everybody. My name is Ajay Shetty from MD Anderson Cancer Center, and it's just really wonderful to see so many people interested in our session on lung cancer in 2023, focusing on disparities. Just a quick note, the session will be a little bit different than what was on the original schedule. We'll start with Dr. Mark Lavercombe, and he's going to talk to us about a global perspective about lung cancer disparities. We'll move on to disparities in lung cancer screening with Dr. Brett Bade, and then I'll talk briefly about how we might incorporate new insights into race-neutral spirometry as we select patients for lobectomy and other thoracic surgeries for resection of lung cancer. And then finally, Dr. Miranda Tan from Stanford is going to tell us how we're going to solve all these disparities in 15 minutes. So it's a pleasure to see everybody, and with that, we'll go ahead and get started. Well, good morning. Thank you for being here. So I wanted to get my head in the right space for this talk, and I asked the Bing image creator to generate a photorealistic image of a typical lung cancer patient. And I don't know if anyone can see the unifying theme here. Okay, so my name is Mark Lavercombe. I'm from Melbourne in Australia, and I'm delighted to be asked to present today. Thank you. I have no financial disclosures relevant to the presentation, but I do want to acknowledge that I'm an able-bodied, white, heterosexual cisgender man with high education and relative income, and I live in an urban center. So my task has been to assess the current state of disparities in lung cancer. I'm going to talk about things from Australia, New Zealand, Canada, the United States, which is only a small portion of the world, obviously. I don't have a lot of data outside of those regions. I suspect it might be worse than what I'm presenting today. So what are healthcare disparities? Typical differences in the burden of disease, injury, violence, or opportunities to achieve optimal health experienced by socially disadvantaged populations. And you can consider those through a number of different spheres. People might be grouped in socially disadvantaged groups by race or ethnicity, gender, education and income, disability, geographic location, sexual orientation. With regards to cancer disparities, you could consider them in the frame of incidence, prevalence, or screening rates, the stage of cancer at diagnosis, survival and mortality, or morbidity, the quality of survivorship, and the financial burden of treatment. This is an excellent article from Carcinogenesis that sort of graphically links these different themes in a way that I think is really useful if anyone wants that reference. So I'm going to talk briefly about each of these spheres, just a little snapshot before we move to the others who are going to solve the problems. In my country in Australia, our indigenous population has terrible outcomes with lung cancer. The incidence is more than twice as high as it is for non-indigenous population. The mortality is almost twice as bad. And the five-year survival, as you can see, is significantly less than the non-indigenous population. In a different study in my country, the five-year survival was estimated at 9.4%. And if we think about our neighbours in New Zealand, just across from Australia, the indigenous Maori population in New Zealand is doing even worse with a five-year survival of only 6.5%, although that data is 10 years old. A paper in Australia looked at who are the indigenous lung cancer patients, and they're much more likely to be female, much, much more likely to be ongoing smokers, live in disadvantaged areas or very remote areas, and to die at home. And they also have much more comorbidity, particularly respiratory diseases and diabetes, which might affect their ability to undergo treatment. In the United States, African-American males have the highest age-adjusted lung cancer incidence and the highest lung cancer mortality. And African-American people generally develop lung cancer earlier and present with more advanced stage disease. Unfortunately, there is also a lower prevalence of EGFR, ALK, and ROS1 mutations, which can affect their eligibility for targeted therapies. Data from last year in the United States shows that black American and indigenous American people were significantly less likely to undergo surgical treatment for their lung cancer. And as we know, that's the most likely path to cure. When it comes to the end of life in non-small cell lung cancer, this paper looked through and worked out odds ratios for black American and Hispanic American populations when compared to a white American reference, and found very significant increase in the likelihood of emergency department attendance in the last month of life, admission to hospital, admission to intensive care, death in hospital, and the degree of aggressiveness of care when compared with white Americans. In terms of gender, women are more likely to be diagnosed at a younger age, more likely to have adenocarcinoma, a family history, and much less likely to be smokers. And worldwide, up to 50% of patients are never smokers. And given that the current lung cancer screening indications include smoking history and age, one wonders what this means for female non-smoking populations. The authors of these two papers, the same authors, actually hypothesized that lung cancer in women is a different disease. Women do survive significantly longer with lung cancer than men, with a hazard ratio for death in men of 1.43 in this paper. From Australia, although a lot of that was explained by treatment factors, and once that was adjusted for, the difference became non-significant. So how about education and income? So again, in my country, in Australia, if you're in the lowest income group and compare that to the highest income group, you have 3.6 times the chance of daily smoking, twice as likely to die of lung cancer, and at least twice the burden of disease measured by disability-adjusted life years. And despite living in a country with government-provided healthcare, there is still significant financial toxicity of having a lung cancer diagnosis in my country in terms of inability to continue working, the costs of treatment, and the costs of life around that treatment. In Canada, again, similar, much more likely to smoke if you're in a lower income bracket than in a higher income bracket, significantly higher incidence, higher likelihood of presenting later, and less likely to receive surgery for an early stage diagnosis than if you're in a higher income group. And the combination of those factors leads to somewhere between 13 and 25 percent lower survival, depending on the stage at presentation. In the United States, this is a graph plotting survival in early stage non-small cell against the eligibility cutoff for Medicaid. So the dark blue is states with a Medicaid eligibility cutoff at 50 percent or less of the federal poverty level, and the sort of aqua-coloured top one is more than 138 percent. So if you live in a state where the Medicaid eligibility cutoff is higher, you have a higher survival. In terms of disability and lung cancer, this paper from Korea demonstrated that if you had severe disability, you're more likely to have an unknown stage of lung cancer, so not staged. And that was higher, again, if the disability was a communication or cognitive disability, and that's a theme that went through all of the different metrics that were measured in this paper. If you had communication or cognitive impairment, you're much more likely to do worse. Patients with disabilities were found to have less surgical treatments, treated less with chemotherapy and radiotherapy, and a higher overall mortality, particularly in the group with severe disability, with a hazard ratio of 1.2. In geographic disparity in Australia, you're eight times more likely to die from lung cancer if you're an Indigenous Australian in a remote area than if you live in an urban centre. In Canada, more likely to be diagnosed with lung cancer if you live in a rural area compared to an urban area, and lower survival compared to urban populations. And in the United States, the hazard ratio for mortality from lung cancer is higher in rural areas than it is in urban areas. Lastly, sexual orientation. So it seems that sexual minority populations do have higher smoking rates, and they will often start their smoking earlier, which means they have a longer risk. This paper suggested that for bisexual women, there was a very significant increase in lung cancer incidence, and the same for bisexual men. But after it was adjusted for smoking rates, it became non-significant, suggesting that smoking was driving a lot of the difference. This author, who has published a lot on cancer disparities, wrote a blog post for Cancer.net two years ago, discussing barriers to lung cancer care for patients in sexual minorities with poor access to prevention, screening, and quality of care, but also reporting significant difference in how they felt about their treatment, difficulties with their relationships, substance abuse, and lower satisfaction overall. I did find a paper that looked at how we feel about our patients from sexual minorities who are having treatment for lung cancer. And 92% of the oncology providers surveyed in this paper reported that they believed that these were populations with specific and unique needs and risks. But unfortunately, only 50% of those respondents felt that they were well-informed. I just note that just two weeks ago, we had World Lung Day, and the theme for this year from the Forum of International Respiratory Societies was access to prevention and treatment for all. And the broader theme was around disparities in lung care. I've just presented a very small and brief snapshot of some of the data and how disappointing, I guess, it is. Hopefully, some of these guys will be able to shed light on how we can address it. Thank you. A bit of musical cheers we weren't expecting. Thanks, Mark. I'll add my thanks to everyone who's attending. And I was going to add to AJ's that we could already hold questions until the end. I think that'd be great. For the second part of our discussion today, we're going to talk about some of the same themes in lung cancer screening. My name's Brett Beatty. I'm an assistant professor in the section of pulmonary critical care and sleep at Northwell, at LIJ, and Lenox Hill Hospital. I do have disclosures related to being a site PI for lung cancer screening and biomarkers for lung nodules. But perhaps the bigger two disclosures are one, this topic needs more than 15 minutes to discuss, and the other is my slides are a little bit busier than Mark's. Apologies for that. Two objectives for me. One is to identify disparities impacting implementation of lung cancer screening, which I'll define in a moment. And then the second part of our group's goals are to look at strategies on how we might overcome some of these challenges. And I'll leave the most part of that to Miranda. But let's start with a question. This is my first ARS question here today, so I hope everybody's more familiar. But here's QRS, and I'll give you a QR code. I'll give you a moment to scan it. So going to the question, which of the following is not a factor impacting rates of lung cancer screening? Here are my definition of implementation when I'm using that term. Eligibility, participation, follow-up, or likelihood of benefit. And that did not work. The answers were going to be age, gender, socioeconomic status, rurality, and ethnicity. And I don't have a good way to salvage this question, other than to say the last one was all of the above. And whether through knowledge or testing psychology, the answer I'm sure all of you would have gotten was all of the above. My job is to convince you that that is the case. Here's the conceptual framework I want to use for today's talk, and it kind of exemplifies the challenges we have in screening overall and in terms of disparities. With the updated USPSTF criteria, somewhere between 18 and 14 million patients in the US are eligible for lung cancer screening. Disappointingly, to date, only somewhere between 5% to 15% of those eligible patients are participating. And further disappointing, only about one in three or less of those patients come back after a baseline scan. So those are the challenges that we see in terms of screening itself. And again, this is going to be this framework that I use for today, because all of these areas have challenges in terms of disparities adding to the challenge. For patients with lung cancer specifically, though, about half are eligible for lung cancer screening at diagnosis. Only a portion of those, somewhere about less than 5% reach their lung cancer diagnosis via a screening pathway. Getting about a 10,000-foot view of disparities for lung cancer screening, Mark mentioned all of the topics that we're going to be discussing today. As far as disparities, they impact all aspects of lung cancer screening, and I have a better slide in a moment. But in terms of contributory factors, as per my question, race and ethnicity, access to care, socioeconomic status, insurance status, and rurality are what we're going to focus today. This is probably a better figure to try and introduce that topic, in that the authors kind of described the figure as a pipeline for lung cancer and lung cancer screening. On the left side of that pipeline, you see the implementation factors that I referenced. From top to bottom, eligibility, participation, and then post-screening adherence to follow-up and outcomes. And then on the far right, probably better demonstrating than my words, that every aspect of that implementation has challenges in terms of race, socioeconomic status, where patients either aren't eligible or aren't participating. Returning to screening itself, I think we're remiss if we don't remind ourselves the basis for screening in the U.S. So the NLST, which randomized over 50,000 patients to low-dose CT versus chest X-ray, and low-dose CT group both found more lung cancers in the top right and had a lower lung cancer-associated mortality reduction, about 20%. Turning back to our topic in terms of eligibility, soon after the implementation of lung cancer screening, we recognized the challenges in terms of eligibility. This is data from the Southern Community Cohort Study that's focusing on some of the differences in eligibility by race. The cohort was over 40,000 patients who were smokers, two-thirds of which African-American in the Southern U.S. And the takeaway is the bottom left, that both in terms of overall patients in the cohort and even patients who had lung cancer in the cohort were less likely to be eligible for screening by race, specifically whereas about one-third of patients who were white were eligible for screening. That number was 17% in patients who were African-American. And in patients who were actually diagnosed with cancer, around one-third for patients who were African-American and almost 60% for patients who were white. The reasons for that are at least two-fold. One, as Mark mentioned, particularly in African-American men, diagnosis tends to be at a younger age. And the second demonstrated here on the figure on the right is pack year history. The vertical line that we're seeing demonstrates 30 pack years, the prior smoking history for lung cancer screening in USPS TF 2013. And we see that in the curve here in patients who were white, significantly higher median pack smoking history, greater than 30 by far, and then slightly lower in median patients who were African-American. The importance of that I think is demonstrated in this slide. This is from one of my mentors in Charleston, Nicole Tanner, who did a secondary analysis of NLST, showing us that if you look at overall survival in patients by race who were in the trial, there was a significantly reduced survival, increased mortality in patients who were African-American, which you see in the hazard ratio here on the left. The part that I wanted to highlight is in those patients who were randomized to screening, there was more benefit in patients who were African-American. So that puts us in a difficult position for screening programs that the patients who are potentially most likely to benefit are the least likely to be eligible. We can tell a similar story by sex. If I call our attention to the defining lung cancer screening trial in Europe, the Nelson trial, the curves you see on the right are similar. Via screening with low-dose CT, we find more cancers, and then there's a lower cancer-associated mortality. The larger bold at the bottom is potentially more benefit in women, recognizing that that study was quite underpowered for women. But the trend is similar to what I showed you in the Southern Community Cohort Study. Pasquinelli and colleagues in 2022 showed us, regardless of the defining criteria, be it USPSTF 2013 or 2021, or risk-based criteria, the women potentially have more benefit from screening, they're less likely to be eligible. So at least to slightly introduce Miranda's topic, what might we be able to do about that? To remind us some of the motivators for updating USPSTF criteria, a busy slide, apologies, but where I wanted to highlight is the left is sensitivity and specificity curves for lung cancer screening criteria for 30 PAC years. And then on the right, the Southern Community Cohort Studies estimates for how we might improve screening eligibility rates if we go from 30 to 20 PAC years. To adjust that slide I showed you before for African-American patients, they're higher risk having a lower PAC year smoking history. And the estimate was about a 20% improvement in eligibility for lung cancer screening, which is a large motivator of the lower age and PAC year smoking history for the USPSTF 2021 recommendations. Transitioning to participation. So this is probably the largest study to date of patients actually undergoing lung cancer screening in the US. Another mentor of mine, Gerard Silvestri, looking at the first million patients submitted to the American College of Radiology Registry. And already we see some signals that there might be inequality in who's undergoing screening, not just eligible for, with higher rates for eligible patients in patients who are women, slightly older, and current smokers. Our group looked at this in terms of race doing a systematic review and meta-analysis last year. And I'll call your attention to the forest plot where we see on the left side of that vertical line showing us in all of the studies patients less likely to undergo screening if they're African-American. And if the rectangles on the right, less likely if patients are white. The part that stood out to us is first of all, it's not already clear that participation in lung cancer screening is lower in patients who are African-American in the published studies. But we saw a significant attenuation in patients once they were recognized to be eligible and got a screening order, which could we potentially resolve this problem. Unfortunately, the story is not quite that simple. In a follow-up study of, cross-sectional study of U.S. veterans undergoing lung cancer screening at a single center, I'll highlight what I want to show you. Three points. It doesn't seem like having the order and recognizing eligibility for screening fixes the problem. Number one, in the study there was a surprising difficulty reaching patients in the screening program. Over 50% of patients couldn't be reached. A bit reassuring is if we were able to reach patients that very few patients declined participation. And then finally, and the part that caught my attention in our follow-up study is we still see persistent differences in obtaining the scan between patients who are black and patients who are white, which tells me there's not only challenges recognizing when patients are eligible, but also system level and individual level challenges getting the scan even after we recognize their eligibility. To kind of add on to that in terms of insurance type, I think it's pretty clear from the prior slides and what Mark mentioned that if there's, you know, insurance versus no insurance, that patients would have additional challenges in obtaining lung cancer screening. But even amongst those patients who have commercial insurance, Medicare fee-for-service, and Medicare Advantage, in administrative databases there are differences in screening acquisition both by insurance type and by race. Looking specifically at fee-for-service, patients who were non-Hispanic white versus non-Hispanic black and other races had higher rates of screening. So clearly across all aspects of acquiring the scan and recognizing eligibility, there are disparities, particularly race. Transitioning to the third part of my talk, adherence to follow-up, it's pretty crucial that we not only recognize patients' eligibility and order the scan, but ensure they come back. So Lopez-Olivio and colleagues' meta-analysis on this topic kind of highlighted the topics that I wanted to focus on today in that redding the boxes that I wanted to show you. One, factors participating in adherence to follow-up are one, current smoking, current smokers less likely to return, non-white race, and education. And again, slightly introducing Miranda's talk, one way to attenuate that is you see significantly higher odds ratios in patients who receive reminders. Returning to Gerard's follow-up study at the first million patients who underwent lung cancer screening, not only obtaining the scan but an adherence to follow-up, some of those same challenges that we see, patients who are younger age, current smoking, Hispanic ethnicity, black and Asian race, and geographic barriers in terms of the southern and western U.S., all challenges to adherence to follow-up. Again, our group looked at this in terms of a systematic review and meta-analysis focusing on race and organized the same way. In the vertical line, the rectangles that you see if they're to the left of the vertical line decrease adherence to follow-up in patients who are black and on the right side, patients who are white. Regardless of how we kind of slice the data, whether it's higher risk groups, lower risk groups, that is presence or absence of nodules or lung RAD score, decrease adherence to follow-up in patients who are black. So clearly race is one of the big challenges in all aspects of screening in single-center data and overall meta-analysis. Here's the last topic that I wanted to introduce in this single slide because it's going to take me a few words here. In terms of outcomes, what we're looking at here are time on the left, different challenges in terms of race and ethnicity, income, and I didn't have space to show education. The authors here did an analysis of SEER and the National Cancer Database and compared early-stage versus late-stage rates of cancer diagnosis by each of these potential contributors to healthcare disparities. If I call your attention to the far left, I think it'll show you trends that we're familiar in seeing, that the most frequent stage of lung cancer diagnosis, regardless of race, socioeconomic status, is stage four disease, which here is that bar in yellow with stage one disease, the bar in purple. If we flash to the right side, that is post-lung cancer screening, I'll call your attention to the red circles. Reassuringly, regardless of race or socioeconomic status, I think you can see the purple bar goes up. That is, we're recognizing more early-stage lung cancer, as Mark mentioned, potentially curative stage. The part that I find most compelling or challenging, I should say, is that we actually see a reversal of the likelihood of lung cancer early-stage. Throughout my early career, stage four disease has always been the most common, whereas in people who have the most access, the least struggle from healthcare disparities, we actually see stage one disease becoming the most common. This is the only group wherein we see it in patients who are white, patients who have a high socioeconomic status, patients who you can't see here have higher education. So I think that demonstrates the point that in terms of eligibility, adherence to follow-up participation, and here outcomes, the patients who are, in this case, not struggling from socioeconomic status, race and ethnicity barriers, and income have the most benefit. So in terms of conclusions, disparities impact all aspects of lung cancer screening, particularly in terms of age, race, ethnicity, sex, and socioeconomic status. And there are several potential mechanisms that I'll leave largely to Miranda, but to introduce updated lung cancer screening criteria, recognizing eligible patients, patient communication, and reminders. I appreciate everyone's time. Okay, we're going to shift gears a little bit, and this is the third talk focusing on spirometry and race with some new insights and some new questions. And so I'm Adesh Choudhury, I'm at MD Anderson, and these are my disclosures here, and these are not relevant to the current topic. The objectives of the talk are as follows. First, I want to present a little bit about the historical reasoning for why race correction became a part of medical practice. We'll review briefly current recommendations for the pre-surgical pulmonary evaluation of patients that are undergoing surgical resection of lung cancer. We'll appraise some new evidence for predictive equations using both a race-neutral and a race-specific normative strategy. And then finally, we'll talk a little bit about what might be the actual implications of switching to a race-neutral strategy if that is the path that we choose. So with that, I do have a very quick question, if you guys can scan the QR code, and we'll have a quick audience response question. Okay, give everybody a few more seconds. All right, let's move on to the question. So which interpretive strategy would you choose to select candidates for thoracic surgery? So A, we have a race-neutral strategy, so by this I mean the use of either multi-ethnic or global equations. Race-specific, so this is where we use specific equations for each self-identified race. Or C, either of these might be appropriate for the selection of patients who will undergo thoracic surgery. So give everyone just a few seconds to fill in their answer. All right, I love the participation, it's great. Give everyone about 10 more seconds. All right, I don't want to screw this up. I'm going to press the slide forward. Oh, perfect. Okay, we have some controversy. All right, hopefully by the end of the talk, perhaps it'll be less controversy, perhaps not. So this is a little bit of an open topic, but I'm glad to see that there's such a wide array of responses. I want to talk a little bit about how race correction and race adjustment worked its way into medical practice. The concept that the lung is important for life goes back several thousands of years, and even the Greeks understood that the lung did something very fundamental, which was necessary for life. And the concept of a vital capacity that you could actually measure probably goes back at least about 300 to 400 years to the 1600s. But at the very least, you could say that it was formalized by John Hutchinson by the time he invented this barometer in the mid-1800s. And the idea that you could measure something and that this correlated with something that had such a huge impact on survival and overall physical fitness was something that was very quickly incorporated to do some very insidious things. And so vital capacity was measured in patients or in people as a way to differentiate between different categories. For example, it was used to differentiate between social classes in England to separate between the nobility and the working class. It was used to justify racial segregation in America, to, for example, to point out that black Americans were not fit for military service. And in South Africa, and this is very recent, it was used to have different standards for who could have workman's compensation if you worked in the mines in South Africa. And so you had to have relatively more pulmonary impairment, but with this race adjustment, it's harder to show that in the indigenous populations of South Africa. So racial differences, part of the reason that this worked its way into medical practice is that these were thought to be real biological differences that were baked into genetic ancestry and that if you had a lower vital capacity, well, this is just normal for certain groups of people. And the idea that a race correction could normalize this is a pretty old concept. And it started with these very insidious roots, but then it became part of routine medical practice. And even the most recent NHANES paper had race-specific equations for Hispanic Americans, black Americans, and white Americans. And in a way, this is a step forward because this actually was a way to sample the lung function of a very diverse population of patients. But unfortunately, this concept that this race adjustment was biological is something that had been in practice now for about 200 years, and it had really become codified by this point. So this is an example of the very first spirometer. And of course, even though our spirometers are much better now, this concept of race adjustment has not necessarily gone away. But I would say in the last 10 years, we finally made some progress towards this. And it comes down to this concept that race is a social and not a biological construct. So when we think about self-identified race, the actual correlation with genetic ancestry is not perfect. It's not 100%. In fact, it's quite short of that. When we model genetic ancestry instead of race, we still don't necessarily get to the root of the problem because we know that people with certain genetic ancestries don't have the same socioeconomic factors that might lead to a lower maximally attained lung function. And there's several things that we know that impact maximally attained lung function, and this includes things like nutrition, exposure to pollution, socioeconomic status, access to healthcare, early diagnosis of lung disease. And if you really want to understand the real association of lung function with important medical outcomes, I think we really need to understand what is the role of race correction and is there a biological role? And I'm glad to say in the last two to three years, there's been some really wonderful papers that have addressed this topic. I will say that it's not necessarily a new concept and that the first paper that did this very comprehensively was by Bernie and Hooper about 10 years ago, and he showed that when you look at black Americans and white Americans and you look at the reductions in forced vital capacities, the relative impact of self-identified race on mortality is much lower when you measure the forced vital capacity without any race correction or race adjustment. More recently, my colleague Aaron Baugh showed that race-neutral normative equations predicted COPD symptoms better than race-specific equations in this baromics cohort, which is predominantly a smoking cohort with a high rate of COPD. In a contemporaneous paper from Mesa Lung, race-neutral and race-specific equations predicted lower respiratory complications more or less similarly, and so there was really no benefit or scientific advantage to using these race-specific equations. And then finally, all these papers came out in a flurry about the same time, and this is using the NHANES dataset. When you look at the impact of spirometry on mortality, and then you look at the additional impact of self-identified race, when you use a race-neutral strategy, the impact of self-identified race is much lower, and so this race adjustment actually obscures the true association of spirometry with mortality. And this led to a recent ATS guideline that is just a few months old, looking at race and at this pulmonary function test interpretation, and the ATS concludes that a race-neutral interpretation with use of either multi-ethnic or global equations is the way to go forward. And so with that, I'm gonna shift gears a little bit, and we'll talk about why spirometry is so important in selecting patients who undergo thoracic surgery. So we use spirometry to measure predictive post-operative lung function, and this is really crucial because this predictive post-operative lung function dictates who we think is at a given risk for complications after thoracic surgery. I won't go into the methods to generate these predictive post-operative equations because there are several methods to be able to do so, but regardless of which method you choose, if you calculate that the FEV1 and DLCO are more than 60% after whatever volume of lung you've decided to resect, those patients are considered to be at low risk for pulmonary complications. If either of those values are between 30 or 60%, the recommendations, which are from a paper by Brunelli and others that was published in Chest about 10 years ago, the recommendations are to use a low-tech test, such as a stair climb or a shuttle walk, to determine if patients are above a certain threshold for eligibility. If either of those are less than 30%, then you need to do a CPET, and once again, there are certain thresholds that you need to meet to be considered eligible. So what you can see is that pulmonary function is crucial to the selection of patients who undergo thoracic surgery, and this is the diagram from that paper by Brunelli and others, and it has the specific cut-offs as well, and it's really an excellent paper and the most recent guideline on how we select patients. So with that in mind, I had a very specific question, and this is work that I conducted with Dr. David Oss and Dr. Ravi Rajaram and others, and the question was looking at, is a race-neutral interpretive approach superior or inferior or similar to a race-specific approach when using spirometry to predict the risk for post-operative lung complications? So this is work that was just published a few weeks ago, and what we did is we took 3,311 patients from our institution over the last 20 years, and these are patients who went surgical lobectomy, so a relatively low volume of lung that was resected. We used GLI reference equations to generate race-specific equations for white American, black American, Hispanic American, and Asians, and then we also used the GLI, other GLI miscellaneous equation to generate a race-neutral equation for everybody, and our main question was, which of these would predict pulmonary complications more completely? So as you can see, about 30% develop pulmonary complications. The range of pulmonary complications included things like pneumonia, prolonged use of chest tube, prolonged mechanical ventilation, and so on and so forth. These are complications that are well recognized by the Society of Thoracic Surgeons. So I'm gonna highlight a few things about the population. First of all, this is a relatively older population. It's in the mid-60s. It's about half and half female and male, and what you'll note is that the vast majority are white Americans, and this, again, is probably a result of several of the disparities that have already been discussed. One thing I'll note is when we look at the race-specific and race-neutral spirometry, if you look at a black American patient, if you use a race-neutral approach, the lung function goes down by five to 7%, and so this does make a difference, although you can see that most of the people in this study had very good lung function, which is to be expected for patients who have already been selected to undergo a thoracic surgery. So 31% of these patients experienced pulmonary complication, and as you might expect, a higher pulmonary function was associated with a lower risk for complications, and it didn't really matter whether you used a race-neutral or race-specific approach. The association was quite similar. The odds ratio was 0.98 per a 1% change, so if you have a higher lung function, you can see that this can compound quite quickly and vice versa, but the interesting thing we found is that there was a further association of race with complications that was only seen when we used the race-neutral approach, so when we used the race-specific approach, this was masked, but there was actually a further protective effect for non-white patients, and their odds ratio for developing complications was about 30% lower, so perhaps different than what you might expect given what we've heard so far. These are the area under the receiver-operator curves, and the only thing I'll really say about this is that they're really broadly very similar. Neither is really excellent at predicting complications, but there's really no advantage to a race-specific approach, and there are some limitations, including that this is just a single-center study with few non-white patients, and I would say the biggest thing is that once you've selected somebody to undergo thoracic surgery, those are fundamentally different than the patients who are not eligible for thoracic surgery as well, too, and this is a difficult problem to address, so with that, is that the end of the story? So race-neutral interpretive strategies are similar to a race-specific strategy when selecting patients for thoracic surgery. Is there anything else we need to consider? So one thing I will note is that with this race-neutral strategy, the lung function of black Americans is gonna be lower. It's gonna come out as a lower value, and especially for those patients who are on a borderline, this may really impact their chance of getting thoracic surgery, getting a lung resection for a potentially curable lung cancer, and if you can't get the best available procedure, your mortality is gonna be substantially higher. So without really reexamining the existing thresholds, we might be worsening disparities by using a race-neutral approach. You know, alternatively, perhaps, some of these patients could be managed in a non-surgical fashion, but at the end of the day, when you have a resectable lung cancer, surgery is the gold standard, and so this is something we really need to think about, and as we think about this, we also need to recognize that we haven't looked at these thresholds in several years, and this might be a good time to take a race-neutral approach and really critically reexamine the thresholds, particularly because surgical techniques are getting better. We don't do as many open-thoracotomies. We've switched to vats and, most recently, robotic techniques, and the surgical techniques get better at a very rapid rate, and so these old thresholds that were used to designate risk may not be valid anymore. And then finally, by reducing some of the social disparities that result in these changes in maximally attained lung function, hopefully over time, this will become less and less of an issue, and so this is a wider goal that requires strategies that are even beyond the topic of the current session. So with this, I have a few conclusions. Hopefully I've convinced you that a race-neutral interpretive strategy is the way to go forward, and what I've shown you, hopefully, is that this strategy is really quite similar to a race-specific strategy, but we do gain additional insights about what leads to complications when we measure lung function as is without applying any further corrections. When implementing a race-neutral strategy, we really need to think carefully about how we can make sure that we are not worsening existing disparities in access to lung cancer care with thoracic surgery. And then finally, the reduction of these disparities should be a common goal amongst the entire medical community. So I'd like to thank my fellow speakers here, as well as the folks who helped us do the work that I presented, and to the Chest Planning Committee, and thanks to everybody for your attention, and this is my email. I am looking forward to comments. So. Thank you. Black smokers, as we all know, are less likely to be eligible due to their infrequency, follow-up, shorter smoking history, and longer time since quitting. The lower SES positioning also reduces access to LCS. And as we mentioned earlier, the post-screening behavior and adherence is also a problem where black men are less likely to undergo resection, and in fact, there are also higher loss of follow-up among these black participants. In fact, black smokers have the highest lung cancer mortality of all racial and ethnic groups in the U.S., and the worst survival. You know, this is likely a multifactorial problem, right? So there's many issues that contribute to this, like, multidimensional problem. You know, we think about SES, we think about insurance status, belief systems, the physician-patient ratio, as well as education and tobacco usage. But somewhat encouragingly, while survival rates are uniformly worse among black Americans, studies also show that when access to care is controlled for, i.e., the VA, survival times are more equitable. But more on this later. So disparities also exist in thoracic surgery for lung cancer. As AJ just mentioned, race-specific spirometry instead of race-neutral spirometry to determine post-op lung function may affect access to care. But what about individuals who undergo surgery for stage 1 non-small cell lung cancer? So what we know is that black men underwent full resection less often than white men for stage 1 non-small cell lung cancer. This is according to the National Lung Screening Trial, about 700 participants. Black women, interestingly, did not differ significantly from white men, as you can see here on the right. So now why is it that black men specifically continue to experience more surgical disparity? Is it because of their social determinants? Do they feel like they are more likely to be stressed and then thus smoke and then perpetuate this cycle? Do they have more distrust in the system? These are all really things to consider. And then finally, for post-operative outcomes, we also know that blacks are more likely to experience surgical complications more than whites. All right. As you can imagine, disparities also exist in management of advanced non-small cell lung cancer. The black race, the uninsured, and the Medicaid patients are less likely to receive recommended lung cancer for advanced non-small cell lung cancer. Next-gen sequencing, as we know, is one of the standards of care, differs significantly between whites and blacks. The whites are more likely to receive next-gen sequencing compared to whites at any given time. Also, what we know is that blacks are less likely to present for clinical trials compared to their white counterparts. And then finally, Medicaid and uninsured patients with non-small cell lung cancer are less likely to receive molecular testing, TKIs, and systemic treatments in general. So now how do we move forward, guys? After listening to my wonderful co-panelists earlier this hour, you may have noticed some unifying themes in the lung cancer continuum. Several interventions and concepts have been introduced earlier to overcome the challenges that patients face in the lung cancer journey. First, let's start with screening. The USPATF, they updated their guidelines in 2021. The updated guidelines lowered the minimum screening age to 50 years and decreased smoking intensity to about 20 pack years. In brief, this helps, but this may not entirely solve the screening problem. So Shustad and colleagues, he performed a cross-sectional study to examine the effects of LCS with the updated criteria. With the updated 2021 guidelines, they found that 161 patients were newly eligible. These patients were younger, African-American, and more likely to be smokers. Sounds good, right? So Pu and colleagues then evaluated the eligibility of patients in the inhaled epidemiology study to compare the 2021 USPATF lung cancer screening criteria with the other lung cancer screening criteria, namely the 2013 USPATF and the prostate, lung, and colon and ovarian screening of 2012. And what she found was that, or what their group found was that there was improvement in racial disparity when we compare the two USPATF guidelines, but essentially there was not that much of a difference in racial disparity when comparing it to the PLCO from 2012. Of note, there is some suggestion that updated guidelines may actually, in fact, worsen the LCS disparities. So Lozier et al., their study raises an interesting question. Will the updated USPATF LCS guidelines improve or exacerbate disparities in LCS provision? So evaluating the behavioral risk factor surveillance system, black individuals undergoing LCS are significantly more likely to be older, above age 65. So what does this imply? They probably have insurance. They probably have Medicare. So since black individuals are more than twice as likely to be uninsured or underinsured, the new LCS criteria will address younger patients. Will then non-white younger patients who are newly eligible for LCS be helped by the updated guidelines if they don't have insurance? Thus, you know, it will be interesting to see over the years if the new LCS criteria may even worsen disparities or help bridge the gap. All right, next. Centralized programs may also reduce disparities in lung cancer screening. Two recent studies highlighted the relationship between race and socioeconomic status in LCS. Kim et al. first studied individuals undergoing LCS at either centralized or decentralized programs to evaluate the association of race with LCS adherence. The annual adherence to LCS in the centralized group was significantly better compared to the decentralized group. In a similar vein, adherence to recommended follow-up was higher compared to the decentralized programs. Of note, in patients with negative baseline scans, adherence to follow-up was 27% lower in black patients in these decentralized programs. So then follow-up analysis of the decentralized LCS programs showed that among black and white participants, adherence improved with the socioeconomic status. And also in black patients, socioeconomic status accounted for 45% of the total effect by race. So how can disparities in treatment for non-small cell lung cancer be eliminated? Perhaps a VA study can shed some light. A retrospective study of approximately 18,000 veterans was performed to evaluate overall survival and lung cancer-specific survival between blacks and whites. As you can see here, there are no differences between blacks and whites in those who underwent A, operation, B, radiation, and C, other or no treatment in this figure. There was a small difference between races for those who underwent surgery for the first couple of years, but the difference later on became marginal. There was also no significant racial difference in patients receiving operation and radiation, overall survival among patients who underwent surgery, overall survival among all patients, and then lung cancer-specific survival among all patients within each treatment category, regardless of race. So now, you know, this suggests does equal access to health care reduce the racial differences? Now finally, quality improvement initiatives can also reduce disparities. The ACURE study was developed by the Greensboro Health Disparities, and that's a community academic medical partnership with expertise in anti-racism community-based participatory research. Their intervention was diagnosed to address systemic racism and reduce treatment gaps between blacks and whites with stage 1 and stage 2 non-small cell lung cancer. Their intervention consisted of a real-time warning system, one, to identify unmet care milestones, two, race-specific feedback of lung care treatment rates, and then three, nurse navigator. So the intervention started by leveraging a real-time registry to flag missed milestones and appointments with daily downloads. This is something that they started many years ago, but I know now many of us in the room have access to EHR. So the milestones were no follow-up schedule within the first 30 days of initial visit, no surgery or radiation within 90 days of schedule, and then no surgery or radiation within 120 days. So then, after creating this alert system, providers and staff were then provided quarterly feedback of clinical care performance aggregated by race every quarter to improve transparency of care. And then next, nurse navigators and physician champions were employed to follow up with patients for missed appointments. You know, the nurse navigator would contact the patient when they had these reminders and understand the reason for the missed appointment. Was it because they didn't realize they had the appointment? Was it because they didn't have a ride, they couldn't take off work, they had issues with child care? Or was it more belief-related? Like they had barrier beliefs, like they thought they would feel better with prayer, or is it because they had distrust in the system? Finally, after that, most patients would convert if it was a distrust issue, and they would help reschedule the patient. The physician champion, what they do is they support the navigator in consulting with cancer teams for patients not receiving the standard of care, and also supporting the nurse navigator wherever else needed. So with all these components, this QI intervention is basically designed to enhance transparency, enhance accountability within the cancer team, and then thus achieve quality improvement. So there were three groups compared. There was a retrospective, which is essentially a baseline group between 2007 to 2012. You can't improve something if you can't confirm there's a problem. Then there was an intervention group, about 300, and this was at UPMC and Cone Healthcare. And then finally, there's a concurrent group. What the concurrent group is there for is to compare if the improvements are related to secular trends. The primary outcome was receipt of potentially curative treatment, such as surgical resection or stereotactic radiation. Both treatments were thought to be superior to no treatment. So we had 28,000, excuse me, 2,800 in the retrospective control group. We had about 360 in the intervention group, and then about 590 in the concurrent controls. It is interesting to note that, as you see here, there are only about 16% of blacks in the retrospective group, compared to 32% in the intervention group, and then 13 in the concurrent group. And you might be wondering, why is that? So this is kind of accepted in quality improvement. So the primary aim of the cure was power to address black-white treatment disparities, hence the oversampling in the intervention group. So the within-group and between-group comparisons were made, and then given the main focus of the study was to examine whether the intervention mitigated black-white disparities, demographic comparisons were made between the blacks and whites, examined here in Table 2. And then what we can appreciate is that the gender, clinical stage, COPD, and Charlson score were similar between both blacks and whites. What I do want to highlight is that blacks were younger, and they had lower income than white patients in this cohort. So for the within-group comparisons, meaning the comparisons of the blacks and the whites in each group, more whites received curative treatment than blacks in the retrospective group. No surprise. That's why we're here today. There was no statistically significant difference between both races in the intervention group. So 96.5% of white patients versus black patients, 95%, underwent and received the primary outcome, which was receipt of either curative treatment for their lung cancer. Adjusted analysis confirmed the treatment parity. There were also similar rates within the intervention group for the secondary outcome of surgery with similar results between races, and that's 75% for whites and then 76% for black patients. When the between-group comparisons were made after the QI intervention, black patients in the intervention group received curative treatment at a statistically better rate than whites in the baseline group. And similarly, the intervention group continued to fare very well compared to the concurrent group. So what does this mean? Timely interventions of missed milestones by teams trained in racial equity can improve outcomes for black patients. So this was a secondary analysis of the cure, where the primary outcome was surgery within eight weeks of diagnosis, and there were similar success rates as the original study, meaning that the intervention group performed significantly better than retrospective and also better than the comparative group as well. The limitations of this QI study, one was the oversampling mentioned, but again, we don't really see that as a limitation. And the second was that most study participants were indeed insured. So this doesn't affect necessarily the, doesn't really help, or we don't know if it helps the uninsured issue. So here's a summary of ways to reduce disparity that we discussed, including updated screening guidelines, a centralized system, equitable health care, nurse navigator with milestone reminders, as well as race equity training. The race equity training, just to note, is that in the CURE trial, it was very brief and just at the initial onset of the study. Obvious, some of these are more achievable than others in a real world sense, but I have more options for you. Additional ideas include the community-based participatory research. So CBPR collaboration has been shown to enhance the reach adherence and ultimate health outcomes by addressing critical determinants of health. They identify health concerns, stress leadership, they engage in intervention design, and they also collaborate to help interpret the findings and what they mean. Mobile screening units are also effective. So LCS mobile units have become available to increase knowledge and uptake at the community level. For example, the uninsured and racial minorities benefit from mobile mammography vans for breast cancer screening. So this should ideally work for lung cancer as well. And then finally, mobile technology, access to care through telehealth and text reminders that they have an appointment coming up. So in closing, multiple disparities exist throughout the lung cancer continuum. Recognition of health and racial inequities, updating guideline practices to adjust for racial biases, socioeconomic biases, improving access to care, centralized care, QI interventions aimed at accountability and transparency. These may all help bridge the disparity gap. And this concludes the discussion portion of our panel. I'd like to thank my wonderful co-panelists, AJ, Brett, and Mark. They're all super knowledgeable, organized, and excellent to work with, five stars on Yelp. I'd be remiss not to acknowledge Sadia Faiz. She's our shadow graphics editor and the thread that really strung our group together, as well as Patricia Vera, a staunch advocate for equitable lung cancer care and making waves in this field with her work. And then finally, Diane Stover, another major proponent of health equities and a 360 role model to anyone treating patients with lung cancer. Thank you all for coming. I'd love to open up the floor for questions. Thank you.

disparities

lung cancer continuum

screening

treatment

outcomes

access to care

race-neutral interpretive strategy

spirometry

healthcare access