low dose radiation and subsequent increased cancer

​​A recent article found a significant increase in cancers, years after exposure to low-dose ionizing radiation in adults with congenital heart disease (see radiation exp kids and malignancy circ2018 in dropbox, or DOI: 10.1161/CIRCULATIONAHA.117.029138)

Details:

--24,833 adults with congenital heart disease (CHD) in the Quebec Congenital Heart Disease Database, aged 18-64, from 1995-2009, with 250,791 person-years of follow-up. Those with genetic disorders were excluded

--12% severe CHD (cyanosis at birth, or requiring early surgical intervention (such as tetralogy of Fallot, truncus arteriosus, transposition complex, univentricular heart), 62% shunt (ASF, VSD, PDA, coarc); 25% valvular abnormalities

--median age 54, 44% male, comorbidities: 15% heart failure/15% atrial arrhythmias/8% CAD/6% PAD/10% DM/33% hypertension/15% obesity/8% hyperlipidemia/12% CHD surgery/20% all cardiac surgery

--patients with cancer were older and had more cardiac comorbidities

--low-dose ionizing radiation (LDIR) exposure: esp catheterizations, CT scans, nuclear procedures, rhythm procedures (regular xrays not included since cumulative LDIR doses were small)

--718 patients had higher LDIR exposure (>5 procedures) and 20,729 had <2

--nested case-control sample: each cancer case was matched with 4 randomly selected controls, matched by sex, CHD severity, year of birth, and age at cancer diagnosis (this matching should eliminate temporal trends in doses of radiation per procedure and management).

    --people who developed cancer during follow-up were treated as controls until the cancer occurred

--dose of radiation exposure imputed by presumed typical effective dose in the literature

--cases vs controls were adjusted for age, sex, year of birth, CHD severity, comorbidities (hypertension, diabetes, hyperlipidemia, obesity, heart failure, atrial arrhythmia, CAD, CKD, PAD, pulmonary hypertension, history of stroke, history of endocarditis)

Results:

--602 cancer cases observed

--patients with cancer had more LDIR-related cardiac procedures: 5.5% of patients with cancer were highly exposed  (>5 procedures) vs 2.8% in the noncancer group; and 74.3% of cancer patients had <2​ procedures vs 83.7% of noncancer ones (p<0.0001)

--cumulative incidence of cancer up to age 64: 15.3% (14.2-16.5%)

--in women, breast cancer was most frequent (34.5% of cancers). another 50% were: respiratory (14%), GU (13%), GI (12%) and hematologic (11%)

--in men, GU in 31%, GI 23%, hematologic 15%

--cancer incidence increased with age: 52% of total in those 55-64 yo; median age of cancer diagnosis was 55.4

--women had more cancer: lifetime risk of 16.1% in women and  14.4% in men

--cases vs controls:

    -LDIR-related cardiac procedures: 1410 vs 921 per 1000 adult patients (p<0.0001)

    --cumulative LDIR exposure was independently associated with cancer: per procedure, OR 1.08 (1.04-1.13)

    --by using dose estimates for LDIR exposure: per procedure, OR 1.10 (1.05-1.15), with possible dose-related response

    --after excluding most smoking-related cancer cases (respiratory, head and neck, bladder):

        --OR 1.10 per procedure (1.05-1.16)

        --in those with >5 procedures, OR 3.08 (1.77-5.37)

        --after a 3-year lag period (to avoid including procedures related to a diagnostic workup of cancer)

            --OR 1.09 per procedure (1.03-1.14)

            --OR 2.58 (1.43-4.69) when exposed to >5 procedures

Commentary:

--overall, patients with CHD are doing better than ever, but at the expense of higher levels of LDIR for their procedures, with estimates of the collective dose of radiation increasing 7-fold from 1980 to 2006. and, of course, the increased longevity of these patients will increase their likelihood of getting cancer in the longrun...

--there is a 1.6-2 times higher cancer prevalence in males vs female adults with CHD

--Quebec’s health care system has much better longitudinal data than we have in the US: there is a uniform registry of all individuals at birth (all get a unique healthcare number) which is linked to all medical services, including all diagnoses, demographics, hospitalizations, inpatient/outpatient diagnostic and therapeutic procedures [if only we had a coherent system of care….]

-- CHD severity was based on anatomic diagnoses, so should be similar in cases vs controls

--this cohort remained in the study for max of 15 years. likely many more cancers would found if followed patients to older ages, or started at younger age

--also, the increased risk in the above CHD study was found even in those receiving <15mSv of radiation exposure

--another Canadian database study of 82,861 pateints who had acute MI, found that 77% had at least one cardiac imaging/therapeutic procedure with LDIR, and there was a dose-dependent relation begween exposure to radiation and subsequent cancer risk (3% increase in age- and sex-adjusted cancer risk per 10 mSv of LDIR esposure over 5 years). see Eisenberg MJ. CMAJ 2011; 183(4): 430

--there have been many blogs over the years on the potential carcinogenic effects of ionizing radiation:

    --see http://gmodestmedblogs.blogspot.com/search/label/radiation for many, many prior blogs on radiation exposure

    --see http://gmodestmedblogs.blogspot.com/2015/03/coronary-angiography-or-exercise.htmlfor the huge radiation exposure by coronary angiography in those with chronic angina

    --see http://gmodestmedblogs.blogspot.com/2016/03/coronary-artery-calcium-scores-from.html which includes one of many studies showing that coronary artery calcium scores are really good predictors of CAD, but (i think) their utility is severely undercut by the need for ionizing radiation

    --see http://gmodestmedblogs.blogspot.com/2014/10/lung-cancer-screening.html for one of many blogs critical of the aggressive approach of the USPSTF on low-dose CT screening in smokers (noting that the actual dose received by patients averages that of high-dose CT scanning, when one adds in the subsequent PET scans/regular CTs done for the vast number of false positive screens; and also speculates that the cancer risk in these patients may be much more than in the average people, from which cancer risk estimates are based, since those with long smoking histories are more likely to have distorted lung architecture/local immunity/etc which might put them at higher risk of cancer than those with underlying normal lungs)

    --see http://gmodestmedblogs.blogspot.com/2013/07/ct-scanning-in-kids-and-radiation.html which highlights the dramatic radiation exposure risks of kids, which may well lead to increased cancers over time (and they potentially live lots longer...)  And, in terms of kids, there are data linkage studies suggesting that of 680,211 exposed to CT scans, there was a 24% increased cancer risk, with a dose-response curve, and more cancer in those exposed at younger ages (see Mathews JD. BMJ. 2013; 346: f2360)

--there are clear limitations to the above CHD study: this assessed a relatively short window of the life of these patients (so, there could be a very different cancer incidence found if included kids and followed longer); this included only CHD patients (and some may have had undiagnosed genetic conditions, for example, that might either increase or decrease their cancer risk); the dose of radiation is guessed at based on historic data (and may not apply to some of these patients, either because different techniques/xray equipment were used, their effective dose may be different depending on their weight/body habitus; even important cofactors such as tobacco/alcohol were not incorporated in the adjusted model of risk). also, was there a fundamental difference, for example, in those patients with severe CHD who required lots of LDIR exposure vs those who did not?

so, i bring up these studies because of the increasingly common use of higher-dose imaging studies with ionizing radiation (though the dose per procedure has gotten lower with improved technology). one positive trend, i think, is the increased recommended interval of mammography screening to every 2 years.  this is counterbalanced by the huge potential increase in radiation by the "low-dose" CT scanning in smokers. And, nary a patient goes to the ED who does not get a CT.  

the issue to me seems to be the effect of the medical culture on both clinicians/patients in this increasingly tech-driven field, ​where CT/PET scans etc are being ordered/requested at increasing rates. the above article for CHD patients and the slew of prior blogs on excess radiation/cancer risks is just an against-the-tide reminder that ionizing radiation has very real long-term risks. As noted in the CT scan blogs for smokers, for example, the mathematical modeling is that we are creating 1 lung cancer in 2500 scans (and maybe a lot more, given the underlying abnormal lung tissue). similar numbers have applied to mammography screening (again, perhaps the cancer risk is higher in women with abnormal breasts, who might well get increased numbers of mammograms, special views, etc). see mammog increasing breast cancer AIM2016 in dropbox or Miglioretti DL. Ann Intern Med 2016; 164(4): 205, for example

i have found that many patients are convinced to forego xrays when i mentioned the real risk of developing cancer [all of this does promote using MRIs, which do not have ionizing radiation, as a better test when appropriate, though that may entail more time spent doing prior approvals.....]

--there are clear limitations to the above CHD study: this assessed a relatively short window of the life of these patients (so, there could be a very different cancer incidence found if included kids and followed longer); this included only CHD patients (and some may have had undiagnosed genetic conditions, for example, that might either increase or decrease their cancer risk); the dose of radiation is guessed at based on historic data (and may not apply to some of these patients, either because different techniques/xray equipment were used, their effective dose may be different depending on their weight/body habitus; even important cofactors such as tobacco/alcohol were not incorporated in the adjusted model of risk). also, was their a fundamental difference, for example, in those patients with severe CHD who required lots of LDIR exposure vs those who did not?

so, i bring up these studies because of the increasingly common use of higher-dose imaging studies using ionizing radiation (though the dose per procedure has gotten lower with improved technology). one positive trend, i think, is the increased recommended interval of mammography screening to every 2 years.  this is counterbalanced by the huge potential increase in radiation by the "low-dose" CT scanning in smokers. And, nary a patient goes to the ED who does not get a CT.  

the issue to me seems to be the effect of the medical culture on both clinicians/patients in this increasingly tech-driven field, ​where CT/PET scans etc are being ordered/requested at increasing rates. the above article for CHD patients and the prior blogs on excess radiation/cancer risks is just an against-the-tide reminder that ionizing radiation has very real long-term risks. As noted in the CT scan blogs for smokers, for example, the mathematical modeling suggests that we are creating 1 lung cancer in 2500 scans (and maybe a lot more, by giving radiation to the underlying abnormal lung tissue). similar numbers have applied to mammography screening (again, perhaps the cancer risk is higher in women with abnormal breasts, who also might get even more radiation from increased numbers of mammograms, special views, etc)

i have found that many patients are convinced to forego xrays when i mentioned the real risk of developing cancer.

[all of this does promote using MRIs, which do not have ionizing radiation, as a better test when appropriate, though that may entail more time spent doing prior approvals.....]

---------------------------------------------------------------------------------- 

if you would like to be added to the regular mailing list for blogs, please email me at gmodest@uphams.org

For access to older blogs:

to get access to blogs since 8/15/17:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list​