Costs & Considerations of Using CROs as a Biotech Startup

Shout out to the Loyal team, who has taught me most everything I know about working with CROs and who put in the hours to ensure our studies are a success. I wrote this to democratize some of this hard-earned knowledge for current and future biotech founders.

There are two high-level models of biotech companies: lab-based and virtual. Most companies are a mix between the two.

Lab-based is probably what you imagine of a biotech company: people in white coats pipetting fluids of various colors and every so often shouting ‘eureka!’.

Virtual biotechs did remote work before it was cool! Most of the physical labor is contracted out, with the intellectual labor largely done in-house. Virtual biotechs are dependent on Contract Research Organizations - CROs - to do the physical laboratory work that they design and pay for.

What is a CRO?

CROs conduct research at the request and payment of you (the sponsor). They may test your drug in an animal disease model, screen molecules for desired activity in vitro, develop antibodies targeting specific proteins. Some CROs participate in the study design and rationale, while others purely do the work you design and pay them to complete.

There are also CDMOs (Contract Development and Manufacturing Organizations), CMOs (Contract Manufacturing Organizations), and I am sure other CxOs that I haven’t run into yet.

CRO - general research work. This category includes most preclinical development, including but not limited to cell-based and animal research and drug discovery and optimization.
CMO - manufacture the product that you have already developed. Expertise is usually in specific types of product manufacturing and meeting the stringent CMC (chemistry, manufacturing, and control) requirements of the FDA, sometimes including scale-up.
CDMO - has the capacity of both a CMO and limited CRO - they work with you to develop (innovate) the product, and have the capacity to manufacture the final product.

Why We Use CROs

At Loyal, the vast majority of our R&D is done in partnership with CROs (we do retain a couple of benches at a laboratory that has a shared core facility[1]).

Loyal is a clinical-stage biotech company developing drugs to extend dog lifespan and healthspan. Our structure makes us especially amenable to relying on CROs for most of our R&D:

• Our core technology is not a drug discovery platform (e.g., a better way to screen for and validate novel drugs).

• Our formulations are technically challenging, but far from unprecedented.

• We are developing dog drugs, so our IND packages are much smaller and we are able to get into clinical studies (AKA, companion dogs) more quickly.

Thes

Lower upfront costs. The primary reason to work with CROs for most biotech startups is that they allow early companies avoid the expense of building, stocking, and staffing a laboratory. Doing so is expensive, and laborious. CROs allow you to quickly conduct an MVP experiment with little more than a laptop.

Minimize risks and variables. CROs and their scientists are procedural, experienced, and their reputation depends on the successful completion of high quality work. While the people variable is always relevant in the sciences, you minimize many by working with CROs, especially in the early days.

Avoid fixed costs. If Loyal were to get into a tough spot financially, we would be able to turn off most of our R&D spend overnight.

Complex studies. Some of the studies we design at Loyal are simply way too complex to be run in-house, even if we had our own laboratory. CROs are able to flex up and down to each study’s needs, without permanently incurring the cost of am

Buy expertise and platform technology. Many CROs specialize in certain assays, methodologies, platforms. They are experts in their domain and think about their platform all day, every day, and therefore can perform the research much better than you would ever hope to. Sometimes, the only way to access these platforms is via working with the CRO that developed the technology.

Regulatory compliance. If you intend to use certain data to support a claim to the FDA, you need to comply with various regulations. For example, your drug product needs to be manufactured in cGMP (current Good Manufacturing Practice) conditions and your manufacturing site needs to be inspected and approved by the FDA. CROs/CMOs/CDMOs will be experienced in adhering to the relevant regulation. This can go the other way - some sites have a bad reputation with regulatory bodies, information they will likely not readily disclose.

Objectivity. No CRO will fake data for you - your measly contract is not worth their business and reputation. This has been helpful in investor diligence.

Considerations and Challenges of Working with CROs

Innovation more difficult. Y

Easier to hire. One of the unexpected benefits of having a CRO-based biotech is that it is easier to hire scientific talent. Many biologists, chemists and other laboratory-based scientists want to move out of the lab and often want to build managerial skills. Additionally, hiring scientists in the biotech hubs (Boston, San Francisco, San Diego) is often challenging - opening the markets you hire from can help you move faster.

Remote work OK. Good scientists are all over; since the majority of our physical work is done by CROs, we hire scientists wherever they are and do not have to convince them to move to one central location. A bonus is that most larger biotechs and pharmas are slower to adapt to shifting work standards, poaching from these companies is easier :)

Harder to troubleshoot. This has

People management skills are necessary. Getting excellent work out of CROs requires excellent people and project management skills. As a startup, you will not have the leverage of bigger ticket clients; you will have to make up for this with people management. The project managers you work with at the CRO will likely not care about your timelines or mission like an employee at your team would, and therefore it is very difficult to motivate them to give 110%,.

Slow. All other things equal, an experiment will take longer with a CRO than at your own bench. CROs pad their time estimates and tetris your work with their other clients. You can’t sprint with a CRO or jump into the laboratory in the middle of the night to get something done.

Expensive. Taking out equipment, facility, etc costs, it usually costs more to run the same experiment with a CRO versus doing the same protocol in house.

IP inventorship. Most CROs agree to assign the ownership of any IP generated to the sponsor. Even thought it is standard to assign ownership to the sponsor, you still need to grant inventorship to CRO employees when you file IP based on the work.

Site checks & quality. For some type of work, especially manufacturing, you will want to tour the facility to ensure they are up to the quality bar. You, as the sponsor, are ultimately responsible for the product and work conducted by the CRO. You are also responsible for the ethics of the work conducted. In addition to being unethical, having stressed animals can and will modify your experiment’s outcomes.

Possible timeline delays. Even if you are ready to start tomorrow, the CRO may not have capacity to run your study. You ultimately cannot control their timelines. This gets better as you build a relationship with the project management team

Can’t ‘play around’ as easily

Scaling

Question & Answer

Some questions, mostly from my tweet thread, that I couldn’t fit into the narrative:

Some CROs we have worked with at Loyal: [2]

• Charles River Laboratories reliably delivers high quality work. They are known for being expensive. We do a ton of varied work with them.

• We have worked with WuXi a couple of times, and have not returned to using them. They are cheaper than CRL, but we found our experience unsatisfactory. Plenty of people love WuXi; ymmv.

• Pyxant Labs does a lot of our bioanalyte work (quantifying the presence or lack thereof of various molecules of interest in serum, etc)

• IDEXX Labs is the animal equivalent of Quest/LabCorp, and we work with them for a variety of bioassays from our canine clinical studies.

• BioStat Solutions supports on biostatistical analysis of the data that comes out of our studies.

  This is an incredibly incomplete list, and some of our best work has been done with highly specialized CROs.

Type of work we contract out:

• Ideating, testing, and validating formulations of our drug products [3] [4]

  • Eventually: manufacturing and packaging of our drug product for the forseeable future

• Most bioassays and bioanalytics that can use an off-the-shelf kit

  • Sometimes: development of novel bioassays (turns out that not many CROs have developed ELISAs for the dog versions of various aging proteins!)

  • We have worked with CROs to validate the cross-reactivity of kits of mouse or human protein for dog, and done this in house. Way cheaper in-house.

• Almost all preclinical efficacy work, especially since these are resource-intense and are usually on proprietary platforms

• Biostatistical design and analysis, although we have brought most of that in-house now.

What we do not contract out:

• The design of our studies (although we do work with the scientists at the CRO to optimize the studies)

• Drug ideation, discovery

• Specialized, aging-specific bioassays

• We have a split CRO/in-house model for our mouse work - the mice live at a Charles River facility, but Loyal staff could come in and perform work at the facility. The CRL staff were available to support on day-to-day care of the mice, which is important when you are running longevity studies!

[1] If you can, I recommend finding a shared lab space, sometimes known as incubators. We use MBC Bio Labs in the SF Bay Area. They come stocked with basic supplies and all the machinery you may need, handle maintenance, and sometimes come with community, too. The only negative has been availability of space - we have been trying, without luck, to get more benches at MBC for months, but the waitlist is long.

[2] Your experience will vary wildly depending on the site, team, and work you ask them to perform.

[3] Formulation is an important part of drug product development. You need to find a drug/API (active pharmaceutical ingredient)

[4] We want to bring the ideation and early POC formulation work in house as soon as we have the resources; the time delay from idea —> data has been incredibly painful.