To follow or not to follow (that is the question)

There are many terms – not all of them complimentary – used to describe drug discovery projects that follow, or partly follow, the path laid by others. These include fast follower, knowledge-based, Metoo, and leapfrog programmes to name but a few, but for this article, the term “follow-on” project will be used.

How does your organisation brand this activity? The subject is taboo within the drug discovery industry; many opt not to discuss it openly, and with good reason. By doing so, they could potentially weaken claims for originality and inventive steps in their own work. Nevertheless, a recent article cited that 59% of new drug approvals came from known chemical matter (https://pubs.acs.org/doi/full/10.1021/acs.jmedchem.3c00521). 

Why we should follow others, some of the time 

Whilst many argue that working on the same pharmacology targets is an inefficient use of resources, different organisations will deliver candidate drugs of differing quality and will solve problems in different ways. As compounds progress to market, these small apparent differences often turn out to be highly significant and ultimately deliver a better product for the patient (best in class). And, if multiple products make it to the market, pricing competition also benefits payers and patients. One feature that all follow-on projects share is that time is always of the essence. By definition, the project is behind at least one competitor, and if the drug is to have a chance of winning a decent market share the competition cannot be allowed to have the field to itself for too prolonged a period. 

A recent example of biotech success in a crowded area comes from Nimbus Therapeutics who sold their TYK2 program to Takeda Pharmaceuticals for $4B in upfront cash. The programme, which has the potential to deliver treatments for multiple immune-mediated diseases was supported by recent Phase IIb data. The expectation is that this potential drug, NDI-034858, will be best in class in a highly competitive field, which includes the likes of Sotyktu from BMS and Pfizer’s less selective Ropsacitinib. It should be noted that there is no reference to these being follow-on programs.  

That said, there have been situations in the past when there have been too many companies on a follow-on approach.  An example is the HCV field in the late 2000s wherein at one stage there were over forty phase-one clinical trials. Clearly, this is a waste of resources and access to the number of patients required for successful clinical trials is rate-limiting.  

Learn more about lead optimization in our latest eBook: ‘Lead optimization: Delivering molecules fit for preclinical development’

It is not that simple 

It may be perceived that delivering a follow-on project merely involves following a roadmap laid down by others, tweaking the molecules here and there to manoeuvre them into IP-free space. In fact, that is rarely the case. Clearly, there is huge de-risking in both medicinal chemistry and biology if others have already trodden a similar path, but first-hand experience tells us that a large proportion of the screening cascade (both biology and DMPK) will need building from scratch. Furthermore, reproducing key literature results often presents a very significant challenge in itself. Synthesising and profiling competitor compounds are useful to a certain point – it can allow teams to find areas where we can improve efficacy or safety. Although, there is always a fine line to tread before such activity could be considered to infringe on others’ intellectual property. Arguably all projects are looking for a better drug candidate (often coined MeBetter), but extensive testing of competitor compounds can be using precious resources. Accessing IND (investigation new drug brochures) where available can save duplication. 

Typical follow-on projects start with published chemical matter, often from patents, and develop a strategy to move the chemistry away from the prior art claims and exemplification. Such strategies could include saturating a ring system, adding nitrogen to the ring system and/or substituting known bioisosteres, this list goes on. However, only a tiny proportion of potential changes succeed in rendering molecules with the desired in vitro pharmacology. It can take more time than expected to find suitable gaps and an up-to-date knowledge of relevant literature is a necessity. Multiple searches across multiple databases are the best way to cover these and reading the patents is essential as many databases can misinterpret claims. 

Alternately, or to complement the above approach, focused or random screening (typically HTS) can add chemical equity to programs, that can, in their own right, provide lead series or offer new inspirations on new designs. Additional screening, for example, fragments and DELs (DNA encoded library) whilst often very useful against challenging targets, is likely to extend lead times and leave the project potentially further behind the competition.  Fragments however, owing to their small size, offer the greatest scope for optimization of physical and pharmacological properties and have the potential to deliver best-in-class molecules. 

With new chemical matter in hand, classical iterative design-make-test processes can rapidly enhance drug-like properties and profiles. As the project, quite deliberately, has ventured into an uncharted chemical landscape, the DMPK (distribution, metabolism and pharmacokinetics) properties of the drug have changed; solving these requires an experienced team of medicinal chemistry and DMPK scientists and is frequently a key bottleneck in projects. 

New technologies, including machine learning/artificial intelligence (ML/AI), can help mine data, and even design new chemical ideas. Follow-on programmes, where significant data already exists, should be ideal for such approaches. However, whilst we should all embrace new technologies, it’s unclear to date whether companies who are wholly focused on these technologies are delivering better drug candidates faster and cheaper, than other organisations. Our experience at Sygnature Discovery is that whilst these design tools can predict molecules with potent in vitro on-target biological activities, there is still a lack of suitable pharmacokinetics and pharmacodynamics data to aid the enhancement of these criteria.  

Now it feels like Schrodinger’s cat  

Highly competitive programs typically lead to an early filing strategy from biotechs, whereas larger companies will typically be more patient, with one eye on the shelf life of their potential product. Whichever strategy is taken, with now first-to file-being recognised internationally as the key timeline, there is a large period, typically up to 18 months from first filling, of uncertainty. During this time projects and teams must essentially cross their fingers that no one else has protected or published work with the same scope. It’s useful to work on multiple series to manage this risk of being gazumped.  

In summary, follow-on drugs bring value to biotech, pharma, payers and patients delivering high-quality medicines and should not be viewed as easy or lacking innovation.