Journal of Clinical Oncology (JCO) Podcast

Host Dr. Davide Soldato and Dr. Aaron Mitchell discuss the JCO article "Quality of Treatment Selection for Medicare Beneficiaries With Cancer"

TRANSCRIPT

Dr. Davide Soldato: Hello and welcome to JCO After Hours, the podcast where we sit down with authors from some of the latest articles published in the Journal of Clinical Oncology. I am your host, Dr. Davide Soldato, medical oncologist at Hospital San Martino in Genoa, Italy. Today, we are joined by JCO author Dr. Aaron Mitchell. Dr. Mitchell is a medical oncologist working at Memorial Sloan Kettering Cancer Center where he is also part of the Department of Epidemiology and Biostatistics. Dr. Mitchell specializes in treating genitourinary malignancy and has a research focus on improving how the healthcare system helps people with these and other cancers. So today, Dr. Mitchell will be discussing the article titled, “Quality of Treatment Selection for Medicare Beneficiaries with Cancer.”

Thank you for speaking with us, Dr. Mitchell.

Dr. Aaron Mitchell: Well, thank you for inviting me. I'm very glad to be here.

Dr. Davide Soldato: So I just wanted to introduce the topic by asking a couple of questions, very general, about the background of the article. So basically you reported the data using the SEER-Medicare to assist to assess the determinants of optimal systemic therapies delivery and selection. So, in particular, you focused on individuals that were diagnosed with cancer who were Medicare beneficiaries and in particular were part of the low income subsidy, which is also known as LIS. So I just wanted to ask you if you could briefly explain to our listeners how this program works, and what was the rationale of the study, and if there is any element of novelty in your study compared to what was done before the study was published.

Dr. Aaron Mitchell: Yeah. So that's a lot to cover, but yeah, a lot of opportunity to introduce the low income subsidy program which is a very important part of the Medicare program for prescription drugs, but often one that flies under the radar a little bit in the policy discussion. So this subsidy was created synchronously back with the Medicare Part D Program, which was created in 2006. There was some anticipation that for some high cost drugs, not all patients would be able to afford them even with the Part D program insurance as it was being created. And so they created a pathway to give an additional subsidy to some patients who had low income, who were anticipated to being at need and needing that assistance to afford high cost drugs. As the number of high cost drugs has really risen since 2006, this program has played an important role in helping patients afford drugs, especially those who need very expensive cancer drugs.

And what this program does is, once you meet the eligibility requirements, which require patients to have both quite a low income. So if you're single, that is at 135% of the federal poverty limit or below, and it also places some restrictions on assets. You also have to have low assets, so low income and low assets in order to qualify for the subsidy. But then once you do, the subsidy is really quite large. Patients who qualify for the LIS at the full subsidy level will pay about $10 per month per drug, even for specialty cancer drugs. So if you think about drugs such as those that we use to treat prostate cancer, my specialty, drugs like enzalutamide or XTANDI that run $15,000 to $20,000 per month, the out of pocket cost for a low income subsidy beneficiary is $10. So that is a huge discount. $10 isn't nothing, but even for someone with a low income, if they've got one or two cancer drugs that are at this rate, it's something that they can often afford.

This program applies to Part D cancer drugs that are prescription drugs basically. By and large, these are oral pills that patients are taking on a daily basis at home. These are the drugs that the low income subsidy program applies to. So if a patient needs a drug like that to treat their cancer, then they are able to receive it at very low cost. And what you'll see is a patient- in the studies that have been done, when a patient has low income, low enough for them to be able to qualify for this program, they then have better access to these drugs. You see increased adherence rates, you see increased prescription fill rates. And then when someone, when their income is just high enough to no longer qualify for this program, and they go back to regular Medicare Part D coverage, that's when the problems arise. So it's like as your income moves up the scale, you actually get more problems affording your cancer drugs. So that's the state of the literature so far.

And what we realize though, is that all these studies that have looked at the low income subsidy have really focused on just the Part D drugs themselves, the oral drugs. And that's certainly not all of cancer care. There is a growing number of oral drugs, but for many cancers, especially when you're talking about immunotherapy drugs or new systemic radioligand therapies, these are not Part D drugs, these are Part B drugs. And so even if you are low income and you're qualifying for this subsidy, it's not going to help you if you need a Part B drug. Yes, there are certainly a whole host of other programs and different avenues that we can get patients assistance, but some percentage of them, even though they're low income and high need, would not have assistance with a Part B drug.

So now, in coming back, the long answer to your question, our rationale was, let's look at these Part D low income subsidy patients and let's see what their access looks like, not just to the oral drugs, but to cancer care writ large. And can we study where they're fitting into the system, not only when they need oral drugs, but when they need any kind of cancer care across the board?

Dr. Davide Soldato: So basically, just to summarize, it was an extension of previous literature, but specifically evaluating whether novel regimens that use, for example intravenous drugs, they were covered at the same level and whether there were any inequities in access to cancer treatment under this specific program, which is the LIS.

Dr. Aaron Mitchell: Yes, I'd say that's a fair summary.

Dr. Davide Soldato: Okay. So more or less, you included 9,000 patients inside of the study and 25% of them were beneficiaries of the LIS program. And you specifically looked at factors that could be associated with not receiving therapies at all, and also whether the quality of care that these patients were receiving were any different compared to those who were not part of the LIS program. So I just wanted to see if you could guide us a little bit in the results, whether you see any kind of differences when we look at access to any type of systemic therapies and whether being a part of the LIS program modified access to the drugs.

Dr. Aaron Mitchell: Let me take this opportunity also to highlight a feature of our study that differentiates us a little bit from previous work that's been done. And this is around the specific definition of quality that we use. I know quality is in the title of the manuscript, but I think it's important to emphasize exactly what we mean in this study when we say quality, and it's something very specific. So our measure of quality references back to the NCCN guidelines, which I don't think our audience needs much of an introduction to that. It's the most worldwide recognized standard of care guidelines for oncology practice. And we specifically looked not only at the NCCN guidelines, but at their evidence block scoring system. So what we did was we looked not only at one set of guidelines, but we looked at guidelines across time. We looked at guidelines across our full study period, which was, give or take, 2015-2018, depending on the cancer. And we looked at each point in time to see what was the treatment regimen that was recommended by the NCCN guidelines as being preferred. Some of them make that designation, some of them don't. If there was not a designation of preferred, then we turned to the evidence blocks. And the evidence blocks, we then apply several different measures to kind of rank treatments from those that get high scores for efficacy and safety to those that get low scores for efficacy, safety and the quality of evidence. So we basically come up with a kind of a rank list of the recommended treatments at each point in time. And then we look at the ones that are the highest, we say which are the most highly recommended treatments at any given point in time. That then becomes our definition of quality treatment. And I'm saying this with air quotes, we use the term “optimal treatment” in the study. Did they get that treatment? If there were ties, you could have gotten either of the two treatments that got the equally good score, did you get that treatment versus did you get anything else?

So then getting back to our analysis, what we really did was kind of a two-stage study. First, we put all of our patients into our pool, into one big analytic model. And we looked to see what are the factors that predict or are associated with a patient either getting no systemic therapy or any systemic therapy. And then as a second question, we look at the patients who got some form of systemic therapy, and then we ask, again, what percentage of those got the optimal treatment or high quality treatment as opposed to one of the more lowly recommended treatment regimens? So that's how we asked it. We found that patients who were low income subsidy recipients, the low income ones, they were both less likely to receive any systemic therapy. And then even the ones that receive systemic therapy, the ones who made it in the door to see their doctor or their part of the system, they still were less likely to get the optimal treatment that was recommended for their cancer type at the time that they were diagnosed.

Dr. Davide Soldato: So basically, even when you are a part of this subsidiary program, you still have a lower access to any type of treatment. And even if you get treatment, you kind of get the ones that were not the preferred according to the NCCN guidelines, or at least they were not scoring as well as those specific regimens. But I think that what our audience might be wondering about is that frequently there are also some other types of characteristics, for example, age or number of comorbidities, which can be associated with having a low socioeconomic status. So I was wondering whether in the analysis you kind of looked specifically also at patient factors, for example, income rather than age or comorbidities, and whether you found any significant association with those and whether it was something that you planned to do in your study.

Dr. Aaron Mitchell: Yes. So we looked at many patient factors and those included age and they included the degree of comorbidity. And what we saw with respect to those characteristics was not too surprising. We saw that patients who were older were less likely to receive systemic therapy. We saw that patients who had more comorbidities were also less likely to get systemic therapy. And then across our different designations of treatments, we saw that those patients were also less likely to get the optimal treatment for their cancer. This result though, we would say it certainly needs more study in the future, but it's not immediately concerning. And that is because for patients who have more age, more comorbidity, those often correlate with frailty. And so it could be that these patients aren't getting optimally treated or it could be that their oncologists are just making clinically appropriate decisions about patient selection.

We saw as we were doing this work that the treatment regimens that are often getting the highest recommendations from the NCCN, hence, it would become our definition of high quality optimal treatment, are often ones that are aggressive. They're often ones that are multi-drug combinations. They're often ones that it's not just your old antineoplastics, it's the antineoplastics plus an additional immunotherapy or plus a targeted drug. So it's the ones that are more aggressive by and large, and that might be in some cases more than a patient who is older, more frail, could be able to tolerate. And so the oncologist might be making inappropriate judgment to say I'm going to do something a little bit less aggressive here and make an appropriate trade off between anti cancer efficacy and safety.

I think we've got kind of a bookmark there and we can look at those trends in the future. So we saw that kind of as expected, and then we turned and looked towards the low income subsidy. And our premise there is, well, your income shouldn't predict what you're getting clinically. In an ideal world, you'd be able to get the appropriate treatment for a patient, and not depend on whether their income is above or below 135% of the poverty limit. So that one seems more like on its face an immediate concern.

Dr. Davide Soldato: Thank you very much for the explanation. I was just wondering, did you make some kind of selection when you were analyzing specific diseases or settings where you included just metastatic patients or you also included patients with early stage neoadjuvant treatments? Because I think that it is also very interesting from the perspective of the objectives that we have as oncologists when we are administering systemic treatments.

Dr. Aaron Mitchell: Yeah, thank you for bringing that up. That was also one of the goals of our study was to be broad. And we wanted to look for factors, whether it be low income subsidy, whether it be age, socioeconomic background, etc., things that would be broad predictors of outcomes, and by which I mean care delivery outcomes across the board. So not just for, let's say, metastatic breast cancer, but also across any cancer that a patient might walk in the door with, what are the systemic predictors. And so when you mentioned before that our overall cohort is approximately 9,000 patients, that's 9,000 patients split over a variety of what we call clinical scenarios or clinical indications. And that includes multiple solid tumor as well as liquid tumor malignancies. It includes both patients who are initiating systemic therapy with palliative intent for metastatic disease. It also includes several groups of patients who are getting adjuvant therapy. So we want it to be as broad as possible. Our selection of those scenarios was really done with the goal of being as broad as possible and really bringing in everything that we could within the constraints of our data source. And that was really the only limitation that we applied in concept was tumor types that are common enough to have a meaningful sample of patients to analyze. So, one, are there enough patients? And then two, are you able to identify this specific group of patients within SEER-Medicare data? Because when the NCCN divides groups of patients by biomarkers that are not available in SEER-Medicare, we can't really say, “Oh, we're going to study this group of patients.” That would then be one that we have to leave on the side and not include. But everything else where one of those things didn't apply, we tried to include it as best we could.

Dr. Davide Soldato: Thank you very much for the explanation. And among the scenarios that you included in the study, were there any striking differences in terms of access to treatment and access to quality treatment the way you define the study?

Dr. Aaron Mitchell: Yes, there were differences between these different cancer types, these different cancer indications, but they're not differences that I want to over interpret or read too much into. Certainly, every cancer indication is going to be different, but when we start getting into the individual cancer types, the sample size does get smaller. And we've not done formal tests of comparison or heterogeneity among cancer types. So I don't want to say that the differences which we certainly do see, like numerically, there are differences in the proportion of patients who are getting optimal treatment versus no treatment. I don't want to say that it's because the low income subsidy status or patient age has a bigger impact, let's say for lung cancer than breast cancer. I want to say that is heterogeneity for potential future study when we are able to do a similar follow up analysis with say a larger sample size. I don't want to over interpret those differences at the moment.

Dr. Davide Soldato: I was just wondering in case there was anything in particular that you wanted to highlight. But in the end, I think that we also have to acknowledge that the data are based on claims data, observational data. So maybe you're right when you say we should not over interpret this type of difference.

And this is just to speculate a little bit, do you think that if you would look at this same specific question in a more contemporary diagnosis frame, like for example, you refer to the fact that most of the diagnoses were between 2016 and 2018. Now that we have more and more of these drugs that would qualify as Part B in the adjuvant or new adjuvant setting, do you think that you would see more differences compared to what you observed in the current study or do you think that it would be more or less the same? Of course this was not part of the analysis that you did, but it's just to have your opinion on the topic in general.

Dr. Aaron Mitchell: My expectation would be that since not much has changed with respect to the low income subsidy program from the time period of our study until now, my baseline expectation would be that those results would hold. On the other hand, it is the case that there have been improvements to the standard Medicare Part D benefit since the time of our study. So the low income subsidy patients would be paying the same low out of pocket costs that I mentioned before, about $10 a month give or take, for a specialty cancer drug. But what has started to happen is that for everyone else, their coverage has improved. Because in the US we're in the process of closing, or I think now we finally finished, but you know, a few years lag in claims data, we've closed what used to be called the donut hole, where there was this big coverage gap where patients had to pay a large amount out of pocket for drugs. So there might therefore be a narrowing of the difference, let's say between our low income subsidy participants, the lowest income patients, and then everyone else. But not so much because the low income subsidy status improved or changed, but just because the baseline level of coverage for everyone else may have improved, narrowing that gap. So I'd say that would be very possible.

And if your question is more geared towards not so much policy changes, but treatment landscape changes, I would say the big thing that I would maybe guess, and again, this is very much speculation, but you introduce the speculation in TBD on follow up. I think the big change in the landscape has been the broadening indication and uptake of immunotherapy drugs, our PD-1, PD-L1 inhibitors, for a variety of cancer types. And I think the way that that would manifest in our data, were we to repeat it in a more contemporary data set, would be, I think that the access for, let's say, that any systemic therapy among older patients might change. And that is because rather than just having your cytotoxics in hand, the clinical oncologists now know that for many cases there's if not first line therapy, then second line therapy for patients who don't qualify, you can go straight to it, to someone who's not a chemo candidate, you've got a much more tolerable treatment in your back pocket. And so I think that for patients who are more old or more comorbid, we might start to see that a greater proportion of them receive some systemic therapy, it just might not be the cytotoxic agent that is still most highly recommended. It might be, say a single agent, PD-L1 inhibitor, because their oncologist wants to be able to give them something. So I wouldn't be surprised if that gap starts to narrow as well if you're measuring no systemic therapy versus any systemic therapy.

Dr. Davide Soldato: And going back to the policy part of the study that you did, do you think that the results of the study that you published in the JCO can better inform policy makers on how to make these treatments more available and be sure that the largest possible proportion of patients gets a systemic treatment and gets the optimal systemic treatment?

Dr. Aaron Mitchell: Yes, I do think that this study has some direct and indirect policy implications. I think that our finding is one to highlight the low income subsidy program and maybe help it not to fly under the radar so much anymore. I think all the work that has been done on how much it has helped patients who need oral cancer medications is great, and it shows how beneficial this program can be. We're now shining the light kind of everywhere else and saying, “Okay. That's great. Here's how well it can work when it covers an oral drug, but we've got this group of low income patients who are still at need and they're still very clearly not able to access everything else. When it's not axitinib that they need, it's a pembrolizumab, they're still very much behind the curve and they need some help.” So I think that's one thing just to call attention to this as an ongoing problem. Low income patients, it's not a solved problem yet. It's something that needs further attention.

And then for direct policy implications that are on the table, I think we're about to see the Medicare program be able to start negotiating not just Part D drugs, but also in future years, Part B covered drugs and try to lower the price for everyone, both for insurance, both for Medicare itself. And then to the extent that that boils over to the patient's out of pocket responsibility, it'll start to reduce the patient out of pocket costs as well. So I think we can look forward to hopefully an aggressive negotiation program by Medicare to start to directly lower the prices of Part B cancer drugs that these patients are unable to afford.

Dr. Davide Soldato: Thank you very much. You did the research you published in the JCO, but you really seem very passionate about the topic of care delivery and quality of care and policy. So I just wanted to ask on a personal note, how did you come to this area of research which is frequently not one that is very cared for by oncologists? It's more frequently something that biostatisticians or public health scientists put their attention to. I just had this curiosity and I wanted to ask you if you could explain a little bit how you came to this area of research.

Dr. Aaron Mitchell: Thank you for asking. That's a great question. I'll tell my favorite story about my journey there. I entered medical school planning to be a clinical investigator or maybe even a basic science researcher, and I had some background in that. I went to medical school at NYU where the teaching hospital is Bellevue, which is a large, well known public hospital within New York City. And my eyes started to open regarding the inequities in the system. You always hear about it, you read about the problems in the US healthcare system, but then when you see it on a day to day basis and you can walk four blocks from a private, very well resourced hospital to see a patient with a similar condition four blocks down the road at a under resourced public hospital getting very different treatments and receiving very different outcomes, the injustice in the system really hits you on a visceral level. And it was really, I would say, as soon as I started my clinical rotations in medical school that I realized maybe that's where I can make the most impact with my career and just really fell into it. By the time I was done with medical school, I then knew that I wanted to do something that was in the health policy space. And then by the time I was done with residency, I was like, “Oh, someone had mentioned the words health services research” and the light went on. It's like, “Oh, that's me. That's what I want to do.”

Dr. Davide Soldato: Thank you very much. That was a nice story. And I really think that we should all work towards trying to make sure that the inequities inside of the system are eliminated as much as possible.

So I think that this concludes our interview for today. So thank you again, Dr. Mitchell, for joining us.

Dr. Aaron Mitchell: You're very welcome and thank you so much for your interest.

Dr. Davide Soldato: We appreciate you sharing more on your JCO article titled, “Quality of Treatment Selection for Medicare Beneficiaries with Cancer.”

If you enjoy our show, please leave us a rating and review and be sure to come back for another episode. You can find all ASCO shows at asco.org/podcasts.

The purpose of this podcast is to educate and to inform. This is not a substitute for professional medical care and is not intended for use in the diagnosis or treatment of individual conditions.

Guests on this podcast express their own opinion, experience and conclusions. Guest statements on the podcast do not express the opinions of ASCO. The mention of any product, service, organization, activity or therapy should not be construed as an ASCO endorsement.