Recent media reports have suggested that an increase in thyroid cancer incidence in women may be due to exposure from scattered x-rays during mammography 1 2. These reports have resulted in a large number of inquiries by women presenting for screening mammography concerning possible ways to shield the thyroid from this exposure. In 2008, Sechopoulos et al. 3 reported on absorbed doses to body organs other than the breast resulting from mammography. Specifically, that study used an anthropomorphic phantom to simulate the human body of an adult female undergoing mammography and used Monte Carlo-based algorithms to determine scattered radiation doses to other body organs. The compressed breast was simulated as a homogeneous mixture of 50% adipose and 50% glandular tissue surrounded by a 4-mm layer of skin, as defined by Hammerstein et al. 4. Both craniocaudal and mediolateral oblique mammographic views were simulated for a wide range of x-ray spectra. The phantom included a total of 66 organs, including the thyroid, and the study reported the relative organ dose, defined as the absorbed dose to each organ per unit of mean glandular dose to the imaged breast. Further details on the methodology used can be found in the study by Sechopoulos et al. 3. Because of space considerations, tables 2 (craniocaudal view) and 3 (mediolateral oblique view) of that report included relative organ dose values only in organs for which the relative organ dose was at least 0.10% for one view and one x-ray spectrum. Thus, the tables did not include thyroid doses. Doses below that threshold were deemed negligible.

Given the current conjecture that exposure during mammography has resulted in a significant increase in the incidence of thyroid cancer, in this article we provide the relative organ dose values to the thyroid from mammography, using the methods of the previous study 3. Because of the current focus on the subject of thyroid exposure, we believe it is useful to provide clinicians with exact figures for these doses. In Table 1, we provide the specific relative organ dose values to the thyroid for each mammographic view for various common target-filter and peak kilovoltage (kVp) settings. The relative organ dose values for the thyroid vary from 0.016% (a factor of 0.00016) to 0.045% (a factor of 0.00045), depending on the mammographic view acquired and the x-ray spectrum used.

Based on the mean glandular breast dose from bilateral two-view mammography of 3.7 mGy for digital mammography and 4.7 mGy for film-screen mammography 5 and assuming the maximum relative organ dose for the thyroid of 0.045%, we estimate that the maximum average dose to the thyroid from bilateral two-view mammography is 3.3 μGy for digital and 4.3 μGy for film-screen mammography, taking into account that exposure of each breast results in scattered radiation to the thyroid.

Using the current International Commission on Radiological Protection weighting factor of 0.04 for the thyroid 6, the thyroid exposure from one bilateral two-view mammography examination has a maximum average effective dose of 0.13 μSv from digital and 0.17 μSv from film-screen mammography. Compared with the annual natural background radiation dose in the United States of 3.1 mSv, the dose to the thyroid from bilateral two-view mammography has equal or less detriment than 30 minutes of exposure to natural background radiation, indicating that the thyroid dose from mammography is indeed negligible.

Based on the age- and sex-dependent risk estimates of the seventh Biologic Effects of Ionizing Radiation report (lifetime attributable risk factor for induction of thyroid cancer due to radiation dose to the thyroid of 14 per 100,000 women exposed to 0.1 Gy) 7 and the higher film-screen mean glandular breast dose of 4.7 mGy, the lifetime risk of thyroid cancer induction due to a single screening examination consisting of bilateral two-view mammography for a 40-year-old woman is six per billion (or 1 in 166 million). Furthermore, this risk from a single examination decreases with age at exposure, as shown in Figure 1.

Adding the risks from multiple examinations, the cumulative lifetime risk of thyroid cancer resulting from serial annual screening mammography examinations between ages 40 and 80 years is approximately 56 per billion (or 1 in 17.8 million), as shown in Figure 2.

We believe the dissemination of actual thyroid doses from mammography based on the 2008 study 3 and risk estimations based on those doses will aid the radiology community in communicating with concerned patients and allaying their concerns about mammography risks. Although thyroid shields may provide patients with some psychologic comfort, they can interfere with proper patient positioning and cause artifacts that obscure breast tissue, requiring a retake if a portion of the shield is included in the x-ray field of the mammogram. An example of these artifacts is shown in Figures 3A and 3B, where it can be seen that the thyroid shield not only covers a substantial portion of breast tissue but also causes the automatic exposure control to use a different technique, resulting in loss of contrast throughout the image.

Finally, it should be noted that thyroid cancer incidence in the United States has increased significantly since 1998 at approximately the same rate in both women and men 8. This increased incidence is likely due to improved imaging techniques that lead to the diagnosis of more subclinical thyroid cancer in both sexes rather than increased radiation exposure in women undergoing mammography.