Shooting down the Star’s story on breakthrough breast cancer drug

The Toronto Star reported, “New breast cancer drug heralded as breakthrough.”

Let’s be clear about this at the outset:  we’ve interviewed countless women with breast cancer and we share their yearning for breakthroughs.  But we also share the healthy skepticism of many of the breast cancer survivors we’ve met about jumping to early conclusions about promising approaches.

In that regard, this headline didn’t pass the sniff test, so we asked ace story reviewer/guest blogger Harold DeMonaco to take a whiff.  Here is what he wrote:


Superlatives are rarely appropriately employed with regard to medical research.  So when a story employs a series of them: “Breakthrough…Rivals the results shown by the miracle drug Herceptin…Transforms the disease into a chronic treatable condition,” in describing the results of a Phase 2 study involving 190 subjects presented in a 15 minute presentation at a scientific meeting, I had to read on.

Breast cancer is a terrible disease and while tremendous steps have been made in the past 10 years, there are still gaps in medicine’s ability to adequately treat some patients.  Newer and hopefully less toxic drugs would be a welcomed addition to the clinician’s armamentarium.  And they would certainly be welcomed by patients and their families. All the more reason for balanced and thoughtful reporting on a potentially beneficial approach to improving remission rates in recurrent disease.  But, that is not what this story provides in my view.

Let’s put this into perspective. The only information available at the moment is from a 15 minute presentation by the senior author of the study at a session of this week’s San Antonio Breast Cancer Symposium.  The results have not been subject to peer review and detailed information is not available at the moment.

Here’s what the related press release provided:

In the first part of this two-part, phase II study, Finn and colleagues randomly assigned 66 postmenopausal women with metastatic estrogen receptor (ER)-positive breast cancer to either the combination of PD 0332991 and letrozole or to letrozole alone. The second part of the study involved 99 patients with ER-positive cancers determined by screening to have certain genomic alterations, specifically cyclin D1 amplification and/or p16 loss, according to Finn.

Results showed that progression-free survival was 26.1 months for those in the combination arm versus 7.5 months for patients treated with letrozole alone. There was also a 45 percent response rate with the combination treatment versus 31 percent with letrozole alone in patients with measurable disease. The clinical benefit rate was 70 percent with the combination treatment and 44 percent with letrozole alone.

Contrary to what the story says, the 165 patients enrolled were from two potentially distinct groups, those with and without genomic alterations.  This is important because the results of the study are not parsed.  Were the positive results seen in both groups of women or primarily in one group?  Is the potential population only those with a specific genomic alteration or all women with ER-positive disease? Thankfully,  Dr. Dr. Jawaid Younus’ comments speak to the issue.

Only after wading through 522 words of optimism of this 806 word story do we find any sense of balance in the story with the remarks of Dr. Younus.   He put the report into the correct context; “This is definitely exciting…It is certainly promising…But, I think it is too early…You don’t know if this humongous difference is going to be sustained.” (we added the emphasis)

The story notes, “Phase 3 trials that could send it into wide-spread clinical use.”  That is quite true but there are no guarantees.  Only about one out of every three drugs that show promise in Phase 2 testing make it through Phase 3 clinical trials.  Between 2004 and 2010, new drug applications were filed for 55% of drugs completing Phase 3 trials and only 80% eventually made it to market.  The road to approval is long and bumpy and approval is not a forgone conclusion.

As I have said on numerous occasions, I am not a journalist so my comments should be viewed in the context of a semi-informed reader.  I appreciate that what looks like promise on one day can look like a complete failure later.  I am not certain what the average reader’s take-away message is from the constant rise and fall of expectations seen in the media for “breakthrough” technologies.  The results of the Phase 2 trial are potentially important.  All the more reason why a story should be balanced and complete.


One observer I follow online wondered “how such a small, preliminary study can establish that ‘the drug is clearly not very toxic’ when it’s well known that toxicities often show up later when drugs are put into widespread use.”

In the US, HealthDay reported on the study.  Flaws that jumped out:

  • They reported, “In one study, of nearly 200 women…”  That’s technically accurate but misleading.  There were 160 people in two groups – only roughly half getting the treatment in question.  So “nearly 200” demands clarification.
  • They reported, “They found that women on the combination had a much longer ‘progression-free survival’ “.  What does that mean?  See this post on the limitations of progression-free survival.  Was this explained to readers of this story? No.


Follow us on Facebook,

and on Twitter:

You might also like

Comments (4)

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


December 6, 2012 at 10:54 am

Well, actually, it is not true that “…progression-free survival was 26.1 months for those in the combination arm versus 7.5 months for patients treated with letrozole alone.” This statement is true of the MEDIAN survival time (most likely, we don’t have details). Why does it matter? An uninformed reader may think “I can expect 26.1 months survival etc etc with the combination.” No, in fact, one can expect a random survival time taken from a distribution which, in this clinical trial, delievered an observed median of 26.1 months.

    Harold DeMonaco

    December 6, 2012 at 12:36 pm


    Many thanks for the correction. I can envision my statistics professor from graduate school rolling her eyes at my omission.