Death is Death but Not All MI's Are Equal

Yesterday morning I attended a session at this year's American College of Cardiology meeting on lipid controversies. Dr. Rita Redberg gave an excellent talk where she presented the case for not starting or *deprescribing* statin drugs for primary prevention in people 75 years or older.

I would summarize her message as “when a drug or intervention can't make someone feel better there should be evidence that it makes people live longer in order to recommend it for *primary* prevention.” If not, the disutility of taking a pill and the side effects that come with it are likely to offset any other potential benefits.

For people over 75, there is not strong evidence that statin drugs reduce the risk of dying. There is evidence that they reduce nonfatal heart attacks but a lot of people, on the order of 50 to 100, must take the drug for 1 person to benefit by having a heart attack prevented. Some people may like those odds and want to take a statin drug but for others, those numbers aren't particularly inspiring and they're happy to go without.

Dr. Redberg, myself, and many others share the latter perspective, especially when it comes to caring for older adults.

So that was her case in a nutshell. Based on my perception of the audience in the room, there were many who seemed to be in agreement with her and others who clearly weren't.

After she presented, several other speakers gave talks on various subjects involving cholesterol management and at the end of the session, all of the panelists got on stage to field some questions and discuss amongst themselves.

Unfortunately the session was running over time and the discussion had to be cut short.

Dr. Ginzberg got the last words in, which he used to criticize Dr. Redberg's emphasis on mortality reduction.

To paraphrase, he said we need to look beyond mortality because for older adults, their lives are changed by having a heart attack. Thus, if we can prevent heart attacks, we should. This is not a fringe view and many, thoughtful physicians share it.

Dr. Redberg didn't have a chance to respond but here is my rebuttal.

Yes, generally speaking, heart attacks are important but in terms of contributing to major morbidity, they are not all equal. In fact, over the last 20 to 30 years our ability to detect smaller and smaller insults to the heart muscle has changed what it means and looks like to have a “heart attack.” So much so that in 2023, a large number of heart attacks cause no major damage to the heart (detectable by echocardiography) and another large percentage (i.e., type 2 NSTEMI’s) require no special treatment at all.

So in response to Dr. Ginzberg, “no, not all heart attacks are life-changing events” and there is good reason to believe that these small, less significant ones are the ones that are reduced by many of our modern therapies.

How could we know?

One way would be to adjudicate heart attacks based on their severity but this isn't what we do in most clinical trials. Heart attacks are adjudicated as all or nothing events.

With this being the case, what would be another way to infer the significance of the heart attacks which are prevented? By looking at other, more concrete measures, that would be influenced by having a major heart attack - like dying!

And this is why I think Dr. Redberg wins the argument. It's simply not valid to assume that a statistically significant but clinically small reduction in non-fatal heart attacks (and other non-fatal events) is enough to justify the tradeoffs of therapy.

Over the last 20 to 30 years, the association between heart attacks and death in clinical trials has gotten significantly weaker. David Brown and colleagues showed this very convincingly. I'll insert the link when I'm back at my home computer.

The endpoints used in clinical trials matter and understanding how they have changed over time (been watered down in most cases) is important when interpreting and translating their results. This is a theme I'll return to often on my substack (statistical significance versus clinical significance is another).

This exchange highlights those themes and thus, seemed like as good a place as any to start writing. I hope you enjoy the content here and I look forward to seeing your comments.