On Thursday, Roche announced that IMpassion131, their phase III evaluating the combination of atezo plus paclitaxel, failed to hits its primary PFS endpoint in previously untreated patients with PD-L1+ metastatic triple-negative breast cancer (TNBC). But wait, you ask. Didn’t they already get approved in that indication? Isn’t it the standard of care? And on Wednesday, the answer was yes; today, it’s more like, Kind of. Because the approval specified nab-paclitaxel (Abraxane). IMpassion131 tested solvent-bound paclitaxel, better known as Taxol. This result is a reminder that these drugs aren’t necessarily interchangeable.
Atezo + Abraxane was granted an accelerated approval in PD-L1+ TNBC, and I would say up front that I don’t think this result creates any risk with the FDA, though label expansion may turn out to be an issue. But there’s a lot about this update that’s intriguing, particularly as it relates to Abraxane and the implications for the next and newest study in the TNBC series, IMpassion132. For patients with TNBC, the news changes the narrative from atezo as a break-from-the-pack wonder drug to its role in a combo and the importance of the chemo backbone.
The three Roche-sponsored phase III atezo TNBC trials are similar in design, with the primary difference being the chemo component:
- IMpassion130: Subjects receive either atezo + Abraxane or placebo + Abraxane. The four primary endpoints include PFS in PD-L1+ subjects, where the drug was most successful and the patient population for which the drug combo was ultimately approved.
- IMpassion131: Subjects receive atezo + Taxol or placebo + Taxol. The study fails to hit PFS in the PD-L1+ population, the primary outcome and low-hanging fruit of this trial (PFS in the intent-to-treat population follows). Unlike -130 and -132, OS is not included as a primary endpoint.
- IMpassion132: This one is a bit of a free-for-all: as with -130 and -131, they’re enrolling previously untreated patients, but the primary endpoint is OS in the PD-L1+ population (again, followed by OS in the ITT population). Patients will be randomized to receive atezo + chemo or placebo + chemo, and the possible chemo choices are Gemzar, Xeloda and carboplatin. Yechhh.
I’m an Abraxane evangelist and am always happy to see this drug get the respect it deserves, but for it to play out in a P3 versus P3 setting where an aging taxane outshines an immunotherapy is a shocker for even me. And yet we’ve seen data that presaged this outcome as recently as December 2018.
Way back in 2008, a big cooperative group trial started enrolling advanced breast cancer patients to receive Taxol, Abraxane or Ixempra with or without Avastin. These kinds of trials are always so boring, even when the FDA doesn’t yank approval of one of your agents mid-stream. But what came out in long-term follow-up was that in patients with TNBC, there was both a PFS and a survival benefit to giving patients Abraxane. The abstract is here, but this is the table of interest:
I know, right? Abraxane just seems to favor TNBC. As I said earlier, the -131 fail certainly won’t jeopardize its FDA status, but it is a reminder of how shaky the path to approval for atezo/Abraxane for PD-L1+ TNBC was and how critical the response to -130 was, despite the fact that there were no good options for this patient population. As for -132, yikes. It’s not like it’s outside the realm of possibility that atezo will boost OS when combined with Gemzar, but who wants Gemzar first-line? The inclusion criteria try weakly to cover for this by requiring recurrence or metastases occur within 12 months of completing adjuvant therapy with an anthracycline and a taxane, but eh. TNBC patients tend to recur fast.
The design of -132 makes me think a little of what happened with Perjeta in 2012; it was approved in first-line metastatic HER2+ disease, in combination with Herceptin and docetaxel. Since there were tons of patients who had already exhausted the taxanes and, further, wanted fourth- and fifth- and ninth-line access to Perjeta, the people revolted. The label stayed where it was, but guidelines quickly usurped it, and insurers went along. There was no issue getting Perjeta + the drug of your choice in the 15th line of therapy. Knowing the atezo + anything data in TNBC are going to be a lot less impressive than the original Perjeta numbers, Roche may be trying to vet various chemo combos in advance. But that strategy would appear today to be increasingly risky, and given the -131 Taxol situation, I wouldn’t expect to find a better pairing than Abraxane + atezo.