radiofrequency volumetric reduction of the tongue.pdf

laboratory and animal investigationsRadiofrequency Volumetric Reductionof the Tongue*A Porcine Pilot Study for the Treatment ofObstructive Sleep Apnea SyndromeNelson B. Powell, MD; Robert W. Riley, MD; Robert J. Troell, MD;Marc B. Blumen, MD; and Christian Guilleminault MD

Study objective: To investigate, in an animal model, the feasibility of radiofrequency (RF)volumetric tongue reduction for the future purpose of determining its clinical applications inobstructive sleep apnea syndrome (OSAS).Design: The study was performed in three stages, one in vitro bovine stage and two in vivoporcine stages. The last stage was a prospective investigation with histologic and volumetricanalyses to establish outcomes.Setting: Laboratory and operating room of veterinary research center.Participants: A homogeneous population of porcine animal models, including seven in stage 2 and12 in stage 3.Intervention: RF energy was delivered by a custom-fabricated needle electrode and RF generatorto the tongue tissue of both the in vitro and in vivo models.Measurements and results: Microultrasonic crystals were used to measure three-dimensionalchanges (volumetric reduction). Lesion size correlated well with increasing RF energy delivery(Sperman correlation coefficient of 0.986; p=0.0003). Histologic assessments done serially overtime (1 h through 3 weeks) showed a well-circumscribed lesion with a normal healing progressionand no peripheral damage to nerves. Volumetric analysis documented a very mild initialedematous response that promptly tapered at 24 h. At 10 days after RF, a 26.3% volumereduction was documented at the treatment site (circumscribed by the microultrasonic crystals).Conclusion: RF, in a porcine animal model, can safely reduce tongue volume in a precise andcontrolled manner. Further studies will validate the use of RF in the treatment of OSAS.

(CHEST 1997; 111:1348-55)

Key words: inreduction

vitro; in vivo; kilojoules; microultrasonic crystals; porcine animal model; radiofrequency; volume

Abbreviations: H&E=hematoxylin-eosin; in OD= inches outer diameter; kj=kilojoules; OSAS= obstructive sleepapnea syndrome; PAM=porcine animal model; RF= radiofrequency; RFe= radiofrequency energy; 2D= two dimensional; 3D=three dimensional

"O adiofrequency (RF) applications have been in-*-" vestigated previously for both medical and surgical use. The biophysics of radiofrequency energy*From the Stanford University Sleep Disorders and ResearchCenter, Stanford, Calif.Supported in part by Somnus Medical Technologies, Inc.,Sunnyvale, Calif.Manuscript received September 16, 1996; revision acceptedDecember 4.Reprint requests: Dr. Nelson Powell, Head and Neck Surgery, 750Welch Road, Suite 317, Palo Alto, CA 94304

(RFe) are unique. RFe is delivered in a unipolar orbipolar fashion from an electrode (needle or antenna). An RF generator provides alternating current tothis electrode to generate low heat energy (40 to90C) sufficient to denature tissue protein. An RFneedle may be placed percutaneously and only thetissue adjacent to the unprotected portion of theneedle will undergo ablation. This fact eliminatessurface destruction. A precise targeting of the tissuesrequiring treatment is also possible with RF. The

1348 Laboratory and Animal Investigations

Downloaded From: http://journal.publications.chestnet.org/ on 03/04/2015

physical properties of RF concerning heat disbursement are favorable for controlling lesion size. RFenergy disbursement is proportional to 1/radius4.This rapid drop of energy from the source is useful intreatment of tissues which contain, or are near, vitalstructures. Lesion size can be further controlled bythe diameter and length of the needle electrode, aswell as treatment duration and total energy7 in joules(wattsXseconds). This study has investigated thefeasibility and application of RF for the treatment ofobstructive sleep apnea syndrome (OSAS). This syndrome affects at least 2% ofwomen and 4% of men.1The disproportionate soft-tissue anatomy of the upper airway associated with OSAS will be the focus ofthis RF application.

Materials and Methods

This study describes the effects of treating in vitro bovine andin vivo porcine tongues with RF. The porcine animal model(PAM) was selected because the tongue size is similar to thatfound in humans. This study was in compliance with the guide forthe care and use of laboratory animals (National Institutes ofHealth publication 80-23). Three staged investigations wereundertaken whose overall purpose was to critically evaluate theinterventional use of RF as an ablative energy for the reductionof tongue mass. The last stage of this investigation was doneprospectively with histologic and volumetric analysis to establishoutcomes.

Medical RF Device

An RF generator with custom-fabricated needle electrodes andenergy algorithm specifically designed for tongue tissue was used(Somnus Medical Technologies, Inc., Sunnyvale, Calif). The RFewas delivered at 465 kHz. The energy algorithm allowed simul

taneous documentation of total energy and test duration, alongwith temperature measurements at separate microthermocoupleson the electrode.

Ultrasonic Volumetric Measurements

Volumetric assessments of the tissues undergoing treatmentwere done using omnidirectional microultrasonic (1.5 mm) piezoelectric crystals with linear accuracy to 0.1 mm (Sonometrics;London, Ontario, Canada). These crystals were positioned tosurround the electrode and were used to quantitate volumetricchanges before and after delivery of RF energy. All volumetricchanges reported in either percent or grams of tissue were madein reference to the region circumscribed by these crystals, andnot to the entire tongue mass. The base line (control), immediateposttreatment, and follow-up volumetric data were measuredapplying multidimensional digital sonomicrometry using customhardware and software for two-dimensional (2D) area and three-dimensional (3D) volume measurements (Sonometrics version4.0; London, Ontario, Canada). Commercial software (Sigma Statand Sigma Plot for Windows; Jandel Scientific; San Rafael, Calif),was used for 2D area calculations along with data presentationand statistical calculations. Volume was determined using asoftware program (Mathematicia version 2.2; Wolfram ResearchInc; Champaign, 111) developed to calculate the volume of anyarbitrary-shaped objects.

Animal Models

Stage 1: Feasibility Study Using the Bovine Model (In Vitro):Two bovine tongues were used for the in vitro evaluations. Thefirst specimen was implanted with four ultrasonic crystals forming a rectangle for a 2D analysis. The second received eightcrystals forming the limits of a cube for a 3D study. In both, asingle 0.050 (inches outer diameter [in OD]) needle electrodedelivered 30 kilojoules (kj) of energy over 20 min. Two separatesites were done on each. An excessive level of power output andtime was selected in order to see an immediate discerniblevolume reduction (shrinkage).

Stage 2: Volume and Energy Calculations Using the Porcine

Table 1.Data From Porcine In Vivo Study Stage 2*

Study Animal (n=7)

Control Porcine Animal Model (PAM) Test

2X

Weight, kgSex, M/FAge, moElectrode (in OD)Electrode length in

tissue, mmEnergy treatment

duration, minTotal energy, kjVolumetric reduction,lh4h

30.0F30.018

21

0

0

-4NA

30.2M

30.018

15

40.0

NA

31.8M

40.01815

27

16.2

14NA

29.5F

30.01815

25

24.0

14NA

28.1F

30.018

21

19

35.0

2631

30.4F

30.035

30

15

7.3

10NA

30.8F

30.035

18

10.5

6.8

NANA

*No volumetric analysis done. NA=not applicable.Control .4% (actual 4% increase in volume).

*Tongue surface cooled to 10C. Studied at 1 and 4 h after energy delivery."Multiple energy7 deliveries (six) with varied low power.

CHEST/111 /5/MAY, 1997 1349


Model (In Vivo): Seven PAMs (Table 1) were used to establishthe effects of varying levels of RFe reported in kilojoules ontongue tissue. Energy was increased from 6.8 through 40 kj. Fivefemale and two male animals had a mean age of 3.1 months(range, 3 to 4; SD, 0.40) and a mean weight of 30.1 kg (range,28.1 to 31.8; SD, 1.10). Volumetric measurements were determined for six of the seven PAMs and were calculated prior to RFdelivery at baseline, and then immediately after energy delivery.PAM 7 was given multiple burns (six) of low energy with variableintensity to produce a small lesion that could be examined on asingle microscope slide. Once the animals were killed, theindividual tongues were immediately sectioned for gross evaluation and measurements.

Stage 3: Volume and Histology Study Using the Porcine Model(In Vivo): The final portion of this pilot study was done prospectively and examined volumetric changes over time combined withthe acute and chronic histologic changes seen with a constantRFe at 2.4 kj. Twelve PAMs (eight male) with a mean age of 3.4months (range, 2 to 6; SD, 1.31) were investigated. PAMs 1through 9 (mean weight, 62.5 kg; SD, 24.4) were implanted withcrystals and treated with RF. Three control animals, A through C(mean weight, 74.4 kg; SD, 23.4) underwent placement ofelectrodes and crystals as in the test models but without RFedelivery. The electrodes were standardized at 0.035 in OD, witha total length of 45 mm. The length of the electrode placed in thetissue was 16 mm (6 mm of protective collar and 10 mm ofexposed tip). Temperatures of each active needle electrode weremaintained at below 100C. All but PAM 1 underwent volumetricanalysis, and this model was excluded since it was killed 1 h aftertreatment. Gross and histologic examinations were done post-treatment on all specimens. The animals were prospectivelydivided into an acute phase (1, 24, 48, and 72 h) and a chronicphase (1, 2, 3, 4, and 5 weeks). The animals were killed accordingto this schedule.

Surgical TechniqueThe RF procedure was performed using general anesthesia.

No antibiotics or corticosteroids were given prior to or after anyin vivo investigation. The tongue was pulled forward to exposethe circumvallate papilla at the junction between the posteriorbase and the anterior portion. Eight ultrasonic crystals wereimbedded for 3D volume analysis. At the end of the procedure,each crystal was superficially buried and sutured into place to be

30

25

20

Immediatevolume 15reduction

(%)10

5

0

-5

20 25 30 35 40

Energy (K joules)

Figure 1. Total energy and immediate volume reduction. Thegraph compares PAM 1 (control) at zero energy with PAMs 3through 6. This demonstrates that as energy increases, an overallincrease in volumetric reduction is seen. PAM 2 was excludedsince surface cooling canceled energy delivery by the process ofconvection. The negative value in the control animal represents atemporary increase in volume due to edema. Needle electrodetrauma is thought to be negligible in all studies; however, themanipulation in crystal placement was sufficient to account formost of the immediate edema.

used again when the predetermined follow-up time was reached,and the animal was killed. An RF needle electrode was placedcaudally at the circumvallate papilla apex in the midline of thetongue. Baseline volume measurements were determined. Thisprocedure was followed rapidly by energy delivery that occurredover a mean time span of 6.06 min (range, 4 to 8.5; SD, 1.29)(Table 2). An immediate postenergy delivery measurement wasthen taken. Follow-up treatment care was done by an attendingveterinarian. The final data collection was also done with theanimal under general anesthesia in the exact manner as the initialtrial.

Table 2.Data From Porcine In Vivo Stage 3*PAM(n=12)

Test model123456789

Control modelABC

TestTime

lh24 h48 h72 h10 d2 wk3 wk4 wk5 wk

72 h2 wk3 wk

Weight,kg

Age>mo

3118511006059805373

1005470

Energy TreatmentDuration, min

4.07.08.55.55.56.05.57.05.5

000

Macro and Histologic Protocol

Hematoxylin-eosin (H&E) and trichrome stains were done toevaluate the tissue response to RF in the acute and chronic

Table 3.Energy and Lesion Size*PAM 7Test

123456

MeanSD

Energy, Lesion Size,cmXcm

Lesion Area,cm2

0.50.60.80.81.72.41.10.75

1.4X0.51.0X1.01.3X1.01.2X1.21.7X1.01.5X1.7

0.71.01.31.41.72.51.40.63

*Test time relates to the time of sacrificing as prospectively planned.

*The Spearman correlation coefficient shows a positive correlation of0.986 between energy levels within the study range (p=0.0003).This would indicate that as energy increases so does lesion size.



Table 4.Volume Reduction Data Over Time*

Test Model

PAM (n=9) 1 2 9

Killing and final test timeBase crystal volume, cm3Immediate crystal volume, cm3Postcrystal volume, cm3Base minus immediate, cm3Base minus post, cm3Immediate shrinkage, %Postshrinkage, %

lhNANANANANANANA

24 h1.311.221.300.090.016.90.80

48 h1.551.451.360.100.196.5

12.3

72 h1.971.821.560.150.407.6

20.8

10 d3.042.832.240.210.807.0

26.3

2wk3.262.96NA0.30NA6.3NA

3 wk1.311.22NA0.09NA8.7NA

4 wk2.071.97NA0.10NA4.8NA

5 wk2.962.71NA0.25NA8.4NA

Control Model

PAM (n=3) B

Killing and final test timeBase crystal volume, cm3Immediate crystal volume, cm3Postcrystal volume, cm3Base minus immediate, cm3Base minus post, cm3Immediate shrinkage, %fPostshrinkage, %

72 h1.151.211.14

-0.060.01

-5.20.9

2 wk4.284.52NA-0.24NA-5.6NA

3 wkNANANANANANANA

*Crystal volume measurements were calculated by using the x, y, z, crystal coordinates. (Mathematica). Base crystal volume measurements weretaken prior to treatment and represent baseline volumes that are used as control values for each animal; immediate crystalvolume=measurements taken within 5 min of RFe delivery; base minus immediate=the volume change immediately after treatment; base minuspost=the volume change from baseline (control) to the time of killing. Calculated data from this table indicate a mean of 7.02% for the immediatevolume reduction (range, 4.8 to 8.7%; SD, 1.24), while the corresponding values in the two control models are .5.54% (SD 0.28). The meansdiffer by t test with a p value of 0.0001. The negative shrinkage (edema) in the control animals is secondary to electrode and crystal placements.Immediate shrinkage was seen in PAM 2 through 5 despite similar edema in this homogeneous sample. The late or postshrinkage group showsfurther volume reduction with time.The negative values listed under immediate shrinkage represent an actual increase in tissue volume. Edema seen in control porcine A is totallyresolved at 72 h. A +0.9% shrinkage is seen and is consistent with a minor loss of volume secondary7 to invasive manipulation. NA=not applicable.

phases. Macroscopic examinations of each treated tongue weredone and the lesion was measured to document the extent andsize (cm2) of each RF lesion. A similar protocol was used for thethree control animals. A systematic evaluation of the histologicfeatures documented the effects of RF over time to the muscle,interstitial space, vessels, and nerves to include the inflammatoryresponse, fibrosis, and muscle repair.

Results

Stage 1: Bovine Model Feasibility (In Vitro)An immediate volume reduction of 12.8% and

13.2% (2D) and 26.7% and 25.2% (3D) were recorded for each tongue. The crystals were left inplace for an additional 4 h on both tongues and thevolume was restudied. A loss of an additional 4% ofvolume was noted in each case for a mean totalvolume reduction of 17% (2D) and 29% (3D),respectively. This additional reduction is consistentwith further protein damage due to undissipatedheat from the central core of the lesion. The difference in volume reduction observed when 2D and 3Dresults are compared is related to the more accuratevolume determination with the usage of 3D measurements.

Volumereduction

(%)

30

25

20

15

10

5

0

-5 H

Stage Three

9 10 11

Time (days)

Figure 2. Volumetric change over time at a constant energy of2.4 kj in stage 3. The graph indicates the volume changes overtime. The mean immediate shrinkage in PAM 2 through 8 was7.02%. This value was used as the origin and graphed at timezero. By 24 h, edema has caused a temporary loss of all but 0.08%(asterisk). At 48 and 72 h continuing to 10 days, further volumereduction is seen.

CHEST/111 /5/MAY, 1997 1351


Table 5.Size Over Time at 2.4 KJ

PAM No.Time UntilSacrifice

Lesion Size,cmXcm

Lesion Area,.,2

lh24 h48 h72 h10 d2wk3 wk4wk5 wk

1.5X1.73.0X1.73.5X2.04.0X1.72.0X1.41.3X1.21.2X0.7NANA

2.555.107.006.882.801.560.84NANA

*Macroscopic investigation using serial sectioning of the treatedtongues immediately after sacrifice demonstrates initial edemafollowed by a reduction in size through 21 days. The 4- and 5-weeklesions were difficult to quantify accurately by macroscopic examination owing to the loss of distinct borders secondary to the healingprocess. NA^not able to accurately determine borders at 4 and 5weeks.

Stage 2: Porcine Model.Volume and EnergyCalculations (In Vivo)Volumetric 3D Analysis: Six of the seven PAMs

(1 through 6) were included in this analysis (Table1). PAM 1 served as a control with placement ofelectrodes and crystals without energy delivery.The results show that crystal and electrode placements in PAM 1 mildly increased total volume by

4%, which was related to edema. PAM 2 wassubmitted to surface cooling (10C) with crystalsand electrodes in place. Despite energy delivery,there was no discernible volume reduction on 3Dreconstruction: this suggested that surface coolingmay help limit heat damage to the immediateunderlying structures. PAM 5 additionally hadmeasurements at 4 h. The results obtained onPAMs 3 through 6 show that volume reduction wasrelated to the energy delivered (Fig 1).

Correlation of Volume, Energy, and Lesion Size:As shown in Table 1, electrodes with variable sizeand length were tried in this stage of the study. PAM7 received six separate lesions of low-variable energyand hence did not undergo volumetric analysis. Theresults of energy and lesion size obtained are shownin Table 3.

Stage 3: Porcine Model.Volume (In Vivo)The volume reduction in PAMs 2 through 9 was

calculated immediately after energy delivery and wasrepeated just before sacrificing these animals (Table4). Minimal edema of 4 to 6% is seen initially aftertreatment, which tapers promptly by 24 h. A progressive volume reduction follows for up to 10 daysof 26.3% (Fig 2). Some of the PAMs had lost oreaten the crystals and a few of those crystals thatremained embedded for close to 2 weeks showed

Stage Three

Time vs. Grams of Tissue

Time vs. Lesion Size

1.0

- 0.9

0.8

- 0.7

- 0.6

- 0.5

- 0.4

0.3

0.2

r- 0.1

0.0

-0.1

Gramstissue(cm3)

n.i.i.i.i.i.i.i.i.i.i.i.i.i.i.i.i.i.i.i.r3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Time (days)

Figure 3. Lesion size and grams of tissue over time in stage 3. The connecting triangles describe thecrystal data shown in Table 4. The difference between the "baseline crystal volume" minus the"posttreatment volume" for each animal was calculated in cubic centimeters. This figure was thenmultiplied by the specific gravity of muscle tissue at 1.05 g/mL (cm3) to yield the grams of tongue tissuereduced over the individual time periods. The data collection allowed only an accurate calculation upto 10 days, and was equivalent to 0.84 g, which was also 26.3% of the sample size for that individualtrial. The solid dots represent the lesion size in square centimeters as measured on gross evaluation atthe time of sacrificing. Lesion size increases from baseline up to 48 h, and then a decline is seen at 72 h,which then steadily drops off as healing and retraction progress.



early evidence of infection. Our concern for possibleinfection at the crystal sites caused us to abandon invivo volume measurements after 2 weeks.

Stage 3: Porcine Model.Histology (In Vivo)Macroscopic Tissue Examination: At 1 h, the

lesion site is clearly distinguished by a targetedcentral area of brown surrounded by a area of whitetissue. At 24 h, the lesion is well defined by a rim ofhyperemic tissue surrounding the white damagedtissue. The 72-h tissue is similar except the centraltissue is now gray-brown. At 1 week, there is whitetissue surrounding the lesion that corresponds histologically to a prominent fibroblastic response, withcollagen deposition replacing damaged muscle tissue. The 2-, 3-, and 4-week lesions show a maturingof the healing process with contraction of the site ofinjury and a stellate white glossy character to thelesion. Measurements of lesion size over time allowed further correlation of the character of woundhealing after RF treatment (Table 5 and Fig 3).

Microscopic Tissue Examination: The microscopictissue examination is outlined in Figures 4 and 5. Thethree control animals were evaluated grossly andmicroscopically with no lesions found at the electrode site. There was evidence of infection at thecrystal site in the 3-week control animal, although it

was not communicating with the electrode site.Subsequently, this control was not used for histologiccomparison. The vascular changes associated withRF injury and scar formation include destruction ofthe vessel walls and thrombosis within the lesion aswell as focally at the periphery. In the early acutespherical lesion, there is interstitial edema and hemorrhage, indicating injury to the microvasculature. Asthe lesion resolves, there is neovascularization of theforming scar and recanalization of thrombosed vessels. The vessels surrounding the lesion remain intactand viable. There is no evidence of extra-lesionalhemorrhage or hematoma formation. The nerves,similar to the vascular structures, remained viable inthe tissues surrounding the lesion. In the earlystages, there is mild edema of small nerves injuxtaposition to the injured tissue. However, as thelesion resolves, there is no evidence of extra-lesionalnerve destruction. Viable nerve structures are identified in the newly formed fibrous tissue at thejunction of the scar and surrounding viable muscle.

Discussion

The unique properties of RF led to their usage inthe treatments of trigeminal neuralgia by Sweet andWepsic2 as early as 1974. RF treatments for cancer

Figure 4. Histopathologic features of the lesion at 24 h (H&E, original magnification X30). At 24 hafter RFe deliver}', this low-power field shows a discrete tract at N, which is the electrode-tissueinterface. There is a rim of hyperemia (H) that defines the peripheral lesion margins. Viable accessorytissues such as salivary7 glands (S), lymphoid tissues (L), and epithelial surface (E) are seen. There ismild interstitial edema and focal hemorrhage with loss of myocyte z-band periodicity and minimalnuclear dropout.

CHEST / 111 / 5 / MAY, 1997 1353


Figure 5. Three weeks after RFe delivery7 (H&E, original magnification X30). Fibrosis and chronicinflammation (FC) have replaced the necrotic myocytes. Viable muscle (M) is seen at the periphery.This 3-week finding is microscopically characterized by a well-formed young scar in the center of thelesion with extensions of fibrotic tissue into surrounding viable muscle. Scattered areas of chronicinflammation are present in the scar tissue.

in animals and patients have also been reported.34 Inthe past decade, cardiologists have used RF catheterablation to treat Wolf-Parkinson-White syndrome.5In 1996, Issa and Oesterling,6 along with surgeons atother centers, reported on RF transurethral needleablation of benign prostate hyperplasia.7 All of theseprevious investigations have demonstrated acceptable feasibility, efficacy, safety and reproducibility oftreatment outcomes for their individual applications.This emerging technology associated with RF tissueablation has provided a vehicle and pathway thatcould assist in shrinkage of the excessive soft tissuefound in the airways of those patients with OSAS.Energy delivered by RF can be precisely placedbelow the surface tissues sparing the entry site, andthe treatment temperature is kept low (40 to 90C);hence its application is minimally invasive. This is indirect contrast to electrocautery or laser treatment inwhich the thermal damage is more extensive sincethe temperatures are two to three times greater.Furthermore, the delivery of energy and subsequentablation is uniquely different than what is found withelectrocautery or laser. RF generates ionic agitationat the cellular level as the ions tend to follow achange in direction with the alternating currentgenerated. This creates frictional heating of thetissues around the electrode and hence the heatemanates from the tissue and not the electrode. A

safety feature with RFe delivery is its biologicalability to self-limit tissue destruction when energydelivery is in excess. As the tissues coagulate at theelectrode tissue interface, water loss occurs (desiccation), and at 90 to 100C, a char forms on theneedle (electrode), which in turn creates sufficientelevation of impedance so that the current flow fallsto zero. This self-limiting biological switch is valuablein limiting excessive tissue damage.

Conclusion

Most specialists have recognized that the future inthe appropriate management of OSAS lies in thedevelopment of new methodologies and technology.This concept is essential to meet the needs of themany individuals who require treatment for sleep-disordered breathing. This means that treatmentapproaches will need to be minimally invasive, easyto perform, short in treatment time with minimalpatient discomfort, and cost-effective with acceptable clinical outcomes. Our investigation of RFtissue reduction could fulfill these parameters. Priorto our investigation, RF had not been considered forthe treatment of OSAS, and hence it was necessaryto establish its feasibility for use in the managementof excess soft tissue in the upper airway. This study



reinforces the previously reported principles of RFablation, and establishes the biological effect of RFwhen specifically applied to the tongue. At 10 daysafter treatment, a volumetric reduction of 0.84 g oftongue tissue at the specific ablation site (26.3% byvolume) was seen. In comparison, tonsil weightgenerally averages 2 to 4 g. It is likely, by extrapolation of the volumetric and histologic data, thatfurther shrinkage would occur with final healing.The gross and histologic investigation of the tonguedemonstrates that RF lesions follow the usual courseof wound healing and are not dissimilar to thepathogenesis of wound injury and healing found, forexample, in an ischemic myocardial infarction. Furthermore, no evidence of neural damage was seenoutside the core lesion in any animal model, and itwas possible to percutaneously target the anatomicsite and limit injury as well as edema. It is probablethat RF can be administered under local anesthetic,and as edema is minimal, an outpatient procedure isa reasonable possibility. We envision that RF couldbe applied conservatively by creating small lesions atsites of excess tissue in the upper airway. Selectedretreatments could be done until the desired resultsare attained. In the future, RF may be used as anadjunctive procedure or eliminate the need for all orpart of the surgeries now used for the treatment ofhypopharyngeal obstruction during sleep and prob

ably some nasal positive pressure treatments. Further investigations will help to support these possibilities. The next step in our continued investigationswill be the evaluation of this technique in otheranimal models and in palatal occlusion associatedwith snoring and/or OSAS.

References1 Young T, Palta M, Dempsey J, et al. The occurrence of

sleep-disordered breathing among middle-aged adults.N Engl J Med 1993; 328:1230-35

2 Sweet W, Wepsic J. Controlled thermocoagulation of trigem-inal ganglion and rootlets for differential destruction of painfibers: I. Trigeminal neuralgia. J Neurosurg 1974; 3:143-56

3 LeVeen H, Wapnick S, Piccone V, et al. Tumor eradication byradiofrequency therapy: response in 21 patients. JAMA 1976;253:2198-2200

4 McGahan J, Schneider P, Brock J, et al. Treatment of livertumors by percutaneous radiofrequency electrocautery. Semin Intervent Radiol 1993; 10:143-49

5 Jackman WM, Wang XZ, Friday KJ, et al. Catheter ablation ofaccessory atrioventricular pathways (Wolff-Parkinson-Whitesyndrome) by radiofrequency current. N Engl J Med 1991;324:1605-11

6 Issa M, Oesterling J. Transurethral needle ablation (TUNA.): an overview of radiofrequency thermal therapy forthe treatment ofbenign prostatic hyperplasia. Curr Opin Urol1996; 6:20-7

7 Schulmun CC, Zlotta AR. Transurethral needle ablation ofthe prostate for treatment of benign prostate hyperplasia:early clinical experience. Urology 1995; 45:28-33

CHEST/111/5/MAY, 1997 1355


radiofrequency volumetric reduction of the tongue.pdf

Documents