May 22, 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. Our featured discussion today centers on the challenges of cardiovascular disease risk evaluation in people living with HIV infection, an important discussion coming right up after these summaries.
The first original paper this week provides experimental evidence that nicotinamide riboside could be a useful metabolic therapy for heart failure. First author Dr. Diguet, corresponding author Dr. Mericskay, from University Paris-Sud investigated the nicotinamide adenine dinucleotide or NAD homeostasis pathways in the failing heart. They found that an expression shift occurs in both murine and human failing hearts in which the nicotinamide riboside kinase two enzyme, which uses the nucleoside nicotinamide riboside was strongly up-regulated for NAD synthesis.
Nicotinamide riboside supplemented diet administered to murine models of dilated cardiomyopathy or pressure overloaded induced heart failure restored the myocardial NAD levels and preserved cardiac function. Nicotinamide riboside increased glycolysis as well as citrate and Acetyl-CoA's metabolism in these cardiomyocytes. Thus, nicotinamide riboside supplemented diet may be helpful in patients suffering from heart failure and may help them to cope with the limited myocardial ATP supply by restoring NAD coenzyme levels and its associated signaling.
In the single ventricle reconstruction trial, one year transplant-free survival was better for the Norwood procedure with the right ventricle to pulmonary artery shunt compared with the modified Blalock‒Taussig shunt in patients with hypoplastic left heart and related syndromes. In the paper in this week's journal, authors compare transplant-free survival and other outcomes between these groups at six years. First and corresponding author Dr. Newburger from Children's Hospital Boston and her group showed that the right ventricular pulmonary artery shunt group had similar transplant-free survival at six years, but required more catheter interventions before the Fontan procedure.
Right ventricular ejection fraction, New York Heart Association class and complications did not differ by shunt time. Cumulative incidences of morbidities by six years included 20% with a thrombotic event, 15% with a seizure, and 7.5% with a stroke. These data therefore emphasize the importance of continued follow-up of the cohort, and the need to find new strategies to improve the long-term outlook for those with single ventricle anomalies.
The next paper presents results of the CREATIVE trial, which stands for Clopidogrel Response Evaluation and Anti-Platelet Intervention in High Thrombotic Risk PCI Patients). First and corresponding author Dr. Tang from Fuwai Hospital National Center for Cardiovascular Diseases, Chinese Academy of Medical Sciences and Peking Union Medical College conducted a head-to-head comparison of the safety and effectiveness of intensified anti-platelet therapies either a double dose clopidogrel or adjunctive cilostazol and conventional strategy in 1078 post-PCI patients at high thrombotic risk as identified thromboelastography, which is a platelet function test.
The primary outcome was the incidence of major adverse cardiac and cerebral vascular events at 18 months post-PCI they find as a composite of all cause death, myocardial infarction, target vessel revascularization, or stroke. The authors found that the primary end point occurred in 14.4% of those in the conventional strategy. 10.6% in those given double dose clopidogrel alone. And 8.5% in those also given adjunctive cilostazol. Now, although both intensified anti-platelet strategies achieved increased platelet inhibition, only the triple strategy with adjunctive use of cilostazol significantly reduced adverse events in the long-term follow-up.
No increased rates of major bleeding was found with the intensified anti-platelet therapy regimes. Thus, in patients with low responsiveness to clopidogrel as measured by thromboelastography, the intensified anti-platelet strategies with adjunctive use of cilostazol significantly improved the clinical outcomes without increasing the risk of major bleeding.
The final original paper sheds light on the prevalence and predictors of cholesterol screening awareness and statin treatment among American adults with familial hypercholesterolemia or other forms of severe dyslipidemia. First and corresponding author Dr. Bucholz from Boston's Children's Hospital and their colleagues used data from the National Health and Nutritional Examination Survey, and showed a high prevalence of screening and awareness above 80%. However, there were relatively low rates of statin use among individuals with familial hypercholesterolemia at 52.3%.
And even lower rates among those with severe dyslipidemia at 37.6%. The discrepancy between the prevalence of cholesterol screening and treatment was most pronounced in younger patients, uninsured patients, and patients without a usual source of healthcare. This study highlights an imperative to improve the frequency of cholesterol screening and statin prescription rates to better identify and treat this high risk population. Additional studies are needed to better understand how to close these gaps in screening and treatment.
And that brings us to the end of our summaries. Now for our feature discussion. The natural history of infection with HIV has completely changed with the use of potent antiretroviral therapies. We now know that people living with HIV actually have morbidity and mortality patterns that really resemble the general population, especially with regards to cardiovascular disease, which is very prominent in this population. And I suppose it's this that has led to the assumption perhaps that risk prediction tools and intervention strategies that we apply in the general population may be used in patients living with HIV.
Is this the case however? Well, this week's feature discussion is going to be so enlightening. And it's so important we are talking across the world here, from South Africa to the United States, and of course with me here in Singapore. I am so pleased to have the authors of this week's feature paper and they are none other than Dr. Virginia Triant from Massachusetts General Hospital, Dr. Ralph D’Agostino from Boston University. And our associate editor, Dr. Bongani Mayosi from University of Cape Town. Thank you so much for joining me for today's exciting discussion. Virginia, could I ask you to first describe your study?
Dr Virginia Triant: As you mentioned in the introduction, we have found that patients infected with HIV have an increased risk of cardiovascular disease. That includes both myocardial infarction and stroke compared to age-matched controls in the general population. And extensive data has suggested that the etiology of this increased risk is related both to traditional cardiovascular risk factors, as well as novel risk factors that are specific to HIV infection. And these include chronic inflammation in the immune activation. So consequently, it remains relatively unknown whether established cardiovascular risk prediction functions are accurate for patients with HIV because they include only risk factors that are traditional factors and they don't reflect the complete mechanism that we know is at play in cardiovascular disease associated with HIV.
So in our study, we assess the performance of three established cardiovascular risk prediction functions, two Framingham functions, and then the ACC/AHA pooled cohort's equations and we applied this to a longitudinal HIV infected cohort that was comprised of men. And we investigated the performance of the risk scores in terms of comparing regression coefficients, discrimination and calibration, which are standard metrics in cardiovascular risk prediction. So I'll briefly summarize our overall results as a start. We found that overall, the risk prediction functions underestimated risk in our group of HIV-infected men.
We found that discrimination was modest to poor, and this was indicated by low c-statistics for all of the equations. And we also found that the calibration or the agreement between observed or predicted risk was also poor across the board for all three risk prediction functions. So our results suggests that simply taking the risk prediction functions and transporting them to an HIV infected group may actually result in mis-classification in terms of patient risk. And in underestimation of cardiovascular risk.
Dr Carolyn Lam: Well, Virginia, beautifully summarized of a beautiful paper. But perhaps at this point, we should take a step back and ask ourselves how exactly were these risk prediction scores originally developed. And I can't imagine asking a better person than Ralph. Ralph, could you take us on a jaunt along history and tell us how were those Framingham risk scores developed in the first place? Who are they supposed to be applied to? And did these results surprise you?
Dr Ralph D’Agostino: After the second World War, what was becoming quite clear is things like cardiovascular disease were becoming very prominent. Things like infections and what have you, we were developing all sorts of ways of handling them with medicines and so forth. But with cardiovascular disease, it's a thing that progresses slowly over the years and it starts wiping out people. And back in those days, one out of three men between the ages of 30 and 60 had some kind of cardiovascular event. Women weren't that bad off, but they were pretty bad off also. And so what happened is the American government and the American Heart Institute set up this study in Framingham, where they took a third of the individuals between the ages of 30 and 60 and actually followed them. They took values of variables like blood pressure, cholesterol, things they thought might be useful.
And took values on them. And they had to come back every two years and after as time went on, they took the data after six years, after 10 years they took the data, and started to look at how each individual's blood pressure related to cardiovascular disease. Does cholesterol, and the answer was yes. And then I started getting involved and we were developing these cardiovascular functions where you could actually take an individual, take their measurements now, and make a prediction that had a lot of validity, good discrimination, high predictability over what was going to happen in ten incidents and then the government, the US Government, started having guidelines and what we did is we ran a study where we took a number of different studies in the US, different cardiac studies, the ARIC studies, number of 'em, and we thought applying our functions how well would they do. And it turned out that for whites in the country, the Framingham functions did very well.
But Japanese-Americans in the country, it over-predicted. Then we found out that you could make a calibration adjustment and what we've gone to, like in China, we have a big study where we had a function and Framingham function it over-predicted but calibration adjustment would make enough corrections and so now with Jeanne and the HIV, our hope was that you could take these functions and see how they work on the HIV population. When we did it we were quite well aware, because people have been looking at different things, there's something beyond the original cardiovascular risk. And what the paper shows, quite nicely, these cardiovascular risks do have some relationship but they don't explain enough. The HIV population have a much bigger burden and a simple calibration adjustment just isn't going to work. We need new variables, we need new insights on what to add to these functions.
Dr Carolyn Lam: Thank you so much for that. That's just such important part of history because I have to thank you for those equations. We apply those definitely in our Asian cohorts with that calibration factor. But I was just reflecting as you were telling that story of how we've come full circle now to actually talk about an infection again. It's the midst of an infection, like HIV infection, that we're now testing these equations once again. What better than to ask than Bongani, you're in the epicenter, if I may, of HIV infection. What do you think of the applicability of these findings to the patients you see?
Dr Bongani Mayosi: Yes. These findings are clearly of great interest to us here in the Sub-Saharan African region because it is really the epicenter HIV pandemic. We found population, in terms of risk factors for arteriosclerosis disease still remains low although there clearly derives, for example, in the incidence of myocardial infarction that's being detected in a number of the leading centers now. And with HIV we have observed cases of myocardial infarction while they tend to be younger men who almost always smoke and who get a lot more of a thrombotic episodes.
When you catch them on a thrombotic load, you do not find arteriosclerosis disease. It's going to be important, I think, as we move forward to make sure that as we develop risk functions that will predict cardiovascular disease in patient HIV that the African epidemiological context is completed teaching that HIV affects younger people, affects large numbers of women, but that, quite clearly, is associated with decreased cardiovascular event and stroke and stroke is well demonstrated. But in terms of actually looking at the risk factor this population was still in the early day and certainly in future studies would have to have a major contribution of the African cohort.
Dr Carolyn Lam: That's true, Bongani, but may I ask how would you, perhaps, advise your African colleagues now to look at these data? Then I'd also like to turn that same question over to you, Virginia. What do we do? What's the clinical take home message of these findings?
Dr Bongani Mayosi: I think the message is true that HIV infection is associated with the increased risk of cardiovascular event, there's no doubt about that. That there are some risk factors that can carry through, such as the smoking population but it's important for all clinicians to be aware of that. The ordinary risk you find in using Framingham and other established risk functions is not going to give us all the information that we need. So that recommendation should come through we need to know that risk factors are unknown, that they're important and we need to learn more about these patients in order to give us a perfect prediction of what will happen in the future.
Dr Virginia Triant: I think the findings have a lot of clinical relevance. This suggests, I think, that there are a lot of clinical implications for any patient who has novel cardiovascular risk factors that may not be accounted for in heart functions. And what our findings suggest is that if functions don't reflect the actual composition of risk factors in the population, that can result in misclassification and thus we underestimate risk, we might miss high-risk individuals, high-risk patients who would benefit from aggressive risk reduction but are not currently receiving it. This is a real clinical challenge as sit in clinic and we pull up the scores and calculate them for our patients, whether that is a trustworthy number or whether we should, perhaps, thinking that it's higher, thinking that it's different than what we're seeing for predicted 10-year risk. I think what this suggests is that the functions may need to be further tailored to different populations and sub-populations to reflect the actual composition of risk factors in that population. Even within HIV patients and populations, the risk factors in South Africa might be different than those in Boston, with different relative contributions.
One of the next stepped planned for our team is to actually look at developing, new risk functions which are tailored to HIV and incorporating both HIV itself as a risk factor, as well as HIV specific variables and to attempt to see if we can improve the performance of these functions for HIV populations. Perhaps HIV or HIV related factors might become sort of a new cardiovascular risk equivalent and we can serve patients in this population as higher cardiovascular risk baseline. I also just wanted to mention, briefly, that I think that there are important clinical implications beyond HIV that extend to other chronic inflammatory conditions. Inflammation is increasingly recognized as important in cardiovascular risk and this way HIV can serve as a prototype population. But these results are likely to extend to a lot of different populations who have chronic inflammation for different reasons.
Dr Carolyn Lam: That's a great point, Virginia. As I'm listening, I'm wondering is there no end to this because now we say HIV and then we put other inflammatory diseases, then we say, "Well, women may be different from men," and then different ethnicities may be different. I think gonna be going closer and closer to precision risk prediction, if I might say. Could I just pick your brain here? What do you think the future is? Where's the room for machine learning approaches for risk prediction, individual almost down to that level? What do you think?
Dr Ralph D’Agostino: I think you're right on target. In some sense, the functions we have there's a sort of massiveness about it, when you come to view this population, back in the 50s and 60s and so forth, cardiovascular disease was such a major ... it still is a major problem ... such a major problem you identify some of the real items like the blood pressure and cholesterol, and you attack and develop functions on that and you'd find that you're affecting positively a huge number of individuals, but now as, like Jeanne was saying, and others have been saying, you start focusing, you've got this massive group of individuals who should have their blood pressure controlled and what have you, but if you go into HIV, you go into a number of other populations and so forth, there are other things that are driving these disease and driving the manifestations of the disease. It isn't that blood pressure isn't important, it's that there's other things that are important. And so it's machine learning and so forth and deep learning that you're gonna have to be dealing with manifestations on very high levels and maybe even get into genetics.
Look in the cancer field ... I do a lot of work with the FDA ... look at the cancer field now; how it's so genetically driven in terms of a lot of the drugs the so-called biomarkers, which are basically driven by uniqueness in populations. I think that's definitely going to be, or is the future of these cardiovascular functions.
Dr Carolyn Lam: Okay audience. You heard it, right here. These are exciting times. In the meantime, thank you so much for this precious, valuable piece of work. Virginia, Bongani, Ralph, it was great having you on the show.