Heart Scans Increase Risk Of Cancer – Or Maybe Not

February 27th, 2011
submit to reddit Share

It’s astounding how much press you can get these days as soon as you start throwing around the word “radiation” and “cancer.” It seems that just by saying that something might possibly maybe perhaps increase the risk of cancer by a tiny bit you can be guaranteed massive coverage. Perhaps it’s that people just love to hate radiation or that they love to fear cancer. Maybe it’s that radiation has become associated with “big corporations” and “unnatural” medicine that should be much more “gentle.”

(and maybe I’m using too many quotes.)

It seems that finding out that diagnostic heart scans cause a tiny increase in the rate of some cancers would not be news worthy. After all, just about every procedure has some risk involved and most of these are preformed on people who are suffering from major cardiac problems, such as those who just suffered a heart attack. In such circumstances these tests undoubtedly save lives.

Never the less, this apparent tiny increase in risk has gotten tremendous press.

Via CTV:

Heart attack survivors who undergo scans and nuclear medicine tests tend to have higher rates of cancer than those with less exposure, a new Canadian study suggests.

The researchers, from the McGill University Health Centre and the Jewish General Hospital in Montreal, note that the use of cardiac imaging tests has exploded in recent years in both Canada and the U.S.

And yet, they point out, little attention has been paid to the cumulative effect of the radiation used in those tests, or on how they might be affecting cancer rates.

So for this study, which appears in the Canadian Medical Association Journal, the researchers looked almost 83,000 patients who had a heart attack between 1996 and 2006, but who had no history of cancer.

About 77 per cent underwent at least one cardiac procedure using low-dose ionizing radiation within a year of the attack.

The tests included a heart imaging test called myocardial perfusion imaging, angiogram procedures called diagnostic cardiac catheterization and percutaneous coronary intervention, as well as a form of nuclear imaging called cardiac resting ventriculography All the tests involve exposing patients to low-dose ionizing radiation.

While most patients received only a low or moderate level of radiation, a substantial group were exposed to high levels through repeated tests — and these patients tended to be younger, healthy men.

The study’s lead author, Dr. Louise Pilote, a researcher in epidemiology, says her team found a distinct link between the cumulative exposure to low-dose ionizing radiation from cardiac imaging and the risk of cancer.

Over the course of the study, the researchers found 12,000 incidents of cancers, with two-thirds of the cancers affecting the abdomen/pelvis and chest areas.

They calculated that for every 10 “milliSieverts” of ionizing radiation, there was a 3.0 per cent increase in the risk of cancer, the study found. (A milliSievert is commonly used to measure the radiation dose in diagnostic medical procedures.)

Since the risk for cancer grows with age, the researchers accounted for that in their calculations.

“These results call into question whether our current enthusiasm for imaging and therapeutic procedures after acute myocardial infarction should be tempered,” she and her co-authors conclude.

But Pilote also cautioned that patients who need to undergo scans after a heart attack shouldn’t be dissuaded from undergoing the tests because of these findings.

“In cardiac patients who just sustained a myocardial infarction [heart attack], it’s clear that the exposure to these radiation (procedures) is warranted,” Pilote told The Canadian Press in an interview.

“And it probably way outweighs the risk of them ever developing a cancer.”

I’m sorry but I simply cannot accept the results of this study as being reliable at all. That’s not to say that it’s impossible that there isn’t a risk, there very well could be, but the study in question has a very large flaw in it. There’s really no proper control group. A proper scientific study needs to have demographic consistency between the groups being compared.

In this case, researchers looked at patients who had been treated for cardiac ailments. Some of those who were treated were given diagnostic procedures that involved ionizing radiation and some were not. The problem is that who received the radiation-based diagnostic procedures were those who doctors had greater concerns over or who had more severe cardiac conditions. These were persons who likely were in worse general health to begin with.

In general, the more severe and prolonged a person’s cardiac conditions are, the more imaging and diagnostics they will receive. A person who shows up at a hospital with a very mild heart murmur or a mild case of ventricular fibrillation is probably not going to get much in the way of diagnostic imaging of the heart. On the other hand, a person who has suffered multiple heart attacks and has a severely enlarged heart, angina and palpitations is going to receive quite a lot of diagnostic imaging.

The groups are therefore not comparable. The cardiac health of a patient is always going to be proportional to the amount of radiation they receive because worse cardiac health means more need for imaging. In all likelihood this will also mean that the patient is in worse overall health and has been under greater stress, possibly also taking more heart-related drugs and needing a variety of other procedures. They may also tend to be older, although the study claims to have compensated for this bias.

Even if this claim was true, it’s hardly newsworthy. A tiny increase in cancer risk from a lifesaving procedure does not alter the risk/reward balance by very much at all. In the case of this study the obvious bias is simply too great to even walk away with that conclusion.

…and yet it was reported like crazy anyway.


This entry was posted on Sunday, February 27th, 2011 at 10:46 pm and is filed under Bad Science, Good Science, Misc. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions



8 Responses to “Heart Scans Increase Risk Of Cancer – Or Maybe Not”

  1. 1
    John Galt Says:

    Or maybe it’s the two populations (“more scans” vs. “less scans”) have the same incidence of cancers, and the increased number of scans makes it more likely that they will be found…..


    Quote Comment
  2. 2
    Mark Says:

    I too have noticed many stories getting into the press. If you believe in LNT (and many think this is discredited, but few currently have a better idea), then a small dose of ionising radiation to the whole body will increase the risk of cancer by a very small amount. The point is that ICRP recognise this and their recommendations account for the risk, suggesting that the benefit of the medical exposure must outweigh the risks generated by it. I think this is what many of the articles are missing – the diagnostic procedure (or even therapy) procedure has some detriment but this is far outweighed by the benefits of finding a problem and then treating it. Taking it to the extreme, having a fluoroscopy examination to place a stent to open a blocked artery may (for example) increase your life-time baseline risk of cancer (from all causes) for a 40 year old from 44.9% to 50%. On the other hand, do not have the procedure and you might be dead by the end of the week.

    It is the ‘radiation’ word at play here – how many people really read the side effects from taking prescribed medicines. If they do, they accept them and move on – what if the side effect said ‘may increase your life time risk of cancer by 0.1%’ – some would worry, but most would not even both to read. However, if the drug (for whatever reason) contained ‘radiation’ and the same risk was expressed then there would be problems!

    That all said, if LNT holds and therefore ALARA (as low as reasonably achievable) holds, then I do have an issue with ‘life style’ CT scans (scans for the hell of it).


    Quote Comment
  3. 3
    Matte Says:

    Well, we could always support exploratory surgery instead of x-rays, PET or CT scans… I mean that would be the alternative, either that or walk away undiagnosed.

    People don’t understand how gentle a scan is compared to exploratory thorasic surgery (I am sure the actual surgeons don’t quite see it that way, but officially they will most likely agree with me here).

    I know what I would choose!


    Quote Comment
  4. 4
    DV82XL Says:

    I acquired an advanced copy of the paper in question, (Cancer risk related to low-does ionizing radiation from cardiac imaging in patients after acute myocardial infarction) when this story was first reported, from http://www.cmaj.ca. This is one of the worst examples of this type of study I have seen to date.

    First the data was gathered solely from administrative databases, which the researchers themselves admit contain no dosage data, thus all such values were estimated. They make no distinction between subjects high acute exposure, from older techniques, and repeated low exposure. They also admit that potential confounding variables were limited to whatever was recorded in these administrative databases which do not contain a medical history for each patient in the cohort. Thus information like smoking, a cause of both cancer and heart disease, was not corrected for in the final conclusions.

    However the most telling is the tabulated data itself. As usual, the confidence level reported was below 2-sigma, and graphed, the results show that the error bars almost always fall within the zero-risk values.

    This type of study is next to useless. It is bad science at every level, done by a group of undistinguished mediocrities, and has garnered more attention than it warrants due to a over zealous local media outlet.


    Quote Comment
  5. 5
    Michael Karnerfors Says:

    A reply to the study… quoting

    http://www.cmaj.ca/cgi/eletters/cmaj.100463v1

    The article “Cancer risk related to low-dose ionizing radiation from cardiac imaging in patients after acute myocardial infarction” by Eisenberg et. al(1) presents results of a study of the radiation risk of patients with myocardial infarction (MI) evaluated with x-ray studies that are provocative but difficult to understand. 82 861 patients were studied and 12,020 incident cancers were found which implied a 3% increase in the risk of age- and sex-adjusted cancer for every 10 mSv of low-dose ionizing radiation. Follow-up for incident cancers began one year after the index admission and it appears the evaluation continued for 5 years. Sensitivity analyses were performed to explore different time lags (one, three and five years) between exposure to low-dose ionizing radiation and ascertainment of incident cancer,

    First, there is considerable experience on the latency of cancer induction from radiation from such diverse areas as patients receiving radiation therapy to survivors from Hiroshima and the evidence indicates that the latency (i.e. time from exposure to incidence of cancer) for solid tumors is at least 5 years and more commonly 10 – 20 years2,3. Eisenberg et. al. observed cancer in the study patients within 1 year and even though the data didn’t appear to change with sensitivity analysis up to 5 years, even this time interval is shorter than the expected latency for cancer.

    Second, cancers were observed in patients who received cumulative doses less than 30 mSv and some at doses even less than 10 mSv. Prior studies, also done in Canada, on patients who received far higher doses as part of their treatment for TB did not develop cancer with cumulative doses up to at least 700 mSv. The study by Howe et. al4. found no radiation risk of developing lung cancer in patients who received at least 15 mSv per week for up to 3 years (for a cumulative dose over 2,000 mSv). A study by Miller et. al. on a similar patient population found no increases risk for breast cancer for cumulative doses less than about 700mSv (although they did find increased breast cancer for higher cumulative doses). In those studies the dose per examination was similar to the doses received in the study of Eisenberg et. al. but the total cumulative dose was at least 70 times higher without evidence of cancer.

    Third, most of the radiation these cardiac patients received was directed to the heart yet most of the cancers were extrathoracic. One might have expected a preponderance of thoracic cancers but this was not the case

    Finally, the experience of daily living makes it difficult to understand that such low doses could give rise to so many cancers in so short a time. The background radiation dose in Baltimore (where I live) is ~3mSv per year. The background radiation dose in Sante Fe, New Mexico is ~7mSv/year. In a little over four years the citizens of Sante Fe (of every age) receive a cumulative dose that is equivalent to that received by over ½ of the patients studied and yet there are no reports of increased cancer in Sante Fe (Actually the cancer rate in Sante Fe is 30% lower than that in Baltimore, but there are too many factors involved to understand this effect).

    This report claims a significant Radiation Risk, but it is based on tumorson tumors that occur too soon after exposure to radiation levels that are too small and with the majority of tumors arising in locations that are remote from the site of radiation exposure. These results are based on a statistical analysis of a very large database and are summarized in Fig. 1 where it appears that the Health RiskHazard Ratio (HR) is significantly increased only for “adjusted” data. There is no significant increase in Health RiskHazard Ratio for unadjusted data. We are told that “The following variables were adjusted for in the models: age, sex and exposure to low-dose ionizing radiation from noncardiac procedures”. We are not told of the nature of the adjustments, or the need for the adjustments. But the observation that Health RisksHazard Ratios only become significant after “adjustment” of the dagta prompts the concern that in fact the suggestion of increased radiation risk are is in fact an artifact of the statistical analysis.

    There is no question that high doses of radiation can cause cancer. There is much controversy as to whether radiation at the levels of diagnostic studies can also cause cancer. To suggest that radiation at levels not much different than the ambient background is a significant risk for cancer begs credulity and needlessly raises fears in the minds of patients and their doctors. To the extent that it mayit may keep some of these patients from receiving needed care it can cause great harm.


    Quote Comment
  6. 6
    Joffan Says:

    Michael, I was going to link to the same letter. I’ll just add that the letter is from Dr Reuben Mezrich, professor and chair of the Department of Diagnostic Radiology & Nuclear Medicine at the University of Maryland School of Medicine and chief of Diagnostic Radiology at the University of Maryland Medical Center.

    Basically the study is simply flawed. The letter describes the issues excellently – the 5-year limit was the clearest indication to me that in effect the results suffer from a divide-by-zero error.


    Quote Comment
  7. 7
    Calli Arcale Says:

    A more serious issue that gets very little press is that people who get a lot of scans are more likely to be DIAGNOSED with cancer — actually, a lot more likely, and this has nothing to do with radiation possibly causing cancer (and in fact includes non-irradiating tests such as MRI). The thing is, modern scans are so sensitive that they can detect quite tiny abnormalities, which can then be biopsied with the aid of fluoroscopy (which subjects the person to radiation, of course, but that’s not the point). All of this lets doctors diagnose cancers that could quite possibly have remained quietly inside the person, biding their time, until the patient died of something totally unrelated. At this point, doctors can very rarely tell whether a cancerous growth is going to grow aggressively or not, and patients quite reasonably tend to prefer the things be removed while they’re small and manageable. This increase in diagnostic sensitivity has been responsible for much of the increase in cancer diagnoses. (Not all of it. Another major factor is that people aren’t dying of other things as much, leading to them living long enough for the cancers to show themselves. And there may be a real increase as well; that can’t be ruled out at this point.)

    How many people get biopsies or even major surgery without having actually needed to? That’s something nobody knows right now, and that puts oncologists in a tricky situation even before having to worry about needless scans.

    Now, I do have concerns about the rise in diagnostic imaging. It’s not an unalloyed good, and the increased radiation risk isn’t something we should just dismiss as fearmongering. It’s not as bad as the general public tends to think of it, but it’s not totally benign either. Doctors are sworn to first do no harm, and that means they do have to at least think about it. But there are other forms of harm than the most direct, and I think more doctors need to consider those as I think they’re more significant. One is the risk of overdiagnosis — or misdiagnosis, as in detecting an anomaly that isn’t even cancerous but which triggers a biopsy. Biopsies are certainly not without risk, and they tend to be painful, regardless of what is being biopsied. The process is also expensive, and the system cannot bear the burden of an ever-increasing panel of tests; the tests may be reaching the point where the financial cost alone exceeds the potential benefit of detection, though that’s a difficult judgment call for a doctor to have to make.

    So there’s a real risk, but this study is completely ignoring it. It might have been more meaningful to look at the rate of accidents in these studies, because even if we ignore the radiation, these studies are certainly not without risks. You can actually kill a person in the cath lab if you screw up, so doctors do not order these tests casually — or at least, they shouldn’t. Mistakes are rare, but humans being humans (and anatomy being variable) I would suspect the run-of-the-mill adverse effects are far more significant than the cumulative radiation exposure.


    Quote Comment
  8. 8
    drbuzz0 Says:

            DV82XL said:

    I acquired an advanced copy of the paper in question, (Cancer risk related to low-does ionizing radiation from cardiac imaging in patients after acute myocardial infarction) when this story was first reported, from http://www.cmaj.ca. This is one of the worst examples of this type of study I have seen to date.

    Thank you. I did not even realize that the full study could be viewed online (many can’t be downloaded without paying a fee).

    I have just had the chance to read it and I concur with your thoughts on the quality of the study data and the conclusions. I also agree with the letter related by Michael Karnerfors.

    I am going to write my own letter to the editor when I get through with work today. This is something I don’t do often for these kind of journals, but given that the Canadian Medical Association is a very mainstream and generally credible publisher, I am deeply disappointed by the quality of this study.

    I will be letting them know that I find the publication of a study with such stretched results and poor data quality to be a poor reflection on the reputation of the journal and that it calls into question the their quality control and peer review process.


    Quote Comment

Leave a Reply

Current month ye@r day *

Please copy the string Qd8Nf3 to the field below:

Protected by WP Anti Spam