Effects of high-intensity focused ultrasound on the intervertebral disc: A potential therapy for...

ABSTRACT: The effects of various exposures (intensity, duration) of high-intensity focused ultrasound (HIFU) on sciatic nerve conduction were inves-tigated in vivo in rats. The objective was to identify HIFU exposures thatproduce biological effects ranging from partial to complete conduction block,indicating potential use of HIFU as an alternative to current clinical methodsof inducing nerve conduction block. In the study, 26 nerves were exposedand treated with 5-s applications of 5.7-MHZ HIFU with acoustic intensities of390, 2,255, 3,310, and 7,890 W/cm2 (spatial peak, temporal peak). Com-pound muscle action potentials (CMAPs), in response to electrical stimula-tion of the nerve proximal to the HIFU site, were recorded from the plantarfoot muscles immediately before and after HIFU treatment and 2 and 4 hafter treatment. Furthermore, a preliminary long-term investigation was per-formed on 27 nerves with the same four sets of HIFU parameters. CMAPswere measured at the survival endpoint for each animal (7 or 28 days aftertreatment). For nerves treated with the three lower exposures, CMAPsdecreased initially within 4 h or 7 days after HIFU treatment and thenrecovered to their baseline level at 28 days after treatment. For the highestexposure, however, CMAPs remained absent even 28 days after treatment.These exposure-dependent effects of HIFU on nerve function suggest itsfuture potential as a novel treatment for severe spasticity and pain.

Muscle Nerve 37: 241–250, 2008

EFFECTS OF HIGH-INTENSITY FOCUSEDULTRASOUND ON NERVE CONDUCTION

JESSICA L. FOLEY, PhD,1 JAMES W. LITTLE, MD, PhD,2 and SHAHRAM VAEZY, PhD1

1 Department of Bioengineering, Box 355061, University of Washington, Seattle,Washington, 98195, USA

2 Department of Rehabilitation Medicine, University of Washington, Seattle, Washington, USA

Accepted 28 September 2007

Spasticity—manifest as uncontrollable muscle con-tractions—and pain are significant complications ofneurological disorders and injuries. Chronic pain isalso associated with various other conditions includ-ing cancer and arthritis. Common treatments ofspasticity and pain include chemical nerve conduc-tion blocks,2,8,9 surgical interventions,14 muscleblocks with botulinum toxin,3,4 and intrathecal ba-clofen injections.1 These invasive methods do nothelp all patients. There remains a clinical need toinvestigate novel alternative treatment methods.HIFU is a therapeutic modality that can noninva-sively achieve focal heating and mechanical disrup-tion of small volumes of internal tissue without dam-aging intervening tissues. By employing this uniquecapability of HIFU to induce a therapeutic effect in

deep-seated tissues, the goal of our research hasbeen to develop a novel method of nerve disruptionthat could potentially provide an effective treatmentfor severe spasticity and certain types of chronicpain. The range of effects, from partial to completeconduction block or axon degeneration, which canbe achieved with varying HIFU exposure, was inves-tigated.

HIFU can be applied percutaneously to achievefocal heating and necrosis of a small volume ofdeeply seated internal tissue.16 HIFU systems pro-duce acoustic intensities greater than 1,000 W/cm2

at the focus, 4 to 5 orders of magnitude higher thanthe intensities used in ultrasound imaging systems(10 to 100 mW/cm2). The great advantage of HIFUis that it can focus energy to a small region (�0.5 �5 mm ellipsoid) inside the body without damagingintervening tissue.18

The biological effects of HIFU are dependent onthe specific tissue being treated and the parametersof HIFU operation, including the intensity of theHIFU beam and the duration of the treatment.HIFU application to a specific structure within the

Abbreviations: CMAP, compound muscle action potential; HIFU, high-intensity focused ultrasoundKey words: conduction block; HIFU; sciatic nerve; therapy; ultrasoundCorrespondence to: S. Vaezy; e-mail: [email protected]

© 2007 Wiley Periodicals, Inc.Published online 26 November 2007 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mus.20932

HIFU Effects on Nerve Conduction MUSCLE & NERVE February 2008 241

body (e.g., nerve) can be described as the HIFUexposure. In this article, HIFU exposure is repre-sented as the ultrasound peak intensity (spatial peak,temporal peak), in W/cm2, and the duration ofexposure in seconds. It is important to investigatethe role of HIFU exposure in the production ofbiological effects in peripheral nerve in order todevelop the most optimal HIFU systems and treat-ment protocols for specific clinical applications.

Previous ex vivo studies have used electrophysio-logical methods to investigate the effects of ultra-sound on action potentials in peripheralnerves.12,15,19 These studies demonstrated that fo-cused ultrasound at intensities above 35 W/cm2 canbe used to suppress axonal conduction in nerves; theextent and duration of suppression varied depend-ing on the HIFU exposure. Work by Lele,10 in bothex vivo and in vivo models, also indicated exposure-dependent effects of ultrasound (frequencies of 600kHz, 900 kHz, and 2.7 MHz) on nerve conduction;specifically, the studies showed the important role ofthermal exposure where higher temperatures re-sulted in more pronounced suppression of conduc-tion. However, this study did not investigate thelong-term effects on nerve conduction or those inresponse to ultrasound of higher frequency, perhapsmore suitable for superficial application to nerves.The study also omitted details of the reported treat-ment of one exposed human ulnar nerve, raisingconcerns about this work and indicating a need forfurther in vivo investigations on the effects of fo-cused ultrasound on the function of neural tissues sothat this technology can be appraised for its clinicalutility. Our previous in vivo investigations have de-termined the feasibility of ultrasound image-guidedHIFU to target and block the sciatic nerves of rab-bits. HIFU with a frequency of 3.2 MHz, 1480–1930W/cm2 intensity, and duration of 5–40 s producedcomplete conduction block in 34 rabbit sciaticnerves.5 Ultrasound imaging was used for targetingand monitoring of the HIFU treatment. In long-terminvestigations, such blocks resulted in distal axondegeneration by 7 or 14 days after HIFU treatment.7

The objective of the work reported here was toinvestigate the short-term and long-term exposure-dependent effects of HIFU on nerve conduction.The sciatic nerves of rats were treated with HIFU ofseveral different exposures using a visually guidedapproach. Specifically, the change in amplitude ofcompound muscle action potentials (CMAPs) in re-sponse to stimulation of nerves treated at all HIFUexposures was observed. The hypothesis was thatHIFU treatment would lead to suppression of CMAPamplitude, which would be more pronounced with

higher exposures of HIFU. There has been no pre-viously reported work investigating the long-termexposure-dependent effects of HIFU on mammaliannerves in vivo, at least to our knowledge. Therefore,the initial work reported here is important as a firststep in identifying the full range of effects of HIFUon nerve conduction.

MATERIALS AND METHODS

HIFU Device. The HIFU device was a 5.7-MHz sin-gle-element transducer with a solid titanium coneattachment, developed for visually guided HIFU ap-plications.11 The titanium cone, attached to the faceof a spherically curved piezoelectric crystal (PZT),acted as a coupling medium between the transducerand the nerve. The cone extended along the HIFUbeam path to the focus so that the focal zone was justbeyond the tip of the cone (�4 mm). This enabledvisual guidance to position HIFU application in thesame location on each nerve (Fig. 1).

It was important that the focal cross-sectionalarea of the HIFU beam was positioned on the nerveso as to enable equal distribution of energy acrossthe nerve. The transducer provided an HIFU focuswith a lateral cross-section no greater than 1 mm.The lateral cross-section of the rat sciatic nerves were�1 to 1.5 mm and a consistent exposure may nothave been received by different portions of thenerve. Therefore, the HIFU focus was positioned inthe center of each nerve (in its lateral dimension) toprovide consistency throughout the studies (Fig. 1c).Thermal conduction of the HIFU focus would allowtreatment of the entire nerve.

FIGURE 1. Visually guided HIFU application to rat sciatic nerve(schematics not drawn to scale). (A) Overview of HIFU applica-tion, (B) close-up view of distance between tip of titanium coneapplicator and HIFU focus, (C) HIFU device positioned such thatfocus is in center of nerve in its lateral dimension.

242 HIFU Effects on Nerve Conduction MUSCLE & NERVE February 2008

The acoustic pressures around the HIFU focuswere measured using a needle hydrophone(TNU001A; NTR Systems, Seattle, Washington) with0.6-mm active diameter. The �6 dB full-width, half-maximum acoustic pressure points were determinedfor axial and lateral beam dimensions based on thepeak pressures measured and the boundaries corre-sponded to the following focal dimensions: 4.4 mm(axial) and 0.95 mm (lateral). The range of electri-cal powers supplied by the system was 4 to 115 W.The transducer had an efficiency of �50%, specifiedas the conversion from input electrical power toacoustic power at the focus (tip of titanium cone);acoustic power was measured using a reflective forcebalance method.11 Using the lateral focal area, cal-culated via hydrophone measurements to be 0.007cm2, the range of acoustic powers corresponded to arange of acoustic intensities of 280 to 8,200 W/cm2

(ISPTP, spatial peak, temporal peak).

Short-Term Study. All procedures were carried outaccording to guidelines of the National Institutes ofHealth (NIH) on the use of laboratory animals andthe procedures were approved by our institutionalAnimal Care Committee. Using a visually guidedapproach, 26 rat sciatic nerves (13 Sprague–Dawleyrats, 350–450 g) were treated with four differentHIFU exposures: 5-s applications with acoustic inten-sities of 390, 2,255, 3,310, and 7,890 W/cm2. Each ratwas anesthetized using an intraperitoneal injectionof ketamine (60 mg/kg) and xylazine (7 mg/kg).The animal was placed on a heating pad on theoperating table to enable maintenance of body tem-perature throughout the procedure. An incision inthe skin on the lateral side of each rat leg was made

to expose the lateral hamstring muscle. The sciaticnerve of each leg was exposed by loosening theconnective tissue between the medial and lateralhamstring muscles. An active stimulating needleelectrode (Teflon-coated except for the exposed tip;Teca, DMG50, Oxford Instruments Medical, Haw-thorne, New York) was placed in direct contact withthe nerve at a level just distal to the sciatic notch; theground electrode was placed in adjacent muscle(Fig. 2). The stimulating electrode delivered a 10%–20% supramaximal stimulus to the nerve: 1–1.2 V,0.5-ms pulse duration, 1 Hz (supplied by a Grass S48stimulator; Grass Technologies, West Warwick,Rhode Island). These pulse parameters would en-able stimulation of all of the nerve fibers and record-ings that represented nerve or muscle response toeach individual stimulus pulse. The CMAPs wererecorded from the plantar foot muscles immediatelybefore and after HIFU treatment and 2 h and 4 hafter treatment; the active recording electrode wasplaced on the plantar foot pad just proximal to themetatarsal joint and the reference electrode wasplaced on top of the toes (Fig. 2). The HIFU devicewas mounted on a 3D stereotaxic positioning systemand positioned �3 mm above the center of thenerve, using a needle guide, so that the focus wouldbegin on the underside of the nerve and progressupward through the entire nerve (Fig. 3). A needlewas attached to the transducer such that its tip was ata distance 3 mm from the focus; the tip was slightly

FIGURE 2. Positioning of stimulating and recording electrodes.Ground electrode positioned near hip bone and stimulating elec-trode positioned on sciatic nerve, proximal to HIFU treatment site.HIFU treatment site located on portion of nerve proximal tobranching to smaller nerve divisions.

FIGURE 3. Positioning of HIFU transducer, using needle guide,to focus on nerve.


off axis from the beam so as not to interfere with theacoustic energy focusing. To achieve the optimalplacement of the HIFU focus the device was posi-tioned so that the needle just touched the outer sideof the nerve. The depth of the needle tip was posi-tioned at the center of the nerve in this same dimen-sion. This method of positioning also placed thecenter of the HIFU beam, in its lateral dimension, inline with the lateral center of the nerve. This proce-dure for alignment of the HIFU transducer resultedin direct treatment of the central portion of thenerve beginning �1 mm below the nerve and prop-agating upward through the entire depth of thenerve. Thermal conduction enabled treatment ofthe entire nerve.

Immediately before and after HIFU the CMAPswere recorded, displayed on an oscilloscope, andsaved on a laptop computer (using ScopeExplorersoftware; LeCroy, Chestnut Ridge, New York). Theanimal was monitored over 4 h with CMAPs re-corded and saved at both 2 h and 4 h. The statisticaldifferences of relative CMAP amplitude between thedifferent exposure groups at 0, 2, and 4 h werecalculated using two-tailed, two-sampled t-tests as-suming unequal variances. Statistically significantdifferences were indicated by P-values less than 0.05.After 4 h the animal was euthanized (intraperitonealinjection of 120 mg/kg of sodium pentobarbital)and a section of the nerve at the lesion site and anormal (untreated) section were harvested and im-mersed in formalin fixative. Representative sectionsfrom each exposure group were embedded in par-affin, sliced in 1-�m sections, and stained with Mas-son’s trichrome. Nerve cross-sections were studiedunder low- and high-power light microscopy.

Long-Term Study. We chose to use the same fourHIFU exposures used in the short-term study: 5-sapplications with acoustic intensities of 390, 2,255,3,310, and 7,890 W/cm2. We expected that thisrange could enable different types of long-term out-comes: (1) neurapraxia (i.e., conduction block per-haps due to demyelination; axons do not degenerateand conduction block resolves perhaps due to axonremyelination); (2) axonotmesis (i.e., axons degen-erate but framework of connective tissue remainsand axons eventually regenerate); or (3) neurotme-sis (i.e., axons degenerate and framework of connec-tive tissue is also disrupted so axons do not regener-ate completely).

Sprague–Dawley rats (n � 27; 350–450 g) wereused in this study. A total of 13 animals were ob-served for 7 days and the remaining 14 animals for28 days following HIFU treatment. For the 7-day

group, three animals were treated with each of thefour HIFU exposures and one underwent shamtreatment (same procedure except no HIFU). Simi-larly, for the 28-day group of 14 animals, three weretreated with each of the four HIFU exposures andtwo underwent sham treatment.

Each animal was anesthetized, the right sciaticnerve was exposed, and a baseline CMAP was mea-sured. The position of the stimulating needle elec-trode with respect to the major branchpoint of thenerve (into tibial and peroneal divisions) was mea-sured for each nerve. This distance was recorded toenable consistent placement of the electrode duringstimulation at the time endpoint for the animal. TheHIFU device was stereotaxically positioned so thatthe focus extended from the underside of the nerveupward through the nerve. The position of the HIFUtreatment site with respect to the major branchpointof the nerve (into tibial and peroneal divisions) wasmeasured for each nerve. This distance was recordedto enable harvesting of the lesioned nerve segmentat the time endpoint for the animal. For the shamtreatment group the nerve was exposed but no HIFUwas applied to the nerve.

Following HIFU or sham treatment of the nervethe CMAPs of the foot muscles were measured againto denote immediate effects on nerve conduction.The two muscle layers and the skin were then su-tured. Buprenorphine, an analgesic, was adminis-tered intraperitoneally to prevent pain during theanimal’s recovery. The rat was allowed to recoverfrom anesthesia, returned to the cage, and moni-tored at regular intervals (twice per day for first 5days, once per day for remaining days). If at any timeduring the posttreatment period the animal ap-peared in distress or pain, or upon recommendationof the veterinary staff, the animal was euthanizedwith an intraperitoneal injection of 120 mg/kg ofsodium pentobarbital. For all animals, CMAP record-ings were performed before euthanasia.

Observations of the animal’s behavior periodi-cally during the survival phase were also used asqualitative descriptions of the HIFU effects on nerveconduction. Rat gait and hindlimb use were ob-served and documented on 3 to 5 different post-HIFU days (for 28-day animals) or on 1 to 2 post-HIFU days (for 7-day animals). Photos of the rathindlimb when the rat was held stationary weretaken at each of these times. The distance betweenthe outer two toes of the treated hindlimb (toe-spread) was also measured periodically. Qualitativecharacteristics of the rat gait and behavior such aslimp, ability to stand on hindlimb, grooming, andself-mutilation were documented at each timepoint.


Sham-treated animals also underwent behavioral ob-servations at these times.

At the designated endpoint for each animal, ratsunderwent general anesthesia (intraperitoneal injec-tion of ketamine, 60 mg/kg, and xylazine, 7 mg/kg)and the sciatic nerves of both legs (treated anduntreated) were exposed surgically. Stimulating andrecording electrodes were placed and CMAPs wererecorded from the foot muscles. The stimulatingelectrode was placed at the same position on thenerve as it was for stimulation before HIFU applica-tion. CMAP amplitude was measured using a MatLabprogram (MathWorks, Natick, Massachusetts); “rela-tive CMAP amplitude” was calculated as CMAP am-plitude at a given time post-HIFU divided by thebaseline CMAP amplitude pre-HIFU. The statisticaldifference in the relative CMAP amplitude betweengroups treated with the same HIFU exposure butmeasured at different time endpoints (4 h, 7 days, 28days) was calculated using two-tailed, two-sampledt-tests assuming unequal variances. For all t-tests astatistically significant difference was indicated by aP-value of less than 0.05. The rat was then eutha-nized with an intraperitoneal injection of 120 mg/kgof sodium pentobarbital and sections of the nerve atthe lesion site (and at sites distal and proximal) wereharvested and immersed in formalin fixative. A seg-ment from the nerve of the contralateral hindlimbwas taken as a control section. Representative sec-tions from each exposure group were embedded in

paraffin, sliced in 1-�m sections, and stained withMasson’s trichrome. Nerve cross-sections were stud-ied under low- and high-power light microscopy andHIFU effects were described qualitatively.

RESULTS

Short-Term Study. Relative CMAP amplitude de-creased with increased HIFU exposure (Figs. 4, 5).There was a statistically significant difference be-tween the relative CMAP amplitude measured 4 hafter HIFU for the intermediate exposures (2,255and 3,310 W/cm2; 5 s) and the low exposure (390W/cm2; 5 s) (P � 0.002, P � 0.005) and between theintermediate exposures (2,255 and 3,310 W/cm2;5 s) and the high exposure (7,890 W/cm2; 5 s) (P �0.001, P � 0.005). However, there was not a signifi-cant difference between the relative CMAP ampli-tudes of the intermediate exposures (2,255 and3,310 W/cm2; 5 s) (P � 0.43). For each exposurethe relative CMAP amplitude decreased slightly overtime. At the highest exposure the nerve was com-pletely blocked immediately after HIFU treatment(in three of four nerves) and did not recoverwithin 4 h.

Histologically, for nerves treated with the highestexposure (7,890 W/cm2; 5 s), there was a clear dis-tinction between the treated nerve sections and con-trol (Fig. 6). The changes present in the HIFU-treated sections included axon swelling, brokenmyelin accumulation in the center of fibers (Fig. 6D,thin arrow), disorganized fiber structure (Fig. 6B,arrow), and vacant regions (possibly indicating celldeath, Fig. 6D, thick arrow). For nerves treated withthe lower exposures, histological differences in the

FIGURE 4. (A) Baseline signal recorded from plantar muscles inresponse to stimulation of sciatic nerve; stimulus artifact andCMAP response identified. (B) Up-close view of baseline CMAPsignal; peak-to-peak amplitude shown. (C) CMAP signal re-corded from plantar muscles in response to electrical stimulationof sciatic nerve 4 h after 5-s exposure of ultrasound of 3,310W/cm2 applied to sciatic nerve; observe diminished peak-to-peakamplitude.

FIGURE 5. Relative CMAP amplitude (CMAP amplitude at time-point post-HIFU/baseline CMAP amplitude) corresponding tonerves treated by four different HIFU exposures, measured attimepoints of 0, 2, and 4 h after ultrasound treatment. Error barsequal SEM.


HIFU-treated and control untreated nerve sectionswere less apparent.

Long-Term Study. Mean relative CMAP decreasedwith increasing HIFU exposure with 7-day survival(Fig. 7A). For nerves treated with the highest expo-sure (7,890 W/cm2; 5 s) there was no measurableCMAP response at 7 days after HIFU treatment. At28-day survival, mean relative CMAP was similar fornerves treated with the three lower exposures, allnear 1, indicating a CMAP response near the base-line response (Fig. 7B). However, for nerves treatedwith the highest exposure (7,890 W/cm2; 5 s) therewas no measurable CMAP response at 28 days follow-ing HIFU treatment. For the lowest exposure theCMAP amplitude at 7 and 28 days was not muchdifferent than the baseline CMAP. For the interme-

diate exposures there was variability in the long-termeffects. For some animals the CMAP fully recoveredwithin 7 or 28 days, whereas for others the CMAPremained suppressed at the time endpoint. For thehighest exposure the CMAP at 7 and 28 days was notmeasurable in any of the animals, suggesting nervedegeneration.

Combining the results for both 7-day and 28-daygroups with the 0-, 2-, and 4-h results, the CMAPamplitude initially decreased up to 4 h or 7 days afterHIFU treatment (longer period of decline as expo-sure increased) and then recovered. For the threelower exposures the CMAP was essentially back tothe baseline amplitude level at 28 days after treat-ment. For the highest exposure, however, the CMAPremained absent even at 28 days after HIFU treatment.The statistical difference in the relative CMAP ampli-

FIGURE 6. Light micrographs of rat sciatic nerves; all sections stained with Masson’s trichrome. (A,B) Low-magnification cross-sectionalviews, (A) untreated nerve section; organized axon structure, (B) HIFU-treated nerve section; arrow indicates region of disorganized fiberstructure. (C,D) High-magnification cross-sectional views, (C) untreated nerve section; organized axon and myelin structure, (D)HIFU-treated nerve section; thin arrow indicates broken myelin accumulation in center of fibers and thick arrow indicates a vacant regionthat could be due to cell death. Structure of nerve sections treated with 5-s HIFU exposures of 390, 2,255, and 3,310 W/cm2 was notdistinguishable when compared to untreated nerve sections.


tude between groups treated with the same exposurebut measured at different time endpoints (4 h, 7 days,28 days) was calculated; differences were not signifi-cant, perhaps due to the small number of animals ineach of the 7-day and 28-day groups (n � 3).

Toe responses of those animals treated with thehighest exposure (7,890 W/cm2; 5 s) remainedclenched and arched even at day 28. All animalstreated with the lower three exposures had a normaltoe response by 28 days after HIFU treatment. Thecharacteristics of the foot after HIFU applicationincluded changes in the distance between the toes(toe spread) and changes in the arch of the toes (flatvs. curved), as compared to the toes of sham-treatedanimals. A quantitative measure of the distance be-tween the outer toes on the foot of the treated legwas recorded for several timepoints. The foot of anormal, untreated leg would have a toe-spread of �1cm between the outer two toes. Figure 8 presents theevolution of toe-spread over time for animals treatedwith the four different HIFU exposures. It appearedthat for animals treated with the lower three expo-sures the toe-spread was approximately normal by 28days after HIFU treatment. However, the animalstreated with the highest exposure (7,890 W/cm2;5 s) maintained a narrower toe-spread that couldindicate hindered function. These behavioral obser-

vations seemed to correspond well with CMAP re-sults, indicating a permanent effect of the highestHIFU exposure on nerve conduction and a minimalto temporary effect of the three lower exposures onconduction.

In animals treated with the highest exposure(7,890 W/cm2; 5 s) at both 7- and 28-day time end-points, all nerves appeared necrosed at the site of theHIFU treatment; the nerve had thickened (�2–3mm as compared to �1.5 mm of normal nerves) andadhesions had formed between the nerve and thesurrounding muscle tissue. Nerves treated with thelower exposures did not have these adhesions.

Light micrographs of nerve sections, both 7 and28 days after treatment with HIFU at the lowestexposure (390 W/cm2; 5 s), did not show any histo-logical variance from control untreated nerve sec-tions (Fig. 9A). Micrographs of nerves treated withthe second-lowest exposure (2,255 W/cm2; 5 s) 7days after HIFU showed minor disruption of someaxons (Fig. 9F); sections of nerves taken 28 days afterHIFU appeared to show a minor decrease in thenumber of axons as compared to untreated sections(Fig. 9I). For the sections of nerves treated with thetwo highest exposures (3,310 and 7,890 W/cm2; 5 s)and with endpoints of both 7 and 28 days, thereappear to be fewer axons and a less organized overallstructure; more Schwann cells could indicate theattempt at repair and regeneration of the nerve (Fig.9B,D,G,H). Hemorrhagic regions and areas of clot-ted blood cells were present as well (Fig. 9C,E).

DISCUSSION

These acute and long-term in vivo studies in ratsdemonstrate that increasing HIFU intensity has astronger and more prolonged effect on nerve func-tion. The results suggest that at the highest HIFUintensity, complete axon degeneration and nervedisruption is achieved (i.e., neurotmesis or axonot-

FIGURE 7. (A) Mean relative CMAP amplitude at 7-day timeendpoint corresponding to nerves treated with 5-s exposures atfour different HIFU intensities (n � 3 for each intensity). (B) Meanrelative CMAP amplitude at 28-day time endpoint correspondingto nerves treated with 5-s exposures at four different HIFU inten-sities (n � 3 for each intensity). Error bars equal SEM.

FIGURE 8. Distance between outer two toes measured at severaltimepoints after HIFU treatment of four different exposures; foreach exposure and time-point, n � 3 and error bars equal SEM.


mesis). At intermediate HIFU intensities a transientdisruption of nerve function is achieved, largely re-solving by 28 days, suggesting neurapraxia or partialaxonotmesis; recovery of CMAP amplitude and toeabduction in walking proceeds via restored axonconduction (e.g., by remyelination) or by motoraxon sprouting in partially denervated muscle. Thus,as a treatment, it may be possible to achieve a tem-porary nerve block lasting weeks using HIFU at in-termediate intensities or permanent nerve degener-ation by using HIFU at high intensity. An HIFU-induced nerve lesion could cause neuropathic orcentral pain, though perhaps reversible neural ef-

fects (i.e., conduction block) may cause such centralpain less commonly than irreversible nerve lesions(i.e., axon degeneration). This hypothesis could betested using the graded HIFU nerve lesions in ani-mals with longer survival time and with specific mea-sures for the level of pain.

The histological results from the short-term in-vestigation (Fig. 6) suggest that immediate damageto the nerve, particularly myelin disruption and axonswelling, is evident in nerves treated with the highestHIFU exposure (7,890 W/cm2; 5 s). When treatedwith low and intermediate exposures (390 to 3,310W/cm2; 5 s) damage was not evident in sections of

FIGURE 9. Light micrographs of rat sciatic nerves; all sections stained with Masson’s trichrome and scale bars equal 16 �m. (A) Sectionof untreated nerve, 7 days after HIFU treatment of opposite leg; many well-defined axons (arrow). (B) Treated section of nerve, 7 daysafter 5-s HIFU treatment of 7,890 W/cm2; presence of Schwann cells (arrow) yet fewer axons. (C) Treated section of nerve, 7 days after5-s HIFU treatment of 7,890 W/cm2; presence of areas of clotted blood cells (arrow). (D) Treated section of nerve, 7 days after 5-s HIFUtreatment of 3,310 W/cm2; presence of Schwann cells (arrow) yet fewer axons. (E) Treated section of nerve, 7 days after 5-s HIFUtreatment of 3,310 W/cm2; presence of hemorrhagic regions (arrow). (F) Treated section of nerve, 7 days after 5-s HIFU treatment of2,255 W/cm2; arrow denotes well-defined axon. (G) Treated section of nerve, 28 days after 5-s HIFU treatment of 7,890 W/cm2; presenceof Schwann cells (arrow) yet fewer axons. (H) Treated section of nerve, 28 days after 5-s HIFU treatment of 3,310 W/cm2; presence ofSchwann cells (arrow) yet fewer axons. (I) Treated section of nerve, 28 days after 5-s HIFU treatment of 2,255 W/cm2; arrow denoteswell-defined axon.


nerves harvested within 4 h after HIFU treatment.For the long-term investigation, qualitative light mi-croscopy of HIFU-treated nerves shows a graded ef-fect of HIFU intensity on nerve structure (Fig. 9).

Darkly staining oval profiles are much increasedin density at the two highest HIFU exposures, both at7 and 28 days after treatment. These dark oval pro-files may represent the nuclei of reactive Schwanncells or other glia or of invading inflammatory cells.Even at the two highest HIFU exposures (3,310 and7,890 W/cm2; 5 s), the gross structure of the nerve(e.g., epineurium, perineurium, endoneurium) wasnot disrupted (Fig. 9B,D,G,H). Thus, by the Seddonclassification of nerve injury this likely representsaxonotmesis, where axons degenerate but the sup-porting nerve structure is preserved. In contrast toneurotmesis, where the nerve is completely severed,in axonotmesis there is more potential for axonregeneration because of the continuity of supportingelements.13

In this study of exposure-dependent effects ofHIFU on nerve, we relied on visually guided, directnerve contact. This method contrasts with the pro-posed future clinical use of HIFU as a percutaneousnerve treatment using ultrasound image guidance totarget a deep nerve. The rationale for direct nervecontact in this study was to attempt to assure that anyvariability due to poor nerve targeting was mini-mized. In our prior work we showed that ultrasoundimage–guided percutaneous HIFU nerve lesions arefeasible.5,7 Future work is needed to demonstratethat image-guided percutaneous HIFU with varyingintensities can achieve temporary and permanentnerve lesions.

The use of HIFU exposure chosen for this studydoes not account for the effect of temperature onnerve conduction. Although intensity and durationof treatment are important parameters in HIFU ap-plications, the temperature change in the tissueplayed a role in previous nerve conduction investi-gations.10 Temperature of the nerve correspondingto the same HIFU exposure parameters (intensityand duration) will also vary between exposed nervesand nonexposed nerves surrounded by muscle; non-exposed nerves should reach higher temperaturesand thus the threshold HIFU parameters for biolog-ical effects (i.e., permanent nerve block) should belower than for exposed nerves. Therefore, futurework will need to include more thorough investiga-tion of the effects of temperature on nerve conduc-tion during noninvasive image-guided HIFU appli-cations.

Ultrasound of different parameters has beenshown to have variable effects on the electrophysio-

logical, functional, and microstructural characteris-tics of peripheral nerves.5,7,10,12,15,19 It is likely thatdifferent acoustic mechanisms, thermal and me-chanical, are predominant in these different scenar-ios of ultrasound application, thus leading to thevariable effects. Due to the coagulative necrosis thatresults from the thermal effect of HIFU,17 high tem-peratures (�70°C), associated with high intensitiesand durations of HIFU exposure, are expected tosuppress nerve function and potentially lead to per-manent or semipermanent degeneration of axons.In fact, work by Lele10 indicated that more pro-nounced biological effects such as nerve conductionblock were due to temperature elevation of thenerve. Mechanical effects of ultrasound, particularlycavitation, may have effects dependent on the extentof the cavitation activity—it is possible that minimalcavitation could produce effects such as decreasednerve signal intensity through demyelination, with-out complete degeneration of the nerve cells,whereas more marked cavitation could completelydestroy nerve fibers, leading to axon degeneration.Our preliminary unpublished investigations (n � 4nerves) have indicated that minimal heating andcavitation were present within the rat sciatic nerveduring treatment with the three lower HIFU expo-sures.6 However, inertial cavitation and heating weredetected during treatments with the highest HIFUexposure (7,890 W/cm2; 5 s). More extensive studieswill be completed to elucidate the predominantmechanisms with different HIFU exposure parame-ters. By determining the roles of these mechanismsin the effects of HIFU on nerve conduction, futureprotocols of treatment could be developed to selec-tively enhance or decrease the mechanisms and pro-duce a desired biological effect.

The studies presented here show the exposure-dependent effects of HIFU on nerve function andstrengthen the position that HIFU could be usedclinically to achieve partial and temporary nerveblocks or permanent and complete nerve degenera-tion. Such different effects on nerve function mayhave clinical applications in the treatment of certaintypes of spasticity and pain. For example, completeand permanent nerve degeneration could be used toirreversibly suppress the severe spasticity of patientswith a complete spinal cord injury and no preservedvoluntary function. Partial and temporary conduc-tion block could be used to suppress more minorspasticity or pain in patients while preserving somevoluntary function. This preliminary work has shownthat HIFU can produce graded effects on nerve con-duction and merits further investigation. Futurework to expand on this research will need to include


optimization of HIFU parameters and the elucida-tion of acoustic mechanisms responsible for differ-ent effects on neural structures. Extensive in vivostudies that determine the full range of biologicaleffects associated with various parameters of HIFUare necessary before safe and effective clinical treat-ments can be developed and accepted.

The authors thank the US Department of Defense, the SeattleFoundation, and the National Institutes of Health for funding thiswork.

REFERENCES

1. Burns AS, Meythaler JM. Intrathecal baclofen in tetraplegia ofspinal origin: efficacy for upper extremity hypertonia. SpinalCord 2001;39:413–419.

2. de Oliveira R, dos Reis MP, Prado WA. The effects of early orlate neurolytic sympathetic plexus block on the managementof abdominal or pelvic cancer pain. Pain 2004;110:400–408.

3. Dunne JW, Heye N, Dunne SL. Treatment of chronic limbspasticity with botulinum toxin A. J Neurol Neurosurg Psychi-atry 1995;58:232–235.

4. Fock J, Galea MP, Stillman BC, Rawicki B, Clark M. Functionaloutcome following Botulinum toxin A injection to reducespastic equinus in adults with traumatic brain injury. Brain Inj2004;18:57–63.

5. Foley JL, Little JW, Starr FL 3rd, Frantz C, Vaezy S. Image-guided HIFU neurolysis of peripheral nerves to treat spasticityand pain. Ultrasound Med Biol 2004;30:1199–1207.

6. Foley JL. High-intensity focused ultrasound as a novel methodof nerve conduction block: dose-dependent effects rangefrom partial to complete block. PhD Dissertation. Seattle:University of Washington; 2006.

7. Foley JL, Little JW, Vaezy S. Image-guided high-intensity fo-cused ultrasound for conduction block of peripheral nerves.Ann Biomed Eng 2006;35:109–119.

8. Gunduz S, Kalyon TA, Dursun H, Mohur H, Bilgic F. Periph-eral nerve block with phenol to treat spasticity in spinal cordinjured patients. Paraplegia 1992;30:808–811.

9. Jang SH, Ahn SH, Park SM, Kim SH, Lee KH, Lee ZI. Alcoholneurolysis of tibial nerve motor branches to the gastrocne-mius muscle to treat ankle spasticity in patients with hemiple-gic stroke. Arch Phys Med Rehabil 2004;85:506–508.

10. Lele PP. Effects of focused ultrasonic radiation on peripheralnerve, with observations on local heating. Exp Neurol 1963;8:47–83.

11. Martin RW, Vaezy S, Proctor A, Myntti T, Lee JB, Crum LA.Water-cooled, high-intensity ultrasound surgical applicatorswith frequency tracking. IEEE Trans Ultrason FerroelectrFreq Control 2003;50:1305–1317.

12. Mihran RT, Barnes FS, Wachtel H. Temporally-specific mod-ification of myelinated axon excitability in vitro following asingle ultrasound pulse. Ultrasound Med Biol 1990;16:297–309.

13. Seddon HJ. Three types of nerve injury. Brain 1943;66:237–288.

14. Smyth MD, Peacock WJ. The surgical treatment of spasticity.Muscle Nerve 2000;23:153–163.

15. Takagi SF, Higashino S, Shibuya T, Osawa N. The actions ofultrasound on the myelinated nerve, the spinal cord and thebrain. Jpn J Physiol 1960;10:183–193.

16. ter Haar G. Ultrasound focal beam surgery. Ultrasound MedBiol 1995;21:1089–1100.

17. Vaezy S, Martin R, Schmiedl U, Caps M, Taylor S, Beach K, etal. Liver hemostasis using high-intensity focused ultrasound.Ultrasound Med Biol 1997;23:1413–1420.

18. Vaezy S, Shi X, Martin RW, Chi E, Nelson PI, Bailey MR, et al.Real-time visualization of high-intensity focused ultrasoundtreatment using ultrasound imaging. Ultrasound Med Biol2001;27:33–42.

19. Young RR, Henneman E. Reversible block of nerve conduc-tion by ultrasound. Arch Neurol 1961;4:83–89.


Effects of high-intensity focused ultrasound on the intervertebral disc: A potential therapy for...

Documents

Transcript of Effects of high-intensity focused ultrasound on the intervertebral disc: A potential therapy for...