Breast cancer is the second most common cancer in US women, after skin cancer. It’s also the second most common cause of cancer death in women, after lung cancer. Though death rates due to the disease have been declining steadily for decades, cases have been creeping up by about 1% annually for the past several years. The fight continues, as scientists continue to press forward with new innovations and discoveries in risk assessment, early diagnosis, personalized treatment, and monitoring for recurrence.
According to the American Cancer Society, US women as a group have a 1 in 7 chance of getting breast cancer over the course of their lifetimes. Women with family history of the disease are at higher risk. To determine exactly how much higher such a woman’s risk might be, clinicians can use any one of several statistical models.
Unfortunately, according to a recent Cochrane review, none of them are terrific. Two of the four they assessed either overestimate risk (Tyrer‐Cuzick / IBIS) or underestimate it (BRCAPRO). The other two (Gail and BOADICEA) don’t have those problems, but neither of them do a great job of telling whether a woman is likely to get breast cancer or not - an ability called discrimination. BOADICEA performed the best, but even it had only moderate discriminatory accuracy.
To learn more about how one company is using AI to predict breast-cancer risk, listen to this week’s On Target podcast, featuring Connie Lehman of Clairity.
Mammograms are the cornerstone of breast-cancer diagnosis. But they can’t tell you everything. When mammography results are abnormal, women often have to wait agonizing weeks for the follow-up testing that tells them definitively whether or not they have breast cancer.
AI triage could change that — by shortening the wait for high-risk women and decreasing the anxiety of those whose risk is lower. In a recent study in press at Nature Digital Medicine, researchers were able to do that using an open-source, previously validated AI model. The model assesses a woman’s risk of having breast cancer within one year, based on their mammogram results.
In this real-world study, women whose risk was in the top 10% were offered expedited follow-up care. Those who accepted got their mammograms read by a radiologist the same day. When necessary, they got biopsies as soon as possible, often on the same day. As the paper noted, “Time to screening results, diagnostic evaluation, and biopsy were significantly shorter in expedited patients versus control patients with screen-detected cancers, with reductions of 99.1%, 99.1%, and 87.2%, respectively.”
The quick care was warranted, too: Cancer was detected in 60/1000 women in the high-risk group, compared to 2.3/1000 in those who were not high-risk.
Getting a mammogram is a yearly ritual for millions of women. However, according to updated guidelines from the American College of Physicians (ACP), women who have an average risk of breast cancer don’t need to get checked quite as often as that.
For women between ages 50 and 74 who are at average risk, the ACP now recommends mammography only every two years (biennially). Average-risk women who are 40 – 49 should talk with their doctor about whether they should undergo biennial screening, and women older than that no longer need screening at all.
So what does “average risk” mean? Here’s the list:
No personal history of breast cancer
No personal history of high-risk breast lesions
No genetic mutations that are known to increase risk
No family history that puts you at higher risk
No history of high-dose radiation to the chest at a young age.
COMMENTARY: Will these new guidelines change clinical practice? We have doubts about implementation, at least initially. The annual mammogram is ingrained not only in the medical community but in popular culture, as well. Since many other preventative screens have not been broadly adopted (think low-dose CT for lung cancer), we are reluctant to see mammogram guidelines that take away from this strong “role model” system. We’ll see what happens.
Biopsy is the gold standard for definitive breast-cancer biopsy diagnosis. But nobody wants to undergo a biopsy if they don’t have to. For women who find a lump in their breast, ultrasound can help with diagnosis — if it shows that the lump is fluid-filled, that lump is unlikely to be cancer. However, “unlikely to be” isn’t the same as “isn’t,” so these women may end up getting biopsies anyway — only to find out that the lump wasn’t cancer, after all.
Work published in Breast Cancer Research showed that pairing another diagnostic technique with ultrasound can give women better answers. The additional piece of the puzzle is diffuse optical tomography (DOT). DOT uses near-infrared light to measure the oxygen levels around the mass as well as the concentration of hemoglobin in the area. High levels indicate that the mass is very active and a lot of blood is flowing to it, which means it’s more likely to be cancer.
This small study (238 women) showed that ultrasound-guided DOT would let 25% more women avoid biopsy than ultrasound alone — it was able to show that those women’s lumps were much less likely to be cancer. While it wasn’t perfect, the false-negative rate was low, at slightly under 2%.
When someone is treated for cancer, the first thing they want to know is whether the treatment is working. For women with triple-negative breast cancer (TNBC), that information is even more critical, as that form of breast cancer is very aggressive. (TNBC cells have neither estrogen receptors, progesterone receptors, nor the HER2 protein, hence the name.) These women need to know whether their cancer is still spreading, and they need to know asap.
Tumor DNA in a patient’s bloodstream is a clear indication of metastasis. But finding that circulating tumor DNA (ctDNA) can be very challenging, because there’s so little of it. Plus, we didn’t have very many biomarkers for triple-negative breast-cancer cells.
We now have four more markers than we used to, though, thanks to work published in Cancer Research Communications. Using all those markers together not only improved the ctDNA detection rate significantly (from 40% to 95% in a mouse model), it also highlighted receptors that future treatments could one day target.
🎙️ New Episode Alert: On Target sits down with Dr. Connie Lehman, Founder and CEO of Clairity, Inc.
In this episode, Dr. Lehman shares how AI-driven insights can improve early detection, personalize screening, and ultimately change the future of breast-cancer care.
Listen in: ontargetpodcast.com
Join the conversation: If you’re leading innovation in this space and want to share your perspective, we want to hear from you. info@ontargetpodcast.com
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.