Christina Lingham: | Hi everyone. I'm Christina Lingham from the PEGS Europe event. I'm really pleased to have the opportunity to speak with John Lambert who is executive vice president emeritus, and distinguished research fellow at ImmunoGen. He'll be talking about Widening the Therapeutic Index: The Next Generation of ADCs, in the plenary keynote session at the 9th Annual PEGS Europe Summit on Monday, November 13th, in Lisbon, Portugal. John, thank you for joining us. |
John Lambert: | You are very welcome. |
Christina Lingham | Can you outline what you consider to be some of the greatest obstacles to developing ADCs? |
John Lambert: | I think the greatest obstacle in my view, is actually identifying the right targets to go after. I think there has been an idea that the ADC technology is so powerful that it will work with almost any target. I think the experience has proven that that's not so easy as first thought. |
Christina Lingham: | Tell us about your approach and what is unique about it. |
John Lambert: | Actually, I think as we have learned, we think of targets in two ways, from in terms of our technology, it's clear that target density is a surrogate for, in the end, target flux, how much antibody and payload, therefore, you can internalize into a cell is a critical factor. We're looking at targets with extraordinary high antigen density on tumors. For example, our folate receptor alpha compound, and in the past, our collaboration with Genentech, HER2 that is massively overly expressed on certain breast cancers. I think these are the types of targets that the ADC technology is well suited for. |
Christina Lingham: | What have you learned about the difficulties of bringing novel constructs to market? |
John Lambert: | The difficulties bringing constructs to market ... I think one of the main difficulties is understanding the relationship between activity and toxicity in animal models versus the activity and toxicities in humans. I think all too often it appears that compounds that work very well in animals, one, in the end in human clinical trials does not get to the doses that one can get to in animals. I think the relationship between toxicity in animals and then in humans is a very complex one and not easy to model. I think in part because ADCs, the toxic component that brings the toxicity ... there were some elements to it's pharmacokinetics and distribution that are like small molecules, and there are others where it behaves and distributes along with the antibody. I think this relationship is very difficult to model going from animals to humans. I think, to me, that's the biggest difficulty in understanding what constructs to bring forward into the clinic and ultimately be successful through the clinic. |
Christina Lingham: | Do you have advice for people introducing new ADC products in the clinical setting? |
John Lambert: | I think the advice is, at least our approach has been, notwithstanding the difficulties of trying to get at the toxicity side of the equation, spending a lot of time on that, I think, to have ADCs well tolerated at doses where the plasma concentrations can give you anti-tumor activity is a critical thing. We do our best with the tools that we have available before bringing them into the clinic. I would say a second important thing about successfully developing ADCs in the clinic is to select the patients well. I think our experience has taught us that it's really important to already have a prototype assay for patient selection. And what I mean by patient selection, that is, selecting those patients whose tumors expressed the target antigen at high levels. I think this is critical to the success of clinical development. |
Christina Lingham: | What do you consider some of the most exciting new developments in emerging applications for ADCs? |
John Lambert: | Some exciting developments are the fact that there are now emerging, as people have learned that compounds in late-phase clinical trials for oncology applications. I can, again cite our own mirvetuximab soravtansine in platinum resistant ovarian cancer where the phase one and expanded cohort data looked very interesting and we have started a phase three trial. I think, looking beyond oncology, I think, some of the interesting work to use ADCs to get at reservoirs of bacteria that are within cells and thereby evade therapy with antibiotics is an exciting possibility. I think some of the other emerging applications can be to learn to modulate the immune response by damping down or eliminating subsets of immune cells, and more delivering, instead of toxic compounds, immunomodulatory drugs. So, there's some exciting new thoughts there and it's still very early, but I look forward to these new applications of antibody delivery. |
Christina Lingham: | Are you aware of ADCs which target both cancer initiating cells and differentiated cancer cells? |
John Lambert: | Yes. There've been a couple of ADCs that I'm aware of, and there may be more, where the primary rationale for the target was that the targets existed on cancer initiating cells, also sometimes called cancer stem cells. 5T4 was one such target, that Pfizer had an auristatin ADC, though that was discontinued in the clinic. More recently, Stemcentrx AbbVie have taken an ADC to a target DLL3 into the clinic and it's in advanced stage clinical trials now. The payload for this is the pyrrolobenzodiazepine dimer. So, as the phase one work is now published and the compound looks quite interesting. I would say though that the clinical activity neither proves nor disproves the hypothesis that going after cancer initiating cells is actually one way of approaching ADC therapy. The clinical data that has come out suggests that the higher the antigen density, or the higher the distribution of the target on the cancer cells, the better the activity, which suggests that actually overall distribution of payload into the cancer as a whole is also an important factor. It's not clear whether the activity's really the elimination of initiating cells only, or the overall effect on all of the tumor. |
Christina Lingham: | What are the most frequent escape mechanisms tumor cells use to avoid destruction by ADCs? |
John Lambert: | The clinical data for why patients who might respond for a while then stop responding, the clinical data for the reasons for that, a few, experimentally in pre-clinical studies you can identify all sorts of mechanisms for resistance, from loss of target to other things. However, my impression is that one has to remember that ADC therapy is really single-agent chemotherapy, albeit a targeted chemotherapy. When one is treating in, especially phase one populations of cancer patients, there is no treatment of cancer cells with single-agent chemotherapy, where the tumors don't eventually become, sometimes in short order, fairly resistant to payload. So, my sense is, that the most likely escape mechanism is that there already exists in the tumors cells that are more resistant to being killed by the mechanism of action of the payload at the concentrations that can be delivered by the ADC. I suspect it's really clinically mostly resistance to being killed by the payload. |
Christina Lingham: | How have perceptions shifted over the last three or four years on ADCs? |
John Lambert: | Well, I think in the early 2010s, so in 2011 and 2013, when Adcetris and then Kadcyla were approved there was a lot of enthusiasm. Many ADCs were taken into clinical trials using many antibodies to many targets. Now, I think, in the last few years there was some discouragement about a lack of success of some of these ADCs. I think what people learned is that one has to be more thoughtful about what targets to go after. Perhaps what diseases to go after thinking of the mechanism of action of the payload. For example, several targets are going after colorectal cancer using tubulin agent payloads, the maytansinoids or the auristatins. Yet, tubulin agents as a class of payload have generally not been effective in colorectal cancer and one might not expect this to change just because you're delivering the payload or payload of these types by an antibody delivery. So, I think there is now, maybe, a new appreciation of some of these challenges to make sure you're targeting the right payload to the right disease, and more thought given as to what targets to actually go after and how to assess pre-clinical data so that the compounds that enter the clinic are those with a higher probability of success. I think perception has shifted now to really focus on target and target disease to make sure that the ADC's designed appropriately and is being developed appropriately in a space, a disease, that success is more likely. |
Christina Lingham: | One final question today. How can ADCs be combined with immuno-oncology approaches? |
John Lambert: | Well, this is really exciting to me. I think the combination of ADCs and immuno-oncology were stimulated by the publications at the end of 2015 from the group led by Professor Zippelius in Basel (Switzerland) using murine systems, syngeneic murine systems using T-DM1 and immuno-oncology approaches using checkpoint inhibition. So, it does seem as if several of the payloads used in ADCs, payloads themselves, can in fact, activate directly the macro-phage or dendritic cells that are found in tumors that are responsible for taking up antigen that tumors may have. Furthermore, because you're actually affecting some cell death, you are creating a flood of such antigens as the cells die, and so this combination of a flood of antigens and activation of the cells to actually take up these antigens and migrate to lymph nodes and present them to T-cells, can create activated T-cells returning to the tumor. Here's where the checkpoint inhibition can then prevent those T-cells from being blocked by the tumor from attacking them, and so I think this is a possibility for Syngene activity and certainly this has lead in the last year to some ten or a dozen clinical trials with several ADCs in combination with several of the checkpoint inhibitors. I really am excited by the prospects of this turning what might be fairly short-lived responses into long-lived responses, which will be really of great benefit to patients. |
Christina Lingham: | John, thank you so much for your time and insights today. |
John Lambert: | Thank you very much for asking me to talk about this field. |
Christina Lingham: | That was John Lambert, executive vice president emeritus and distinguished research fellow at ImmunoGen. He'll be speaking in the plenary session at the PEGS Europe Summit this November in Lisbon. If you'd like to hear him in person, go to pegsummiteurope.com for registration information and enter the key code, podcast. I'm Christina Lingam. Thank you for listening. |