A conversation with Dr. Adam Dicker on the multi-disciplinary aspects of cancer care
I had the pleasure of sitting down with Dr. Adam Dicker, a leading oncologist who focuses on digital health. He’s Senior Vice President and Chair of Enterprise Radiation Oncology and Director of the Jefferson Institute for Digital Health in Philadelphia. We discuss the complexity and multi-disciplinary aspects of cancer care and how it’s changing with the 21st century cures act. Bottom line: we can get better results with more complete data.
Leonard: Hi, this is Leonard Kish on Real World Evidence Today, that's RWEtoday.com. I'm here with Dr. Adam Dicker who's an oncologist and Senior Vice President and Chair of Enterprise Radiation Oncology and Director of Jefferson Institute of Digital Health. He's also Professor of Radiation Oncology and Pharmacology and Experimental Therapeutics at the Sidney Kimmel Cancer Center at Jefferson Health which is in Philadelphia. He's also a fellow at the American Society of Clinical Oncology and a leading expert on prostate cancer and brain tumors with the focus on digital health and translational sciences applied to oncology. You're also an advisor to Self Care Catalysts platform that helps patients manage their conditions including cancer report on the progress through clinical trials including their quality life and patient reported outcomes. So thank you for joining us today, Dr. Adam Dicker.
Dr. Dicker: My pleasure.
Leonard: So obviously (you have) very diverse interests and things that you’re working on a daily basis. They bring a lot of complex elements together but oncology and cancer is really obviously one of the most complex diseases out there. The Emperor of All Maladies it's been called. So we're seeing in your work just how complex it is but also how complex the best treatments are going to need to be in order to combat this disease. So tell me a little bit about how you're bringing together now immunologists, oncologists, data sciences, software developers, machine learning, and AI people, all these different disciplines together and how that's working today to combat cancer.
Dr. Dicker: I think it's taken me about 25 years to kind of get to a certain point in terms of the way I think about helping patients and taking care of them with cancer and initially in your training if you have a very narrow type of focus and you master certain skill sets and clinical skill sets and develop expertise and then you develop other skill sets and as you get older and a little more comfortable then you can start taking more of a macro view. I think it's the macro view. So you could have the greatest immunotherapy for cancer but if one of the challenges is toxicity for example of that immunotherapy and if the patient is experiencing this toxicity as an outpatient so they may get an infusion or they may get an injection of some radiation then they go home. So it's in between clinic visits that they're experiencing all this toxicity. Because thus area is variable, meaning there are different levels of expertise in administering cancer therapy around the world, some patients may be more stoic, “they may not necessarily ask for help” if they're experiencing toxicity.
There are resource limitations. Some places around the world may have more resources, nurses physicians. (Some may) be in contact with the patient more and some maybe less. So that means that how you're monitoring patients is going to vary. When you add all that up you could have a terrific therapeutic for cancer, but not everyone's going to be getting exactly the same type of care for all sorts of reasons. Some of these relate to systems and processes. Some of these relate to experience of the health care providers, and some relates to the patient themselves. So it's that type of complexity that drew me to Digital Health Solutions.
The reason to bring in diverse stakeholders is because everyone looks at the problem a little differently. (If you) are thinking about what's the biomarker. How can we predict which patients are going to have more toxicity? Behavioural psychologists might be interested in the digital health platforms by which we ask patients about what they're experiencing and how they describe it in their own voice through patient-reported outcomes. The technology people are more interested in is the user interface. The user experience of these technologies. Then there's the analytics of a lot of stuff and how you take these disparate sort of datasets and lay them on top of each other to kind of put together a more complete picture is this part of the challenge. So I don't think I've been successful with all the different stakeholders in putting together some unified holistic view, but I'm starting to appreciate that the approach I may have had five years ago probably needs to be modified because of all the complexities that I mentioned earlier.
Leonard: When we look at clinical research with patients traditionally that hasn't included what happens at the home Patient-reported outcomes and surveys that are done in the clinic. Do you see research and trials changing now that we have wearables and mobile devices and ways to monitor patients not just for those 10 minutes every month or so but more continuously? Or have you seen areas where we're starting to see results in that aspect?
Dr. Dicker: Yeah certainly work by Ethan Basch at UNC Chapel Hill which demonstrated that if you ask patients, you engage patients, and ask them about their symptoms and you act on the information, you can make a difference in their lives and that actually results in improved survival. I think that's one body of data. There are other bodies of data. There was a French study in advanced lung cancer called the Move It Study. They did a great job there with a software company. I think there's greater interest now at the patient level to in part in terms of where cancer research is going and in part because of kind of the consumerism of digital health in terms of there are patient support groups. There are patient communities online whether it be Reddit or Twitter or Patients Like Me or Self Care Catalysts or others and then they’re finally at the regulatory level. There's interest by the FDA and others in things that go beyond a controlled clinical trial and reflect more of what occurs in the real world.
Leonard: Are we seeing… you and I we're talking earlier… there was a study recently about lipids and if you look at real world outcomes around statins that they only appear to work in 50 percent of patients to get their desired effects. Are we starting to see other discrepancies maybe in your work and other studies you've seen where what happens in the clinical trial doesn't hold up when we start looking at real-world evidence?
Dr. Dicker: In general drugs only work if you take them… with oral drugs, so one of the issues with many medications is the adherence by patients and the adherence is (often) due to the toxicity of the treatment. Sometimes the adherence is due to the financial toxicity. Sometimes the lack of adherence is it's just difficult to remember to take it four times a day and then finally in some patient populations adherence can be complicated by mild cognitive impairment or visual or hand dexterity et cetera and I'm sure there are a few other reasons why adherence is not always superb.
Leonard: So tightly monitored and clinical trials but then when you go to the real world there's all sorts of reasons why somebody might not take the medication.
Dr. Dicker: And then finally there's a great interest nowadays with pharmacogenomics because patients may have a particular genome where they metabolize fast, they metabolize slowly whatever it is but that's another confounder and that may be affected by race, ethnicity, geographic location, social determinants of health, et cetera. So I think there's great interest although there isn't great clarity yet. There's great interest by the FDA and I think there'll be further guidance over time as to what that means and how the pharmaceutical industry or the digital therapeutics industry or the data science industry…how can you use it so that at least in the pharmaceutical world you can expand the label for certain drugs and if there's an unmet need not have to spend significant amount of dollars creating a Phase 3 trial where you might be able to create a synthetic control arm from data mining. In our particular case oncology databases to come up with the kind of what's the representative baseline control rate or response rate or whatever it is to show that you're doing that.
Leonard: What do you see as the biggest hurdles to moving towards a more patient centric more real world aspect to research so we get a little more clarity about how things work in the real world?
Dr. Dicker: In United States in adults there are a few hurdles. I think one is making it easier for people to participate in research. It's interesting if you pay either your genome to get sequenced or you participate by companies like Nebula Genomics or you participate in certain services like 23andMe or you can opt in or opt out even if you buy a Fitbit. Fitbit is running their own trials or their own studies. So all these are examples of how one can lower the bar to participate in research. So every person is like a citizen scientist. So in oncology research there are some opportunities to do that but there aren't a lot of opportunities because right now only a very small percentage of adult oncology patients are participating in clinical trials. I mean it's certainly less than 10 percent in this country, so I think that's one issue. And the other is there's further data to be meaningful it usually requires resources at the provider end and those cost money and that represents an additional burden in addition to collecting information.
Leonard: You mean meaningful in that it's been vetted like if I was a cancer patient and I just started using might Fitbit or my Apple Watch or whatever monitoring tool I wanted that data might not necessarily be interesting on its own unless it was validated by a clinic.
Dr. Dicker: Yeah I don't know if all data has to be “validated” but I'll give you one anecdote. One of the first apps for patients using Apple Research Kit was an app called Parade and this was developed by a woman at GSK, GlaxoSmithKline. Her name is Michelle Crouthamel and this was an app and what they wanted to do in patients with osteoarthritis was use the phone itself. The phone has an accelerometer, a smartphone can measure steps and measure altitude and a variety of other things. They wanted to use the accelerometer feature of the phone to understand the baseline characteristics of patients with osteoarthritis and it was really well designed. They had a very nice short video about why you should do it and in two weeks they were able to enroll over 300 patients or more. I had the pleasure of meeting Michelle in Philadelphia and I said, “wow really impressive app. I went through the first couple of screens but I wasn't going to enrol because I don't have osteoarthritis I don't think I do.” And she said, “thank you” and what she meant by that was I could have enrolled and pretended to be a patient with osteoarthritis just how the app was developed. So I think there are some aspects of real-world evidence that how do you know that the person is participating in some survey or who's contributing or their Fitbit data, do they have the disease that you're looking to study. So there are ways to solve that problem but that little anecdote was just an example of how they tried to, they crowd sourced with a soft launch they got over 300 people or 3000 people I forget exactly how many in two weeks time but they'll never know who really had osteoarthritis and who didn't. They just put it out there. So most people are going to behave honestly and there's no incentive for people to mislead and misrepresent themselves but it just highlights the issues in the field.
Leonard: Yeah, hopefully well to a point probably a long way off where you go into your doctor and you get your diagnosis and you get some kind of digital signature that validates what you have and that belongs to you but it’s probably a long way off. But as you say there's probably solutions to that problem but not the way things are done right now until there's a big impetus, so it might be a while. At the end of the day each of us (maybe is) tracking but each of us manages our own care and that has maybe as big of an impact as you suggested, taking your medication (on schedule) for instance has big of an impact as any treatment (that’s developed). So some of that's behavioural science and some of that economics as you mentioned (financial toxicity) but what are you looking at in terms of behaviour science and how that integrates with clinical care and with digital medicine?
Dr. Dicker: So I guess the older I get the more I appreciate that behavioural science and behavioural economics has probably a greater impact on my own health and wellness than anything else. There's a great podcast by Katy Milkman of Wharton called Choiceology and it has a slightly financial focus looking at behavioural economics and how people make decisions but behavioural economics is fascinating in terms of how it frames the way we look at things for our own care, so I enjoy listening to that. I think if you want to make an impact for patients, most of us behave not as rational beings, I mean that's the work of Nobel Prize …
Leonard: Predictably Irrational.
Dr. Dicker: Yeah and we do most things because of emotion so kind of recognizing that. Actually right before this call I was on the phone with a colleague and we were talking about cancer survivorship and what is important for different folks at different parts of the cancer journey to have different interests and different motivating factors and that's part of the work of, I'm sorry I'm blank out his name but the psychoanalysts who went through the Holocaust and then.
Leonard: Man's Search for Meaning?
Dr. Dicker: Yeah.
Leonard: Yeah, I forgot his name too, oh, yes… Viktor Frankl.
Dr. Dicker: Yeah, Viktor Frankl. So what he recognized in his own personal experience in concentration camp is that if he could identify what was important for a person and had them focus on that that could get them through frequently really pleasant situations and the other key point of his work is that he couldn't control what was being done to him but he could control how he would choose to react to it. So I think all these aspects are critical whenever you're facing adversity and traditional western medicine generally doesn't incorporate any of this when you're taking care of patients because most of the western medicine is focused reactively. If there's a problem, you react to the problem, you fix the problem if it's fixable and then you move on. So I think just to circle back to a point you made a few minutes earlier, I think in the med school course that I teach in digital health and the graduate work that were courses we’re trying to figure out how we could use all these predictive analytics to be more proactive instead of reactive. So I think what a good doctor will do taking care of a cancer patient whatever part of the journey is ask certain questions that may relate to depression or use certain types of survey instruments to pick up on affect and other things like that and then if there seems to be a trend or if the feedback suggests the person is depressed then look more into that and can use technology to try to pick up on that, can you sentiment analysis to pick up on texts and email and other things that might pick up on depression earlier. There are some companies working in that space. So I think that's where you can get to be potentially more proactive than reactive.
Leonard: I know from Dr. Eric Topol mentioned in his (most recent) book he had been in pain from his (knee) surgery where they’re recommending the wrong kind of physical therapy for him (and gave him a lot of ongoing pain) and made him depressed. There's another article recently about high deductibles were forcing people to delay treatments and that would have an impact and there's so many things that can happen to you that can drive depression-like symptoms, not because of depression but also because things are bad especially when you're in oncology and other kinds of conditions where you know the prognosis may not be that great. So hopefully we can differentiate between those as well.
Dr. Dicker: I think the uncertain factor is a big one in oncology and not knowing and having this kind of cloud over your head, I mean those are biggies.
Leonard: Well thank you so much, it's been a really interesting conversation. Thank you so much for the work that you're doing and bring in really more of a patient-centric view into cancer care and cancer treatment. It's great work.
Dr. Dicker: Well thank you. It's been my pleasure.