The seemingly never-ending battle between PD-1 inhibitors Opdivo (nivolumab) and Keytruda (pembrolizumab) rages on.

Keytruda stuck first with a 2014 approval for advanced or unresectable melanoma patients who are no longer responding to other drugs. A few years prior, the drug was sitting on Kenilworth, N.J.-based Merck & Co. Inc.'s shelves, having arrived via an acquisition, but after stunning early clinical trial data from Bristol-Myers Squibb Co.'s Opdivo, Merck quickly dusted off the program and restarted it quickly enough to edge out Opdivo to the first U.S. approval. (See BioWorld Today, Sept. 5, 2014.)

New York-based BMS wasn't too far behind, gaining FDA approval for advanced melanoma patients in late 2014, although it technically beat Keytruda to the first global approval with a nod from Japanese regulators earlier that year. (See BioWorld Today, July 9, 2014.)

Both drugs have now been approved for about a dozen indications each. Last year, BMS racked up $4.95 billion in sales of Opdivo, besting Merck's $3.81 billion for Keytruda, but Merck appeared to be catching up, with fourth-quarter revenue of $1.3 billion, just a sliver behind Opdivo's fourth-quarter take of $1.36 billion.

Investors will get to update the revenue score on Tuesday when Merck reports earnings for the first quarter. Last week, BMS put up the number to beat: $1.5 billion for its first-quarter Opdivo sales, a jump of 34 percent compared to the year-ago quarter.

Looking ahead to additional indications, the American Association for Cancer Research (AACR) 2018 meeting and concurrent publications in The New England Journal of Medicine gave doctors and investors additional data points to rank the two drugs.

In BMS' Checkmate 227 trial, non-small-cell lung cancer (NSCLC) patients taking Opdivo had a median progression-free survival (PFS) of 4.2 months compared to 5.6 months for patients receiving chemotherapy, producing a hazard ratio of just 0.95.

"In other words, BMS' PD-1 monotherapy failed to beat chemo in Checkmate 227," Evercore ISI analyst Umer Raffat concluded, "Recall BMS's PD-1 monotherapy failed to beat chemo in [the] Checkmate 026 trial also. Meanwhile, Merck's PD-1 monotherapy has beat chemo in two lung trials: Keynote 024 and Keynote 026."

Drug or trial?

One of the problems with comparing drugs across trials is that companies can set up the trials differently, making comparisons difficult, as Raffat pointed out in a follow-up note to clients.

Patients in BMS' first-line lung cancer trials received Opdivo for a shorter time, about four months on average, compared to about seven months for patients taking Keytruda in Merck's trials.

"Perhaps the answer is simply patients in BMS' trial progressed faster and thus got lesser dosing," Raffat hypothesized, while quickly noting that, "perhaps there's more to it."

As it turns out, BMS' Checkmate 227 and Merck's Keynote 189 used different definitions of progression. Patients were allowed to continue on the drug past their initial scan if they were deemed stable until a second scan confirmed the progression.

Both trials used RECIST version 1.1 criteria for identifying the initial progression scan, which Checkmate 227 also used for the confirmatory scan, but Keynote 189 used the more-lenient immune-related RECIST (irRECIST) for the second scan to confirm progression.

While it's hard to tell if the progression rules affected the results of the two trials, Raffat found data from a Merck trial in melanoma and NSCLC patients presented at the 2015 American Society of Clinical Oncology (ASCO) meeting showing that patients who had positive scans as measured by RECIST 1.1 and negative per irRECIST lived longer than patients who progressed based on positive scans for both.

In addition to the different criteria for progression, Raffat noted that Checkmate 227 required patients to sign an additional written consent form if they wanted to continue dosing after the first progression scan while Keynote 189 only required a verbal consent by patients. It's possible that extra barrier resulted in fewer patients even continuing on drug until the confirmatory scan.

Interestingly, the FDA is scheduled to present a poster at ASCO 2018 evaluating using time to treatment discontinuation as a pragmatic endpoint based on pooled data from eight metastatic NSCLC trials, Raffat noted.

Combinations

"From BMS' perspective, if the company feels strongly that this is just trial to trial differences, perhaps they may consider running a head to head vs. Keytruda in lung cancer," Raffat speculated.

Considering the number of combinations being tested with PD-1/PD-L1 drugs, it seems unlikely BMS would get the opportunity to run a head-to-head monotherapy trial. By the time the results were available, some combination would already have taken over as the new standard of care.

At AACR, Merck presented data showing that first-line nonsquamous NSCLC patients receiving Keytruda plus pemetrexed and a platinum-based chemotherapy lived longer than those on pemetrexed and a platinum-based chemotherapy alone with a hazard ratio of 0.49. (See BioWorld, April 17, 2018.)

While the data are still early, Toll-like receptor 9 (TLR9) agonists appear to be having success in combination with PD-1 inhibitors. At AACR, Checkmate Pharmaceuticals Inc., of Cambridge, Mass., and Dynavax Technologies Corp., of Berkeley, Calif., both reported data showing their TLR9 agonists, CMP-001 and SD-101, respectively, combined with Keytruda produced responses in melanoma patients. (See BioWorld, April 18, 2018.)

TLR9 stimulates the innate immune response system, which can act with PD-1 inhibitors to activate T cells, stimulating them to attack the tumor.

"It's nice to see two different molecules that stimulate the same target seeing early responses. I think that's reassuring and can build confidence in this as a class that has activity," Robert Janssen, chief medical officer and vice president of clinical development at Dynavax, told BioWorld at the conference earlier this month.

Companies have also started segmenting patients based on their tumor mutational burden (TMB). While Checkmate 227 failed to show an effect comparing Opdivo monotherapy to chemotherapy, a predefined subset of patients with high TMB in the trial responded substantially better to a combination of Opdivo and Yervoy (ipilimumab) compared to standard-of-care chemotherapy with a PFS hazard ratio of 0.58. (See BioWorld, April 17, 2018.)