Self-inflicted wounds hurt the most.
Bristol-Myers Squibb's immune-boosting cancer drug Opdivo had been dominating Merck's rival drug Keytruda because it is marketed to a broader patient population. Opdivo outsold Keytruda by more than $500 million in the second quarter.
But Bristol's attempt to use the same broad-brush approach in treating newly diagnosed lung-cancer patients blew up in its face back in August. In a trial with those patients, Opdivo did not outperform chemotherapy, while a more-conservative trial by Merck succeeded. That result carved $30 billion from Bristol's market cap at the time.
The news got worse this weekend: Updated data from the same trial show Opdivo didn't even beat chemo among a more-promising patient group. Merck's drug, on the other hand, reduced death in those patients by 40 percent compared to chemo. Markets shaved another $9 billion in Bristol's value on Monday, as shares fell to a two-year low.
Observers who picked Bristol as the surefire winner in this drug class were premature. Now, instead of defending its throne, Bristol needs to scramble to keep up.
The fight for market leadership in this space, known as immuno-oncology, has only just begun.
The first approvals for these drugs were in patients that had failed other lines of treatment. It's where cancer medicines often start -- it's easier to show superior benefit in tougher-to-treat patients with limited options. That may have helped Bristol take an early lead, as did its aggressive clinical-trial strategy -- Bristol didn't limit trials to patients whose tumors have a very high level of a protein (PD-L1) thought to predict success for this drug class.
But Bristol's head start may mean less going forward. Merck's recent trial success gives it a jump on treating newly diagnosed lung-cancer patients. Roche's Tecentriq, meanwhile, is expected to get an approval in previously treated lung cancer patients in the next few weeks.
And all of the companies in this space will likely be fighting for market share in treating various cancers for years to come.
The real prize is the 75 to 80 percent of newly diagnosed lung cancer patients not eligible for Merck's drug based on their PD-L1 levels. And there are many patients with other types of cancer that don't benefit from this drug class on its own. To reach them, firms are testing combinations of multiple drugs, with the Opdivo/Keytruda class as a backbone.
Bristol got the first approval for an IO combo, its older drug Yervoy plus Opdivo to treat melanoma. It's testing that combination in a variety of other cancers. But AstraZeneca will have trial results for a similar combo in lung cancer ahead of Bristol. And Merck, Roche, and others are trying other approaches to the combo game that might end up being better. In one very small trial, a combination of Keytruda with an early-stage Incyte drug had triple the response rate of Bristol's Opdivo/Yervoy duo in melanoma, and appeared safer.
If Bristol loses more of its lead in single-drug treatment and can't find a way win in the combo game, then it may be in real trouble. Analyst estimates for Opdivo's sales in 2019 have fallen by more than $2 billion since August and may have further to go.
Even with these lowered estimates, Opdivo is expected to account for more than 30 percent of the company's sales in 2019. The company's valuation compared to its peers was sky high when many assumed it would lead this market for years to come, but it has eroded dramatically along with expectations for Opdivo.
It seems unlikely Opdivo is a fundamentally worse drug than Keytruda, given its success over the past few years. Bristol reported promising Opdivo trial results in head and neck cancer over the weekend, in fact.
But in trying to differentiate Opdivo from its rivals, Bristol's approach has been too aggressively broad, raising the possibility of more design failures to come. It needs to find a new way to expand Opdivo's opportunity, instead of watching it diminish.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
To contact the author of this story:
Max Nisen in New York at email@example.com
To contact the editor responsible for this story:
Mark Gongloff at firstname.lastname@example.org