Conversation with Elizabeth Jaffee: An Immuno-Oncology Update

May 21, 2018, by Jim Hartley

Can you give us an overview of the application of immunology to cancer, the history and where we are in 2018?

Sure. Cancer immunotherapy is really the result of 30 to 40 years of basic immunology science. In the 1980s we had T cells being identified and the ability to identify cytokines, that was a major area of progress that went on into the 1990s.  Viral diseases became an area of high inquiry during the 90s, and we were learning a lot about T cell recognition of viruses and how an antigen is processed and presented to a T cell.  Ralph Steinman was working on dendritic cells, which at first we thought were activating T cells but later realized they were regulating them. Then in the 1990s people like Jim Allison and Lieping Chen were identifying regulatory signals on T cells, realizing that T cells are regulated by a number of different signals, CTLA4 and B7 1 and 2, and now we recognize that there are a panel of important T cell regulatory checkpoints. I think the early ones were real efforts in science to try to isolate and sequence, but with the development of technologies that were able to sequence the human genome, those were also applied to immunology.  So during the early 2000s we were isolating all of the different checkpoint-activating and down-regulating signals on the T cell that interact with dendritic cells and in some cases tumors. That really allowed us to develop these antibodies that can alter the function of T cells through these signals. And in 2011 anti-CTLA4 was the first to be approved and we've had now over 30 FDA approvals for multiple indications in different cancers.  We would have expected melanoma being the first but what we weren't expecting was lung cancer and cancers such as head and neck and bladder and some of the others that were just not even on our radar to respond to immunotherapies.  We were shocked.

And then what happened was a happy marriage between genetics, genomics, and immunotherapy, where we started realizing that current responses had to do with the mutation burden in tumors. We went from this concept of oncogenes regulating cancers to this concept that these non-synonymous mutations, which probably don't necessarily have anything to do with the cancer growth and progression, can be recognized by the immune system and used to kill cancer cells. It has turned out that being able to readily sequence in 2-3 weeks any tumor to determine the mutation burden predicts better than any other biomarker whether a patient will respond to our current immunotherapy.

And that brings us up to 2018, where the biggest challenges now are to figure out how to get beyond the low hanging fruit, cancers with a high mutational burden.  It's not that other cancers don't have mutational burdens, they do, it's just that we're talking about 10 - 100 fold differences in burdens that turn the cancer cell into something like a virally infected cell. How can we activate T cells against tumors that aren't naturally responsive?  Clearly we need more science to better understand how to turn a low-mutation burden tumor into one that looks like a high-mutation burden tumor.

It sounds as if you are convinced that sequencing tumor DNA will become a part of the standard of care for many cancer types.

We definitely think that precision medicine is really what we're doing here. And in fact, one of the approvals of immunotherapy and the anti-PD1 antibody was for genetically associated cancers, the microsatellite instability cancers.  And that is really a tissue-agnostic disease now, we're talking about many different cancers that occur as a result of the same genetic problem. We are now moving to smaller populations of patients that have specific genetic alterations that make them susceptible to immunotherapy.  We're just starting but I think over the next 5-10 years we're going to subtype cancers based on their genetics and on their inflammatory and stromal microenvironment. And as new drugs become available, immunotherapies and non-immunotherapy targeted agents, we'll be able to think more about patients in smaller categories that are determined by their tumor makeup. That's where I think we're going.  In fact we already have clinical trials designed with this in mind, the umbrella trials that enroll patients based on the genetics of their cancers rather than their cancers' sites of origin.  More and more we're going to be able to do smaller, more efficacious clinical trials and offer patients a higher chance of responding to new drugs. And we're going to get approvals faster. The MSI-High approval, from start to finish, from first patient enrolled to approval was three years.  When did we ever have that rapid a result in cancer medicine? As clinical researchers, as oncologists, we're excited about the science, but the real excitement is that we're really helping patients.

Doesn't that raise a difficult issue?  Suppose I'm a patient that falls outside of this MSI category, or whatever the criteria of the moment are. What do you say to me? Even if the chances are low, I want to have some chance.

That's a tough question, because the overall response rate is still only 20%. We have the tools to really understand what's happening in different tumors, and now we need to do the science. There are many good clinical trials where we're looking at combinations now, not just combinations of the existing couple of agents that are already approved, but combinations with stromal targeting agents, combinations with targeted agents, so I think that's where we're headed.  And I would say to you, please find a clinical trial that has some chance, because who thought the MSI cases would have a chance 5 years ago?

A cancer that was full of hundreds or thousands of mutations was the worst news of all, wasn't it?

Yes! And suddenly we took patients who were resistant to chemotherapy, with metastatic disease going home to hospice, and turned them around to where they're living normal lives.  About 65 - 70% who got this treatment responded. That's remarkable.  So I hope that patients today can at least feel there is more hope than there was even ten years ago. And going on an early phase trial today is very different from going on an early phase trial even ten years ago.

The combination of pembrolizumab and chemo is widely used now. The chemo is killing cancer cells but it is also making their antigens more accessible, or educating T-cells, is that correct?

We don't really know the full mechanism.  The hypothesis is that the chemotherapy does a couple of things, it reduces the tumor burden quickly, because when you have a large tumor burden, immunotherapy can take a couple of months to see a response.  So that's one thing.  The other thing that chemotherapy will do is, as you said,  we believe it releases antigens, similar to how we think oncolytic viruses work. Chemotherapy may also be releasing toxins, cytokines and chemokines,  which bring in inflammatory cells to clean up the tumor damage, and these inflammatory cells may in turn present antigens to the T cells. So we think that's how it works but to be honest with you, there isn't great data.  What I can tell you is that preclinically there is data for the different chemotherapy agents, and they all have different effects on the immune system.  For instance, we've published on the taxols and they enhance the antigen-processing machinery in dendritic cells.  They're specific for dendritic cells, and probably for a specific population of dendritic cells.  And cyclophosphamide and gemcitibine reduce the suppressive cells, the myeloid-derived suppressor cells in mouse models. There may be many mechanisms, and there are not a lot of studies yet, particularly in humans, showing specifically how chemotherapies affect the immune system. We've always thought, you give chemotherapy and it kills cancers, it stops cells that are proliferating, but we now know they have a lot of effects, and I think it's a whole area that we need to investigate more. I suspect that looking at different dosing, looking at different sequencing, of these agents and what they're targeting would be very helpful for figuring out how to combine these with the newer agents that are being developed and approved as therapies, whether they're immunotherapies or targeted therapies.

What are your thoughts on "local immunotherapy", which harkens back to Coley's toxins.

There are studies that are ongoing injecting adjuvants, very targeted adjuvants, adjuvants that target certain pathways that will secrete chemokines to attract the dendritic cells and the T cells.  For instance, there are STING [stimulator of interferon genes] pathway activators that are being injected into tumors, and there's been some interesting data.  We have a paper I helped with that was published last year, injecting the STING adjuvant into a mouse model of mammary tumors, and we showed that although it had some effect you had to give a systemic vaccine with STING as well to fully eradicated the tumor.  But again, we're early in this area and combining these intratumoral agents with anti-PD1 or anti-CTLA4 would be worth trying.

There are also agents which propose to improve access of immune cells into tumors, and those are going to play a role too, I suppose.

Yes, and this all has to be explored.  We're in an amazing time where we understand a fair bit, but we need to understand more, and we have a lot of agents available to really gain access and attack the tumor.  I tell my students and postdocs "You're so lucky, you're starting out now.  Cancer's not going to look like what it looks like today, you're going to get rid of it, or it's going to be a chronic disease that people don't even think about most of the time as they go about their daily lives."

What do we understand about the emergence of resistance? 

That's an area we're pursuing now, how does a cancer that's responding become resistant, and we think that has to do with T-cell quality and T-cell exhaustion. There is some work now by immunologists like John Wherry and Raffi Ahmed, who are T cell biologists.  They are showing that T cells can get exhausted after an initial positive reaction to anti-PD1.  So that's one area of resistance that develops after sensitivity to these agents.  You can have loss of the signaling or loss of the HLA molecule that presents the antigens, and tumors escape that way from the treatment that they're getting. These are examples that have already been shown in patients, particularly in melanoma. But melanoma is different from other cancers and other mechanisms may be at play in other cancers.

If I were a cancer patient and not scientifically sophisticated, how would I find the best care, or a good clinical trial?

Very important question.  I was one of the co-chairs of the Blue Ribbon Panel and worked on the Moon Shot and continue to work with the Biden Initiative, and one area that the Moonshot under Vice President Biden's leadership was to try to bring all the communities together to have an easy access for patients, to help them figure out what the clinical trials were, and even be able to access clinical trials based on their genetic information and other information. That's a work in progress, there are some pilots studies for that, but in the meantime, it's very hard with the lists [of clinical trials] that are available in different places. If you have a cancer that has a foundation, they can tell you what priority clinical trials are available, and I always encourage people to contact these foundations. The American Association for Cancer Research will also be helpful in connecting patients to the right places.  Having a National Cancer Institute-designated cancer center in your area, I would always look there, they always have lists and those are going to be the most scientifically rational studies. 

Congratulations on your election as the president of AACR.

Thank you. It's an opportunity to really help in ways that I could not help before, so this really cool.

Would you like to mention any of your priorities?

One of the nice things about being a president of the AACR is that they provide an amazing number of resources that allow you to prioritize a few areas. One area I will focus on is enhancing convergence science.  This is an emerging area and we need to recruit young investigators who are interested in computational biology, mathematics, engineering, machine learning, artificial intelligence, to come think about cancer biology problems. We are thinking about ways to increase the workforce but also to recruit and enhance the more established work force in these areas that weren’t thinking about cancer.  We're going to have a think tank of leaders in the different areas at AACR to come up with a five year plan to to provide grant money and develop symposia and workshops to enhance this area. I've already been talking with people in tech fields and at foundations, people are very interested in working with us on this initiative.  It goes right with all the big data issues we have, all this big data we need to be able to have the technologies and the expertise to put them into biological frameworks.

Another area I want to continue is what Mike Caligiuri has started, looking scientifically at how we enhance our understanding of health disparities by this 20-20 initiative, bringing in 2020 African American patient cancers into the genetic databases, data that will also go into the Commons database at the NIH.  I'm also hoping to come up with ways to enhance the diversity of the cancer work force.  Again, I've been talking with pharmaceutical companies to see if we can provide mechanisms to recruit young people of diverse backgrounds into cancer research. I think it's time we do a bit more than what we've already been doing. 

Another developing issue is cancer survivorship.  With all these great treatments I think we need survivorship research at a higher level. So we're also organizing at the AACR to figure out a 5 year plan of how to support more research in that area, implement it and get it out to people who need it.

Can you clarify what you mean by "survivorship"?

As patients are getting therapies they're going to have long term consequences from those therapies, even immunotherapies. We need to figure out how to ameliorate those symptoms so that their quality of life is excellent. Also, we need to understand the best ways to integrate them back into the normal health care system.  With deadly cancers you stay with your oncologist, and you forget about heart disease and hypertension and all the other things that normal healthy people get that are less of a priority with cancer patients. But with patients living years and decades we need to figure out what are the special issues that may change how these non-cancer diseases develop, they may develop earlier, they may develop in different ways, how do we predict them, how do we treat them, how do we prevent them. We need research in this area so that people who undergo cancer treatment and now become cured or in long term remission don't get consequences that are non-cancers.

I have read that cardiotoxicity is an issue, with both chemotherapy and also with targeted therapies.

That's something we want to prevent in the future. We don't want to overtreat, and we don't want to give the wrong treatments if you're predisposed.  As treatments become more diverse we need to do more understand their long-term consequences.