Health News Review

There was much news coverage of the announcement yesterday that Merck won’t try to get its cholesterol drug Tredaptive approved in the US after new questions about efficacy and safety were raised in a study.

Bloomberg reported:

Tredaptive was designed to raise levels of HDL, or “good” cholesterol.

The findings call into question the benefits of raising good cholesterol, one of the main methods pharmaceutical companies including Eli Lilly & Co. and Whitehouse Station, New Jersey-based Merck are pursuing in their efforts to develop heart drugs. The pill, which generated less than $20 million in 2012, could have hit sales of $1.1 billion in 2020, said Tim Anderson, an analyst at Sanford C. Bernstein in New York. He removed all sales expectations for the drug from his model.

“There are millions of people taking this drug, niacin,” said Steven Nissen, chairman of cardiology at the Cleveland Clinic in Ohio. “Those people are going to be calling their doctors today. This result certainly undermines the likelihood that any benefit from niacin will ultimately be proven.”

CBS News’ story also touched on the important theme when Nissen said in their interview with him:

“It raised the good cholesterol, it lowered the bad cholesterol, it didn’t improve clinical outcomes.  That is a stunning finding.”


This study and this announcement are strong reminders for journalists – and for the public they serve – about focusing on outcomes that matter – like whether people lived longer or better – rather than focusing on test scores or blood test values or increases or decreases in certain blood test values.

Findings like lowering bad cholesterol or raising good cholesterol can be called surrogate markers, surrogate endpoints or intermediate endpoints.  They are surrogates for things we want to know but may not know yet.  They are intermediate messages but they are not the final message of meaningful “truth” about the impact on peoples’ lives that we need to know about an intervention.

Outcomes like whether people live longer, safer, better are what we care about.

If you focus only on the test scores, it’s like the old saying that “You can’t see the forest for the trees.”

Read our little primer, “Surrogate markers may not tell the whole story.”


Follow us on Facebook,

and on Twitter:


Dan Keller posted on December 24, 2012 at 12:07 pm

As I posted in a comment to the Tredaptive story on
Based on accumulating evidence of failed niacin and other HDL-raising trials (eg, torcetrapib, etc), I am not ready to ditch the HDL hypothesis in its entirety. But not all HDL is created equal. It is not enough to just raise levels of “HDL,” so the particular form of HDL (particle “species”) needs to be considered. Framingham and other epidemiologic trials, as well as molecular and clinical experiments, have established a protective role for HDL, but it’s a far more complex story than originally imagined. Consider for example the case of ApoA-1 Milano, a variant of the apolipoprotein moiety of HDL found in a proportion of the population of a small village in northern Italy. There is a low prevalence of heart disease among people with the variant form even though they have low levels of HDL. Their variant form appears to be very protective against cardiovascular disease. Without knowing anything else about HDL, one could have studied them and come to the conclusion that a LOW level of HDL was a good thing. Red herrings abound. How to raise the right kind of HDL will be the big question going forward in HDL research.