May 27, 2019
Dr Carolyn Lam: Welcome to Circulation on the Run, your weekly podcast summary and backstage pass to the journal and it's editors. We're your co-hosts, I'm Dr Carolyn Lam, associate editor from The National Heart Center and Duke National University of Singapore.
Dr Greg Hundley: And I'm Greg Hundley, associate editor for Circulation and director of The Pauley Heart Center at VCU Health in Richmond, Virginia.
Dr Carolyn Lam: Guess what Greg? Right after this we have a double feature discussion. It is all about dapagliflozin with some really, really important self-analyses from the DECLARED-TIMI 58 trial and about heart failure in Type 2 Diabetes with dapagliflozin. But, all of that coming right up only after we have our chat. So Greg, what do you have for us today?
Dr Greg Hundley: My first article is going to be from Dr Mintu Turakhia at the VA Palo Alto healthcare system at Stanford University and is going to discuss the practice variation in anticoagulation prescription and outcomes after device-detected atrial fibrillation. It's a study that has insights from the VA Health Administration. This study evaluated the relationship between oral anticoagulant prescription practice variation in response to new device detected atrial fibrillation and the association to outcomes.
As you know Carolyn, there are no clearly defined thresholds of AF burden, for which to initiate oral anticoagulation.
Dr Carolyn Lam: Interesting, so what did they find, how did they do this?
Dr Greg Hundley: Carolyn, the investigators performed a retrospective cohort analysis using data from the Veterans Health Administration linked to remote monitoring data that included day level AF burden. They included patients with cardiac implantable electronic devices and remote monitoring from the years 2011 through 2014. A CHA2DS2-VASc score of greater or equal to 2, and no prior stroke or oral anticoagulant receipt in the preceding 2 years. They determined the proportion of patients prescribed oral anticoagulants within 90 days following new device-detected AFib across a range of AFib thresholds. Greater than or equal to 6 minutes, all the way up to greater than 24 hours. And they examined sight variation in oral anticoagulation prescription.
Dr Carolyn Lam: And so? What did they find?
Dr Greg Hundley: Well, you ask among 10,212 patients with defibrillators, proportion receiving oral anticoagulation varied based on device detected AF burden. For example, for those greater than or equal to 6 minutes, it was roughly 13% of individuals, for those greater than 24 hours, 27% of individuals received oral anticoagulants. Importantly, there was a substantial sight variation in oral anticoagulation prescription after device-detected atrial fibrillation, for example, greater than one hour. The median was 16%, but it ranged from as low as 3% up to highs of 67%. And so, in adjusted models, oral anticoagulant prescription after device-detected AFib of greater than 24 hours was associated with reduced stroke risk and has a ratio of 0.28, p-value's 0.02, although, the propensity adjusted model was significant when AFib lasted at least 6 minutes.
So, in conclusion, among veterans with implanted devices, device-detected atrial fibrillation is common. There is large practice variation in 90-day oral anticoagulation initiation after new device-detected AFib with low rates of treatment overall, even for episodes greater than 24 hours. Remember, we said that rate was 27%. The strongest association of oral anti-coagulation with reduction in stroke was observed after device-detected Afib of greater than 24 hours. And what this study shows, is that randomized trials are needed to perform these observational findings.
So, Carolyn, how about your next study?
Dr Carolyn Lam: Well, from anti-coagulants to anti-hypertensives. I'm going to tell you about the 6-month results if the RADIANCE Hypertension Solo Trial.
Dr Greg Hundley: Oh, so, what was the RADIANCE Hypertension Solo Trial? Can you remind us?
Dr Carolyn Lam: Glad you asked. So the trial was the one that demonstrated a greater reduction in daytime ambulatory systolic blood pressure at 2 months by endovascular ultrasound renal denervation compared with a sham procedure among patients who were not treated with anti-hypertensive medications. So the current paper, led by Michel Azizi from Université Paris-Descartes and colleagues, now report the 6-month results following the addition of a recommended, standardized, stepped-care anti-hypertensive treatment to the randomized endovascular procedure under continued blinding to the initial treatment.
Now, remember these were patients with uncontrolled combined systolic and diastolic hypertension who were initially off medications for two months following randomization. Now, between two and five months, if the monthly measured home blood pressure was more than 135/85, the stepped-care antihypertensive treatment approach was recommended and consisted of sequential addition of, for example, amlodipine 5mg a day, then a standard dose of an angiotensin-converting-enzyme inhibitor, or an ARB, and hydrochlorothiazide at 12.5mg a day, followed by sequential uptitration of the hydrochlorothiazide and amlodipine.
So, what did they find? At 6 months, 65% of the patients in the original renal denervation group were being treated by this stepped-care approach, versus 84.5 in the sham group. And the average number of antihypertension medications and defined-daily doses were all less in the renal denervation group than the sham group.
Now, despite less intensive antihypertensive treatment, the renal denervation group had reduced daytime ambulatory systolic blood pressure to a great extent than the sham group. Importantly, there were no major adverse events in either group through 6 months. The blood pressure lowering effect of endovascular ultrasound renal denervation was maintained at 6 months with less prescribed antihypertension medications compared with the sham control. And what this means is if safety is maintained in larger studies with longer follow-up, renal denervation could be a promising adjunct therapy for patients with hypertension.
Dr Greg Hundley: Wow, so we're getting back toward renal denervation? How about that?
Carolyn, my next paper jumps into the world of basic science. This is a study from Kari Alitalo at the University of Helsinki, and it involves endothelial cells and how they regulate physiological cardiomyocyte growth versus VEGFR2 mediated paracrine signaling. The study evaluates the role of bidirectional endothelial cells and cardiomyocyte cross-talk via cardiokine and angiocrine signaling as it pertains to the regulation of cardiac growth and homeostasis in pathological cardiac hypertrophy. The expansion of the cardiac vasculature to maintain adequate supply of oxygen and nutrients is a key determinant of whether the heart grows in a physiological compensated manner, or a pathological decompensated manner.
Understanding how an excess of angiogenesis induces cardiac hypertrophy and how endothelial cells regulate cardiomyocyte homeostasis, could provide novel therapeutic targets for heart failure.
Dr Carolyn Lam: Ah, this is something very close to my heart. So Greg tell us, how did they establish the link between the endothelial cells and cardiomyocytes?
Dr Greg Hundley: The investigators demonstrated that both endothelial cell deletion of vascular endothelial growth factor receptor 1 and AAV-mediated delivery of the VEGFR1's specific ligands, VEGF-B or BGIF, into the myocardium increased the coronary vasculature and induced cardiomyocyte hypertrophy in adult mice.
The resulting cardiac hypertrophy was a physiological as indicated by preserved cardiac function and exercise capacity and lack and pathological gene activation. Also, the investigators demonstrated that the reported changes were mediated by increased VEGF signaling via endothelial VEGFR2 and found that the notch and ERBb pathways are involved in transducing signals for endothelial cell cardiomyocyte cross-talk in response to angiogenesis.
So clinically, the relevance of the findings are highlighted nicely in an editorial by professor Issei Komuro at the University of Tokyo Hospital. First, he emphasizes that cross-talk between the endothelial cell VEGFR2 and cardiomyocyte ErbB signaling pathways coordinates cardiomyocyte hypertrophy with angiogenesis and contributes to physiological cardiac growth. And understanding whether factors could modify this process may impact the treatment down the road of pathologic hypertrophy.
Dr Carolyn Lam: Oh interesting! Well you know what, Greg, I've got a preclinical one for you too, and this time looking at the role of inflammation in atherosclerosis and specifically at the role of the adaptive immune response and T-cells.
So, Greg, let me remind you that when we looked at CANTOS and Canakinumab we were actually looking at the role of the innate immune response. And here is where I had planned this very nice, complicated quiz for you, Greg, about the innate versus the adaptive immune response in the various cells. Would you like to take the quiz?
Dr Greg Hundley: You know what? I think I'm going to pledge that I'm already going to get a D or an F, so why don't you enlighten us?
Dr Carolyn Lam: Now alright, remember that the CD4 T-cells are assumed to be activated by our antigens derived from modified proteins such as oxidized LDL, and these are presented via MHC class II molecules in the context of cytokine signaling, remember those? What I didn't realize is that it hadn't been assumed that atherosclerosis involves a loss of tolerance against these modified self-antigens, generated in response to hypercholesterolemia and that presentation of such antigens on these MHC class II cells, then lead to activation of proatherogenic Th1 cells. So, that was the assumptions, but this was really studied in detail by the authors, Dr Wigren from Scania University Hospital and Lund University in Sweden and their colleagues, who addressed the role of CD4 T-cells in a real novel, unconventional way. And they did this by crossing MHC class ii deficient mice with atherosclerosis-prone ApoE-deficient mice.
Now the result of these double deficient mice was almost complete void of CD4 T-cells. However, despite the lack of these T-cells and inflammation, these mice developed larger atherosclerotic lesions in the aortic root area of the heart than their ApoE-deficient counterparts. Cell transfer and blocking antibody studies also, then supported these findings and suggested that loss of regulatory T-cells is the most important cause of aggravated atherosclerosis in the double-deficient mice.
So, overall these observations demonstrate that deficiency of activation of the adaptive immune responses through MHC class ii is associated with increased development of atherosclerosis, and the findings have important implications for our understanding the possible risks and benefits of immunosuppressive therapy in patients with cardiovascular disease. Now this is discussed in a beautiful editorial by Dr Slütter and Kuiper and they are from Leiden, the Netherlands.
So, Greg, interesting stuff, huh?
Dr Greg Hundley: You bet! Let's go on and here a little bit more about diabetes.
Dr Carolyn Lam: And dapagliflozin coming right up.
Today's feature discussion is all about SGLT2 inhibitors both in heart failure and atherosclerotic disease. A huge discussion because we have two papers, and they're all coming from the DECLARE-TIMI 58 trial. I am so pleased to have the corresponding author, Dr Stephen Wiviott from the TIMI study group at Brigham Women's hospital in Boston, Massachusetts, as well as the first author of one of the papers, and that is Dr Eri Kato, who was at the TIMI study group and is now at Kyoto University, as well as the editorialist for these two papers, Dr Subodh Verma from University of Toronto, and our deputy editor, Dr Darren McGuire from UT Southwestern.
All-star cast, all-star papers. So, Steve, could you start by telling us about the DECLARE-TIMI 58 trial, just to set the background please?
Dr Stephen Wiviott: So DECLARE, for people who don't know, was a large, randomized trial of 17,000+ patients with diabetes, comparing the SGLT2 inhibitor dapagliflozin to placebo. And the patients could be enrolled if the patients had either an established cardiovascular disease, meaning secondary prevention, or simply risk factors for cardiovascular disease with primary prevention.
Patients that were treated with dapagliflozin or placebo were followed for a period of just over 4 years and there were co-primary endpoints. Those were cardiovascular death and hospitalization for heart failure, and the second co-primary endpoint was MACE, major adverse cardiovascular events, a combination of cardiovascular death, MI, or stroke.
And what we saw, initially, this was a safety trial to demonstrate the safety of this diabetes agent according to worldwide guidelines. We saw that there was certainly non-inferiority for MACE, so it was safe with regard to MACE, but we did see a statistically significant reduction in cardiovascular death and hospitalization for heart failure driven predominantly by a large reduction in hospitalization for heart failure, and we also saw consistent with the other SGLT2 inhibitors, a significant reduction in the progression of renal disease. And so we had the opportunity to follow up with a couple of important papers that were published in circulation.
Dr Carolyn Lam: Thanks, Steve. And at this point I would love to invite Eri to tell us, because we just heard that the heart failure hospitalization signal is very strong. What did you do in your analysis?
Dr Eri Kato: So previously, SGLT2 inhibitors including dapagliflozin have shown to reduce hospitalization for heart failure, now we wanted to take a step further and explore those who are at high risk. So, the aim of our study was to evaluate whether the clinical benefit of dapagliflozin is greater in patients with HFrEF, heart failure with reduced ejection fraction, compared with patients without HFrEF.
So, we used data from the DECLARE-TIMI 58, which you just heard, which included a broad spectrum of patients with Type 2 diabetes, and was also unique that it is the only SGLT trial to date that has detailed the information of these ejection fractions. So, for this study, for our study trying to find patients by the presence or absence of HFrEF, which was defined as having ejection fraction less than 45%, which is pre-specified ejection fraction couplings, and the key outcome in each was cardiovascular death or hospitalization for heart failure, its components, and of course, mortality. But we also additionally looked at MACE and renal composite endpoints.
There are several interesting findings. First, is that dapagliflozin reduced the risk of hospitalization for heart failure regardless of ejection fraction, including those with preserved ejection fraction.
Second, is we have observed lower rates of cardiovascular death in all-cause mortality with dapagliflozin in patients with HFrEF, but not in those without HFrEF.
So, in patients with HFrEF, there was a significant 45% reduction in cardiovascular death, and 41% reduction in all all-cause mortality with dapagliflozin. And I'd like to highlight that these were achieved on top of high-proportional use of conventional evidence-based heart failure therapies, and that it did not increase any adverse events.
And third, and finally, there were lower rates of renal composite endpoints with dapagliflozin regardless rejection fraction, and once again, it improves patients with preserved rejection fraction.
So, to summarize, our results showed a robust mortality benefit in patients with HFrEF, but also showed that dapagliflozin is beneficial in full spectrum patients with diabetes, regardless of ejection fraction.
Dr Carolyn Lam: Thank you Eri, that's beautifully summarized, but could I just clarify? These were patients not just with a reduced ejection fraction, but with heart failure? And how was that determined?
Dr Eri Kato: We collected data at the baseline and the heart failure was collected based on the medical record.
Dr Carolyn Lam: So, I just wanted to clarify that it wasn't just rEF, but HFrEF, but the HF part was a medical record. So, Darren, I know you thought a lot about this stuff, so what do you think? Is there still equipoise for these heart failure trials, or how do you think this adds?
Dr Darren McGuire: First, as deputy editor of Circulation, I'm thrilled that were attracting these excellent diabetes-related publications, we've had a track record of several years now of capturing many of the key analyses and certainly these two papers we're talking about today qualify among the very best we've had, and I've also had the privilege of working with these investigators on the executive committee at DECLARE. And to the investigators and the credit of the sponsor, we observed these heart failure signals in other trials as DECLARE was ongoing, and we actually made a modification during the trial to begin to collect as much as we could pre-randomization ejection fraction data. And we were able to capture on roughly one-third of the patients, pre-trial EF data and we took any way it was measured and any time of when it was measured, and there are some limitations to that, but this now represents the largest data set where we can stratify the outcomes by some measure of ejection fraction.
And I have to say I was really surprised by these results; that the cardiovascular death benefits were amplified in patients with heart failure with reduced ejection fraction, but not in those with heart failure with preserved ejection fraction, as these medications are relatively modest, diuretic agents I anticipated the opposite, honestly, that heart failure would preserve ejection fraction that is much more volume-sensitive may have incremental benefits from these medications.
So, I was surprised by this, it was a little bit upside-down from what I expected. I know Subodh and Carolyn you've both thought a lot about this as well, I'd be interested in your opinions. Did you expect that heart failure with reduced ejection fraction would drive these clinical results?
Dr Carolyn Lam: Subodh, I'm going to let you go first.
Dr Subodh Verma: First and foremost, I appreciate the opportunity the circulation gave both myself and Professor McMurray to write the editorial to these very important pre-specified analyses from DECLARE.
I actually see the results not only as interesting and tantalizing as you already discussed, but I actually see a lot of consistency between the two phenotypes, if I may, in that there is a heart failure signal or reduction in heart failure hospitalizations that appears to be consistent between the two groups, right? People with an EF of less than 45 with or without heart failure and then on the other side, people without reduced ejection fraction. They're both responsive in terms of reductions in heart failure hospitalization, so it brings into question that is this differences that we're seeing with respect to mortality, a reflection of a difference in phenotypic responsiveness to an SGLT2 inhibitor, or is this simply a reflection of increasing placebo event rates and a response based on baseline of entry in one group versus the other.
So, as has been nicely outlined by the authors, the placebo event rate for CB death and heart failure and the placebo group would have pass, if I may, was about 5 times lower than those with heart failure with reduced ejection fraction. And it might be that as we go up the pyramid of risk, whether that risk is defined based on a TIMI risk score, whether it's based on a post-MI versus stable CAV risk score, or whether it's defined based on GFR, or whether, finally, it's defined based on the event rates for CV death and heart failure, that the higher the event rate, the higher the probability of demonstrating a CV death benefit, but that old strategies are actually demonstrating a consistent benefit on the overall driver of that outcome, which in this case, is a reduction in heart failure.
So, that's what we sort of said in the editorial as well that we think that it may be a bit premature at this point to reach a conclusion that one group is responsive, and the other group is not responsive. But, as you rightfully said, Darren, it is entirely feasible through these analyses to hypothesize that one of the alternative hypotheses could be that there is a greater responsiveness in HFrEF compared to HFpEF. I actually don't understand the mechanisms of it, if that was the pieces I would have a difficult time explaining it based on the overall biology and sort of current understanding of these agents.
But, I would say let’s wait: dapa heart failure is just around the corner. That trial will enroll people with documented heart failure with reduced ejection fraction. I think 4,774 patients that are being randomized on top of base, on top of RNA, on top of MRA, etc. who still have heart failure and who have a BP that's elevated so the definitive proof for this, at least from a rough standpoint, will be forthcoming. And then there are numerous HFpEF studies that are ongoing. There's Emperor Preserve and there's Deliver, and they have characterized the HFpEF population with a little bit more granularity and clarity. And I think we will be able to then look at, specifically, is there a HFpEF group that has the same event rate for CV death and heart failure, and compare that population to a HFrEF group at the same level of risk and whether there is differences in the responsiveness to be definitive about whether this is a matter of risk, threshold, or whether this is a true representation phenotypically.
Dr Carolyn Lam: Subodh is a hard act to follow. So, I will answer your question directly, but maybe not with so many words, because it's already been said, Darren. And I'll just say I expected this benefit to be in both, I wouldn't have said one versus the other, but because we do know that in trials and with prospective studies that HFpEF outcomes are lower, especially mortality is lower, compared to HFrEF. I do wonder if it's a power issue, but the most important message--and this is coming from me also being on the steering on the committee of both DELIVER and EMPEROR Preserve--that please, this doesn't mean that we don't need the trials. I really really think that there's equipoise there still and we need to look at the DEDICATED HFpEF trials.
But, moving on from the concept of risk stratification, I would like to go on and talk about the next paper. About the DECLARE sub-study of those with a prior MI. So, Steve, could you tell us, why did you do this, and what did you find?
Dr Stephen Wiviott: I think that what we've seen as a pattern across the three SGLT2 inhibitor trials including CANVAS, EMPEROR, Outcome, and DECLARE, was that there seems to be, as Subodh has said, reductions that are relatively consistent in heart failure and renal outcomes. But there was what appears to be ischemic outcomes, the MACE outcomes, cardiovascular death, MI and stroke.
In fact, in a meta-analysis that we published at the same time as the primary paper for DECLARE, we demonstrated that there was an interaction between the primary prevention in the population, those without established cardiovascular disease, and the secondary prevention population as it relates to MACE, where the benefits for MACE seem to be in the secondary prevention population.
So, this was seen in DECLARE as well, and so we hypothesized that the population of patients who had myocardial infarction as their entering condition may be particularly at high risk for MACE and it may potentially be that that was driving the benefit. And so, what we did was we stratified the patients based on history of prior myocardial infarction versus none, turned out that there was about 3,500 patients in the trial who had had a prior myocardial infarction. As would be expected from what's known about the conditions, the event rates for those patients in the placebo arm were much higher, about 2.5 times higher than patients without myocardial infarction for MACE, also true for CV death and hospitalization for heart failure.
And then what we saw when we looked at the treatment outcomes was that there tended to be a greater reduction in MACE for patients with prior myocardial infarction. In fact, for the MACE outcome, we saw about a 16% reduction in MACE with patients with prior MI compared to reduction with patients without prior MI. And so, the combination of this higher risk, also a tendency towards a greater relative benefit lead to a much greater absolute benefit, where, in fact, we saw about a 2.5% reduction in MACE over the four-year period for patients with prior MI, compared to a 0% reduction for patients without prior MI.
And, in fact, when we broke this down to three groups: patients with prior MI, patients with atherosclerotic cardiovascular disease without prior MI, and then patients with no atherosclerotic vascular disease, essentially, we saw the same thing, which is that patients with MI were the ones who had the greatest benefit in terms of MACE. And this was almost entirely driven by reductions in myocardial infarction.
Now, I would say in contradistinction, as we look at heart failure reductions, the relative benefits for heart failure were similar among these groups, but because the risk was higher in the patients with prior MI, and of course the absolute benefits were greater in the MI population, and similar for renal outcome. So, I think that this sort of extends what we have previously known that patients with atherosclerotic cardiovascular disease were at higher risk intended to have greater benefit with the SGLT2 inhibitors on MACE events that the core of that appears to be those patients with myocardial infarction.
Dr Carolyn Lam: Thanks, Steve, that was just so clearly explained. Darren, in the last couple of minutes could I ask you to give us the take-home messages from these two studies, and maybe just, what next?
Dr Darren McGuire: I think the take-home message from these two studies in the context of the overall field of SGLT2 inhibitor data, I think the picture's becoming relatively clear and Subodh stated in eloquently before and is reviewed in the editorial is that I think across the board, and independent of how you define higher versus lower risk subsets, this class of medications in general and dapagliflozin, and these studies appear to have augmented benefit, the greater the risk. Whether that greater risk is defined by prior myocardial infarction, or heart failure with reduced ejection fraction, or decreased EGFR, these are all states where various sub studies have consistently shown across the three compounds where we have outcomes data that the treatment benefits are amplified in the higher risk patients.
And it's not just an absolute risk reduction that's augmented based on baseline risk, but there appears to be an interaction where the relative risk reduction is also amplified. And so, it's really a remarkable field and it's providing therapeutic options in these really high-risk subsets of patients where we've really been handicapped up until now with these antihyperglycemic therapies for type ii diabetes.
Dr Carolyn Lam: Thank you everybody for joining us today. This was truly a bonanza feature discussion, didn't I tell you?
You've been listening to Circulation on the Run, thank you for listening today and don't forget to tune in again next week.
This program is copyright American Heart Association 2019