Design and Development of In Vivo Sensor Systems: The Long and Tortured Road to a Self-Contained, Implantable Glucose Sensor for Diabetes Management - Designing Receptors for the Next Generation of Biosensors - page 211

enzyme [ 9 - 13 ], fluorescent [ 14 - 18 ], and NIR spectroscopy [ 19 - 23 ]. Despite various
attempts, successful development of a fully functional implantable, noninvasive
continuous monitoring device has remained elusive due to critical deficiencies of
these detection techniques. Each method has physical and/or chemical limitations that
make them impractical for use in a long-term, implantable device. Enzyme-based
techniques function on reagents that are consumed and require a continuous reagent
supply during the process of detection. The by-products of the reagent reactions are
undesirable and cause detection interference. In addition, enzyme-based detection
techniques experience reagent degradation and inactivation over the long term,
eventually causing inaccurate readings and sensor drift [ 24 - 26 ]. Similar to problems
with enzyme-based techniques, there are also reagent limitations for long-term
fluorescence-based systems. Current fluorescence-based sensors cannot remain at an
implantation site and respond to blood glucose concentrations over an extended period
of time [ 27 ]. Over the lifetime of the sensor, denaturation, relaxation, or poisoning of
the fluorescent molecular recognition element occurs [ 28 ]. Gradual deterioration of
signaling reagents results in sensitivity and signal shifts that subsequently require
continual readjustment and calibration in order to achieve accurate measurement [ 7 ].
Using NIR spectroscopy to decipher glucose levels by way of absorption
measurements through or at tissues, however conceptually simple, is equally imprac-
tical. This approach is currently not acceptable for clinical use due to the fact that a
number of factors such as tissue hydration, blood flow, temperature, light scattering,
and overlapping absorption by non-glucose molecules cause read-out precision errors
[ 7 , 29 ]. It is no surprise that the search for the ideal glucose detection system continues
to motivate the scientific community. However, past efforts in designing an implant-
able and self-contained glucose sensing system have not been successful because
developers have given only partial consideration to the long-term impact and
limitations of the in vivo environment.
A technically and commercially successful implantable glucose sensor requires
the integrated design and development of several critical components (Fig. 1 ). The
mission-critical self-contained and closed-cycle sensing component must be
designed to interface with an appropriate signal transduction/signal processing
device that, in turn, is coupled to the sensor's electronics and communication
function. Further, the entire device must be enclosed in a porous, biostable, and
biocompatible material that simultaneously prevents biofouling of the device and
allows biotransport of the glucose analyte in and out of the device. Failure to
integrate any of these components into the implantable device invariably leads to
product development failure. Our integrated design for the implantable device, as
illustrated in Fig. 2 , envisions signal transduction using a MEMS cantilever [ 30 - 33 ]
that will respond to bound/unbound mass changes of the reporter construct with
subsequent processing of the resulting signal on a device-specific ASIC chip
[ 34 - 37 ]. Signal export to the external environment will be via RFID communica-
tion [ 38 - 40 ] with signal processing to provide the diabetic patient and their medical
team with glucose concentration and rate-of-change information both onboard the
RFID reader module and wirelessly exported to an external database. Additionally,
the biocompatible/biotransport membrane [ 41 , 42 ] will: (1) protect the device from