Spring Update on Medical Science

Some good news and some not so good mews

I have recently written and spoken about a number of published studies. Some of these papers brought good news; some have been flawed.

And, as always, the randomized controlled trial has shown its awesome power to inform clinicians and patients alike.

Earlier this month, the American College of Cardiology held its annual meeting in Washington DC. That 11,000 people attended in-person was a hopeful sign. There were even hugs!

My colleagues at theheart.org | Medscape Cardiology have awesome meeting coverage for healthcare peeps.

Here are four short chapters on my latest work.

A New Drug for A Bad Disease

ACC brought news on a novel drug called mavacamten which seems to help patients with a condition called hypertrophic cardiomyopathy or HCM. This is a bad disease where the heart becomes super-thick, often in the septum. This blocks blood flow out of the heart and limits patients’ ability to do anything physical.

Previous studies have shown that mavacamten reduces the obstruction and improves exercise tolerance. The study presented at ACC found that the drug helps patients avoid surgery for HCM.

FDA is likely to approve the drug soon. The normal caveats apply: we don’t have long-term data on efficacy or safety. It will be expensive. And it’s only been tested in severe disease.

I put this in the potential good news category. My column: Mavacamten in Hypertrophic Cardiomyopathy: Reasons for Both Optimism and Caution

A Common Dogma Felled by Randomization

For my entire career, we’ve been telling patients with heart failure to strictly restrict salt. (Split infinitive seemed ok there.)

At ACC, Canadian researchers presented results of a trial called SODIUM-HF, which randomized out-patients with heart failure to standard advice on salt vs a 1500 mg/day low-salt diet. They found no differences in outcomes—including admissions to the hospital for fluid excess.

The authors concluded with just one spin-free sentence: "In ambulatory patients with heart failure, a dietary intervention to reduce sodium intake did not reduce clinical events."

This doesn’t translate to patients with heart failure should eat tons of salt. In fact, the control arm of this study ate only 2000 mg on average. The point was that we don’t have to spend time and energy getting patients down to 1500 mg per day.

I love trials like this. My column: Sodium Restriction in Heart Failure: Another Dogma Felled by Randomization

Left Atrial Appendage Closure Looking Worse

I have been consistently negative about this procedure. The idea of putting a plug in the left atrial appendage is that preventing clots in that structure would help decrease stroke rates and allow patients to get off anticoagulant drugs. The problem is that trials haven’t really shown that actually happens.

Yet the procedure has gained popularity and is increasingly done in many US hospitals.

At ACC, a group from the Mayo Clinic presented a paper that found…sit down for this…one in four patients who have this procedure in the US do not get full occlusion of the appendage. We call incomplete closure, leaks, and even if small, they associate with higher rates of strokes and bleeding.

Two other papers presented at the meeting, one of which I co-authored, reported that patients getting this procedure in the US are older, sicker and frailer than those enrolled in the regulatory trials. This is important because patients with more underlying disease are least likely to benefit.

My column: Percutaneous Left Atrial Appendage Occlusion: ACC Data Challenge Promise vs Reality

On last week’s This Week in Cardiology podcast, I once again reiterated that I feel this procedure may be cardiology’s biggest mistake of my generation. (BTW: I hope you listen to TWIC every Friday.)

Spin Alert

Spin in medical literature is defined as language that detracts from a nonsignificant primary endpoint. In other words, when a medical intervention fails to show reduction in the pre-defined endpoint of a study, researchers find other ways to cast their findings in a good light.

This is weird because science is supposed to be about neutral reporting of an experiment’s results. A good experiment should declare its endpoints of interest before enrolling any patients.

And the value of an experiment should not turn on whether it was positive or negative for a new therapy.

In short, medical science should be about gaining knowledge. Instead, too much of medical science is pre-designed to show that some-(usually expensive profitable)-thing works.

Perhaps the most notable case of spin I’ve seen recently involves a trial named GUIDE-HF, which studied a device used to measure pressure inside the pulmonary artery of patients with heart failure. CardioMEMS is a paper-clip sized monitor that records pressures and then wirelessly transmits this data to clinicians caring for these patients.

The idea is that having this extra data will help keep patients out of the hospital and living longer. A smaller trial with the device had found encouraging data. And FDA had approved the device for use in a narrow group of patients.

GUIDE-HF compared device-led care to standard care in a broad population of patients with heart failure. The authors measured a reasonable endpoint of death, urgent heart failure encounters and heart failure hospitalizations. So far so good.

Then came the results: Despite having more than 250 events in each arm of 1000 patients, the 12% lower rate of events in the device arm did not reach the statistical threshold of significance. Not even that close really.

If science were neutral, as it should be, the authors would have declared that the ($20,000) device did not show superiority over standard care. (Similar to the SODIUM-HF authors I mentioned earlier.)

But that is not what happened.

Instead, the authors decided to change their analytic plan and look at the results in the smaller portion of patients who were followed pre-COVID. In this group, the lower rate of events in the device arm, barely met that statistical threshold. And this allowed them to declare victory. FDA approved the expanded indication, and estimates have about 1 million more patients eligible for the device.

(For grins, multiply $20,000 by 1 million and you will get a hint at why US healthcare spends more than any other country, but does not get the best outcomes.)

In my column on this study, I go over the many reasons why this was a dubious analysis.

Conclusion:

Science can be beautiful in its ability to discover new therapies and prove old dogmas wrong.

Science can also be ignored by its consumers (doctors), as in the case of left atrial appendage closure.

And worse, science can be used in doubtful ways to promote therapies that at most provide low-value care.

The goal of this newsletter is to get people to Stop and Think about medical evidence.