Biotech can be a very risky business, but embracing the risk can be the key to success. At our last Labiotech Refresh in Heidelberg, we got to talk to Friedrich von Bohlen, scientist, entrepreneur and investor, about how he approaches risk and what he thinks about the future of biotech.
Friedrich von Bohlen is the founder and now the CEO of Molecular Health, a company mining data to help the development of personalized medicine. Back in 2005 he met Dietmar Hopp, who is well-known for the fortune he made as co-founder of SAP. Thanks to that chance meeting, von Bohlen became one of the founders and Managing Partner of Dievini, a firm that invests in key areas of the life sciences.
“I’d say I’m not a typical investor. I’m more an entrepreneur who really likes to found and build companies,” von Bohlen said. “For me it’s more important and exciting to find the right people and create a team that really believes in what it does and really goes the extra mile to do what no one has done before.”
Taking risks to maximize success
The success of Dievini stems from daring to take huge investment risks. “Biotech is very dangerous, everything can go wrong,” said von Bohlen. “We can do what everyone does, which is already very dangerous. Or we can do what no one does, which is even more dangerous, but if it works, it’s even better.”
Following this model, Dievini has a strategy of making big investments in just a few companies. “If we believe in the company we do as much as we can,” he explained. “This is why on average we invest about €100M in per company, so the risk is very high but if it goes well the success is high.”
CureVac is a great example. This German company, a pioneer in mRNA technology, is one of the rare private European biotechs valued at over €1Bn. According to von Bohlen, Dievini owns around 80% of the company.
The investment firm focuses on several key areas where it can fully exploit its high risk-high reward strategy, including precision medicine, neurodegenerative and immuno-oncology.
Today these are hot biotech topics, especially with the huge market immuno-oncology has created. However, they were not obvious investment choices when Dievini started, as von Bohlen remembers from a conversation back in 2005 with the President of the German Cancer Research Center, or DKFZ.
“I told him about immuno-oncology, and he put his hand on my shoulder and said, ‘Friedrich, I wish you good luck with all the other companies, because immuno-oncology won’t work.’”
The strategy has worked really well so far for Dievini, as only 5 of the 15 companies it has invested in have failed — a really good performance for any investor specialized in the life sciences.
The next big thing in life sciences
Of the many areas of biotech that von Bohlen finds promising — gene therapy, gene editing, cell therapy, mRNA… — precision medicine is the one that he believes could have most impact. Especially as it can be applied to all sorts of different areas of medicine.
Much like with an investment, running clinical trials comes with a high risk of failure. The success rate of a drug passing all the trials required to win FDA approval is around 15% and it is almost impossible to predict which trials will succeed. That’s where big data could make a big difference.
“There is no small data anymore. Looking forward there is only bigger data,” von Bohlen said. But while in other industries the challenge is mainly having enough capacity for larger volumes of data, in healthcare a big challenge is that there is still a lot of information missing. These knowledge gaps that can lead researchers and doctors to make mistakes.
In addition, a lot of the data available is not accurate — von Bohlen estimates that between 60% and 80% of scientific publications are wrong, and not necessarily intentionally.
It is also not enough to just collect information. To make use of it, the data needs to be processed and structured properly. That is clearly exemplified in the case of IBM Watson, an artificial platform that failed to meet huge expectations that it would prescribe the right treatment for each patient with cancer.
“IBM Watson works for what it’s supposed to work,” said von Bohlen. “IBM Watson is a capturing technology, but then the work starts. After you capture all the data, you have to clean it, sort it and understand it.”
That is a problem that his company, Molecular Health, aims to tackle. Using artificial intelligence, the company aims to predict the likelihood of success of a drug in clinical trials. “We collect, clean and contextualize data from molecular, clinical, medical and drug sources,” he explained. “What good part of our energy goes into is not big data, but clean data. Clean data allows you to still do mistakes, but less mistakes.”
Another challenge, especially in the pharma industry, is that it’s difficult to extract information out of data that is stored in vertical silos. “If you have a limited data set horizontally you get mileage out even big vertical silos don’t get,” von Bohlen said.
Using this approach, the company is successful in its predictions 80% of the time, compared to around 55% for current alternative tools. “If you take a portfolio of 10 and you are 8 times right and you are a pharma company, that is great. If you are an investor, you are rich,” joked von Bohlen.
Preparing for the future
As we venture into the future, von Bohlen believes data security is an issue we need to start paying more attention to. He recalled the case of Flatiron Health, a US-based company that collected oncology data in the US. Last year, Roche acquired the company, which gave the big pharma access to the data of people that did not have a say in the decision. In Europe that would never have been possible.
“Data is the new currency, not only in healthcare,” von Bohlen said. “Since patients never see money if a hospital sells their data to a commercial organization, there must be other reward systems for the patient… It has to be a circle where the individual who allows to use their data gets a reward.”
He sees a risk in new generations not being fully aware of the value and the risk of sharing their data. But risk should be embraced. The sharing of data is not something that should or can be stopped, as it is inevitably part of future progress. What we need to do instead is to come up with a system that deals with and makes the most of it.
Images via Shutterstock