Last week, Roche announced that the FDA was unwilling to accept T-DM1’s accelerated approval filing for breast cancer, sending Immunogen’s (IMGN) shares tumbling almost 40%. The filing was based on impressive results from a single arm phase II trial in 3rd line HER2 breast cancer. Typically, gaining regulatory approval requires a large, randomized phase III study but in cases of highly unmet need where patients are bereft of effective treatment options, the FDA may consider approval in a limited patient population.
Surprisingly, it turns out that the FDA was unwilling to even consider approval for T-DM1 based on the phase II. T-DM1’s data set is one of the most impressive ever generated in late stage breast cancer, which is why the market (myself included) was confident T-DM1 would win FDA approval. The reason for the FDA’s response as appears in the official press release was that “all available treatment choices approved for metastatic breast cancer, regardless of HER2 status, had not been exhausted in the study population”. In other words, T-DM1 did not qualify for accelerated approval based on the phase II results. This got people scratching their heads as no further details were provided by the FDA, Roche or Immunogen.
Regardless of the reasons for the rejection (to be discussed below), it is important to note that it had nothing to do with the clinical data, which remains promising. The next attempt to get T-DM1 approved will be based on a large randomized phase III trial (the EMILIA study) in mid 2012. This is a major setback for T-DM1 and Immunogen, which is eligible for a 4-5% royalty rate, but it has no real impact on T-DM1’s long term potential.
Defining available treatment options
The FDA’s statement that not all available treatment options had been exhausted probably refers to a chemotherapy regimen, as the patients in the trial had progressed on both Herceptin and Tykerb, which are HER2 targeted therapies. According to the data presented last December, patients in the phase II trial had prior therapy with a median of 7 agents, including Herceptin and Tykerb. All but one of the patients had been treated with the three most commonly used chemo drugs (or classes of drugs): An anthracycline, a taxane, and Xeloda. So what were then the additional treatment options the FDA was referring to?
There are many chemo drugs that can be used in breast cancer, some have been approved for decades, however, patients rarely receive all of them. Therefore, it is unrealistic to expect every patient to receive every drug that was ever approved. If I had to guess, the single drug the FDA wanted to see as a prior therapy was BMS’s (BMY) Ixempra, which is (to my knowledge) the only drug that is specifically approved in breast cancer patients who relapse after Xeloda.
Ixempra was approved in 2007 for Xeloda-pretreated patients as a single agent. Intriguingly, its approval was very similar to that pursued by Roche for T-DM1. The filing was based on a single arm phase II study in 126 patients who had previously been treated with an anthracycline, a taxane, and Xeloda. The response rate in 113 response-evaluable patients was 12.4% and the median progression free survival (PFS) was 3.1 months. Of note, BMS pursued approval of Ixempra in combination with Xeloda for an earlier treatment line in the same filing, which probably increased chances of approval.
If the FDA had Ixempra in mind when it decided to issue the “refuse to file” response, this makes the decision even more puzzling. T-DM1 led to a response rate of 32.7% and preliminary PFS of 7.3 months (final PFS should be higher as this figure represents the first 50 patients who experienced disease progression). Although T-DM1 and Ixempra have been evaluated in different trials in non-overlapping patient populations (only 7% of patients in the Ixempra study were HER2 positive), T-DM1 looks much more effective than Ixempra in HER2 breast cancer patients. T-DM1 also seems to have a better safety profile, which could be an issue for heavily pretreated patients who are often frail and more sensitive to side effects.
The FDA’s dilemma
The FDA probably faced a dilemma: On the one hand there was a highly efficacious and safe drug in a well defined patient population. On the other hand were objective criteria that had to be met in order to receive tentative accelerated approval. Technically, the patients in the T-DM1 study were not last line patients because they were eligible for single agent Ixempra. Nevertheless, Ixempra’s efficacy seems limited, which may explain the drug’s modest sales (~$110M per year) and the fact that it was not approved in Europe.
As the gatekeeper of healthcare in the US, the FDA must be very protective and prevent ineffective or unsafe drugs from reaching the market. This time, it seems the FDA chose (or had to) follow the rules rather than bend them by not giving T-DM1 an opportunity to be evaluated by an expert panel. Unfortunately, by sticking to the formal rules, the agency lost sight of the patients’ best interest, depriving them of access to an excellent drug. Instead, patients are left with an ineffective, toxic drug they are not likely to receive anyway.
Putting investment aspects aside, the real challenge right now is getting T-DM1 to patients who need it as part of a compassionate use program. 3rd line patients are not eligible for the majority of ongoing clinical trials with T-DM1 and at the moment, there are only 9 sites across the US that offer the drug as part of an expanded access study, according to clinicaltrials.gov. It is the moral obligation of all parties involved to provide 3rd line patients with T-DM1 as an experimental drug.
Tomorrow’s post will discuss Immunogen and T-DM1’s prospects following the rejection.