5 Key Lessons: Boston Breakfast Meeting

In May 2019 we held our first Breakfast Meeting in Boston, Massachusetts.

We had two brilliant speakers in Nadeem Sarwar, PhD (Founder & President, Eisai Center for Genetics Guided Dementia Discovery) and Max Bylesjo, PhD (Technical Director, Fios Genomics). With two differing and complementary talks, the morning allowed attendees to gain insights into the drug discovery and analysis processes from two different commercial perspectives.

From the quality of data collected to using a multi-omics approach, we’ve pulled out our five key lessons from the meeting.

1 – Quality over quantity for collected data

Many times throughout Nadeem’s talk, the emphasis was placed on the quality of data in the drug discovery and development pipeline. He highlighted four key areas through the development process where it was important for collected data to be of high quality:

Identifying specific targets; prioritising therapeutic hypotheses; gaining more information from data gathered (to enrich clinical trials); and selecting the patient populations that are most likely to benefit from a new drug.

In diseases such as oncology, there are different underlying causes. By identifying specific genetic targets, therapeutics can be targeted to a much higher degree. This ensures that the quality of collected data is kept high (as data collected is much more focussed on the specific target), and reduces the amount of data that is collected and unable to be used.

By prioritising the therapeutic hypotheses for a study, it helps to drive resource allocation. While in itself this does not cause high-quality data to be collected, by prioritising research needs this again gives focus to the study and what type of data is collected. Once the hypotheses have been prioritised, this then leads to gaining more information from the data that is gathered due to the focused and quality approach.

Finally, while population sizes for studies are important, the quality of the patients in the study is also as significant. Finding out which patients are most likely to benefit from the therapy on trial allows for stratification in the samples, and gives rise to the ability to select the subsequent patient populations who are most likely to derive benefits from the therapy.

2 – Targets with genetic support are more successful

This isn’t a small increase in success – targets with genetic support are nearly twice as successful in trials. 73% of trials that link the target with the disease indication through genetics are either active or successful, compared with 43% of trials that don’t. By backing up the trials with human genetics, they are more likely to succeed than those without the genetic foundation.

“In exchange for greater odds of success, we must use all tools at our disposal to overcome challenges inherent in genetics-guided drug discovery”

It isn’t an easy route to finding prospective genetic targets. For new therapeutic targets, genes may be poorly characterized and the mechanisms behind physiological processes also unknown. By having firm genetic support in your research, you are able to better predict how a drug will perform, and also potentially why it didn’t perform as expected in patient populations. Without that genetic support, it is almost a case of trial-and-error as to why a specific therapy has not met its targets.

3 – Not all treatments work for all patients

Perhaps this is an unsurprising point, however, it is one that is a key issue for many therapies. Max highlighted in his talk the protein PD-L1; programmed death ligand-1 (PD-L1) is a regulatory molecule expressed in immune cells that can dampen the immune response. A number of diseases exploit this regulatory pathway by expressing PD-L1 on their surface to dampen the immune response and evade the immune system.

There are at least three PD-L1 inhibitors on the market with more in the experimental phase and under development. The key market for PD-L1 inhibitors is oncology, whereby blocking the PD-L1 pathway produces anti-tumorigenic effects in a number of cancer types. Only a subset of patients actually responds to anti-PD-L1 treatment, however, meaning that clinical trials could be adversely affected if those patients are included in a trial. This comes back to Nadeem’s point of selecting the patient populations that are most likely to benefit from a new drug.

Stratification strategies using genetic biomarkers enable those patients who do respond to anti-PD-L1 treatment to be separated and enable future therapies to be trialled on patients that it would have an effect on.

Max spoke specifically about a case study on PD-L1 and therapies surrounding oncology, however, this can also be applied to other diseases and treatments. Stratifying patients through genomic strategies allows for treatments to become more specialised and targeted for those that it would work for.

4 – Proteins alone do not give the full picture

Processes in the human body do not occur in silos. Proteins and proteomic pathways are affected by other areas – to get the entire picture of what occurs when therapies are tested, multi-omics approaches are needed. Again, this complemented a point from Nadeem about gaining more information from gathered data to enrich clinical trials.

By combining data including RNA-Seq and clinical data from patients, a more rounded picture of response is able to be formed. Each ‘omics dataset comes with its own benefits (and drawbacks), and so being able to combine these allows those drawbacks to be overcome as other datasets can fill in the gaps. A patient’s response to a drug is not only shown in RNA-Seq output, nor does the demographic data of the patient give 100% accuracy on how they will respond. Building the full picture of response to a drug allows the results to be used to their full extent.

5 – Keep your promises to attendees

We promised good food and thought-provoking speakers – and we delivered on both. Nadeem and Max spoke on different but complementary topics and in doing so sparked thought-provoking conversations with attendees. It was also a breakfast meeting, and we couldn’t let attendees go without a proper start to their day. We know from our own experiences that taking 4 hours out of your day to attend a meeting can be difficult, so we made sure that it was worthwhile for those who joined us.

Didn’t attend the breakfast meeting? If you’re looking for bioinformatics assistance – from large pharma to small biotechs – take a look at our Services pages or send us an email for more information on how we can help your research.

Further reading

Lessons learned from the fate of AstraZeneca’s drug pipeline: a five-dimensional framework

TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells