Research Status

Pioneering research in the field of high-precision out-of-plane microneedle array fabrication and their medical applications were first studied by research groups at the Berkeley Sensor and Actuator Center, University of California (Berkeley, CA, USA), the Royal Institute of Technology (Stockholm, Sweden), MESA+ Institute for Nanotechnology, University of Twente (Enschede, The Netherlands) and the Georgia Institute of Technology (Atlanta, GA, USA) [3, 34–41]. A large number of publications have appeared over the last 15 years concerning the fabrication and testing of microneedles in a diversity of applications, including brain–computer interfaces, microneedle probing and stimulating for various brain regions, scalp electroencephalography, drug delivery, diagnostics and applications where spiked microstructures are used as nozzles and pipettes.

MNAs containing microneedles of diverse shape, length, diameter and density have been demonstrated to be capable of bypassing the skin barrier. Both in-plane and out-of-plane fabrication are well suited to the creation of microneedle arrays, but each requires a different approach if it is to be integrated into a skin patch. Only the out-of-plane method will directly produce a two-dimensional array seamlessly connected to a back plate, which is preferable for integration into higher functionality [3]. MNAs are also fine-mechanically manufactured by assembly from conventionally produced hypodermic needles, and successfully demonstrated in certain types of transdermal delivery studies [42]. Nonetheless, lithographic exposure techniques utilizing microsystems technology (MEMS), (i.e., reproducing a structural element multiple times from a mask or mold), are now established in drug delivery research.

Lithographic techniques result in arrays with a tip radius of < 100 nm, microneedle outer diameters below 300 μm, very well-defined length in the range of a few to hundred micrometers, and varying needle densities. Batch processing allows some cost reduction. Hollow MNAs require a flow-through delivery system and have been tested by gluing the back plate of the MNA to a syringe. These devices are not yet applicable to self-administered transdermal insulin delivery, but they demonstrate a clear step forward. Figure 8.2 summarizes complementary examples of out-of-plane MNAs, and demonstrates early prototypes of microneedles coupled to a syringe for intradermal drug delivery [33, 43–46]. Recent publications have discussed microfabricated microneedle technology, and evaluated their performance against other insulin delivery techniques [24, 47–49].

Fabrication and Design Concepts

Some processes form the MNA device directly, while others first produce a master. Silicon micromachining and stereomicrolithography are two examples of direct techniques forming three-dimensional work pieces [3, 37, 51]. SU-8 photolithography on pre-patterned silicon substrates combining lithography with micromolding, titanium micromachining, and the use of two photon induced polymerization of organic–inorganic hybrid biomaterials [52–54] have also been attempted experimentally.

Some remarkably simple methods have been recorded for micromolding microneedles. For example, Lee et al. introduced drawing lithography for the three-dimensional fabrication of an ultrahigh-aspect-ratio microneedle, and Zhu et al. recently demonstrated a fabrication process that uses chemically etched silica needles as a template for a replication process [55, 56]. Unfortunately, Zhu et al. [56] omit a detailed description of the fabrication of the silica template. Although their paper does not describe the fabrication method of the template the authors clearly identify micromolding as a low-cost manufacturing scheme, which is suitable for further drug delivery research.

Many different processes have been tried to make extremely sharp microneedles. Unfortunately, there is no consensus on how one should define sharpness in this context. Many different research groups have elaborated on various fabrication processes to generate particular needle geometries, with the intention of reducing the tip radius, which will obviously lower the insertion energy. The density, the total number of needles within a defined array area, and the mechanism of insertion (manual or impact) are likely to be the major factors influencing insertion performance. Knife-like features, such as that introduced by Gardeniers et al. [38], should be assessed against other designs with respect to pain sensation, skin reaction, intrinsic skin variations from volunteer to volunteer, place of incision and relevant parameters of wound healing in relation to the size and geometry of the incision. Development of a benchmark for these parameters is not trivial because the skin is an anisotropic biomaterial, which changes its mechanical characteristics under stress. Moreover it is unclear how the tip radius of a microneedle (i.e., sharpness) can be translated into a meaningful parameter for such an incision experiment.

Some studies have explored needle radius together with other geometrical and application parameters through mechanical modeling [57, 58]. However, the ease of insertion of the microneedles or the successful introduction of drug does not just depend on the individual needle tip, or the configuration of the array, but also on the apparatus that is used to administer the MNA device to the skin and the drug formulation [59]. Furthermore, a trade-off between sharpness and microneedle stability must be sought for a clinical application. The insertion of microneedle arrays in human subjects was reported in a study by Gill and Prausnitz [60].

Materials

Microneedles are made either directly from active agents embedded in a dissolving material, or the microneedle platform acts as a carrier, which releases active compounds from a coating (solid microneedles), through a fluid conduit from a reservoir (hollow microneedles) or from the material’s intrinsic pores (porous microneedles). The range of materials used in research or commercial activities includes glass or silica, glass-like plastics and ceramic composites, silicon, silicon nitride, stainless steel, nickel-plated needle arrays and titanium, amongst various types of polymers, carbohydrates and micro-nanoparticle hybrid materials. Each of these materials may require their own fabrication techniques, for example, stainless steel or titanium microneedle arrays are cut by a laser, which limits their physical dimensions to the resolution of the laser cutter. This limited resolution will also affect the radius of the needle tips, and the overall geometrical dimensions of the needles.

Silicon, which forms a native oxide layer, is much used in microsystems technology and has been accepted for use in transdermal pre-clinical research, although no commercial application yet exists. Ji et al. have done some preliminary work in the area of porous microneedles. They used selective electrochemical etching, with a photoresist as a protective mask to form the porous silicon tip, thus forming one of the first silicon out-of-plane microneedle arrays that contain tips of two types of materials with very distinct properties [61]. Other publications also discuss the range of materials used in out-of-plane micro- and nanomanufacture of MNAs. For example, Park et al. investigated a polymer double-layer molding process, as shown in Figure 8.2f, which was generated from a mold structure. The mold was fabricated by means of a combination of SU-8 photolithography, reactive ion etching and polydimethylsiloxane soft lithography [45, 50]. They produced tapered tips, using poly(lactic-co-glycolic acid) as a material for the formation of dissolving microneedles and encapsulated calcein in the tips. The particle structure remains intact inside the molded microneedle.

Similarly, Moon et al. fabricated a mold by soft-lithography, from a master made by inclined deep x-ray exposure, then the soft-mold is used for the replication of MNAs into biodegradable poly-L-lactide acid [62]. Carbohydrates have also been used for molding MNAs, and Donnelly et al. discussed the processing difficulties and instability of the material in MNA production [63]. Kolli and Banga, and Li et al. have found maltose to be more easily adapted to molding microneedles [64, 65]. Both Kolli and Li are members of Banga’s group affiliated with the Mercer University (Atlanta, GA, USA), which holds an impressive record on microneedle array technology.

Ceramic Nanoporous Microneedles

Within microneedle research, some groups are attempting to make ever sharper and smaller point-like microfeatures, while others are looking further on in the development chain, towards low-cost production and applications. One particularly exciting concept is the possibility of making arrays of microneedles in one platform, so combining a multitude of microneedles in parallel without the need for additional assembly steps. Such arrays could be used to make hundreds of miniscule incisions into the skin in one step, without causing pain. These incisions could then be used to transport molecules across the skin barrier. Figure 8.3 shows a ceramic-based nanoporous MNA platform technology developed at University of Twente’s MESA+ Institute for Nanotechnology (patent pending WO2009/113856). Here, the MNA is replicated from a soft-mold that is generated by polydimethylsiloxane soft-lithography from a master – this technique was discussed in Chapter 3, Section 3.3.4. A specific master was designed and manufactured by a combination of SU-8 photolithography and silicon micromachining that also integrates knife-like features in the tip with similar dimensions as introduced in previous work by Gardeniers et al. [38].

A nanoporous microneedle has a porosity of approximately 30%, and an average pore size of 100 nm (see inset Figure 8.3, top right), which is large enough to hold model molecules such as fluorescent ovalbumin (inset Figure 8.3, bottom right). Further investigation is necessary to determine the extent to which such non-dissolving, nanoporous platforms can be of use in the delivery of macromolecules such as insulin.

In general, transdermal MNA devices are classified by the Food and Drug Administration as an invasive device, since the skin barrier is bypassed and also probably ruptured at deeper layers than the stratum granulosum. In most cases applying such devices to the skin surface will allow substances other than the drug to enter the body. Any skin rupture event will interfere with the body’s defence mechanism, generating additional adverse reactions in the skin, or even unfavorable systemic effects through infection or poisoning, hence this classification is necessary. All components of the microneedle administration device that come into direct contact with tissue, or risk contact during operation of the device must, therefore, obey stringent rules to gain medical approval.

At this stage of development, the exact material properties of a MNA are not fully understood, and have to be assessed as they stand and also when they are packaged within a drug-applying device. Despite their general potential being recognized in the drug delivery community the benefits of a specific material for insulin delivery still have to be addressed for all the different novel platform technologies described above.

Figure 8.4 gives an overview of the nanoneeedle fabrication process developed at University of Twente’s MESA+ Institute for Nanotechnology. This process starts with a hard mold, manufactured by photolithography and silicon micromachining, to combine a pre-patterned silicon substrate and a thick-film photoresist (SU-8) into one hard-mold structure for the fabrication of a soft mold for subsequent multiple production of ceramic green bodies. The individual MNAs are cut from the green body and sintered, forming the semi-finished product ready for integration into a delivery patch.

Figure 8.5 depicts the integration of the ceramic MNA into a patch-type system. With a resulting materials porosity of approximatly 30%, a single microneedle within the array has a liquid loading capacity in the order of 1–2 nl, and the backing plate can contain up to 5 μl. These minute amounts may suffice for vaccination purposes. Currently, the application of University of Twente’s MNA technology is being investigated in collaboration with Prof. Dr. Scheper’s and Dr. de Gruijl’s groups at the VUmc Free University Medical Center, Amsterdam, The Netherlands, for the purpose of vaccine delivery. This type of nanoporous MNA is probably not appropriate for insulin delivery, but one needs further research to be conclusive.

Summarizing the Technology Requirements from a Commercialization Point of View

Microfabricated devices compete with the delivery of insulin by standard, ultra-fine needle systems, either integrated in a patient-friendly, pen-like delivery system, or as a classical hypodermic needle mounted on a syringe. This type of needle costs about 0.05–0.20 US Dollars per disposable needle. MNA replication technology can probably compete with such a low pricing regime, but the pharmacodynamics of insulin uptake from the epidermal layer needs much more research, and is not yet clinically identified as a beneficial route for insulin delivery.

Small-molecule-releasing skin patches are already far better understood, and have been marketed at a large volume in industrial pharmacy. Fitting microneedles into one of the existing skin patch platforms may accelerate its integration into a technological solution. An example of successful, interdisciplinary, transdermal, drug delivery co-development is evident in current products in the iontophoretic patch market. These products combine microelectronics integrated current-controlled power supplies with skin patch technology into a patient-controlled iontophoretic transdermal system, providing on-demand systemic delivery of substances such as fentanyl (Ionsys™) and lidocaine (LidoSite^®) [66]. Clinical trials are under way for both these systems. These electrically-controlled systems are quite complex, so it is surprising that passive microneedle arrays in a skin patch lag behind them in the market.

Another very promising example of MNA technology in a transdermal application is being commercialized by Zosano Pharma Inc. Zosano Pharma’s Macroflux^® transdermal microprojection delivery system combines the convenience of a needle-free administration process with room temperature stability for various therapeutic peptides, proteins, small molecules and vaccines. The Macroflux^® microneedle array component has been described in the scientific literature for more than 10 years.

Upgrading MNA technology from a device that just creates microscale aqueous conduits into the skin barrier into a fully packaged and safe drug delivery system comparable to Macroflux^® will clearly require more than just engineering ingenuity. A device platform must be designed and manufactured, that will not only allow testing of a small number of demonstrators for transport studies, but must also present a better solution for diabetes management, justifying patient trials supported by the shared risk of private–public partnerships. Once such a microneedle technology platform is confirmed by evidence-based clinical translational research, it can be re-engineered into a more flexible system that will accommodate and address individual patient’s needs through the appropriate industrial channels for market introduction. The state of the art in microneedle array technology as it exists for insulin delivery in the scientific literature is discussed in the next two sections.

Fabrication Attributes and Their Clinical Relevance

Shaping of the MNAs directly, by silicon micromachining techniques (category I, Figure 8.6), is considered to be an expensive technical solution due to the relatively high cost of high-precision silicon wafers as a base material. These techniques allow a high level of integration, and are advantageous when a certain amount of electronic integration is needed for the device. These processes will also be capable of producing very sharp tips, with radius less than 100 nm. This reduces the insertion force needed for such devices, thus enabling ease of use for manual administration. A first comprehensive trial, benchmarking important delivery parameters in the diabetic rat model, was carried out by Nordquist et al. [32]. Figure 8.8 shows two of their findings: (1) the decreasing plasma glucose level, showing the effect of the route of administration by microneedles (Figure 8.8a), and (2) the effect of the rate of intradermal administration on blood glucose (Figure 8.8b).

The second category (II), Figure 8.6, covers replication processes from a master. In general, these are cheap production schemes at a relatively low level of integration. Polymer micromolding processes incorporating insulin into a matrix material were introduced by Ito et al. This group used a relatively expensive, lithographically-defined master to yield very well-defined features in a biodegradable matrix [78]. As an alternative, batch-replication by electrodeposition from a relatively cheap sacrificial master, made by laser machining is less accurate but generates a very specific design with a through hole [74]. These two process types may each have benefits and opportunities for cost reduction. From the affordability point of view, high-volume replication from a sustainable master structure is the most feasible route.

Clinically, however, these two replication concepts are very different. The dissolving polymer microneedle can only allow a fixed amount of insulin release per time interval. On the other hand, the hollow metal microneedles can be coupled to an insulin reservoir [74, 78], which can in turn be linked to a closed-loop feedback sensor, which triggers release of insulin on demand. This is similar to continuously controlled release from an implanted or percutaneous drug delivery pump, and thus should bring high patient compliance. Such concepts anticipate high efficacy and reduced side effects during therapy. Other hollow micromachined microneedles, e.g., as given in category (I), also offer this kind of advantage.

The devices in category (III), Figure 8.6, are made by fine-mechanical machining and are perfectly suitable for a relatively small number of devices that bring a technical solution to a drug delivery niche market, but they remain rather expensive for high-volume manufacturing, so their widespread application is limited. A variety of studies are described that use fine-mechanically manufactured out-of-plane microneedles. Martanto et al., for example, experimented with different insertion configurations [33]. The microneedles were inserted by a specific impact device and then removed after 10 s, 10 min, and 4 h, respectively. Leaving needles in the skin for longer led to smaller reductions in blood glucose levels, and multiple insertions of the device created more pathways into the skin. In these studies the area of skin that was treated with microneedles was covered with an insulin-containing reservoir. Although some variations in transport were observed during these different application procedures, the general conclusion was that the microneedles created aqueous pathways. Utilizing this fairly simple pre-treatment mechanism, an overall pharmacodynamic response to insulin delivery was demonstrated, in which blood glucose levels decreased in a similar manner to subcutaneous hypodermic injection.

Table 8.1 compares the different microneedle configurations that were used in the rat model, with reference to the parameters illustrated in Figure 8.1. So far, there is no clinical evidence that the more precise shapes lead to a higher delivery efficacy. The device design depicted in category (III) of Figure 8.6 allows very simple and straightforward integration with a drug-filled adhesive patch by lamination [83]. The patch releases its cargo through openings in the back plate. The release profile will depend on the interplay of the type of incisions made by the spikes with the opening in the back plate, and varies quite significantly amongst different skin types or age groups of patients. It is difficult to say how robust the different design categories actually are for insulin delivery in the real world, and if they could produce the best type of incision. Further clinical research is required.

Table 8.1 Microneedle Arrays for Insulin Delivery in the Diabetic Rat Model

The various tips depicted in Figure 8.6 all present design varieties that create a “mechanical cut” and may be defined by their main function: “cutting”. However, their actual function is to release a controlled dose of insulin, and clinicians need to know how the exact size, type and geometric shape of the microconduits into the skin influence its delivery mechanism. When studying the pharmacodynamics of insulin release, one should not forget that the skin is an immunosensitive organ. When incisions are made in it, allowing foreign bodies to enter, an immune response may be triggered, which may alter the efficacy of insulin uptake, and also provoke unfavorable side-effects. This may also be an explanation for the smaller reduction in blood glucose levels observed by Martanto et al. when microneedles were inserted for a longer time or more than once [33]. Quality control of sterility, biocompatibility and efficacy of drug delivery are key for the final product.

Quantification of insertion and delivery processes, as well as the visualization of the incisions made by the individual types of microneedles in combination with the use of specific applicator devices, are therefore very important factors in these devices. A comparison on different types of insertion procedures and microneedle designs in transdermal drug delivery has been made by the Bouwstra group, but as yet not with respect to its effect on controlled insulin delivery [84]. The cost of statistically relevant clinical studies are high, and require a sustained supply of the devices which need approval by an ethical-medical committee. All the design varieties introduced and discussed above show a certain level of technological readiness that supports this request.

Clinical Perspective

From a clinical point of view, the microneedle array devices summarized schematically in Figure 8.6 are all of preliminary but potential interest, since the only studies that have used them have been small-scale animal trials (n < 10). Figure 8.9 shows two extreme design alternatives that have achieved a certain full proof-of-principle for insulin delivery, using out-of-plane microneedles as depicted in Figure 8.1(B) and (D). Figure 8.9(A) illustrates one such device integrated with a special actuator for the active release of insulin that is stored in a reservoir in liquid form. Figure 8.9(B) shows a simpler device, which releases insulin by a phase change, dissolving the polymeric material into the skin without any further active device components [78]. These self-dissolving microneedles encapsulate the insulin in a chondroitin sulfate matrix during micromolding of arrays, which is done by using flexible silicone (PDMS) molds. Both design alternatives demonstrate a viable solution for ultra-minimally invasive insulin release.

Computer-assisted design studies, in combination with the appropriate testing in a clinical laboratory, are needed to prove the microneedle concept. For diabetes care management, the devices introduced by Stemme’s group offer stop-and-go delivery, and keep the device attached to the skin [32, 43, 76, 79], which may be of particular importance during food intake or exercise. It is not yet clear which type of device offers the best number and type of pathways through the skin. One technical solution for the investigation of the pharmacokinetics of insulin is to use a hollow microneedle array together with sensor augmented active pumping. The device introduced by Roxhed et al. currently shows the highest level of integration for such a class of drug delivery device [43]. Measuring glucose by a user-independent integrated sensor is also in development, as briefly mentioned in the previous section [70]. The augmentation of the released insulin dosage will require such feedback parameters, which can significantly vary with the patient’s specific health status.

Any decision taken in the conceptual design phase of a new device will impact the type of experiments being carried out in drug delivery research. The choice of technology will therefore influence the data that can be collected for a clinical model. Insulin release by microneedle array pre-treatment has been taken to a fairly high level of testing, yet there are little or no data available that compare this novel transdermal mechanism of insulin release with the gold standard (bolus injection) within a clinically relevant number of human diabetic subjects. Although the working principle of insulin microneedle delivery has been confirmed by Gupta et al., utilizing a single, pulled glass needle at various insertion depths [49], there is still insufficient evidence to show that a delivery system equipped with an out-of-plane microneedle array can deliver insulin in a clinical application.

Delivery technology research can be carried out for the creation of a multitude of parallel injection sites with the preliminary devices presented here, but some level of concern remains regarding the degree of control that is exerted over the device insertion mechanisms. For example, applying high pressures during insertion will affect the skin properties, and thus the results of the transport studies. A delivery platform that includes sensor-augmented release is clearly desired, but at this stage of development many kinds of out-of-plane MNA devices can be already translated from the pure engineering sciences to research in drug delivery technology. Patient trials, especially those related to comfort, user-friendliness and pain sensation during the use of such devices, have by now demonstrated an appropriate level of technology readiness for clinical translational research. The initial studies by Gupta et al. [49], photographed the insertion sites to qualitatively evaluate the difference between injection depths. Much more indepth studies of the efficacy of MNA injections will be needed to prove their clinical effectiveness. For micromachined microneedles, new benchmark methods must be adopted to evaluate the clinical relevance of the different designs and materials. Both of the two concepts presented in Figure 8.9 provide a starting point for a further analysis of the specific design criteria for the application of MNAs in insulin delivery.

From the viewpoints of microengineering, drug delivery technology and clinical translation, the current microdevices show clearly that microsystems engineering presents certain benefits compared to those designs that use simple fine mechanical solutions or even mechanical assembly. A formal design approach can offer an exquisite tool for finding appropriate technological solutions for the development of transdermal insulin delivery, based on the aforementioned out-of-plane microneedle designs [85].