocketed, from about 3 million in 1980 to about 62 million today. This is hardly surprising because CT is such a quick and accurate diagnostic tool. The concern here relates to the fact that CT results in a radiation dose typically 100 times that from conventional X-ray exam such as a chest X-ray or a mammogram, and there is now definitive epidemiological evidence that there is a small risk of radiation-induced cancer at CT doses. Having said that, the individual risk is very small, and so the benefit/risk balance is generally by far in the patient’s favor. However, it is generally accepted that perhaps one-third of all CT scans could be avoided altogether, or replaced with a different diagnostic tool. So even though the individual risk is small, if, say, 10 million scans could be avoided each year, there would be a significant public health advantage – particularly for children, who are much more sensitive to radiation than adults.

How has the medical community reacted?
The response has largely been very positive, but the study has had its critics. These seem to fall into two categories: The first are those who have commented that no studies have as yet shown significant increases in risk from diagnostic X -rays. This is true, but given that CT scans really only took off in the 1990s, and given latency periods of 20 years and more, it has only now made sense to start an epidemiological study to look at this – and indeed we are involved in such a study just starting in Europe with regard to pediatric CT scans. But to wait to draw attention to the fact that too many CT scans are being done in this country until there is (or is not) direct epidemiological evidence would be irresponsible.
      The second group fears that many people who really need CT scans will not have them because of these cancer risk projections. I think the evidence is that this is not the case. For example, in a recent 2007 survey [Am J Roentgenol 2007;189:271] where parents were informed about CT risks, their willingness to have their child undergo a CT recommended by their doctor did not significantly change, but they became less willing to have their child undergo CT if another modality was equally effective. No parent refused or asked to defer CT after receiving the information.

How did you calculate the CT risk?
The radiation risk estimates were made using standard methodologies endorsed by the National Academy of Sciences [Health Risks from Exposure to Low Levels of Ionizing Radiation, National Research Council, 2006]. We used long-term epidemiological studies of individuals exposed to the same doses as those from a couple of CT scans (most patients undergoing a CT scan of the abdomen or pelvis get more than one CT scan on the same day). In particular, about 40,000 individuals were about 2 miles from the Japanese atomic explosions, and they received radiation doses similar to those from two CT scans. This large group of people have been followed for more than half a century and show a small but statistically significant increase in cancer risk. Using risk estimates from this low-dose group means that we do not have the uncertainties associated with extrapolating risks from high doses to low doses.
      This standard methodology of using risk estimates derived from atomic bomb survivors who received low doses, and applying them to a low-dose radiology scenario, has certainly been criticized. The situations seem very different, but if one considers only the survivors who were a long way from the explosions, they really aren’t. If both groups got similar types of exposure, it would be hard to see why A-bomb survivors would be at risk for cancer and CT scan patients would not be.

How might CT usage be made safer?
There are a number of areas in which CT usage could be decreased without affecting patient care. There is the potential for more evidence-based clinical decision rules about when CT is used. Such guidelines do exist, such as the American College of Radiology Appropriateness Criteria, but they could be extended and applied more often. This will surely require changes in attitude, both among physicians and patients, so broad discussion of the issue is most welcome.
      Another way in which CT usage could be decreased is by using alternative imaging modalities, where appropriate. A good example is diagnosis of pediatric appendicitis. In the past this was generally done using ultrasound and clinical observation – which has now been largely replaced by CT, certainly an excellent tool but arguably not a better one.
      A third area relates to the duplication of the same CT scan, simply as a patient moves through the medical system – for example, from one hospital to another. There is no reason in principle why a patient should not have access to his or her CT scan, on a CD or jump drive, which could go a long way toward eliminating most needless repetitions of a scan.

—Anna Sobkowski