I’m in shock. Dr. Nortin Hadler of the University of North Carolina has proclaimed that the era of coronary angioplasty is over.
Poor, poor angioplasty…you were barely 36 years old, but you’re no longer needed. Guess we’re going to have to find a new name for our web site!
Of course, my first thought upon seeing this Op-Ed piece posted today on The Health Care Blog was that it was yet another article railing against the overuse of stents in patients with stable angina. Whenever I encounter this line of reasoning, I point to the fact that, according to the National Cardiovascular Data Registry (NCDR) 71% of the 600,000 angioplasties performed annually in the United States are done urgently: that is, for patients in the midst of a heart attack or acute coronary syndrome. Virtually every practitioner in the field of heart disease agrees that the gold standard treatment for ST-elevated Myocardial Infarction (STEMI) is angioplasty.
But, as I continued to read the article, I saw that Dr. Hadler’s premise is that angioplasty may not be necessary even in these severe cases, that clot-busting drugs could achieve the same outcomes at less expense and less risk for the patient. He spends most of his article discussing the recent STREAM study. I’ll get back to Dr. Hadler in a minute, but first a little background on STREAM, which I first heard presented at the American College of Cardiology meeting in March and which was subsequently published in the New England Journal of Medicine.
STREAM
The STREAM (Strategic Reperfusion Early after Myocardial Infarction) study compared Primary PCI (angioplasty as the initial treatment for STEMI) with fibrinolytic therapy (clot-busters) followed by diagnostic angiography and, if deemed necessary, angioplasty — but non-urgent angioplasty, done up to 17 hours later. Almost 2,000 patients in 15 countries were enrolled in the STREAM study. Very importantly, these were patients who were not able to undergo Primary PCI within one hour after ambulance pickup or presentation at the Emergency Room.
In other words, these were patients outside the optimal therapeutic window in which angioplasty is known to be the best way to stop a heart attack. These were patients who were not near a PCI-capable hospital and had to be diagnosed at a local Emergency Department and then transported to a hospital with a cath lab. Half of these patients received clot-busters before transport. If they arrived at the PCI-capable hospital and the clot-busters had not opened up the blocked artery, they underwent urgent angioplasty. This was the case in 36% of the clot-buster cohort. This is an important figure to remember when I get back to discussing Dr. Hadler which, I promise, will be in a minute.
The results of STREAM at 30-days were that there was no difference in outcomes for the two therapies. There was one difference actually: in the beginning of the study, older patients (over 75) in the clot-buster group experienced significantly higher rates of intracranial hemorrhage, a serious type of stroke caused by the drugs. Once discovered, the researchers halved the dose of the clot-busters and the excessive bleeding risk was lowered.
Are Reports of PCI’s Demise Greatly Exaggerated?
Yes. To say that the STREAM trial supports Hadler’s conclusion that angioplasty may be replaced by fibrinolytic therapy is completely incorrect. The population studied was very specific and excluded a significant number of STEMI cases (for example, those who could get an angioplasty quickly). Even the authors of the study wrote, “This trial was designed as a proof-of-concept study. All statistical tests were of an exploratory nature.” And, oh yes, the clot-busters didn’t work in 36% of the patients who received them.
An excellent commentary accompanying the study in NEJM was written by Dr. Deepak Bhatt of the VA Boston Healthcare System, Brigham and Women’s Hospital, and Harvard Medical School. In his editorial, titled “Timely PCI for STEMI — Still the Treatment of Choice,” Dr. Bhatt makes an important point: the clot-buster strategy was judged to be non-inferior. It was not better than PCI; it just was not worse. There was also the increase in intracranial hemorrhage in elderly patients, even though the dosage was altered during the trial (this is not news — elderly heart attack victims are rarely given clot-busters for this reason). Most importantly, why not strive “to design systems that allow patients to receive rapid PCI uniformly” since that is known to be superior?
These seem like obvious points. But instead Dr. Hadler seems intent on targeting angioplasty as an unnecessary and overused treatment and he makes over-the-top statements such as:
“If angioplasty, with or without stenting, was a pharmaceutical instead of a procedure (device), is there any regulatory agency that would license it without a randomized controlled trial demonstrating adequate efficacy?…If some august governmental body were to declare interventional cardiology and cardiovascular surgery for coronary artery disease a failed experiment that should no longer be supported, I would applaud and display the scientific basis for the shift in policy.”
Of course, Dr. Hadler, the author of a number of books and a highly respected academic clinician in the field of musculoskeletal disorders, quotes several other sources to support his opinion, and you should read his piece (and by all means post your opinion in the comments below). But my guess is that the millions of people whose heart attacks have been stopped by stents, who lives have been saved and who are able to live full and productive lives without the morbidity imposed on them by a severely damaged heart muscle, my guess is that they might have a somewhat different opinion.
Of course, Hadler’s article is just an Op-Ed on The Health Care Blog, certainly not the largest circulation journal — and I’m probably giving it even more readership by posting this on my blog. However, of more import and concern is the fact that Hadler recently provided the team at Bloomberg News (they have much much higher circulation) with some juicy anti-angioplasty thoughts for their series on over-stenting. (He seems to be one of the “go-to” sources on this subject.)
I’ll be writing soon and in more detail about the Bloomberg series, but I would like to point out that in two weeks, 12,000 interventional cardiologists from around the world will be gathering in San Francisco at the 25th Annual TCT meeting, where they will spend 5 days discussing studies, parsing data and attempting to refine the practice of interventional cardiology, without doubt one of the most data-driven specialties in medicine today. Unfounded pronouncements such as today’s obit for angioplasty not only insult this professionalism, but do a disservice to patients by dissuading or delaying them from seeking the most immediate and best treatment for their heart attack.
And such a delay can cost lives. And that’s not just my opinion….
Agree that the analysis and opinion are flawed. Getting tired of the witch hunt against PCI that essentially removes all clinical judgement and expertise from the hands of experienced clinicians. There is no longer any ‘art’ in the practice of medicine. We are missing the forrest through the trees by attemmpting to apply population based guidelines to predict individual outcomes losing site of the fact that not all clinical opportunities are the same and that sometimes, the physician experts do indeed know what to do.
A clear and helpful dissection of the limitations of the STREAM study.
The language of Hadler’s original piece on THCB sounds somewhat shrill, the data marshalled in support somewhat selective and the conclusion obvious (even to a non-expert in the field) as somewhat overblown.
That said, the one thing it seems you cannot do in refuting his claims is point to a clear and unambiguous RCT of emergency PCI versus medical management that shows where PCI is superior as you claim. Or at least you dont cite such evidence, so one might reasonably assume it doesnt exist.
On that basis, the point I found most interesting from Hadler’s original post still stands: major pathway decisions in interventional cardiology remain based on the experience and skill of doctors (which varies widely from place to place) rather than on a solid and irrefutable evidence base that would be necessary to see a new drug (as opposed to a new surgical procedure) approved and then widely adopted.
David – Thanks. In order to keep the article to a manageable length, I decided not to get into quoting all the studies that have been done over the past decade-plus that have shown a clear benefit for pPCI (Primary PCI), and I probably should have! Most clinicians in the field know these well.
One of the most important ones was published in Circulation in 2003: the DANAMI-2 study which is even referenced by Dr. Hadler in his article. Interestingly, however, Hadler doesn’t footnote it, perhaps because, if you read the study, you’d have seen that in a cohort of 1,572 patients, the primary endpoint of reinfarction, stroke or death at 30 days was reached in 8.5% of the patients in the angioplasty group, as compared with 14.2% in the fibrinolysis group, a very significant difference. The difference between the therapies held at the 8-year followup analysis.
Dr. Hadler dismisses the DANAMI-2 study somewhat, saying that it was biased against lytic therapy because the more stable the patient the more likely they were to be transferred expeditiously. The only problem with his point is that the most critical patients specifically were excluded from DANAMI-2 for that very reason. Moreover, the DANAMI-2 trial was stopped prematurely on ethical grounds because the interim results showed such a clear benefit — in other words, it was judged unethical to continue randomizing patients to lytic therapy when pPCI had shown such a pronounced benefit. So Dr. Hadler’s call for a randomized controlled trial seems uninformed by the known data. And there have been other trials, almost all of which have mirrored these results.
Additionally, the DANAMI-2 trial enrolled patients during the early era of pPCI, from 1997-2001. Transport, procedural and pharmacologic pPCI therapies have advanced significantly since then; most likely the mortality benefit of pPCI is underestimated by DANAMI-2.
A final point is that, as with all medical issues, it’s not either-or. It’s not pPCI vs. fibrinolytic therapy. The two are both strategies and in specific locales, patient populations, length of time since symptom onset, etc., etc., etc. both should be considered or combined.
Thanks for the additional clarification.
It boils down to the matching of the patient population. Hadler, it seems, believes the patients in DANAMI-2 were not matched (or choses to draw that conclusion, anyway) – and you argue they were well matched.
Certainly, my pre-existing bias without fully reading the literature was that emergency PCI was well accepted to be beneficial not just as a matter of opinion but as a matter of fact. Your citations seem to indeed confirm that, and suggest that Hadler is more missionary than visionary.
My guess is that 98% of the people who’ve been stented would be of the opinion that it saved them from death or “severely damaged heart muscle,” even though statistically many of them must be wrong.
Jane – Thanks for your comment. I would point out that we are talking about STEMI patients here, not about all people who have had angioplasty. In stable patients, angioplasty and stents are a way to relieve symptoms that aren’t being adequately relieved by medical therapy alone. In STEMI or urgent cases, studies have shown (see my reply to David above) that pPCI is in fact a “life-saver” and reduces both mortality and morbidity, when done quickly.
After many years working in ICU and seeing the miracle of modern cardiac care, it is so discouraging to see what is happening in the insurance controlled health care we receive. Trying to discourage mammograms, PSAs, and now cardiac stents, is just another effort by the insurance companies to avoid paying claims for these services. It’s always about the money…
If memory serves, there was an editorial in JACC, written in perhaps in the mid 1990s, perhaps commemorating a major anniversary of what was initially known as a Greuntzig procedure (like the piano player in “Casablanca”, I don’t rightly remember) that noted amongst other things, how fortunate it was that angioplasty had worked out pretty well since its inception, as it had not been subjected to the scrutiny of clinical trial. It would appear that from the outset, concerning PCI, that it was better to be lucky than good. From a usefulness standpoint, it helps that it turned out to be good.
How then is Dr. Hadler’s statement, “If angioplasty, with or without stenting, was a pharmaceutical instead of a procedure (device), is there any regulatory agency that would license it without a randomized controlled trial demonstrating adequate efficacy?” so over the top?
It is certainly appropriate to ask questions about the efficacy of any therapeutic measure. It’s equally appropriate to determine for PCI when “time is muscle” begins and ends with respect to risk of intervention vs. risk of non-intervention.
Is the issue of referral, less pressing in STEMI treatment than elective PCI, addressed anywhere in the literature? Those with the greatest likelihood of remunerative benefit are the biggest proponents of procedures which are self referred.
Jon – Not sure which JACC article you are referring to. There were many around 1997 which was the 20th anniversary of the first PTCA, performed by Andreas Gruentzig in Zurich in 1977 — it was also the year Angioplasty.Org went online in celebration of that milestone. In my above reply to David, I cited studies that showed clear benefit for pPCI in STEMI. As for proving the benefit of angioplasty, not just in STEMI but in the whole spectrum of coronary artery disease (CAD), there are many studies that have been and are being done. For example, even in the COURAGE Trial, oft-quoted by those who want to show superiority of medical therapy over angioplasty for stable heart disease, fully one-third of the medical therapy only patients crossed-over to angioplasty for symptom relief. Importantly, this whole issue is a moving target because of ongoing and constant advances not only in angioplasty techniques, protocols and devices but also in the efficacy of new drug therapies. And new diagnostic tools, such as fractional flow reserve (FFR), are changing the way the decision to stent or not is made.
To repeat what I said in my blog, the practice of interventional cardiology is without doubt one of the most data-driven specialties in medicine today. Of course, it’s good to be lucky, but the importance of hard proof is paramount. This was Gruentzig’s mission also. He knew that he needed solid data to prove to the medical community that angioplasty worked. That’s one reason he brought Dr. Gary Roubin to Emory to head up research (he ran the first surgery vs. angioplasty trial). If you were to attend the TCT meeting in two weeks in San Francisco, you’d see hundreds of presentations and posters with charts, tables, graphs, p-values, etc. There’s no lack of data in this field.
For now. What will “kill” PCI is precise dose titration of fibrinolytics to maintain them in therapeutic range. Universal dosage is very crude action: in some patients it results in suboptimal firbrinolytic level, while in others to dangerously high hyperfibrinolysis, leading the hemorrhagic complications.
One of the values of fibrinolytic therapy is that it can be applied quickly and easily in a local non-PCI-capable hospital, in cases where the cath lab is just too far away. In the STREAM trial, lytic therapy could be started 78 minutes sooner than PCI. But a question I have is how a “precise dose titration” can be determined in an emergency situation at a regional community hospital that would still make it “quick and easy.” And how that will “kill” PCI.
With apologies for the commercial plug-in: our company has the first portable device for near instantaneous measurement of fibrinolytic activity in blood. Our goal is to integrate it with an infusion pump for automatic real-time dosage adjustment, to keep therapeutic range.
Obviously, I have a strong interest in the field, but also am biased because of this product. Should’ve disclaimed in the original comment. Sorry, again.
Re: death of angioplasty….easy to talk about or write about the statistical nuances of such therapy. Harder when you have a 99+% mid-LAD lesion, a large peri-apical nearly akinetic LV segment on ECHO, and exercise induces terrible anginal pain and shortness of breath. Will my stent allow me to live longer? Perhaps not as Dr. Hadler opines (and he may simply be right).
I don’t care nor did I care the day we decided (yes, we, my doctors, my wife, and I) to procede with angioplasty and stent placement. Have I looked back? Yes, but when I do it is not over my shoulder and I see as clearly behind me as ahead of me. I am grateful the technology exists and that it was available to me. I am certainly grateful for the time, concern, empathy, and just plain listening that my cardiologist provided that day (first day I met him) and every meeting since. But I am most grateful that I “lost” the pain and SOB promptly and was able to return to exercise, work, and full life in a few days. FYI, the food in hospitals is still pretty mediocre. Perhaps we could work on that?