■ site map
BLOG: June 2010 - December 2013
II - Mammography
17. Mammography radiation exposure
Ever since the beginning of X-ray mammography screening, decades ago, it was assumed and publicly asserted by both, official medicine and the government, that the health risk from screening radiation is negligibly low. Very few voices back then warned that there is no sufficient evidence to support such assumption. They were effectively sidelined and neglected. Do we have enough evidence today to resolve this question? Probably not to some high level of accuracy, but sufficiently for a meaningful answer.
Unlike the average figures thrown around, actual exposures to radiation in the standard X-ray mammography screening can vary significantly, depending on the intensity of radiation output (so called skin entrance exposure, varying with breast thickness, usually between 0.5 and 5 roentgen), thickness of the aluminum filter, and glandular fraction, representing the weight fraction of glandular (sensitive to radiation) vs. adipose (fatty) tissue.
Table below shows radiation
delivered to breast tissue at near minimum, median and near
maximum levels (based on Duke University's online
TABLE 1: radiation dose delivered by X-ray mammography (breasts radiation dose equals per view dose multiplied by number of views per breast)
Note that the calculator gives the absorbed dose in millirads (mrad) per roentgen; to convert to milliSieverts (mSv), i.e. equivalent dose, 1mSv=100mrad, so dose in mSv equals the dose in mrad divided by 100. Breasts radiation dose is a product of the absorbed radiation per view and number of views (it does not get multiplied by two for two breasts, because the radiation risk assumes dose delivered to both breasts).
The main variable is the breast thickness, on which is based radiation output, but the level of radiation delivered also varies significantly with the value of glandular fraction (from near zero for "fatty", or "fibrous" breast to near 1 for "dense" breast; the more of glandular tissue, the lower its irradiation for given output level), and the beam "hardness" , expressed as half-value layer of aluminum. Peak voltage is kept constant, at 28 kV (the voltage is also a variable in actual setups, but its effect on irradiation dose is relatively small within the usual 22-35 kVp range).
This technical calculation only approximates the actual irradiation level - which can vary unpredictably and significantly with the specific machine, operator and breast - but does give a usable general estimate.
For the common 2-view screening procedure, it gives the range of possible exposures anywhere from 2 mSv to 15 mSv absorbed/equivalent dose. For 2+2 views, which can be either a single procedure with two extra views, or additional 2-view procedure, the radiation dose is doubled, ranging from 4 mSv to 30 mSv.
Note that the actual maximum could be still higher, since the compressed breast thickness in this consideration is limited to 8 cm.
The magnitude of mammography radiation dose is often illustrated by a comparison with the ordinary chest X-ray dose. But it may not be neither as simple, nor accurate, as one would imagine. Commonly cited effective doses for chest X-ray are 0.02 mSv (posteroanterior, or PA) and 0.1 mSv (posteroanterior and lateral, or PA/LAT). Reported range for the former is 0.007-0.050 mSv, and for the later 0.05-0.24 mSv (Mettler et al. Effective Doses in Radiology and Diagnostic Nuclear Medicine, 2008).
However, according to the U.S. Office of Research Services, the effective dose for the PA/LAT chest X-ray is 0.40 mSv - nearly double the highest reported doses, and four times higher than what is commonly cited (http://drs.ors.od.nih.gov/services/rsc/tables/CXR-PA-LAT.xls ).
Who's got it right? Where does the difference come from? First glance at the agency's list of irradiated organs suggests the answer: among irradiated organs during chest X-rays, it lists as many as 18 organs directly (including a few that receive near-zero radiation) plus 10 more within "reminder" tissues/organs. Lungs actually receive
less than one sixth of the total dose,
and by far the largest dose goes to - guess where - the breasts: 44% of the total effective dose, or over three times more than to the lungs.
The most likely scenario is that those other, commonly cited figures originate from estimates including only a partial list of organs irradiated during the conventional chest X-ray exam. And if the correct figure for the PA/LAT test is four times higher than the commonly cited value, it is to expect the same for the PA chest X-rays as well, which would put their average effective dose closer to 0.1 mSv.
If you find such sloppiness in assessing and presenting radiation risk of such a common procedure surprising and unacceptable, you are in for more frustration. According to a recent study focused on cardiac diagnostic imaging, wide discrepancies in dose estimates from one study to another, and form manufacturers' vs. those based on ICRP (International Committee on Radiation Protection) standards are rather common.
Part of it is due to the incomplete list of irradiated organs and tissues serving as the basis for estimates; also, some still use outdated methods for radiation dosimetry (for instance, effective radiation dose for commonly used radio-isotopes in gamma-ray imaging, like Technecium99, is routinely based on methodologies dating back to 1975, even 1968, in addition to be limited to only a handful of affected organs/tissues). On top of that, plain errors due to insufficient knowledge about radiation dosimetry are fairly frequent.
This little digression was needed to bring to reader's attention the fact that the present day radiation risk assessment and presentation in medical imaging is, for these and other reasons, often
inconsistent, unreliable and not seldom erroneous.
Back to mammography vs. chest X-ray radiation risk. Taking the 0.4 mSv effective dose figure for PA/LAT chest X-ray, and 0.72 mSv effective dose for the average mammography session (from 3 mSv absorbed dose per view multiplied by 2 for two views, and by 0.12 breast tissue weighting factor), the latter delivers a nearly double the effective dose delivered by PA/LAT chest X-rays.
With nearly half of the effective dose with the chest X-ray procedure going to the breasts, average breast cancer risk of the standard 2-view mammography is about
four times higher than breast cancer risk with PA/LAT chest X-rays.
Assuming the actual effective dose for the common posteroanterior (PA) chest X-ray higher than the routinely cited 0.02 mSv figure by the same ratio of 4 as for PA/LAT, it comes to 0.08 mSv total effective dose, and nearly 0.04 mSv for the breasts alone. Thus the average breast cancer risk with the standard 2-view mammography is about
20 times higher than with the standard PA chest X-ray.
accuracy of the radiation weighting factor, officially set at 1
for the wide range of radiation energy levels from low-energy
X-rays to high-energy gamma-rays, the actual risk with
mammography is probably somewhat higher. The reason is that the
standard chest X-rays are performed at a peak voltage exceeding
The lower-energy X-rays have more interactions with tissue molecules, thus causing more damage; assuming, very approximately, that their actual effective dose is 50% higher, puts the average risk of developing breast cancer with standard 2-view mammography at about 9 times higher than that for any cancer in the chest cavity (from 0.72 vs. 0.08 mSv effective dose), and about 40 times higher than for developing breast cancer due to a single regular PA chest X-rays (from 0.72 vs. near 0.04 mSv).
In other words,
a single average
mammography procedure a year would pose
Obviously, it cannot be assumed neither harmless nor insignificant, particularly for women with above average vulnerability to ionizing radiation, undergoing regular longer-term screening.
Evidently, thicker breasts are exposed to higher levels of glandular irradiation (while the average compressed breast thickness in the U.S. is 4-5cm, those of up to 12cm, or more, are not uncommon). A woman with thick breasts could be irradiated with a dose more than twice higher than 0.72 mSv in a single 2-view mammography session, and double that in a single 4-view session. That is comparable to the risk from
over 3 and nearly 7 standard chest X-rays a month, respectively, for a year,
assuming annual mammography screening.
Damage to the cells exposed to X-ray irradiation is caused by the transfer of radiation energy to the electrons within tissue molecules. These electrons are pushed to higher-energy orbits, or out of atoms altogether, altering molecular structure of the surrounding tissue and/or causing formation of reactive chemical species (free radicals). Being ionizing radiation, X-rays have energy high enough to break atomic bonds, causing direct damage to the cellular structures, including the central control molecule of the cell, the DNA (single and double strand breaks, gene fusions/deletions, translocations, etc.).
By altering molecular structures, X-rays also cause cross-linking between body molecules and twisted molecules formed by radiation; this causes both, structural and functional damage (to gene expression, cellular enzymes and regulatory proteins, hormones, etc.).
How, specifically, the level of breast tissue irradiation by X-ray mammography translates into increased risk of developing cancer? Will try to answer this question next.