Are CT Scans More Dangerous Than We Thought?

[Tsu]

In the USA, MQSA regulations limit the maximum mean glandular dose to 3 mSv (300 mrem) per exposure. The majority of mammography units typically achieve about half this value. Digital mammography units, once regulators get a feel for how they behave and stop treating them like film/screen mammo units, i believe will be able to provide acceptable mammograms at even less exposure.

DR/CR is, no doubt, the future of radiology. But even so, aren't we still dealing with the extremely low KV ranges? 20-30? Can you convince me that those 'soft' rays are not CAUSING the very cancers we are trying to detect?

(Hey, guys! Sorry I hijacked this thread. We'll get back to CT in a minute. OK? I tried to start a thread about mammography about a year ago, but no one would touch it. This is a subject that has been eating at me for 20 years and it IS (somewhat ) relative to the topic...)

anyone who comes up with an imaging method that's as sensitive as x-ray mammography, specific as MRI, cheap and fast will probably have it made. very challenging task.

Are you up for it?

More to the point, how can this get pushed to the forefront in R&D? I've written to most of the major producers of mammography equipment trying to nudge them along, but...

BACK TO CT!

I'm really questioning that 1,300mr dose for an average CT scan. I thought it was about half that. In fact... Yeah! Pervect showed that on page one. I'll try to check my scanner tonight and get a few dose readouts.

Here's the interesting thing about CT dose. (Imabug, correct me if things have changed!) If you have a chest CT, you'll receive a dose of 500mr along the length of the chest scan. (NOW think about a dose of 300mr to each BREAST in an annual screening mammogram! - sorry - had to throw that in... ) If you have a chest AND abdomen CT at the same time, you'll still receive a dose of... 500mr! - but it's along the length of the chest and the abdomen. They aren't added together to get a dose of 1000mr. However, it is still cumulative. Two chest CT's in one week will yield a dose of 1000mr.

 

[imabug]

DR/CR is, no doubt, the future of radiology. But even so, aren't we still dealing with the extremely low KV ranges? 20-30? Can you convince me that those 'soft' rays are not CAUSING the very cancers we are trying to detect?

I ran across a paper a while back discussing the cancer induction risks associated with screening mammography. As I recall, they were relatively low (not insignificant though). However, I think the benefits of finding a cancer early via screening mammograms vastly outweighs the slight increase in risk of developing a cancer years later because of screening mammography. it will probably take some time to dig it up, but I'll see if I can find a reference.

BACK TO CT!

I'm really questioning that 1,300mr dose for an average CT scan. I thought it was about half that. In fact... Yeah! Pervect showed that on page one. I'll try to check my scanner tonight and get a few dose readouts.

CT doses vary significantly from machine to machine, but anywhere from 10-30 mGy (1-3 rad) is probably typical. Body CT doses can range anywhere from 3-4 rad for older single slice machines to less than 1 rad for newer machines with automatic technique adjustments (these have software that adjust the x-ray technique throughout the scan to compensate for changing body thickness). A lot of places are starting to modify their CT protocols to lower patient dose, particularly for pediatric studies.

Here's the interesting thing about CT dose. (Imabug, correct me if things have changed!) If you have a chest CT, you'll receive a dose of 500mr along the length of the chest scan. (NOW think about a dose of 300mr to each BREAST in an annual screening mammogram! - sorry - had to throw that in... ) If you have a chest AND abdomen CT at the same time, you'll still receive a dose of... 500mr! - but it's along the length of the chest and the abdomen. They aren't added together to get a dose of 1000mr. However, it is still cumulative. Two chest CT's in one week will yield a dose of 1000mr.

yes, this is correct, but only if the slices don't overlap. doses quoted for CT scans are typically dose/slice, but having a 40 slice CT scan doesn't mean you get 40 times the dose.

Not very intuitive, but obvious if you know what dose represents.

Dose (in rad/Gray) is a unit of absorbed energy measured in J/kg. The dose from 1 CT slice might be 1.3 rad. The dose from 2 CT slices is still 1.3 rad. You've deposited twice the energy (2 slices vs 1 slice), but at the same time the total mass you've irradiated has doubled. The factors of 2 cancel out.

However, if you irradiate the same volume of tissue twice, you've deposited twice the energy, but the total mass of tissue you've irradiated stays the same. Therefore dose adds up.

 

[Math Is Hard]

So, to those of you who may be doing research in the field or if you really want to help to make a difference - PLEASE FIND A WAY OF IMAGING THE BREASTS WITHOUT RADIATION! I have huge concerns that the very method we are using to detect breast cancer is also CAUSING some of those cancers. I don't have these concerns about CT scans as much because they are done in the 120KV range instead of the more damaging 20-30KV range. BUT - you DON'T want to go to one of those Body Imaging places in the local mall to have your annual screening CT scans. Most of the medical community are STRONGLY against those places. They are a source of needless radiation and the exams are generally useless.

I am glad you weighed in on this thread, Tsu. Your comments have really made me think. I can't even imagine how difficult it would be to do a study on the correlation of mammograms and incidence of breast cancer. Women who are getting annual (or more frequent) mammograms are going to be in a group that has a greater risk for the disease, so choosing groups for a study would be very problematic, I'd think.
Interesting that you mentioned those "Body Imaging" places. I hear them advertised on the radio frequently and I was very curious as to what traditional radiologists had to say about them.

 

[imabug]

Interesting that you mentioned those "Body Imaging" places. I hear them advertised on the radio frequently and I was very curious as to what traditional radiologists had to say about them.

FDA's http://www.fda.gov/cdrh/ct/
American College of Radiology's http://www.acr.org/s_acr/doc.asp?CID=2192&DID=16014

The American College of Radiology (ACR) recognizes that an increasing number of computed tomography (CT) screening examinations are being performed in the United States. Much CT screening is targeted at specific diseases, such as lung scanning for cancer in current and former smokers, coronary artery calcium scoring as a predictor of cardiac events and CT colonography (virtual colonoscopy) for colon cancer. Early data suggest that these targeted examinations may be clinically valid. Large, prospective, multicenter trials are currently under way or in the planning phase to evaluate whether these screening exams reduce the rate of mortality.

The ACR, at this time, does not believe there is sufficient evidence to justify recommending total body CT screening for patients with no symptoms or a family history suggesting disease. To date, there is no evidence that total body CT screening is cost efficient or effective in prolonging life. In addition, the ACR is concerned that this procedure will lead to the discovery of numerous findings that will not ultimately affect patients' health but will result in unnecessary follow-up examinations and treatments and significant wasted expense.

The ACR will continue to monitor scientific studies concerning these procedures.

 

[Math Is Hard]

Thank you, imabug.

In addition, the ACR is concerned that this procedure will lead to the discovery of numerous findings that will not ultimately affect patients' health but will result in unnecessary follow-up examinations and treatments and significant wasted expense.

Sounds like the advice is: Don't borrow trouble.

 

[russ_watters]

Unfortunately, Russ, I am not so sure this is true. The Medical "professionals" must justify and pay for expensive equipment Tsu's comments reinforce my believe that profits come before long term patient safety in the US medical system.

Well, like Tsunami said, I think that's more about money and CYA than risk/reward. But I was talking as much about the science that went into designing these machines and the approvals that had to be gotten before they could ever use them. Its not just the doctor ordering the test that you are putting your trust in. But still, doctors are human, so there are times when doing your own homework is a good idea. But a standard test done in a standard way doesn't ring my warning bell (and trust me - I'm a very distrustful person).

Even I have a story about a doctor who wasn't motivated by a patients' best interest: My grandfather at an aneurism of the aorta about 15 years ago. Standard treatment at the time was to crack the chest and wrap some dacron around it. A surgeon in Allentown (where it was diagnosed) wanted to perform the surgery. He'd done 5 - 3 successful. This wasn't good enough for my mom, who started asking questions and researching and found that the best place to have this done was in Dallas, by a surgeon who had done hundreds. So that's where they went and my grandfather is still active today at 88.

I know some doctors are pushing yearly full-body scans for profit, but AFAIK, few reputable doctors' groups/medical journals say that's a good idea (though probably at least as much because of the unnecessary expense). We may yet get to the point where your yearly physical includes some type of full-body scan, but we're not there yet. A Star Trek style tricorder that can non-invasively diagnose anything would be the holy grail of diagnostic tools.

 

[Loren Booda]

In graduate school we were once asked to determine a lower limit for proton decay halflife by considering the effect of such radiation on ova production!

 

[imabug]

In graduate school we were once asked to determine a lower limit for proton decay halflife by considering the effect of such radiation on ova production!

interesting, although I'm not quite seeing the connection. how did you go about solving the problem?

 

[Tsu]

Hi. Wow. Did the ER kick our butts last night! I checked my scanner dose - an average chest scan runs a dose of 456mGy.cm. An abdomen is around 620 while a pelvis is around 600-650. Heads are less than 100. So your dose for a mammogram is higher than for most CT work.

WONDERFUL!

edit: on the other hand, what is the conversion factor from rems to Grays? I'm actually going to have to finally LEARN the new Rad Safety Speak!? Why the hell did they change everything anyway? I know. It was just to mess with the heads of all of us old dogs, right??

 

[Loren Booda]

imabug,

I'm sorry. If it does have a legitimate answer, I have long since forgotten it.

 

[Tsu]

imabug,

I'm sorry. If it does have a legitimate answer, I have long since forgotten it.

Welcome to the Old Dogs Club, Loren!

 

[Loren Booda]

Tsu,

I'm even enrolled in obedience school, taught by my girlfriend's 14 year-old Cocker, Dyna.

 

[Tsu]

Good boy!

 

[pervect]

As I mentioned earlier, this is a population risk, not an individual risk. This means that in a population of 10,000 individuals, exposure at this level could be expected to result in as many as 8 additional deaths over what would normally be expected. It doesn't mean that your individual risk has increased by .08%

Are you a frequentist or something? If 10,000 people play russian roulette, and 1000 of them die, one can reasonably estimate that if one plays russian roulette, one's individual risk is 1 chance in 10 of dying.

[add]
The best thing I can in support of this position say is that if you have some reason to suspect that your radiation resistance is different from the rest of the population, your risk would be different from the population risk. But I can't see any reason one would believe that they were "radiation resistant" (or, for that matter, radiation prone).

Unless there is some reason to believe you are not representative of the population, population risk IS your risk.

 

[pervect]

If you're that concerned about radiation, stay away from the following:
Airplanes
Microwaves
Tobacco
Televisions
Smoke detectors
Lantern mantles
Building materials
Luminous watches and dials
Water
Rocks
Sunshine
Other people

This advice is extremely vague, and tends to divert the discussion from the actual risks, IMO (falsely downplaying them). Microwaves, for instance, aren't even ionizing radiation, so there is no radiation risk associated with using them.

One chance in 2000 of dying is significant. Think of your life as being worth at least $10 million dollars (perhaps more). You're talking about an insurance risk of say, $5,000 per exposure to the risk.

$10 mil isn't much, actually, when you consider how much the treatment for cancer will cost if you develop it, plus lost income & lost wages, even if you don't put any intrinsic value on your life (which would be distinctly odd).

 

[imabug]

Hi. Wow. Did the ER kick our butts last night! I checked my scanner dose - an average chest scan runs a dose of 456mGy.cm. An abdomen is around 620 while a pelvis is around 600-650. Heads are less than 100. So your dose for a mammogram is higher than for most CT work.

the numbers you quote are for a quantity called dose-length product (CT dose/slice multiplied by total scan length), which is a measure of the total energy deposited during the scan. AFAIK, there is little information yet as to how this number relates to risk. To me though, it's one of those bean-counting quantities used to keep track of someone's total radiation exposure. Useful, but not quite sure how useful.

The mean glandular dose (dose to glandular tissue in the breast) from mammograms is typically 1-3 mGy.

edit: on the other hand, what is the conversion factor from rems to Grays? I'm actually going to have to finally LEARN the new Rad Safety Speak!? Why the hell did they change everything anyway? I know. It was just to mess with the heads of all of us old dogs, right??

can of worms! can of worms!

convering from absorbed dose in gray (or rad) to dose equivalent/effective dose equivalent in sieverts (or rem) is a non-trivial task and involves many factors such as the type of radiation, duration and frequency of exposure and the types of organs exposed.

Different types of radiation (x/gamma, electron, proton, alpha) deposit energy at different rates along their track through matter (linear energy transfer). This gets bundled into a factor called the Quality Factor (QF) with x/gamma rays and slow electrons having a QF of 1 and alphas being much higher.

The same amount of each radiation also has a different effect on biological tissues, which gets bundled into a factor called relative biological effectiveness (RBE). For x/gamma and slow electrons, RBE is 1 while alphas have an RBE as high as 20.

Put those together with your absorbed dose (in gray) and you have a value known as Dose Equivalent (DE).

In addition, when it comes to radiation exposure in people, body parts and organs have different sensitivities to radiation. Organs such as the brain and skin are relatively radioresistant while other parts such as bone marrow and digestive tract lining are very radiosensitive. In general, any part that has rapid cell turn over will be the most radiosensitive. The radiosensitivity of the different organs gets bundled into a weighting factor (there is a list compiled by the NCRP in a document somewhere).

Apply this weighting factor to Dose Equivalent and you end up with a value known as Effective Dose Equivalent (EDE), which reflects the radiation risk associated with that exposure.

Lots of hand-wavy values that change over time as people do research and studies.

 

[Tsu]

<groan> oh, god!... I hate conversions...[/size]

So, have you got that breast MR unit built yet?

I still don't follow how the Duke dose calculator can show 400mr for an abdomen CT and 600mr for a mammogram while the mGy for these are 10-30 and 1-3 respectively. This does not compute. What are your comments on the 120 KV for CT exams vs. the 25KV for mamms? Over the years this has been a key point in my discussions with Radiologists and physicists. They have all agreed that 25 does far more cellular (and subcellular) damage than 120.

 

[imabug]

I still don't follow how the Duke dose calculator can show 400mr for an abdomen CT and 600mr for a mammogram while the mGy for these are 10-30 and 1-3 respectively. This does not compute.

The conversion factor between the SI (gray) unit and old (rad) unit of absorbed dose is 1 Gy = 100 rad. Similarly, 1 Sv = 100 mrem.

The absorbed dose delivered by most CT scanners is in the range of 1-3 rad (10-30 mGy). When you take into account the region being scanned and the radiosensitivities of the organs in the region, you will end up with another value in mrem (or mSv) reflecting the radiation risk from exposure. if the organs being scanned are radioresistant, then the calculated value in mrem (mSv) will be low. if the organs being scanned are more radiosensitive, then that value will be a higher.

Most people are taught that 1 Roengten (exposure) ~ 1 rad (absorbed dose) ~ 1 rem (biological risk), which is ok for a 1st order approximation. But to go from absorbed dose to biological risk accurately, you need to consider the region being exposed, the radiosensitivity of the organs being exposed and duration of exposure.

What are your comments on the 120 KV for CT exams vs. the 25KV for mamms? Over the years this has been a key point in my discussions with Radiologists and physicists. They have all agreed that 25 does far more cellular (and subcellular) damage than 120.

The reason low kVp is used for mammography is to get image contrast. This allows fatty tissue to be distinguished from glandular tissue, and soft tissue masses to be distinguished from the rest of the breast tissue. Using a higher kVp would result in less radiation dose, but less contrast and decreased visibility of soft tissue masses.

There is no question that the low kVp gets absorbed more and results in an increase in cancer risk. In most other diagnostic modalities, we are using aluminum and copper filters to remove low energy x-rays from the beam to reduce radiation dose to patients.

The real question is whether this risk of developing a future cancer a decade or two down the road due to mammography screening outweighs the risk of dying in a couple of years from an undetected cancer growing now.

 

[Lambert]

Bump ;-/

(Reuters) - Radiation from CT scans done in 2007 will cause 29,000 cancers and kill nearly 15,000 Americans, researchers said on Monday.

http://www.reuters.com/article/2009/12/14/us-cancer-radiation-idUSTRE5BD4VD20091214


Diagnostic tests after a heart attack linked to increased cancer risk

http://www.theglobeandmail.com/life/health/scans-boost-cancer-risk-for-heart-attack-survivors-study-finds/article1897387/

 

[ZapperZ]

Wow. This must break some record in terms pf necroposting.

Zz.