Science vs. Marketing

A data-driven analysis of the differences between Opdivo® and Keytruda®

A Wall Street Journal article on March 13 reported that Bristol-Myers Squibb’s cancer drug Opdivo® has outperformed Merck’s similar drug, Keytruda®, by falling back on an “old, mass-marketing approach.” Merck’s product is paired with a diagnostic test used to identify patients expressing the PD-L1 protein, in whom the drug would be the most effective. Merck’s trials only included patients expressing this protein, and the FDA approved the drug only in those who test positive for it.

Bristol-Myers Squibb considered the approach used by Merck, but ultimately found that their drug benefited even patients who did not express the protein. As such, they included a broader population in their trials and the FDA approved the drug for use in patients regardless of PD-L1 status.

The article asserts that this difference in approach—marketing to a broader population instead of only to patients expressing the PD-L1 protein—explains why Bristol-Myers Squibb’s product has sold far better than Merck’s. While this may be a relevant factor, the clinical trials themselves show that there are likely more significant differences to consider. Using the Context Matters data model we were able to compare the FDA labels for the two drugs and the clinical trials conducted to identify the clinical points of distinction.

Opdivo’s FDA label shows that it demonstrated a statistically significant improvement in overall survival (OS) compared to docetaxelin patients with non-small cell lung cancer (NSCLC). Keytruda, on the other hand, did not address OS and instead focused on overall response rate (ORR), which it only demonstrated in a single-arm (non-comparative) study.

Opdivo’s trial, then, had two major advantages over Keytruda’s:

  • OS is more clinically and patient-relevant than ORR.
  • An active comparator trial provides stronger evidence than a non-comparative single-arm trial.

The article proposes that physicians are less likely to prescribe Keytruda primarily because of the time it takes to complete the PD-L1 diagnostic test, but the clinical data suggest a deeper reason. Physicians may not be able to rationalize the time and expense of conducting the test because Opdivo has demonstrated survival benefits based on stronger clinical evidence in patients regardless of PD-L1 status. It’s not a matter of taking the time to identify the optimal treatment—the evidence base in favor of Opdivo is stronger regardless of the test results.

The Wall Street Journal article suggests that Merck’s decision to use a precision medicine approach to market its drug to those expressing the PD-L1 protein is the primary reason for its failure to compete successfully with Bristol-Myers Squibb’s drug. It also suggests that this was primarily a difference in marketing decisions rather than hard clinical evidence. There is a place for precision medicine, and Keytruda’s weak performance compared to Opdivo should not be taken to mean that the “old, mass-marketing approach” is universally superior, or even a true marketing (rather than clinical) decision. It is always important to consider the evidence—Opdivo’s stronger primary outcome and active-comparator data provide compelling alternative reasons for the drug’s success.