Bristol-Myers Squibb’s clinical trial bet on its immunotherapy treatment, Opdivo, backfired when the study missed its primary endpoint.

The Phase-III trial, dubbed CheckMate-026, investigated Opdivo as a therapy for previously untreated patients with non-small cell lung cancer and compared those patients to another group receiving chemotherapy, the standard of care. Opdivo failed to demonstrate that it was a superior treatment to chemotherapy in terms of progression-free survival.

Bernstein analyst Tim Anderson gave the edge to Merck’s competing treatment, Keytruda, saying that the drugmaker now will be allowed to promote its therapy as “the only one that has been proven to improve survival in first-line lung cancer,” in direct-to-consumer advertising. “This puts BMS on its back foot from a commercial perspective, as it tries to defends its franchise,” he noted.

See also: As immuno-oncology therapies evolve, so will the marketing

Keytruda will likely gain “significant” traction in the overall market, and that traction in the long-term could mean more combinations with chemotherapy agents or other immunoncology therapies over time, Leerink Partners analyst Seamus Fernandez said in a research note. He added that Merck will “dominate” the first-line lung cancer market for at “least the next 12 months” in certain lung-cancer patients.

Mark Schoenebaum, an analyst for Evercore ISI, wrote Friday that the first-line lung-cancer market is worth more than $12 billion. He estimates that Merck’s current label for Keytruda in this setting opens the drug up to 30% of that market.

Opdivo has been the market leader in terms of sales. Opdivo brought in $942 million in 2015, compared to Keytruda’s $566 million. The marketing spend for Opdivo was also significantly higher last year, to the tune of $125.9 million for Opdivo compared to $7.3 million for Keytruda, according to Kantar Media.

See also: Merck’s Keytruda versus Bristol-Myers Squibb’s Opdivo

The trial’s failure is considered a serious disappointment by both Bristol-Myers Squibb’s management and analysts. The misstep can be traced back to BMS’ attempt to pursue the broadest possible label by including too vast a patient population in the study, which ultimately created a bigger risk. This setback will likely open the door for Keytruda to gain market share and pins Bristol-Myers Squibb’s hopes for a first-line treatment on its combination trial studying Yervoy and Opdivo. That trial is not expected to be completed until January 2018.

Anderson wrote in the same analyst note that “the missed result reflects the fact that BMS pushed the envelope too far in designing the trial.”

Merck, on the other hand, studied a narrower patient population in its monotherapy for Keytruda — a move that increased its chance of success. BMS, for this trial, chose a population of patients with tumors that express a specific protein — PD-L1 — at levels equal to or greater than 5%. Merck, on the other hand, chose patients for its trial with much higher levels of expression: with 50% and greater expression of PD-L1. The level of PD-L1 within lung-cancer tumors is considered an important bellwether for how well immunoncology agents will perform.