Primer on Safety of Imaging Studies: There may be scarier things under the imaging bed than GDD

July 6, 2018



Many sufferers have called for informed consent prior to administering GBCAs for MRI studies. This may not be an unreasonable request at all. In my opinion, at the least the patient should leave with an information sheet, as happens after surgical procedures, which describes adverse events, with a focus on the symptoms of GDD (1,2), and to call the imaging facility if they have problems. Ideally in the near future I would like MRI facilities equipped with chelating agents so the patient can return to the center and be chelated for the Gd. Essentially it is not much different than the original GBCA administration in the first place.


Some sufferers vigorously claim that all GBCAs should be taken off the market, and maybe some form of penalty to all involved with the administration, which I don't believe in either.


It would be fair to say that informed consent should be performed whenever a certain level of risk is present for whatever is being done. I have frequently referenced a paper in the New England Journal of Medicine (3), which describes a large Veterans Administration Investigation and Report in which they recommend that information about any event that carries with it a risk of a significant adverse event greater than 1 in 10,000 individuals should be provided to the individual. I have used this as the baseline for my recommendations, dating back to 2007.


I suspect that most GDD sufferers would not also have studied the adverse events or frequencies of other imaging modalities. In my opinion a level playing field is essential to understand the fullness of risks, not only for MRI but for other imaging modalities, for both referring physicians and patients, to decide what they should undergo based on risks, benefits, and cost. Accuracy of the test must always be factored in - if it is not accurate then why undergo a test no matter how safe it is.


I will provide a short primer on those considerations. I do not believe that many observers who have been askance at my writing about Gadolinium Deposition Disease realize that I have written about 380 papers with at least 18,000 citations, and of them perhaps 320 papers with 13,000 citations (an estimate)  on the value of GBCAs in imaging studies - perhaps as many as anyone (4). They also do not likely realize that I was the first major academic radiologist to write about safety aspects in CT with ionizing radiation, and on the subject of ionizing radiation since probably 1980, when I wrote our first peer-reviewed work that appeared in JMRI in 2007 (5). I would like to acknowledge that Eugenio Picano MD, who has become a good friend, who as an echocardiologist wrote about radiation risk about 1 year earlier than my 2007 paper, although at the time I was not aware of his work.


Radiation Safety.  We identified 4 issues with the subject of safety and CT: do only studies that are necessary, chose safer alternatives when appropriate, reduce radiation doses, and patient information. Since we wrote this paper, radiation doses have been lowered and tailored in CT, especially important reduced radiation doses for children, new CT machines constructed that delivered lower and safer doses of radiation, and more effort spent by radiologists to select safer imaging modalities. Another important paper we wrote as follow up was entitled the Information Imperative (6). In that paper we focused on informed consent for patients prior to CT imaging.

In very brief, radiation is often measured in milliSieverts (mSv). Radiation societies, perhaps foremost the National Academy of Sciences Biological Effects of Imaging Radiation (BEIR) VII report, which described a 1 in 1,000 chance of developing cancer from 10 mSv of radiation (7). This was the standard radiation delivered for an Abdomen-pelvis CT scan at the time. Since then, into the present time, considerable work has been performed to reduce radiation exposure for most imaging studies. So many studies which in 2007 exposed patients to 10 mSv of radiation, at the present time expose them to 5 mSv or lower.

I have considered that 1 mSv should be used as a cut-off for informing patients about risk from cancer from x-rays. Studies with lower doses (e.g.: chest x-ray) not requiring informed consent, and studies with higher dose (CT, PET-CT, nuclear medicine) requiring informed consent.

This risk relates to all forms of medical imaging that use x-rays - from tomography, to fluoroscopy, to interventional radiology, to cardiac fluoroscopic procedures (e.g.: coronary angiography and angioplasty).

       In the years 2012-2013 large nation-wide studies have been written from UK and the Australia defining risk of cancer from CT, the UK study focusing on children (8,9). They found that the use of CT scans in children to deliver cumulative doses of about 50 mGy might almost triple the risk of leukaemia and doses of about 60 mGy might triple the risk of brain cancer (8). Also, they found that the overall cancer incidence was 24% greater for CT radiation exposed than for unexposed people, after accounting for age, sex, and year of birth (9).

       So, 10 mSv results in a 1 in 1,000 risk of cancer. Amy Bennington Gonzalez and her team estimated that based on the number of CTs performed in the US in 2007 alone, 29,000 cancers would develop in the future. Double-stranded DNA breaks is the defect most often described with radiation, and theoretically these should be detectable in cancers that develop from radiation. To date though, radiation-induced cancers have been indistinguishable from standard cancers, so I have termed medical radiation a silent-killer, as opposed to gadolinium issues below.


Contrast Media Induced Acute Kidney Injury (CI-AKI) formerly called Contrast Induced Nephropathy (CIN). If one thinks that risks and causes of NSF or GDD are challenging and uncertain, CI-AKI has been recognized since the 1950s and to the present time it remains uncertain why this happens. Also confusing, the diagnosis refers to a transient increase in serum Creatinine (sCr) level, but the negative consequences are really only important if it is long-lasting decrease in renal function. CI-AKI has been largely associated with the use of Iodine contrast, used in CT, cardiac studies, angiography, and intravenous pyelography (the last rarely used now, imaging of the kidneys, largely supplanted by CT). Only recently, after maybe 60 years, has the qualification been used of 'persistent' CI-AKI (the only important circumstance), as opposed to 'transient'. For many years CI-AKI has been considered the fourth most common cause of renal failure in hospitalized patients. Even more troubling, and few papers on this, the conversion from stage 3 renal failure, effective glomerular filtration rate (eGFR > 60 mL/min/1.73 m2, basically doing ok) and stage 5 renal failure (eGFR < 15 mL/min/1.73 m2, requiring hemodialysis or renal transplant) was associated with a 1 year 40% mortality, which is in the range of the mortality from NSF.  Most recently, skeptics (some CT users) feel that maybe CI-AKI does not exist at all, since a couple of studies have shown changes in sCr in patients after CT who did not receive contrast. The problem with those studies is that patients who do not receive contrast are generally sicker patients who also have poor renal function; so true patient-matched studies do not exist for noncontrast and contrast-enhanced CT studies. 

     What common sense should tell us is that AI-AKI does exist, but it is not likely as bad as reports from the 1970s to 2000s suggested. Perhaps recommendation to administer contrast to patients with renal function eGFR > 30 mL/min/1.73 m2 (previously limited to not giving iodine contrast to patients with eGFR <60 mL/min/1.73 m2) does not seem unreasonable. What is also true, and a positive step, is that for perhaps at least 1 decade imaging centers with newer CT equipment have given just 75 ml of contrast, instead of 150 ml which was the standard in earlier years. The saying attributed to Paracelsus that 'the dose makes the poison' applies to contrast media. Also High Osmolar Contrast Medium (HCOM), used primarily earlier than 1990 also has a much higher association than the Iodine contrast currently used, termed low osmolar (LOCM, with one iso-osmolar agent). LOCM largely replaced HOCM in the late 1980's.

The reason that CI-AKI is not seen with GBCAs is largely because the volume of GBCA used for MRI is much lower than the volume used of iodine contrast media with CT.

Be aware of the risk of CI-AKI with the use of iodine contrast agents, notably with CT. A simple preventive measure is for patients to be well hydrated (not thirsty) before receiving iodine contrast.  As with medical radiation and CT, there is at present no kidney markers for contrast as the cause of the injury. So as with medical radiation, iodine contrast is the silent assailant and at times becoming a silent killer in potential patient harm.


Nephrogenic Systemic Fibrosis (NSF). I have discussed NSF earlier. Basically,  it only arises in patients with poor renal function, primarily eGFR < 15 mL/min/1.73 m2, who had received one of the weaker gadolinium chelates: Omniscan, Optimark, and Magnevist. With patients having poor renal function largely not receiving GBCA any more (stable macrocyclics could still be used) the disease disappeared in 2009, and essentially disappeared as early as late 2007 when warnings of the risk were widely distributed.

Be aware of NSF if your renal function is eGFR <15 mL/min/1.73 m2 and a linear chelate is being offered to you.


Gadolinium Deposition Disease (GDD). My major focus in blogs. Any agent, any level of renal function. Rare.

Be aware of this entity. If you have symptoms of GDD after a GBCA-enhanced MRI study, do not undergo GBCA administration again. Unlike medical radiation and CI-AKI, which are silent as causes for injury, GDD is the opposite: it is loud after the exposure, and has markers for the injury with distinctive features. However it is important to note that the loudness of GDD does not necessarily make it worse than the silent killers, infact the silent killers may long term be worse (ie: they are killers). This is something I am certain the great majority of sufferers do not appreciate.


Ultrahigh Field MRI (7T). HIgh field MRI (1.5T), which is the most common field strength used for clinical imaging in the US appears to be safe, regarding cancer induction and general severe disease development. At present I am not sure the same can be said about ultrahigh field MRI (5-7T). Patients often report dizziness when walking too fast in the region of the scanner, and heating of metal objects within the body is greater. I am not sure we know enough about how safe ultrahigh field strength MRI is.


MR contrast agents not using Gadolinium. Both iron-based and Manganese based contrast agents have been developed and FDA approved, but abandoned some years back. With issues related to gadolinium toxicity, revisiting of these atoms for contrast agents, and other atoms, are being considered. One has to be clear thinking about other agents- what could their safety features be? Gadolinium may still be better using the 'devil you know vs. the devil you don't know' mental calculation. Non-gadolinium MR contrast agents too have, and will be shown to have more yet unknown, adverse events associated with them. It had not been until atleast 100 million GBCA administrations had been made until the risks began to be more fully appreciated. Many of the alternatives may have been administered in 100,000 patients at the very most, and maybe 100 administrations in some. … It would be folly to blindly believe using new agents may be safer without realizing these numbers.


Severe acute hypersensitivity reaction. Occurs with any administered contrast agent, iodine contrast and gadolinium contrast agents. Fatal acute reactions to iodine contrast estimated at 1 in 150,000 (may be 1 in 300,000). Fatal acute reactions to gadolinium contrast estimated at 1 in 300,000 (may be 1 in 1,000,000). Does dose explain the difference in fatality between iodine and gadolinium contrast, with a much lower dose administered with gadolinium agents (10-20 mls with gadolinium contrast vs. 75-150 ml with iodine contrast)? Maybe, I think so.


Ultrasound. Ultrasound may be the safest of the imaging modalities. The problem is, it is also for many applications the least accurate. Microbubble contrast used in ultrasound introduces the issue now of safety of air being present in blood vessels. Stroke and heart attack have been reported but appear to be rare.


Nuclear Medicine/ PET / PET CT. President Obama famously said that worrying about 1 country (I won't say which one, as I have enough trouble with the radiology community) keeps him up at night. I have written a fair amount about CT, safety, overuse, and a fair amount about MRI regarding NSF and GDD, but I have never written peer-reviewed articles about Nuclear Medicine (I have on PET CT). Yet what keeps me up at night is worrying about radiotracers, PET radioisotopes, and couple CT to that. I am very nervous about papers that describe using repeat PET CT imaging, especially in children and especially with benign disease. The doses of radiation of some of these radiotracers is truly enormous, with Gallium scans and Thalium scans easily exceeding 50 mSv. But how is this measured, as the tracers’ remains in the body for some time continuing to emit radiation?  One should also read how PET tracers generate images – that will make you frightened. The one important aspect of radiotracers in all these studies is that the volume administered is very low, so we can rely on Paracelsus observation above, to feel there is some safety. If CT now has administered 75 ml of contrast, MRI 15 ml of contrast, typical volumes of radiotracers are in the 1 -2 ml range. This is their saving grace. The radiotracers emit radiation in the patient's body, as the source of the images, and this radiation stays with them for some time. I had edited an on-line radiology issue on safety on imaging a few years ago. I could find CT radiologists to write on CT risks, MR radiologists to write on MRI risks, ultrasound radiologists to write on ultrasound risks. I could not find any nuclear medicine physician around the globe to write on nuclear medicine/ PET safety and risks.

There are important roles for PET imaging and new exciting tracers in development. I would cautiously use this technique in children, for benign disease, and when multiple repeat studies are recommended. For many studies the radiation exposure is much higher than with CT.


Summary. There are safety issues with essentially everything in Radiology, as with Medicine in general, and there are safety issues with even over the counter drugs. If one decries MRI and gadolinium, one has to be aware of the alternatives, and many may be either worse or much worse. Informed consent is a reasonable discussion with GBCAs, but this must then be applied to all other imaging that exceeds for example what I described as the 1 in 10,000 chance of serious adverse event. In my own career, when patients ask about safety and related issues about MR, CT, ultrasound, PET, etc., I can describe brief comments as I have done here, but what underlie this are volumes of knowledge. Unfortunately one also encounters the situation (frequently) that physicians or other health care workers who represent that they have expertise in a subject, like gadolinium safety, may only have cursory knowledge; the same is true for iodine safety, CT safety, etc. So I can't tell any one person 5000 pages of knowledge on a subject of comparative safety, so what I have done for the last 10 years is tell them "this is what I would do if I were in your situation", and it often is no imaging at all. Unfortunately it has only been for about 3 years that I have been fully aware of GDD, so prior to that I did believe GBCAs were safe in all patients with normal renal function... but they are not - a small subpopulation of patients will get very sick from GBCAs.



1. Gadolinium deposition disease: Initial description of a disease that has been around for a while. Semelka RC, Ramalho J, Vakharia A, AlObaidy M, Burke LM, Jay M, Ramalho M. Magn Reson Imaging. 2016 Dec;34(10):1383-1390.


2. Gadolinium Retention and Toxicity-An Update. Ramalho M, Ramalho J, Burke LM, Semelka RC. Adv Chronic Kidney Dis. 2017 May;24(3):138-146.


3. The disclosure dilemma--large-scale adverse events. Dudzinski DM, Hébert PC, Foglia MB, Gallagher TH. N Engl J Med. 2010 Sep 2;363(10):978-86.




5. Imaging strategies to reduce the risk of radiation in CT studies, including selective substitution with MRI. Semelka RC, Armao DM, Elias J Jr, Huda W. J Magn Reson Imaging. 2007 May;25(5):900-9.


6. The information imperative: is it time for an informed consent process explaining the risks of medical radiation? Semelka RC, Armao DM, Elias J Jr, Picano E. Radiology. 2012 Jan;262(1):15-8.




8. Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Pearce MS, Salotti JA, Little MP, McHugh K, Lee C, Kim KP, Howe NL, Ronckers CM, Rajaraman P, Sir Craft AW, Parker L, Berrington de González A. Lancet. 2012 Aug 4;380(9840):499-505.


9. Cancer risk in 680,000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians. Mathews JD, Forsythe AV, Brady Z, Butler MW, Goergen SK, Byrnes GB, Giles GG, Wallace AB, Anderson PR, Guiver TA, McGale P, Cain TM, Dowty JG, Bickerstaffe AC, Darby SC. BMJ. 2013 May 21;346:f2360.


Please reload

Our Recent Posts

Please reload


Please reload