Immuno-Oncology Lottery Ticket Response

An Interview with Michael Seiden, MD, PhD


Michael Seiden, MD, PhD, is Chief Medical Officer at McKesson Specialty Health and the U.S. Oncology Network. He is responsible for leading the development of a clinical care delivery system that integrates community physician care with the clinical, operational, and technological expertise of McKesson Specialty. Previously, Dr. Seiden served as the CEO and President of Fox Chase Cancer Center. Dr. Seiden is also an ICLIO Advisory Committee member.

Could you talk about your current role and responsibilities?

Michael Seiden, MD, PhD: I’m the Chief Medical Officer for McKesson Specialty Health and the U.S. Oncology Network. I am also the leader of the research executive committee for the U.S. Oncology Network.

I have a couple of different roles. One of them is to keep track of how and what the 1,100 doctors and the several hundred advanced practice providers are doing in the network. This includes what they need to deliver high-value care in the communities, and how we, at McKesson, can help them, particularly in the evolving world of alternative payment models and value-based care. We are a major participant in the Oncology Care Model. We have a number of new requirements as far as how care is delivered to patients and what’s documented. So, that’s one of my roles.

My second role is I sit on the McKesson leadership team, the corporate leadership team. My role is to represent the voice of the physicians and the voice of the patients, as far as what are the key strategic investments the corporation should be making for cancer care today.

My third responsibility is to participate in things like ICLIO, and the American Association of Cancer Research (AACR), the American Society of Clinical Oncology (ASCO), the Society for Translational Oncology (STO), and NRG Oncology, and various healthcare think tanks to try to develop a better understanding of where cancer care and where healthcare, in general, is going. In the words of Wayne Gretzky, “I try to skate where the puck is going to be, instead of where it is currently.”

The Innovation Factor

You mentioned that you sit on the corporate McKesson leadership team to represent the voice of clinicians and patients in cancer care. What kinds of issues do you see in that role in relation to immuno-oncology, specifically?

Dr. Seiden: I think there are two issues. One is how do we educate physicians in a progressively more complex world of cancers, which are being progressively subdivided? Molecular testing, for example, which helps perhaps select the best therapies for individual patients, and with more than a dozen drugs being approved now every year, and many of them getting secondary indications.

How do we educate in this area, especially for community physicians who, in general, see every disease that walks through the door? How do we provide decision analysis tools, educational tools that are in the workflow, at their fingertips, so they are as smart as possible about the best way to treat the patient in front of me, now? That’s one issue that applies to immuno-oncology.

The second issue is the cost of care. I was recently looking at our cancer drug spend for immuno-oncology (I-O) agents. In August of 2015, our I-O drug spend was 6.7% of our overall drug budget. Now, one year later, the network has gotten a little bigger—of course, there’s drug inflation—so our drug spend for the network a year later is up about 20%. But, our I-O drug spend is up from the number I gave you; it’s now 15.8%.

So, our I-O drug spend in one year is up about, in gross dollars, 320%, where overall growth spend is up 30%. Now, in a world of alternative payment models, where you are paid for global cost savings, there is this challenging balance of wanting to try to win in these models. The biggest part of the cancer bill, in many cases, is the drug spend.

How does a physician take the best care of the patient in front of them and, at the same time, understand the evolving world of value-based therapy?

These exciting agents are now available in melanoma, lung, head and neck, Hodgkin’s, bladder, renal, and there are at least three or four more that are close to approval. I think the expectation is that the growth spend next year will be significantly more for I-O agents. And the largest of the alternative payment models is the Oncology Care Model, and they’re using the drug spends or the global healthcare spends from a couple of years ago as the baseline.

They are building in an innovation factor to represent how costs will go up because of innovation, but that innovation factor is likely going to be in single percentage points, not tens of percentage points. How does a physician take the best care of the patient in front of them and, at the same time, understand the evolving world of value-based therapy?

I can’t say there are any easy answers. But I think it’s a very active dialogue. Can we make up for the increasing costs associated with drug innovation by keeping patients, otherwise, healthier and out of emergency rooms, and out of hospitals? And if protective therapy isn’t available for patients, can we provide more effective ways for best supportive care so they don’t spend time in the last months of their lives in intensive care units or somewhere that isn’t going to help them live longer or better?

As you think about this, this is very hard to do in, say, a hospital setting. One of the challenges is that we have 450 sites of service in 22 different states, and the average age of the physicians in the community is about 54. Many of them trained right around the time we were trying to figure out the difference between B-cells and T-cells, and well before the era of cancer genetics and cancer immunotherapy existed, including things like high-dose interleukin 2 and high-dose interferon, but not much else. It’s an incredibly fascinating kind of medicine. A lot of innovation, both payment innovation, therapeutic innovation, and I’d also say care innovation.

As patients become better consumers, and as pharmaceutical companies become more aggressive about going directly to the consumer through a variety of different media, the discussion is, “What are all the options; what are the best options?” If I have a $10,000 a year out-of-pocket copay, what are the financial ramifications of these options? What are the documentation issues that are now required by Meaningful Use to not get penalized by payers?

All of this innovation, each on its own—laudable—but in concert together, they’re all playing slightly different symphonies. Trying to figure out a way that a physician and a patient can successfully navigate through this is becoming progressively more complicated, unfortunately.

Managing Cost

What is your sense of how the physicians in your network and physicians specifically are responding to all these immunotherapy innovations in particular? What sense do you get of what it means for them, what they’re thinking about, what it means for their practice, their attitude towards the rapidity of change?

Dr. Seiden:
We have a pretty robust research portfolio and, as you might guess, a larger and larger portion of that portfolio includes I-O agents. We do have the advantage that many of the physicians in our network are getting some familiarity with these agents in a well-structured research environment, which is useful. I will say that it depends a little bit on how you look at it. On the one hand, I’d say that at times uptake has been a bit slow and that’s been for two or three reasons, and I’ll go in reverse order.

We have to manage cost. We have to be careful how we use drugs… The amount of money we spent on IO drugs this year compared to last year is up 300%.

One is a concern by some people that, “Wow, these drugs are incredibly expensive, and how are they going to get paid for?” I’d say that is the least of the three, but occasionally. Two has been either an early fantastic experience or an early “unfantastic” experience as biased. So, if you talk to some of the people who are using a ton of it, they will say, “Hey, the second patient I treated is in really horrible shape. . . [I] gave them this drug, it’s been two and a half years; they’re playing golf. I mean, it’s amazing. I’ve seldom seen anything like this in oncology. I now have a lot of people on this drug.”

The flipside is the slow adopter who said, “I hadn’t really used these drugs very much, but I was on call last weekend and I rounded on a person who was on, let’s say ipilimumab (Yervoy®) and nivolumab (Opidvo®) together, and they’ve been in the hospital for 21 days with refractory diarrhea, and had to be put on high-dose steroids, and had a complication, and even though I’ve never personally used it, it was a patient of my colleague, and I said, ‘Ooh those drugs are toxic.’ I’ve never used them.” Both of those are probably not quite right, but as far as the typical experience, those can definitely bias adoption.

I’d say the other thing, which probably is the number one thing, is the quickly rising cost. Physicians now have this in almost every one of their monthly meetings; we have to manage cost. We have to be careful how we use drugs. Now, with that as a caveat, I want to go back to an earlier point. The amount of money we spent on I-O drugs this year compared to last year is up 300%. So, while those have been some headwinds, clearly by the statistics the tailwinds are much, much stronger than the headwinds.

Broader Scope of Malignancies

So, there are many challenges in terms of adoption and using these agents in practice, but the overall enthusiasm for the response rates and the efficacy means that they are being used?

Dr. Seiden: Oh, they’re being used three times more than they were just a year ago. Part of that is there are more approvals. A year ago they weren’t approved in head and neck cancer; they weren’t approved in bladder cancer. But a year ago, they were approved in renal and melanoma and lung cancer. I think part of the increase is a broadening scope of malignancies, which are now FDA approved.

I will say one thing about our network—and I have sat on some panels with some of the academic colleagues— is that we are very, very much using them on-label, almost exclusively, which is how you’re supposed to use them. But I’ve sat on panels with academic colleagues where, in many cases, they said a large portion of the drug use was off-label. Not made up, but you know, “There’s a clinical trial that shows it has activity in disease X, and my patient had disease X, so I am using it.”

In most of those academic institutions, almost all of them, the physicians are shielded from the fact whether the drug is paid for. In the community, they’re not.

Treatment Decision Making

Which tools and resources are physicians using to help them make decisions about immunotherapy treatments?

Dr. Seiden: I would say, in general, right now they’re using no tools. When you look at the PD-L1 expression, for instance, as the biomarker, and the fact that you don’t need that test for Opidivo (nivolumab), but you do need it for Keytruda® (pembrolizumab) in lung cancer. That has added a lot more steps to the community oncology’s workflow.

Meaning, they’re often not in a hospital, so they have to request the tissue block; they’ve got to get the tissue block; they have to get PD-L1 expression; they have to probably get insurance approval to do the PD-L1 test. Then they have to wait for the test to come back, and then if the test is positive, they can give them Keytruda. If the test is negative, well maybe they can give them Opidivo.

Well, you know what the faster way is? Just give them Opidivo and then you can skip all those other steps. So, where time and efficiency is so important, I would say most of our physicians have not tested. It adds to the workflow, adds to prior authorization, adds to cost.

Immuno-oncology Lottery

Mutational analysis is interesting. It’s biologically logical, but it has not really been stringently validated as a useful marker. So, while I’m sure there’s a couple of physicians who are using it, right now, in general, we’re not in the community. I do think that if a validated, highly predictive biomarker came out, especially in this value-based world, it would likely be embraced. I would say the performance of that biomarker is really going to have to be good because—and I wrote an article about this recently, where I talked about the immuno-oncology lottery ticket.1

At least many of the studies have shown that even if you’re PD-L1 negative or PD-L1 low, there still is a chance that you’ll have a very good response. And there’s some variety of reasons why that might be the case, but the point I’m going to make is if you have a patient in front of you, and they have recurrent lung cancer or recurrent melanoma or recurrent kidney cancer…first of all, for half of these diseases, it’s not good. PD-L1 expression is important. For even those with PDL-1 expression, it does seem to enrich for a response.

If it was me, and I had a first recurrence or a second recurrence or progression of my non-small cell lung cancer, and I was PD-L1 negative, I think I’d probably say, “Look, I understand the chances are lower, but they’re not zero. And some of these responses are really pretty durable and pretty fantastic, and particularly for the PDL-1 inhibitors, for most patients, the toxicity is quite manageable, so I want a chance to win the immuno lottery.”

Now whether the FDA or insurance companies ultimately say unless you have a certain expression with this particular diagnostic test, we’re just not paying for it. If you’re wealthy and you can pay the $10,000 or $12,000 a month to get two or three months, even though you’re PD-L1 negative, and see what happens, then go for it. But, otherwise, sorry. You know, I don’t know. It’s a little hard to guess how that’s going to work out.

There is a high need to identify a test that shows this person has a high chance of responding.

Could you talk about where you see immunotherapy going in ovarian cancer? Because obviously there are some kinds of interesting clinical investigations going on at the moment that might have some practice-changing implications. What’s your perspective on that?

Dr. Seiden: I think it’s definitely worth more investigation. Clearly, the very first studies that have come out have shown modest activity,2-4 but as you start thinking about patients, which have different types of DNA repair mechanisms, defects, whether it’s in BRCA-1 or patients with Lynch syndrome, who have ovarian cancer, you could imagine using a molecular marker for therapy selection, whether it’s PD-L1 or mutational load or DNA repair statuses determined by HRD (homologous recombination deficiency biomarker) score, which is a predictive marker for PARP inhibitor therapy.5,6

There are other ways that people are looking at genomic noise in ovarian cancer or genomic instability. You could imagine us attempting to sort out which patient population will benefit. My reviews of the field are that there’s a couple of ultrahigh responsive diseases, like Hodgkin’s disease and Hurthle cell tumor and hyper-mutated Lynch syndrome-associated colon cancer, but for a vast portion of the other solid tumors, somewhere between 1 and 10, and 1 and 5 patients will have really pretty good, pretty durable response stages.

It’s pretty hard to find a tumor—maybe pancreatic cancer and a few others—but it’s pretty hard to find a tumor where that isn’t the case. The challenge is it’s a pretty expensive drug for most of the cancers you treat, four or five patients won’t get a great response. So, I think there is a high need to identify a test that shows this person has a high chance of responding. Well, that’s good. You make sure those patients get it.

The challenge is, what will save you money is coming up with a test, which will give you nearly 100% confidence so you can look the patient in the eye and say, “Listen, I understand there’s a lot of people with your disease who have had great responses to these drugs, but I promise you, you will not be one of them.” Then you could save a lot of money.

Now with the tests that are evolving, you have the group where it’s very, very likely they’re going to respond and they say, “Oh my God, well get them the drug.” But then you have the group where it’s much less likely, but they still might have a really good response. If you have a life-threatening cancer, and you hear that it’s, “The chemo stuff you’ve given me hasn’t worked, so why don’t we try that? [immunotherapy].”

The amount of discretionary time is decreasing rapidly, and at the same time, we’re where the pace of innovation and new knowledge is increasing exponentially.

Are there other things about your role that you would want to share with our readers?

Dr. Seiden: I do think we have to become progressively more clever in how we educate physicians. I think the pressure on physicians to skip conferences, to skip what I call the rubber-chicken dinner, to skip inviting pharmaceutical agents into the office to provide lunch…that drumbeat is getting loud. It’s partly because of [Physician Payments] Sunshine Act,7 but it’s probably even more so because with all the electronic documentations, and all the other work that’s now required, there just isn’t much time in the day.

And the world we live in and…10 years ago, I was, wow. I have email where I’m sitting at my computer. Now I carry two smartphones with two active email systems. I take phone calls in airports. The amount of discretionary time, essentially all busy professionals have, is decreasing rapidly, and we’re—at the same time—where the pace of innovation and new knowledge is increasing exponentially. So, how do you deal with that? Is it video vignettes, is it text messaging, is it some social media thing that flashes up on your Google glasses?

As I said, I really don’t have an answer, but I do think we are getting pushed, whether we’d like to admit it or not, to a point where much, much more of the medicine that gets delivered is delivered by computer, and not by humans.

I’m not saying you wouldn’t see a human being. But instead of me having to review your 40-page charts, some computer has to review your 40-page chart, and they’ve provided me the best solution in whatever world we live in. I just need your electronic signature. And, oh, by the way, if it’s under a certain acuity level, I don’t even need your signature; the computer will just sign it.


  1. Seiden M. Immuno-Oncology 2016 and Beyond: The Opportunities, Challenges, and Risks. AJMC.com. Posted February 4, 2016. Available at http://www.ajmc.com/journals/evidence-based-oncology/2016/february-2016/immuno-oncology-2016-and-beyond-the-opportunities-challenges-and-risks/p-4#sthash.o4Hz30GH.dpuf
  2. National Cancer Institute. Clinical Trials Search Results. Ovarian epithelial cancer. Available at https://www.cancer.gov/about-cancer/treatment/clinical-trials/search/results?protocolsearchid=6119596 Accessed September 23, 2016.
  3. Cancer Research Institute. Cancer Immunotherapy: Checkpoint Inhibitors and Immune Modulators studies. Available at http://www.cancerresearch.org/cancer-immunotherapy/impacting-all-cancers/ovarian-cancer Accessed September 23, 2016.
  4. Rodriguez-Freixinos V, Mackay HJ, Karakasis K, Oza AM. Current and emerging treatment options in the management of advanced ovarian cancer. Expert Opin Pharmacother. 2016;17(8):1063-76. doi: 10.1517/14656566.2016.1159295. Epub 2016 Mar 16.
  5. Helwick C. Homologous Recombination deficiency score correlated with response to platinum in breast cancer. The ASCO Post. Posted January 25, 2016. Available at http://www.ascopost.com/issues/january-25-2016/homologous-recombination-deficiency-score-correlated-with-response-to-platinum-in-breast-cancer/. Accessed September 23, 2016.
  6. Telli ML, Timms KM, Reid J, et al. Homologous recombination deficiency (HRD) score predicts response to platinum-containing neoadjuvant chemotherapy in patients with triple-negative breast cancer. Clin Cancer Res. 2016;22(15):3764-73. doi: 10.1158/1078-0432.CCR-15-2477.
  7. American Medical Association. Physician Financial Transparency Reports (Sunshine Act). Available at https://www.ama-assn.org/ama/pub/advocacy/topics/sunshine-act-and-physician-financial-transparency-reports.page. Accessed September 23, 2016.