Merck’s CTLA4 move – insurance policy or knockout bid?

In spite of the toxicity burden, the idea of CTLA4 blockade as part of the immuno-oncology armamentarium does not look like going away any time soon. Last week Merck & Co quietly launched a phase III trial that could go further than any other to demonstrating whether the approach has legs.

On the face of it the Keynote-598 study, in first-line lung cancer, looks like a hedge against the outside chance of Bristol-Myers Squibb’s and Astrazeneca’s similar trials succeeding. But its design can also be seen as a way of showing that in blocking PD-(L)1 and CTLA4 the competition has got it all wrong.

Clearly this is not something that Bristol or Astra want to risk proving in their respective first-line NSCLC studies, Checkmate-227 and Mystic. These two trials compare combination and monotherapy approaches against chemotherapy, and are not designed to see whether a CTLA4 combo adds anything to PD-(L)1 blockade alone.

In Keynote-598, however, Merck is asking precisely this question: the two-arm trial pits Keytruda plus Yervoy versus Keytruda alone, measuring overall and progression-free survival as co-primary endpoints.

Selected phase III studies in 1st-line NSCLC
Study PD-(L)1 + CTLA4 combo Enrolment Note Trial ID
Mystic Imfinzi + tremelimumab 1,118 Failed first PFS analysis; multiple OS analyses due in 2018 NCT02453282 
Checkmate-227 Opdivo + Yervoy 2,220 OS & PFS readouts expected in early 2018 NCT02477826 
Neptune Imfinzi + tremelimumab 960 All-comers trial; OS readout due in late 2018 NCT02542293 
Keynote-598 Keytruda + Yervoy 548 Patients with PD-L1 >50%; completion expected in 2022 NCT03302234

Of course, Merck has little to lose here. Keytruda already has a strong hold on front-line NSCLC by virtue of its accelerated US approval in combination with chemotherapy in all patients, irrespective of PD-L1 status.

And the odds of success are low for both Bristol, with Opdivo plus Yervoy, and Astra, with Imfinzi plus tremelimumab. Bristol crashed and burned with its disastrous Checkmate-026 trial, while Astra’s Mystic study has failed its first analysis (Mystic falls at the first hurdle, July 27, 2017).

Bernstein analysts, however, have long argued that it made sense for Merck to hedge against the possibility of rivals' success, however remote. Earlier this year Merck started a phase I study of Keytruda and MK-1308, its internal anti-CTLA4 MAb, in solid tumours, though this small trial barely registered with investors.

Too high a price?

While CTLA4 inhibition clearly yields strong efficacy the problem has always been toxicity. The rationale behind its continued combo use is the theory that it can broaden the activity of PD-(L)1 inhibition by priming naive immune cells and turning tumours immunogenic – in a low enough dose for toxicity to be manageable.

A recent EP Vantage report on immuno-oncology found 75 active trials combining Yervoy with Opdivo, and 41 comprising combos of Imfinzi and tremelimumab. And clearly, judging by competitor pipelines, the industry is not ready to give up on CTLA4 just yet.

Industry anti-CTLA4 projects
Project Company Status Note
Yervoy Bristol-Myers Squibb Marketed Launched in 2011
Tremelimumab Astrazeneca Phase III Licensed from Pfizer
AGEN1884 Agenus Phase I NK cell-mediated
BMS-986218  Bristol-Myers Squibb Phase I Afucosylated MAb
MK-1308 Merck & Co Phase I Sourced extrenally, refined internally
CTLA-4 Probody Cytomx Preclinical Licensed to Bristol-Myers Squibb
AGEN2041 Agenus Preclinical Macrophage-mediated
XmAb20717 Xencor Preclinical Anti-CTLA4 & anti-PD-1 bispecific
ATOR-1015 Alligator Bioscience Preclinical Anti-CTLA4 & anti-Ox40 bispecific
Onc-392 Oncoimmune Preclinical Option to Pfizer
FPT155 Five Prime Therapeutics Preclinical Immune modulator vs CTLA4, CD28 & PD-L1
Anti-CTLA4 CAB project Bioatla Research project Conditionally active biologic; option to Pfizer
Immuno-Oncology II Olipass Research project Oligonucleotide technology
ADU-1604 Aduro Research project "Differentiated" MAb 

Not only is Merck now tackling the opposition head on with Keynote-598, it is also sending a message about how active a player it is in immunotherapy combinations – having previously been dismissed as something of a one-trick pony.

While that might have been true once, the EP Vantage report found that Keytruda was in more combo studies than any other anti-PD-(L)1 MAb – 268 to be precise, including IO/IO and with small molecules, chemo/radiotherapy, vaccines and other advanced approaches. These included three Yervoy combos, though none in first-line NSCLC.

And, if Merck does think that CTLA4 blockade has a chance, its choice in Keynote-598 of the Bristol drug rather than MK-1308 looks smart: if the study fails to back the addition of CTLA4 the company can put the blame squarely on its rival.

A full version of EP Vantage’s PD-(L)1 combination report can be downloaded for free here.

To contact the writer of this story email Jacob Plieth in London at or follow @JacobPlieth on Twitter

Share This Article