skip to content

Blog Post

Landmark UCSF study uncovers a better way to screen for breast cancer

Hannah Hoban, Research Program Manager, Color

Risk-based, individualized approach to breast cancer screening reduced the risk of late stage cancers, found early stage cancers at greater frequency, and safely stratified those at lower risk, improving efficiencies and reducing costs.

Today, UCSF and co-authors from a number of world-renowned institutions published long-awaited results of the WISDOM breast cancer screening study in the Journal of the American Medical Association. It’s a milestone almost a decade in the making, and one that points to a more effective, more accessible future for cancer prevention.

The findings from the study involving 46,000 participants are clear: a risk-based approach to breast cancer screening improves outcomes, reduces unnecessary interventions, and uses resources more efficiently. Importantly, WISDOM also shows that personalized screening can work at scale when paired with strong clinical and technological infrastructure.

Color partnered with UCSF to support essential components of the study, including at-home participant logistics, genetic risk data generation, and infrastructure to support nationwide engagement. Our role helped UCSF operationalize a modern research model that centers on access, risk understanding, and timely clinical clarity. 


Why we need a smarter screening model

For decades, most women in the U.S. have been offered the same recommendation: begin annual mammograms at age 40. While simple to administer, this traditional screening one-size-fits-all approach overlooks well-established drivers of breast cancer risk including breast density, personal and family history, genetics, and lifestyle.

As a result, some people with elevated risk are underscreened, while many with low risk undergo more screening than necessary. Evidence has long suggested that an individualized approach could improve detection and reduce harms, but large-scale data was needed.

WISDOM set out to generate the evidence needed to determine whether personalized, risk-based screening could safely match or outperform traditional screening while remaining acceptable to participants.

Instead of basing screening on age, WISDOM created a risk-based cohort that tailored screening timing and frequency based on each participant’s individual risk profile. The goal was the same for everyone: detect cancer earlier for those at highest risk and reduce unnecessary interventions for those at low risk. 


When screening matches risk, outcomes improve

The WISDOM results show the power of aligning screening with individual risk:

  1. Risk-based screening was as safe as annual screening while improving key outcomes
    • The rate of stage IIB or later stage cancers was noninferior in the risk-based cohort compared with the annual screening cohort meaning risk-based screening was not worse than annual screening in detecting advanced cancers.
    • There was a 37% reduction of stage IIB cancers in the risk-based cohort vs. the annual screening cohort.
    • Women in the highest risk category, assigned to screen every 6 months, had no stage IIB cancers.
  2. Screening intensity appropriately matched risk
    • Cancer incidence, biopsy rates, MRI use, and mammography use increased in step with assigned risk category, demonstrating effective risk stratification.
    • Although risk-based screening led to fewer mammograms overall, biopsy rates did not decrease. Biopsy frequency closely followed the risk category, with frequency much higher in high-risk groups, and lower in low-risk groups.
  3. Scalable testing uncovers unknown risk
    • 30% of participants who completed at-home genetic tests were carriers of high-penetrance variants and had no family history, meaning traditional screening guidelines would not have identified them.

This is exactly what risk-based screening compared to annual screening is designed to accomplish: put the greatest attention where the risk is highest, reduce over-screening where the risk is low, and improve outcomes in a way that is both clinically sound and resource-efficient.


A model for the future of population-level screening

What makes WISDOM especially transformative is not just the data, it’s the delivery model. The study required three critical components to work at scale:

  1. Convenient, at-home testing that made participation and risk assessment accessible across all 50 states
  2. A scalable, digital platform to receive results 
  3. A clinical team capable of delivering guidance, timely follow-up, and clear next steps

Together, these elements enabled UCSF, with operational and bioinformatic support from Color, to stratify risk across a diverse, nationwide population, personalize screening pathways, and deliver care efficiently.


A turning point for breast cancer screening

The WISDOM study’s first decade demonstrates something important: precision screening is possible at population scale. The WISDOM results provide a roadmap for modernizing breast cancer screening across the U.S. and makes a compelling case for expanding risk-based models to other diseases.

Health systems, employers, and public health agencies now have evidence that:

  • Personalized screening is safe and effective
  • Targeted approaches can improve outcomes
  • Population-level infrastructure can make precision care scalable

The next phase of WISDOM, WISDOM 2.0, is already enrolling participants nationwide and expands risk assessment to women as young as 30, a critical shift, as identifying high-risk individuals earlier can change the course of disease.

Researchers are also developing models to distinguish fast-growing cancers from slow-growing ones, further refining when and how screening should occur. This work will continue moving the cancer community toward more efficient risk-based screening, and Color will be involved with WISDOM 2.0 researchers closely, utilizing our expertise to drive the future of precision medicine.

Color and Google Cloud partnership logo
Cancer Care

Color and Google Cloud: Using AI to help tens of thousands of women get screened for breast cancer