Statement of Dr. Diana Zuckerman before the FDA Advisory Committee Regarding Dapagliflozin (Farxiga)

By Diana Zuckerman, Ph.D.
December 12, 2013

Thank you for the opportunity to speak today.  I’m Dr. Diana Zuckerman. I’m president of the Cancer Prevention and Treatment Fund.

Our center is dedicated to improving the health and safety of adults and children, and we do that by conducting and scrutinizing research and explaining the findings to health professionals, patients, and the general public.   Our non-profit center does not accept funding from pharmaceutical companies, so I have no conflicts of interest.

Today I am speaking from my perspective as someone trained in epidemiology at YaleMedicalSchool. I also was on the faculty at Yale and at Vassar and conducted research at Harvard, and was a fellow at the Center for Bioethics at the University of Pennsylvania.  I’ve also discussed the data with an expert on our staff who previously worked at NCI, Dr. Anna Mazzucco.  I’m putting together all of those perspectives, as well as having worked for the U.S. Department of Health and Human Services, as a consultant to the Institute of Medicine., and as someone with several close family members with diabetes, one of whom died from the disease..

My concern about this drug is that there are just too many unanswered questions – and those unanswered questions are frightening ones.  That was true when the FDA rejected this application for approval 2 years ago, and it is still true today.

Question #1: How well does it work?

The results indicate that DAPA is NOT effective for patients with moderate to severe renal impairment.  As the FDA has noted, 20-40% of diabetes patients have compromised renal function.

DAPA is better than placebo but does not provide a significant improvement over currently available drugs.  Specifically, glipizide was superior in the short-term and comparable to DAPA at week 52.

So, in the context of these limited benefits what are the risks?

FDA stated that the animal studies conducted could not rule out the possibility that dapagliflozin contributes to bladder cancer.   We completely agree.  Experiments done using cell lines and tumor models which are not implanted in the bladder cannot answer questions about the risks to humans.  Because in humans, changes in the bladder microenvironment or urine flow may be most relevant to carcinogenesis.  FDA has already suggested use of an appropriate animal model which could address these concerns, specifically a “a (BBN) chemically-induced rodent bladder cancer model (4-hydroxybutyl(butyl)nitrosamine).

Why didn’t the company do the right kind of animal study?  That’s an important question to keep in mind.  But, until the correct type of animal studies are done, DAPA must be considered a possible cause of bladder cancer.  Since other effective diabetes drugs are already on the market, and those drugs are more effective for more patients, bladder cancer is an unacceptable risk.

But bladder cancer isn’t the only concern.  What about breast cancer?  The FDA consultant gave several reasons why DAPA is unlikely to have caused the breast cancers that were observed.

These reasons included:

  • short treatment time prior to onset,
  • the decline in incidence risk ratio over time, and
  • the fact that the breast cancers were estrogen receptor positive.


Let’s start with the short treatment time before onset.  We all know that cancer usually takes years to develop.  So, it would be easy to assume – or perhaps hope – that the increase in cancer happened by chance.  But wishful thinking isn’t enough.  We need to know.

When the percentage of women taking hormone replacement therapy for menopause dropped dramatically, the rate of new cancers in the U.S. went down for the first time ever.  Experts in the field agree that this unexpected quick effect was because of the drop in use of hormone therapy.  So, we know from that experience that even a slow growing cancer can be stimulated by a drug.

We challenge the other 2 assumptions as well.  Hormone receptor status is irrelevant because the potential mechanism for this drug to cause breast cancer is unknown.  It is entirely possible that dapagliflozin could act in multiple biological pathways, including hormone receptor signaling, or in a biological pathway that is common to breast cancers regardless of hormone receptor status.

Similarly, long-term biological responses to dapagliflozin and potential feedback mechanisms in breast tissue are also unknown.  Thus, the decline in incidence risk ratio over time doesn’t mean the drug is safe — because the body could respond differently to dapagliflozin over time.

Diabetes and the prior use of hypoglycemic medications were also listed as reasons to doubt potential breast cancer risk, but these factors are common among all patient groups in these trials.

In summary, these arguments cannot rule out the possibility that dapagliflozin causes breast cancer.  The only way to answer this question is with appropriately designed animal models of mammary tumors, or human patients.

Other Safety Issues:

Renal impairment/failure adverse events were associated with dapagliflozin treatment in data based on 13 short-term studies (3.2% in dapagliflozin-treated patients vs. 1.8% with placebo) and the 9-short-term plus long-term studies (6.7% vs. 4.2%) studies.  As many diabetes patients already have compromised renal function, this risk represents a significant health hazard primarily for the patients most likely to receive this drug.

At two different doses, the DAPA patients had elevated LDL levels, while placebo patients did not.  This increase raises concerns that any potential cardiovascular benefits – which are questionable — cannot justify the additional risks.

Lack of Diversity in the Studies

In the entire clinical program, less than 4% of the patient s were African American.  And yet 13% of African Americans have type II diabetes.  They should have been tested if the intent was to approve the drug for all Americans, not just white Americans.

But, given the risks, it would be difficult to justify recruiting African Americans or whites for more research at this time.

When I speak at FDA advisory committee meetings it is always as a scientist, but today I also want to speak on behalf of patients.  My Dad developed diabetes at the age of 90, probably because he took lupron for prostate cancer.   In retrospect, that was probably not a good treatment decision, since prostate cancer is rarely fatal.   It would certainly be ironic if a diabetes patient developed bladder cancer, which is often fatal, or breast cancer, which can be fatal, as a result of taking a diabetes medication that is no more effective, and often less effective, than other diabetes treatments on the market.

It’s hard to imagine a well- informed patient choosing DAPA, but unfortunately, patients are rarely well-informed.   I assume that the physicians on this panel would clearly and carefully inform their patients of the risks of this drug, but in the real world, many doctors won’t be well-informed because the company that makes DAPA is telling doctors very clearly that DAPA is safe and does NOT cause cancer, when the truth is we don’t know, but it might.

We can’t solve this with a black box warning on a label, because many doctors and patients don’t read the labels.  We can’t ethically rely on post-market studies, because it would not be ethical to prescribe this medication with all these unanswered questions.

I sympathize with the companies’ repeated efforts to get this drug approved, but I wonder why they didn’t do the animal study that the FDA recommended.   And I would not want anyone I care about to take a drug with such serious potential risks when so many alternatives are on the market that do not cause cancer.