The United States Preventative Services Task Force—one of the country’s major independent organizations that drafts recommendations for breast cancer screening—made headlines last month when it announced it was updating its guidelines to recommend that all average-risk women in the U.S. start biannual mammograms at age 40.
To explain these new recommendations and why they’re important, we spoke with BCRF investigator Dr. Constance (Connie) Lehman, professor of radiology at Harvard Medical School and chief of Breast Imaging at Massachusetts General Hospital. Dr. Lehman, who is researching ways to use artificial intelligence to improve breast cancer screening and risk assessment with BCRF support, also discussed why the field needs to move away from age-based screening to personalized, risk-based screening.
Watch the full video above or read an edited transcript of her interview below.
Throughout my career, my focus has been, how can we leverage the strength of imaging of the breast tissue—whether it’s with MRI or mammography or ultrasound—to really tackle every aspect of the breast cancer cycle, from early detection through more accurate diagnosis, more effective treatment, and then bringing women back into surveillance after successful treatment. So that work has been very exciting for me.
However, it was with generous support from the Breast Cancer Research Foundation that we’re now at the brink of a paradigm shift, where the stars are aligning, and we have experts from every domain tackling breast cancer and bringing together the best tools that they have. And this new tool we have in artificial intelligence is pretty incredible. It’s going to enable us to shift from a very crude, imprecise, age-based screening paradigm to a very precise, AI-empowered, risk-based screening paradigm. It’s really going to change the way that we detect and prevent breast cancer.
It’s so important that these experts and committees we have for our different screening guidelines continue to keep up with the research and continue to update their recommendations. [Questions about] screening mammography: When do I start? How often should I be screened? When can I stop? [They have] been an area of controversy for over 50 years.
It’s amazing to think we had screening mammography out in clinical practice in the late 1960s, and in the 1970s, the American Cancer Society started a very strong campaign to get women in for their mammogram. And they said [in the ad copy], “If you’re 35 or older and you haven’t had a mammogram, you need your head examined.” It’s a crazy statement to make to women now. But over the decades, we found that different organizations—whether it was the American College of Radiology, or the American Cancer Society, or our United States Preventive Services Task Force [USPSTF]—they had different perspectives on the same research.
They were all trying to decide what the right age is so that we maximize the benefits of screening and minimize the risks—or the potential harms—of screening. And so, when those groups got together, they came up with different recommendations. Some said start at 40, others at 45, others at 50. That’s why these new recommendations from the USPSTF are exciting for all of us, because we’re coming to a consensus that we haven’t had before [on age].
It was very interesting—the process that the USPSPF used in making their new recommendations—that, rather than having all average-risk women start at 50, they are really recommending starting at 40. And it [was influenced by] two pieces of evidence. Interestingly, it wasn’t a new study on screening mammography that showed women in their 40s do really well with screening mammography. Rather, it was two pieces of evidence that the taskforce brought into their recommendations.
One was the observation that the number of women who are being diagnosed with breast cancer in their 40s is on the rise. They had been following it carefully, and noticed that between 2015 and 2019, it really increased. We had more younger women being diagnosed with breast cancer. We’ve all noticed that, and many women say, “I couldn’t believe I had another friend who was so young—in her 40s or even late 30s—being diagnosed with breast cancer.” And there are a lot of reasons why that trend may be happening. But the USPSTF wanted to acknowledge that and incorporate that into their new recommendations, because part of their risk benefit profile was: If you don’t have a lot of cancers, that benefit drops down. But if now we see that we have more women in their 40s being diagnosed with breast cancer, the benefits of early detection increased because there are more women with cancer [in that group].
There was a second observation from the research that they wanted to incorporate into their recommendations. And that is that we know that Black women tend to be diagnosed at younger ages than white women. And their outcomes are worse.
They wanted to update and have a fresh approach to modern screening recommendations.
The benefits of screening mammography are earlier detection. When we detect breast cancer early, not only can we cure it, but we can cure it with less negative health impacts on the woman who needs to undergo treatment. And those are two amazing results that we have been very proud of in the area of screening mammography.
But there are also harms and potential risks. There are false positive exams, where a woman who doesn’t have breast cancer needs to undergo additional imaging or testing or maybe even surgery—even though there’s no cancer in her breast tissue. There are also the potential harms of false negative exams [where] a woman feels very confident that she’s getting her mammogram regularly, and then is surprised to find out that she has a late-stage disease diagnosis. There are a lot of reasons we could have a false negative.
[Screening] technology itself isn’t perfect; it can miss cancers when we just use mammography. But also, there are some cancers that are very fast-growing, so the interval between the screens is really important. These are some of the benefits and risks that we’ve all grappled with.
I certainly think that the USPSTF got a lot right. They acknowledged that we have racial inequities that they wanted to address, and they acknowledged that more younger women are getting breast cancer than used to be the case. Incorporating that [research] into their new recommendations and stating that women should all start screening at 40 was a significant advance. Really excited about that.
I still think we’ve got some work to do on the intervals in which women are screened. It’s great to tell a woman to start screening at 40. But I’m not sure that [the recommendation for] every two years for women in their 40s is the best science. We need to think about that. The American Cancer Society [and American College of Radiology] strongly believe that screening every year in younger women is really going to save more lives. I think that makes a lot of sense. We need to think about that interval.
Then at the other end of the spectrum: Is there any age at which women can stop screening? I mean, we sometimes have women in their late 90s coming in for screening. Is that a good approach? Do they still need to do that?
I think we have the three questions: when to start, when to stop, and how frequently to be screened. But the when to start, I think we’ve got it right now, and that’s great news.
The biggest takeaway from this announcement is if you’re 40 or older, we want you to be screened with mammography. And I think it’s important that women hear that and understand that. Now, there are going to be a lot of women that are going in for their first mammogram that have never had one before. They may have heard that it’s painful, and it’s terrible. They should talk to their friends and go to a center that has a lot of experience in screening mammography. Let the technologist know, of course, that it’s their first mammogram. And if it’s uncomfortable, have that conversation with the technologist. Also, they should be prepared that they may be called back for some more imaging. It doesn’t mean they have cancer. It means there may be an area or spot that the radiologist just wants to study a little bit more carefully—usually, just with another type of a mammogram, maybe ultrasound. For most of those cases, they can then clear the patient until the time that they come in for screening. In the event that the patient needs a biopsy—again, not to panic—but just to walk through each step.
I am so excited that we’re at a point in medical history where we have new tools to actually, finally move from a very crude, imprecise, age-based screening for breast cancer to the precision of risk-based screening to detect breast cancer. It’s unbelievable that we have this opportunity now. But we have a lot of work to do to start to have women following a risk-based protocol.
Thankfully, we have some experience in this. We’ve known that if a woman has a strong family history, or if she has a genetic mutation, we’re going to screen her differently. We don’t care that she’s 35, she needs to start screening earlier. We also know that mammography won’t be enough; that we need to do vascular imaging with MRI, or even possibly, at some centers, screening ultrasound is integrated into the screening program. These are really exciting domains that we’re right at the edge to engage in.
What I’m most excited about are the many different physicians and scientists that are bringing their knowledge into improved risk assessment. In imaging, we used to think the best that we would do is when a woman comes in for a mammogram, we could ask her questions: How many members of your family have had breast cancer? Have you had a biopsy before? What were the results of that biopsy? Did you have children? Did you breastfeed them?
But now with AI, we have found that we can leverage the power of artificial intelligence and machine vision to see into the mammogram and predict a woman’s future risk of breast cancer. This is really exciting. I sort of always felt like this was going to be a possibility, because I knew that many of those questions that we ask women, I could confirm on the mammogram. I could see the clips of her biopsy, I could see what kind of surgery she had, I could notice that she had gained weight or lost weight, or that she had very dense breast tissue or really fatty tissue. A lot of these clues were already there. But my human eyes and my human brain couldn’t begin to process those [clues] to deliver a precise risk assessment score for the individual woman. And that’s right where we are in our research with the Breast Cancer Research Foundation. And I’m so excited to see how we can start to put this into practice.
When we’re able to move from age-based to risk-based, we’re going to capture more of these women with cancers that occur in their 30s and even, unfortunately, their 20s. And we’ve all seen those. We know that women in their 20s and 30s are likely to have more advanced cancers. They’re not part of regular screening unless they’ve already been identified as high-risk. We need to expand our approach to early detection to find those groups of women that have a cancer burden—those women that have cancer. It’s not all 20-year-olds. It’s not all 30-year-olds. But we’ve got them. They’re there.
And we can now, I believe, identify them more accurately. It’s an exciting time, because many groups now recommend that every woman at the age of 25 know her risk. It’s a great statement we’ve had for a while now, [along with] federal legislation [to] know your breast density. But that really only applies to women 40 and older. And we’ve had a really strong message about making sure you’re screened with mammography, but again, 40 and older. So for these other women in their 20s and 30s. We’re saying, “OK, now we want you to know your risk.”
We’ve got a lot of work to do to make sure that the tools that we’ve been using historically to assess a woman’s risk are really the best we can offer from our research. We know they’re not. We know they have a lot of problems. And traditional risk models, I predict, will be replaced by our image-based AI models, as well as bringing in genetic information from patients and just leveraging the strength of radiomics and genomics to be more precise in finding those women with breast cancer in their 20s and 30s as well as 40s and older.
Please remember BCRF in your will planning. Learn More
