Hello, I'm Dr. Stephen Nathan. I'm the Medical Director of the Advanced Lung Disease and Lung Transplant Program at Inova Fairfax Hospital, which is in Falls Church, Virginia, really the Washington DC suburbs. I'm going to be talking to you today about pulmonary hypertension in association with interstitial lung disease, and we'll be covering the basic epidemiology diagnosis, and we'll touch on the prognosis. The objectives of this talk is to develop an understanding about the prevalence of pulmonary hypertension complicating various forms of interstitial lung disease, also develop an appreciation for the prognostic implications of PHALD, and then to gain knowledge about when to suspect pulmonary hypertension in patients with interstitial lung disease. The European Respiratory Society, in conjunction with the European Cardiology Society, just came out with some new guidelines pertaining to pulmonary hypertension towards the end of August 2022, and in that, they did touch on group 3 pulmonary hypertension. This cartoon depiction, I think, is a very nice depiction, really, of what happens in group 3 pulmonary hypertension. You can see the three main group 3 conditions at the top, COPD, interstitial lung disease, and combined pulmonary fibrosis with emphysema. Of course, we're going to be focusing on ILD alone. What they regard as an estimate of the prevalence of pulmonary hypertension is around 30 percent, non-severe PH being around 20 percent, and severe being around 5 to 10 percent. In the guidelines, they changed the definition of severe from greater than 3 to now greater than 5, based on, I would say, pretty skimpy data. In any event, what is the prevalence of PH complicating these various forms of lung disease? It's a wide variation, anywhere from as low as 1 percent in COPD to as high as 50 percent in IPF, which is the prototypical interstitial lung disease, anywhere from 8 percent to as high as 86 percent, and in sarcoidosis, which is really group 5 pulmonary hypertension, anywhere from 6 percent up to 74 percent. It really depends on when in the course of the disease you look for it. If you look for it early, the prevalence is going to be low. If you look for it later, it's going to be significantly higher. These are various estimates from multiple studies through the years that were done in IPF. A lot of the data about PHILD emanates from the field of IPF. So here you can see anywhere from 14 to as high as 84 percent, so very wide range once again. This histogram distribution of the mean pulmonary artery pressures was taken from a study done a good number of years ago. It was called the Artemis IPF study, which looked at ambrosantin for its antifibrotic properties. This was a negative study. Ambrosantin is now contraindicated in PHILD, but we did get some very good data about the prevalence and severity of PH in patients with mild to moderate IPF. From the study, you can see that about 15 percent of patients with mild to moderate disease had pulmonary hypertension. With those who did have pulmonary hypertension, the MPAP mostly sat in the 25 to 30 range. This was done at a time where we still had the old definition, two iterations of the definition ago when an MPAP of 25 or more defined pulmonary hypertension. So all these folks over here had pulmonary hypertension. This is a paper that our group put out, I have to do the math, 22 minus 6 is 16 years ago, and we showed very similar data. In this particular population, a little bit more advanced in their disease, about just under 40 percent had pulmonary hypertension, but once again, most of it was centered in the 25 to 30 range. Very few patients had more severe pulmonary hypertension. From the same paper, there were patients who had six-minute walks, 10 with no pulmonary hypertension, 24 with pulmonary hypertension. And what is notable is that those patients with pulmonary hypertension walked significantly less and desaturated more. And those are some of the first clues as to the presence of pulmonary hypertension. Patients who need more oxygen, desaturate more, and are more functionally limited. So the six-minute walk gives you that data. Now, despite most of these patients having mild pulmonary hypertension, those who do have a significantly worse outcome. As you can see here on the yellow line versus the blue line, at one year, about 40 percent of these patients had succumbed. So the combination of ILD in the context of IPF plus PH is a pretty lethal combination for many of the patients. Now, this is a concept slide. We spoke about mild to moderate 25 to 30. Where is the right cut point to be looking for it, and what is the impact of these various cut points? If we think about the most severe, almost approaching Group 1 type of pulmonary hypertension, it's around 35 to 50. 25 is our old definition. 20 is our new definition. As we increase the threshold, the number of patients gets less. And if we take this to the bottom, the question to pose here is, when we do Rheumatoid Arthritis, these are patients laying flat on their backs at rest. What happens with exercise? So you exercise these patients, and invariably, you get significant increases in their PA pressures. I'm sure all of you are going like this at the moment in unison, so we can move on to the next slide. This is a Japanese study looking at rate of change in MPAP over time. These patients had mostly mild to moderate restrictive disease, mild to moderate disease. And the rate of change that was figured out from this study was around, on average, 1.8 millimeters of mercury per year. So the progression tends to be insidious, but once these patients have it, it doesn't stop. It doesn't go away. This is from another study by our group, once again, where we looked at patients at the time of their transplant evaluation, and then at the time that they were transplanted. So we extracted this data from the OR before any incisions were made. The prevalence was 38% at baseline when they were evaluated for transplant, and this had increased to 87% at the time of transplant. So it seems like it's almost invariable that if you follow these patients long enough, if they live long enough, the vast majority will develop the vast majority will develop pulmonary hypertension. And towards the end of the disease course, it seems like in many cases, it almost goes parabolic because if you look at the time interval between our pre-transplant and post-transplant, it wasn't that great. The median was around three to six months. So right towards the end, the PA pressures can really take off. Looking at another disease, the less evil cousin to IPF is NSRP or nonspecific interstitial pneumonia. Generally, the prognosis is better in NSRP, but the prevalence of pulmonary hypertension is very similar in NSRP. And in fact, once patients with NSRP develop pulmonary hypertension, they behave as if they have IPF, their prognosis becomes significantly worse. Another disease, chronic hypersensitivity pneumonitis. Now we don't have MPAP, but we have the PVR. And you can see to the left how the PVR in this group of patients discriminates outcomes. If you compare the prognosis of PHIIP, which is really the same as PHILD, it stands for idiopathic interstitial pneumonia. So it's a large subgroup of ILD or the idiopathic interstitial pneumonias of which IPF and NSRP sit under the IIPs. But in any event, these patients have a significantly worse prognosis compared to patients with idiopathic pulmonary arterial hypertension. If you look, if you take out, if you go out to five years, let's take a look, 60% survival versus less than 20% survival. This is taken from the Comperi registry, which is a large European registry. This is a busy slide, and I'm not going to focus on the bottom. I'm going to draw your attention to the top. This was from the World Symposium in 2018, where we came out with a statement regarding group 3 pulmonary hypertension. When do you suspect? How do you support? And if you do support, when should you confirm the diagnosis? Group 3 pulmonary hypertension, same as group 1, always need a right heart cath to confirm the diagnosis. So let's get into that a little bit in more detail. This is a graphic depiction of the symptomatology of ILD and how it overlaps with pH, which makes it very difficult because the predominant symptom for both of these is shortness of breath. The shortness of breath gets worse over time, sometimes associated with cough and fatigue. But how can you discern if the shortness of breath is from progressive ILD versus the onset of pulmonary hypertension complicating the ILD? And then of course, towards the end of the disease course, you can have other symptoms of severe pulmonary hypertension, right heart failure, dizziness, syncope, unusual in the context of ILD, but certainly for those patients who do go on to severe pH, they can develop these symptoms as well. So let's go through some of the other assessments. Patients that can clue us into the presence of pulmonary hypertension, disproportionate shortness of breath. Those patients who you look at their lung function, you look at their radiographic findings, and my gosh, they're much more short of breath than you would suspect looking at those objective parameters. So disproportionate shortness of breath. Physical signs of pulmonary hypertension, a loud P2 heart sound, heard best in the second left intercostal space. That's where I always listen for that P2 heart sound. Signs of right heart failure are really quite rare in the context of interstitial lung disease, only in a small minority of patients. PFTs can provide important clues. The single breath diffusing capacity for carbon monoxide or DLCO. The lower the DL, the more likely it is that the patients have pulmonary hypertension. If you look at a DL less than around 40%, then probably at least a quarter to one third of those patients will have pulmonary hypertension. We will sometimes look at the FEC in relation to the DL, and if the FEC is more preserved than the DL, then that can be an indicator of pulmonary hypertension. There are various cut points, 1.5, 2, 2.5. The higher the ratio, in other words, the greater the maintenance of the FEC in relation to the DL, the more likely it is that they have pulmonary hypertension. Exercise. I mentioned the six minute walk. The less they walk, the more oxygen they need, the more they desaturate, the lower their pulse rate recovery, the more likely it is that they have pulmonary hypertension. Pulse rate recovery is the difference between the pulse rate at the end of the six minute walk and one minute into recovery. So if someone ends the six minute walk and their heart rate is 120, they rest, one minute later it's 110, then the pulse rate recovery would be 10. So the lower the pulse rate recovery, the worse the prognosis, the more likely it is that they have pulmonary hypertension. CT. I always look at the parenchyma, but I also always look at the pulmonary artery segment. If the PA segment is bigger than the ascending aorta, that's a good sign of pulmonary hypertension as well. We have our biomarkers, either the BMP or the NT pro BMP that can indicate pulmonary hypertension. So the trick is putting all these parameters together and then risk stratifying patients for likely pulmonary hypertension or not. And then of course, our best screening tool is echocardiography. There are multiple parameters we can look at. RBSP, evidence of RV dysfunction, right? Atrial enlargement, TAPC. These are all things that we can gain from the echo, but we know in ILD particularly, echo can be notoriously inaccurate. Here's a CT. You can see diffuse parenchymal changes, but what a lot of people sometimes neglect to look at is in the middle over here, the mediastinum. This is the ascending aorta, the descending aorta, and this is your pulmonary artery. We measure the pulmonary artery diameter, and clearly that's bigger than the aorta. So we have a ratio here that is greater than one, indicating possible underlying pulmonary hypertension. This is also from the World Symposium Task Force. I mentioned all these things in blue already. I mentioned echo, suspect support, confirm. We always confirm with a right heart catheterization. Now drilling down a little bit more on echo. Echo can both overestimate as well as underestimate what the true pressures are. This is from a study that we did in conjunction with a group at UCLA, showing the variation in echo RVSP compared to the right ventricular systolic pressure measured on right heart cath, and we called inaccurate if there was a 10 millimeter or more difference between the two. One thing also to bear in mind is that the echo will only give us the estimated systolic pressure. It doesn't give us a mean pressure, and those shouldn't be confused. I'll show a slide in a couple of minutes time. The higher the estimated RVSP, the more likely it is that the patient does have pulmonary hypertension. But once again, just to reiterate, in lung disease, echo can be inaccurate. This was a recent Delphi that was published in CHEST earlier this year. Delphi is where when you get a group of experts together, we answer a bunch of different questions where there's no existing guidelines. This is very helpful, and then we come up with recommendations based on a strict hierarchy of voting. But from this paper, we developed a very nice algorithm that is shown over here, going through many of the things that I mentioned already. Risk factors, signs, symptoms, pulmonary function tests we measured, we mentioned six-minute walk parameters, CT, biomarkers, and then in our minds, and this is the art of medicine, we develop an index of suspicion for underlying pulmonary hypertension. Either you're high suspicion, low suspicion, or no suspicion. No suspicion, no further workup necessary, and then we do an echo to suspect and support. Now, here we have our high and our low, high and low, no suspicions gone away, no further workup, and then we do an echo. And from the echo, we have a high index of suspicion just based on the echo, intermediate or low. And based on this, you can decide, okay, next step, should I do a right heart cath or not? So high, high, yes, for sure. High suspicion, intermediate, yes. High suspicion, low echo probability, consider, don't rule it out. If all these parameters point towards pulmonary hypertension, sometimes no matter what the echo shows, I'll still go ahead and get a right heart cath. Let's put it, try and put it all together now. How do we use these things and what are the performance characteristics? Very busy slide. This slide was developed for another reason, for shortness of breath, but we have some of the parameters we look at here, typically PFTs, PA segments, six minute walk. I mentioned all these things already, the DL, the ratio, what about composites, six minute walk. Let me show you some data, how we use these and how we can develop a pre-test versus post-test probability for pulmonary hypertension. So let's go. Here's the PHA auto ratio, 0.6. We're going to pick up all patients with pulmonary hypertension, but the sensitivity is very good, but the specificity is very poor. As our PHA auto ratio goes up, our sensitivity goes down, but our specificity goes up. What that means is if you have a ratio of more than 1.2 or more, you can be pretty certain that that patient has pulmonary hypertension, but it's going to be very poor in picking up all patients with pulmonary hypertension. Pulse rate recovery. These are the performance characteristics here using a break point of 13. Sensitivity for pulmonary hypertension, 52%. And if they do have a low pulse rate recovery, there's a three quarter chance that they're going to have pulmonary hypertension. This one I like, RBSP, and hopefully this resonates with you. Same concept. We have different cut points, 30, 35, 40, 45, 50, 55, 60. The lower, the greater the sensitivity, but the less specific. As you go up on the RBSP, you can see you're losing sensitivity to pick up all cases, but the specificity is going up. So you can see the value in using all of these in conjunction in terms of trying to decide if someone is low risk, intermediate risk, intermediate risk, or high risk for pulmonary hypertension. How do we put these together? This is the same data that I showed you previously. And here we have DL, resting set, exercise set, six minute walk distance. So let's work with the DL. Less than 50, good sensitivity, low specificity. As we go down on the DL, we lose sensitivity, but we gain specificity. Same with the resting set. High sensitivity, low specificity. Low sensitivity, high specificity. Exercise, walk distance. You can see how as you change thresholds, you're changing the sensitivity and specificity. But here's the value of looking at things in conjunction with one another. Let's say you, by way of example, let's pick one over here. Let's say you have a DL less than 50%. Here are the performance characteristics, sensitivity, specificity. And then you also, by the way, have an RVSP of greater than 50. Now the sensitivity changes to 36.8%, but your specificity goes up to 81%. So if you have someone whose DL is less than 50, and RVSP is greater than 50, then 81% chance that they will have pulmonary hypertension. So you can see how you can use two different parameters in combination to risk stratify patients. Sorry, this table is overwhelming and a little bit confusing. I'll just show that to you as an example. This is a paper that looked at CT measurements of the PA segment with echo. This is with echo alone. In terms of the accuracy, you can see that echo is pretty good in estimating how good the RVSP is in predicting MPAP, but it gets a little tighter. You can see the dots kind of come in on the line if you use the CT PSRs in conjunction with echo. So another example of using two parameters together. Now there have been a bunch of different attempts looking at coming up with formulas and scoring systems to risk stratify for pulmonary hypertension. This is one that was developed in a small group of patients. This was from Dave Ziesman, who was at UCLA at the time, where they came up with this regression formula of MPAP, and no one's going to remember this, but it includes saturation and FEC to DL ratio. That's since been tested and found not to be a great predictor of pulmonary hypertension. This is actually one from my own group, where we looked at the same kind of concept in 105 patients. No one's going to use this, but this is the regression equation we came up with, including PA to aorta size, FEC to DL ratio, and RVSP. So this is kind of one of the holy grails that we're still going after, how best to predict underlying pH in patients with ILD. This is a Japanese scoring system. It's a very simple scoring system that patients can get zero, one, two, three points. They get a point if the DL is less than 50, a point if the PAO2 is less than 80 millimeters of mercury, and another point if their PA to aorta ratio is greater than 0.9. So these are the performance characteristics over here. Suffice it to say, if patients have all three, then there's a two-thirds chance that they're going to have underlying pulmonary hypertension. I'm going to end off with the definition. Let's get rid of the task force three definition. This was the definition that came out of the Sixth World Symposium a couple of years ago in terms of pulmonary hypertension. What we're talking about primarily is pre-capillary pulmonary hypertension, which was the threshold was changed from 25 to greater than 20, got a rule out group two. So the wedge has got to be 15 or less and PVR three or more. So this is pre-capillary pulmonary hypertension. A lot of what I showed you previously was based on definitions of 25 or more, but now they've changed the definition again. At least the Europeans on the other side of the pond have decided to change the definition as of the end of August, 2022. They want the PVR now down to greater than two rather than greater than three. So I feel like sometimes we end up chasing our tails with these definitions because obviously that changes the prevalence and has potential implications on clinical trial design and who goes on therapy. So this is what I think we're doing. We're chasing our tails, going to get your pH, got your pH, now what? Let's play chase again. So I feel like when we change definitions, sometimes it's helpful, but it can be confusing as well. So this is another slide with moving parts. This is my second to last slide here. We have PVR anywhere from normal being less than two up to five wedge pressure. We used to have a wedge cut off of 12 to define pre-capillary pulmonary hypertension. NPAP just spoke of how that's changed. And this is what we're doing. We're changing all these different thresholds. So I wonder if people end up being a little confused. And then throw in group three. When is a group one? When is a group three? Bang, bang. How much lung disease is permissible to call a group one versus group three? So this is why us physicians also tend to somewhat be a little confused about these different issues. I think of the PA pressures, the PVR, how much lung disease as a continuum, we try to box patients into these distinct categories, but in actual fact, it's just all a continuum. So are we there yet? Are we there yet? Yes, we are. Thank you for hanging in for the duration of this talk. It's been my pleasure speaking to you and look forward to the last component as well. Have a good rest of the day, everyone. Thank you very much.

pulmonary hypertension

interstitial lung disease

prevalence

prognosis

early diagnosis

diagnostic tools

PH