May 29, 2016
6 min read

Model used to estimate link between radiation, cancer must be re-evaluated

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Previous investigations conducted by the founders of the linear no-threshold model with respect to genetic effects and carcinogenesis represent “one of the greatest misapprehensions and oversight failures within the scientific research community,” according to authors of a study review published in American Journal of Clinical Oncology.

The controversy continues surrounding the carcinogenic risk suggested to be induced by low-dose radiation exposure.

James Welsh

James S. Welsh

Some researchers suggest a surge in use of CT scans has significantly increased the risk for radiation-induced cancers. Others contend the risk cannot be assessed using epidemiologic methods alone, as the data on radiation at low doses conflict and the linear no-threshold model — the model by which low-dose radiation has been measured for years — is not accurate and should be re-examined.

HemOnc Today spoke with James S. Welsh, MS, MD, FACRO, president of the American College of Radiation Oncology and one of the lead researchers of the American Journal of Clinical Oncology paper, about the potential association between radiation exposure and cancer risk.

Question: What was the basis for your research?

Answer: Several things, actually. For one, I have always been fascinated with the concept of background radiation levels over geological time. So, if the earth is about 4.6 billion years old, knowing the half-lives of various primordial radioisotopes — such as potassium-40, thorium-232, uranium-238 and uranium-235 — we know that there were much greater amounts of these isotopes in the distant past and, therefore, much higher background radiation levels million and billions of years ago. Life on this planet has been bathed in much higher natural background radiation levels than it has today. In fact, some estimates have suggested a sevenfold increase in background radiation levels, and I have estimated for it to be even higher than this. The point is that life must have evolved under circumstances that had far higher background radiation levels and, therefore, had to have efficient DNA repair mechanisms to handle all of that. To think that today — if we get an X-ray or a mammogram or a CT scan — that our bodies have forgotten how to deal with radiation damage to DNA just does not seem logical. This is why I wanted to explore the idea of whether low-dose radiation exposure is truly as harmful as what I was taught in school. Or, is it possible that organisms can manage with these low levels of radiation from medical procedures? We started to look at some of the earlier studies that led to the birth of the linear no-threshold (LNT) model and, when we did this, we were very surprised. Some of these very early studies on the subject that led to the model did not even show anything that would support the supposition of it at all. The data actually suggested a linear threshold model. Another thing that got me thinking about this whole concept is that I was taught in medical school, graduate school and college that all doses of radiation are potentially hazardous, and the only way to estimate how hazardous it is, is by looking at the amount of damage or carcinogenesis at Hiroshima/Nagasaki and then drawing a straight line from there down to zero. You just plot it out and figure out where on the line you are and this will give you an estimate of the likelihood of cancer. This just seemed a little too naive and too simple to me. Biological systems do not function this way. As an example, we all know that too much ultraviolet light gives you a bad sunburn and can lead to melanoma, but if you have no UV, then you can end up with vitamin D-deficiency.

Q: What is the take-home message of your study published in American Journal of Clinical Oncology ?

A: Number one, the linear no-threshold model does not have any scientific basis. The studies that led to its proposition did not even show a linear no-threshold model, but in fact showed a linear threshold model. Continued use of the LNT model could be harmful for a variety of reasons. For example, people become afraid to get medical imaging studies, emergency departments grow more reluctant to order a CT scan for someone who might benefit from one, and people will become very anxious and concerned from low levels of radiation exposure — and possibly even move from their homes because of the fear of radiation. In many ways, this radiation phobia that low levels of radiation are dangerous is based off of a model that actually can be harmful. Many thousands of individuals were evacuated from regions in Japan following the tsunami that affected the Fukushima Daiichi nuclear power plant because of the fear that any and all levels of radiation are dangerous. There was far more real danger associated with the evacuation than the hypothetical danger associated with low levels of radiation.

Q: Are you suggesting that radiation doses should not be limited , and that radiation from CT scans and other medical imaging should not be limited because they pose no harm to patients?

A: In general, based upon the evidence we have, this is correct. I think we may be a little bit more nervous about this than we need to be. The concerns that have been raised over the past 5 to 10 years that we are going to ‘give ourselves cancer’ are based upon models and predictions that use this erroneous LNT hypothesis rather than from sound data. I am not convinced that radiation doses from medical imaging are nearly as concerning as they are being made out to be.

Q: It has been suggested that radiation doses from various medical imaging technologies vary, and that institutions across the country do not all use the same dose. Would you recommend one universal radiation dose be used ?

A: This is a very interesting question. It may depend on the various technologies from one vendor to another and from one institution to another. The bottom line is that the dose needs to be whatever is required to provide an adequate diagnostic image. If the dose is too low and the image is inadequate, then the study may need to be repeated or the patient will have to undergo exploratory surgery. This would be far more harmful and wasteful of resources, and we could ironically end up giving more radiation to the patient in the long run. So, in my opinion, there is no need to make the dose uniform from one place to another. It should be whatever is necessary to provide the diagnosis at the specific institution using a specific technology within the dose range commonly used in diagnostic imaging today.

Q: W here do we go from here?

A: I would love to see more research going on in the very low-dose radiation range. There is not enough going on presently, and part of the reason for this is for practical reasons. It is so much easier to get a result if you use very high doses of radiation. This guarantees that you will get some sort of data. In contrast, if you use very low doses and expect to see something but you do not see it, then it could be very difficult to publish these negative results and gain credibility. It could also be very difficult to get a follow-up grant. But in my opinion, a negative result with very low doses of radiation is actually an important result, because it would be consistent with my hypothesis regarding these very low doses.

One area that I believe should be investigating this in greater depth is space exploration research, and I know that NASA is researching this quite aggressively now. In the past, it was more difficult to get a grant to investigate low-dose radiation effects than it was to get a grant for high-dose radiation effects because of the guarantee of some data — even if those data were irrelevant to the question of what the real effects of low-dose radiation are. This is true to some extent even today. An agency like NASA should really be focusing on the genuine effects of low-dose radiation exposure, such as what would be the effect on an astronaut on a mission to Mars or another extended space expedition. Studying what happens when you get exposed to the very high doses that clinicians use for radiotherapy, for example, is not going to provide an answer to the question of what happens to you upon very low-dose and low-dose–rate radiation exposure during a protracted space mission. One cannot simply extrapolate results from high doses down to low doses because the LNT model is not truly valid. I hope that more research will go into this area.

Q: Is there anything else you would like to add ?

A: As I mentioned earlier, the things I am learning as I re-read the old literature from over 50 years ago that led to the LNT model contradict everything I was taught in medical school, during my radiology rotations and during my residency in oncology. I would recommend that people who are interested in this subject not just trust the textbooks and believe what is in there, but go back and read the original references. They may come to surprisingly different conclusions from what we were taught in medical school — I certainly did. – by Jennifer Southall


Siegel JA, et al. Am J Clin Oncol. 2015;Published online ahead of print Nov. 3.

For more information:

James S. Welsh, MS, MD, FACRO, can be reached at Stritch School of Medicine, Loyola University Chicago, Cardinal Bernardin Cancer Center, 2160 S. 1st Ave, Maguire Center, Room 2932, Maywood, IL 60153; email:

Disclosure: Welsh reports no relevant financial disclosures.