Jun 19, 2018
Dr Carolyn Lam: Welcome to Circulation on The Run, your weekly podcast summary and backstage pass to the journal and its editors. I'm Dr Carolyn Lam, associate editor from the National Heart Center and Duke National University of Singapore. This week's issue is so special. It is an autopsy issue. I think it's actually the first of its kind in the history of Circulation. I am so pleased to have with me today Dr Jeffrey Saffitz from Beth Israel Deaconess Medical Center, who's the content editor for Pathology for Circulation and the guest editor for this entire autopsy issue. Welcome, Jeff.
Dr Jeffrey Saffitz: Thank you.
Dr Carolyn Lam: We also have Dr Lee Goldman from Columbia University Medical Center who wrote a beautiful perspective piece on autopsy. Thank you and welcome, Lee.
Dr Lee Goldman: Morning.
Dr Carolyn Lam: Jeff, could you start us off? I mean, an autopsy issue. How in the world did this come about?
Dr Jeffrey Saffitz: I think it really began by coincidence. The journal received submissions from several authors, each involving studies of autopsies, and the editors approached me and asked if we might consider grouping them together in a special issue focused on the role of the autopsy and cardiovascular medicine. I thought that would be a very interesting idea and this evolved into actually something much greater. Two additional papers came in focusing on the autopsy and I think looking at these papers in the aggregate, they represent what we can now consider to be the contemporary utility of the autopsy in understanding the way cardiovascular disease works. So I was particularly pleased that the editors agreed to group these papers into a single issue focused on the autopsy. We were really delighted that Lee Goldman agreed to write a perspective. He has had a longstanding history of studying the role of the autopsy and I hope the readers will find this to be a really interesting and useful issue which will, I hope, chart the course for future discovery.
Dr Carolyn Lam: Just listening to you, I love the way you say it's a contemporary look at autopsy. I mean, we covered things like molecular genetic, proteomic, autopsies, even like electronic autopsies using device. That's really cool. Lee, thank you again for sharing your time and incredible perspectives with us. The long history of autopsy. Do you think it's still necessary now?
Dr Lee Goldman: Maybe give some perspective. I first got involved in this a number of decades ago, when as a junior faculty member, I was assigned to be on the medical audit committee of the hospital where I saw patients as a cardiologist. And two of the senior people in the committee got into a debate about whether autopsies were still important given the advent of CT scans and other modern diagnostic technology. And to listen to them debate for 15 or 20 minutes, I finally had the temerity to pipe in and say we can actually study this, and so we did. We looked at autopsies in three different decades: 1960, 1970, 1980, and much to everyone's surprise, I think found, A. that the rate of which autopsies found diagnoses that doctors had missed and for which treatment would almost certainly have prolonged life was about 10 percent, and it was 10 percent, 1960, 10 percent, 1970, 10 percent, 1980.
But the difference was that doctors were missing different diagnoses. The things that got missed in 1960, and where autopsies showed there were being missed led to better diagnostic approaches and those things were rarely missed in 1980. But since people stayed alive longer, they got new things that we didn't really know much about in 1960. A big difference, fewer people missed heart attacks, pulmonary emboli, and things of that sort, but far more people had missed infections, especially fungal infections that were complication of multiple antibiotics or immunosuppressive therapies.
And so, as I followed this in 1980, if you will, to now, 2018, we find this gets recapitulated over and over again. Medicine moves forward, things we used to miss, we no longer miss, but people still die, and they still die from things that we don't always diagnose. We've done statistical analyses to show that probably the rate of misdiagnosis is going down a little bit, but it's still substantial and we still estimate that thousands of people each year die in the U.S. from things that are not what the doctors thought they had, and if that diagnosis had been made, the patient would have lived longer.
Dr Carolyn Lam: Lee, I just love that perspective. I have to say, it's really humbling. I mean, 1960s and so on would predate me as well, so I'm really humbled, and I love that reminder. Jeff, in fact, quite a number of our papers illustrate exactly what Lee said. We have four papers just dealing with sudden cardiac death, and that is still what diagnosis was struggled with. Could you tell us a little bit more about those?
Dr Jeffrey Saffitz: Yes, of course. I think we all recognize that sudden death remains a huge public health issue. We also realize that most people who die suddenly and unexpectedly don't do so in the hospital when they're being followed and monitored; rather, they die out in the community, and in many cases, these are individuals in whom major risk for coronary disease or other potentially lethal cardiovascular conditions was really not known. So I think it remains a major public health issue, and we still have a great deal to learn. So perhaps it's not surprising that four of the five papers involved autopsy studies of sudden death victims of individuals who died out in the community. A couple of them focused on sudden death in young people.
We know that these individuals often will have familial diseases, and the autopsy has been one mechanism for studying these individuals, so one of the papers from Michael Ackerman at Mayo Clinic, advanced the concept that they started many years ago, the so-called molecular autopsy in which they apply a whole exome sequencing in cases of sudden unexpected death in young people defined here as age under 40, and they identified some rare variants which were likely to be of potential pathogenic significance in sudden death. A related paper from Junttila et al in Finland looks at the finding of myocardial fibrosis in young victims of sudden death. They identified several cases in which that was the only structural change in the myocardium, and when they applied next gen sequencing, the identified variance that we typically associate with the familial non-ischemic cardiomyopathies, arrhythmogenic, dilated, and hypertrophic cardiomyopathy. But the key insight here is that we traditionally think of these diseases as having rather characteristic structural changes which we can recognize at autopsy. What they showed is that those structural changes might be limited to nothing more than some fibrosis. And so the key here is that this expands our potential opportunity to recognize these familial cardiomyopathies, and the overarching idea is we use the autopsy to serve the living. This is a way to gain information at autopsy that we can then use to help family members and other individuals by virtue of the insights gained at autopsy.
Dr Lee Goldman: When we did the estimates in my editorial, and I estimated that roughly 28,000 people die each year in America with diagnoses that doctors missed and for which treatment would have been different if they hadn't missed it, that's really based on, I'll call traditional autopsy methods, which are anatomical, include microscopic evaluation, include culture, but it's not historically included genetic testing. I believe, as these articles show, that the advent of genetic testing, which you could argue could have been done while the patient was alive, but we're not quite there yet in terms of testing everyone's genome, now help you autopsies find even more things that might've been missed. And as you just heard, also can have important information for the family. So, one of the issues you often get into in autopsies is what's in it for the family, and one of the problems here is that the pathologists don't get paid. For the family members, it's mostly an aggravation. The doctors are worried they're going to get sued if something that gets found. And so, to make this work you need to bring in some incentives. Doctors not getting sued if they find things because they should get credit for trying to learn more, some way to reimburse reasonably pathologists and hospitals who do the autopsies, and the understanding of family members that they not only will perhaps be more reassured about what happens to the loved one, but also may learn things that will affect their future, because certainly, these cardiomyopathies, instead of them being diagnosed, are familial and oftentimes will lead to testing and hopefully interventions in family members that'll be to their benefit.
Dr Carolyn Lam: Lee, what great comments about bringing this into the clinical perspective and I just love what you said, Jeff, about autopsy for the living. That is just a quotable quote. That's so cool. I noticed that you did ask Dr Judge to write an editorial specifically about bringing autopsies into the molecular genetic era. So I just want to encourage all our listeners to make sure you read that as well. But Jeff, back to you about the other two papers.
Dr Jeffrey Saffitz: Well, I think one that I found particularly significant is this idea that nowadays, patients come to autopsy with implantable cardiac electronic devices, and the point of this paper is that interrogation of these devices postmortem can provide really important information about the cause and timing of those deaths. I think the reality is that most pathologists who do these autopsies are entirely unprepared or ill equipped to do such interrogations, and so I think the point of this paper is simply to encourage pathologists who do these autopsies to develop partnerships and relationships with cardiologists who are able to get this type of information from these devices. And again, it not only provides information about what happened to that one individual and what the death was all about, but it provides important information to the family and potentially information that allows the family to recognize particular risks that might impact the living members. So I thought this was just another really interesting example of how information that is potentially available at autopsy may not be fully utilized, and I hope that this paper will have an impact in that regard.
Dr Carolyn Lam: That's great. Lee, did you have any perspectives on devices and its role in autopsy now?
Dr Lee Goldman: I guess that the point that I would just reinforce would be that diagnostic technologies, including the ability to monitor someone's heart rate, have helped us diagnose things that were missed in previous eras, but medicine is always pushing the frontier forward, and as long as we develop new therapies, develop new devices, there'll be new things to learn. I want to make one other point about what I'll call overconfidence in diagnoses. The published statistics for the accuracy of most diagnostic tests are based on what doctors think the diagnosis ends up being, not the autopsy, which is the ultimate gold standard. So, if you actually go through some not-so-complicated arithmetic, you'll find that many of the tests that we think are almost perfect at finding things really aren't because the people who die with those things found that autopsies that the test missed. There's something called a virtuous circle, there's also a vicious cycle. There's a bit of a vicious cycle here that if you don't do autopsies to be sure you aren't missing things, you become overconfident in the tests that you think are finding them, and therefore think you already know everything and don't need to do an autopsy. To me, in some ways, that's the most perverse result of the plummeting autopsy rate, which, by the way, can be linked directly to changes in how hospitals get accredited, that in prior years there was a minimal autopsy rate required for accreditation. When that was removed, not surprisingly, autopsy rates plummeted, and now, most autopsies done in the US are not done in hospitals because doctors aren't sure what's going on. They've done by medical examiners as part of the laws for autopsies least being considered and people who die without having had a medical attention to some degree.
Dr Jeffrey Saffitz: You are exactly right on all of these points. I'll just say this is the point of one of the other papers from Tseng et al. This was a prospective autopsy study of sudden death in the city and county of San Francisco, and what they showed here is that only about half of the deaths that were considered to be sudden cardiac deaths as defined by the conventional criteria actually turned out to be deaths due to a rhythmic disorder. So Lee's point is exactly right. Doctors think they know a lot of things, but they're not always right about that, and the autopsy is probably one of the best ways to bring some quality control to this, and to really provide, I think, objective data that often is the case flies in the face of what the previous thinking was, and I think this paper in this issue of Circulation really brings that point home very clearly.
Dr Carolyn Lam: Yikes. OK, so here I am, I practice in Asia, and I think the autopsy rates are even lower, so this is a great wake up call for me just listening. Let's switch gears a little bit. How about the paper by Dr Herrington? Now this goes to a proteomic bisection almost of maybe preclinical disease and atherosclerosis. Would you like to comment on that on, Jeff?
Dr Jeffrey Saffitz: In the perspective that I wrote with Gaetano Thiene, in addition to looking at the history of the autopsy, we looked to the future and just considered briefly what role will the autopsy play going forward, and I think the paper by Herrington is a great example of how we can use the autopsy to learn so much more about the way human disease works. The basic idea here is that something like coronary artery disease or atherosclerosis, we think of as being a disease that only involves the blood vessels, and we tend not to recognize it until it is rather advanced and clinically manifest, but we recognize that these diseases begin decades before they become clinically manifest. We really don't know how to identify the earliest antecedents, and without knowing that we really, I think, very much limit our ability to identify the disease way early before it becomes clinically manifest, and then be able to practice preventive measures and intervene to prevent the disease from occurring.
So, what this paper showed is that it's an application of high-throughput proteomics looking at coronary artery and aortic samples obtained at autopsy, and these authors identified particular changes in proteins that they then were able to show in a prospective independent clinical cohort were able to predict the development of coronary artery disease. So I think going forward, we are going to redefine our understanding of human disease by learning about its earliest expressions and its full systemic distribution, and in doing so, we'll be much better prepared to diagnose earlier and intervene and prevent disease. So I think this was a great example of how the autopsy can help in that effort.
Dr Carolyn Lam: I feel like we are going full circle in history and going back to learn about how to go forward. I don't know if I expressed that well, but I am just in awe of what I've learned from both of you. Thank you so much, Jeff, for putting together this amazing issue, and thank you so much, Lee, for sharing your perspectives. Thank you, audience, for joining us this week. You've been listening to Circulation On The Run. Don't forget to tune in again next week.