ESMO preview: Has Merck's Keytruda lost its mojo, and what to expect from Padcev combo in bladder cancer?

For Merck & Co.’s Keytruda, medical conferences have in the past served as platforms to showcase new clinical trophies. But at this year’s European Society for Medical Oncology meeting, the PD-1 king is instead presenting a series of failures on top of a closely watched trial win in collaboration with Seagen and Astellas.

The trial flops that will be featured at ESMO 2022 cover various Keytruda combinations in liver cancer, head and neck cancer, and prostate cancer, raising questions of whether the Merck immunotherapy has reached a bottleneck or if it was just bad luck.

As for the much-anticipated cohort K of the phase 1b/2 EV-103 trial, detailed data will be shared for Padcev monotherapy or a combo of Padcev and Keytruda in newly diagnosed metastatic bladder cancer patients who are ineligible to receive the chemotherapy cisplatin. The readout is not just important to gauge Padcev’s market potential but also interesting in light of Merck’s reported Seagen acquisition discussions, which seems to be falling apart.

Before ESMO 2022 officially raises the curtain, Fierce Pharma caught up with SVB Securities analyst Daina Graybosch, Ph.D., to walk through those important trials and on what to look for in the data and their implications for Keytruda, its competitors and collaborators.


Is cohort K a clear win?


Keytruda is in need of redemption in metastatic bladder cancer. After the phase 3 Keynote-361 trial raised a red flag, Keytruda’s use in frontline bladder cancer was recently limited to patients who are ineligible for platinum-based chemotherapy.

The addition of Padcev could give Keytruda a boost. As Seagen and Astellas previously announced, in cohort K of the EV-103 trial, the Padcev-Keytruda combo shrunk tumors in 64.5% of newly diagnosed patients, and as of data cutoff, the median duration of response wasn’t reached.

“The question will be, how picky is FDA going to be about subgroups,” Graybosch said.

Pfizer and Merck KGaA’s Bavencio received a full approval after showing it could extend lives as a first-line maintenance therapy in patients who’ve responded to one cycle of chemo. And the design of cohort K in EV-103 leaves a patient subgroup that—though not eligible for cisplatin—is eligible for carboplatin as an overlap with Bavencio’s indication. Both cisplatin and carboplatin are platinum chemotherapies.

Daina Graybosch, Ph.D.  (Leerink Partners)

So the question is, as Graybosch noted, will the FDA only give Padcev-Keytruda an accelerated approval in patients ineligible for any platinum chemo based on tumor shrinkage data when there’s a standard-of-care PD-L1 regimen for patients who’re up for platinum chemo?

Moreover, an accelerated approval isn’t exactly a done deal for the Padcev-Keytruda combo, and a closer look at the data is warranted, Graybosch said. 

Citing a regulatory expert SVB’s Jonathan Chang, Ph.D., recently interviewed, Graybosch noted that the FDA has recently been giving guidance to companies, asking that the lower bound of their tumor response data in the statistical analysis be greater than an approved standard treatment.

Based on that observation, Graybosch argued that Padcev monotherapy and the Padcev-Keytruda combo must show better tumor response rates and durations of response at the lower bound than Keytruda’s median numbers in its own trial to potentially win their respective accelerated approvals.

In their original announcement of EV-103’s topline results, Seagen and Astellas didn’t offer any information on how Padcev monotherapy fared. Those data are important because the FDA wants to see contributions of each drug in considering a combo therapy, Graybosch said.

Between tumor response rate and duration of response, Graybosch argued that the latter could be even more important because “if you have a 90% response rate but a duration of one month, it’s not that clinically meaningful.”

There, Keytruda has set a very high bar. In the Keynote-052 trial that got Keytruda monotherapy its original accelerated approval in frontline bladder cancer, the Merck drug showed a median 30.1 months of response duration in cisplatin-ineligible patients.


Has Keytruda lost its magic?


Keytruda recently posted at least four late-stage trial flops, three of which will be presented at ESMO 2022. Among them, the Leap-002 failure of Keytruda and Eisai-partnered Lenvima in newly diagnosed liver cancer “feels like a real miss” and is a “really big deal” for Merck’s relevance in liver cancer, Graybosch said.

“It may be a trial design miss,” Graybosch observed. The Leap-002 study found that adding Keytruda didn’t improve upon Lenvima in terms of extending patients’ lives or stalling tumor progression or death.

Keytruda’s disappointing results come in stark contrast to positive readouts from two China-made PD-1 inhibitors. At ESMO 2022, Jiangsu Hengrui Pharmaceuticals will detail data showing a combination of its camrelizumab and a VEGFR inhibitor called apatinib (also known as rivoceranib) topped Bayer’s Nexavar in newly diagnosed liver cancer. BeiGene and Novartis will also report how tislelizumab monotherapy showed similar power as Nexavar did in the Rationale-301 study.

In Merck’s original announcement, the New Jersey pharma noted an unusually strong performance for Lenvima monotherapy in the control arm of Leap-002. So one couldn’t help but wonder whether the trial would have gone positive had Merck used Nexavar instead of Lenvima as the comparator.

“Can you now go forward and do a [Nexavar] control arm? I think it’s going to be really hard,” Graybosch said.

The other problem for Merck is that the FDA hasn’t converted Keytruda’s problematic accelerated approval as a monotherapy in previously treated liver cancer. After a narrow confirmatory trial miss, Merck has proposed an Asian-only trial and Leap-002 as the new confirmatory trials. The lack of a final FDA decision after the modestly positive Asian trial suggests the agency might still be waiting for Leap-002. After all, the FDA has recently been pushing back against trials without participant diversity, and as Graybosch noted, liver cancer has different causes in Asia than in the U.S.

But Graybosch still thinks Keytruda may eventually keep its second-line nod because the original confirmatory trial was a very close miss, and its results were very consistent with the Asian trial. And despite Leap-002's overall failure, it would be even better for Keytruda's second-line chances if the trial shows at least a positive trend in favor of the Keytruda-Lenvima combo, she added.

To Graybosch, BeiGene and Novartis’ noninferiority trial design is also approvable because it is a randomized study. So, even if Keytruda loses its monotherapy nod, tislelizumab could potentially get one and save patients an immunotherapy-only option.

But Graybosch would still like to see detailed Leap-002 data because it might well show that Lenvima monotherapy is a very competitive option. And BeiGene’s monotherapy trial also offers an indirect comparison between a PD-1 inhibitor and Lenvima.

Leap-002 represents the second failure for the Keytruda-Lenvima combo, coming after Leap-007 in newly diagnosed non-small cell lung cancer. Still, Graybosch argued “maybe not reading so much into it” for the remaining 14 or so ongoing phase 2 or phase 3 trials of the combos given Leap-002’s trial design.

Besides liver cancer, Merck will also elaborate on how Keytruda’s combination with chemoradiation failure to top chemoradiation alone for locally advanced head and neck squamous cell carcinoma in the phase 3 Keynote-412 trial.

For that disease setting, Graybosch pointed to phase 2 trial results, suggesting that using immunotherapy after chemoradiation—instead of concurrently—might be the way to go.

The third Keytruda flop to be presented at ESMO 2022, Keylynk-010, found Keytruda and Lynparza failed to beat either Johnson & Johnson’s Zytiga or Astellas and Pfizer’s Xtandi in metastatic castration-resistant prostate cancer. Graybosch said she didn’t really expect a positive readout, because her team hasn’t seen any positive signals from PARP inhibitor combined with anti-PD-1/L1.

All things considered, Graybosch suggested she wasn’t concerned about these trials portending a bad trend for Keytruda: “I think this is a bit of bad luck and a bit of bad trial design, with a retrospective hat on.”