Life science industrials are filling gaps in biopharma by providing all the critical tools, disruptive technologies, and services needed to advance programs from drug discovery through biomanufacturing. Reinhard Vogt and Sebastien Latapie highlight opportunities to improve manufacturing processes and explain how Dynamk Capital is selectively financing start-ups, realizing the promise of Industry 4.0, and championing accessibility.
David Alvaro (DA): What are the most critical gaps in the industry that require innovation to advance more quickly?
Sebastien Latapie (SL): There are two areas that demand the industry’s attention. Principally, existing processes need to be improved, whether for vaccines, biosimilars, or any modality. There are also gaps in technology and a need for innovation related to new therapeutics, such as cell and gene therapy and for protein expression.
DA: How has COVID-19 impacted the industry, especially in terms of supplying key resources?
SL: COVID was a stressor in many ways, and it caused research to pause. However, the need to refocus resources during the pandemic also highlighted the potential of new modalities and approaches, which has generated excitement and brought attention to the critical need for life science industrials.
DA: What’s the clearest path forward for improving processes for mAbs and biosimilars?
Reinhard Vogt (RV): AI technologies are playing a crucial role in making processes more effective and efficient. Especially in the clinical trial space — failure in later phases is hugely expensive. AI can be applied to all phases, beginning with process optimization. These intelligence technologies are eliminating much of the guesswork in development; if there’s an incorrect element, software exists that can either help fix it or confirm in advance that it won’t work on an industrial level.
SL: Process analytical technologies bring manufacturing analytics online, so information is captured without human interference. There’s a need for that in the market, and a few companies are working on it, but a lot of what we’re seeing are single-point solutions that only work at the bioreactor level, for instance. Taking a holistic approach to the Industry 4.0 model is the holy grail for optimizing and improving processes.
This is key for cell and gene therapies in terms of quality control and release testing, as these processes are extremely difficult (especially for autologous cell therapies) via available technologies. We’re constantly exploring how to streamline these processes to reduce the burden on the industry.
RV: When you look at allogenic therapies, it would be ideal to have real-time online measurement of critical quality attributes. The current process is to take samples offline and bring them to the lab for analysis. Ultimately, software is useless without the right sensors to collect the data.
DA: Do you ultimately see a role for continuous processing in biomanufacturing?
RV: Continuous and intensified processing are interesting but only with measurable results — it doesn’t always make sense to change a currently effective process. Going forward, it’s likely that single-use technologies could comprise 90% of future manufacturing so they will need to consider a continuous format.
DA: What comes next?
RV: The end goal is producing efficacious therapies faster and cheaper. It’s clear that we are on our way, although out-of-control costs are affecting people internationally. Affordability is paramount to accessibility, especially as the population ages.
DA: When you think about the next generation of drug development and manufacturing, how important is precision medicine?
RV: While precision medicine is fantastic, again, it’s cost-prohibitive. It hasn’t reached a stage that allows most people to benefit from it. Another issue stemming from precision-based therapies is that monoclonal antibodies rely on a mammalian cell expression system, which often leads to contaminants, such as viruses. I believe replacing this system with plant-based or microbial technologies could potentially be a breakthrough over the coming years, which is why we have already invested in a company playing in this space.
DA: More tools are always needed for finding new drug candidates aimed at existing targets and to discover new targets, but what else can be done to really make R&D more dynamic?
SL: The cost of drug discovery and development is overwhelming the pipeline. Capturing more information at earlier stages is helping to reduce this burden, leading many companies to leverage machine learning and AI for drug discovery. A lot of companies are active in this space, but I think that Exscientia is one of the pioneers in validating this model. We’ve also observed other tools that allow more information to be captured in a higher-throughput manner.
There are additional gaps further down the line, in the pre-IND stage — two of our portfolio companies focus on this space. Curi Bio replaces animal model testing with human stem cells that are much more representative of what’s actually happening in the human body. It’s a more reliable approach leveraging data analytics to better capture and process those large data sets.
We’ve also invested in Lightspeed Microscopy, which uses an open-top light-sheet microscope to visualize tissues in 3D, instead of the typical thin two layers of obscured cells, allowing connections to be made that were impossible to see before. There’s a lot of computational integration of engineering and biology that’s happening at the discovery stage.
DA: Replacing animal models with tissue models and systems could address both ethical issues and help overcome the translation gap. How do you envision the progression to achieve both of those goals?
SL: iPSCs have unlocked much of that, and being able to differentiate iPSCs into the different tissue types is critical. We’ve seen that a lot of the initial focus tends to be around neurons and cardiac tissues, which has had excellent traction, but there’s still much work to be done with other tissue types. A few companies are working on different ways to do just that, however.
Another focus is looking beyond a single tissue to understand how tissues systematically interact with each other. Emulate comes to mind as the most established player in that space, though there are quite a few others.
Companies are also approaching this from a patient-centric, precision medicine perspective, taking patient samples and using those to create organoids that accurately represent disease models.
DA: Do you feel that the fundamental AI and machine learning technology necessary to enable all of this already exists and it’s a question of figuring out how to layer the biology into these technologies, or is more development needed on the actual technology itself?
SL: We have seen great traction around machine learning algorithms in imaging, which has been more robust and can be leveraged effectively in biology. There’s a massive effort in terms of codifying, cleaning up, and organizing data such that it’s truly useful for these algorithms. The unique data sets that can be acquired and the amount of information needed, however, create a gap in leveraging ML to the fullest scale. Additionally, there’s a talent shortage, which further complicates integrating augmented intelligence.
One of the biggest gaps is the issue of disjointed systems and siloed data. We regularly speak to data scientists and researchers on this topic specifically, and they all explained that, aside from curating and combining data, both groups have to speak the same language to create actionable results.
The data scientists and researchers also emphasized how disjointed the automation platforms and sensors are. The biggest obstacle to the Industry 4.0 vision is harmonization, largely due to issues around communication. Every large-scale strategic LSI supplier has its own proprietary software, and each has a unique communication and automation platform, plus all the sensor companies around it.
RV: Another issue is scalability. If a process succeeds at lab scale, it won’t necessarily work in commercial manufacturing.
SL: There’s a scale-up and scale-down model, and then there’s also a digital twin, which is very interesting.
DA: Fundamental to in silico machine learning and tissue models is additional input needed from upstream, including a better understanding of the biology of these systems to generate the inputs that will strengthen these algorithms. Is there an industrial way to accelerate initial discovery even before thinking about drugs?
RV: A lot of the fundamental research that leads to better understanding of biology happens in the academic setting. Linking the two is critically important to effectively bringing those discoveries to market. There’s definitely a schism between academia and industry, as a lot of potential avenues aren’t being utilized. In my experience, there’s a much stronger link between these institutions in the United States than in Europe, but considerable work remains to be done.
DA: Is there a need to transfer industry tools to academic researchers to ensure that they’re dealing with the right kind of machine learning to perform analogous optimization?
SL: These technologies are designed for industry, but that often comes with an industry price tag. This means that the tools are not always accessible for fundamental basic research. On the flip side, from a business model perspective, industry customers are more attractive. Despite this, having access to that technology at an earlier stage and even in the academic setting can be really valuable to push research forward and unlock new insights that may eventually translate into new approaches and therapies.
At Dynamk, Lucid Scientific falls into that bucket. The company is allowing cell culture monitoring and direct measurement of oxygen consumption, which used to require a device that cost upwards of $200,000. Their sensing lid fits perfectly on a 96-well plate and capture all that information while sitting in the incubator. This is an example of how accessing that information earlier and at a much lower price point generates greater possibilities. While that’s tremendously exciting and encouraging to see, it’s not easy to do.
There has been a momentous push toward collaboration between academia and industry, and a lot of the tech transfer offices in universities are harnessing their energy into making that transition happen. I’ve been privy to discussions around reimagining what the traditional academic career path is, as, historically, those who enter academia remain in academia. There are a lot of bottlenecks associated with rising through the ranks and moving forward in academia, which has led to more drive among academic researchers to launch a company or create a start-up. People in academia are becoming more excited about business prospects and are finding more resources available to pursue them.
DA: In terms of the real source of the innovation needed to drive these life science industrials — is that happening at the bigger providers, or is the innovation primarily happening at these small companies that are inevitably absorbed into the larger ones?
RV: In general, 80% of innovation is driven by smaller companies that are then acquired. We strategically invest in these organizations and allow them to use our programs, making Dynamk a cornerstone of this innovation.
SL: It’s analogous to how much of the interesting science and R&D in pharma actually happens at biotechs, and then pharma companies acquire or partner with those companies. Likewise, in the life science industrials, most of the exciting innovation emerges from smaller players.
RV: When you venture outside of the United States, when we look at the companies that are relevant for Dynamk in life science technology, there might be 500 on our radar. Plus in Europe, it’s still not easy for a start-up to raise venture capital.
DA: Can you share a closing thought on Dynamk’s mission and vision and how that differentiation is ideal to enable life science industrials?
RV: Dynamk is a unique fund, because it’s 100% concentrated on the life science industrials segment. We’re helping young companies and start-ups grow. In a short time, the market has appreciated what we have to offer. Dynamk was started in 2017, at that time we had to look for targeted companies. At the moment, we have more incoming inquiries than we can handle — we have to be very selective about the companies that we finance.
A strong network is critical in this industry, and Dynamk helps facilitate contacts, adding credibility to these start-ups. I’m very proud that companies like JSR, Lonza, and Pfizer have invested in us and support our vision.