Skip to Content
24Sep 2012

Did researchers just identify four new types of breast cancer?

Pink baseball bats

[Update: A colleague pointed me to the NPR story by Richard Knox, which says, "Scientists have known for a while that breast cancer is really four different diseases..." And in the comments below, I added a link to a JAMA study that described the four types in 2006. So I'm now more confident saying that reporters who said researchers "identified four types of breast cancer" were incorrect.]


Sorting out what is new about the breast-cancer-genetics study published Sunday in Nature is proving to be a daunting task. Reading the coverage, however, it seems most of the press made a mistake. 

Here's what I know from the press release issued by the National Institutes of Health: 

One subtype of breast cancer shares many genetic features with high-grade serous ovarian cancer...The findings suggest that the two cancers are of similar molecular origin, which may facilitate the comparison of therapeutic data for subtypes of breast and ovarian cancers...Analyses of genomic data have confirmed that there are four primary subtypes of breast cancer, each with its own biology and survival outlooks.

The findings come from a detailed genetic analysis of breast cancer genes as part of what's called The Cancer Genome Atlas, a government project that has already analyzed the genes of certain ovarian, colorectal, lung, and brain cancers.

What I read in much of the copy, however, was not included in the release, nor could I find it in the study itself. 

Gina Kolata at The New York Times reported that researchers "have identified four genetically distinct types" of breast cancer. But the press release says researchers "described new insights into the four standard molecular subtypes" of breast cancer. Kolata's lede is apparently wrong, and so is the headline that reflects the error. A Google search turned up many stories that talked about the identification of four types of breast cancer--all, apparently, incorrect. Victoria Colliver at The San Francisco Chronicle seems to make the same mistake, if it is a mistake, in a lede that says researchers "have redefined the disease into four main classes." So did the AP.

Liz Szabo at USA Today avoids that phrasing, writing that scientists "have finished mapping virtually all of the genetic mutations in breast cancer," which may be overreaching a bit. The press release says the researchers identified a lot of genetic mutations, but it doesn't say that the study covered "virtually all" of them. She also writes that the resarch "could soon change the way patients are treated." In contrast, Harold Varmus, the head of the National Cancer Institute, says in the release, "This treasure trove of genetic information will need to be examined in great detail to identify how we can use if functionally and clinically." In other words, I think, it will not affect the way patients are treated any time soon.

A lede that noted the similarity between ovarian cancer and one kind of breast cancer would have been safe. So would a lede that said resarchers made a major advance in understanding the genetics of breast cancer, which might one day lead to better care for patients.

-Paul Raeburn







Our disagreement seems to be over the meaning of "soon."

You do indeed attribute the assertion that the findings "could soon produce real benefits" for patients. But in your comment here, you focus on the ovarian cancer link, which was, in my view, the real news. And you did not lead with that, nor with the idea that the findings could soon help only some patients. You led with "soon change the way patients are treated."

In your comment, you say the findings give doctors a rationale to conduct a clinical trial. The results of that would likely be years away. I would not call that "soon."

It was a good story. As you can see from my post, I was much tougher on others. But my judgment is that the story would raise false hopes for too many patients. It's a judgment call; we can disagree.



Paul, I wrote the USA TODAY version of the breast/ovarian cancer news this week.  I don’t understand your critiques, though. Are you saying that our story is off-base only because it differs from the official press release? Gosh, that’s a good thing. I actually read the study, interviewed two of the authors, the NIH project director and a couple outsiders, plus lots of patients with triple-negative disease. A researcher was the source of the quote about mapping virtually all the mutations.  I wrote: "Scientists announced Sunday that they have finished mapping virtually all of the genetic mutations in breast cancer, an effort that could soon change the way patients are treated and eventually help researchers develop better treatments. 'The catalogue of human breast cancers is nearly complete,' says study co-leader Matthew Ellis of the Washington University School of Medicine in St. Louis. "

Also, it’s really NOT far-fetched to say this finding could change treatment quite quickly. As we say in my story, above, that’s NOT  because there will be a new treatment tomorrow. And we certainly never mention "cure."

Instead, it’s because this finding could help these women make better use of alternative regiments that are already approved.

 Previous research indicates that women with triple-negative disease don’t seem to benefit from anthracyclines, which are currently a mainsty of their treatment, but which are very toxic. Because their tumors are genetically very close to ovarian tumors, researchers told us that it would make sense to see if these women could better on the chemo regimen currently used in ovarian cancer. One doesn't switch chemo regimens willy-nilly. But this data gives doctors a rational for at least conduting a clinical trial to see if the ovarian regimen might work better. Why? Because the current therapy for triple-negative cancer is very, very toxic, involving anthracyclines that are known to damage the heart and even cause leukemias or "pre-leukemias." We want these women to live. And if they do survive breast cancer, it would nice if they didn't later get leukemia.

Ovarian cancer docs stopped using anthracyclines about a decade ago with no decrease in effectiveness, doctors told us.  Now that docs know that triple-negative cancers are so similar to ovarian cancers, there is a good reason to believe that their tumors would respond to chemo in a similar way.

One key point to make to readers is that genomic research can be used to help target the best currently available treatments, getting the right drug to the right patient at the right dose and best time. Obviously, everyone wants a cure. And scientists are unlikely to develop another Herceptin-like targeted therapy without this sort of basic knowledge. But until we get to a cure, which may never come, it looks like women could benefit, fairly soon, from a slightly less toxic treatment that is no less effective.

I take my responsibility to cancer patients very seriously. I had patients in mind when writing this story, with every line that I wrote.

Here is my attempt to sort it out at The Cancer Chronicles:

Kenneth: "Identify" and "confirm" are not quite the same thing, I think.

Victor: Thanks for reminding us to look at the study as well as the press release, which of course I did (and I linked to the study).

Here's why I think Sunday's paper did not "identify" four new categories of breast cancer:

1. From the abstract: It says the authors "demonstrated the existence of four main breast cancer classes," but it also says the research "provided key insights into previously defined gene expression subtypes." It's unclear to me what that means, without interviewing them. But the coverage should have clarified that.

2. From the press release: The researchers "described new insights into the four standard molecular subtypes." 

3. And from a 2006 JAMA paper: "Gene expression analysis has identified several breast cancer subtypes, including basal-like, human epidermal growth factor receptor-2 positive/estrogen receptor negative (HER2+/ER–), luminal A, and luminal B." 

Therefore, I don't believe the authors of the new study "have identified four genetically distinct types of cancer," as Kolata and others report. 


Before we go too far with this slam on the coverage, we should start with the paper itself, not the press release.  In line with Chang's comment, here's the abstract: We analysed primary breast cancers by genomic DNA copy number arrays, DNA methylation, exome sequencing, messenger RNA arrays, microRNA sequencing and reverse-phase protein arrays. Our ability to integrate information across platforms provided key insights into previously defined gene expression subtypes and demonstrated the existence of four main breast cancer classes when combining data from five platforms, each of which shows significant molecular heterogeneity. Somatic mutations in only three genes (TP53, PIK3CA and GATA3) occurred at .10% incidence across all breast cancers; however, there were numerous subtype-associated and novel gene mutations including the enrichment of specific mutations in GATA3, PIK3CA and MAP3K1 with the luminal A subtype. We identified two novel protein-expression-defined subgroups, possibly produced by stromal/microenvironmental elements, and integrated analyses identified specific signalling pathways dominant in each molecular subtype including a HER2/phosphorylated HER2/EGFR/phosphorylated EGFR signature within the HER2-enriched expression subtype. Comparison of basal-like breast tumours with high-grade serous ovarian tumours showed many molecular commonalities, indicating a related aetiology and similar therapeutic opportunities. The biological finding of the four main breast cancer subtypes caused by different subsets of genetic and epigenetic abnormalities raises the hypothesis that much of the clinically observable plasticity and heterogeneity occurs within, and not across, these major biological subtypes of breast cancer.

Am I missing something? From the press release, you quote, "Analyses of genomic data have confirmed that there are four primary subtypes of breast cancer, each with its own biology and survival outlooks."

Which sounds, to me, a lot like, "researchers 'have identified four genetically distinct types' of breast cancer."

Login or register to post comments