Comparison of radiation exposure and associated radiation-induced cancer risks from mammography and molecular imaging of the breast.
Level 5 - mechanism / opinion, no new human data
Mathematical risk modeling and simulation study using BEIR VII risk projection models rather than empirical clinical data.
PubMed 21302775 · doi:10.1118/1.3512759
What was done
Researchers modeled and compared lifetime radiation-induced cancer incidence and mortality from digital mammography, screen-film mammography, molecular breast imaging (MBI), breast-specific gamma imaging (BSGI), and positron emission mammography (PEM). Using excess absolute risk and lifetime attributable risk models from the BEIR VII report, they simulated a population of 100,000 females followed from birth to age 80. They evaluated annual screening regimens (ages 40–80 and 50–80) assuming a 15%–32% screening-related mortality reduction, testing standard doses of 925 MBq Tc-99m sestamibi for MBI/BSGI and 370 MBq F-18 FDG for PEM against mammography and natural background radiation.
What was found
At conventional doses, cumulative cancer incidence and mortality from MBI, BSGI, and PEM were 15–30 times higher than from digital mammography. The benefit-to-risk ratio for annual digital mammography exceeded 50:1 for screening across ages 40–80 and 50–80, but dropped to 3:1 for the 40–49 age subgroup. To achieve benefit-to-risk ratios comparable to mammography for women with dense breasts, molecular imaging doses would need to be reduced to 75–150 MBq for Tc-99m sestamibi and 35–70 MBq for F-18 FDG.
Why it matters
Standard molecular breast imaging imparts substantially greater radiation exposure than mammography, limiting its viability as a primary screening tool. For these modalities to serve safely in screening populations, radiotracer activities must be lowered substantially through improved camera and detector sensitivity.
Limits
The findings are derived entirely from mathematical simulations and theoretical risk models (BEIR VII) rather than observed clinical cancer endpoints in human cohorts. The model relies on linear non-threshold extrapolations from low-dose radiation, and diagnostic accuracy differences across breast density categories were not modeled directly.
Cited by
- contradicts Molecular breast imaging (MBI) delivered approximately 20 to 30 millisieverts of radiation.