DR MCCULLOUGH: There is a critical concept that needs to be in every clinical mind when seeing a patient with diabetes and it’s called a heat map. A heat map organizes patients according to their estimated glomerular filtration rate (GFR), which is calculated by creatinine, sex, age, and race on one axis and urine albumin-creatinine ratio on the other axis, which is just done from a urine sample. In that heat map, the lower the GFR and the higher the albumin-creatinine ratio the greater the risk. They actually have calculated risks not only for the progression of kidney disease but cardiovascular outcomes including HF. We’ve published and organized, as an example, the SGLT-2 inhibitor trials along one of these heat maps to really show where the patients were recruited, and the epidemiology of the heat map is exactly matching what we’re seeing in clinical trial outcomes.8 So, that is a great way to size up a patient’s risk and then to take it from there in terms of clinical suspicion, and, if patients have effort intolerance and some of the other cardinal features of HF, I think a clinician really needs to move forward with natriuretic peptide testing and echocardiography to identify patients.
DR PUNZI: Great. You mentioned that when you look at the microalbumin-creatinine ratio, this is a marker of inflammation within the blood vessel, which should be considered one of the earliest and easiest risk factors to assess. This correlates nicely with what you just said, Dr McCullough. Dr Verma, would you care to give us your perspective on those HF risk prediction tools?
DR VERMA: So, first and foremost, I think there is no perfect score in people with diabetes to say this is how we predict who’s going to develop HF. I think there have been two recent approaches. One is based on clinical features and they sort of divided this into diabetes-specific risk factors and then general risk factors, and when you’re dealing with a patient with diabetes, duration of diabetes of more than 10 years, people who are insulin-requiring diabetics, those who have any evidence of microvascular disease––again, not just nephropathy but also retinopathy and neuropathy––are good surrogates for HF, and then those who have persistently elevated hemoglobin A1c. From a general standpoint, elderly patients, people who have multiple risk factors such as hypertension and obesity, people who have a history of coronary disease, people with atrial fibrillation––those are other ways of identifying people at risk for developing HF. Also people with sleep apnea, for example.
So, we’ve written about this sort of a general guide to predicting HF in people with diabetes, but more recently there’s been the thrombolysis in myocardial infarction (TIMI) diabetes HF risk score that has been published from the DECLARE-TIMI 58 trial.9 It was developed through the SAVOR-TIMI 53 study10 and then validated in DECLARE. That has identified, I believe,5 risk factors11 and assigned a numeric risk score––a number for each one. If you have a prior history of HF, you get 2 points. That’s pretty straightforward, right? You don’t need to be a rocket scientist to say this patient is at risk for HF because he had a prior HF event, right? But, then the other ones are quite informative and that is if you have microalbuminuria, macroalbuminuria, reduced GFR, atrial fibrillation, then you get, in some cases 2 points, in some cases 1 point. This is essentially just like all of us using the CHADS2-VASc score or other types of scoring systems to determine risk and other situations. This is a numeric score that tracks for a higher risk of HF.
What was also apparent with the TIMI HF risk score was that it didn't matter whether a patient was at low, intermediate, or high risk for developing HF, the strategy that was used in the trial, which was an SGLT-2, dapagliflozin, had a similar relative risk reduction across the entire spectrum. So, I think you don’t really have to identify patients at high risk––everybody can derive benefit.
DR PUNZI: Very good.So, the issue that we are still dealing with when looking at data from the AHA, they break down the causes of the 647,457 people who die yearly from cardiovascular disease: coronary heart disease 43%, cerebrovascular accident 17%, and then HF is 9.3%. It always seems to be that HF is going to be at the lower spectrum of what is at the forefront in clinicians’ minds. My question to you is: what is it that we have to do or should do to make clinicians more aware? Dr Verma, you just mentioned the issue of the spectrum, right? So, if you treat these diabetic hypertensive individuals with an SGLT-2 inhibitor such as dapagliflozin, we now know we can reduce that risk. The question there is, should we be telling clinicians that for the diabetic hypertensive individual, they need to use dapagliflozin to decrease that risk of HF?
DR MCCULLOUGH: That’s a fair statement, as Dr Verma pointed out, that the SGLT-2 inhibitors are associated with a reduced risk for incident HF as well as HF hospitalizations and cardiovascular death in those treated. So, I think that’s a fair point that HF can clearly be prevented. I think this new genre of drugs plays a big role. I can tell you that blood pressure control as you implied from the SPRINT trial and others have shown in clinical trials, that HF is a blood pressure–
responsive end point. Far more responsive than coronary heart disease end points. There are some emerging data now that are quite interesting that HF with preserved EF appears to have a large preventable fraction based on maintenance of a normal body weight and cardiovascular fitness, both cardiorespiratory fitness and strength training.12 So, I think being lean and fit through the course of life really drops the risks for HF with preserved EF fairly dramatically.
DR VERMA: I would add to that and ask, should people with diabetes be treated with an SGLT-2 inhibitor if they have multiple risk factors? The answer to that, in my opinion is, yes, yes, and yes, because what are the other alternatives? You have to treat diabetes. We now know that there’s no threshold. If you look at the TIMI HF risk score it doesn’t matter whether you have a high or low score, the bandwidth of these agents is quite broad. If you start with a patient who’s older than 55 who has multiple risk factors, who has preserved renal function, or if you end up with a CREDENCE13-like patient who actually has significant renal disease and an GFR that’s low and is macroalbuminuric, or someone like in DAPA-HF14 who has established HF, it doesn’t matter where in that continuum you intersect with an SGLT-2 inhibitor, the relative benefits seem to be the signature of preventing HF and preventing renal disease are observed. So, I think that it is, in my opinion, the European Society of Cardiology guidelines15 that came out in 2019 suggesting that people with risk factors or those with established atherosclerotic cardiovascular disease, the use of these agents should be considered even potentially ahead of metformin. There’s been sort of two camps where some people have actually questioned the evidence base for that, and others think it’s an appropriate recommendation to be using these therapies early as disease-modifying therapies in patients with diabetes.
DR PUNZI: Wonderful. Right, I think we’re in agreement on that. So, in the hypertension realm I’ve been a lead investigator in numerous clinical trials. We mentioned the SPRINT trial. We know that when it comes to BP, lower is better. If we look at a BP reduction of only 2 mm Hg, you reduce ischemic heart disease by 7%, you reduce stroke by 10%, and you further reduce HF by 15%. As you mentioned, Dr McCullough, HF improvement is really related to that BP reduction. So, for the last comments here, can you just give me a perspective on what you both think the potential mechanism of action is, because I think we established that we’ve got to intervene early in these patients. If we’re going to treat their diabetes, now the question is metformin or SGLT-2 inhibitor, and in this case with dapagliflozin, should we utilize it first?
DR MCCULLOUGH: Dapagliflozin, like all SGLT-2 inhibitors, causes a loss of glucose and sodium in the urine and predictably lowers BP. There’s also a predictable reduction in body weight, and body weight reductions in patients with T2DM work to improve glycemic control from other agents. So, weight loss is very assistive in the T2DM condition. However, my view is that the cardiovascular benefits, the really large reductions we’ve seen in HF hospitalizations and cardiovascular death, and I’d also parenthetically say that the large reduction in the progression of chronic kidney disease with SGLT-2 inhibitors, cannot be explained just by reductions in hemoglobin A1c, BP, and body weight. So, there’s great interest in searching for additional mechanisms by which these drugs confer the benefits.16
DR PUNZI: We have been collaborating with Dr Gabby Navar from the physiology department at Tulane University performing the first-in-human experiments. In their studies in spontaneously hypertensive rats that subsequently developed diabetes they observed an upregulation of intra-renal angiotensinogen with suppressed systemic renin levels. Even though diabetic hypertensive patients are characteristically low renin, we found in their 24-hour urine collections increased angiotensinogen. This may be another potential mechanism of action of the SGLT-2 blocking intra-renal angiotensinogen with an effect similar to renin-angiotensin system blockers, which would make sense because––as you mentioned, Dr McCullough––all of this is tied to the issue of the kidney, the issue of HF, the issue of BP.
What do you think is a possible contributing factor to the HF benefit that is seen with dapagliflozin? The FDA has required manufacturers to include a black-box warning (increase cardiovascular morbidity and mortality) on all sulfonylurea drugs. As part of the drug approval process, the FDA has the pharmaceutical companies do long-term clinical outcome trials with the SGLT-2 class with these showing an improvement rather than a neutral effect on cardiovascular disease. Dr Verma, can you give us your opinion on from where you think that benefit derives?