Low Radiation Doses: Small Risks? No Risks? Or Risks to Only a Few?
See also the article by Sakane et al in this issue.
Whereas the benefits of CT in almost every sphere of medicine are indisputable, over the last 2 decades the potential cancer risks associated with low doses of radiation from CT have been much discussed and debated. Undoubtedly doses from one or two CT examinations carry smaller risk for cancer than higher radiation doses, but is the risk actually zero? More than 90 million CT examinations are performed each year in the United States and so the sheer numbers of individuals exposed makes this a pertinent question.
A number of different approaches have been used to estimate possible cancer risks associated with the low doses of radiation from CT examinations. The modelers were first, estimating CT cancer risks on the basis of survivors of the atomic bomb who were exposed to low radiation doses (1). Next, epidemiologists performed direct studies of cancer risks in individuals who underwent CT as children (2,3). Results from these two approaches are in broad agreement (4), suggesting small but nonzero risks. Both approaches have some methodologic issues, which is not surprising when the estimated risks are so much smaller than background cancer risks. In such situations, biases are inevitable, and the significance of these biases needs to be carefully assessed (5).
In this issue, Sakane et al (6) have taken a third approach to the problem. They measured DNA double-strand breaks and unstable chromosome aberrations in the blood of over 100 individuals who underwent standard-dose chest CT examinations (5-mSv effective dose, 13.3-mGy blood dose) and over 100 individuals who underwent low-dose chest CT examinations (1.5-mSv effective dose, 3.5-mGy blood dose). Measurements were taken both before and after the examination. Although not a new approach, it represents a substantial advance because of the scale of the study, its use of two doses, and two DNA damage end points.
For both dose groups, the authors observed increased DNA damage after the CT examination versus before the examination. However, whereas the increased DNA damage was statistically significant in the standard CT dose group, the increase in the low CT dose group was smaller and no longer significant. This pattern reflects low-dose radiation epidemiologic studies of cancer risk, in which as the dose and the risk decrease, exponentially more participants are required in the study to achieve statistical significance (7). Sakane and colleagues (6), however, studied approximately the same number of individuals in both dose groups, so loss of statistical significance at the lower dose is not surprising.
Another way of looking at these results is in terms of the numbers of individuals affected: 64% of the participants in the standard-dose CT group showed increased DNA double-strand breaks, whereas among participants in the low-dose CT group (in whom the dose was decreased almost four-fold) about 47% still showed increased DNA damage. So, even with a substantially reduced dose, the number of affected individuals changed comparatively little. This hint of a radiosensitive subpopulation dominating the risk is reminiscent of the groundbreaking study (8) of North African families who emigrated to Israel and underwent low-dose scalp irradiation for tinea capitis. That study strongly suggested that sensitivity to low-dose radiation–induced meningioma was highly familial and thus also largely limited to a radiosensitive subpopulation.
Sakane et al (6) are conservative regarding the implication of their DNA damage results, although they do reference one epidemiologic study suggesting a potential link between DNA double-strand breaks and cancer risk; however taking all the available evidence to date, it would be fair to say that such a causal link is not well established at this time (9). If such a link did indeed exist, then the results from this study would of course imply that we should optimize or reduce CT doses per examination as much as is practically possible, which is a conclusion that few would debate. A second conclusion, again not unfamiliar, is that whereas the study is admirably large, it is not large enough at the lower dose to provide good evidence either way regarding “zero risk.” But because most CT examinations represent an enormous benefit to the patient, whether the radiation risk is zero or just very small is neither here nor there for a patient who needs to undergo CT. It is, however, relevant for the 20% (or possibly 10% or 30%) of patients who undergo CT without a good clinical rationale (10). For these patients, the benefit-risk balance is different and must be much smaller.
Finally, there are hints in these results (6) that support earlier studies (8) suggesting that perhaps we should not be thinking of the general population as having a single bell-shaped distribution of radiosensitivities. It may be better described as two distinct populations, the larger one less sensitive to low radiation doses and the smaller one much more sensitive. If only we could identify the individuals in each group!
- 1. . Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176(2):289–296. Crossref, Medline, Google Scholar
- 2. . Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet 2012;380(9840):499–505. Crossref, Medline, Google Scholar
- 3. . Cancer risk in 680,000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians. BMJ 2013;346:f2360. Crossref, Medline, Google Scholar
- 4. . Cancer risks from CT scans: now we have data, what next? Radiology 2012;265(2):330–331. Link, Google Scholar
- 5. . Is there unmeasured indication bias in radiation-related cancer risk estimates from studies of computed tomography? Radiat Res 2018;189(2):128–135. Crossref, Medline, Google Scholar
- 6. . Biological effects of low-dose chest CT on chromosomal DNA. Radiology 2020;295:439–445. Link, Google Scholar
- 7. . Estimating cancer risks from low doses of ionizing radiation. Science 1980;209(4462):1197–1203. Crossref, Medline, Google Scholar
- 8. . Genetic predisposition for the development of radiation-associated meningioma: an epidemiological study. Lancet Oncol 2007;8(5):403–410. Crossref, Medline, Google Scholar
- 9. . Fifth Warren K. Sinclair Keynote Address: Issues in quantifying the effects of low-level radiation. Health Phys 2009;97(5):394–406. Crossref, Medline, Google Scholar
- 10. . Analysis of appropriateness of outpatient CT and MRI referred from primary care clinics at an academic medical center: how critical is the need for improved decision support? J Am Coll Radiol 2010;7(3):192–197. Crossref, Medline, Google Scholar
Article HistoryReceived: Jan 26 2020
Revision requested: Feb 5 2020
Revision received: Feb 10 2020
Accepted: Feb 11 2020
Published online: Mar 10 2020
Published in print: May 2020