Letters Incretins and pancreatic cancer

Investigations editor’s reply to Holt and to Barnett and O’Hare

BMJ 2013; 347 doi: http://dx.doi.org/10.1136/bmj.f4383 (Published 10 July 2013) Cite this as: BMJ 2013;347:f4383
  1. Deborah Cohen, investigations editor1
  1. 1BMJ, BMA House, London WC1H 9JR, UK
  1. dcohen{at}bmj.com

Holt emphasises the importance of balancing risks and benefits of treatment.1 Barnett and O’Hare point out that other well established diabetes drugs are not without risk.2 They also comment that more reassuring safety analyses concerning pancreatitis were not cited, including a Cochrane review; that there is no human evidence of pancreatic cancer; and that the autopsy findings in eight patients on incretins require further validation.

In response, the main point of my article was that important safety data had not reached the public domain and had therefore been excluded from the current debate. These data included evidence of raised pancreatic enzyme concentrations in patients on incretins and unpublished findings from monkey studies performed by two of the drug companies.3 I argued for full transparency of all safety and trial data, and this call has been echoed by the American Diabetes Association and American Endocrine Society, partly in response to my article.4

There is as yet no firm evidence that incretins offer cardiovascular benefit, and market approval was based on the surrogate measure of glycated haemoglobin. The meta-analysis cited by Holt and by Barnett and O’Hare concludes that there is a “possible protection from cardiovascular events.” However, it then says that this “should be interpreted with caution, as those events were not the principal endpoint, the trial duration was short, and the characteristics of patients included could be different from routine clinical practice.”

Over four months, I asked companies marketing incretins questions about the safety and efficacy of the drugs. None said that incretins offered cardiovascular benefits (see questions and answers with the companies at: www.bmj.com/content/suppl/2013/06/17/bmj.f3680.DC1).

Trials of cardiovascular outcomes are under way, and since publication of my investigation one has reported. Saxagliptin did not outperform placebo in terms of cardiovascular benefit.5

The Cochrane review cited by Barnett and O’Hare reviewed 6899 participants in 17 short term clinical trials and did report a low rate of pancreatitis. However, the review’s conclusions pointed to evidence from the US Food and Drug Administration and European Medicines Agency (EMA) of an increased risk of pancreatitis. It also said that the studies included in the review were too short to remove concerns about such risks.6

Barnett and O’Hare do not mention that there is now a clear signal for pancreatitis and pancreatic cancer coming from the world’s three major safety databases—the FDA, EMA, and World Health Organization—as described in the BMJ investigation.3

With respect to the human pancreas pathology study, abnormalities were detected in all eight transplant quality human pancreases from patients taking incretins.7 Both the FDA and EMA are undertaking their own independent reviews of this study. Doctors prescribing these agents and patients using them have a right to know that relevant safety concerns have been hidden from their view.


Cite this as: BMJ 2013;347:f4383


  • Competing interests: None declared.