Objective
Our ultimate goal for this project is to develop a working prototype of a pumping mechanism capable of providing an unprecedented measure of control over delivery accuracy relative to other patch pump devices. This is made possible by the geometrically-defined volumetric displacement of our electrostatic/electrokinetic microactuators, a unique and highly desirable characteristic feature among MEMS devices. These actuators are theoretically capable of meeting the ideal patch pump requirements of relatively simple and cost-effective fabrication, low power, sufficient actuation force, and accurate volume delivery. Moreover, the normally-closed multi-chamber design we propose inherently acts as a built-in one-way valve to prevent backflow and deliver fluid only when actuated, possibly eliminating the need for additional fluidic complexity such as check valve integration.
Through previous funding from the Helmsley foundation, we demonstrated an innovative approach of using induced-charge electrokinetics to actuate a soft diaphragm with a reduced footprint and at relatively low power compared to existing approaches. Unfortunately, the specific mechanism we studied (DC-driven bipolar electrokinetics) could not reliably achieve the requirements needed in terms of stroke power for a given input power. However, after investigating other mechanisms and undertaking a detailed numerical analysis/optimization of micropump geometries to meet these criteria, we arrived at viable designs made from conventional MEMS materials (silicon, glass, gold, and PDMS/other polymers) capable of fixed 500 nL actuation strokes under applied voltages on the order of ~10 V or less, with actuation either achieved either through 1) electrostatic zipping or 2) ac electroosmotic flow (ACEOF)-induced hydraulics. Therefore, to achieve the ultimate goal outlined above, we plan to pursue the following specific aims:
Project Aim 1: Validation of Theoretical Pumping Systems
In this aim, we propose to experimentally validate the two aforementioned novel pumping systems. Specifically, we aim to first achieve the robust fabrication of the simplest embodiment of each device. Next, we will not only confirm the repeatability and reliability of our fabrication process, but also demonstrate the ability to actuate each device with the application of an electric field. This can be done both visually (by optically tracking fluorescent particles on the diaphragm) and electronically (via current measurements). Successful achievement of this aim will not only allow for fine-tuning and optimization of our numerical models towards a viable commercial pump, but will also provide a plethora of fundamental research insights into such systems.
Project Aim 2: Characterization, Optimization, and Viability Determination
The ultimate goal of this project aim is to determine the viability of our proposed pump mechanisms in terms of ability to meet the system-level requirements of our end vision. To do that, we plan to iterate on pump embodiments to achieve the following objectives:
1. Understanding flow accuracy, precision and repeatability
2. Assessment of ability to pump against downstream hydrodynamic resistances
3. Initial trumpet curves with both water and off-the-shelf insulin
4. Flow diodicity and ability to build without a valve (and/or design of a valve, if necessary)
5. Practical fabrication and integration ability
Project Aim 3: Prototype Development and Commercialization Plan
As we hope to commercialize our mechanism in a fully integrated AP system, our goal is to garner commercial interest by the conclusion of the grant period. Therefore, we aim to have some sort of initial working prototype and commercialization plan by the end of this period. Objectives for this project aim include:
1. Determination of any additional valving requirements to achieve system level requirements
2. Fluid delivery/trumpet curves of prototype under physiological hydrodynamic conditions
3. Fabrication/manufacturing cost analysis
4. Major remaining technical obstacles and plan to overcome
5. Aggregation analysis with higher concentration insulin and assessment towards achieving delivery of concentrated insulins
Background Rationale
The pumping mechanisms we propose are truly innovative and have not yet been implemented in any sort of drug delivery system. As mentioned previously, the volumes of fluid that pumps on the market are currently designed to deliver (10-100 μL) are still sufficiently large to preclude the miniaturization and seamless integration onto a comfortably small form factor wearable “patch-style” device. All other existing non-microfluidic innovations, both macroscale and microscale, require significant volume for accurate pumping and their performance either deteriorates or becomes more difficult to precisely control at smaller scales. This is the arena in which on-chip microfluidic electrostatic/electrokinetic actuators shine, however; the ability to geometrically define fixed stroke volumes in microactuators which actually provide more force when downscaled enables an unprecedented measure of flow control in nanoliter-scale volumes and provides a key advantage over alternative macroscale (and microscale) competitors. Piezoelectric actuators can provide many of the same advantages, for example, but tend to require large applied voltages when fabricated on-chip and necessitate much more complex fabrication and integration steps than electrostatic/electrokinetic actuators.
While electrostatically actuated polymers have been widely studied and implemented in microfabricated chips and soft robotics applications, previous iterations have not been specifically designed to meet the key criteria of simplicity, low voltage/power, and precise actuation control that we strive for and which are necessary for a successful patch pump product. Conventional electrostatic actuators have been integrated into micropumps in research environments, but these pumps either required excessively large voltages >100 V or provided condition-dependent strokes and volume delivery rates too large for the delivery of concentrated therapeutics. The key innovation of electrostatic zipping actuators allows us to overcome these shortcomings, however, as they reduce the actuation voltage required to a manageable level and provide predictable, fixed-volume strokes due to the complete and conformal collapse of the actuated diaphragm onto the curved substrate during pull-in. These zipping geometries have demonstrably improved the performance of ~mm scale actuators for wearable haptics, for example, but have yet to be further downscaled and see their full potential realized by integration in microfluidic pumping systems - as we propose to do.
Electroosmotically-driven actuators have also drawn renewed interest of late, but none of the existing approaches have specifically targeted the application of micropumping for drug delivery. The concept of using electroosmotic flow to hydraulically drive fluid and subsequently deform a soft material was introduced in so-called “nastic” actuators, but these devices were designed to operate at a much larger scale and thus required significant voltages (~ kV) to generate the electric fields necessary to drive the fluid. This is a key disadvantage of hydraulic pumping with DC EOF, as the flow speed that determines the induced pressure is proportional to the electric field, which is dictated by the length of the actuation channel (E = V/L). Nonuniform DC-voltage-driven electroosmotic flows have, for example, recently been demonstrated to hydraulically actuate thin-film elastomer diaphragms, exploiting fluid-solid instabilities driven by patterned surface-charge-governed flows of opposing directionality (with instability-based actuation behavior not dissimilar to zipping actuators). However, the applied potential (~100 V) in these DC EOF devices is also sufficiently large to diminish their potential for useful, practical pumping. Our innovation of hydrodynamic flow generation using optimized AC electroosmotic flow (ACEOF) electrode pairs (with actuation voltages of order ~1 V and large electric fields maintained by the closely-spaced electrode pairs), however, allows us to overcome this significant drawback and downscale the devices to the degree necessary to achieve our desired power, pressure, and flow specification metrics.
Description of Project
Continuous Subcutaneous Insulin Infusion (CSII) pump systems have transformed the quality of care for patients with Type 1 diabetes in recent decades. However, current state-of-the-art commercial pump systems still leave much to be desired. First, existing pump systems are large and cumbersome, usually requiring an infusion set and separate tubing. Those that do not (e.g., patch pumps) still tend to be large and cumbersome, mainly because of the footprint of the pumping mechanism, batteries required, and the volume of insulin required to last 3 days. Additionally, pumps cause site loss, skin irritation, and other issues due to a variety of potential problems, including the continuous injection of phenolic compounds (required for preservation, reducing aggregation, and sterility) into the body at the insulin infusion site. Moreover, there are no products on the market that fully incorporate into the pump a continuous glucose monitor (CGM), which means that most patients with Type 1 diabetes need to wear multiple devices to take advantage of all the artificial pancreas (AP) algorithms that have been presented in the market. Finally, AP systems to date all have a long lag time because subcutaneous systems are known to be ~15 min delayed from blood glucose.
Our vision is to overcome all of those challenges, which will truly differentiate our system from anything currently on the market. PI Pennathur has a company that is developing a dermal CGM (with less than one minute delay from blood glucose) that can easily be incorporated into a compact patch pump. Furthermore, PI Pennathur has been working on novel pumping mechanisms using nonlinear electrokinetics (a specialty in her lab over the last 15 years) to be able to pump 10x less fluid with improved accuracy; these fixed-volume microfluidic pumping mechanisms allow for flow control and fluid delivery of nanoliters of solution with unprecedented accuracy in the wearable pump market, enabling the inevitable transition to higher concentration insulin formulations and thus reducing the volume of insulin required per device (leading to a smaller form factor patch pump) while also extending lifetime. Finally, we will leverage our recent work on the electrooxidative removal of phenolic compounds from insulin using novel chemical modification of electrodes to avoid passivation, which enables all toxic excipients to be extracted within the cannula immediately before delivery into the body.
For this proposal, we will focus on developing the patch pump insulin delivery mechanism idea to a point beyond proof of concept, as we believe this to be the highest risk and most novel piece of the puzzle, and we have parallel programs existing for the CGM (through PI Pennathur’s startup company). Specifically, our aims are to: 1) evaluate two different microfluidic pumping mechanisms, both numerically and experimentally, to determine which is most suitable for insulin pumping, 2) characterize and optimize the chosen pumping mechanism with the goal of producing equivalent or better trumpet curves (i.e., basal delivery accuracy) in-vitro than pumps currently on the market, and 3) develop a robust prototype capable of meeting our desired pump specifications and determine the commercial outlook for the concept in terms of COGS and additional development needed to bring such a system to market. We believe that these specific aims are the most important aspects to de-risk a miniaturized patch pump for higher concentration insulin delivery and, if successful, we can license the IP to a commercial development entity for production.
Anticipated Outcome
The expected outcome of this research is to have a prototype pumping mechanism that can satisfy the requirements needed for a next-generation dermal patch pump. If either of our proposed mechanisms is deemed viable, we aim to develop it to the point where we can demonstrate a prototype with both off-the-shelf insulin and then more concentrated insulin formulations. At the conclusion of the proposed work, we hope to be well-positioned to be able to partner with a commercial entity and/or spin-off our own company to handle the more practical aspects of prototype development. This includes occlusion studies, embedded pressure sensors for accurate monitoring, investigating priming of the complete pump system, reservoir pressure conditions, and gas intrusion into the reservoir. Although we believe all of these to be very important to the development of a pump, we believe them to be better suited to the final development stage than the research we are proposing here; we also have partners to assist us through the prototype development phase, including partners at CASS pharma and Diatech Diabetes. Once we show good in-vitro performance, we will subsequently develop the more mechanical aspects of the pump (e.g., connections, interfaces with the injection system, hermeticity/packaging, etc.) This will likely take 1-2 years of further development time, while in parallel we will continue to perform inverter trials with all prototype systems. Once a fully validated model prototype has been manufactured and sterilized, we plan to perform pre-clinical trials through Integrium IRB with facilities either in-house at UCSB, Sansum Clinic in Santa Barbara, or a third-party vendor. Finally, once we show pre-clinical trials to be safe, we will apply for an IDE from the FDA to perform clinical trials to fully substantiate the commercial viability of our micropump mechanism.
Relevance to T1D
Our main vision is to revolutionize care for patients with Type 1 diabetes by offering a fully automated Artificial Pancreas (AP) system. Indeed, to date, continuous subcutaneous insulin infusion pump systems have transformed the quality of care for patients with Type 1 diabetes; however, current state-of-the-art commercial pump systems still leave much to be desired and have yet to offer a fully automated AP system. First, pump systems are currently large and cumbersome, mainly because of the pumping mechanism footprint, batteries required, and the volume of insulin required to last 3 days. Additionally, pump systems are known to cause site loss, skin irritation, and other detrimental effects, thought to stem from the continuous injection of phenolic compounds (required for insulin preservation, reducing aggregation, and sterility) into the body at the infusion site. Moreover, there are no pumps on the market that fully incorporate a continuous glucose monitor (CGM) into the pump system itself, forcing most Type 1 diabetes patients to wear multiple devices to take advantage of the various AP algorithms that have been presented in the market. Finally, AP systems to date all have a long lag time because subcutaneous systems are known to be ~15 min delayed from blood glucose, and therefore there is nothing currently on the market that is fully autonomous.
To overcome these challenges, we envision a CGM-integrating dermal patch pump that delivers high-concentration insulin and removes toxic preservatives and excipients upon delivery in the cannula to extend wear lifetime. Such a system would significantly increase the quality of care for patients with Type 1 diabetes. As we already have programs and/or partners focusing on building a dermal CGM, formulating higher concentration insulins, and removing toxic excipients, we are focusing this proposal on overcoming one of the highest risks of this main vision: developing a novel pumping mechanism that is much more compact and accurate than those on the market. The Pennathur Lab’s expertise in microfluidics and microfabrication places us in an ideal position to innovate and develop the technology needed for this next generation of patch pumps.
To date, micropump commercialization efforts have been largely centered around piezoelectric, mechanical, and electrochemical actuation methods. While these techniques have proven effective at moving relatively “large” volumes of fluid (μL-mL), fabrication constraints and a number of other hurdles have limited the development of affordable technology for more precise manipulation of the increasingly small (nL-μL) fluid samples demanded by devices delivering concentrated therapeutics. For example, piezoelectric and electrochemical MEMS micropumps require relatively large power inputs and can have variable flow rates under adverse operating conditions, while more conventional mechanical actuators are bulky and difficult to downscale to the size necessary to meet desired precision specifications. We have been developing fixed-volume microactuators which are theoretically capable of meeting the ideal patch pump requirements of relatively simple and cost-effective fabrication, low power, sufficient actuation force, and accurate volume delivery.
The Pennathur Lab has recently been developing two microactuation mechanisms in particular (electrokinetic and electrostatic) that have shown promise towards being incorporated into a pump; however, our two candidate actuation mechanisms have not yet been fully characterized and optimized. The development of these innovative microfluidic pumping mechanisms will allow for a smaller pump footprint due to the simple on-chip microfabrication processing. Additionally, the pumping mechanism itself will be compact, as we will be using higher concentration and fast acting insulin formulations with smaller delivery volumes. This micropump concept could truly revolutionize patient care, and our goal is to demonstrate a working prototype of such a pumping system with commercialization potential by the end of tenure of the grant period.