
The incidence of metastatic cancer has remained stable over the past 40 years, a sign that breast screening isn't catching cancers earlier and before they begin to spread and become symptomatic, according to an opinion article published October 29 in the New England Journal of Medicine.
Breast cancer screening's effectiveness may be limited by an understanding of the science behind cancer dynamics, wrote the group led by Dr. H. Gilbert Welch of Dartmouth College.
The incidence of metastatic breast cancer has been stable since 1975, suggesting that screening mammography is unable to identify disease before symptoms appear, according to the authors. In contrast, the incidence of prostate cancer screening has decreased by half since 1988 as a result of prostate-specific antigen (PSA) screening.
Why the difference? Breast cancer is often already a systemic disease by the time it is detectable, whereas prostate cancer progresses in a way that allows time for screening to identify it, Welch and colleagues wrote. The screening approaches are also different: Mammography screening involves an anatomical search for structural abnormalities, while PSA screening uses a biochemical assay to detect a tumor marker.
"It's possible that the latter is a much more sensitive indicator of disease burden," the group wrote. "Were a similar breast cancer assay discovered ... then perhaps fewer women would present with metastatic breast cancer."
Some cancers will be systemic at the outset, some will progress, and some won't progress at all -- and not taking this into account results in screening programs that do not help and leads to ineffective or unnecessary treatment, according to Welch and colleagues.
"Because early detection efforts will never be successful for patients with [systemic disease], disease dynamics can have a profound effect on the efficacy of screening," they concluded.
![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=100&q=70&w=100)







![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)










