New DCIS study, news release lead to (very) mixed messages: ‘And we wonder why patients get confused’

The following is the first contribution to the blog by Dave Mosher, a deputy editor at Tech Insider who joined the team of reviewers earlier this year. His writing about science, technology, and innovation has appeared in outlets such as WIRED, Popular Science, and Scientific American

mammogram scanIt’s an all-too-common theme in the theater of health news: scientists publish a nuanced but potentially field-rattling medical study, journalists and press officers zero in on too-sharp angles that set aside crucial information and context, and readers and patients get lost in the shuffle.

That’s exactly what happened last week, when Dr. Steven A. Narod, who studies breast cancer at Women’s College Research Institute in Toronto, and his colleagues dropped a doozy of a study in JAMA Oncology.

The work focused on ductal carcinoma in situ, or DCIS. It’s an abnormal collection of cells inside the milk-producing ductwork of a breast, and it constitutes 20% of breast cancer diagnoses by mammogram. About 60,000 American women are diagnosed with DCIS each year, according to the American Cancer Society, and it’s often referred to as “stage 0 breast cancer.”

But whether or not DCIS should have “breast cancer” attached to it by default is a controversial topic. The vast majority of DCIS is contained in the breast’s ductwork and hasn’t broken out to become invasive breast cancer. Yet, despite not being cancer, it gets that “stage 0” label.

Does it mean cancer? Pre-cancer? A risk of cancer? Something else? Even doctors struggle to agree. Yet they routinely recommend surgery, followed by radiation and hormonal therapy, as part of a “better safe than sorry” approach, even though the efficacy of that standard of care — which is often invasive and traumatizing — is not well-understood.

Hence, the reason Dr. Narod et al. ask with their study: What happens to women after diagnosis and treatment? What works? What doesn’t? And should we change up the standard of care accordingly? The researchers took the medical data of 108,196 women diagnosed with DCIS, then analyzed it for their survival (among other factors) at 10 and 20 years after diagnosis.

The New York Times titled its article “Early-Stage Breast Condition May Not Require Cancer Treatment” and took this focus: “Patients with this condition had close to the same likelihood of dying of breast cancer as women in the general population, and the few who died did so despite treatment, not for lack of it.” (Note: NYTimes later changed its headline to “Doubt Is Raised Over Value of Surgery for Breast Lesion at Earliest Stage.”)

But a news release published by Dr. Narod’s institution declared “Ductal carcinoma in situ carries a higher risk of death than previously thought,” and it took a decidedly different tack than the NYTimes‘ angle: “Women diagnosed with ductal carcinoma in situ (DCIS) are twice as likely to die from breast cancer compared to the general U.S. population, according to a new study led by Dr. Steven Narod.”

Time magazine may have picked up this angle in the news release and run with it. Their headline declared the study “A Major Shift in Breast Cancer Understanding” but focused on the dangers of DCIS not being “as benign as doctors once thought.”

Where it starts to get really confusing, however, are prominent quotes from Dr. Narod.

He told the NYTimes, “I think the best way to treat D.C.I.S. is to do nothing.”

But in his own news release, he says, “Our work shows that DCIS has more in common with small invasive cancers than previously thought. In these cases, we’ve found that there’s an inherent potential for DCIS to spread to other organs.”

In effect, Dr. Narod seems to be saying that most women have no need for surgery, radiation, or other treatment for DCIS — then he turns around and calls it a potentially dangerous cancer that we aren’t addressing.

Dr. Michael Wosnick, who runs a blog called “Cancer Research 101,” summed it up nicely on his Facebook page: “And we wonder why patients get confused.”

So, what gives? The devil, as always, is in the details.

One key point, which a lot of news coverage glossed over, is that the study focused on progression of DCIS to invasive and deadly breast cancer. As it turns out, breast cancer killed 956 (out of 108,196) women with DCIS over 20 years. But of those, less than half (41.3%) experienced an in-breast invasive recurrence prior to death and 54.1% did not experience an in-breast invasive recurrence prior to death. This begs an important question: How could the latter group have died of breast cancer without having an in-breast invasive recurrence? Doesn’t DCIS need to progress to an invasive breast cancer before it can become deadly?

We reached out to Dr. Deanna Attai, president of the American Society of Breast Surgeons, and one of’s editorial contributors, for some help interpreting this confusing finding, especially since Dr. Attai has focused on the issue of surgery and DCIS on her own blog. Dr. Attai emphasized that women don’t die from cancer in the breast, even if it’s invasive. Death results from metastatic disease (cancer that spreads elsewhere in the body), she added, so ALL patients that died had metastatic disease. The patients in question here just had metastatic invasive cancer without any documentation of invasive cancer in the breast. This could mean a few things, she said:

1. The original DCIS actually had an invasive component that wasn’t detected at initial pathology. It is well known that only a small fraction of an entire pathology specimen is evaluated. If there was truly an invasive component not detected, would that have led to a recommendation against systemic chemotherapy or hormonal therapy, leading to a higher rate of metastatic disease? There’s no way to know. But this was Monica Morrow’s comment in the NYT piece. There’s no way to know how many cases this might have accounted for. It’s important to note that this is not sloppy pathology. It’s highly impractical to examine every micron of a specimen.

2. Residual cancer cells left after treatment (we know radiation doesn’t kill everything) became invasive before being detected on a mammogram and then metastasized. It’s a common misconception that we can detect all cancers. We pick up most, but not all, and invasive cancer doesn’t have to be a lump. At some point, it’s just a few cells. So for the patients who developed metastatic disease without detecting invasive cancer in the breast, this could simply mean that the cancer was there but not detected. The more common situation is that we detect an invasive recurrence and subsequently the patient is diagnosed with metastatic disease. Metastases without a new primary or recurrent lesion in the breast isn’t something I’ve seen commonly, but this study suggests it’s not uncommon. In this database, radiation reduced the risk of locally recurrent (in the breast) or new primary cancer, but did not prevent death. So on some level, the cancers that are bad are bad — no matter what we do.

There are a few other crucial nuances:

Black women with a DCIS diagnosis face a lower survival rate (94%) over 20 years than other ethnic groups (97%); women under age 35 faced a mortality rate 17 times higher than average within 9 years of diagnosis; and of those women who did develop invasive breast cancer, treatment with surgery or radiation helped prevent recurrence but ultimately didn’t seem to help prevent death by breast cancer. (Let that sink in for a moment.)

In a JAMA Oncology editorial that accompanied the paper, two researchers not involved with the study — Dr. Laura Esserman and Christina Yau of the University of California, San Francisco — tried to parse these details and, in the process, made some bold statements about the future of healthcare.

Essentially, they write, 80% of women who are diagnosed with DCIS, and increasingly so (thanks to improving mammogram image technologies), likely don’t need surgery, radiation, drugs, or any interventional therapy. Instead, they and their doctors should keep an eye on the situation. The other sub groups — women diagnosed under 35, black women, and a few others — should pursue the current standard of care.

Here’s Dr. Attai’s analysis of the editorial’s main message:

Esserman focuses on avoidance of overtreating … while acknowledging that high grade disease … in young women is different and should be treated as per standard. She also pushes for more research to understand the biological behavior and diversity of DCIS which she has been saying for years. For quite a while she’s been ahead of her time but many are starting to agree with her.

Suzanne Hicks, a cancer survivor and another editorial contributor, also found Dr. Esserman’s analysis sound:

Specifically, she speaks to the usefulness of the DCIS [diagnosis] in focusing on prevention of fatal breast cancers […] identifying those most at risk, re-evaluating the best means of prevention of lethal [breast cancer], incorporating current information about the risks as well as benefits of radiation options, and forming a structured evaluation of DCIS within the larger field of invasive cancer. What are the treatment/prevention alternatives and their value, vs. surgery and radiation? And for which patients?

Hicks’ view of press coverage, however, wasn’t as shiny:

The NY Times article could have better clarified the term [DCIS] and been less alarmist in its generalizations. For instance, the statement that “some women who died of breast cancer ended up with the disease throughout their body without ever having it recur in their breast” ignores the clearly stated fact in Narod that “The SEER registry records multiple primary cancers but not recurrences.” The most startling question in Kolata’s piece is an old question I thought we’d moved past: “If treatment does not make a difference, should women even be told they have the condition?”

Alicia Staley, a cancer survivor and prominent advocate in the breast cancer social media community, called the messages surrounding DCIS “very confusing” for women who receive this diagnosis.

There never seems to have been a true consensus on DCIS. Is it cancer? Is it pre-cancerous? It’s considered non-invasive, but upon recurrence it can become invasive. There so many conflicting information sources on DCIS, it’s hard to help people that are newly diagnosed find the best information to make a treatment decision. DCIS is a complex situation. Sweeping generalizations make it harder for patient advocates to help their communities navigate through information presented in this story. It does a disservice to all those working hard to uncover the best possible information sources for their communities.

Other coverage, including ABC World News, glommed onto the “stage 0 breast cancer” label to miscast DCIS as “the earliest stage of breast cancer.” It’s tempting to group all cancers into one bucket, like this statement does, but in reality every cancer and its progression is unique for every patient. In the case of DCIS, fewer than 1% of the women in the study died from breast cancer over two decades.

From the perch of a journalist-slash-editor, it’s not hard to understand how such mistakes — factual, contextual, and otherwise — routinely get made in health news. The media industry is a high-pressure vessel that incentivizes publishing as many stories as quickly as possible to garner reader attention and advertiser impressions and revenue. The data also suggest news consumers don’t spend a lot of time reading, so packing in extra information is frowned upon.

To some extent, this is helpful: A push for speed and brevity makes stories economical and helps readers get the latest news in the timeliest fashion. But unlike stories about the next iThing or famous celebrity’s off-camera moment, those about health carry an entirely different gravity: influence on medical decisions, however slight or strong, and ultimately the length and quality of peoples’ lives.

We owe it to readers to slow down, parse the information overload, ask the right questions, and deliver something truthful and useful.

Editor’s Note: Mosher would like readers to know that one of his writers at Tech Insider also covered the recent DCIS study, editorial, and related news coverage, titling the piece (without his involvement) “We might be treating one of the most common cancers in completely the wrong way.”



Listen to our new podcast on the DCIS dilemma, with Dr. Laura Esserman, surgeon and breast cancer researcher.

You might also like

Comments (6)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Donna Pinto

August 26, 2015 at 12:50 am

Thank you for this excellent review of the extremely mixed messages and media spin. Have a listen to my interview (from a patient perspective on DCIS overtreatment) on KPBS in San Diego along with a local oncologist: Be sure and listen to the longer radio interview as well — link just below. I was also interviewed in a front page article in the San Diego Union Tribune: Thanks again for a great article! Donna Pinto

Dawn Burke

August 27, 2015 at 7:52 am

I was diagnosed nearly 2 years ago with low grade DCIS and was told I needed to have a mastectomy. Something did not sit right with me, so I started researching and found exactly what this study says. That basically DCIS may or may not turn into invasive cancer and it was unlikely that I would die from it anytime soon. I decided to decline the standard of care treatments and cleaned up my life instead. As of my last MRI I have been able to reduce the area of DCIS by 27%. I plan to continue my healthy lifestyle and use of alternative methods and monitor. After reading this study I was so glad I did not have a mastectomy 2 years ago.

Ignatius Brady

August 28, 2015 at 11:15 am

This is a wonderful piece. Thank you for trying to straighten out some of the mess. The conclusions in the original article in JAMA Oncology last week do not match the data as presented. It’s impossible to understand how they wound up the article advocating more treatment and more worry about DCIS.
H. Gilbert Welch covers this topic very clearly in a 2006 book titled, “Should I be Tested for Cancer? Maybe Not and Here’s Why.” Which I would encourage anyone reading on this issue or similar screening issues to consider:


August 28, 2015 at 1:27 pm

What ARE the right questions to ask?

Annette Bar-Cohen

August 28, 2015 at 3:51 pm

Thanks for confirming how confusing, inconsistent and at cross purposes some of the data and conclusions are of this otherwise important study. It does reenforce the growing body of evidence that mammography creates harms along with benefits and our focus must move to prevention and preventing women fom dying of metastasis Another take on the article and on this issue is by Fran Visco, president of the National Breast Cancer Coalition which is focusing urgently on these two areas to reduce mortality from breast cancer.


August 31, 2015 at 9:09 am
The Natural History of Invasive Breast Cancers Detected by Screening Mammography
From 2008, note that 22% of invasive breast cancers resolved on their own! It appears that they spontaneously regressed. I know…heresy. These are invasive cancers, not DCIS. This is an area that deserves a look more consideration: what DOES make cancer go on to metastases and kill and, conversely, what makes it seemingly disappear in some?