Randy Dorian is the Principal Inventor at the Hanuman Medical Group, the inventor of the Islet Sheet. He has various degrees but is best understood as a jack-of-all-trades without being an expert in any, a true 19th-century biological amateur. We asked him to how he approaches the task of invention, especially his approach to the Islet Sheet.
We’ve been successful in bringing a number of medical products to market. We thought it might be useful to some to share the way we go about solving problems to produce useful devices and methods.
We consider it important to keep things simple and to avoid, as one former collaborator described it, “paralysis by analysis”. In other words, to avoid getting lost in mountains of minutiae and losing sight of the obvious. I’ve often said, only half in jest, that I protect my ignorance as a pianist protects his hands. By focusing excessively on what has gone before, one sometimes runs the risk of becoming persuaded by the apparent impossibility of certain approaches or being lured by the promise of tantalizing findings leading ultimately down dead end pathways.
In the case of the Islet Sheet, we began by defining our end goal as simply as possible and then establishing a set of theoretical obstacles which would need to be overcome to achieve the goal. Put most plainly, the goal is to keep islets alive and functioning in the recipient. What could get in the way of achieving that end? We made a list of requirements and reviewed the list with the question in mind, “what else might be important?” When the answer to that question finally came back “nothing”, we checked to be sure that none of the identified “requirements” was actually unnecessary. Thus the list was of criteria both necessary and sufficient for success. Next we set out to as simply as possible address and solve (at least to theoretical perfection) the entire list of requirements. That the list is comprised of requirements means that all items must be simultaneously met or failure is certain. We sometimes joke that what distinguishes our approach from many other approaches is that ours could work (there is at least in principle no “fatal flaw”).
There are a number of basic directions one might follow to keep islets alive and functional, including immunoisolation, immune suppression, tolerance induction, islet regeneration and mechanical approaches (feedback controlled insulin pumps, which of course don’t require islets at all). For a variety of reasons (e.g., shunning impracticality, improbability, patient risk, etc.), we chose to narrow our focus to immunoisolation.
What’s required to keep immunoisolated islets alive in the recipient? Here’s the list of criteria:
1) The encapsulation process must not damage the islets
2) Immune cells of the recipient must be prevented from physical contact with the islets
3) The islets must be protected from destruction by humoral immune attack
4) The islets must be sufficiently nourished to maintain viability and function.
Once the requirement of keeping the islets alive is achieved, the second part of our simply defined goal is to ensure their functionality in the host. As it turns out, this goal is automatically realized by fulfilling the conditions required to maintain viability. In a word, the conditions necessary to assure sufficient nourishment include a requirement for rapid diffusion of small molecule nutrients (like sugar and oxygen) into the sheet and small molecule waste products (like lactate and ammonia) out of the sheet. Given sufficient small molecule permeability to maintain viability, the diffusion of little glucose molecules into the sheet and of the small hormone insulin out of the sheet will be sufficient to allow rapid response and correction to changes in blood sugar, i.e., adequate functionality.
So now we have a set of four very simple requirements which are necessary and sufficient to keep islets alive and functioning in the recipient. Just as we approached the overall goal by simply defining it and identifying the requirements for success, we can next turn to each of the four identified broad criteria and again ask the question of what is sufficient and necessary to satisfy each subgoal. Here’s how we broke each subgoal down:
1a) The reagents used must all be non-toxic
1b) The process must not induce excessive mechanical trauma
2a) The coating must be impenetrable to host cells
2b) The coverage of islets must be complete
3a) The material contacting the host must be stable and unrecognized as foreign
3b) Diffusion of immune molecules into the sheet must be sufficiently retarded to prevent immune destruction
3c) Islet coverage must be complete (same as 2b)
4a) Small molecules, notably oxygen, must diffuse very rapidly from adjacent tissue to the encapsulated islets
4b) Larger nutrient molecules (such as transferrin) must be able to diffuse through the sheet, albeit at a slow rate.
The rationale for some of these sub-requirements may not be immediately apparent, but thorough contemplation in light of basic facts of immunology and cytochemistry will reveal their importance and sufficiency.
Of course satisfaction of some of these sub-requirements necessitates amplification to the sub-sub level of analysis. As long as the criteria identified at each level of analysis are sufficient and necessary, we are certain to finally end up with the simplest system capable of achieving success. As long as we’ve taken sufficient care in breaking the problem down into sufficient and necessary individual criteria the only possibility for failure would be if the ultimate goal is impossible. The possibility of impossibility resides entirely in a very few sub-sub criteria for which there is no known answer. For example, we know the molecular dynamics of cellular destruction by humoral immune processes and so we can confidently set an appropriate permeability profile to prevent immune rejection, but we don’t know every single nutrient which might be necessary for long term islet functionality nor the rate at which they are consumed or utilized, so determining ideal permeability to maintain long term function while avoiding rejection cannot be done on purely theoretical grounds. For these known unknowns we can only rely on intuition and what relevant empirical data we can gather from our own work and the work of others. Interpolation and extrapolation from what is known in these areas provides us with a high level of confidence that our ultimate goal is not impossible. For example, with regard to sufficient nutrient supply to the islets, we have previously maintained viability and functionality of microencapsulated dog islets in a dog recipient for seven years, using an immunoprotective permeability profile similar to that of the Islet Sheet. Because of the way we tackle our goal, we are confident that if success is possible (as we believe it is) we will succeed.
Figure. The movie illustrates our approach to capsule permeability. Blue circles show the size of seven components from essential nutrients (left) to an essential component of humoral immunity. The green line is our goal for diffusivity of these components.
I agree with your solution-oriented approach. When Bayer developed aspirin to treat inflammation and headaches in 1879, there was absolutely no basic research demonstrating how it worked, but only empirical and epidemiological evidence that it would have this effect at non-toxic levels. Only a century later did the first explanations emerge for why it worked as it did. Today the general approach is to pour billions of dollars and decades of research effort into studying the basic science of diseases in the hopes that a cure for them will become apparent as a side-effect, but unfortunately, as is abundantly evident in cancer research, vast amounts of basic science data can be collected with little or no corresponding gains in clinical treatment outcomes. Perhaps a more functional approach, just looking for what works whether we understand why it works or not, would prove more useful, especially given the slowing of medical progress over the past half century.