Apr 2, 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.
Today's feature paper is about statins, and it's the first population-based study to show a dose-dependent benefit on amputation and survival in peripheral artery disease. Very important data and a very important discussion coming right up after these summaries.
The first original paper this week indicates for the first time that the natural history of coronary stenosis is better predicted by physiologic information by FFR, or fractional flow reserve, than by anatomic information from angiography. First author, Dr. Ciccarelli, corresponding author, Dr. DeBruyne, from OLV Hospital in Belgium compared the values of angiographic diameter stenosis and of fractional flow reserve in predicting the natural history among 607 patients from the FAME 2 trial who had documented stable coronary disease and in whom no revascularization was performed. The primary end point was defined as vessel oriented clinical end point at two years, and this was a composite of prospectively adjudicated cardiac death, vessel-related myocardial infarction, vessel-related urgent and non-urgent revascularization.
The overall results showed that FFR predicted the natural history better than diameter stenosis. In addition, among the stenosis with mismatch between diameter stenosis and FFR, more than half had a low FFR in the presence of an angiographically mild stenosis and the rate of primary outcome was higher in those with reduced FFR regardless of whether diameter stenosis was significant or not. The take-home message is, therefore, that measurements of FFR should be considered not only an angiographically intermediate stenosis but also perhaps a mild or severe stenosis by visual evaluation.
The next study provides population-based data on cardiovascular outcomes and risks after initiation of a sodium glucose cotransporter-2 inhibitor, or SGLT2 inhibitor. First and corresponding author, Dr. Udell, from University of Toronto, and his colleagues, performed population-based cohort study among type 2 diabetes patient with established cardiovascular disease and newly initiated on antihyperglycemic agents within the US Department of Defense Military Health System between 2013 and 2016. After propensity matching, more than 25,250 patients were followed for a median of 1.6 years. Initiation of SGLT2 inhibitors was associated with a lower all-cause mortality, lower hospitalization for heart failure events, lower major adverse cardiovascular events, but higher below-knee amputation risk. Findings underscore the potential benefits and risks to be aware of when initiating SGLT2 inhibitors. Importantly, it remains unclear whether the risk of below-knee amputation extends across a class of medications as the study was not powered to make comparisons among individual treatments.
The next paper reports results of the redefined trial, which is the first trial to study the effects of renin-angiotensin-aldosterone system inhibitors in adults with tetrology of Fallot and mild right ventricular dysfunction in the absence of severe valvular lesions. First author, Dr. Bokma, and corresponding author, Dr. Bouma from Academic Medical Center Amsterdam, and their colleagues, studied 95 patients in the redefined trial and found that 150 mg of losartan daily did not significantly improve the primary outcome of right ventricular ejection fraction change compared to placebo. There were no significant treatment effects on secondary outcomes of left ventricular ejection fraction, peak aerobic exercise capacity or NT-proBNP. However, in a post hoc analysis, losartan was associated with improved right ventricular ejection fraction in a subgroup of 30 patients with nonrestrictive right ventricles and incomplete remodeling. The conclusion is, therefore, that losartan had no significant effect on right ventricular dysfunction or secondary outcome parameters in repaired tetralogy of Fallot. Future larger studies may determine whether there might be a role for losartan in specific vulnerable subgroups.
The final study reinforces that vesicle trafficking plays an essential role in the signal regulation of pathologic hypertrophy and identifies a novel potential target in this process. This novel target is the transmembrane BAX inhibitor motif containing 1, or TMBIM1. First author, Dr. Deng, corresponding author, Dr. Li, from Wuhan University in China, and their colleagues, found that TMBIM1 expression levels were substantially decreased in both clinical and experimental hypertrophic hearts. Mechanistically, TMBIM1 interacted directly with tumor susceptibility gene 101 and accelerated the formation of multivesicular bodies to degrade activated toll-like receptor 4. Toll-like receptor 4 degradation in turn was essentially for the progression of cardiac hypertrophy. Importantly, expressing TMBIM1 in monkeys via lentivirus protected their hearts from aortic banding induced cardiac hypertrophy. In summary, these findings shed light on the role of vesicle trafficking in signal regulation during cardiac hypertrophy and provide a novel therapeutic target for treating hypertrophy.
That wraps it up for our summaries. Now for our feature discussion.
Peripheral artery disease, a disease that affects more than 200 million individuals worldwide and associated with a high risk of cardiovascular events and death and, of course, the much feared amputations. Yes, statin guidelines for peripheral artery disease are largely based on coronary artery disease or stroke data. Well, today's feature paper really addresses an important knowledge gap between statins, doses, amputation survival in peripheral artery disease. I'm delighted to have the first and corresponding author, Dr. Shipra Arya from Stanford University School of Medicine and, of course, our favorite, Dr. Josh Beckman, Associated Editor from Vanderbilt University.
Now, Josh. I understand there's a bit of a back story of how this paper came to circulation. Want to share?
Dr Josh Beckman: Oh, absolutely. First of all, I have to say that one of the jobs of an associated editor is someone who kind of goes antiquing in every single store. Every place I am where people are presenting really good science, I'm kind of scoping it out. I'm interested. I want to see what's going on. I like to talk to the people who are doing the work to see how they're thinking about it, and I was lucky enough to see Dr. Arya's presentation. I think it was at an ATVB meeting, wasn't it?
Dr Shipra Arya: That's right.
Dr Josh Beckman: I thought that this is an incredibly cool piece of work, and I basically hoped, I prayed, I asked. I said, "You know, maybe you should send this to us because we would really like to see the full manuscript," because inside I hoped that it would be just as impressive when it was written out as a full manuscript as it was when she was discussing it at the meeting. And, lo and behold, we were lucky enough that she submitted it to us and you can see the results online right now.
Dr Carolyn Lam: Indeed! Well put. Shipra, with that kind of lineup, please, tell us about your study and what you found.
Dr Shipra Arya: Thank you for that invitation to submit to Circulation because initially I wasn't sure if Circulation would be interested in my work, so it was really great to hear when Josh said that this is something that it would certainly consider. The basic premise was to try and find out whether high-intensity statins as defined by the 2013 lipid guidelines, they would also have limb protective effects for PAD along with reduction mortality. As you said in the introduction, most of the data comes from either coronary data or comes from small groups of PAD patients, but never from such a large population.
We identified about 150,000 veterans in the National VA database from 2003 to 2014 and excluded people who didn't have a diagnosis of PAD before 2003, and why this was such a labor of love was also to figure out how to identify the certainty that people had PAD and then getting into their pharmacy files and trying to parse out whether they were on high-intensity, low, moderate, or no statin. Initially, I had done the analysis of no statin, but then after review and discussion, it became clear that we needed a control group, which was people who were also on some guideline-directed therapy and not just no statin because they could be patients who were the noncompliant patients and who don't show up to the doctor's visits, and that's why they do poorly.
That's why we chose a control group which were on antiplatelet therapy, at least aspirin or Plavix, any other antiplatelet agent. Even in that comparison, we find that after risk adjustment, patients who are on high-intensity statin had a more than 30% risk reduction of amputation as well as about a 24, 25% risk reduction of mortality compared to people who did not take a statin but at least took an aspirin. Low to moderate intensity statins were also effective, about 20%. Risk reduction in both amputation and mortality, but high-intensity statins when directly compared to the low to moderate intensity statins outperformed them.
Just to be sure of our findings, we did it so many different ways. We did the Cox modeling. Then we did propensity matching that which person is more likely to receive the statin versus the other. Then we did subgroup analyses where we put people in different subgroups that people who had coronary artery disease as an indication, maybe that's why they were on these statins. But, people without coronary artery disease also same association [stack 00:11:12]. We were pretty confident in our findings, and that's why we sent it to Circulation.
Dr Carolyn Lam: Wow. You know, Josh, you are the best at putting papers like this into context and really expounding on the significant. Tell us, why did this catch your attention so much?
Dr Josh Beckman: Every time I think that statins have become just a standard part of therapy for patients with atherosclerosis, the first thing I noticed in this paper was that there were so many people who were still not on any statins or people who were on homeopathic doses of statins, and I can't understand how that happens. I think the mortality data was nice and consistent, but the amputation data is what really made a big difference. I'll ask Dr. Arya, but in my impression, the literature has been sort of back and forth as to whether or not statins really reduced limb outcomes. Your paper, I think, was clearly the largest sample that had taken a look at that question. Can you sort of separate out your papers from some of the previous work in that area?
Dr Shipra Arya: Sure. I would add that a lot of work about amputations has been coming out from vascular surgery data, and a lot of that work just focuses on short term outcome for limb loss. They look at 30 days. Maybe they'll go look up to six months to a year, but actually patency of bypasses, patency of vessels is a long-term phenomenon. Much like mortality that can happen years later, your amputation risk can happen years later, too. I think what separates us is the lifetime followup for these patients, and we are looking in a cohort of patients who are in this veterans' healthcare system so the data is automatically getting captured even if they get their care outside. Records do make it back and diagnoses do make it back. It's the VA [inaudible 00:13:03], and we did some sensitivity analysis to show that, yes, most of the veterans we have in [inaudible 00:13:09] actually get their care and have data being added continuously into the corporate data warehouse.
That was something I think that lent to the power of making the [sure 00:13:20] conclusion and that's where previous studies have not been able to show a significant association with amputation. The studies, if they are single center or they are focused from electronic medical records or perspective followup, either the patients get lost to followup or go see other doctors or other healthcare systems, and that information doesn't get back to the researchers, while mortality data you can get from Social Security Death Index or other sources. I think that's what makes the study different than other studies in this similar field in terms of followup.
Dr Josh Beckman: I don't think you're giving yourself enough credit. There's a whole bunch of things that make the study unique. One of the things that I was most taken with right upfront was the way that you defined peripheral artery disease for this population. There has been, as far as I know, at least seven or eight different definitions that people have used with administrative data to try and ferret out who has PAD, and in contrast to coronary disease and stroke, it's a much more complicated endeavor to do that. So, when I saw the way that you did it ... I'm going to say this in a way that I know is going to sound funny, but you made the complicated look really simple. Your definition is not something that required 3,000 lines of ICD-9 codes within inclusion and exclusion criteria and speaks, in my opinion, to the power of the large sample because, basically, they needed one ICD-9 code and either two ABIs, a visit to a vascular surgeon or procedural code. Now, I know that this definition comes from some of your work, so can you tell us how you derive this and then let's talk about what that means.
Dr Shipra Arya: Absolutely. We looked at practice patterns for patients with vascular disease across the VA, and most patients who undergo procedures for PAD, we can confidently say that they do have PAD. When we look at the specificity of just that occurrence, it's pretty high, like [90% 00:15:23]. Then what we did was we did some random sampling in the VA data, about 300 patients, and used different codes to see if patients came back to the vascular surgeon within ... We used 14 months because it's usually one year followup that most people prescribe, so whether they went two months before or after because the appointment hours. We found that that was again a high specificity of about 80%. Then, when you look at patients who come back with ABI followup. So, we looked at CPT codes for ABI. We found out it's like a 99% specificity. If you have ABI followup within a year, and we relaxed it to 14 months, you could be 99% confident that this patient does have PAD.
We just combined all those three together, and this is ... If Circulation is interested, I can send you this, too. We are working on this manuscript where we are giving researchers different algorithms that they could use to identify PAD because I wanted a more specific sample because I was looking at PAD outcomes. I wanted the PAD definition to be tight. Our specificity is greater than 80% combining all these three together, about 84%. We are fairly confident in this that, yes, these patients truly have PAD, so when we follow them up for outcomes, we can be confident in our results. If researchers wanted a more relaxed definition of PAD, they could use other algorithms that we are putting in that paper where they could say, "We will only use one ABI measurement, or we would use a combination of these."
Dr Josh Beckman: That brings up two points. You talk about this brings up the power of large data and the ability to tone down on people who really, truly, absolutely have PAD without any question. So, number one, are you worried that you're missing people that probably do have PAD and would benefit from therapy, and number two, do you worry that you're basically concentrating on the sickest right end of the curve of the group of PAD patients?
Dr Shipra Arya: Right. That's a great point, and I discussed that with my coauthors and mentors and we wanted to be sure about our outcomes and not want to include people who did not have PAD, and then we are kind of including the effect size of what we may find, but yes, these are truly what we are calling a symptomatic PAD, and I think I mentioned that in the manuscript somewhere, that we probably would be missing people who are asymptomatic and not really being followed up. If we extended this analysis to people who are not regularly being followed or being under surveillance for their PAD, the results could be different. So, yes, it does not generalize the whole of that population. If we had gone that route and relaxed our inclusion, my worry was that we would get ... Because of large data setting up, as you say, if we include a bunch of people who are truly not PAD, we would be a [threading 00:18:17] risk in non-PAD patients.
Dr Carolyn Lam: Josh and Shipra, I loved the paper, but after this discussion I'm even more in love with the paper and impressed, so I think I just have a question for both of you. Is there any excuse not to give statins now? Do we actually think a trial is going to come on this topic? Is this the best data that we have? Is it going to enter guidelines? What do you think?
Dr Josh Beckman: I can give you my opinion first, if you want, because you're the person who actually has control of all the data. I would say this. I think it's been well known that statins should be used in all the patients with PAD for their cardiac outcome. My guess is that there are two things that are going to happen that are going to make people consider statins for limb outcomes.
One, data like this and there's never going to be a trial, a prospective randomized trial at this point, I mean unless you disagree, but there's no way people will randomize to not statin. I think the second reason is the recent data on the PCSK9 inhibitor, evolocumab, which showed that on top of statins in PAD patients, there was a further reduction in limb events. I think we're heading towards getting the LDL to zero. It may take a couple more steps, but that's basically what's going to happen.
Dr Shipra Arya: I agree. I think there has been time and time again data that shows, especially those already data supporting the mortality benefit for larger cohorts of patients with cardiovascular disease including PAD. I think this study really nails down the limb protector effects of statins, and doing a trial of this magnitude would be very difficult to do because to get that would be effect size that you have. You would need a huge cohort of patients, and you probably won't find statin-naïve patients because you have already half the patients with PAD have coronary artery disease, as well. So, not every study needs a trial. Not every question needs a trial, in my opinion. I think that's the power of large data sets. I think the evidence is overwhelming, and I would agree with Josh.
Dr Josh Beckman: I have always had a hard time explaining to people who came to see me for legs problems that they have to take a drug for their heart. It's sort of a weird two-step that people have a hard time accommodating. Do you think by telling them that this drug will also save their leg that they're going to be more likely to take the medicine by the end of the year?
Dr Shipra Arya: Yes, absolutely. That's what I tell my patients who come and see me, that this medication works on arterial plaques, and it stabilizes them. It's not just the same plaque that you have in your heart is the one you have in your leg. Maybe a little different, but to oversimplify, yes. This is not just a heart medication, and this is not just a cholesterol medication. This is a medication for your plaques, for your blockages. That's how I explain it to them, and I think the uptake would be more if we explain to them that, yes, this will help you keep your leg, stay ambulatory and stay at home and not end up in assisted living or nursing home.
Dr Josh Beckman: Carolyn, this is so much fun, especially when we get to talk to the people that do so much hard work to put stuff in circulation, so I just want to say thanks again to Shipra and her coauthors.
Dr Shipra Arya: Thank you so much, and thank you for giving us the opportunity. I think the comments from Circulation really made our paper better, so thank you for doing that.
Dr Carolyn Lam: I wish that we could just keep going on and on because I just know that Josh has even more great questions up his sleeve. See, Shipra, I told you, he's amazing. But, there you go. You're amazing, too. Your paper is amazing. Thank you so much for joining us today.