This post is part of our series on the National Science Foundation I-Corps Lean LaunchPad class in Life Science and Health Care at the University of California San Francisco. Doctors, researchers and principal investigators in this class got out of the lab and hospital and talked to 2,355 customers, tested 947 hypotheses, and invalidated 423 of them. The class had 1,145 engagements with instructors and mentors. (We kept track of all this data by instrumenting the teams with LaunchPad Central software.)

Mira Medicine is one of the 26 teams in the class. The team members are:
  • Pierre-Antoine Gourraud--Ph.D., M.P.H., UCSF neuroscientist, and co-leader of the MS BioScreen project
  • Jason Crane--Ph.D.; UCSF Manager, Scientific Software Development
  • Raphaelle Loren--Managing Director, Health Practice at the Innovation Management Institute

Todd Morrill was the diagnostics cohort instructor. Matt Cooper, CEO of Carmenta BioSciences, was the Mira Medicine team mentor.

Multiple Sclerosis - MS
Multiple sclerosis (MS) is an immune system disease that attacks the myelin, the fatty sheath that surrounds and protects nerve fibers of the central nervous system (brain, spinal cord, and optic nerve). T-cells, a type of white blood cell in the immune system, become sensitized to myelin and cross the blood-brain barrier into the central nervous system (CNS). Once in the CNS, these T-cells injure myelin and secrete chemicals that damage nerve fibers (axons) and recruit more damaging immune cells to the site of inflammation.

multiple sclerosis and therapeutic targets

There are currently 10 FDA approved medications for use in relapsing forms of MS. None of these drugs is a cure, and no drug is approved to treat the type of MS that shows steady progression at onset. MS disease management decisions are complex and require a patient's neurologist to figure out what drugs to use.

Mira Medicine and Multiple Sclerosis
The team came to class with the thought of commercializing the UCSF Multiple Sclerosis BioScreen project, a precision medicine application that integrates a patient's medical records with the latest population-based data from hundreds of other multiple sclerosis patients (including their 3D MRI scans), and that using predictive algorithms makes it possible to chart a unique course of treatment for each patient. (Mira Medicine team member Pierre-Antoine Gourraud was the project co-leader.)

MS BioScreen

Mira wanted to commercialize the UCSF Multiple Sclerosis Bioscreen project and to add additional neurological diseases that require multiple types of data (including biomarkers, clinical, and imaging). It wanted to help medical centers and large providers assess disease progression to guide therapeutic decision-making. Over the course of the class the Mira Medicine team spoke to over 80 customers, partners, and payers.

Here’s Mira's two-minute video summary.

If you can’t see the video above, click here.

Then reality hit. First, the team found that their multiple sclerosis bioscreen application (which they used as their MVP) was just a "nice-to-have," not a "must-have." In fact, the "must-have" features were their future predictive algorithms. Next, they found that if their tool can enable a diagnosis (even without claiming it could), then it was likely that the FDA would require a 510(k) medical device clearance. Then they found to get reimbursed they need a CPT code. And they had to decide whether to code stack using multiple codes for "one" diagnosis, thereby getting multiple reimbursements for one test. (The rules have changed so that code stacking is hard or impossible.) Or they could get a new CPT code, or use miscellaneous code.

To get a new CPT and a 510(k) they would have to perform some sort of clinical study. At a minimum, they would be required to do a one-year prospective study (a study to see if the neurologists using the application had patients with a better outcome than those who didn’t have access to the app). Getting approval to use an existing (aka old) CPT code means showing equivalence to an existing dx process or test, and the requirements are code-specific. Finally, to get access to data sources of other MS patients they would need to have a HIPAA Business Associate Agreement.

Watch their Lessons Learned video below and find out how they pivoted and what happened.

If you can’t see the video above click here.

Look at their Lesson Learned slides below 

If you can’t see the presentation above, click here.

Lessons Learned

  • Researchers and PI’s come in believing "My science/project/data are so good that people will immediately see the value and be willing to pay for it.  It will sell itself."
  • A business is much more than just good science: It is about customers seeing value and being willing to pay and proper validation and reimbursement coding and
  • A successful business is the sum (and integration!) of all the parts of the business model canvas.
    • It includes reimbursement, regulation, IP, validation, channel access, etc.