Thank you, everyone, for coming to today's session. It's going to be focused on checkpoint inhibitor pneumonitis. It's going to be divided into three separate sessions. I'm going to start off with the background and the risk factors. Dr. Patel will talk about the differential diagnosis and the diagnostic workup. And then Dr. Johnson will talk more about the management of checkpoint inhibitor pneumonitis. I'm Kevin Ho. I'm an assistant professor in pulmonary critical care and sleep medicine at The Ohio State University, and I have no disclosures. Our lesson objectives today will be to, one, identify patients at high risk for checkpoint inhibitor pneumonitis, two, to recognize the recommended multidisciplinary approach to diagnosis, and then, three, understand the treatment for checkpoint inhibitor pneumonitis, including up-to-date recommendations on steroid refractory disease. To start, I wanted to give a little bit of background just on immunotherapy in general. Pictured on the left is Dr. William Coley. He's often known as the father of immunotherapy. He was a sarcoma surgeon back in the late 1800s, early 1900s. And frustrated by the mortality associated with sarcoma, he had delved back into the literature to see how physicians in the past treated solid tumors and malignancies. He found a series of case reports on patients with syphilis and fever in the 1700s who had either partial or complete remission of their solid tumors. And a couple years before he started practicing, a different sarcoma surgeon had tested injecting streptococcus in a patient with sarcoma, which led to complete remission of the solid tumor. So kind of inspired by these other case reports, the use of fever in the immune system to fight cancer cells, he invented this Coley's toxins, which is pictured on the right. Initially contained streptococcus and then serration and other bacteria. The idea to induce an immune system response via infection to fight the cancer cells. So this kind of bought him the title as the father of immunotherapy. I wanted to step back a little bit and cover why the immune system needs activation in the first place. If cancer cells are foreign to the body, shouldn't the immune system be able to recognize and target these cancer cells without any additional help? And it's a complicated topic and a whole field of study in and of its own. But the cancer cells are thought to create a tumor microenvironment that makes it very difficult for the immune system to recognize and destroy the cancer cells. So immunotherapy kind of gives the immune system a little boost to try to overcome some of these barriers to better fight the cancer cells. I've divided it up into five broad categories of immunotherapy, including the checkpoint inhibitors that we'll be covering today. The first is cytokine therapy, and these are typically pro-inflammatory cytokines. Consists of high-dose interleukin-2, an interferon alpha, and then BCG. There's T-cell manipulation, and this encompasses things like CAR-T, where you can manipulate the T-cells to target certain hematologic malignancies. And then the use of tumor-infiltrating lymphocytes, or extracting the lymphocytes effective at a specific tumor and re-infusing it back into the patient. There's oncolytic viruses, which involves manipulating things like the herpes simplex virus to be oncolytic against certain malignancies like melanoma. There's the use of cancer vaccines, which is a pretty active field of study. The one I've seen most common being the prostate cancer vaccine. And then finally, the topic for today, checkpoint inhibitors. And these are proteins that are responsible for damping down the immune system. So I've included a few of the ones that are most commonly the targets for medications we have today. The PD-1, PD-L1 interaction is responsible for the apoptosis of T-cells. And then CTLA-4 is just another negative regulator of the immune system. So these checkpoint inhibitors inhibit these checkpoints, allowing the immune system to be more active and hopefully overcome some of these barriers that the cancer cells present. So a consequence of taking away some of the physiologic checkpoints are an overactive immune system. And this can lead to side effects and mostly autoimmune diseases. So when this happens in the lung, this is termed checkpoint inhibitor pneumonitis. And I won't delve too far into the clinical presentation and the incidents. Dr. Patel is going to go a little more in depth in her part of the talk. My last name is Desai. Oh, sorry. Sorry. Desai. Sorry about that. Desai. But the one thing I wanted to stress was the timing of pneumonitis. Typically it's thought to occur after a few doses or a few months. But it can occur even after a single dose and even months to years after the drug has been discontinued. So it's important to keep a high suspicion or an index of suspicion in patients who have received checkpoint inhibitors in the past. So that's going to take us to our first audience response question. Which of the following increases the risk for checkpoint inhibitor pneumonitis? Is it A, previous use of checkpoint inhibitors, B, pretreatment interstitial abnormalities, C, moderate to severe asthma, or D, baseline oxygen requirements? So, I think the best answer is probably B, pre-treatment interstitial abnormalities. The caveat being this is, none of these studies are very robust, but I'll cover each of the answer choices in subsequent slides. To start, does the type of checkpoint inhibitor impact the risk of pneumonitis down the line? This is a meta-analysis on lung cancer patients. They received either PD-1 checkpoint inhibitors and PD-L1 checkpoint inhibitors, and this group found that PD-1 checkpoint inhibitors were associated with an increased risk of all-grade pneumonitis and also more severe pneumonitis, grade 3 or higher. Going back to answer choice A, they also looked at patients who had received cancer treatments in the past and compared them to treatment-naive patients, and they found that treatment-naive patients were at a higher risk for pneumonitis, 4.3% compared to 2.8%. I think this point is somewhat controversial, as you'll see in the subsequent slides. The way they encapsulated previously treated cancer patients included chemotherapy, radiation, checkpoint inhibitors, surgery. So when they lumped them all together, it seemed like treatment-naive patients had a higher risk, but we'll talk about some of the other previous treatments that may impact pneumonitis. So if one medication can cause pneumonitis, does getting two at one time increase that risk? Typically, melanoma patients are the patients most commonly getting dual checkpoint inhibitors. So Nishino and his group looked at whether receiving two at once would increase your risk of pneumonitis, and it seemed like it did, 6.6% incidence compared to about 1.5% for monotherapy, and then also increased the incidence of severe pneumonitis, grade 3 or higher, 1.7% compared to about 0.2% in melanoma patients treated with just one agent. It's unclear how this translates to other types of cancer, but at least in melanoma patients, it appears that combination treatment does increase that risk of pneumonitis. Building upon the previous treatment, whether that remains a risk or not, so chest radiation, there was always a thought that patients that received chest radiation may have higher baseline inflammation in the lung, you're releasing a lot of tumor antigens in the lung, and then when you get these patient's checkpoint inhibitors, are they at increased risk of inflammation following chest radiation? And there are a lot of papers that have come out recently on this. I tried to highlight more of the lung cancer population. This is the Pacific trial, which looked at non-resectable stage 3 non-small cell lung cancer patients. They received either chemoradiation alone or chemoradiation followed by a checkpoint inhibitor, and they were given pretty much back to back. And in the adverse events of the study, they found that pneumonitis or radiation pneumonitis, these were grouped together, so it can be pretty difficult to kind of tease out whether it's radiation pneumonitis or checkpoint inhibitor pneumonitis, but the combination group had a pneumonitis incidence close to 34% compared to 25% with chemoradiation alone. So kind of suggesting that adding checkpoint inhibitors on top of thoracic radiation kind of in a short interval can increase your risk of pneumonitis. In a different study, this is checkpoint inhibitors in non-small cell lung cancer patients who had previously received thoracic radiation in the past. They also found that thoracic radiation seemed to increase your risk of pneumonitis and severe pneumonitis, both separating it out at grade 2 or above or grade 3 and above. I think this study is a lot more heterogeneous. The way they defined previous thoracic radiation was at any point in the past, so kind of contrasting to the Pacific trial where you received it back to back, taken together, it seems like any previous thoracic radiation can increase that risk of pneumonitis down the line. And what about the type of cancer? So if you have lung cancer or maybe pulmonary metastatic disease, does releasing the immune system and it's coming to the lung, does that increase your risk of pneumonitis? In the same paper, they looked at the combination compared to monotherapy for pneumonitis risk. They also looked at the different types of cancer and tried to tease out different types of cancer at higher risk. And lung cancer had around a 4% incidence of pneumonitis, and this was back in 2016, compared to melanoma in renal cell, which was around 1.6%. So some suggestion at the time that lung cancer patients may be a higher risk. Some of the more recent studies coming out in the last couple of years and more focused on lung cancer patients alone suggested the incidence may be as high as 10%. So again, taken together, lung cancer patients seem to be at higher risk for pneumonitis compared to patients with other types of cancer. And this is question choice B, which I thought was the best answer for a risk for pneumonitis, interstitial abnormalities and interstitial lung disease. So a lot of cancer patients, especially lung cancer patients, have some sort of chest imaging abnormality before cancer treatment. And the question that some of these papers were trying to answer was, do these imaging abnormalities increase your risk of pneumonitis down the line? As immunotherapy and checkpoint inhibitors are being brought up more, as adjuvant and new adjuvant therapy, it's important to see how risky it is for these specific patients. So on the left, some papers on interstitial lung abnormalities. The way they defined interstitial lung abnormalities was the same, but the patients they included were a little different. They included lung abnormalities as two or more lobes that were involved with some sort of interstitial lung pattern, whether that's consolidation, ground glass, central lobular nodules. And if they met those criteria, they were included. On the right are patients with full-blown interstitial lung disease, so a radiographic diagnosis of some sort of ILD. And all of these papers seem to suggest that treatment, either interstitial abnormalities or full-blown interstitial lung disease, both increase the risk of pneumonitis in the future. A caveat to a lot of these papers, a lot of them were out of Japan, and Japanese patients are known to have a higher incidence of drug-related pneumonitis, so it's unclear how this translates to a non-Japanese population. Also some of these papers included previous thoracic radiation, some of them excluded thoracic radiation. There's a lot of confounding, obviously, with thoracic radiation and causing interstitial abnormalities. But I think taken as a whole, it's important to pay attention to pretreatment imaging abnormalities before checkpoint inhibitors. And finally, I wanted to cover pulmonary function tests. There's not a whole lot of literature on this, but I think the best study to date, also coming out of Japan, these were patients with lung cancer who had pulmonary function testing done before being treated with checkpoint inhibitors. And this group found that reduced forced vital capacity, reduced forced expiratory volume within the first second, and reduced total lung capacity increased your risk of pneumonitis in the future. The same caveats with this study as the imaging ones, this was done on a majority of Japanese patients, so unclear how it would translate to a non-Japanese population. They also didn't include chest imaging, so we don't know if this restrictive appearing PFT risk factors were in patients with actually interstitial lung abnormalities or interstitial lung disease, or rather that risk was driven by extra thoracic causes. But again, some suggestion that impaired spirometry and lung capacity may increase your risk down the line. And to go to answer choice C and D, I don't think we know a whole lot about the risk of asthma and baseline oxygen requirement. But in the same paper, they looked at the diffusion capacity, and that didn't seem to be a risk for pneumonitis. So suggested that oxygen requirement itself is not necessarily a risk factor for pneumonitis in the future. To summarize my part of my talk, several immunotherapy options for cancer today with checkpoint inhibitors being increasingly used for solid tumor malignancies. Checkpoint inhibitor pneumonitis usually presents after a few doses in a few months of drug initiation, but really can occur at any time, even many months to years after the drug is stopped. And what we think are potential risk factors are pre-treatment interstitial abnormalities and interstitial lung disease, prior chest radiation, lung cancer over non-lung cancer, PD-1 over PD-L1, combination over monotherapy, impaired spirometry and reduced lung capacity in a definitely in a Japanese population, but potentially in a non-Japanese population as well. So I'm going to hand it over to Dr. Desi to talk a little bit more about the differential diagnosis and diagnostic workup. I'll just get this set up here. Good morning, everyone. I'm Alpha Desai, and I'm going to focus on the initial evaluation and differential diagnosis of our patients with checkpoint inhibitor pneumonitis. So I'm here from Stony Brook on Long Island, where I am the director of our Center for ILD and also run our outpatient practice. My disclosures are listed here, none of which are really relevant to today's talk. So I think we've gone over a little bit as far as how checkpoint inhibitors work, but it's important to review this. Most of us are not immunologists in the room here. One of the things, one of the mechanisms where immune checkpoint inhibitors function is what they do is they block the normal signaling pathways that would suppress T-cell function. That allows T-cell activation, which results in tumor destruction. And of course, that's where they've become so important in cancer therapies. Here are the available immune checkpoint inhibitors. I certainly just want to point out that there's multiple different indications for the use of these medications, and the journey really started with ipilimumab in 2011, which showed increased survival in melanoma and has expanded dramatically since that time. Of course, as much as we want to target our therapies, and we know that our immune checkpoint inhibitors work really well as far as tumor destruction, what ends up happening is that there is an immune dysregulation and a cytokine release syndrome that can result in an immune-related adverse event. Now this can affect any organ in the body. Of course, our focus for today is pneumonitis, but you have to be thinking about rashes, myocarditis, liver abnormalities, endocrine abnormalities, and even arthritis that can be seen as a side effect of these medications, and sometimes a combination of these different organ manifestations. How exactly immune checkpoint inhibitor pneumonitis occurs is not fully understood. We do know that, to a certain degree, checkpoint mechanisms work to suppress autoimmunity. So if you're then inhibiting that checkpoint mechanism, you're going to have some presence of autoimmunity, and the thought is that that's what is going to result in your checkpoint inhibitor pneumonitis, as well as some of these other complications. We know for a fact that there is increased T cell activity against cross-antigens that are expressed in the tumor tissue, which is, of course, of interest, as well as other normal tissues, which can result in some of the side effects. There are also increased levels of preexisting autoantibodies, and of course, this cytokine release syndrome, where there's an increase in inflammatory cytokines that results in the findings that we see in the lung. Now as far as the epidemiology of checkpoint inhibitor pneumonitis, much is mentioned by Dr. Ho, our data is really limited. We know that the incidence is reported at between 3 and 6 percent, with variability in the different meta-analyses and retrospective studies out there. One study reported as high an incidence of 19 percent in retrospective reviews, so it may be that we're not as good at recognizing it as we hope to be. We know that incidence is going to be higher in PD-1 inhibitors compared to the other treatment modalities, and time to pneumonitis is so variable. Anywhere from days to over a year and a half into therapy, patients can develop pneumonitis. The median is right around just shy of three months, but again, anywhere in that spectrum is a possibility. Earlier onset of pneumonitis is seen in our patients with lung cancer compared to melanoma, and particularly those who use combination immune checkpoint inhibitors. What makes this so challenging is that you can develop signs of checkpoint inhibitor pneumonitis even several months after discontinuing therapy, so vigilance, even in patients who were previously treated, is essential to really pick up on this diagnosis. And fatality, as you can imagine, is certainly not insignificant at between 10 to 17 percent in our patients, so early recognition and, of course, appropriate treatment is really the key here. There were some studies, well, there was a real-world analysis that demonstrated a fatality of as high as 35 percent. I think one of the confounders there is that that was not a clinical trial setting, and so there may have been some diagnostic discrepancies there in that population. But nevertheless, I think we need to be thinking about this because fatality is clearly something that is significant in this population. The challenges to checkpoint inhibitor pneumonitis diagnosis are numerous, and I think you're going to be left with as many questions as answers after my portion of this talk is complete. Many patients are found with surveillance imaging, which means that they may be completely asymptomatic at the time of, you know, onset of disease. Symptoms are also, you know, very nonspecific. So Zhang and colleagues looked at about 43 patients who had, who developed checkpoint inhibitor pneumonitis in the meta-analyses and essentially found that fever was seen in a quarter of patients, but just for clarification, that's 10 people. Cough was seen in about a third of patients, and dyspnea in two-thirds. Of course, when you see respiratory failure, that tends to affect a more aggressive treatment and diagnostic pathway. But in general, the symptoms really are so nonspecific. Clinical exam findings may be completely normal. In some cases, you might find evidence of crackles, but again, this is not a guarantee. And as you'll see, radiographic findings are also not pathognomonic for the disease process. So I have a question for all of you as well. I'd love your input on this. All of the following are important in the initial workup of checkpoint inhibitor pneumonitis except, A, a thorough history and review of medications, B, cultures to rule out infectious epidemiologies, C, dedicated chest tomography, and D, lung biopsy. We don't get a timer, so I'm not really sure when this is ending. I'm going to see if I can advance the slides in a second. I'll give it another couple of seconds. All right. So it looks like the majority is favoring lung biopsy as being less essential to the workup of checkpoint inhibitor pneumonitis. And we're going to talk about that coming up. So the diagnostic workup. Much like any other lung process, of course, a physical exam is essential. That doesn't always mean it's going to answer our questions, of course. If you find crackles, you may feel a little bit reassured that there's something wrong with the lungs, and it may prompt you to look into this further. Pulse oximetry to find even subtle differences from baseline can be helpful. Again, not in isolation, but supporting everything else. I want to draw your attention to the table on the right, which refers to the common terminology criteria for adverse events. What this does is, particularly for pneumonitis, it grades pneumonitis from 1 to 5, 1 being asymptomatic and the least severe disease, and 5 being death. Moderate pneumonitis really is around grade 2, just to give you an idea where there is some limitation to ADLs. To go back to our diagnostic workup, if you have somebody who has some limitation in their ADLs and has some symptoms, then that would prompt you to do additional workup to rule out other entities. And infection becomes really important. As a start, of course, we're talking about culturing various sites, including nasopharyngeal swabs to look for viral entities as well. Labs, again, are more supportive. Things like CRP and ESR might indicate some inflammation that, again, would support, if everything else follows in that vein. PFTs, as we already discussed this morning, the data is very scarce. As far as Suzuki and colleagues, the same paper presented by Dr. Ho, what they did is they did look at patients who went on to develop checkpoint inhibitor pneumonitis, and they found the group that had pneumonitis had developed, compared to their non-pneumonitis colleagues, a decrease in FEV1, FEC, and total lung capacity. But, again, their DLCO stayed intact as well. So it's hard to know, when we think of pneumonitis, how much that population can be applied to our patients in general. Imaging, we're going to review in the coming slides. Bronchoscopy and BAL, I would argue, has a fairly decent role in the diagnostic workup. And biopsy is really equivocal, and I'll show you why. So the radiologic manifestations of checkpoint inhibitor pneumonitis are many. So just going to the top left there, that's organizing pneumonia, where you can see these sort of peripheral nodular capacities. You may see air bronchograms, and you'll often see what we call a reverse halo sign or an atal sign. NSIP in the top middle there is going to give you ground glass changes. Often in the mid to lower lung fields, you may see reticular abnormalities in the periphery. And in an ideal case, as you can see with the white arrow there, some subpleural sparing. The top right represents probably the most fulminant form of pneumonitis, which is acute interstitial pneumonia, an ARDS. These patients do not present subtly, so they probably get the most aggressive workup up front. We're going to see multifocal ground glass changes, as well as some mass-like pulmonary consolidations that you can see there. The bronchiolitis pattern, again, more subtle, but that's going to reveal more central lobular and ground glass nodularity that you can see there as well. Interestingly, this patient also has a left lower lung tumor, so that's their original malignancy. Radiation recall is really when patients have inflammation at the site of prior radiation. Ideally, it's at the same site of their initial radiation treatment and is usually caused by a new inciting agent. In this case, checkpoint inhibitors can do that. And you can see what looks like almost like a linear differentiation between normal lung and the affected lung parenchyma. So what do these patterns on imaging tell us about grading and severity of disease? So this was put together by Delany, and what their group did is took essentially three of the larger groups, and again, large is a relative term in this world, and started to understand the severity grade and then the most common patterns radiographically present in checkpoint inhibitor pneumonitis. So Naidoo and colleagues had 43 patients, Nishino and colleagues had 20, and Delany and colleagues had 64 patients. So again, this represents still a very small volume in terms of patients, but I think it's very thought-provoking. On the top, you can see that, you know, as the color gets darker towards that darkest blue, that's an increase in grade and therefore severity of disease. And on the bottom are the different radiographic patterns that were seen in this population. In red is organizing pneumonia. This was the most common pattern observed, and the patients who had this generally had grade 2 toxicity, which is intermediate toxicity. So these are symptomatic patients that had some limitations in their ADL. NSIP, which is denoted by the yellow, that mustard color, was the second most common pattern, and this was often seen in patients with more low toxicity grades, like a grade 1, who were often asymptomatic and found on just routine surveillance imaging. What is interesting to me on this is that there was a large proportion of ground glass capacities in gray on the bottom left, as well as the purple that has no suggestive pattern. And I wonder if some of these ground glass capacities might otherwise fall into the NSIP group if they had been delineated further. And, of course, like all ILD, there is going to be an unclassifiable portion, and it's hard to know what to do with that component as far as where they fall into severity of disease and management. Bronchoscopy with BAL, though it's recommended for anybody with, you know, grade 2 or higher checkpoint inhibitor pneumonitis, really isn't done as often as we'd like. The data is really limited for this. So Delany's group of 64 patients only had slightly above half of their patients undergo bronchoalveolar lavage. What was seen and what's been reported in the literature is generally lymphocytosis, often more than 15% in the BAL. Flow cytometry when performed can show a decreased CD4, CD8 ratio. In some study populations, increased IL-6 was reported. And in Delany's case, they had 15 patients who had immunochemistry that was sent, and in 12 of those patients, they had PD-L1 tumor expression. So this may be something that in the future can be an option to give us some more clear and defined information in patients that we're suspecting have checkpoint inhibitor pneumonitis. Surgical lung biopsy or lung biopsy in general really isn't recommended unless we feel it's going to move forward our diagnostic and treatment approach. So if there's some kind of inconsistencies in radiology and in clinical findings, this may be indicated. In an ideal world, and we all know this, surgical lung biopsy is going to be more important for diagnostic purposes. But I would argue that transbronchial biopsies have a very clear role as well. When you're looking to rule out infection, malignancy, sarcoid-like reaction, I think these are situations where if being able to rule out other entities will allow you to move forward with a treatment plan, it's reasonable to bypass the surgical lung biopsy and have the patient undergo a transbronchial biopsy. Now, transbronchial biopsy may show us just lymphocytosis, but again, it's what it doesn't show us that may be important here. However, with surgical lung biopsy, even in the best of scenarios, we're talking about multiple different pathologic presentations. You can see organizing pneumonia, diffuse alveolar damage. Of course, those cases, again, are far sicker. Eosinophilic pneumonia in rare cases, there have been interstitial pneumonitis, nonspecific findings, much like the radiology that I presented earlier, and granulomas inflammation has been well reported in this population. What's really important in understanding about surgical lung biopsies is much like the ILD literature, we know that a certain portion of patients can have an acute exacerbation in pursuing surgical lung biopsy, so you really need to weigh the risks and the benefits to see if it's appropriate and it will further your patient's treatment plan. And I would argue that given the number of different pathologic presentations, it may or may not really clinch your diagnosis if you're otherwise confused about that. The differential diagnosis, and I'm sure there's many more, but this, I thought, was the most reasonable list to start with. We know that immune checkpoint inhibitors by activating T cells do cause some immunosuppression in individuals that are using them, so we need to start looking at all kinds of infections, be it bacterial, viral, fungal infections, PCP, and of course reactivation of tuberculosis in this population. Tumor progression, which may be seen as a mass lesion or even lymphogenic carcinomatosis, can present with some of these same symptoms. Pseudoprogression, which is an immune phenomenon seen with the immune checkpoint inhibitors, results in essentially inflammation surrounding the tumor site. What that means when you're looking at imaging is you're going to see what looks like growth of tumor or new lesions, and that can also be very challenging as far as understanding whether somebody is failing therapy or whether this is something that is an expected component of immune checkpoint inhibitor therapy, and quite frankly, it often resolves if it is truly tumor pseudoprogression. Radiation pneumonitis, we know radiation pneumonitis can develop anywhere between two and six months following radiation treatment, and ideally we expect to see pneumonitis occur in the site of the field of radiation, but that's not always the case, and so that can create some diagnostic uncertainty as well. Of course, other drug-induced pneumonitis, various chemotherapeutic agents and other antibiotics can result in pneumonitis, and we need to take these off the differential by really doing a thorough history and investigation. Pulmonary embolism, more so than actual findings on imaging, it's the symptoms that may confound your diagnosis, so patients can have a lot of the same dyspnea and chest pain that may be seen in checkpoint inhibitor pneumonitis. Pulmonary edema can give you a pattern of ground glass, which can be seen in NSIP and can be confused with pneumonitis, and neuromuscular disease, which can give you respiratory symptoms, needs to be considered and is unfortunately also a side effect potentially of immune checkpoint inhibitors. Sarcoidosis, like granulomas disease, is well established as another potential complication of, or finding rather, of immune checkpoint inhibitors, and this can present with some similar radiographic and symptomatology, and bronchiolitis airways disease also needs to be looked at in patients who both have underlying disease and may have developed it along the way. So I just want to show you this proposed decision tree for patients who have checkpoint inhibitor pneumonitis. If you have a patient who develops new symptoms while they're receiving immune checkpoint inhibitor, and again, those symptoms can be vague, you want to start with a high-resolution CAT scan. I think that's sort of our standard workup for anybody who we need really close, parenchymal information on. If you see no lesions, you can just continue doing what you're doing. If you start to see some lesions that require further investigation, you can run through your differential diagnosis and perform investigations as you feel are appropriate to rule out other entities and rule in what you believe is going on. Of course, if you are still left confused, then you have to decide whether, you know, whether further surveillance or treatments are needed, and if you have a strong suspicion, then you have to decide whether continuing treatment or other modalities, as are going to be discussed in the next talk, are appropriate here. So just to summarize, immune checkpoint inhibitors have really changed the paradigm of cancer treatment for both lung cancer and other cancers, and they work through T-cell activation, resulting in antitumor activity. However, as a side effect, checkpoint inhibitor pneumonitis and other immune-related adverse events can be seen. Checkpoint inhibitor pneumonitis can be seen in up to 19% of individuals and has a variable time to onset. Diagnosis is challenging and requires a high index of suspicion, and workup has to exclude other entities that can give you similar findings. Radiographic patterns, though variable, often include organized pneumonia and NSIP, and BAL, when performed, can show lymphocytosis as well as rule out other entities. Biopsy may be an option for certain individuals, but there really isn't a lot of data to support that at this time. And that's the end. Thank you so much for your attention. Thank you. Well, thanks for the opportunity to speak with you all. My name is Doug Johnson. I'm actually an oncologist from here at Vanderbilt. I take care of melanoma patients. So thanks for indulging and welcoming an outsider, a non-pulmonologist, to the group. So here are my disclosures. One slide, just very quickly, to give you a little bit of context on why we use these therapies. I know, obviously, I know my pulmonologists look at me like I'm crazy when we, you know, have some of these patients in their ICU suffering from pretty significant checkpoint inhibitor toxicities. And so just a quick letting you know why we use these, and they actually do help patients. So this is a five-year survival curve from patients with metastatic non-swell cell lung cancer that have high PD-L1 expression. You can see this is a 30% plus five-year overall survival. And so, you know, that's really a pretty significant improvement compared to some of our historical therapies. And I think even more, you know, remarkable for something like metastatic melanoma, who, you know, as many of you have taken care of some of these patients when they're at very end stages of their disease, you know, this used to be a universal death sentence with a five-year survival of maybe 5% in a subset of patients where you could, you know, do some sort of heroic surgical intervention. But really there was almost no chance of having these patients have long-term survival. Well, this is metastatic melanoma, which now has a over 50% melanoma-specific survival at seven and a half years. So, I mean, we're starting to get to the point where more patients are dying of other causes, you know, at least if they're responding. Obviously, there's still plenty of patients who don't respond. So I'm not presenting this as a complete panacea for our cancer patients, but many patients have incredibly remarkably durable responses and really that probably amount to a cure in many cases in solid tumors. And the other thing that's important to think about is that we're, as has been mentioned already, is that we're using these therapies in much earlier disease courses. So, you know, patients who perhaps have stage three melanoma that may have a, you know, 10 to 15% risk of the melanoma coming back after surgery. So clearly managing the toxicities from these patients is incredibly important because these patients have an incredibly high chance of being cured with or without the therapy. The therapy increases those odds. But then, you know, even if those patients were untreated, they would have a pretty high cure rate. Certainly the same goes for very early stage lung cancer and many other kinds of cancer where we're using these therapies earlier on. So very quickly, one of my interests is severe toxicities from immunotherapy, both including pneumonitis and outside of pneumonitis. And we published a paper looking at the fatal toxicities. Clearly, you know, if you've got somebody with low risk stage three melanoma and the patients are dying from toxicity, this is a real problem. Obviously it's a problem for metastatic disease too, but it certainly seems to be a special tragedy when it's somebody who has a very high cure rate. And we looked at patients who had what the different types of fatal toxicities were. We saw that they were pretty rare overall. So single-agent anti-PD-1, fatal toxicity rate of only about 1 in 300. So clearly that's pretty good compared to things like multi-agent chemotherapy, for example, and certainly pretty good compared to, you know, untreated metastatic malignancy. And that raised up to about 1 out of 80, 1.2 percent, for patients who were treated with a combination of PD-1 plus CTLA-4 inhibitor. But what I just want to really quickly draw your attention to is that pneumonitis is particularly important here because it's not only fairly common, as mentioned, you know, in the range of 5 to as high as 20 percent, but also the death rate is reasonable. And so this is probably the disease that will get the patients in the most trouble. Certainly things like myocarditis may have a higher fatality rate, but they're much more rare. So this one does have the sort of, you know, unfortunate combination of factors being common and also potentially deadly. And the other important thing to quickly note is that some of these fatal tuxes, these tend to happen really early. So it's, as was mentioned, that the risk of pneumonitis is pretty wide. You know, it can happen two years into treatment for a patient. It can happen a few days into treatment. But the really severe tuxes, these seem to happen very early. And the way, again, my sort of simple non-immunologist mind just thinks about this is if you have a pre-existing immune response, hopefully in the tumor, you see a pretty quick response in the tumor. Well, you can kind of see the same thing. When you have an already pre-existing immune response in the lung, that's potentially going to be much worse of a toxicity than perhaps a later environmental trigger that could be managed more readily. So here's the audience response question. So what's the primary treatment for checkpoint inhibitor pneumonitis? So holding checkpoint inhibitor therapy, steroid inhaler, bronchodilator, oxygen, and close monitoring, kydosteroids, or IVIG. And again, this is the primary treatment. Man, this live thing, it just like totally influences answers, I have no doubt. Whoever the first person who answers is like the tastemaker of the group. So I think both of these answers are correct. So I think there's really, if you think about treatment here, there's really three different cornerstones of checkpoint inhibitor pneumonitis treatment. So the first is, and you can kind of think about this like a fire. So this is sort of untrammeled inflammation going on in the lungs, so what do you want to do? Well, first you want to remove the source. You don't want to keep going with the checkpoint inhibitor. The problem is, these therapies are monoclonal antibodies, they have a very long half-life, even, you know, they have variable by different therapy, but they're all going to be hanging around in your system. Not only the drug itself, but then also the pharmacodynamic effects. If you think about some of these slides I just showed you that show patients responding seven years later to therapy, and they've been off therapy for five plus years at this point, clearly there's some effects that hang around from checkpoint inhibitors for quite a long time. So just stopping the drug is really insufficient to manage most of these patients. And so really, one of the other cornerstones from an active standpoint is to mitigate the inflammation. Clearly, you don't want the lungs to burn down, so to speak, so you're going to have to actually put out the fire, and that's high-dose steroids is kind of the cornerstone for treating checkpoint inhibitor pneumonitis. And then obviously, as we all do, supportive management is key as well. While the steroids are taking effect, clearly the patient needs close monitoring, oxygen, et cetera, if needed. Okay. So all cases are not the same. As was noted, there are some cases ranging from asymptomatic to all the way to fulminant respiratory failure. And so what happens if you get a referral into your clinic, and, you know, the oncologist has seen some inflammation on the chest CT, and he wants your help? Well, the guidelines, there's a couple of different guidelines. These are really sort of expert opinion-driven, so NCCN and ASCO have one, the American Society of Clinical Oncology, and the Society for Immunotherapy and Cancer. And again, these are sort of expert guidelines. These are not based on rigorous phase three clinical trial data, but this is sort of the consensus of a group of oncologists and pulmonologists who have made these guidelines. So for asymptomatic disease, we tend to hold therapy. Again, this is not something absolute. If somebody has maybe really focal inflammation, totally asymptomatic, early in their therapy you really want to keep giving therapy. You can probably continue therapy and watch closely, but most of the time if it's more diffuse or you're concerned about the patient at all, you want to hold therapy. And then probably repeat imaging pretty closely, three to four weeks, something like that. And then you can consider following them a little bit more closely clinically to make sure they don't develop symptoms. I've had plenty of patients who start off asymptomatic, you watch them a little bit, and all of a sudden they're coughing and short of breath and so forth, and you're having to manage them. And if this doesn't improve, if you get your second CT, and boy, there's still a lot of inflammation, you may want to think about treating these patients as a higher grade pneumonitis. Okay, so what about grade two? Again, Dr. Desai kind of went through the grading for these events, and really grade one we consider asymptomatic, grade two we consider mildly symptomatic. These are patients that are not hypoxic, but maybe they're symptomatic. They're coughing, they have a little shortness of breath. Maybe they have a fever, although we don't see it as noted, we don't see that as often. And then this also covers some patients that have more diffuse disease. Occasionally I'll see a patient where their CT just looks pretty terrible, but they're walking around and they say, yeah, I'm totally fine. So you may think about treating those patients a little bit more aggressively too if their CT looks bad. So again, you want to hold therapy for these patients, and then you probably want to think about steroids for these patients. And really the starting dose tends to be one to two milligrams per kilogram per day of prednisone. And then you can also think about competing diagnoses. As we've mentioned, there is a broad differential, so you can think about empiric antibiotics if you're concerned about some sort of atypical bacterial infection or a more classic lobar pneumonia or something like that. And then you can consider more diagnostic workup, which I think can be key, especially for these mildly symptomatic patients when you're thinking about potentially permanently discontinuing therapy. And so if this improves, then you want to taper the steroids over a four to six week range. There's really no science to this. Everybody's got their own steroid taper. But something in this range is reasonable. You don't want to go too fast, but obviously you want to get them off steroids at some point. And then if the patient really only had mild symptoms to begin with, you can consider re-challenging the patient. And we'll go over that more in a minute. And then if they really don't improve, then you're thinking about treating this as a higher grade of pneumonitis. Okay, what about more severe cases? Obviously, this is the time when they're more inpatient and potentially have multidisciplinary input. So this is for patients who have severe symptoms. And really, I think the line here is the patients require oxygen. That's kind of what puts them over the edge into grade three or four pneumonia. Certainly also if they have really severe radiographic findings, even if they don't require oxygen. And then, of course, grade four is life-threatening and requiring intubation or really high doses of oxygen. In most cases here, we're going to permanently discontinue checkpoint inhibitor therapy. This is something that obviously is quite life-threatening and quite severe and would be at very high risk to recur if it were tried again. Probably want some IV steroids here to get rapid action and probably a higher dose, something like two milligrams per kilogram per day. Empiric antibiotics probably because they're probably pretty sick and maybe as the diagnosis is being established and certainly consider more diagnostic workup. And certainly, these are sick patients. So sometimes the risk benefit of going to a procedure is it certainly has to be weighed. And sometimes the really sickest patients, we sometimes will do something like a pulse dose of steroids, something like a gram per day of solumedrol. The guidelines don't necessarily say this, but we certainly do this occasionally for patients with myocarditis or encephalitis or some other sort of really catastrophic checkpoint inhibitor toxicity. So I think this can be considered for your patients that are hovering on the need, needing intubation or something like that. But again, that's just my opinion, not necessarily in the guidelines. And of course, as this improves, you also do want to do the steroid taper. And if this doesn't improve, then we need some sort of additional agent. So what kind of additional agent? Well, I wish one of you guys would tell me because I think this is a really hard area. This is, I think, a really challenging question. So in general, we've kind of extrapolated our experience from other toxicities. So for example, a patient who gets checkpoint inhibitor colitis, those patients, and they don't respond to steroids, those patients actually have pretty good success rates with infliximab, something like 80% response rates. Even if it doesn't work, then something like vetilizumab, which is an integrin inhibitor, can work as well. So we have better data for things like colitis. Also, hepatitis tends to respond very well to mycophenolate, second line. But pneumonitis, I think, has been a really tough one. And I think everybody's kind of got a different second agent. So here are some of your options. I mean, mycophenolate, infliximab, IVIG, tocilizumab, cyclophosphamide. I think the guidelines just give one of those shrug emojis when they give the recommendation for second line. You know, I've talked to some of the Hopkins group, some, and some of you actually may be in the crowd, so you probably know better, you certainly know better than me. But I know that some of those papers that Dr. Desai had cited, and I know they're actually running a clinical trial of infliximab versus IVIG for patients who fail steroids with pneumonitis. But I think they had published a small series that showed that, like, something like six out of six patients that they gave infliximab ended up dying. So they've kind of moved off of that a little bit. So I guess, you know, based on that, I tend to lean towards IVIG. But I think Celcept is very reasonable, too. So those are probably my two go-tos. But you know, if you pick something else, like, you know, definitely more power to you. I think that's totally reasonable. Until we get more data. So I'm sorry not to have any better insights than that, because I think this is a really challenging issue. Of course, you know, when these patients are on high doses of steroids, especially if, you know, patients worsen and have to get more steroid and have a longer taper and so forth, you want to think about PJP prophylaxis. You know, these patients are sometimes immunosuppressed if they've had multiple previous treatments. But then also, you know, their immune system is also a bit revved up, so to speak, too. So, you know, I've actually only seen one case of PJP, you know, with all the toxicities that I've induced and all the steroids I've given. You know, so I think it's probably a bit less common than certain other scenarios. And that single patient was actually on steroids for brain mets and had CLL and so forth, so, you know, had some underlying conditions. But clearly, you still want to think about these things as well, obviously, PPIs as well, and again, secondary opportunistic infections from prolonged steroids. You don't see this so much with checkpoint inhibitors, but certainly, you know, with prolonged steroids or other immunosuppressive reasons. Okay, what about re-challenge? Let's say you've successfully brought this patient through their pneumonitis, and then, you know, the oncologist asks you, what now? Should I give more therapy? The guidelines say if it's grade three to four, so if the patient was sick enough to require oxygen, don't do it. Now, I think in reality, it's a bit more nuanced, a bit more complicated than that. There's been a couple of series that show somewhere around a 40 to 50% risk of recurrence, and that seems to go beyond pneumonitis, that seems to go to other toxicities as well, somewhere in the range of 40 to 50% recurrence. It probably has something to do with duration of time off therapy, that probably improves the risk. But I think some things to think about, you know, certainly, you could ask the oncologist, well, how much does this patient need therapy? If this patient's two years into treatment, they've had a complete response, you know, that patient may actually not need any more treatment. On the other hand, if the patient is progressing on treatment, you know, you probably don't want to re-challenge them because they're, you know, they're not benefiting from therapy. Of course, the pulmonary considerations are really important, too. You know, the sicker the patient was, the more hesitant I personally would be to re-challenge them. You know, again, there's not a ton of data for that, but, you know, presumably, if they flare, they're probably going to have worse disease than if they had, you know, very minor, mild symptoms. And then I think the third thing is really important, too, is was there diagnostic uncertainty? You know, if you were actually leaning towards, ah, this may have actually been more of a bacterial infection, the patient really needs more therapy, well, maybe that makes us, pushes us a little more towards re-challenge, whereas if it's a clear-cut diagnostic, yeah, this is obviously pneumonitis that, you know, maybe leads you away from it. So I think there's not a hard and fast right answer of whether to re-challenge or not, but I think these are some principles that could potentially help you make that decision and help your oncologist do that. So to summarize, you know, really three cornerstones of management, holding treatment, giving steroids, and supporting the patient through it. I didn't really mention this, but I should have, but basically, if patients don't improve over 48 to 72 hours on steroids, that's when you're starting to think about adding another agent or if they're on lower-dose steroids, escalating the dose of steroids. 48 to 72 hours, I think, is a pretty good rule of thumb, and high doses of steroids are the cornerstone of management and should be tapered over four to six weeks after improvement. So with that, I'll thank you for your time, and I don't know if you have time for questions or not.

checkpoint inhibitor pneumonitis

immunotherapy treatment

cancer

long-term survival rates

solid tumors

lung cancer

melanoma

high-dose steroids

supportive care