• Performed due diligence on an approach for non-invasive glucose monitoring using Raman Spectroscopy.   A fingernail bed was irradiated through the fingernail using a A NIR (Near Infra Red) laser.

In late 2005 and early 2006, after my experience at Predicant Biosciences, I consulted for Skymoon Ventures where one of the projects I worked on was a due diligence dive on a business plan by Company X to market a non-invasive glucose monitoring system.  The system measured glucose levels using Raman Spectroscopy by irradiating the fingernail bed, through the fingernail, with a NIR (Near Infra-Red) laser.

An accurate, reliable, inexpensive, non-invasive blood glucose measurement system has been diabetes management holy grail for a long time – and given the explosive growth in Type-II obesity related diabetes, its a market with a big CAGR.

I had just completed 3+ years at a cancer biomarker detection startup called Predicant Biosystems, and was able to leverage the connections I had made there in the biotech space to quickly assemble and meet with a number of experts in the both the FDA device regulatory approval process and in blood glucose measurement and management.  These individuals included Anna Longwell (regulatory expert), Ephraim Heller (the brother of Jonathan Heller -who was VP Bioinformatics at Predicant- and co-founder and CEO of TheraSense), Karen Drexler (blood glucose), Diane Ward (regulatory expert) , Syrous Parsay (glucose monitoring), and Roger Rudoff (engineering).

The most important characteristic of any blood glucose measurement system is accuracy.  Accuracy is almost always better if glucose levels are measured directly rather than indirectly, and Company X got a check mark there as they utilized a direct measurement approach.

However, the downsides of Company X’s approach were:

1. It was not portable – which severely limited approachable markets
2. A fairly long measurement time was required (this had to do with low gain and detector sensitivity of the Raman-Stokes system)
3. It was expensive.  The laser alone for the prototype was $7K in single units quantities.  The prototype unit cost was ~$75K, and it was felt that a BOM for a marketable instrument had to be in the $700 range.
4. It would require a PMA (PreMarket Approval) certification path through the FDA as there were no “substantially similar” predicates to justify a more direct 510K application approach.   The PMA path is longer and more expensive as the applicant must prove efficacy and safety via studies.
5. I high power (class 3B) “invisible” NIR laser was required (this is a significant component of the high cost in #2).  Approximately 400mW was delivered to the nail in the initial study – which alone was thought to potentially violate ANSI skin MPE (Maximum Permissible Exposure) limits.   Because the beam is invisible there is no blink reflex, which makes safety criteria more stringent.
6. Technical risk.   The original study detector was cooled with liquid nitrogen, and there was not yet much empirical data regarding the performance of a much less expensive non-LN2 (thermo-electrically) cooled detector and delivery of the laser beam through a fiber-optic cable (instead of free space).

Given the above issues, my recommendations were to discontinue attempting to refine the existing design center, and instead focus – with perhaps some out of box thinking – on an approach which could have a much better chance of meeting the BOM cost necessary to be commercially viable.