In (Gasp) Defense of the Coronary Stent

A kerfuffle ensued recently when an oncologist and expert on evidence based medicine took the field of cardiology to task over the evidence for placement of the ubiquitous coronary stent.  What started with a lengthy article in Propublica that included coronary stenting for stable coronary disease as a prime example of a procedure done without evidence to back it up turned into this fiery twitter exchange between Drs. Kirtane (cardiology) and Prasad (oncology).

The crux of the debate revolves around placement of coronary stents in patients with stable coronary artery disease.  Stable coronary artery disease refers to narrowing of the arteries by a build of plaque that has occurred slowly over time.  Unstable coronary artery disease refers to eruptions that occur within the coronary vessel when a plaque ruptures, quickly leading a patent vessel to become completely occluded or nearly occluded.  Unstable coronary artery disease, otherwise referred to as an acute coronary syndrome is regarded as an emergency that requires urgent intervention by skilled operators (interventional cardiologists) who must race against time to abort a process that if left unchecked may lead to death or severe damage of the heart muscle.

Figure 1. Stable angina/Acute Coronary syndromes

Stable coronary artery disease on the other hand is not considered an emergency, but can result in patients being symptomatic because of diminished blood flow through the culprit artery. Angina pectoris is the descriptor one uses to describe chest pain that relates to a mismatch between the blood flow the heart muscle needs and what it receives. It is almost always the case that angina in stable coronary disease is triggered by activities such as physical or emotional stress that require more blood flow than the narrowed artery can supply.

Patients currently get stents placed for both stable and unstable coronary artery disease, but the debate centers around the evidence for placement of stents in stable coronary disease. Vinay Prasad, an oncologist who apparently spends much time thinking about the heart, believes that there is no data to support placement of stents at all in this setting. The charge is also made implicitly, and explicitly that the only reason for stents being placed in this setting relate to underlying financial incentives rather than the best interest of the patients. Cardiologists, of course, almost universally disagree with these sentiments.

There certainly is blame to be laid at the feet of Cardiologists, but as is usually the case with simple narratives this particular story is not so simple and is deserving of context.

The story of coronary interventions goes back 40 years to 1977.  A 37 year old insurance salesman named Adolph Bachmann with severe exercise induced angina had been found to have severe narrowing of a major coronary vessel.  At the time the treatment slated would have involved a cardiothoracic surgeon cracking open his chest and bypassing the blockage.  Andreas Gruentzig, a cardiologist who had been working on opening up blockages in arteries in the legs by blowing up balloons in narrowed arteries had other ideas.  One year earlier Gruentzig had presented the results of his artery opening technique called angioplasty in dogs at the American Heart Association meeting, and had been waiting for the ideal patient to attempt this on.  When offered the opportunity to avoid bypass surgery, the insurance salesman enthusiastically leaped at the opportunity.  With a cardiothoracic surgeon on standby, Gruentzig guided a catheter across the narrowing, and inflated a balloon across the blockage, proving for the first time in man that a blockage in the heart may be safely treated percutaneously.  The patients symptoms improved markedly and remained symptom free almost thirty years after his procedure.  The success was not surprising to Gruentzig – the coronary work was simply an extension of his prior work.

By the time Gruentzig attempted this procedure in the coronary artery he had become a master of angioplasty in the larger diameter vasculature that fed the legs by building on the work of an interventional radiologist named Charles Dotter.   Dotter had accidentally recanalized a stenosed iliac artery while doing a routine diagnostic angiography, and immediately understand that the lowly catheter ‘if used with imagination’ could become an ‘important surgical instrument’.  Dotter’s first non-accidental patient was an 83 year old woman named Laura Shaw, bedridden for months with a cold, painful left leg  admitted for amputation due to progressive gangrene.  Deemed too sick for surgery, and with no available options, Dotter proceeded to progressively dilate the artery with successive catheters. The gangrene healed and the patient left the hospital without an amputation.

Gruentzig’s genius was in taking Dotter’s technique and perfecting a balloon catheter to allow for dilation of narrowings, rather than the progressive recanalization Dotter had demonstrated.  The first patients he helped were patients like Laura Shaw, with severe disabling pain in the legs from blocked arteries that didn’t allow blood to reach the muscles of the leg.  In 1974 – a full three years prior to the first coronary angioplasty, Gruenzig started successfully using his hand made balloon catheter to open up these leg arteries and provide relief to formerly crippled patients.

It was his success in these patients that made Gruentzig so confident about interventions in the coronary arteries.  After all the principle was the same –  leg pain was a result of poor blood flow to the muscles of the leg, and chest pain was a result of not enough blood flow to the heart muscle.  If relieving blockages in the legs gave function back to patients with leg pain, relieving blockages in vessels of the heart should similarly reduce angina.  In Gruenzig’s mind, all that was needed was a smaller catheter and a delivery system to access the coronary tree.  He was right.

After his first resounding success, Gruentzig performed four more procedures and published his seminal work as a letter to the editor in Lancet in 1978.  The procedure was more successful than anyone could have imagined but still required emergency rescue with coronary artery bypass surgery in 9 of the first 60 cases.  Over the next few decades interventions in the coronary vascular tree became safer and easier with ever higher rates of long term success with better techniques and stents to keep narrowed arteries open longer.

Unstable coronary disease, or acute coronary syndromes did not become a focus of cardiologists until a man named Marcus Dewood did what prior had been considered heretical– inject contrast into the coronary artery of a patient having a heart attack.  Up to this point no one really understood what happened to patients who had heart attacks.  This all changed when Dewood demonstrated it was a blood clot (thrombus) that was completely occluding these important vessels. Cardiologists who had in the past been bystanders to heart attacks as they happened, became active combatants who rushed to open these blocked arteries using the same techniques Gruentzig had pioneered.  Averting a heart attack didn’t just relieve chest pain, it prevented death, and heart failure that used to be a natural consequence of heart attacks.

All good things eventually get screwed up, and the same can be said of the field of interventional cardiology.  What started in a kitchen with catheters made by hand (Figure 3) became a very big business.  And where there is lots of money nefarious interests seem to intervene.  Lots of stents started being placed for a variety of reasons.  Cardiac catheterizations were being done not in patients with severe exercise induced angina but in grandpa John with chest pain after his grandson used his sternum as a trampoline.  The finding of a narrowing in Grandpa John’s artery should not have warranted a stent, but sometimes did.  The impression was left by many cardiologists, either explicitly or implicitly that disaster in the form of a heart attack had been averted.

It was visually easy and enticing to fall prey to the idea that opening up nearly blocked arteries was in some way preventing a heart attack.  Of course, averting a future coronary event in patients with stable CAD had never been the premise of coronary stents, and the early data that attempted to answer this question suggested the tightest blockages may counterintuitively be the least likely to rupture and cause an acute blockage.  Studies at the moment suggest that the likelihood of a plaque rupturing- its stability- is a function of how thick of a fibrous cap overlies it. (Figure 4)   Patients with stable CAD who had serial coronary angiograms to study the natural history of coronary plaques appeared to show that the area of highest narrowing, may also be the area that has the thickest fibrous cap.

Not surprisingly, studies that randomized patients to medications or stents for stable CAD, showed that patients with stents live no longer.  Dealing with the epidemic of unnecessary care prompted guidelines on the appropriate use of stents.  Efforts to prosecute and review high volume operators made the national news and sent an especially strong message to the interventional cardiology community.  Perhaps as a result, the number of coronary stents for stable CAD fell by 50% from 2009 to 2016.

Critics, however, smelling blood in the water interpret the studies to say that no stents should be placed unless you’re having a heart attack. These are evidence purists who only accept changes in practice supported by randomized controlled trials that preferably show a mortality benefit.  They are unimpressed by the effect of stents on angina because the patient is not blind to the intervention. A placebo is a powerful thing – and angina is a subjective complaint.  Convincing these puritans requires a control arm that’s a sham – the patient must believe a stent was placed to know if stents have benefit.

Regrettably, what stands in the way of generating perfect data for the apostles are ethical cardiologists who refuse to randomize patients they believe will be harmed by no intervention.  Randomizing a patent requires uncertainty on the part of the randomizer – a doctor who believes one arm of a trial may be harmful to his patient cannot be expected to put the trial before his patient.  Importantly, primum nocere (do no harm) applies regardless of how wrong the doctor may be proved in some future era.  The physician who is out of step with a future he does not know carries no more blame than the Italian mother who refuses to let her son voyage to India by sailing West because she fears he will fall off the edge of a flat earth.

Randomized control trials thus fail to apply to patients that were never included in the trial because cardiologists refused to randomize them.  The patients excluded in this case are patients with severe disease – those who were unable to be stabilized with medical therapy, or those with disease that jeopardized a large amount of heart muscle. These decisions are not at the fanciful whim of cardiologists but instead are data driven, albeit the non randomized kind.  While most significant narrowings may have a thick fibrous cap that makes them stable, not all do.  Some highly stenotic plaques in critical areas do rupture, and the consequences of this are severe.  To understand what I mean by blockages in critical areas, consider the consequence of an obstruction of the water main that leads to a house versus one that leads to the basement powder room. One makes the house unlivable while the other is a minor nuisance.   Navigating a 1% chance of a water main obstruction is all together different than a 1% risk of losing use of a powder room.

The problem with coming down too hard on stents is that the trials to date have specifically excluded patients with ‘water main’ like obstructions.  On top of that the whole idea that coronary occlusions/heart attacks evolve from mild lesions may be wrong.  The original data on this came from studies that evaluated serial coronary angiograms.  This has the small problem of excluding patients who die before they can get their next angiogram.  Post mortem studies of subjects dying from cardiac arrest or an acute MI note that the % luminal area of narrowing was >75% in two-thirds of cases, and the mean stenosis of likely culprit lesions causing an MI was 90%.

My intent is not to convince the reader or the purist that overzealous cardiologists may actually have been doing the right thing in their stent orgy, but rather to attach reasonable doubt and fear to doing nothing to the 45 year old father of three who has a water main like obstruction.

In addition, the studies in patients with stable CAD don’t really show that stents are ineffective for angina ( Gruentzig’s original indication ). The most famous study hailed as evidence of the futility of stents – COURAGE – actually reported that one-third of the patients randomized to medical therapy ended up needing a stent.  The actual conclusion of the trial was that in stable, low risk patients with angina, an initial strategy of medical therapy was safe and effective and would avoid a stent in two-thirds of patients.  This was a worthy conclusion, but a far cry from the conclusion that no stents should be placed for stable coronary artery  disease.

I have no doubt skeptics mean well and have patients in mind when they question the motives of cardiologists, but these ad hominem attacks are devastating precisely because of the vast information asymmetry that exists between patients and doctors.  Patients simply can’t hope to know as much as cardiologists know about stents.  The basis of any argument a physician makes relies heavily on the truthfulness of the person making the argument.  Absolutist critics of coronary stents know what they are doing. Having made up their mind about stents from their review of the available randomized control trials, they favor a moratorium on all stent placement outside the realm of those having a heart attack.  Since they can’t convince the larger cardiology community of this, their campaign necessarily involves casting doubt on the motives of cardiologists.  It’s an understandable but unfortunate approach. It makes for meaty headlines to suggest cardiologists who place stents purposefully ignore evidence to line their pockets – but it would be a gross mischaracterization. I don’t mean to suggest patients should accept their doctors assertions about invasive therapies without question.  A recent review of a registry of cath labs found 14% of stents placed for stable disease could be inappropriate.  This continues to be too high a number, though I wouldn’t focus too much on that absolute number because what is defined as appropriate has a tendency to change from year to year.   I do mean to suggest that when it comes to stents in stable coronary disease, don’t throw tomatoes at the next cardiologist who recommends one.

Anish Koka is a cardiologist based in Philadelphia

Categories: Uncategorized

25 replies »

  1. Steve, one cannot serve two masters so one has to choose between the one at the bedside and the other that puts bread on your table. Anyone that doesn’t believe that can ask the uneducated organ grinder. ” He who pays the piper calls the tune ”

    I referred to who I thought was best for my patient. Sometimes that individual was in a group, but that didn’t mean I referred to all in the group because all members of a physician group are not the same. They are not interchangeable widgets though that is what physicians are becoming.

  2. Good points. When I moved to a new job in a new city socially I met a pathologist who recommended a primary doc and helped me get into his closed practice. For 15 years I saw him felt he listened well and was carefully analytical. On the occasions he referred me to a specialist I felt I was with a very competent doctor (granted, who knows for sure). Since that doctor quit clinical practice ten years ago I have had no interaction with a primary care doctor that compared….and it keeps getting worse as the salaried doc. just seems glued to the EHR screen and pushes checklist things. Luckily I’ve had no serious conditions…..and perhaps if I develop one the engagement with a competent and caring and analytical doc would happen. Hope I don’t have to test that soon.

  3. Paul,

    I am likely older than you are. I know very few people who had a personal relationship with an interventional cardiologist in the old days. Yes, it was possible to have a good relationship with your PCP, but people joining up into larger groups was well on its way before the ACA so that was going away regardless. If you had a PCP you liked, based upon my experience there wasn’t much correlation between quality and referrals so I am not sure that helped very much. (Still doesn’t. The second busiest surgeon in one of our specialties in our state has an awful complication rate, but he still gets referrals like crazy.) Docs have been looking to become employees rather than risk joining a group to become a partner for quite a while, a trend which precedes the ACA. At the risk of sounding old, I think the young docs today are a bit different.

    Alan- I am in the trade. I have access to our network’s stats on outcomes and complications. However, if I had to go out of network, other than a few areas, I doubt that I would be much better at ascertaining who is a good provider than the average person and whether they were offering advice free from economic influence without calling friends. (When you have watched your fellow docs and institutions follow the fads and payment trends for a long time you get cynical about this stuff.)


  4. So many places to go with this piece. Fun reading. My comment is this; the only EBM expert is the patient. It does not matter a whit what an oncologist, cardiologist, primary care MD thinks; it is what that patient thinks. My wife’s oncologist keeps presenting guidelines and she keeps altering their choices. Patients are really smart about the value of information once informed. Here is one example of a high school student’s comments on a paper. http://www.sharedmedchoice.com/blog.html I consult with patients all the time and let them decide; I spend my time showing data, showing flaws, showing differences of opinion, and then let them decide. They do so much better than docs; their incentives are their’s only. Time for docs to quit talking about docs.

  5. Michael, HMO’s of today aren’t expanding at the rate I saw when I was in practice. There are reasons for that and some of those reasons might have to do with the fact their prices went up when the public demanded better care and less denials. People do gravitate to what suits their needs professionally and morally so things are rather fluid.

    I know all too many horror stories.

  6. Steve, Paul provides a short response. Do you believe your personal choice for your personal care doesn’t matter?

  7. I am very fortunate to have a great cardiologist who I initially met as one of the doctors who perform our annual corporate physicals. He told me that close to 80% of his practice is primary care. The guy is a genius, in my opinion, who seems to know almost everything about medicine. He has a terrific referral network and once rattled off a list of doctors from hospitals and private practices around NYC that he refers patients to for everything from cardiology to cancer to orthopedics to urology, to neurology, to dermatology, etc.

    When I asked him how he came to know all these high quality doctors to refer his patients to, his response was “you just do.” Exactly how he and other high quality PCP’s actually do that and how long it takes them to build a quality referral network, I have no idea.

    I’ve also read that good PCP’s generally try to make it their business to build a high quality referral network because they don’t want to send their patients to lower quality doctors. While I usually don’t ask, when I meet a specialist for the first time, I wonder how many doctors and nurses he or she has as patients.

  8. Steve,
    In the old days….before ACA and Macra and HiTech etc. etc it was much easier to have a personal relationship with a doctor who knew who he/she trusted to refer for a consult. It wasn’t markets that eroded this…..it was all our policy wonk reformers with each wave of acronyms passed by congress foisted on docs and patients.

  9. Barry – I am not a cardiologist and cannot comment on any of the care situations being described (with any authority), however as to who you can trust… that is precisely the job of your PCP. Anish is correct many just write for a generic cards consult, however there are diligent docs out there (you know at least one 😉 who spend time with their complicated cardiac patients discussing whether or not to have reparative surgery at which time, how to obtain second opinions from pediatric cardiac surgeons etc. at different hospitals throughout the country. As I often say, find a darn good PCP and you will know him or her when you see them. Otherwise, it is not terribly comforting as Anish has suggested.

  10. It’s probably not a willful delay in treatment. Rather, they operate on a “capitated” basis and haven’t no immediate incentive to invest in the “means of production.”

  11. There are criteria that distinguish unstable from stable angina. These relate to the temporal onset of the symptoms, and workload required to get angina. There is clearly some judgement required as well. With regards to knowing who to trust – I don’t know the right answer. Used to be that a trusted pcp who as very invested in you would send you to the right person – I’m not sure if that model really exists anymore. PCP now rights for a generic cardiology consult – and you see whoever in clinic can see you first available. It then becomes a luck of the draw – not terribly comforting..

  12. There will be no perfect study to convince you I fear. I can only point to the fact that there is a reasonable failure rate for fibrinolysis that is much higher than stenting in the setting of acute coronary syndromes. I’d also note that the safety of coronary stenting continues to improve, especially as operators have move to using the radial artery in the wrist as an access point. Having watched the sequelae of large myocardial infarctions that went untreated, my general feeling from review of the data as well as experience is that I would favor the most effective, safest therapy that can be provided the most expeditiously. In 2017 that would seem to be a coronary stent. I freely admit, however, that being trained in cardiology is certain to induce bias that is hard to overcome.

  13. Sounds like this particular patient is in good hands. There is so much nuance to these decisions that I haven’t touched on – how well collateralized the area at risk is – what exactly optimal med. mgmt is (not just pills)… Calls for a very individualized approach that may be poorly suited to trials. Not to harp on it but only 9% of the folks who were screened in the COURAGE trial were randomized..

  14. I saw a patient today who, without signs or symptoms had developed a total coronary occlusion of his left anterior descending. On maximal medical therapy he still had effort angina and very poor exercise tolerance. After stenting, he is gaining stamina and is angina free off his long acting nitroglycerin. He doesn’t think his stent was unnecessary, but we waited six months to see what medications would do for him – he had some collateral blood supply.

    Of course this man is also on maximum statin and blood pressure control, he is working on his weight and he was recently started on C-PAP for sleep apnea.

    Without a trial of conservative therapy and without lifestyle and other secondary prevention measures, a stent would have been like a urologist treating a patient for one kidney stone after another without analyzing the stones and guiding the patient toward preventive measures based on the composition of the stones – profitable procedures and withholding information about how to prevent needing them again.

  15. You can’t very easily. And you can’t take it back if you don’t like it. Yet, somehow markets are supposed to work when we such information asymmetry. Go figure.

  16. Delay in treatment means profit.

    Apparently I saw a different side of HMO’s.

  17. “how can I be reasonably confident that the care I’m getting (or not getting) is appropriate?”

    You can never be sure and even the experts will frequently disagree. However, today there is a political push to do less and politics isn’t very scientific. Neither is the one shoe fits all methodology. We have loads of pundits here and elsewhere without training that can say whatever they want, but their useful knowledge is near nill. They pick and choose ideological statements that suit their positions.

  18. The way HMOs try to control costs is by making fewer physical resources available (increasing wait times).

    Physician incentives don’t play a major role, in my opinion, especially when the 2 systems coexist in the same community. HMO docs are unlikely to be that much more restrained than FFS docs

  19. I’ll offer a patient’s perspective. In 1999, after an angiogram showed diffuse disease, I had a quintuple CABG. I’ve been on aggressive medical management ever since. However, in 2005, I developed some chest discomfort. A stress test showed what my cardiologist described as a change from prior test results that was significant for me. An angiogram six days later found that one of the arteries re-occluded and a DES was inserted to clear the blockage. My question is how can doctors tell at a given point in time if the patient in front of him or her has stable or unstable angina?

    Even if it’s stable angina, I certainly understand that there is a huge difference between a problem in the Left Main or the LAD as opposed to one of the distal branches that’s the equivalent of the basement powder room. I also understand that cardiac procedures are reimbursed quite well even at Medicare rates. So, aside from having or not having a good level of trust between me and my interventional cardiologist, how can I be reasonably confident that the care I’m getting (or not getting) is appropriate?

  20. Interesting. The free market guys are always telling me that the HMO docs are heavily incentivized to justify withholding care so as to increase profits. I guess there are always exceptions.

  21. If the two physicians mentioned at the start of this blog post had engaged in this kind of respectful presentation of the evidence, perhaps there would have been less acrimony. I’m a strong EBM advocate well aware of the history of overuse, but this made a good case for thinking carefully about what overuse means today.

    Of course, this IS a long blog post, not a Twitter argument, so that may explain why there’s actual subtlety involved.

  22. Thank you, Anish.

    On the question of financial incentives, I began my interventional cardiology career working for a large HMO where stents were implanted by doctors who are on fixed salary. I then practiced for three years in a fee-for-service, private hospital environment. I have known the local community of interventional cardiologists quite well over the last 15 years, including those in private practice, those on salary at the large local HMO, and those on salary at the academic medical center where I trained.

    It’s my observation that in each sector, cardiologists vary widely in regards to their enthusiasm for stenting patients with stable angina. I am fairly confident that the local area cardiologist who would likely be labeled as the most prolific implanter of “inappropriate” stents is one who actually works for the HMO, although he might be tied with a cardiologist who is in fee-for-service private practice.

  23. Stent or medical management? Seems like there is another option you should mention. That said, since I practiced in the heyday of stents, with so many of our patients walking around with double digits numbers of stents, that I grew to be pretty skeptical about them. What you said here echoes what i have heard the cardiologists I trust also say. There is still a place for them, but we should probably be much more selective than we were 15 years ago.

  24. Excellent discussion of the controversy and its history, Anish. I agree with your conclusions — and certainly with the idea that ad hominem attacks have no place in discussions of the appropriateness of therapies.

    At the same time, PCI is far from the only therapy that we need to re-examine, or whose overuse may be pushed for financial reasons. The science is weak to non-existent to even negative on many tests and procedures that we take for granted and do in large volumes.

    It is a difficult discussion to have, especially to the extent that one’s professional identity and personal ego can get wrapped up not just being a doctor, but in being really good at a particular therapy. Someone who has a practice that mostly consists of putting in coronary stents several times a day for years on end is likely consciously or not to see themselves as an expert in putting in stents and a champion of their use — because, after all, look at all those patients who are doing better now. When someone else questions or even attacks the usefulness of the therapy, it is easy to feel personally attacked, as if what you have been doing all these years is not only useless, it is actually damaging to patients, and you are morally bankrupt for doing it.

    Ad hominem attacks make the discussion of proper use of therapies much harder.