Early last summer, the pecking order in immuno-oncology looked set. Bristol-Myers Squibb had staked out an early lead as its star checkpoint inhibitor Opdivo (nivolumab) jumped ahead of Merck & Co.’s rival Keytruda (pembrolizumab). Roche, AstraZeneca and others appeared to be fighting for third place.

Yet, Opdivo’s surprise failure last August to beat out chemotherapy in first-line non-small cell lung cancer (NSCLC) has given competing drugs a shot at capturing the lucrative indication.

AstraZeneca, in particular, is back in the mix. Once seen as a laggard in immuno-oncology, the British pharma has been taking steps since early 2016 to improve the chances of success for its own checkpoint inhibitor, durvalumab.

Knowing a trial failure could mean being sidelined as an also-ran in immuno-oncology, AstraZeneca is taking full advantage of the flexibility it built into its closely watched MYSTIC study, and changes made to the trial last month show the company is doing all it can to sidestep the pitfalls that did in Opdivo.

Benefit to playing catch-up?

Bristol-Myers made a big, ambitious bet with Opdivo, when the company opted to test the PD-1 inhibitor in patients expressing PD-L1 at levels higher than 5%, instead of the higher threshold used by many competitors. Bristol-Myers hoped to win approval for the broadest indication possible, which would unlock billions in market opportunity.

That approach, however, backfired and Opdivo failed to beat out chemotherapy in the Checkmate-026 study. Merck, on the other hand, was able to secure an approval for Keytruda in first-line lung cancer for the roughly 30% of NSCLC patients expressing high levels of PD-L1 (greater than 50%).

To date, AstraZeneca has been playing catch-up to Bristol-Myers, Merck and Roche. While late entry comes with plenty of costs, AstraZeneca has been able to benefit from the successes and failures of its competitors and change course accordingly.

Modifying MYSTIC

MYSTIC, a Phase 3 study testing durvalumab both as a monotherapy and in combination with the CTLA4 inhibitor tremelimumab, is AstraZeneca’s most important trial in immuno-oncology. Not only is it aimed at first-line NSCLC, but MYSTIC will give investors the first real look at how tremelimumab performs as a combo agent.

At the beginning of February, AstraZeneca modified the trial so it would assess progression-free survival (PFS) and overall survival (OS) in patients with PD-L1 expressing tumors as well as in "all-comers" for both durvalumab monotherapy and the combo.

MYSTIC was originally directed at evaluating the combo rather than monotherapy, with a focus on PFS, and had been expected to read out early this year.

The changes pushed back an initial readout to mid-2017. If all goes well, a regulatory submission could come sometime in the second half of the year, but final OS data won’t be available until 2018.

Reaction to the shift in trial strategy was mixed, with some industry watchers wondering if the delays might spell trouble for a company already late to the game.

While it remains to be seen how MYSTIC will turn out, the changes do reflect a careful approach that aims to give durvalumab the best shot possible at allotting time to show benefit, ensuring sufficient power and building in flexibility.

Optimizing MYSTIC

February’s changes caught the attention of many. Yet, AstraZeneca had been priming MYSTIC’s chances for success for months.

In the spring of 2016, the pharma elevated OS to a co-primary endpoint with PFS. While that lengthened the timelines for the trial, AstraZeneca felt OS better captured the benefit of immuno-onology drugs.

"One of the things that we also learned, not us as a company, but as an industry, I believe is, over the last maybe 12, 18 months, is that overall survival is really the end-game in immuno-oncology,” AstraZeneca CEO Pascal Soriot told investors on a fourth-quarter earnings call.

That modification also required a larger trial size to ensure clinically meaningful differences are detected. Patient enrollment was bumped up first from 675 to 780, then from 780 to 1,092.

In the wake of Opdivo’s setback, some had questioned whether the smaller size of Checkmate-026 — which only had 541 patients — left the study under-powered to detect a benefit. AstraZeneca’s enrollment boost appears aimed at avoiding that mistake.

"I do think they increased some of the powering on the study as it went as an insurance to try to make sure that they set themselves up for success," said Damien Conover, an analyst at Morningstar

Patient selection

The boost in statistical power also gives MYSTIC enough heft to be diced and analyzed across the three arms of monotherapy, combo and doublet chemotherapy.

In AstraZeneca’s eyes, the elevation in stature for durvalumab monotherapy also opens up different paths for MYSTIC to be successful. For the study to read out positively, either the monotherapy or the combination has to beat standard of care chemotherapy.

Boosting the attention paid to durvaluamb monotherapy hedges AstraZeneca’s bets, guarding against the possibility of unacceptably high toxicity from the combo.

If durvalumab monotherapy alone shows benefit, AstraZeneca can move forward with a regulatory submission with the monotherapy. If the combination shows better results that outweigh the likely higher toxicity of the two drugs combined, AstraZeneca can go with the combo.

The two arms could also show benefit among different groups. Monotherapy PD-L1 inhibitors seem to work best in patients with high levels of PD-L1 (although not always). The combo, on the other hand, could be effective in patients who weren’t benefiting from monotherapy alone, those with low PD-L1 levels or non-expressors.

This last part is key because Keytruda doesn’t yet serve this population, presenting a window of opportunity still open for AstraZeneca (and its rivals) to seize.

It is still unknown how AstraZeneca plans to delineate PD-L1 expression levels. Pushing the cut-off down to 5% appears to have been part of why Opdivo failed to show benefit. A similar broad target could be risky, while an narrower cutoff could limit AstraZeneca’s chances to offer a compelling alternative to the already approved Keytruda.

MYSTIC could still fail, or durvalumab may be too late for AstraZeneca to catch its rivals, but the trial optimization of the past year makes clear AstraZeneca doesn’t plan to repeat Bristol-Myers' mistakes.

"Our confidence in MYSTIC is unchanged," AstraZeneca CMO Sean Bohen said on the fourth quarter earnings call. This year will prove whether AstraZeneca's preparation will bear fruit.