Reproducibility, Technology Transfer, and Biotech Innovation
The reproducibility problem in biomedical research is well established. Several well publicized studies have shown that as low as 10–15% of discoveries in prominent journals can be reproduced by independent investigators. This problem has brought the discussion of incentives and checks and balances to the forefront, and certainly has implications for the speed at which the scientific community can move forward - ultimately affecting our ability to treat patients. One interesting player in this ecosystem is technology transfer offices (TTOs) in universities, as they are attempting to out-license technology that that is coming under increasing scrutiny from industry veterans who are skeptical the results will be reproduced.
Two articles last week highlighted the plight of technology transfer offices in lieu of this reproducibility crisis. Bruce Booth of Atlas Ventures highlighted that technology transfer offices needed to set reasonable expectations around licensing terms for technology that needs to be reproduced and be willing to grant exclusive options while this work is being done before a full license is taken out. Michael Rosenblatt from Merck went a step further and suggested that Universities be willing to reimburse companies (to the tune of up to $2MM!) who are unable to reproduce research after they have taken out a license.
Rosenblatt’s views are certainly extreme. As David Winwood from the Pennington Biomedical Research Center put it, “Few if any public schools would have either the (financial) capacity or, I suspect, the legal authority, to enter into such an agreement.” However, both authors identify a serious issue with technology licensing in biomedical research: the reproducibility problem leads to challenges in evaluating the potential of technologies for people on both sides of the table, ultimately leading to fewer technologies being licensed, less licensing revenue for TTOs and fewer successful products on the market.
TTOs funding validation studies
The challenges inherent in the reproducibility problem also present an opportunity for TTOs to significantly differentiate and add value to their IP portfolio. Right now, TTOs are likely spending $20–50K to patent a technology once all is said and done, and likely even more if including international filings etc. Furthermore, many TTOs are devoting additional funds to help create an entrepreneurial ecosystem and resources for new ventures that to increase the number of patents making it into industry. If the above cited estimates of reproducibility frequency are correct, the vast majority of the TTOs budget is going towards supporting inventions or technologies that are doomed from the start. It may be worth-while for TTOs to redirect some of their resources towards reproducing key experiments in an invention disclosure to validate the results. This could be done through another academic lab, or through a CRO much more cost effectively than through a large pharma, and would likely cost far less than the $2MM price point that Rosenblatt quotes.
If the above cited estimates of reproducibility frequency are correct, the vast majority of the TTOs budget is going towards supporting inventions or technologies that are doomed from the start.
Taken to the extreme, a TTO could try to reproduce the work BEFORE a patent is even filed for. For inventions in which the results of a key experiment drives the value of the patent, it may be worth it to put that money towards reproducing the results first, then file for the patent if the work is validated. (Similarly, the TTO could file a provisional patent, then reproduce the work over the next 12 months before converting to a full patent application). Given the low rate of successfully validated experiments, this could result in far fewer patents being filed, yet it would significantly increase the value of each one without increasing the budget of the TTO that much.
Certainly, there are a lot of variables at play here (the cost of reproducing the experiments being a major one) and they would not be relevant for every invention disclosure. However, devoting more funds towards validating great science rather than filing for superfluous patents, may be a useful strategic shift in for TTOs to consider. It has the potential help surmount one of the major inefficiencies in biomedical technology commercialization today and create a system where parties on both sides of the table are happier.
