neuroprotection trek—the next generation : the measurement is the message

1

Ann. N.Y. Acad. Sci. 1053: 1–11 (2005). © 2005 New York Academy of Sciences.doi: 10.1196/annals.1344.001

Neuroprotection Trek—The Next Generation

The Measurement Is the Message

RUSSELL J. ANDREWS

Smart Systems and Nanotechnology, NASA Ames Research Center, Moffett Field, California, USA

ABSTRACT: Animal trials of many pharmacological neuroprotective agentshave been quite successful, whereas trials in humans have been uniformly dis-appointing. A major difference between laboratory research in animals andclinical research in humans is the amount and/or quality of data obtained. Thegoal of this presentation is to argue that when clinical studies consist of morevalid, objective data—that is, as our measurement capabilities in clinicalresearch become as robust as they are in laboratory research—we are likely togain new insights into both (1) injury to the nervous system and (2) neuro-protective treatment strategies. Technological advances (in data acquisitionand analysis)—often novel even in the laboratory—will be the “scale” that willenable progress in measurement. As examples of such technological advances,two projects initiated at NASA Ames Research Center are cited. The NASASmart Probe Project, with the goal of combining multiple microsensors andneural networks for real-time tissue identification (e.g., for tumor detection),has recently moved into the clinical realm, with a prototype being used to diag-nose breast cancer in women “on the spot”. The NASA Nanoelectrode ArrayProject has fabricated nanoscale devices that can simultaneously monitor elec-trical activity and neurotransmitter concentrations, while providing electricalstimulation focally and precisely (and potentially in a closed-loop fashion basedon the input from the nanosensors). The large amounts of data that such tech-niques can acquire and analyze—separated spatially and temporally throughoutthe nervous system, if necessary—will provide insights not only into neuro-protective strategies, but also into the workings of the nervous system itself.

KEYWORDS: cancer diagnosis; deep brain stimulation; electrical stimulation;nanotechnology; neuromodulation; neuroprotection; optical spectroscopy

INTRODUCTION

Why have so many pharmacological neuroprotective agents that appeared quiteeffective in stroke or brain injury in animal models proven to be so ineffective inhuman trials? This topic was addressed in part by my contribution to the FourthInternational Conference on Neuroprotective Agents (1998), where the argumentwas made that failure to achieve neuroprotection in humans was in large part due to

Address for correspondence: Russell J. Andrews, 555 Knowles Drive (#112), Los Gatos, CA95032. Voice: 408-829-1700; fax: 408-866-8842.

[email protected]

2 ANNALS NEW YORK ACADEMY OF SCIENCES

(1) our failure to understand the complexity of the “ischemic cascade”—the almostcertain need for a multidrug “cocktail” approach to neuroprotection—and (2) thepolitics of the pharmaceutical industry in betting on short-term success with a singleagent “magic bullet” drug that might turn a substantial profit before patent expira-tion. At that time (1998), it appeared that technological advances were making agreater contribution to clinical neuroprotection than pharmacological advances—two examples given were cardiovascular stenting and deep brain stimulation formovement disorders.1

The present contribution carries on that theme of technology driving neuro-protection. The argument is that, in the laboratory, the data are “cleaner” (if not moreaccurate) and the endpoints are more quantitative (e.g., “percent of the rat hemi-sphere infarcted”, in contrast to “Uncle Harry’s ability to function independentlyafter his stroke”). However, we are now gaining some tools that may enable us torealize more substantial progress in clinical neuroprotection. The theme is that dra-matic improvements in our ability to measure the events of neural injury and neuro-degeneration, as well as measure the neurorepair and neuroregeneration followingintervention, should result in substantial improvement in clinical neuroprotection fora variety of disorders from stroke to movement disorders to Alzheimer’s disease.

MARSHALING THE EVIDENCE FOR MEASUREMENT

“Evidence-based” has become a mantra of modern medicine, an incantationchanted by the medical literati and the health maintenance organizations (HMOs)alike. Research “evidence” has been categorized into four classes:2

• Class I: prospective, randomized, controlled trials;

• Class II: case control, cohort, prevalence studies;

• Class III: retrospective series, registries, databases;

• Class IV: case reports, “expert” opinions.

In class I, measurement accuracy is certain; in contrast, in class IV, measurementaccuracy is uncertain. Laboratory studies entail well-controlled, well-measuredresearch conditions, well-measured treatments (e.g., amount of drug, blood/brainlevels, duration of treatment), and well-measured outcomes (e.g., volume of braininfarcted). Clinical studies tend to be messy (even if not tainted by political oreconomic pressures)—or, if well measured, very expensive.

Consider the following quote, and the adaptation—by substituting measurementfor electric light :

“The electric light is pure information. It is a medium without a message.… Whetherthe light is being used for brain surgery or night baseball is a matter of indifference. Itcould be argued that these activities are in some way the ‘content’ of the electric light,since they could not exist without the electric light. This fact merely underlines thepoint that ‘the medium is the message’ because it is the medium that shapes andcontrols the scale and form of human association and action.”3

MARSHALL MCLUHAN, Understanding Media: TheExtensions of Man, 1964

“Measurement is pure information. It is a medium without a message.… Whether mea-surement is being used for brain surgery or night baseball is a matter of indifference. It

3ANDREWS: MEASUREMENT AND NEUROPROTECTION

could be argued that these activities are in some way the ‘content’ of measurement,since they could not exist without measurement. This fact merely underlines the pointthat ‘measurement is the message’ because it is measurement that shapes and controlsthe scale and form of human association and action.”

With Apologies to Marshall McLuhan

The effect of the “information explosion” (e.g., the Internet) needs little elabora-tion. With a click, we can “google” any notion that can be expressed as keystrokes.The ability to marshal large amounts of objectively acquired data has fueled themove toward evidence-based medicine. To adapt another quote:

“In the fields of observation, chance favors only the prepared mind.”

LOUIS PASTEUR, Inaugural Lecture,University of Lille, 1854

“In research, publication and funding favor only the well-measured data (and lots of it!).”

With Apologies to Louis Pasteur

In the following sections, two projects at NASA Ames Research Center are con-sidered. The common theme is the use of “cutting-edge” technology to gather largeamounts of data—objectively, accurately, minimally invasively, and dispersedspatially and temporally—which enables us to measure with much more validity thevarious processes that are occurring in tissues that are undergoing (or have under-gone) undesirable changes (e.g., ischemic, neoplastic, neurodegenerative). The firstis the NASA Smart Probe Project, originating with the Smart Systems Group atNASA Ames and carried into clinical trials by the start-up company, BioLuminate(Dublin, CA). The second is the NASA Nanoelectrode Array Project, involving theNanotechnology Group at NASA Ames and the nearby Parkinson’s Institute(Sunnyvale, CA).

THE SMART PROBE PROJECT FOR REAL-TIMETISSUE RECOGNITION

The NASA Smart Probe Project and the refinements developed by the NASAtechnology licensee, BioLuminate (Dublin, CA), demonstrate that a unique “signa-ture” for any tissue—indeed, a unique “signature + address” (e.g., the substantianigra pars compacta of the brain)—can be acquired in real time.4

The Smart Probe combines continuous data streams from multiple microsensors,with the data being processed in real time by neural network/fuzzy logic algorithms(FIG. 1). Various “off-the-shelf” microsensors, each < 1 mm in diameter, have beentested at NASA Ames in rodents: a microstrain gauge, a laser-Doppler blood flowprobe (Vasamedics, St. Paul, MN), a fiber-optic neuroendoscope (Codman/Johnson& Johnson, Raynham, MA), a combination CO2/O2/pH monitor probe (NeuroTrend,Codman/Johnson & Johnson, Raynham, MA), a standard microelectrode, and a light-scattering spectroscopy probe (PC2000, Ocean Optics, Dunedin, FL). Various typesof optical spectroscopy have been shown to be extremely powerful in differentiatingtissues in vivo; typical spectra for several tissues using light-scattering spectroscopyare illustrated in FIGURE 2.

The BioLuminate probe for breast cancer diagnosis is illustrated in FIGURE 3.Breast “biopsies” in 24 women with suspected breast cancer have been performed at


FIGURE 1. NASA Smart Probe schematic.

FIGURE 2. Rodent in vivo tissue spectra, 350–900 nm. From top to bottom at 475 nm:brain, nerve, fat, artery, muscle, blood.


the University of California, Davis and San Francisco, Medical Centers; the resultswere compared with histopathological samples acquired by standard breast biopsytechniques. The probe simultaneously assesses (1) oxy- and deoxyhemoglobin con-centrations (using infrared and blue laser spectroscopy), (2) broadband (white light)spectroscopy, and (3) electrical impedance (FIG. 4). With a repetition rate of 100times per second, more than 500 MB of data are collected for each patient (fromthree or more probe tracks per patient through the lesion and adjacent tissues).

FIGURES 5–7 present data collected by BioLuminate. FIGURE 5 presents white(broadband) spectroscopy data for 2 women with breast carcinoma, contrastingtumor oxyhemoglobin at the edge versus the center. FIGURE 6 presents data contrast-ing the edge and the center for both blue laser (fluorescence) and white (broadband,oxyhemoglobin fraction) spectroscopy. FIGURE 7 presents impedance data in normal

FIGURE 3. BioLuminate probe for breast biopsy: (left) comparison with standardbreast biopsy needle; (right) end view of beveled tip of needle (<1 mm diameter).

FIGURE 4. BioLuminate probe sensors. Left to right: infrared laser spectroscopy sen-sor, white light (broadband) and blue laser spectroscopy sensors, and electrical impedancesensor.


FIGURE 5. Broadband spectroscopy data from 2 patients (D-007 and D-012) withhistologically verified breast carcinoma.


FIGURE 6. Patient with histologically verified infiltrating ductal breast carcinoma,grade III, 2.2 cm diameter: (top) blue laser (fluorescence); (bottom) white/broadband (oxy-Hbfraction).

FIGURE 7. Impedance measurements in a patient with histologically verified breast carcinoma.


breast, edge of breast carcinoma, and center of breast carcinoma over the frequencyrange of 2 to 2000 kHz, with the maximum tissue differentiation being at the lowestfrequency (2 kHz).

NANOELECTRODE ARRAYS FOR SIMULTANEOUSNEUROTRANSMITTER AND ELECTRICAL ACTIVITY

MONITORING, PLUS ELECTRICAL STIMULATION

Neuromodulation—the use of controlled electrical stimulation of the central orperipheral nervous system—is a treatment modality being explored for variousnervous system disorders, for example, movement disorders, epilepsy, chronic pain,depression, and eating disorders. Three of the most common forms of neuromodula-tion—deep brain stimulation, vagus nerve stimulation, and transcranial magneticstimulation—were reviewed in the last conference.5 At present, each of these tech-niques is an “open-loop” system, that is, the stimulation is performed without thebenefit of input from the brain to guide the timing or characteristics of the stimula-tion. There is evidence from neuromodulation for intractable epilepsy that a “closed-loop” system (with monitoring of the brain’s electrical activity—to detect animpending seizure) can greatly increase the effectiveness of neuromodulation.6

Combining continuous monitoring of neurotransmitter levels and electrical activitywith precise focal electrical stimulation [electrochemical closed-loop neuromodula-tion (ECN)] would be a significant advance over currently available techniques.Additionally, reducing the scale to the nanolevel (1) improves signal-to-noise ratios,(2) permits greater precision (down to the subnuclei level), and (3) opens the possi-bility of multiple recording/monitoring and stimulation sites throughout the centralnervous system.

FIGURE 8. Nanoelectrode array fabrication. See text for abbreviations and details.


The NASA Ames Nanotechnology Center is developing microchip nanoelectrodearrays for neuromodulation that exhibit the above properties.7 The microchipconsists of two types of vertically aligned multiwalled carbon nanotube (MWCNT)arrays on multiple individually addressed microelectrode pads. The first type is aforestlike MWCNT array that presents a large surface area and is used as a stimula-tion electrode. The second type is encapsulated with an insulator, leaving only thevery end of the MWCNTs exposed at the surface to form an inlaid nanodiskelectrode array.

A schematic for the fabrication of the nanoelectrode arrays is given in FIGURE 8.To a silicon wafer, a nickel catalyst film is deposited. The vertically alignedMWCNT arrays are grown on the nickel catalyst by a plasma carbon vapor deposi-tion (CVD) process. For the uninsulated stimulation electrode, the fabrication stopsat this point. For the insulated recording electrode (electrical and electrochemical),two more steps are required. First, a silicon oxide layer is created by tetraethylortho-silicate (TEOS) CVD. Next, chemical mechanical polishing (CMP) removes theinsulation so that only the tips of the MWCNT arrays are exposed. FIGURE 9 illus-trates both the insulated and uninsulated electrodes, plus one potential configurationwith a single recording electrode surrounded by eight stimulating electrodes.

The insulated nanodisk electrode array has shown extraordinary electrochemicalproperties, with a detection limit of redox species down to a few nanomolars and anextremely high temporal resolution down to milliseconds, which are ideal formeasuring simultaneously (1) extracellular neurotransmitters (e.g., dopamine) and(2) focal electrical activity. The uninsulated array is ideal for focal electrical stimu-lation, potentially as a focal (or regional) closed-loop stimulation system dependingupon variations in the monitored electrical activity and/or neurotransmitter concen-tration (ECN). FIGURE 10 summarizes these advantages of nanoelectrode arrays.

FIGURE 9. Uninsulated and insulated carbon nanotube (CNT) arrays. Scanning elec-tron micrograph (SEM) image of (a) a 3 ×3 microelectrode array, (b) a high-magnificationimage of exposed CNT array electrodes used for electrical stimulation, and (c) an embeddedlow-density CNT array used for recording changes in local neurotransmitter concentrations.Scale bars: (a) 200 µm; (b) 1 µm; (c) 2 µm.


CONCLUSIONS

The NASA Ames Smart Probe and Nanoelectrode Array Projects represent stepstoward the goal of using “cutting-edge” technology to gather large amounts ofdata—objectively, accurately, minimally invasively, and dispersed spatially andtemporally. The Smart Probe is primarily a software concept of employing neuralnetworks and fuzzy logic techniques to integrate multiple data streams in real timeto provide immediate feedback—in the form of either a tissue diagnosis or informa-tion to guide an intervention (e.g., a closed-loop neuromodulation system). The Nano-electrode Array allows the simultaneous and continuous gathering of data—bothelectrical and electrochemical (e.g., neurotransmitters)—from multiple sites in thebrain. When combined with neural network software, the Nanoelectrode Array alsocan form both the recording and stimulation ends of a closed-loop neuromodulationsystem (ECN).

A major difference between laboratory and clinical neuroprotection trials is thequantity and precision of the data collected in the former in comparison with thelatter. With advances such as the Smart Probe and the Nanoelectrode Array, themeasurement of neuroprotection processes and outcomes in clinical trials may even-tually rival those of laboratory studies. As we trek beyond the neuroprotection “basecamp”, advances in measurement may prove to be one of the trustiest “guides”.Importantly, just as a seasoned guide may lead to unexpected vistas, advances inmeasurement will likely lead to novel neuroprotection insights and strategies.

FIGURE 10. The spatial, temporal, and sensitivity advantages of nanoelectrodes incomparison with microelectrodes.


ACKNOWLEDGMENTS

The projects reported here involved the following individuals: R. Mah, S. Jeffrey,M. Guerrero, R. Papasin, and C. Reed from NASA Ames Smart Systems; J. Li,H. Chen, B. Nguyen-Vu, M. Meyyappan, A. Cassell, J. Koehne, H. Purewal, andO. Ye from NASA Ames Nanotechnology; L. DaSilva and U. Kasthuri from Bio-Luminate (Dublin, CA); and N. Huang and L. Chen from the Parkinson’s Institute(Sunnyvale, CA).

REFERENCES

1. ANDREWS, R.J. 1999. Neuroprotective “agents” in surgery: secret “agent” man, orcommon “agent” machine? Ann. N.Y. Acad. Sci. 890: 59–72.

2. KNOPMAN, D.S., S.T. DEKOSKY, J.L. CUMMINGS et al. 2001. Practice parameter: diagnosisof dementia (an evidence-based review)—Report of the Quality Standards Sub-committee of the American Academy of Neurology. Neurology 56: 1143–1153.

3. MCLUHAN, M. 1964. Understanding Media: The Extensions of Man. McGraw–Hill.New York.

4. ANDREWS, R., R. MAH & L. DASILVA. 2004. The NASA Smart Probe for real timemultiple microsensor tissue recognition. Proc. SPIE (Prog. Biomed. Opt. Imag. Opt.Biop. V) 5326: 92–97.

5. ANDREWS, R.J. 2003. Neuroprotection trek—the next generation: Neuromodulation I.Techniques—deep brain stimulation, vagus nerve stimulation, and transcranialmagnetic stimulation. Ann. N.Y. Acad. Sci. 993: 1–13.

6. ANDREWS, R.J. 2003. Neuroprotection trek—the next generation: Neuromodulation II.Applications—epilepsy, nerve regeneration, neurotrophins. Ann. N.Y. Acad. Sci.993: 14–24.

7. NGUYEN-VU, B., H. CHEN, A. CASSELL et al. 2004. Carbon nanotube nanoelectrode arrayfor electrophysiology. Presented at the NIH Workshop on Neural Interfaces,Washington, D.C.

neuroprotection trek—the next generation : the measurement is the message

Documents