Teva’s phase III rejig leaves some unanswered questions

The changes disclosed last week to Teva’s phase III trial with the Mesoblast-licensed CEP-41750/rexlemestrocel-L stem cell therapy in chronic heart failure are ostensibly designed to produce an earlier result in this large event-based study.

But the decision to cut back enrolment by a third and change the interim analysis plan at this stage is highly unusual, since these actions would be expected to raise the statistical hurdles for success. So the question then is what has prompted the revision. One possibility is that Teva may have found a way to improve its chances of an early re-partnering of the programme.

Teva has never been seen as the most likely party to commercialise this product in the event of success, given its respiratory/neurology focus, and this will be especially true after planned strategic changes.

Reluctant partner

The Israeli group has been at best a lukewarm development partner for Mesoblast, even before the most recent rediscovery of its roots with the proposed acquisition of Allergan's generics business. Teva originally inherited the development rights to cardiovascular, cancer and neurologic indications for Mesoblast’s mesenchymal precursor cells (MPCs), along with a sizeable 20% equity stake, as a result of its acquisition of Cephalon in 2011.  

The deal left Teva with a contractual commitment to conduct the phase III study. However, it took a year longer than expected to get the trial up and running – it finally started in January 2014. At the time, it was thought that the Israeli company had been stalling and was exploring whether it could sell the asset on.

Indeed, many commentators thought it only did start the study as the lesser of two evils – a decision to commit $150m or more to undertake the CHF trial was preferable to a negotiated handback of rights, with the writedown of the up-front payment and the probable knock-on effect on the value of its equity investment in Mesoblast.

The latest changes could therefore be seen as a way to get to an efficacy/futility result earlier and thus lower the financial risk in the programme, while perhaps providing a new opportunity to secure a partner, especially now that new data suggest that CEP-41750 might be more effective in CHF than was initially thought.


The basic design of the new trial remains the same: recruiting patients with NYHA class II or III systolic heart failure, who will receive a single injection of CEP-41750 directly into the left ventricle or undergo a cardiac catheterisation involving a sham cardiac mapping and cell delivery procedure.

The primary endpoint remains a time-to-event analysis of heart failure-related major adverse cardiovascular events, or HF-MACE. The study was originally powered to show a 25% relative reduction in HF-MACE versus control.

An interim analysis is expected later this year, but the key will be the second analysis, which Teva has changed to a combined interim efficacy/futility analysis now triggered at 50% of the HF-MACE events, and which could occur in mid-2017.

Part of the quid pro quo negotiated with the FDA is that a second, confirmatory study will now be required to support approval; this is expected to be conducted in around 500 patients in an identical patient group using recurrent HF-MACE as the primary endpoint.

Mesoblast said use of recurrent HF-MACE as primary endpoint was supported by a new analysis of the completed phase II trial, where patients treated with CEP-41750 had no HF-MACE over 36 months' follow-up, compared with 11 (of 15) recurrent HF-MACE in the control group (p<0.001). This recurrent HF-MACE analysis will be presented at a future medical conference, possibly the American Heart Association meeting in November.

Mesoblast says this second study will run in parallel. But it is a good bet that Teva is hoping that it does not need to start it until after an interim analysis, so that a future partner will pick up the tab. 

To contact the writer of this story email Robin Davison in London at [email protected] or follow @RobinDavison2 on Twitter

Share This Article