–The slippery slope of noninferiority.
Last week the big news in interventional cardiology was the approval of Abbott’s bioresorbable stent, Absorb. There’s no doubt that this represents a remarkable technical achievement, propping open a clogged artery and then gradually disappearing. There are all sorts of theoretical reasons why this might represent an important advance over traditional stents that stick around forever, leaving open the possibility of future harm.
But there’s often a big gap between theory and reality, and there are good reasons why interventional cardiologists need to remain cautious. Unfortunately, many of the leaders in the field appear to have fallen in love with the future promise of this technology and are not focusing on its current weaknesses.
In a Medpage Today story Sripal Bangalore said he would use the device in 20-30% of his patients; Cindy Grines said she would use it in as many as half of her patients. “The Absorb bioresorbable scaffold represents a major advance in the treatment of coronary artery disease,” touted Gregg Stone, chairman of the ABSORB clinical trial program, in an Abbott statement. . “This novel technology appeals to both physicians and patients alike because after treating the underlying blockage it is completely absorbed, leaving nothing behind. No metal means the treated artery can pulse and flex naturally as demands on the heart change with everyday activities. No metal may also reduce the potential of future blockages that occur with permanent metallic stents, and allows easier access to other treatment options should they prove necessary in the patient’s future.”
But a careful look at the data suggests that Absorb, at least in its current form, may not deliver any advantage to patients. The most we can say for now is that the device met the criteria for noninferiority in its major clinical trial, but this did not actually rule out the possibility that the risk of stent thrombosis with the device– the most feared outcome with a stent– is twice as large with Absorb as with traditional stents.
The problem of “non-inferiority trials with inferior outcomes,” as Eric Topol tweeted, leads to the danger of the “slippery slope in the clinic.” Topol asked: “Who would want a bioabsorbable heart stent given the data that led to FDA approval”?
Exemplifies the problem of non-inferiority trials w/ inferior outcomes
a slippery slope in the clinic
lots of patients asking for these 🙁
— Eric Topol (@EricTopol) July 12, 2016
I asked a clinical interventional cardiologists with no ties to Abbott for his thoughts about Absorb. He did not wish to be identified but as @jeddacath on Twitter he has gained a reputation for his reliable and independent opinions. He thinks Absorb in its current form is an inferior product. In addition to the higher risk of stent thrombosis, he notes that procedures to implant the Absorb stent are technically more difficult and will require more imaging and lesion preparation, all of which will make the procedure more expensive. Here are his thoughts:
The current data with the ABSORB bioabsorbable stent (BVS) is very consistent that it is actually an inferior product compared to the best current metallic DES (EES).
- The stent thrombosis is definitely twice higher. No one can dispute this. (Stone. Metanalysis Lancet 2016) 1.3% vs 0.6% .
- You may try to slice and dice data (Post hoc analysis) that it was the smaller stents
- You may argue the doctors were not experienced. Heck no, these were the hand picked chosen ones based on their experience to place BVS. These were not rookies.
- You may argue the technique was not optimal. These interventionists did what was asked of them.
- More expensive: the BVS and lesion preparation (increased need for cutting balloons or NC balloons). Direct stenting is out of the question
- More expensive if imaging which is advocated by the so called experts to prevent ST.
- More time consuming (lesion preparation + imaging). Instead of the slam-bam-thank you DES.
- After all the above more expensive, more difficult it is only non-inferior at best. If I am buying a car I don’t want a Toyota for a price of a Mercedes.Albeit, I am not a nihilist. BVS may give us certain advantages, which we don’t know yet. To use a cliché it is a device looking for a niche. As the next generation devices are produced they will be certainly better than the first generation. Just look at Xience vs Cypher. Would I want one in myself right now? no. So we are at an existential dilemma, should we use it in our patients? With all due respect patients can not decide when we after 10+ years of training and 10+ years of experience don’t know the answer. It definitely should not be sold to patient as a great new product, and I do agree to the thought it should be used part of a phase 4 registry where all the implants are recorded and followed systematically.