There is a growing understanding of the detailed reactive tissue responses to implantable probes. The early work in this area made important seminal observations concerning tissue responses to implanted microscale probes (Stensaas 1978; Agnew 1986; Woodford 1996) and the damage associated with inserting these devices into the brain (Edell 1992). Recent studies utilize sophisticated immunohistological procedures combined with advanced laser microscopy and visualization to provide finely detailed observations of cellular and chemical responses to implanted probes over multiple time scales (Shain 2003; Szarowski 2003). The basic working hypothesis suggests three different tissue responses: an immediate response that includes cellular and chemical events occurring at the time of insertion, early reactive responses representing the sequellae of events resulting from the insertion process itself, and sustained responses that are due to cellular interactions with the device itself (Figure B.3). The degree of the reactive responses seem to be a function of inter-related factors including tethering forces (Biran, Roy 2003) and insertion techniques (Szarowski 2003).

Insertion of microfabricated devices into the brain produces injury. The extent of injury is proportional to the cross-sectional area of the shaft of the device (Szarowski, et al., 2003). Other factors that may influence the extent of damage are tip design, rates of insertion, and shaft orientation as a function of insertion trajectory (Szarowski). The brain is highly vascular containing a matrix of small (<10 m diameter) capillaries. In rat brain, we have observed that the median distance between any one neuron and its nearest capillary is ~ 13 m (Lin et al., 2005; see, FIGURE 4.C.3.i). Numerous studies have demonstrated that a similar density exists in human cortex. Insertion of devices with shafts that have cross-sectional areas of 10,000 – 100,000 m², e.g. 15m x 60 m to 100 m x 100 m, will produce damage by stretching, breaking, and/or cutting vascular components. Leakage of blood-borne biomolecules or cells can initiate inflammatory processes that may result in transient loss of function, e.g. neuron excitability, or celldeath. The longer-term consequences is development a series of reactive responses (FIGURE B.i). The early reactive responses may result from the immediate damage produced at the time of insertion; however, the sustained reactive responses that are observed around devices for as long as they are inserted may be maintained by direct device-brain interactions. The sustained response is characterized by a compact sheath of cells and extracellular matrix material. This sheath completely surrounds inserted devices resulting in reduced device function and possibly compromised function of neurons immediately adjacent to the device. Controlling these consequences of device insertion may lead to improved device function. One of the goals of the Center is to develop multifunctional devices that will use device design to reduce responses and develop methods for local drug delivery to control these responses, thus ensuring long-term device function.

The field is in transition from being device-centered to being centered on the biological system and how the device interacts with that system. Perhaps one reason is that current empirical descriptions of successful—albeit limited—long-term probe function are juxtaposed with their contemporary reports of varying degrees of deleterious microscale tissue responses to implants. Probe developers are often asked, “Do your probes work?” Unfortunately, at this time, the most appropriate answer when factoring in detailed tissue responses is usually: “Yes, but no.” The dichotomy resulting from this transition is reflected in the current literature. Depending on their perspective, publications tend to either emphasize the fact that the device under study was found to work throughout the period of investigation (Serruya 2002; Kipke 2003; Nicolelis, M. A. 2003; Vetter In press) or that it ultimately failed or elicited a pronounced tissue response (Turner 1999; Szarowski 2003).   These two perspectives are likely to converge as the field progresses.

Most groups are working along several overlapping themes to improve long-term neural interfaces. While these themes are presented here in an approximate hierarchy, most of the work is actually done in an ad hoc manner following the expertise of the respective group. The first theme involves probe-centered design refinements that are intended to reduce insertion damage, reduce tissue reactive responses, and place recording sites in the appropriate target region. True optimization is difficult and time consuming. The typical strategy is to identify a probe configuration that works and then lock it in. Several designs have been considered or are under study, including making the penetrating shanks smaller (Stice 2003), optimizing their tip shapes (Woodford 1996; Hofmann 2002), varying their spacing, alternative materials (Edell Sept 1999), improving surgical techniques (Rousche, P.J. 1992; Maynard 2000; Biran, Roy 2003; Szarowski 2003), and developing coatings to passivate the probe surface (Ehteshami 2003). Among our group, Kipke investigated probe flexibility (Rousche, P. J. 2001) and surgical techniques (Vetter 2003). Martin is investigating the use of hydrogel coatings to make softer probes to reduce the mechanical mismatch with the soft brain tissue. Martin and Kipke are collaborating on recording sites coated with conductive polymers (Cui, X. 2001) as a means to improve signal transduction. With the exception of coatings, these types of probe refinements have resulted in the various types of ‘stock’ implantable probe systems used in neurophysiological studies today. There is a large body of biomaterial science attempting to ameliorate the foreign body reaction but most result in temporary and not a steady-state reduction in the tissue response (Ratner 2002). Early on in biomaterial science, material properties were given the most attention. More recently, others have begun to explore the importance of geometry and the potential damage induced by implant micromotion.