Gilead and Arcus want you to believe again in TIGIT. But mixed ARC-7 data are rolling out to a tough crowd

Bioregnum Opinion Column by John Carroll

Roche let the air out of the TIGIT party balloon with a pair of crushing Phase III failures. Now Gilead $GILD and Arcus $RCUS are back with details of their closely-watched Phase II data in Stage IV non-small cell lung cancer, hoping to spark some renewed enthusiasm for their contender and the rest of the field.

But despite some positive results in frontline therapy for patients with high PD-L1 levels, plus a claim on a potential “new standard of care,” the mid-stage readout from the ARC-7 study won’t end the running debate that has blighted expectations for the target.

For the development partners, ARC-7 offers clear proof-of-concept evidence on TIGIT’s potential. But after the Roche setbacks, anything in Phase II looks iffy at best. And investors immediately turned a cold shoulder to what they were seeing on display, as Arcus shares dived 21% on the data, with a much larger Gilead sliding 1.6%.

Arcus and Gilead will lay out their study at the ASCO Plenary Series session on Tuesday. The high point comes from the anti-PD-1 monoclonal antibody zimberelimab combined with anti-TIGIT domvanalimab, with a hazard ratio of 0.55 and 12 months of median progression-free survival — compared with 5.4 months in the checkpoint monotherapy group.

“There have been mixed signals from prior studies looking at TIGIT targeting,” says Lauren Byers, a cancer expert at The University of Texas MD Anderson Cancer Center quoted in the ASCO release. “The promising results of the combination immunotherapy treatment in this study support further investigation, and if confirmed, could lead to a new standard of care for patients with advanced lung cancer.”

After that, things start to get more problematic.

The overall response rate for the two-drug combination at about a year comes in at 41%, compared to 27% for the solo control.

The triplet, which adds etrumadenant — a dual adenosine receptor antagonist — to the mix, gets worse, with a hazard ratio of 0.65, a 10.9 month median PFS and a comparable ORR of 40%.

From ASCO:

Additionally, for PFS at six months, 43% of patients showed no signs of disease progression compared to 65% of patients in arm DZ and 63% of patients in arm EDZ. Grade three or greater treatment-related adverse events occurred in about half of all groups. Significantly, there wasn’t an increase in immune-related adverse events for the DZ arm.

But even the double is likely to raise more questions. SVB analyst Daina Graybosch, writing ahead of the release, noted that investors should be looking for at least a 50%-55% ORR/9 months of mPFS in Phase II for a win — preferably 60%-plus, with 11 months of mPFS.

SVB based its analysis on a 35% expected ORR and a likely mPFS of 6 months in the PD-1 solo control group, which as we noted came in at a lower rate: 27% and 5.4 months.  The data come without any p-values, simply a conclusion of clinically meaningful results at the interim.

Falling short

For Jack West, a thoracic oncologist at City of Hope, the data from ARC-7 fall well short of the mark for a game-changing approach to lung cancer, especially after the Roche failures.

“We’d need highly compelling data to overcome the negative results with anti-TIGIT strategies we’ve learned thus far from Roche,” West tells me via email.

Jack West

“These results don’t suffice to counter that, particularly when we consider pembrolizumab as the benchmark the vast majority of the clinical oncology world would use to gauge contemporary expectations, noting that zimberelimab is just a placeholder here,” he said. “What we see is that the response rates in the combination arm fall short of the KEYNOTE-024 ORR for pembrolizumab alone (about 45%), and the PFS results for the combinations aren’t clearly superior to pembrolizumab alone. To really impact practice, I’d hope to see a clinically significant improvement in overall survival relative to a regimen we use, in a phase III randomized trial. These results fail by all of those criteria. Moreover, every trial that fails to demonstrate remarkable benefits for anti-TIGIT therapy add to a broader conclusion that this strategy is marginally effective at best and unlikely to warrant a change in practice, let alone be worth the anticipated cost of another immunotherapy.”

Truist’s Robyn Karnauskas, though, is watching the hazard ratio, where the profile for the double looks better. Notes Karnauskas:

We believe the HR is the correct endpoint to focus on in the upcoming read out. HR will tell us the degree of benefit over an active zimberelimab control arm. The previous trials highlighted in Exhibit 1 did not include a PD-1 inhibitor as a control. Furthermore, the nuances described below may make HR the more relevant focal point than to compare PFS across studies.

For Arcus’ new clinical development chief, Dimitry Nuyten, who stepped in last August after joining the exodus at a struggling Nektar, the data offer a solid snapshot of efficacy and relative safety.

“Progression-free survival data and then later with longer follow-up survival data are really the critical endpoints for patients, physicians, and also from a regulatory perspective,” Nuyten tells me in an advance of the data drop. “Survival data is still immature. Progression-free survival looks very strong. I’m particularly happy with the hazard ratio of 0.55 for our double versus the single.

“So, I think the totality of the data between the different benchmarks for PFS, and the delta in response rate show me that we have a very active combination that’s shown meaningful additive activity.”

Expectations game

Some analysts will also look for evidence of commercial potential with a comparison of the data with the KEYNOTE-189 data for Keytruda plus chemo, among other studies, where they can anticipate future results with a chemo add-on. There the two ORRs cited by SVB stood at 62% and 64% for non-squamous and squamous NSCLC, with a mPFS of 11.1 and 8 months.

Gilead and Arcus are up against a set of low expectations created by Roche, which whipped up considerable fervor for TIGITs with their positive Phase II data — which so far hasn’t held up in the much larger pivotal programs. That’s created a great deal of uncertainty around Phase II TIGIT studies in general, where researchers are exploring the impact of adding a checkpoint therapy to the first-gen pioneers that have reaped megablockbuster returns. And Gilead and Arcus don’t expect to report Phase III data for about three or four years, according to the CMO.

Roche’s experience with back-to-back failed Phase IIIs will cast a long shadow over Gilead and Arcus. In their Phase II readout in the spring of 2020, they saw a doubling of the objective response rate of the combo over Tecentriq alone: 31.3% versus 16.2%. There was a 43% reduction in the risk of disease worsening or death, with a median PFS of 5.4 versus 3.6 months. The hazard ratio was 0.57.

The high point for Roche came in patients with high levels of PD-L1, a significant subset. The data were better, with a 55.2% ORR for the combo compared to a 17.2% rate in the monotherapy. The hazard ratio hit 0.33.

Asked about the Roche experience — hitting in Phase II but collapsing in late-stage tries — Nuyten believes they have a better look at mid-term data, with more patients in the study (133 evaluable results from 150 patients) to base expectations on.

The single-arm PD-1 data that they have also came in line with their expectations, he adds, which hit at a variety of points throughout previous studies, including the 20s and mid-30s.

All the comparisons, though, will likely only encourage a reluctance to do anything but wait for the definitive Phase III data.

This is no sure thing.

ket
therapy