- Bristol-Myers Squibb/AstraZeneca received a complete response letter for dapagliflozin from the FDA, requesting “additional clinical data to allow a better assessment of the benefit-risk profile.”
Last night, BMS/AZ announced the receipt of a complete response letter from the FDA for their SGLT-2 inhibitor dapagliflozin, in which the agency requested “additional clinical data to allow a better assessment of the benefit-risk profile,” including data from ongoing studies (see below), and potentially data from additional trials. While BMS/AZ’s press release is vague, we suspect that the FDA is largely concerned with the drug’s potential bladder and breast cancer risk, and to some extent its potential hepatotoxicity, the paucity of data for specific subgroups, the potential use of dapagliflozin in renally impaired individuals, and the lack of PK/PD data – concerns voiced by members of the Endocrinologic and Metabolic Drugs Advisory Committee (EMDAC) during their discussion of the benefit-risk profile of dapagliflozin in July. The downsides, of course, of trying to get a drug approved with a new mechanism. As a reminder, the EMDAC ultimately voted 6-9 against approval of dapagliflozin – those who voted for approval emphasized that a post-marketing study was the only realistic way to assess dapagliflozin’s potential cancer risk, while many of those who voted against approval requested additional pre- approval data, without providing concrete recommendations on how to characterize these risks in apre-approval setting (see our July 19, 2011 Closer Look at http://www.closeconcerns.com/knowledgebase/r/a12eacdc for more details). Notably, during the discussion, Dr. Sanjay Kaul (Cedars Sinai, Los Angeles, CA) estimated that a large trial with 30,000 to 100,000 patients would be needed to properly assess the cancer uncertainties, and would thus be more appropriate in a post-marketing setting. Dr. Zan Fleming, a respected expert on regulatory matters and former FDA head of drug approvals has emphasized his view that the cancer signal is not a reason to hold up dapa – see his detailed comments below.
While BMS/AZ have stated that they remain committed to the development of dapagliflozin, we suspect that the two companies would reconsider if in the most extreme case they are asked to undertake such an onerous study in a pre-approval setting. We hope to hear more details from both BMS and AZ regarding the complete response letter and next steps i.e., what trials or additional data are needed, whether a stronger REMS program will suffice to support post-approval outcomes studies, etc. during their respective 4Q11 financial updates on January 26 and February 2. As a sidenote, we assume a stronger REMS program would not be sufficient, based on the additional CRL sent. As a reminder, dapagliflozin has also been submitted for approval in Europe, Brazil, India, China, and Russia. Given a late 2010 filing in Europe, we expect to hear the European Medicines Agency’s (EMA) decision.
While the CRL for dapagliflozin increases the effort, time, and resources BMS/AZ will need to invest to see their SGLT-2 inhibitor through approval, the decision does not necessarily bode negatively for the drug class as a whole. On the contrary, this piece of news could even be viewed positively by other companies in the field, as they now could have the opportunity to have the first SGLT-2 inhibitor to market, especially since cancer concerns do not appear to be a class effect thus far. Or, they could be further ahead than they had expected. Other selective SGLT2 inhibitors in development include J&J/Mitsubishi’s canagliflozin (phase 3), Astellas/Kotobuki’s ipragliflozin (phase 3 in Asia), BI/Lilly’s empagliflozin (phase 3), Chugai’s tofogliflozin (phase 3 in Asia), Pfizer’s PF-04971729 (phase 2), Taisho’s luseogliflozin (TS-071; phase 2), and BHV Pharma/Kissei’s Remogliflozin Etabonate (phase 1). Additionally, Lexicon is developing a dual SGLT-2/SGLT-1 inhibitor (LX4211; phase 2) and GSK/Dainippon/Kissei are developing a selective SGLT-1 inhibitor (GSK-1614235, phase 1).
- While BMS/AZ’s press release is quite vague, we suspect that the FDA is largely concerned with the drug’s potential bladder and breast cancer risk. Encouragingly, even though the risk ratios of bladder and breast cancer with dapagliflozin were four-to-five-fold higher than placebo (but not statistically significant), no clear mechanism could explain these observations, and several members of the EMDAC pointed out that detection bias could have been at play – mammograms are easier to conduct when obese patients lose weight, hydration could affect imaging, and increased frequency of urinary tract infections with dapagliflozin treatment may have led to increased follow-up analyses and greater detection of hematuria.
- A search of clinicaltrials.gov identified 14 active or currently recruiting trials for dapagliflozin. Although it is still unclear which of these trials the FDA may have requested data from, notable trials include: 1) a 104-week phase 2/3 trial examining dapagliflozin in people with type 2 diabetes and moderate renal impairment (NCT00663260); 2) a 52-week plus 156-week extension phase 3 study examining dapagliflozin in people with type 2 diabetes inadequately controlled on metformin (NCT00660907); 3) a 24-week plus 80-week extension phase 3 study examining dapagliflozin in people with type 2 diabetes and CV disease (NCT01042977); 4) a 24- week plus 80-week extension phase 3 study examining dapagliflozin in people with type 2 diabetes, CV disease, and hypertension (NCT01031680); 5) a 52-week phase 3 study examining dapagliflozin in people with type 2 diabetes on a background of metformin and a sulfonylurea (NCT01392677); and 6) a PK/PD study examining dapagliflozin in people with type 1 diabetes (NCT01498185A).
- In a conversation with regulatory expert Dr. Zan Fleming, CEO of Kinexum and former head of diabetes review at FDA, he emphasized that concerns about cancer risks related to almost all diabetes drugs (metformin is one of the few exceptions) have been conflated and confused, thereby causing more harm than good. Addressing possible signals of drug-induced increased cancer risk is a very long-term proposition, he believes, pointing out that even known carcinogens or promoters like ionizing radiation, alkylating agents, and smoking have latencies of five to 25 years. Thus, he stressed that the imbalances seen in bladder and breast cancer in the dapagliflozin program are almost certainly not related to drug induced cancer promotion or induction. By contrast, Dr. Fleming says that the most plausible explanations for these imbalances are treatment effects on increased cancer detection rates or “play of chance”.
- As others have, Dr. Fleming also emphasized in our discussion that people with diabetes have far more to fear from the complications of diabetes than from small increased risks of cancer (see The Emerging Risk Factors Collaboration. NEJM 2011;364:829-841). Dapagliflozin is likely to reduce those non-cancer risks, he stressed, which could substantially outweigh small increases in absolute risks of cancers. For drug approvals, he shared that FDA has traditionally taken cancer risks signals into the overall estimate of the drug’s benefit to risk relationship—virtually never as a sole reason for non-approval. For example, statin labels carry descriptions of positive carcinogenicity studies in rodents, he pointed out - we don’t know whether there are any similar warnings for drugs that have had “imbalances” in clinical studies but are doing some research on this front. Overall, and unsurprisingly, Dr. Fleming believes the overall benefit to risk relationship for dapagliflozin is positive. “FDA should certainlyrequire a long term plan to address these signals but cannot meaningfully expect to resolve the signals with data and/or analyses available now or any time soon. In the meantime, the drug should be approved and made available to well-informed prescribers and patients.”--by Vincent Wu, Ben Kozak, and Kelly Close Note: this report was updated with comments from Dr. Zan Fleming on January 20, 2012.
--by Vincent Wu, Ben Kozak, and Kelly Close
Note: this report was updated with comments from Dr. Zan Fleming on January 20, 2012.