Between 20 years researching leukemia at Harvard and four years directing Merck's oncology program, Gary Gilliland, MD, has seen a lot of ups and even more downs. Now, as the head of precision medicine at Penn, he's fairly optimistic about emerging therapies like immunologics, but just as concerned about how to pay for them.
Late last year, Gilliland joined the University of Pennsylvania Health System in the newly created position of vice dean and vice president for precision medicine. At the same time, a wave of litigation was erupting over patent rights to technology for a promising T-cell leukemia immunotherapy co-developed by Penn, St. Jude and other researchers -- a sign of just how high the stakes are in the quest for innovation.
While taking a long-view and remaining cautiously optimistic, Gilliland sees a bright future in immune system-based therapies and genetic-guided treatments in general, in oncology and other areas.
In his new role, he's trying to synthesize and translate biomedical research into clinical settings. Now, for instance, pretty much every Penn cancer patient can have their malignancies sequenced to see if they're driven by mutations that might respond to any of the dozens of targeted therapies on the market or in the pipeline.
Gilliland envisions personalized applications beyond oncology and genetics -- although it's those two areas primarily where research, treatment and financing for targeted therapies are being tested and will likely come to define how the medical community thinks of personalized medicine in the coming decades. In an interview with Healthcare Payer News not long after moving to Philadelphia, Gilliland shared his thoughts on navigating that journey.
Q. You've been working at the intersection of genomics and cancer for a while. What would say is one of the best examples of success in the field?
A. Oncology, although it's very challenging, is probably one of the more straightforward opportunities for applying personalized care, at least when considering a genetic perspective. It's based on the understanding that non-small cell lung cancer, for example, is 80 to 100 different diseases based on our understanding of complex genetics. The one example in non-small cell lung cancer that I think is worth noting is the development of a tyrosine kinase inhibitor, generic name crizotinib, made by Pfizer. Crizotinib targets a translocated ALK tyrosine kinase called EML 4-ALK that's present in about five percent of patients that have non-small cell lung cancer. The five percent that do have it have a very impressive response to crizotinib.
The treatment costs for crizotinib are around $100,000 per patient per year. Many oncology drugs are priced around that structure. That's clearly a benefit for the 5 percent of patients who have the translocation. But if you calculate the cost of giving crizotinib to all comers of non-small cell lung cancer, you're treating 95 percent of patients who are destined not to respond. If you do the rough math based on the prevalence of non-small cell lung cancer in the United States and the cost of the drug, that's somewhere in the order of $20 billions. Cost avoidance and benefit to patients by not treating patients who are not going to respond is as important as identifying those patients who are going to respond. Other recent examples include a BRAF inhibitor for treating BRAF-mutated melanoma that works in the 60 percent of patients who are BRAF mutated.
Q. The way the federal government approves new therapies has drawn criticism for various reasons, and likewise for the way pharmaceutical companies develop therapies. Do you think the clinical trial needs to be updated, and do you think there'll be more partnerships along the lines of Merck and AstraZeneca's combination drug testing?
A. I think combination therapies will be essential for improving outcomes; we've known that for decades since the '60s when we started treating children with acute lymphoblastic lymphoma. I think in the current climate, which is as challenging for pharmaceutical industry as it is for everyone involved in healthcare, partnering with other companies in a cost-sharing and risk-sharing approach makes a lot of sense.
I would say that the (Food & Drug Administration) has done a very good job of supporting those types of approaches. This is the HIV paradigm, where you start combining drugs once the single agents are registered, understanding that either drug alone may not have adequate activity for registration but the combination might. It is a difficult development path because that also requires a companion diagnostic and you've got a lot of moving parts. The question is whether you have to prove that the combination is better than either monotherapy in a randomized trial where you're pre-selecting patients with a specific genetic subtype. I think that's something we'll have to solve for. It's gratifying to see that many other companies have followed suit since then, looking for strategies to combine at the pre-registration phase rather than trying to develop their own internal programs for doing combinations.
The other part of that paradigm in terms of the interaction with the agency is the increasing importance of consideration for registration paths that might not require randomization, if the new medicine looks to have such a high level of efficacy that it might not even be appropriate to consider having a randomized trial. One example is dabrafenib, the BRAF inhibitor, where a randomized trial was done but was actually stopped early because it was so clear that the patients getting the drug were doing so much better than the patients that were not.
So the agency is willing to consider registration of drugs based on single-arm, non-controlled expansion studies. Another good example of that is the registration of crizotinib made by Pfizer, where they preselected patients that had the ALK translocation. It was a huge amount of work. I think they had to screen in excess of 1,500 patients to get the 80 that were ultimately enrolled in the single-arm trial. But those 80 patients had a staggering effect size, and that trial plus another single-arm trial was adequate dfor the agency to approve it.
Q. The emergence of new molecular diagnostics and even increasingly precise imaging has raised a dilemma in oncology: balancing the benefits of early detection with the costs of treating what may actually be benign malignancies, like the so-called incidentalomas. How do you think health organizations should approach this?
A. I would say that we're actively engaged in bringing that technology into patent. We have active discussions with a number of different parties. I think it's a very promising area of investigation. The reason it's an area of investigation and interest is that we're still looking for the value in terms of clinical decision making and benefit for patients outcomes. We do have a lot of examples of biological markers of disease like the CA125 test used for some assessment of disease burden for ovarian cancer. It turns out not to be a very useful test for things like early detection or even monitoring response to therapy in predicting outcomes. So everything that glitters isn't gold. And I think the burden is on us especially as we think about the value to third-parties to show that doing these tests, which like most are very expensive, has an impact.
And you could break that down into two buckets. One is can you use it to monitor response to therapy? I think that's where the best data is for this technology -- without having to have expensive and frequent high-resolution MRI, for example, to follow a tumor's response. If you can pull a tube of blood from the periphery and purify the circulating tumor cells and show that it's decreasing or increasing, that could be a very helpful marker. The question you raise that I think is more challenging, that I'd be more cautious about, is whether you can use it to make treatment decisions based on early detection or early relapse. And if you find out that you have early relapse from a tumor the next step in standard of care would be to use high-dose cytotoxic chemotherapy, and I think you'd have to think carefully about whether it's warranted and whether it improves outcomes. Those types of studies will be very difficult to do; they'll probably take a lot of patients and long period of time.
Q. 2013 saw something of backlash brewing against the pharmaceutical industry over the high cost of oncology drugs, particularly for the highly-successful chronic myeloid leukemia treatment Gleevec, which is now priced around $100,000 a year, although not every patient is paying that. Do you think those criticisms are fair?
A. I think this is one place where I have perspective, both having been at Dana Farber and at Merck's oncology program. We need to work towards a common goal. I would say the pharmaceutical companies in general are not interested in highway robbery, as some make it sound. I think we're all interested in understanding how to generate a positive return on investment and the investment in R&D in the pharmaceutical industry is for various reasons currently running a little bit less than one. That's not a tractable business model. There's ways we could cut costs on the development path, but for every imatinib (Gleevec) there's literally nine failed drugs.
I think we should focus a lot of attention on the next step, curative therapy. I think the new immunotherapies that are evolving from Merck and from other companies actually have the potential in combination with standard of care cure certain types of cancers. I've been around a long time. I've heard people say that a lot. I've said it myself. But this really looks real to me. The response rate we've seen with immunological checkpoint therapy is stunning.
It's the ultimate personalized care, you're actually taking a patient's own T-cells and reengineering them to their benefit. There's a lot of room for improvement in the therapeutic, but the proof of concept is extraordinary. You see patients who were on the edge of death in remission without evidence of disease.