Changing the Nature of the Game

What motivated your career in pathology?

I knew all along that I wanted to do mechanistic medical research. I was sitting in medical school, thinking, “Okay, what’s my quickest route back into the lab?” so I asked my mentors, who were doing lab-based work and patient care, “How can I do that?” They said, “Don’t. Go into pathology.”

Although choosing pathology meant I gave up seeing patients, I never gave up interacting – I see the doctors and their patients as “my patients”, which I find just as rewarding.

Because the Air Force paid for my medical school, I had to give them time back after I finished my residency – and, in retrospect, that was a phenomenally good thing. The people I saw starting out as junior faculty had to simultaneously get up to speed as clinicians and start building a lab. I don’t know how they did it; both of those things are overwhelming. But in the Air Force, I was in a really good clinical setting that let me come up to speed as a practicing diagnostician – so by the time I started my first academic job at the University of Rochester, my clinical chops were fine and I could really focus on building my lab.

I arrived at the University of Rochester fairly pluripotent as a pathologist; I could have developed in a number of different ways. I wanted to join a group that I could learn a lot from, so I re-tooled and became an immunologist. Immunologists and pathologists speak different languages, and by virtue of having a foot in each world, I was able to translate things better. It set me on the path of understanding what I call immune subversion – that is, how tumors block the immune system.

Why the move to pharma?

When I was writing my first R01 grant, I got some critiques that forced me to build a drug for the very first time. It was a polymeric drug delivery vehicle to get my molecule to my target cells, so I started working with a friend in biomaterials, and at that point, I got the bug. I think that was probably the seed of the demise of my academic lifespan, because then I wanted to take what I was discovering and apply it to actually making a drug.

Personal reasons took me to the Bay Area and, out of the blue, a fantastic job emerged at DNAX, which is legendary in immunology. In my mind, the place was a scientific utopia.  Schering-Plough, amazingly, funded this research institute and never really put much pressure on the drug development aspect – so for almost 20 years, it produced great science. It’s formed so much of what we know in immunology today. And when they decided to transform DNAX into a drug discovery enterprise, they brought in Dr. John Curnutte from Genentech, who was a real proponent of translational medicine.  He felt they needed pathology to really understand the biology and mechanisms of disease at the level of tissue architecture and cellular organization. I took the job in a heartbeat and I’ve been happier than a pig in slop ever since.

What was the story of PD-1?

When I was at DNAX (then Schering-Plough Biopharma), the PD-1 program came to us through the acquisition of Organon. My understanding was that Organon was preparing to move forward in development (i.e., humanization of the monoclonal antibody and Investigational New Drug application) without a “companion” mouse antibody and mouse-based IND enabling studies.  This was a pretty bold, forward-thinking approach.  Although it was clear from the literature that anti-PD-1 was a strong candidate molecule for immuno-oncology, Schering-Plough took a more conservative approach and we were tasked with building the mouse surrogate program. The immunohistochemistry antibody we developed is still moving forward as a companion diagnostic at Merck; it’s a brilliant antibody, one of the best out there. But when Merck and Schering-Plough merged in 2009, the PD-1 program was deprioritized, at least until Bristol-Myers Squibb published Phase I data with their candidate. Then it was like Lazarus – raised from the dead! It’s amazing that Merck still got the first approval in the US. I think they benefited from going after refractory melanoma patients as their main indication; that triggered the breakthrough therapy designation.

It’s exciting that we have such a good idea who responds to anti-PD-1 and who doesn’t. That’s critical to why PD-1 development is going so fast – in large part, we understand the mechanism of action, so we can come up with a rational understanding of how to use it and what to pair it with in combination therapy.

My first a-ha moment came when I saw tumors IHC-stained with PD-1 and PD-L1. Patients who respond to anti-PD-1 have cytotoxic T cells in their tumors, and you only need immunology 101 to say, “Wow! The T cell coming in is generating a cytokine,” which upregulates PD-L1 to shut off the T cells. It’s a homeostatic mechanism we evolved to avoid excessive immune reactions. Every reaction contains its own braking mechanism and tumors hijack these brakes to the immune system.

It took a long time to convince the scientific and medical community that immunotherapy would work.  In fact, it’s been more than 100 years of chasing Dr. Coley’s vision of harnessing immune responses to treat tumors.  If you think about where we are with anti-PD-1 and other immunotherapies today, where might we be if this transformation had happened earlier? Where would patients and industry be if that were the case?

What new treatment strategies hold most potential?

The future is in combination immunotherapies – I predict they’ll become the backbone in many indications. We just need to figure out how to use current targeted agents and chemotherapies judiciously.

I think the most important question that we need to answer in immuno-oncology now is: how do you make PD-1 non-responders into responders? That’s our current strategy at OncoSec – and our primary candidate is intratumoral delivery of IL-12. As a cytokine, it’s a chief driver of anti-tumoral immunity; it turns on all the mechanisms of antigen presentation and processing that tumors try to suppress, so it’s a good choice for targeting PD-1 non-responders and we’re seeing very good systemic anti-tumor response effects in the clinic so far.

Immunotherapies in general are not innocuous. PD-1’s safety profile is pretty good, but when we combine therapies, we’re going to have to be sensitive to synergistic immunotoxicity. That’s a benefit of moving toward a multimodal therapeutic paradigm, including intratumoral therapy – we can harness the treatment efficacy without the systemic exposure and toxicity. IL-12, for instance, used to be given as a recombinant protein drug; it was efficacious, but quite toxic, which is why it wasn’t pursued further. Giving it intratumorally, we don’t see any IL-12 in circulation, so we aren’t seeing the systemic toxic effects that we know systemic IL-12 administration would produce.

Some of the real luminaries in the literature today are beginning to talk about intratumoral therapy. I think we’ll see these therapies come of age in the next decade; Amgen’s T-VEC, a virus that encodes GM-CSF, has met with some success, but there are a lot of different ways you can approach it and I think others will follow suit.

I also think we’ll come back to DNA damage and repair. Tumors are dependent on their ability to mutate; they have to, because that’s how they defeat the immune system and the chemotherapies we throw at them. But this can be their Achilles heel; they become dependent on DNA checkpoints, unlike normal cells, so we have a therapeutic leverage point. I think we’ll see a resurgence of this as a field to explore.

I think one of the benefits of immunotherapy is that you’re not going directly after the tumor, but instead you’re educating the immune system.

How do you see the role of pathologists evolving?

It’s going to be increasingly important for pathologists to perform and interpret companion diagnostic tests in clinical labs. But that gets into a really sticky wicket about companion diagnostics – the FDA guidance seems very arbitrary. For instance, we might use immunohistochemical stains to diagnose melanoma, but then if we want to use the PD-1 or PD-L1 stain as a companion diagnostic for a PD-1 drug, suddenly it has to go through FDA approvals. I know the FDA’s mandate is to protect the public health, and they obviously want to alleviate the risks associated with an unregulated test. Unfortunately, their process was set out to make drug development more expeditious and cost-effective, but in fact, it makes costs higher and takes more time. I think there are cases where it stands in the way of being innovative.

On the research side, it’s clear that we can’t just grind up tumors irrespective of what cells are in there and get an answer. We need to understand what interactions are occurring within the tumor. And those are the sorts of insights that come out of a sensitivity to tissue architecture and cell function; that’s why we need pathologists to further our understanding of tumor behavior, discovering potential mechanisms and research angles. That’s where I’ve spent my entire career. The technology is coming, but you still need the brain behind the scope.

I think we really are at this transformative moment in oncology. We’re no longer just trying to out-poison the tumor; we’re changing the nature of the game, and I’m amazed that the message hasn’t gotten out. People just don’t know that we’re doing that. I think everyone should be as excited as I am – it’s a brilliant time to be in this field!