Personal reflections on World ADC, San Diego

San Diego in October is a delightful time and place to spend time outside, in the sunshine. But for the 1,300 or so attendees at World ADC in mid-October, the only place to be was inside the walls of the Town and Country Resort, where the latest advances in ADC discovery and development were presented, debated, and reflected upon.

Whilst much in the ADC space still remains true to the established “folklore” pathway of “identify a disease-relevant target, then attach a cytotoxic chemotherapeutic or a Topoisomerase I inhibitor,” seeds of dissent and ideas of alternate approaches were being sown, with novel discovery pathways being presented and some difficult questions being asked.

Key to the underlying debate was the recognition that, even after four decades, we still do not have a good understanding of what really makes a good ADC, and how to best deliver patient benefit. In the opening plenary, Suzanne Everdettal (Amgen) eloquently articulated that our approach tends to distill down to “ADCs are dosed, cells die,” with little real comprehension of the chain of events in between. The outcome of this is that, when our ADCs fail in the clinic (as most of them do), we have no real understanding of the underlying cause, nor how to remedy the situation. Rather, we try again, tweak the ADC, but without the same fundamental understanding of the pharmacological audit trail that we apply to small molecule, systemic therapeutics.

On several occasions, this mechanistic disconnect was highlighted as one of the key areas to be addressed as the field moves forward. This highly reductionist approach fails to account for the multiple complexities of internalisation, lysosomal degradation, payload escape, target engagement, and downstream phenotypic effect, making dissection of failure even more complex.

Key causes of failure in the clinic

Beyond this, it was also made clear that our models of disease remain inadequate. As I overheard one attendee state in passing “the model is not the disease” – a concept that we often overlook when trying to predict efficacy and tolerability. In one panel debate, Jay Harper, director of target discovery at AstraZeneca, reiterated the point, noting that “preclinical effects do not guarantee clinical success – identifying the cause of failure is the priority.” Indeed, failure to adequately predict therapeutic window, or resistance emerging from disease heterogeneity, remain key causes of failure in the clinic, and remain unaccounted for in many of our models. In response, Sharsti Sandall (VP, ADC Discovery, Pfizer) made the bold assertion that accelerated clinical entry is the most effective way to demonstrate efficacy and tolerability. That’s not to say we should rush haphazardly to the clinic, but that we can over-rely on models to the detriment of asking the key question: Do we deliver patient benefit? By learning from translational models, selecting the best candidate and then asking the clinical questions as expediently as possible, and being prepared to fail fast, reverse translational studies can help us elucidate where we need to improve and accelerate the path to better therapies.

One argument that resurfaced several times was that our ADCs would be improved if we could simply deliver more to the intended disease site. On average, only 1-2% of the total dose localises in the tumour – which is rather woeful for a modality frequently described as a targeted ‘magic bullet’. Yet, we often fail to consider the fate of the other 98-99% of the dose. Counter-arguments raged – it was asserted that this systemic exposure is similar for orally dosed systemic small molecule therapies, but we don’t concern ourselves with that distribution. Whilst true, this opinion was frequently counteracted with the observation that small molecule therapeutics rarely resided in the system for several weeks, and thus long-term tissue accumulation was much less of a concern than it was with an ADC, particularly one bearing a potent, non-specific cytotoxic that accumulated largely in a single target organ such as the liver.

Perhaps this argues towards an alternate pharmacodynamic approach, such as the shorter-lived but still targeted agents like the bicyclic peptides from Bicycle Therapeutics, which “get in, hit hard, and disappear,” potentially committing a diseased cell to death, without the prolonged systemic exposure.

In pursuit of novel ADCs

In contrast, with more targeted payloads that lessen the impact on healthy tissues, perhaps some of these challenges can be circumvented. Elegant novel payload stories from AstraZeneca (PI3K) and Genentech (Mcl-1) point towards an increased excitement in this space, and we saw the first set of proposed medicinal chemistry guidelines for developing these ADC payloads disclosed at the meeting – an ADC equivalent of Lipinski’s Rules, perhaps? These disease-specific payloads tend to follow a more traditional discovery pathway, with robust target engagement studies leading to a better mechanistic understanding of what occurs after ADC internalisation.

As a long-standing small molecule medicinal chemist now involved in the ADC space, it’s heartening to see the lessons learnt in the small molecule space now being employed in the pursuit of novel ADCs. And their application is timely. As Mohit Trikha, CEO of Kivu BioScience, frequently (and quite rightly) reminded us, we owe it to patients to learn and apply these lessons to deliver “kinder, safer, more efficacious therapeutics” into the clinic.

The path to clinic is complex, and our understanding remains limited. But this year at World ADC, it felt that we are realising that these complexities need to be unravelled and understood, if we are to improve our ADC therapeutics. To paraphrase Bill Clinton, “It’s all about the biology, stupid!”

About the author

Allan Jordan, PhD, is VP, oncology drug discovery, at Sygnature Discovery. With 25+ years’ experience in medicinal chemistry and drug discovery, he has actively participated in almost 60 drug discovery projects, resulting in the delivery of fifteen pre-clinical candidates, of which ten have entered human clinical trials. Jordan has also co-authored over 60 scientific publications, four book chapters, and is a named inventor on 22 patents related to drug discovery. Jordan is an honorary Senior Lecturer in Medicinal Chemistry at the University of Manchester and remains passionate about improving science communication and research engagement, particularly with early career scientists and the general public. He has a BSc (Hons) in Chemistry and a PhD in Organic and Medicinal Chemistry from UMIST / The University of Manchester.