The Peter Attia Drive - June 15, 2026


#396 ‒ Breast cancer screening: understanding risk, deciding when to start and how often to screen, and choosing the right imaging strategy


Episode Stats


Length

50 minutes

Words per minute

149.88

Word count

7,539

Sentence count

428

Harmful content

Misogyny

26

sentences flagged

Hate speech

7

sentences flagged


Summary

Summaries generated with gmurro/bart-large-finetuned-filtered-spotify-podcast-summ .

Transcript

Transcript generated with Whisper (turbo).
Misogyny classifications generated with MilaNLProc/bert-base-uncased-ear-misogyny .
Hate speech classifications generated with facebook/roberta-hate-speech-dynabench-r4-target .
00:00:00.000 Hey, everyone. Welcome to the Drive podcast. I'm your host, Peter Atiyah. This podcast,
00:00:16.540 my website, and my weekly newsletter all focus on the goal of translating the science of longevity
00:00:21.520 into something accessible for everyone. Our goal is to provide the best content in health and
00:00:26.720 wellness. And we've established a great team of analysts to make this happen. It is extremely
00:00:31.660 important to me to provide all of this content without relying on paid ads to do this. Our work
00:00:36.960 is made entirely possible by our members. And in return, we offer exclusive member only content
00:00:42.720 and benefits above and beyond what is available for free. If you want to take your knowledge of
00:00:47.940 this space to the next level, it's our goal to ensure members get back much more than the price
00:00:53.200 of a subscription. If you want to learn more about the benefits of our premium membership,
00:00:58.040 head over to peteratiyamd.com forward slash subscribe.
00:01:04.200 Welcome to a new episode of The Drive. Today, we're diving into breast cancer screening,
00:01:09.940 why women are still dying from breast cancer despite effective screening tools,
00:01:15.100 where current screening strategies fall short, and how to think about personalizing your own
00:01:19.760 screening. This episode is really about one central question. How do you give yourself the
00:01:25.200 best possible chance of not dying from breast cancer? As we consider this to be a really
00:01:31.280 important public service announcement, the full episode and the detailed show notes for this
00:01:37.080 discussion will be available to everyone, regardless of whether or not you're a premium
00:01:41.020 subscriber. So without further delay, please enjoy this episode of The Drive.
00:01:49.760 Most of us have heard that terrible statistic. About one in eight women will develop invasive
00:01:57.560 breast cancer over the course of their lifetime. In the United States, roughly 42,000 women die 0.58
00:02:04.000 every year from this disease. That makes it one of the leading causes of cancer death behind only 0.51
00:02:10.560 lung, colorectal, and pancreatic. And yet, despite how common and consequential this disease is,
00:02:17.820 Many women have questions about screening, including when it should start, how often to do it, and what factors actually matter.
00:02:26.100 Even women who have looked at the guidelines often come away more confused than before because the guidance differs between organizations, and it's also shifted over time.
00:02:35.780 And that confusion has real consequences because breast cancer screening works.
00:02:40.740 Cancers found through screening are more likely to be caught early, before they've spread, when treatment is easier and outcomes are better.
00:02:47.820 And that stage shift, as it's called, matters enormously. When breast cancer is caught at
00:02:53.620 stage one, the 10-year survival is over 96%. By stage four, five-year survival is only around
00:03:03.500 30%. It's no surprise then that women who screen regularly are up to 40% less likely to die from
00:03:11.380 the disease. Now, of course, screening is not without trade-offs. For example, over-diagnosis
00:03:16.500 of lesions that would never progress to cancer can increase healthcare burden with no real benefit
00:03:21.740 and remains an area of ongoing uncertainty. We'll get into these considerations later,
00:03:27.180 but they do not negate the core point. Screening is one of the most effective tools we have for
00:03:32.680 reducing breast cancer mortality. And if you're optimizing for your individual risk of dying from
00:03:38.460 breast cancer, not population efficiency, not total societal cost, but your own outcome,
00:03:43.960 the default should be to err on the side of more effective screening, and certainly not less.
00:03:49.780 Which raises the obvious question. If screening works so well, why are so many women still dying 0.99
00:03:56.620 from breast cancer every year? Part of the answer is biology. Some breast cancers are simply more 0.96
00:04:03.380 aggressive than others. They grow quickly, they spread early, and can be difficult to intercept,
00:04:08.780 even with a very good screening system. Some back of the napkin math suggests that somewhere
00:04:15.220 around 7 to 10 percent of cases are the ones we are unlikely to catch even with perfect screening.
00:04:24.080 But those biologically aggressive cases are not the only reason this disease is still taking so
00:04:30.820 many lives. A major and much more solvable part of the problem is far more mundane. We are still
00:04:37.920 under screening. And I don't just mean that some women never get a mammogram, though that is
00:04:43.520 certainly part of it. Even among women who have been screened before, screening may be inconsistent.
00:04:50.640 Roughly a third of women over 40 have not had a mammogram in the past two years. And even among
00:04:57.860 women aged 50 to 74, where the evidence is most universally agreed upon, about 20% are not up to
00:05:07.240 date. Underscreening has two layers. The first is pretty basic. Most women are not even getting
00:05:14.620 routine mammography at the right time. But the second is more nuanced. Some women are getting 1.00
00:05:20.920 screened, but not with the right strategy for their risk profile. This is one statistic that
00:05:27.020 I think captures this perfectly. According to the criteria laid out by major screening guidelines,
00:05:33.460 at least 9% of women meet the threshold for breast MRI as part of their screening protocol.
00:05:42.180 And yet the actual utilization rate is just 0.4%. It's not that MRI is unproven or controversial
00:05:50.600 for these women. This is a pure execution failure. We already know who these high-risk women are, 1.00
00:05:58.560 and we already have a tool that materially improves detection. We're simply not connecting
00:06:04.020 the two. So what I want to do today is not simply tell you to get screened. You already know that.
00:06:10.220 I want to give you a practical framework for smarter screening, one that starts with understanding
00:06:15.860 your actual risk and helps you make informed decisions about when to start, how often to
00:06:21.980 screen, and what imaging to use. Population guidelines form a great starting point, but
00:06:28.020 they're not the finish line. But to get there, we need to start with the guidelines themselves.
00:06:33.720 What do the major organizations actually recommend, and how should we talk about personalizing them?
00:06:39.720 If you're confused about breast cancer screening, you're not alone. A recent survey found that
00:06:45.660 roughly 50% of women aren't sure when to start mammography. And frankly, that confusion is
00:06:52.540 understandable. The guidance has shifted multiple times over the past two decades, most recently
00:06:57.800 in 2024. There are several organizations that publish screening recommendations, and they do
00:07:04.360 not all agree. Rather than walking through all of them individually, let me first give you the
00:07:09.900 composite picture of the most rigorous guidelines. We'll include a full comparison table in the show
00:07:15.520 notes. Here's the big picture. Every woman should have a formal risk assessment by about the age of
00:07:22.540 25. If you are average risk, annual mammography should begin at 40. If you are high risk, and I'll 0.81
00:07:31.060 define that in a minute, you may need an MRI and eventually mammography much earlier, and screening
00:07:37.960 should continue for as long as you would be willing to pursue treatment if cancer were found.
00:07:44.300 That composite comes from groups like the American Cancer Society, the National Comprehensive Cancer
00:07:50.560 Network and the American College of Radiology. The one notable outlier is the U.S. Preventive
00:07:57.140 Service Task Force, or USPSTF, whose guidance tends to inform insurance coverage decisions.
00:08:05.980 The USPSTF currently recommends mammography only every other year for average-risk women
00:08:12.880 aged 40 to 74, with no explicit recommendation for high-risk women. Now, these are population-level
00:08:21.440 guidelines. They're created by looking at large studies and evaluating the optimal strategy for
00:08:28.180 maximizing cancer detection while minimizing false positives across millions of women. And
00:08:33.760 of course, they take into account the societal cost of screening. But when you move from populations
00:08:40.640 to actual people, personalization matters. When we think about how to personalize screening,
00:08:48.520 for any cancer, the first question we need to start with is what is my baseline risk? Then,
00:08:56.000 given that risk, how much false positive burden am I willing to tolerate in exchange for potentially
00:09:03.120 earlier detection? And third, which screening plan, both modality and frequency, best match
00:09:11.640 both of these objectives? This is not about trying to outsmart the guidelines. It's about deciding
00:09:18.520 whether the default plan actually fits you. For some women, it will, but for others, it will fall 0.99
00:09:24.940 short. The aim is not to get as much imaging as possible. It's to land on a strategy most likely
00:09:32.540 to help someone with your particular risk profile. So let's start with the first question. What
00:09:38.960 actually determines your baseline risk? While most people immediately think of BRCA level risk when
00:09:45.420 they think about breast cancer, most women who develop breast cancer do not have one dramatic
00:09:51.000 obvious risk factor. It is far more common to have several smaller risk factors that together
00:09:58.000 add up to a higher risk. So we are not just looking for the rare woman with the obvious red 0.97
00:10:04.540 flag. We want to identify the women whose overall risk is meaningfully above average. However,
00:10:11.700 risk assessment is only useful if it happens early enough to change the plan. You do not want to find
00:10:19.080 out at age 42 that you should have been the patient who started MRI years earlier. That is
00:10:25.680 why some groups recommend formal risk assessment by age 25, not because everyone needs imaging at
00:10:31.960 25, but because by then you want to know whether or not you are truly average risk or not. I think
00:10:39.280 this is exactly right. A risk assessment should be done in your mid-20s. So what are the risk
00:10:44.700 factors we need to consider? The most basic are the ones we don't usually think of as risk factors,
00:10:50.980 sex, and age. Breast cancer is overwhelmingly more common in women, one versus eight versus
00:10:56.000 one in about 750 for men. So keep in mind, men can develop this as well. Then there's age,
00:11:02.620 which is one of the strongest breast cancer risk factors overall. The median age of diagnosis is
00:11:08.540 around 62, and the vast majority of breast cancers are diagnosed after age 40. While age alone is an
00:11:15.760 imperfect guide, it is important to keep in mind that risk accumulates over time. But once you get
00:11:20.940 past these two obvious risk factors, the most obvious high-impact category is genetics. Mutations
00:11:28.180 in genes like BRCA1 and BRCA2 are the most recognized inherited breast cancer risk factors,
00:11:34.300 and for good reason. They can substantially increase risk and often shift that risk earlier
00:11:39.900 in life. But these mutations are actually much rarer than people tend to think. In the general
00:11:46.320 population, only about 1 in 400 people carry a pathogenic mutation in BRCA1 or BRCA2,
00:11:54.180 though prevalence is higher in some groups, including people of Ashkenazi Jewish ancestry.
00:12:00.100 There are also other important genes, but the broader point is that inherited mutations matter
00:12:06.460 a great deal for screening, even if they account for a minority of breast cancer cases.
00:12:12.780 And genetics ties directly into the next factor, family history.
00:12:17.180 I separate these because family histories capture more than just the known single gene mutations.
00:12:24.060 Yes, family history is a proxy for high-impact mutations like the BRCA mutations,
00:12:28.720 but it also reflects the cumulative effect of lower penetrant genetic variants,
00:12:35.460 shared environmental exposures, and other inherited factors that no single genetic test currently covers.
00:12:42.420 Having multiple first or second degree relatives, parents, siblings, grandparents, aunts, with breast cancer can substantially increase your risk.
00:12:52.420 But keep in mind that lack of family history does not automatically equate to low risk.
00:12:57.180 Some families are small, some relatives may have died young of other causes, and, importantly, some mutations that increase breast cancer risk can show up in the family as other cancers, like prostate or pancreatic cancer, rather than an obvious pattern of breast cancer.
00:13:15.920 So family history matters, but it should be interpreted thoughtfully, and the absence of a family history of breast cancer does not guarantee that you are of average risk.
00:13:25.160 In addition, ancestry-related differences can influence both risk and tumor biology.
00:13:32.420 In the U.S., for example, black women are more likely to be diagnosed younger and with more
00:13:37.700 aggressive subtypes, which has implications for how early and how aggressively screening should
00:13:43.540 be considered. Another important category is prior chest radiation. The classic example is
00:13:50.360 radiation treatment for Hodgkin's lymphoma in adolescence or early adulthood. Cumulative high
00:13:56.160 dose exposure, particularly around the time when breast tissue is still developing, carries the
00:14:01.720 greatest risk. Now, to be clear, this does not mean you should avoid a diagnostic x-ray if your
00:14:06.600 doctor recommends one. A single chest x-ray or CT delivers a tiny fraction of the dose used in
00:14:13.020 cancer treatments, and the evidence that the occasional routine chest imaging study meaningfully
00:14:18.240 increases. Breast cancer risk is very weak at best. Then there's breast density, which is important
00:14:24.180 for two separate reasons. First, dense breasts are associated with a somewhat higher baseline risk of 0.97
00:14:30.840 breast cancer. But just as importantly, dense tissue makes mammograms harder to interpret 0.98
00:14:36.600 because both dense tissue and tumors appear white on the image. So density is not just a biological
00:14:43.420 risk factor. It is also one of the main reasons our most common screening test becomes less
00:14:49.740 effective. And this is where things get a little tricky. You usually cannot actually know your
00:14:56.460 breast density until you've had an imaging study, which for most otherwise average risk women does 0.85
00:15:02.140 not happen until screening begins at around age 40. This creates a bit of a catch-22. Density
00:15:08.280 matters for screening decisions, but you often do not learn until it's potentially too late.
00:15:14.520 There's a partial workaround though. Breast density is fairly heritable, roughly 60 to 70%.
00:15:20.200 So if your mother or grandmother was told she had dense breasts, that's worth knowing. It is just
00:15:26.340 not something most women tend to think about. Density matters enough that the FDA now requires 1.00
00:15:32.340 imaging centers to notify women of it. That is typically reported with the BIRADS density
00:15:38.640 category. A and B are considered non-dense, while C and D are considered dense. While density
00:15:47.300 declines with age, about 50% of screening age women have dense breast tissue. Then there are 1.00
00:15:54.580 the reproductive and hormonal factors, which tend to be more cumulative in nature. In general,
00:16:00.680 the factors that point toward higher risk include earlier onset of menstruation, later menopause,
00:16:07.660 never having had a full-term pregnancy, or having a first pregnancy after age 40 and not breastfeeding.
00:16:14.580 Individually, none of these risk factors typically change the screening plan on its own,
00:16:18.740 but they do contribute slightly to overall risk, especially when several are present together.
00:16:25.120 The same is true, of course, of modifiable risk factors such as alcohol use,
00:16:29.160 obesity, poor metabolic health, and physical inactivity. You don't need to change your
00:16:35.560 screening protocol based on these alone, but they can shift risk and are factors you should
00:16:41.840 actually do something about. So the big picture here is that risk is usually not one thing. It is
00:16:48.320 the sum of multiple inputs, some large, some small, that together determine whether someone
00:16:54.760 is truly average risk, somewhat above or below average, or clearly high risk.
00:17:02.320 And because no one is going to accurately estimate all of that in their head, this is
00:17:08.160 where formal tools can help. 0.70
00:17:10.940 Calculators like Tyra Cusick combine family history, personal risk factors, and breast
00:17:16.940 density to estimate 10-year and lifetime risk.
00:17:20.280 These models are not perfect, but they are much better than guessing, and they help
00:17:24.320 identify women whose risk is high enough to justify earlier or more intensive screening.
00:17:30.160 Most screening guidelines classify lifetime risk above 20% as high risk, though factors such as
00:17:36.460 ancestry may shift that threshold. You can find these calculators online, but we're going to link
00:17:40.960 to one in the show notes. And it's a good idea to complete one so you can get a more concrete
00:17:46.120 understanding of where your risk falls. So if you take anything away from this, know your risk and
00:17:51.920 know it early enough that you still have time to act on it. But risk alone doesn't tell you what
00:17:57.740 to do. The next question is, given that risk, how aggressively do you want to be screening?
00:18:03.620 Because every time you push screening towards higher sensitivity, that is, finding more cancers,
00:18:09.920 you also accept more downside, more false positives, and more follow-up testing,
00:18:15.360 not to mention the anxiety that can come with those. In a perfect world, screening would find
00:18:20.500 every meaningful cancer early and spare everyone the downside of a false alarm. But in the real
00:18:26.420 world, the balance is never that straightforward. In the U.S., about 10% of screening mammographies
00:18:33.180 lead to a callback for additional testing. But only about 5% of those callbacks end in a cancer
00:18:40.980 diagnosis. In other words, for every thousand women who undergo a screening mammogram, about
00:18:48.020 100 will be called back because something looked abnormal, but about 95 of those 100 will not
00:18:54.720 actually have cancer. And those numbers add up over time. More than half the women screened annually
00:19:02.300 for 10 years will experience at least one false positive result. The most common harms from breast
00:19:09.320 cancer screening are not major physical injuries from the test itself, but the anxiety and
00:19:14.960 uncertainty that comes from a return visit for more imaging. When a callback leads to an earlier
00:19:21.460 cancer diagnosis, it's obviously worthwhile. But when most callbacks do not end in cancer,
00:19:27.620 we need to think carefully about what screening strategy makes sense for each individual woman. 1.00
00:19:33.560 So how should a woman think about that trade-off? The higher your baseline risk, the easier it is 1.00
00:19:38.020 to justify accepting more false positives in exchange for finding more cancer earlier. If
00:19:44.000 you're very high risk, say you carry a BRCA mutation or have a combination of factors that
00:19:49.020 pushes your lifetime risk well above average, then a more aggressive strategy is easy to defend.
00:19:55.360 If you're truly average risk or even lower than average risk, the decision is based largely on
00:20:00.900 preference. That does not necessarily mean less screening. It means being thoughtful about how
00:20:07.020 much extra testing you're willing to accept for what may be smaller incremental benefits.
00:20:12.120 And that's really the key here. More screening is not automatically better screening. The right
00:20:18.260 question is not how much imaging can I get, but which strategy is most likely to help someone
00:20:23.980 with my risk profile and am I comfortable with the trade-offs that come with that? There's no
00:20:28.940 universal answer to that question, of course. It depends on the person. But once you know your
00:20:33.180 baseline risk and you have a sense of your tolerance for false positives, you're ready
00:20:37.180 for the next piece. Which screening tools actually exist and how should you think about choosing
00:20:41.820 between them. When we talk about breast cancer screening, it's often treated as if breast imaging
00:20:46.840 were just one thing, but it's not. Breast imaging is a toolkit consisting of different varieties of
00:20:53.400 mammography, MRI, and ultrasound. Some of these are best for routine screening, while others are
00:20:58.980 more appropriate for follow-up on an abnormal finding or as a supplemental option in higher
00:21:04.400 risk women. We'll go through each of these here, and I'll also place in the show notes a detailed
00:21:09.420 comparison table. The foundation of screening, of course, is the mammography. Mammograms use
00:21:14.820 low-dose x-rays to look for breast cancer, and for most average-risk women, they remain
00:21:19.360 the starting point. One important strength of mammography is that it is particularly good at
00:21:25.060 detecting calcifications, such as those seen in ductal carcinoma in situ, or DCIS. DCIS is
00:21:32.600 sometimes called stage 0 breast cancer. The abnormal cells are still confined to the ducts 0.69
00:21:38.400 rather than invading surrounding tissue. If left untreated, estimates suggest that somewhere
00:21:43.960 between 25 and 60% of DCIS cases may eventually become invasive cancer, a wide range that reflects
00:21:51.020 how much we still don't know about the natural history of DCIS. But that uncertainty is precisely
00:21:57.300 why early detection matters, and mammography is well-suited for finding it. That said,
00:22:03.020 not all mammograms are the same. In 2000, the FDA approved the first full-field digital mammography
00:22:10.380 system, and this is what most of us are thinking of when we hear mammogram. Standard digital
00:22:17.500 mammography, sometimes called 2D mammography, is the technology that many research studies have
00:22:23.760 historically used. Most recently, in 2011, digital breast tomosynthesis, or DBT, commonly referred to
00:22:32.060 as 3D mammography was rolled out. DBT takes multiple images from different angles to create
00:22:40.080 a more layered view of the breast. This results in better cancer detection with lower recall rates, 0.51
00:22:46.440 particularly for women with dense breasts. Not every imaging center offers DBT, however, and
00:22:52.940 sometimes there is still an added cost. However, this is the version of mammography I would
00:22:57.880 prioritize, again, especially for women with dense breasts. Mammography is the right foundation for 0.61
00:23:04.140 virtually everyone, but for women at higher risk, mammography alone may not be sensitive enough.
00:23:09.480 This is where MRI comes in. While not a complete replacement for mammography, MRI is an important
00:23:16.100 supplemental option. MRI uses magnetic fields and intravenous gadolinium-based contrast to detect
00:23:24.200 abnormal blood flow or tissue behavior that mammography can miss, making it the most sensitive
00:23:29.720 screening tool available. It is better for detecting very small invasive tumors and atypical cancers
00:23:35.780 than any other imaging modality. But it's not perfect for everything. Mammography still does
00:23:41.000 a better job with certain calcifications, so MRI is generally used in addition to, but not instead
00:23:48.400 of mammography. The downsides are cost, access, and the use of IV contrast, along with a higher
00:23:56.300 callback burden because of the higher sensitivity. But if your goal is maximum cancer detection,
00:24:02.940 MRI's sensitivity tends to outweigh these downsides. The full breast MRI is what most
00:24:09.680 very high-risk women are recommended for initial screening and may also be used for diagnostics 0.99
00:24:15.360 after an abnormal test. More recently, many patients are opting for the abbreviated breast 0.99
00:24:21.920 MRI, which in my mind is the most underutilized tool we have. Based on current diagnostic accuracy
00:24:29.960 data, the abbreviated protocol preserves nearly all of the sensitivity of the full exam, but takes
00:24:36.220 only 10 to 15 minutes compared to 30 to 60 minutes for the full exam. That makes it cheaper, faster,
00:24:42.580 and more scalable while still providing dramatically better cancer detection than
00:24:47.480 mammography alone. For women with extremely dense breasts, adding MRI after a negative mammogram 0.76
00:24:53.960 cut the rate of interval cancers, meaning cancers found between screens, in half from 5 per 1,000
00:25:00.980 to 2.5 per 1,000 with mammography alone. So for women who are high risk, have dense breasts,
00:25:07.920 or simply want a more sensitive complement to mammography, MRI is the strongest option we have.
00:25:14.720 In many screening settings, an abbreviated MRI is likely sufficient, while the full protocol
00:25:20.480 may be more useful for diagnostic workup and select edge cases. If MRI is not feasible,
00:25:27.640 the next best modality is contrast-enhanced mammography, or CEM. CEM is a newer modality
00:25:35.660 introduced in 2011 that is basically mammography plus intravenous iodine-based contrast. It gives
00:25:43.380 more functional information than a standard mammogram and can be a very reasonable alternative
00:25:49.300 when MRI is unavailable or contraindicated. This technology isn't yet widely available,
00:25:56.460 but I do think it will become more popular in the future. Then there's ultrasound. Ultrasound
00:26:01.360 uses sound waves rather than radiation or magnetic fields to create images of breast tissue as with
00:26:07.600 the other imaging modalities ultrasound comes in different flavors handheld where a technician or
00:26:13.360 radiologist moves the probe manually and automated where the machine acquires images more systematically
00:26:20.560 both are more operator dependent than mammography or mri and both carry a higher false positive
00:26:26.560 burden than you would see in those modalities. Ultrasound can be useful as a supplemental
00:26:32.120 imaging modality, but its value is highly dependent on two things. Who performs the
00:26:38.280 ultrasound and what baseline imaging you are adding it to. In one study, adding physician-performed
00:26:46.600 handheld ultrasound to a standard 2D mammography increased the rate of cancer detection by 4.2
00:26:54.780 per 1,000 women screened. A second study paired technician performed ultrasound with DBT,
00:27:02.580 the more sensitive mammogram, and the detection boost was much smaller, only about 1.1 per 1,000.
00:27:10.140 So when the base imaging is better or the technician is less experienced, the incremental
00:27:16.040 benefit of ultrasound shrinks. That doesn't mean it's not worth doing, but it does mean its benefits
00:27:22.420 are more variable than what we see when adding MRI to mammography. So where does ultrasound best
00:27:29.100 fit into practice? Because handheld ultrasounds allow for real-time visualization, they're an
00:27:34.920 excellent option for getting more clarity on something suspicious seen on other imaging or
00:27:39.680 guiding biopsies. And while ultrasound can boost cancer detection over mammography alone,
00:27:45.900 it's not going to do it to the same extent seen with adding MRI. Ultimately, ultrasound is a
00:27:52.080 viable tool. But for average risk women, it is not a substitute for mammography. And for high risk 0.89
00:27:57.840 women, it's not a substitute for MRI. So how do we actually pull all of this together in a way
00:28:04.640 that is useful? At a practical level, the hierarchy looks like this. The foundation for all women is
00:28:10.820 mammography, ideally digital breast tomosynthesis or DBT. If your risk is elevated or if dense
00:28:19.020 breast tissue is likely to reduce the sensitivity of mammography, MRI is the most effective
00:28:24.600 supplemental tool. If MRI is not feasible due to cost, access, or contraindication,
00:28:30.480 contrast-enhanced mammography is the next best option. Ultrasound can add incremental detection,
00:28:36.980 but its benefit is more variable and highly dependent on operator skill and on, of course,
00:28:43.240 baseline imaging that it's paired with. The key point is that these tools are not
00:28:48.280 interchangeable. They have a clear hierarchy in terms of sensitivity and consistency. And your
00:28:53.460 goal is to choose the combination that best matches your risk. But regardless of the screening tool
00:28:59.340 you decide on, keep in mind that where you get screened matters. Not every imaging center offers
00:29:05.100 every modality, and not every center has the same level of experience with the tests it offers.
00:29:11.100 Simply having imaging done is not the same thing as having high-quality imaging done.
00:29:16.380 For mammography, positioning and complete tissue capture matter. MRI and CEM are more standardized,
00:29:25.100 but they still depend on good protocol execution, contrast timing, and experienced interpretation.
00:29:31.620 For the reasons I discussed a moment ago, ultrasound is the most operator-dependent
00:29:35.580 modality. So a routine mammogram can usually be done well in many places, but the quality of
00:29:42.260 imaging and interpretation still varies. High-volume centers and dedicated breast imaging centers
00:29:47.780 tend to have more experience, better protocols, and more consistent interpretation.
00:29:53.080 If you are pursuing more advanced imaging, such as MRI or contrast-enhanced mammography,
00:29:58.500 or even ultrasound in dense breast tissue, you should strongly consider going to a center that
00:30:03.760 performs these studies frequently and has specialized expertise. The difference is not
00:30:09.300 trivial, and in some cases, it can directly affect whether a cancer is detected or not.
00:30:14.700 Bottom line here is that each imaging modality has strengths and weaknesses,
00:30:18.900 and which one or which option you choose is a personal choice based on your risk,
00:30:24.660 your preferences, and your access. Once you think about imaging in that way,
00:30:28.780 the next question becomes, how often should you screen? As I mentioned previously, the USPSTF
00:30:35.500 recommends biennial mammography for average risk women aged 40 to 74. Most other groups recommend
00:30:42.140 annual mammography starting around the age of 40 or a hybrid approach with annual screening
00:30:48.320 from 45 to 54 and biennial screening thereafter, though I've never really understood that given
00:30:54.580 that risk is always going up with age. It's worth acknowledging up front that no randomized
00:31:00.380 control trial has ever directly compared annual versus biennial screening with mortality as its
00:31:07.400 primary endpoint. Every interval recommendation we have is based on modeling studies and
00:31:13.060 observational data. That does not mean we're flying blind, but it does mean we are working
00:31:17.600 with a slightly different kind of evidence than we have for, say, the question of whether screening
00:31:22.200 reduces mortality at all. The most important modeling data come from the Cancer Intervention
00:31:28.880 and Surveillance Modeling Network, or CISNET, a consortium of independent modeling groups
00:31:34.940 funded by the National Cancer Institute. CISNET has been commissioned three times to run
00:31:40.740 simulations for the USPSTF, and the task force has leaned heavily on these models to justify
00:31:48.520 biennial mammography recommendations. Interestingly, other cancer and radiology groups
00:31:54.140 have looked at the same CISNET data to conclude that annual screening is better. So which is it?
00:32:00.680 The answer depends on your question. If the question is how to maximize efficiency across
00:32:06.980 an entire population, balancing cost, false positives, and resource utilization,
00:32:12.880 biennial screening is a defensible answer. But if the question is what gives an individual woman 0.98
00:32:19.200 the best chance of avoiding death from breast cancer, the answer is different. And that
00:32:24.720 distinction is critical for everything that I'm about to say. The argument for biennial mammography
00:32:31.100 traces back to the original 2009 CISNET analysis, which found that biennial screening retained
00:32:38.060 about 81% of the mortality benefit of annual screening with roughly half as many false
00:32:44.460 positives. This was done using data from film-based mammography, the technology that came before
00:32:50.680 digital mammography and certainly before DBT was introduced. The conclusion from this modeling
00:32:56.260 study was not that biennial mammography was best for mortality, but that it offered the best
00:33:02.280 trade-off between benefit and resource use at a population level. In 2024, CISNET published its
00:33:10.420 most comprehensive analysis using multiple models and imaging strategies, including both
00:33:16.080 two-dimensional digital mammography and DBT. A secondary analysis of those results showed,
00:33:22.640 in essence, that annual screening is better for saving lives. When compared to no screening,
00:33:28.780 annual screening of women aged 40 to 79 produced a 42% mortality reduction, while biennial screening
00:33:36.220 produced only a 30% mortality reduction. In absolute numbers, this corresponded to 230 life
00:33:45.660 years gained per thousand women for annual screening versus 165 for biennial. The cumulative
00:33:54.000 number for false positives is higher for annual screening, as we would expect, given that we're
00:33:58.760 performing twice as many tests. However, the rate of false positives and benign biopsies per exam
00:34:05.360 is actually lowest with annual screening, likely because the radiologist has a more recent image
00:34:12.080 to compare to, making it easier to distinguish cancerous changes from normal variation. I think
00:34:18.340 this is a very important statistic that gets overlooked. The bottom line is that the case
00:34:23.700 for biennial screening rests on population-level efficiency, not on maximizing the benefits for
00:34:29.260 any individual woman. If the question is what gives you the best chance of not dying from breast
00:34:34.640 cancer, CISNET's own data answers it clearly, screen annually. The observational data tell a
00:34:42.640 similar story. Among women aged 40 to 84 who develop breast cancer, those screening annually
00:34:49.440 had far fewer interval cancers, 11% versus 38% for biennial screeners, and were more likely to
00:34:57.780 have an early stage diagnosis, 76% stage one with annual screening versus 56% with biennial.
00:35:06.520 Taken together, the modeling and observational data, the conclusion is straightforward. If your
00:35:11.600 goal is to maximize your chances of avoiding death from breast cancer as an individual,
00:35:16.800 annual mammography is the better strategy. But all of these studies have focused entirely on
00:35:23.240 mammography. What about when to add in other imaging modalities such as MRI? The most common
00:35:29.740 approach in clinical practice is to alternate the two tests every six months, say mammography in
00:35:36.760 January, MRI in July, rather than stacking them at the same time. The logic is kind of intuitive.
00:35:42.740 By spacing them out, you're effectively creating a six-month screening interval instead of a 12-month
00:35:48.240 one, which in theory gives the fastest growing cancers less time to develop between the screens.
00:35:55.740 But the evidence here is thinner than you might expect, in large part because these studies are
00:36:00.500 genuinely hard to do well. When two tests are given at two different times, it becomes very
00:36:05.500 difficult to fairly credit a cancer detection to one test versus the other, which makes comparing
00:36:11.240 the two scheduling approaches unreliable. One small study found that the combined sensitivity
00:36:17.340 of MRI plus mammography was higher than either test alone, while another found no significant
00:36:23.480 difference between stacking and alternating. Neither is definitive. For ultrasound plus
00:36:29.040 mammography, the timing question has essentially never been studied directly. All major ultrasound
00:36:34.260 trials performed both tests simultaneously and annually. Any rationale for staggering ultrasound
00:36:40.380 is purely extrapolated from the already limited MRI literature. The bottom line on multi-modality
00:36:47.400 scheduling is rather unsatisfying. We do not yet have compelling evidence to favor one
00:36:53.300 schedule over another. I might lean towards the alternating schedule for high-risk women 1.00
00:36:59.420 who are more likely to develop rapidly growing cancers, but until better prospective data exists,
00:37:05.400 the scheduling decision is really more about logistics and personal preference.
00:37:09.380 To summarize the practical takeaway here, all women should screen annually with mammography.
00:37:14.380 If you are high risk and you are getting multiple imaging modalities, alternating them every six
00:37:20.740 months is reasonable, but the evidence is not strong enough to insist on that schedule.
00:37:24.960 Right now, the bigger issue is not whether the tests are staggered. It's whether both tests are
00:37:29.800 getting done consistently each year. Okay. At this point, we know we should do a risk assessment in
00:37:36.680 mid-20s and we know how often to screen and which imaging modalities make sense, the next question
00:37:44.820 is when should we actually start screening? I saved this for last because this is the area
00:37:49.800 where we have the least concrete data beyond the whole scheduling thing. The intuitive assumption
00:37:54.980 is that if screening saves lives, then starting earlier should save more of them. But whether
00:38:00.800 earlier screening actually helps depends on many factors, and the answer is more nuanced than you
00:38:05.520 might expect. First, the risk for developing breast cancer before age 40 is genuinely low.
00:38:12.160 Only about 5% of breast cancer diagnoses occur in women under 40. The cumulative risk through age
00:38:17.880 90 is about 13%, which is where that familiar 1 in 8 statistic comes from. But the cumulative
00:38:23.400 risk through age 40 is less than 1%. Now the obvious next question, how do we know incidence
00:38:30.700 is actually low if we're not routinely screening women under 40. Maybe we're just not finding
00:38:36.240 cancers that are there. But these data come from cancer registries, which capture all diagnosed
00:38:42.540 cancers, including cancers found symptomatically and cancers caught in women who began screening
00:38:48.020 earlier because of elevated risk. On top of that, the age incidence curves rise smoothly and
00:38:54.960 continuously through the 30s and 40s. We've included a figure in the show notes so you can
00:39:00.340 take a look at this for yourself. But there is no sharp spike at age 40, which is what you would
00:39:05.740 expect to see if screening were suddenly uncovering a large hidden backlog of disease. So the low
00:39:12.560 incidence under 40s appears to be real, not an artifact of under-detection. But low risk does
00:39:19.740 not mean no risk. So could there still be a case for screening before 40? This is where we run into
00:39:26.660 the limits of evidence. We have no randomized trials and no mortality data on screening average
00:39:32.680 risk women under 40. The best study we have looked at about 6 million mammograms from facilities
00:39:39.840 across the U.S. About 12% of these mammograms were from women under 40. And the researchers
00:39:46.340 compared cancer detection rates across age groups, broken down by whether a woman had at least one
00:39:52.540 risk factor, defined as a personal history of breast cancer, a family history, and a first-degree
00:39:56.920 relative or dense breasts. Among women aged 35 to 39 with at least one risk factor, the cancer
00:40:04.440 detection rate was 2.1 per 1,000 women screened. For average risk women in the same group with no
00:40:11.880 family history, no personal history, and non-dense breasts, the rate was just 0.59 per 1,000. This is
00:40:20.080 about a three and a half full difference driven by risk factors, not age alone. And here's where
00:40:26.380 the comparison gets especially interesting. For average risk women age 40 to 44, the detection
00:40:32.660 rate was only 0.71 per thousand. In other words, women in their late 30s with risk factors were
00:40:42.260 being diagnosed with cancer at roughly three times the rate of average risk women who were screening
00:40:49.240 in their 40s. Those risk factors are not just nudging the needle. They are effectively shifting
00:40:56.580 a woman's screening profile forward by a decade. Once you get to women 45 and older, incidence
00:41:04.020 rises enough that detection rates exceed what we would see in women under 40, regardless of risk.
00:41:10.860 But below that threshold, risk factors may matter more than age. This brings us back to the point
00:41:17.720 I've made throughout this discussion. Knowing your baseline risk is what makes the difference.
00:41:23.280 The age 40 cutoff for screening makes sense on average, but for women with one or more of these
00:41:29.760 three risk factors we've just discussed, earlier screening may be the right call. And the women who
00:41:35.560 are diagnosed with breast cancer in their 20s and 30s are not a random cross-section of the
00:41:41.320 population. They are disproportionately women who could have been identified as above average risk
00:41:47.180 earlier on. On top of this, breast cancer doesn't look the same in younger women as it does in
00:41:53.060 older women. Younger women are much more likely to develop aggressive subtypes. About 20% of
00:42:00.040 breast cancer in women under 40 are the so-called triple negative, the most aggressive form,
00:42:06.560 compared with roughly 6 to 12% in women over 40. And these tumors grow fast. Triple negative
00:42:14.120 cancers can double in size in under four months, which means even annual screening may not catch
00:42:19.800 them in time. On the other end of the spectrum, the slower growing cancers that screening is best
00:42:25.520 at detecting, the ones with doubling times closer to a year, make up only about a third of cases in
00:42:31.860 women under 40, compared to well over a half in older women. For high-risk women, this biology
00:42:38.900 has direct implications for which screening tool to use. Mammography alone may not be enough. 1.00
00:42:46.380 Consider how risk from a BRCA1 mutation plays out over a lifetime. It's not spread evenly. It's
00:42:52.700 heavily front-loaded. A woman in her late 20s carrying a BRCA1 mutation has a breast cancer
00:42:59.620 risk roughly 100 times that of a non-carrier. By her 30s, it's about 44 times. And by her 60s,
00:43:08.140 it drops to about three times. In other words, these mutations don't just increase risk,
00:43:13.720 they shift the entire risk curve earlier. And because the cancers that develop in these younger
00:43:20.560 high-risk women tend to be fast-growing and harder to see in mammography, MRI is the more
00:43:26.400 appropriate screening tool for this group. There is one more practical consideration worth
00:43:32.320 mentioning. One of the three risk factors in the study I just discussed was breast density. And as
00:43:38.740 I mentioned to you earlier, you cannot know your breast density without imaging, at least
00:43:42.900 definitively. About half of women have dense breasts and density is higher in younger women
00:43:48.960 than older women. So there's a reasonable argument for getting a single baseline mammogram in your
00:43:54.720 30s, not primarily to find cancer, but to establish whether you have dense breasts. If you do, that
00:44:01.780 changes your risk profile and may change your screening strategy. To be clear, this is less
00:44:07.140 about cancer detection and more about risk stratification. And while there is no direct
00:44:12.400 evidence supporting this approach, it's something that I think is worth considering. So when should
00:44:18.100 you start screening? As with everything else we've talked about, it depends on your risk.
00:44:22.600 If you are truly average risk, beginning annual mammography at 40 is well supported, though you
00:44:29.040 may want to consider getting a baseline mammogram in your 30s at a minimum to establish your breast
00:44:34.340 density. If you are above average risk, there is a small amount of evidence that mammography in
00:44:41.020 your 30s can be worthwhile. And if you're clearly high risk, a known BRCA carrier, a strong family
00:44:46.800 history, or prior chest radiation, the conversation is different entirely and more aggressive screening
00:44:51.700 protocol should be started in your 20s or early 30s. Now everything we have discussed so far is
00:44:57.880 about routine screening, finding cancer before it causes symptoms. But there's one type of breast
00:45:04.180 cancer that does not follow those rules, and I want to make sure we address it before we wrap up.
00:45:09.740 Inflammatory breast cancer is a rare type of cancer, roughly 1 to 5 percent of all breast
00:45:15.140 cancers, but it is aggressive and it does not present the way most people expect breast cancer
00:45:20.500 to present. There is often no discrete lump. Instead, what you may notice is rapid swelling
00:45:26.100 or heaviness of the breast, redness or rash, warmth, or changes in skin texture or thickness.
00:45:32.900 Because these symptoms can look like skin irritation or something else relatively benign,
00:45:37.980 diagnosis is frequently delayed. And this is the key point. Normal screening mammography does not
00:45:44.160 rule out inflammatory breast disease. These cancers may not be visible on mammography,
00:45:48.840 so a diagnostic workup is necessary. This is an important reminder that screening tests are
00:45:54.940 designed for women without symptoms. If you notice something new, a lump, skin changes, nipple
00:46:00.940 discharge, pain that does not resolve, do not wait for your next scheduled screening. A recent
00:46:07.700 normal screen does not guarantee everything is fine. Go and get evaluated in person by your
00:46:13.620 doctor. And this applies to men too. I have focused this discussion on women because the vast majority
00:46:19.260 of breast cancers occur in women, but men do develop breast cancer. Because we do not routinely
00:46:25.420 screen men, symptoms are typically the only path to diagnosis. A new symptom in a man should not 0.60
00:46:31.560 be dismissed. It should be evaluated just as it would be in a woman. All right, so where does all 0.99
00:46:37.440 of this leave us? We started this episode with a question. If screening works so well, why are 42,000
00:46:45.960 women still dying from breast cancer every year? Part of the answer is biology. Some cancers are
00:46:51.260 aggressive enough that they will evade even the best screening. But the far larger part, and the
00:46:56.820 one that we can actually do something about right now, is that we are not screening intelligently
00:47:02.040 enough. A quarter of eligible women are not up to date on basic mammography. The vast majority of 1.00
00:47:09.280 women who qualify for MRI are not getting it, and most women have never had a formal risk assessment.
00:47:15.960 The science here is not the bottleneck. The tools exist. The evidence is strong. What is missing is
00:47:23.460 the bridge between what we know and what women are actually doing. And to be clear, this is not an 0.93
00:47:29.580 individual failure. Access to MRI is limited in many areas. Insurance coverage is often tied to
00:47:36.100 the USPSTF recommendations, which do not fully address high-risk populations. Imaging quality
00:47:42.940 varies across centers. All of these system-level factors contribute to underscreening and suboptimal
00:47:49.080 screening. But there is still a great deal that is within your control. So if you take nothing else
00:47:56.160 from this episode, let it be this. First, complete a risk assessment using a validated risk calculator
00:48:02.800 like the one we'll link to in the show notes, so you have a quantitative sense of your baseline risk.
00:48:08.960 Second, find out your breast density from prior imaging or plan to establish it when you begin screening.
00:48:16.240 Third, choose a cancer screening strategy, both modality and frequency, that matches your level of risk and your tolerance for false positives.
00:48:25.360 And fourth, execute that plan consistently over time.
00:48:30.800 None of these steps are complicated, but taken together, they are what separates passive
00:48:35.860 screening from intentional, personalized screening.
00:48:39.200 With the current technology, we cannot reduce breast cancer deaths to zero, but far too
00:48:45.500 many lives are still being lost because we are applying the right tools too late, too
00:48:50.080 inconsistently, or to the wrong people.
00:48:53.000 At least at the individual level, this problem is more solvable than most people realize.
00:48:57.960 If risk is assessed, the right screening strategy is chosen and the screening is actually carried out.
00:49:04.980 Thank you for listening to this week's episode of The Drive.
00:49:08.420 Head over to peteratiyamd.com forward slash show notes if you want to dig deeper into this episode.
00:49:16.180 You can also find me on YouTube, Instagram, and Twitter, all with the handle peteratiyamd.
00:49:21.660 You can also leave us a review on Apple Podcasts or whatever podcast player you use.
00:49:26.920 This podcast is for general informational purposes only and does not constitute the practice of
00:49:32.200 medicine, nursing, or other professional healthcare services, including the giving
00:49:36.040 of medical advice. No doctor-patient relationship is formed. The use of this information and the
00:49:42.360 materials linked to this podcast is at the user's own risk. The content on this podcast is not
00:49:47.980 intended to be a substitute for professional medical advice, diagnosis, or treatment. Users
00:49:52.660 should not disregard or delay in obtaining medical advice from any medical condition they have,
00:49:57.580 and they should seek the assistance of their healthcare professionals for any such conditions.
00:50:02.760 Finally, I take all conflicts of interest very seriously. For all of my disclosures and the
00:50:07.800 companies I invest in or advise, please visit peteratiamd.com forward slash about where I keep
00:50:15.060 an up-to-date and active list of all disclosures.