Jim Mulshine is a thoracic medical oncologist by training, who spent 25 years at the National Cancer Institute in Bethesda, Maryland. After a decade researching management and treatment options for advanced non-small cell lung cancer, Mulshine came to the conclusion that the best path to treating the most patients successfully is not necessarily trying to find a cure for late stage lung cancer – which is when the disease is most typically found – but rather trying to determine how best to detect the disease in its early stages when it’s more easily managed. For Mulshine, one answer lies in early detection via a thoracic spiral CT scan. To confirm this, a 2011 trial by the National Lung Screening trial found that using a spiral CT scan to perform three rounds of annual screenings of high-risk populations reduced mortality by 20%. In practice, if the treatment were to be adopted, screening would not be stopped after three annual screens because a heavy smoker’s risk of lung cancer persists. All evidence shows that over time, the mortality reduction would be much more.
Now at Rush University Medical Center, Mulshine joined Bruce Pyenson, a principal and consulting actuary at Milliman, for a Q&A to discuss lung cancer and provide a medical and actuarial take around this alternative form of disease management.
Milliman: Let’s just start laying a little bit of a foundation. If you could talk briefly about lung cancer—when typically is the disease found in patients and what does a typical prognosis and progression look like?
Jim Mulshine: So lung cancer is the most common lethal cancer in the United States as it is throughout the rest of the world. Peak incidence begins in the fifth decade of life and carries over around the eighth decade of life. It’s a disease that is typically diagnosed late. Even today in the United States, it’s associated with an overall mortality rate of 85%. It has a very, very lethal course, and many of the people that are first diagnosed succumb within the first year. And so, from a public health perspective, it’s certainly a very, very big cause of death in relatively young adults.
Bruce Pyenson: Lung cancer is primarily associated with smoking tobacco and kills more men than women, although it kills more women each year than breast, ovarian, and cervical cancer combined. From a numbers standpoint, it’s hard to claim progress against cancer overall unless there’s progress against lung cancer. I’m of the generation of actuaries that had to learn how to construct mortality tables and the mortality difference between early and late stage lung cancer is just so dramatic. And it’s much more dramatic than most of the other common types of cancers.
Milliman: There’s been a lot in the news lately about disease management using immunotherapy - can you tell me a little bit about lung cancer treatments in general?
Jim Mulshine: Well, the first thing is that we are learning that the immunological response is incredibly important in the development of lung cancers. For lung cancer, like other cancers, the tumor emerges opportunistically due to a selective defect in the body’s response to injury. The new generation of emerging immunotherapies work by restoring the full spectrum of the body’s immune response. With smoking, you have a very toxic substance that’s being inhaled into the airways of individuals, and for the period that they smoke, they have extensive chronic injury in their lungs. Their body is desperately trying to heal these airway wounds through various immune mechanisms. Close to 70% of all pharmaceutical development that is happening right now in targeted drug development is happening in lung cancer and there are definite signs of progress. Still while the array of agents that are being tested and the wealth of clinical trial data that is emerging every day is hopeful, consistent curative outcomes are not yet emerging.
Bruce Pyenson: The cure rate for early stage lung cancer is very high. That is, someone with a Stage 1 lung cancer can be cured—that cancer is not going to have a recurrence, or if another lung cancer emerges, CT screening will detect that at an early, curable stage. But we’re still in a pretty sobering situation when it comes to lung cancer treatment because so much lung cancer gets detected at a late stage.
The Swiss Army Knife of medical tests
Milliman: Can you talk a little about CT screening for lung cancer and its benefits?
Jim Mulshine: For individuals who’ve smoked the equivalent of 30-pack years, which could be a pack a day for 30 years or three packs a day for 10 years, when you screen these individuals [with a thoracic CT] some 20 to 30 years later, what you find is that one in a 100 may, per year, be found to have an early lung cancer. And fortunately, the surgical management of early lung cancer has vastly improved. The recovery rate is pretty quick. People are back to full activity within a week. However [with CT screening], you’ll also find 25 people of the 100 who turn out to have significant but asymptomatic chronic obstructive pulmonary disease (COPD), which is one of the top four causes of premature death in our society. And, in addition, about 15 to 20 or so, depending on the population, have asymptomatic but significant coronary artery disease, in which the top 25% of that group have a relative risk of having a major coronary event of about seven-fold the rate in that same age group of general population over the next 12 months. CT screening finds the consequences of this chronic tobacco ingestion in various outcomes that are identifiable by radiologists looking at these screening images.
If you could have one test of a population, in a hypothetical world, to try to get a sense for what’s the distribution of health or disease in that population is – if you had a CT of the chest and you could look at lung cancer, heart disease, COPD, osteoporosis, a variety of other diseases in the chest in which to assess the overall health risk of that population, that would be the one test that I think is going to be the—far and away the most informative of all the tools out there. It’s kind of a very interesting Swiss Army Knife proposition.
There is so much clinical information in these thoracic CTs. We’re using CT scans to reconstruct life in Egypt, Peru, Italy from thousands of years ago; this tool just gives us a portal into medical and biological issues which are just remarkable.
Bruce Pyenson: The costs of treating cancer keep getting higher and higher. I often tell people that very few treatments ultimately save money. When we looked at lung cancer screening, it’s very cost-effective and it doesn’t cost very much, but it’s still an added cost to the system. So the typical ways that people look at costs are cost-per-life-year saved. And when we and others have looked at that for lung cancer [CT] screenings, it’s one of the best deals around.
Jim Mulshine: If you take lung cancer screening and you just apply best practices smoking cessation in the context of that annual screening and you just use the best standard approach (which turns out to be one of the most cost-effective measures that we have in our overall medical armamentarium), overall that reduces the cost of screening by about a third. So whatever Bruce has just said about the cost effectiveness of screening now, if we just implement our smoking cessation best practices, we drive it down even further. So this is a remarkably robust public health approach that is going to get traction and is going to really, hopefully, improve our overall health in a very measurable kind of way.
Milliman: The thing that I think about is breast cancer screening, mammograms. Every woman gets a mammogram at a certain age. Do you envision CT screening ever getting to a point where it would be routinized like that?
Jim Mulshine: Well, the case for routinizing lung cancer screening in a high-risk population is much stronger than the case for routinizing breast cancer screening (even in women), although breast cancer screening does work and it can be highly effective. But, yes, absolutely. I think, you know, across the board-- Bruce is also very involved in studying colon cancer screening. And if you look at the adoption of colorectal cancer screening, which is a much less politically charged area than smoking—we’re still struggling to get screening to the people who need it most. So, yeah, breast cancer, colon cancer, lung cancer screening, some of the other screenings that are coming along are much more capable than tools we’ve had in the past.
Also, we are talking about annual, fairly regular screening in older populations, but is there a role 10 years from now or more for some kind of sentinel analysis of 40-year-olds to look and see exactly how their body is working and is there something they should know about that? I don’t know. But a lung cancer screening—the medical radiation is quite modest. It’s not associated with significant disease consequences. The cost is quite modest. It takes three seconds. We could develop technologies in which it could take even less time. How this powerful visualization technology will be used in the future, I think, is a very interesting question to consider.
Milliman: So what do you envision this conversation looking like five to 10 years from now?
Jim Mulshine: Well, I think we are already seeing it. We’re seeing that drugs are being developed to affect inflammation in people who have statin-resistant problems, coronary artery disease problems. And it turns out when you looked at the long-term consequences of heart-directed drugs that shut down that aspect of the inflammatory response, there was a 40% to 45% reduction in lung cancer. And so, you’re going to see people starting to think about this host response to this incredibly toxic substance in the lungs in a much more systematic way and in a much more public health-oriented way to just improve overall health outcomes. And I think that’s, you know, going to be very interesting, because things that modulate inflammation have impact across the board-- lung cancer, COPD, Alzheimer’s—and we’re going to start understanding how we can modify our lifestyle and our diets and our other pharmacological interventions to short circuit the chronic disease consequences that have been so plaguing our society.
Bruce Pyenson: CT imaging is digital and lends itself to the revolution in digital medicine that’s not accessible to many other areas of medicine. Because CT screening can identify lung cancer and COPD and cardiac conditions as they emerge over years and it is digitized, I see imaging as leading the way in computerized diagnostics, machine learning, and other advanced methods. We have the basis not only for some very significant improvements with public health areas, but also a hint at what the future of medicine might look like.
Jim Mulshine: You know there’s just a real reluctance on many people’s part, especially people who are not as aware of how these technologies are moving forwards. And that is that, you know, “I don’t want to know I have Alzheimer’s because there’s nothing they can do about it. So why should I look for it?” You know, “It’s not going to bother me, I’m not going to bother it,” type of thing. Whereas the digital technologies that you’re talking about are finding asymptomatic disease, you know, maybe two years, maybe five years, maybe 10 years before it will come to clinical attention and at that point in time, we actually have some very meaningful things that we can do for these people that could alter that natural history, such that they don’t have to be doomed to that problem. And some of it may be as simple as lifestyle modification: diet, exercise, a more active lifestyle, it could be the getting a statin therapy, it could be a variety of other anti-inflammatory strategies that kind of come forward. But these diseases of lifestyle, you know, these major chronic diseases we think are amenable to relatively inexpensive and relatively non-dramatic interventions that if applied early enough are going to have a lot of people avoid very pernicious diseases.
From that perspective, it should be a different dynamic than people have dreaded from the kind of early detection technologies we’ve had in the past where it’s kind of like the light on your dashboard that goes off just a second before your engine blows up. And we, hopefully, kind of have to educate the public out of that dynamic and see if we can use this as a much more enlightened way to just allow people to be a lot healthier in terms of their lifestyles and their ability to do their daily living.
Bruce Pyenson: Ultimately, I think we’re heading into a different world of diagnostics and hence a different world of treatment. It will improve outcomes and change the entire healthcare industry.