It isn’t so much what we don’t know that gets us in trouble. It is the information we are certain is true, which turns out to be wrong, that creates the most harm. Ambrose Bierce, in the Devil’s Dictionary, defined this feeling of being positive about something as “the art of being mistaken at the top of one’s voice.”
Medicine is humbling. A medical school attending physician at OHSU summed this up with his encouragement to remember “today’s dogma is tomorrow’s dog shit.”
I will summarize 4 major studies in Cardiology that will challenge your assumptions about present day management. These studies are presented in more detail in the blog, Sensible Medicine. I was struck by:
How compelling the data was and -
How completely the data has been ignored in practice.
Coronary heart disease is a leading killer. CAD can be silent until it presents with sudden heart attack or death. The consideration of using technology to identify blockages in arteries prior to those severe events makes perfect sense. Dilating or bypassing the severe stenosis will change the natural history of the disease by preventing death and MI and eliminating angina.
Except it doesn’t.
Nearly all the empirical evidence refutes this. Consider these four studies:
COURAGE trial: Stenting severe coronary lesions did not improve survival over medical management.
This study was published in the New England Journal of Medicine in 2007. Over multiple centers, patients with severe but stable coronary lesions and known ischemia proven on stress testing were randomized to two groups: 1. PCI or stent and medical therapy 2. Medical therapy alone. The primary endpoint was death or MI and follow up was 5 years.
The results sent shocks throughout the cardiology world. NO DIFFERENCE. No reduction of death. No reduction of MI.
So, the investigators looked at sub studies, searching for a subgroup of patients who would benefit from PCI. Maybe multi-vessel disease or LV dysfunction. They carried their study out to 10 years. Same result. No difference in outcomes.
Then they turned their attention to quality of life. In the first two years there were more angina free patients in the PCI arm. By three years there were no differences. That conclusion was weakened in the study design. Both groups knew they had bad blockages. One group were told they were “fixed” with the stent and the other knew they were not. Angina is pain and pain can be subjective. Could a placebo effect explain these results?
ORBITA trial: The caring signal of placing a stent is massive.
Habits are hard to break. And fixing clogged pipes is so intuitive. Post-COURAGE, cardiologists didn’t stop doing stress tests, angiograms, and stents. They just changed the reason. The PCI/Stents were not to extend life but to relieve angina. They could point to the COURAGE trial as evidence. Ten years passed.
Then a brave woman named Rasha Al-Lamee asked the obvious question. What if the relief of angina seen in the COURAGE trial WAS due to placebo effect?
ORBITA may be one of the most brilliant trials of our generation. The investigators found 200 patients with single vessel partial blockage. Half were randomized to have the stent. The other half had a placebo (sham) procedure in which a flow wire was placed across the lesion, but it was left unfixed. No one in the trial knew the treatment arm. A repeat treadmill stress was completed at six weeks after either the real stent or the sham procedure.
The difference in exercise time did not reach statistical significance. There also was no difference in the proportion of patients with improvement in angina symptoms. There were no differences in the physical limitations scores.
Using an echocardiogram, investigators found that muscle downstream from the stented vessel looked a lot better in the treatment arm. The stent worked. It just made no difference in symptoms.
ISCHEMIA trial: There is no advantage to the common practice of immediate coronary angiography with intervention.
Modern cardiology does a lot of things, but highest on its goals is the search for ischemia. The main weakness of COURAGE is that it randomized patients after the angiogram. Skeptics argued that doctors did not allow the most worrisome patients in the trial. So, ISCHEMIA randomized patients before the angiogram.
The second uncertainty was the question of going right to the Cath lab for angiography after a positive stress test. ISCHEMIA randomized two strategies: 1. Go to the Cath lab if the stress was positive versus 2. Conservative strategy wherein patients were treated with medical therapy and angiograms were only done for recurrent symptoms. ISCHEMIA was not a stent versus medical management study, but a conservative versus early invasive approach trial.
One other nuance was a blinded coronary CT scan was done to 1. make certain there was coronary disease (eliminate patients with false positive stress tests) and 2. Exclude patients with left main coronary artery disease. 2600 patients were randomized in each of the two groups with an average age 64 years. About half of these patients had at least 3 vessel disease.
The primary endpoints were death due to cardiovascular causes, MI, unstable angina, heart failure or resuscitated cardiac arrest. The results: NO DIFFERENCE. Median follow up 3.2 years. Cardiovascular death or MI – no difference. Rates of all-cause death nearly identical. Importantly, the subgroup with the most severe disease did not benefit from early invasive strategy.
At the end of the day, patients with positive stress tests, many of whom had severe multi-vessel coronary artery disease, achieved the same outcome, same rate of events, with 70% fewer procedures. But this data does place the clinician in a bind. How can you know if a positive stress test will turn out to have left main disease.
This data doesn’t sit well. We’ve all seen stents save lives in the setting of MI. Why is it different in the chronic stable situation? The answer is treatment of sick people is always different from prevention. While we know coronary occlusions exist, we don’t know the lesions will fracture and create MI.
REVIVED-BCIS trial: Revascularization with stents did not improve LV function.
Percutaneous revascularization for ischemic left ventricular dysfunction is common. Patients are frequently referred for angiography because of unexplained heart failure. It makes perfect sense not to leave a patient with a weak heart muscle with major obstructions to flow.
REVIVED BCIS gave stenting every chance to win. Patients had to have weak heart muscle and viable heart muscle and lesions had to be suitable for PCI/stents. Half the study population received medical management only while the other half received stents and medical management. The primary outcome was death or hospitalization for heart failure.
After three and ½ years there was not a shred of difference. No difference in heart function. No difference in quality of life. And they found no subgroup which benefited more from the intervention. This was a trial of 700 patients – of which more than a third had a primary outcome event. There was no difference.
These are my conclusions after reviewing these studies:
Atherosclerosis is systemic disease. Focal narrowing is one manifestation of a diffuse process.
Stenting addresses the focal lesion. Medical therapy treats the diffuse disease.
Primary care would do well to encourage positive lifestyles including diet, weight loss and exercise, use of statins, control blood pressure and diabetes, discourage harmful habits such as smoking – and reserve revascularization for patients with refractory symptoms.
Just as a small percentage of patients drive most medical expenditures, it is often select expensive technology and interventions that may provide limited benefit, which makes Western medicine so unaffordable for the masses. This defines inequality.
Despite a mantra of “follow the science”, we often don’t. We think with our hearts. Me included. When I have my 95% stenosis of my LAD, I’m going to get it stented.
Tim Powell MD