Practical tools and resources for clinicians

Dr. Raymond Benza:

Welcome to the Pulmonary Arterial Hypertension Initiative podcast. This podcast is sponsored by and the presenters are being compensated by United Therapeutics. In this series of podcasts, we will talk about how PAH treatment guidelines and risk calculation drive treatment decisions in pulmonary arterial hypertension, as well as their own experience in treating our patients with PAH. My name is Dr. Raymond Benza. I'm a practicing cardiologist in the state of Ohio and currently serve as the division director for Cardiovascular Medicine at the Ohio State University. And I have the pleasure of being with one of my longtime colleagues, Dr. Vallerie McLaughlin.

Dr. Vallerie McLaughlin:

Hi Ray. It’s great to be here. I’m looking forward to this podcast. I am also a cardiologist and endowed professor at the University of Michigan, and I direct the Pulmonary Hypertension Program here.

Dr. Raymond Benza:

Thanks Val. In today's program, we will be discussing the importance of risk assessment and pulmonary arterial hypertension management. According to the PAH treatment guidelines, risk assessments should be driving our initial therapy choices as well as any changes in therapy. So let's get started. So Val, can you describe to me what risk assessment is all about in pulmonary arterial hypertension?

Dr. Vallerie McLaughlin:

Yeah. Ray. I think that we as cardiologists have really incorporated risk assessment into really many different diseases. We use different scores for NSTEMIs and STEMIs, and CHA2DS2-VASc for AFib. We live this, this is really something that is important to objectively care for patients. And over the years, we have learned about the important parameters that are prognostic in patients with pulmonary arterial hypertension. We've learned from many databases that have included thousands of patients, the parameters that assess a patient's risk or the severity of their illness and what their likelihood is of having an event over time. I think the multiparameter approach is really critical because this is a very complex disease and there are many different predictors, including how the patient is feeling, whether or not their right ventricle is failing, what their symptoms are like, syncope, how dyspneic they are, what their Functional Class is. Those are all clinical assessments of the patients. Echo has not been incorporated as much as I would like to see it.

We are really just more recently getting savvy about quantitating right ventricular function. So if you look at the registries, the Echo parameters that are included most often include right atrial area and the presence of a pericardial effusion. I think we both know how that right ventricle function is really critical in terms of the outcomes of PAH patients. And then of course, hemodynamics are important. They define the disease and also reflect the function of the right ventricle. And so when we talk about a multiparameter risk assessment, we're looking at many of those different parameters and trying to assess whether we think our patient is at low, intermediate, or high risk over the ensuing years. And we learned from each registry that no matter what the treatment is, if we can get a patient to a low-risk status, their 5-year mortality is very minimal and that should be our goal for the patients to reassess their risk and to drive them into that low-risk status. Ray, is this your approach as well?

Dr. Raymond Benza:

It is my approach. And I think you really eloquently described the available nuances that we take into consideration when we do these assessments of risk, including very importantly, the multiparameter approach. We never hang our hat on just one variable. And I also wanted to really emphasize the piece about imaging because, as we’ll talk about it a little bit later, many of the contemporary tools that we have lack sophisticated imaging parameters as part of them. And I think the tools that we have and that we use and that we will describe later, have to be really used in conjunction with some of these other newer things that are coming out that we think are important, but achieving low risk and using the multiparameter approach, I think of the 2 salient issues that I really would like our practicing clinicians to take away from this.

Dr. Vallerie McLaughlin:

Yeah. I agree. Let’s talk about how often we should be performing those risk assessments. What’s your approach to that?

Dr. Raymond Benza:

I think at least at baseline is probably one of the most important parts of risk stratification. Really, when a patient comes to your clinic for the first time, they may not have been on therapy. They’re just newly diagnosed. They’re really assessing their baseline risk; it’s very, very important. Not only for the treatments that we decide to place them on, but also for the patient’s information and their own means to plan their lives for the next year or 2. And then after that initial assessment, and I think at minimum, every 3 to 6 months, we should be evaluating our patients, but I’d like to get your opinion on a more detailed approach and maybe peaking at a patient every time we see him in our clinic. What do you think about doing risk assessments even in our routine clinic?

Dr. Vallerie McLaughlin:

Oh, Ray. I think that’s really key. I do that every single time I see a patient, I think pretty much every encounter we have with a patient, it’s an opportunity to assess their risk and consider whether or not we’ve gotten them into that low-risk status because that’s where we want them. So we can look forward and have confidence that they’re going to do well. So Ray, there are lots of ways to assess risk. Maybe we should go on and talk about some of those methods and you were really the genius behind the REVEAL risk calculators. Why don’t I talk about the ERS approach, ESC approach, and the French approach. And you talk about REVEAL, if that’s okay with you?

Dr. Raymond Benza:

That’s perfect Val.

Dr. Vallerie McLaughlin:

Yeah. The ESC and ERS guidelines published in 2015, this table, the very famous table. Green, yellow, red stoplight sort of table about a number of different prognosticators in pulmonary arterial hypertension and cut points that put them at low, intermediate, or high risk. Most of these are derived from registry data. Some of them are derived from gestalt, right? Like syncope is really a gestalt thing and an observation that we’ve made, but not really included in registries. So the important determinants include Functional Class. So we know patients who have Functional Class 4 symptoms do very, very poorly and Functional Class 1 and 2 do well, and Functional Class 3 sort of in the middle. We have discussed hall walk already, the exercise tolerance, objective exercise tolerance is something that’s very important and that can tell us about a patient’s prognosis. So greater than that sort of magic 440 number, patients do well, less than the mid-100 patients do poorly.

The Echo data and well, this is different; something I actually would like to talk to you about. The Echo data in all of these is really very limited. The presence of a pericardial effusion, which has been published a few times and the size of the right atrium—that’s really the data that we have from registries that go into these scores. And I think that’s a bit of a pity actually and personally, when I do an Echo on a patient, I go look at that right ventricle. It really is very meaningful to me and, I think, anyone who has experience in this disease. I think that it’s unfortunate that many registries don’t have a quantification of right ventricular function. Not all Echo labs quantify right ventricular function. And so really we have very limited imaging of the right ventricle in any of these risk scores, but I think it’s so, so important. And I guess sort of in my head, I incorporate that when I see patients as well. I just thought maybe you’d want to comment on that, Ray, before we move on.

Dr. Raymond Benza:

Oh no, I agree with you 100%, Val. I think all the tools that we have developed were developed in eras when imaging wasn’t as prominent as a known risk player in prognosticating. And I agree with you that as these risk scores and equations evolve, that I would love to see more imaging data in the scoring systems. And I think the important thing about that is if you really look at the spectrum on how a patient evolves or progresses with PAH, we have changes in pressure that then lead to changes in the right ventricle. And these changes can occur very early, even before many of the other symptoms or variables that are accounted for in these 3 scoring systems happen. So it even might add a level of greater sophistication and predictability for early decompensation by including these imaging parameters, entities, contemporary scoring systems that we have.

But ultimately, I think that it’s the combination of these things that are important, as you mentioned earlier, the ESC and ERS guidelines and even the French method, which use a number of variables that are within the guidelines. There are some differences between the calculators that we have developed in REVEAL, but I was curious if you might want to go into a little bit more about some of the French methodology for calculating risks before I went into what we do with REVEAL.

Dr. Vallerie McLaughlin:

Yeah, for sure. In fact, let me just make a couple more comments on ERS/ESC. So we talked about the syncope, right ventricular failure, Functional Class, hall walk or cardiopulmonary exercise testing PCO2 is listed. Biomarkers are also included, BNP less than 50, NT-proBNP less than 300. Very good prognostic indicators puts the patient in a low risk. Whereas if the BNP is greater than 300 or NT-proBNP is greater than 1400, that puts the patient in a higher risk. And then of course, hemodynamics are also important in terms of risk. And we’ve known for many years, and this point is emphasized in the risk assessments. It’s not what the pulmonary artery pressure is. It is how the right ventricle is coping with that pulmonary artery pressure. So, right atrial pressure, cardiac index, and SvO2 are the hemodynamic parameters that are most predictive. So, that’s the ERS/ESC approach.

Now the French approach, which I really like, is very simple. They took a different approach. They basically said, these are the 4 factors that are most important to us. And it has to do with being in Functional Class 1 or 2, having a hall walk over 440, having a cardiac index greater than 2.5, and a right atrial pressure less than 8. And so those are the 4 factors using the invasive French approach that they look at. And if you have 3 or 4 of those 4 factors, your prognosis is very good. Whereas if you have less than that, the prognosis is quite poor.

And then when they had a subgroup of patients that also had biomarkers and they can incorporate the BNP or NT-proBNP, when you put that into the multivariate analysis, the right atrial pressure and cardiac index falls off. And so the noninvasive French is really just Functional Class, hall walk, and biomarkers. And at least for me, that’s something I do in clinic every time I see a patient. I have all 3 of those parameters every time I see a patient. So that’s the noninvasive French approach. Ray, you want to tell us about REVEAL?

Dr. Raymond Benza:

Yeah. Thanks Val. So REVEAL calculators are in essence, very similar to the ESC/ERS guidelines and the French method in that they both use very similar variables. And that’s good because that tells us that all the contemporary risk stratification scoring and systems that we use are complimentary. And we could feel safe using them because many of the variables are shared. REVEAL takes it just a little bit further down kind of the statistical pathway, than the guideline or French method in that REVEAL methodologies are really derived from very standardized statistical modeling. And very importantly, that the variables that are used in these scoring systems are weighted against each other so that you know the relative importance of one variable versus another. For example, the Functional Class may be more or less important than an NT-proBNP level. So it kind of gives you a way to really summate the risk with a little bit more sophistication than some of the European methods.

So REVEAL Lite 2 is the derivation of an earlier model called REVEAL 1.0, and what REVEAL 2.0 did is, it took a lot of the same variables that we had found in the original REVEAL 1.0 calculator, which contains both demographic vital signs and other more objective measurements and add it to that hospitalization and the risk that is imparted by recent hospitalization for pulmonary hypertension and heart failure. It gave us a more objective means of measuring renal function, which we know comorbidities are very important part of risk stratification and renal function is one of the very important ones that have been highlighted and uses a GFR and then set of a subjective assessment of renal insufficiency. And also changes some of the scoring points and actually adds a more of a dynamic nature to the changeable risk factors like BNP and the 6-minute hall tests.

So it takes those variables and then weights them, and you can calculate a score that very nicely discriminates between low, intermediate, and high risk. Now, the difference between REVEAL 2.0 and REVEAL 2.0 Lite is just the number of variables that’s required to make the calculation. The parent score, which is the REVEAL 2.0, scores 13 variables that are associated with it. And these contain some immutable risk factors. The factors that don’t change over time, like a patient’s type of pulmonary hypertension that they have and their gender and age, even though age does change the risk points, here it is a kind of a combination of age and gender. And then it has a number of easy-to-discern variables, including vital signs like low blood pressure or high heart rate combined with the Functional Class NT-proBNP levels in which there are several cut points, and 6-minute walk distance, which also has several cut points.

And then add to that the Functional Class and Echo estimates of pericardial effusion that you mentioned earlier, and then hemodynamics, which include importantly, the right atrial pressure and the pulmonary vascular resistance. And then REVEAL Lite, takes the parent calculator and really strips it down to the easily obtainable, changeable variables that you can assimilate in each clinical encounter. Very similar to the French method in that it uses vital signs, Functional Class, the hall walk test, the NT-proBNP levels, and the vital signs. But the big difference, as we mentioned earlier, is that these factors continue to be weighted against each other, such that the discrimination index with these more statistically derived methods are a little bit higher than those that don’t incorporate weighting. Well, I hope those descriptions of the algorithm and calculators were helpful to the audience, but I’d like to ask Dr. McLaughlin, if she has any further opinions or statements that you wanted to make about the formal risks calculations that we make?

Dr. Vallerie McLaughlin:

I think that was a really elegant discussion of how REVEAL was developed. And one thing I want to highlight is the difference between 2.0 and 2.0 Lite, including the nonmodifiable risk factors. I think when we have nonmodifiable risk factors in a risk calculator, it’s an excellent way to predict prognosis. In fact, I think there’s probably nothing more accurate at actually predicting prognosis than REVEAL 2.0. But sometimes when we think of driving people to low risk, we get a little challenged because there are so many nonmodifiable risk factors in it that, sometimes our patients are going to be at very high risk, no matter what we do. And that I think is one of the things that we’ve discussed over the years about using calculators and driving patients to the low-risk status. And I think REVEAL 2.0 Lite very nicely addresses that issue and really includes the risk factors that you can treat with our medical therapies and have success at trying to drive patients into a low-risk status.

I think that’s a really important differentiation and in fact, REVEAL 2.0 Lite and French are very similar with the exception of the additional parameters of vital signs and kidney function. I think we’ve discussed 4 different tools and they’re all very good. They’re all very meaningful. If you get into low-risk status with any of those tools on therapy, no matter what the therapy is, the patient’s prognosis is good and that’s what we’re looking for—to improve the patient outcomes. And so that leads us to the next discussion point. We have this really wonderful data from all of these risk calculators that show if we can get them to low risk, they do well. So let’s talk a little bit about why it’s so important to use a formal risk calculator rather than just gestalt. Do you have any comments on that, Ray?

Dr. Raymond Benza:

Yeah. Thank you Val, for pointing that out. I think that is really important. And as we mentioned earlier in the talk, these tools are very complimentary to each other, and I think that can be used together—a more formalized, detailed risk assessment with REVEAL 2.0, perhaps at baseline and maybe at a year. And then when you’re doing the peak of patient, like you mentioned earlier, watching them at each one of their clinical visits, maybe that’s the time where you can implement the French method or REVEAL Lite. To get those on the fly risk assessments, just so that you make sure that you are actually plotting the trajectory correctly. Clinical gestalt, they think can really mislead us in certain circumstances. Now we even have a recent study that compared clinical gestalt and the formal risk calculation, and by reviewing patient’s charts from clinicians who treat patients with PAH. And both clinical gestalt and the calculated risk were aligned in less than half of the 365 charts that they examined.

I found that very surprising; 80% of the patients that are estimated to be low risk by clinical gestalt were actually reassigned to a higher risk category after formal risk calculation. That’s really important because underestimating risk is where patients can really be hurt because that’s where you wouldn’t use as an intensive mode of therapy, as opposed if they were really low risks. So that’s a really important thing is just not to miss the intermediate or high-risk patients by utilizing a less nuanced way of predicting risk. And I think there have been other studies that looked at formal risk calculation in these earlier patients functional two class patients.

And I believe a retrospective chart analysis of 153 Functional Class 2 patients who were either on mono- or dual combination therapy. And more than half of the Functional 2 patients were classified as intermediate or high risk when their risk was calculated. So again, these are patients whom if we use the single variable of assessment like Functional Class or even clinical gestalt, we would have underestimated their risk and perhaps not put them on the intensity of therapy that the patients deserved to be on.

Dr. Vallerie McLaughlin:

I think those are great points, Ray, and I think we always have to have some common sense, some gestalt in medicine, but this is an example where risk calculators are helpful. But I might point out Functional Class, as we’ve talked about, is subjective. Maybe a patient sounds Functional Class 2 because they’ve limited their activities so much that they don’t get short of breath. A really skilled historian can try to dig that out, but sometimes patients just adapt. And it also leads to the next topic of discussion is relying too much on one thing such as Functional Class. I mean, for many years before we had all of these additional tools and all of the additional data about BNP and what have you, Functional Class was the holy grail and it’s something that’s easy. It’s inexpensive, we do it every single time we see a patient. But it’s not enough, is it?

Dr. Raymond Benza:

No, it’s not. And it’s like you said, we were both weaned on Functional Class, as cardiology fellows. And it’s been ingrained into the way that we evaluate our heart failure patients, but it is, the bottom-line, subjective. It really relies on patient input, their memory, and even their honesty at some point. And so this subjective evaluation really has to be weighted and thought of in a different manner. The formal risk assessment is objective. We’re talking about multiple parameters. Some that are objectively measured, but not subjectively measured like hemodynamics or vital signs or pericardial effusion on an echo or an NT-proBNP level. So, balancing these subjective with objective factors I think is very good, but I don’t want to short come the Functional Class because even despite its subjective nature and every risk calculator and every assessment we have done, it’s always peaked its head as something that is important. It may not be as discriminatory as the others, but it certainly remains important.

Dr. Vallerie McLaughlin:

Yeah, no question. Now, some people who are critics of this say, “Oh my gosh, it takes so much time. You have a limited amount of time with a patient appointment and epic or the electronic medical record makes it all so complex. I don’t have the time to do this over the course of a visit.” And I think that’s really not a very good excuse to do something that is so important and so prognostic for a patient. A French invasive or noninvasive is really 4 or 3 variables. It really doesn’t take that much time. And Ray, I know that you’ve actually gone to a lot of trouble trying to develop apps to make the REVEAL calculation easy as well.

Dr. Raymond Benza:

Yeah. So, there are apps that are currently available. There are websites that are available that can calculate the risk for you. We've made a lot of headway in some of the EMRs in programming them to calculate risk on the fly. So I think the nature of the calculation is getting easier and easier for practitioners to perform. But like you mentioned, doing a risk calculation using REVEAL Lite 2 or the French method, that can be done in less than 30 seconds. So it really is very, very easy to do. In my practice, my nurse coordinators calculate the risk before I even walk into the room. And so it's readily available for me to talk with the patient about and discuss and make critical decisions on.

Dr. Vallerie McLaughlin:

Yeah. Ray, I think you've contributed so much to this area. It's really been a delight to have this conversation with you. I think to wrap up, we've really emphasized that formal risk calculations are important to help us monitor our patients and choose the appropriate therapy for them. So it's been such a pleasure and this concludes our first PAH Initiative podcast. Please join us next time when we will discuss initial therapy choices based on risk calculation.

If you would like more information on risk assessment, PAH treatment guidelines, or to use an online risk calculator, please visit pahinitiative.com/hcp.