Acute Differential Modulation of Synaptic Transmission and Cell

Survival During Exposure to Pulsed and Continuous

Radiofrequency Energy
Alex Cahana,* Laszlo Vutskits,* and Dominique Muller†

Abstract: Pulsed radiofrequency, in which short bursts of radiofrequency energy are applied to
nervous tissue, has been recently described as an alternative technique devoid of nerve injury, a
subsequent side effect of thermal lesions created by continuous radiofrequency lesioning. Yet the
mechanism of this effect remains unclear. In this study we compared the acute effects of pulsed
versus continuous radiofrequency energy on impulse propagation and synaptic transmission in
hippocampal slice cultures and on cell survival in cortical cultures. A differential effect was observed
on both systems, with pulsed radiofrequency producing a transient and continuous radiofrequency a
lasting inhibition of evoked synaptic activity. In addition, although both continuous radiofrequency
and pulsed radiofrequency treatments induced a distance-dependent tissue destruction under the
stimulating needle, the effect was more pronounced in the continuous radiofrequency group. These
findings suggest that the acute effects of pulsed radiofrequency are more reversible and less
destructive in nature than the classic continuous radiofrequency mode, even in normothermal con-
ditions. This model might help elucidate the importance of various parameters for the clinical
application of radiofrequency lesioning and might open new horizons for the role of pulsed radio-
frequency lesioning in cases of neuropathic pain.
© 2003 by the American Pain Society
Key words: Radiofrequency, neuropathic pain, organotypic cultures, cortex, hippocampus.

R
adiofrequency treatments are offered for a variety pulsed radiofrequency energy on cell cultures induces
of pain syndromes: cervicogenic headaches,27 oc- immediate early gene expression1 not mediated by tissue
cipital neuralgia,5 whiplash injury,11 intercostal heating.9
neuralgia,25 mechanical low back pain,26 discogenic Although preliminary clinical investigations have
pain,10 and pain associated with the sacroiliac joint.6 shown that pulsed radiofrequency can be used safely as
The rationale for the application of radiofrequency de- an alternative interventional treatment in patients with
nervation is the assumption that selectively heating ner- refractory pain syndromes or with intolerable side ef-
vous structures can impede nociceptive input. Practically fects resulting from their conservative treatment,19 it is
this is achieved by percutaneous application of small size still not clear what are the differences or advantages
electrodes at target neural tissues, resulting in size-con- between pulsed radiofrequency and continuous radio-
trolled lesions at different anatomic positions, which de- frequency, in terms of both clinical outcome and biologic
pend on patients’ symptoms. mechanisms involved.
However, other results8,18 have questioned the utility The objective of our study was to investigate whether
the acute application of radiofrequency energy, either
of thermal lesioning, which is essentially neuroablative,
on the pulsed or continuous mode, resulted in similar or
in the presence of neuropathic pain. Accordingly, pulsed
different effects on functional parameters such as im-
radiofrequency has been recently described as an alter-
pulse propagation, synaptic transmission, and cell sur-
native technique to apply a relatively high voltage near a
vival. To this end, we used in vitro preparations (hip-
nerve without nerve injury, a subsequent side effect of
pocampal organotypic slice cultures and dissociated
thermal lesions.3,13,21 It is known that the application of
cortical neurons) that are ideally suited for such analyses.

Received September 30, 2002; Revised December 17, 2002; Accepted

December 17, 2002.
Materials and Methods
From the *Interventional Pain Program, Department of Anesthesiology, All animal procedures were performed after obtaining
Geneva University Hospital, and †Division of Neuropharmacology, Uni- approval from the Animal Care Committee of the Uni-
versity Medical Center, Geneva, Switzerland.
Address reprint requests to Alex Cahana, MD, Interventional Pain Pro-
versity Medical Center in accordance with the Swiss Fed-
gram, Department of Anesthesiology, Geneva University Hospital, Rue eral Animal Welfare Act.
Micheli-du-Crest 24, 1211 Geneva 14, Switzerland. E-mail: alex.cahana@
hcuge.ch
© 2003 by the American Pain Society
Electrophysiologic Studies
1526-5900/2003 $30.00 ⫹ 0 Hippocampal organotypic slice cultures (n ⫽ 6 to 7)
doi:10.1016/S1526-5900(03)00554-6 were prepared as previously described24 from 7- to 14-

The Journal of Pain, Vol 4, No 4 (May), 2003: pp 197-202 197

198 Modulatory Effect of Pulsed Radiofrequency
day-old Sprague-Dawley rats. Briefly, hippocampi were The effect of continuous and pulsed radiofrequency
dissected in an ice-cold dissection medium consisting of energy was tested on cultures maintained 2 days in a
50% minimal essential medium supplemented with 25 serum-free medium. A 22-gauge SMK 54-mm, 5-mm ac-
mmol/L N-2-hydroxyethylpiperazine-N-2-ethanesulfonic tive tip needle was placed perpendicularly to the mono-
acid; 300- to 400-␮mol/L thick slices were cut on a McIll- layer in contact with the serum-free culture medium to
wain tissue chopper (Mickle Laboratory Engineering Co., allow energy exposure. The distance between the mono-
Ltd., Gomshall, England) and then placed at the interface layer and the needle tip was set to 2000, 1000, 500, and
on a porous, translucid membrane (Millicell-CM; Mili- 250 ␮m, respectively, by changing the amount of me-
pore, Bedford, Mass) in a CO2 incubator. The culture me- dium in 3 treatment groups. Group I was exposed to
dium contained 25% horse serum, 25% Hank’s solution pulsed radiofrequency energy at 500 kHz at a pulse rate
buffered by 5 mmol/L Tris and 4 mmol/L NaHCO3 at a pH of 2 Hz for a period of 120 seconds, with an adjusted
of 7.2, with penicillin and streptomycin. Cultures were voltage of 35 to 45 V, to permit maximal temperature of
allowed to recover for 4 days at 36°C and then trans- 38°C. Group II was exposed to pulsed radiofrequency at
ferred to a 33°C incubator. an adjusted temperature of 42°C. Group III was exposed
For electrophysiologic measurements, 1- to 2-week-old to a 42°C continuous radiofrequency energy for 120 sec-
cultures were placed in an interface-type recording onds. Electronic impedance, voltage, current output, and
chamber continuously perfused with 126 mmol/L NaCl, 3 the tip temperature were recorded every 30 seconds.
mmol/L KCl, 24 mmol/L NaHCO3, 1.25 mmol/L NaH2PO4, After stimulation, cultures were washed and cells were
1.5 mmol/L MgCl2, and 2 mmol/L CaCl2 at a pH of 7.4. kept in culture for an additional 24 hours in serum-free
Excitatory postsynaptic potentials were elicited with a medium before fixation.
stimulation electrode placed on CA3 neurons to activate Cultures were fixed in cold (4°C) 4% paraformaldehyde
the Schaffer collateral pathway and were recorded in the in 0.1 mol/L phosphate buffer for 60 minutes and then
stratum pyramidale of area CA1. Stimulation intensities washed several times in a phosphate-buffered saline
were adjusted so as to evoke below threshold excitatory (PBS) solution. Cultures were incubated with primary an-
postsynaptic potentials. The excitatory postsynaptic po- tibodies at room temperature for 2 hours and at 4°C
tentials slope and amplitude were continuously moni- overnight. The following primary antibodies were used:
tored and digitized on-line. (1) rabbit polyclonal antibodies to glial fibrillary acidic
Slices were exposed to radiofrequency energy (RF3C protein (GFAP) (1:400 dilution) (Dakopatts, Copenhagen,
plus; Radionics, Burlington, MA) by using 22-gauge SMK Denmark) to identify astrocytes; (2) mouse monoclonal
54-mm, 5-mm active tip needle (COTOP BV, Amsterdam, antibodies to isotype III ␤-tubulin (1:400 dilution) (Sigma,
The Netherlands). Cultures were divided into 3 treat- St Louis, MO) to identify neurons. These antibodies were
ment groups. Group I was exposed to pulsed radiofre- diluted in PBS/0.5% bovine serum albumin/0.3% Triton
quency energy at 500 kHz at a pulse rate of 2 Hz for a X-100 solution. Bound antibodies were revealed with
period of 120 seconds, with an adjusted voltage of 35 to rhodamine-conjugated sheep anti-mouse immunoglob-
45 V, to permit maximal temperature of 38°C. Group II ulin G (dilution 1:40) (Boehringer Mannheim Biochemi-
was exposed to pulsed radiofrequency at an adjusted cals, Rotkreuz, Switzerland) or fluorescein-conjugated
voltage to allow maximal temperature of 42°C. Group III sheep anti-rabbit immunoglobulin G (dilution 1:80) sec-
was exposed to 42°C thermal lesion by continuous radio- ondary antibodies (Boehringer Mannheim Biochemicals)
frequency energy for 120 seconds. diluted in PBS/0.5% bovine serum albumin solution.

Morphologic Studies Results

Primary cultures of newborn Sprague-Dawley rat cere-
bral cortices were prepared as described previously.28 Effects on Impulse Propagation and
Briefly, after removal of meningeal tissue, the cerebral Synaptic Transmission
cortex was mechanically dissociated in Hank’s calcium- Hippocampal organotypic slice cultures are well-suited
and magnesium-free medium and centrifuged at 1000 for quantitative analyses of the effects of various treat-
rpm for 10 minutes. The pellet was resuspended in Dul- ments on synaptic transmission. In these experiments,
becco’s modified Eagle’s medium supplemented with pulsed radiofrequency energy was applied through a
10% fetal calf serum and plated on polylysine-coated 0.5-mm needle tip placed on the Schaffer collateral fi-
coverslips in 35-mm Petri dishes with a seeding density of bers, while continuously monitoring synaptic responses
106 cells per Petri dish. Under these conditions, cultures generated in CA1 cells and evoked by generating an ac-
became confluent at the sixth day, and at this time, the tion potential in CA3 neurons (Fig 1A). The amplitude of
serum-containing medium was replaced by serum-free evoked synaptic responses thus correlated with the num-
medium (DMEM, Gibco, Basel, Switzerland; 15 ␮g/mL in- ber of activated fibers or the efficacy of synaptic trans-
sulin, 20 ␮g/mL transferrin, 20 nmol/L progesterone, 100 mission.
mmol/L putrescine, and 30 nmol/L sodium selenite) in the Cultures were exposed to 3 types of radiofrequency
presence of Ara-C (2.5 mmol/L). Under these conditions, energy: 38°C pulsed radiofrequency, 42°C pulsed radio-
neurons remained viable, assessed by morphologic crite- frequency, and 42°C continuous radiofrequency. Tem-
ria such as arborization pattern and cell shrinkage perature profiles for all groups are shown in Fig 2A. Tis-
throughout the experimental period. sue temperatures increased from 28°C (temperature in
ORIGINAL REPORT/Cahana et al 199

Figure 1. Modulation of evoked synaptic transmission by

pulsed or continuous radiofrequency (RF). (A) Experimental de-
sign illustrating the placement of the stimulating and recording
electrodes as well as the needle for pulsed or continuous RF on
top of an organotypic hippocampal slice culture. RF was applied Figure 2. Temporal course of temperature (A), impedance (B),
to Schaffer collaterals originating from CA3 neurons and mak- and current (C) after exposure of an organotypic hippocampal
ing synaptic contacts with CA1 pyramidal cells. (B) Changes in slice culture to pulsed or continuous radiofrequency (RF). Tem-
amplitudes of excitatory postsynaptic potentials (EPSPs) evoked perature and impedance were stable and did not differ be-
in CA3 and recorded in CA1 before and after application of tween the treatment groups (A, B). Note that values of imped-
pulsed and continuous RF at the indicated temperature. Results ance were due to the characteristics of the electrical circuit. As
are mean ⫾ standard error of mean of 6-7 different experiments expected, peak current in the pulsed mode was significantly
per condition. The traces are examples of field potentials re- higher than in the continuous mode (C).
corded before and 15 minutes after application of pulsed (top
traces) or continuous (bottom traces) RF in a representative ex-
periment.
pulsed radiofrequency produced only a transient change
followed by a fast and complete recovery, the reduction
the perfusion chamber) to 38°C, 42°C, and 42°C, respec- observed in the continuous radiofrequency group was
tively, within 15 seconds from stimulation and returned pronounced and lasting, without signs of recovery for at
to baseline after the conclusion of radiofrequency appli- least 15 minutes (Fig 1B) but up to 40 minutes in some
cation. In all groups, tissue impedance decreased pro- longer experiments. These results thus suggested that
gressively over time, but the changes were not signifi- continuous radiofrequency, but not pulsed radiofre-
cantly different between the conditions tested (Fig 2B). quency, had produced prolonged alteration of synaptic
As could be expected, the peak radiofrequency current transmission.
was markedly higher in the pulsed radiofrequency
groups, averaging 50 ⫾ 19 versus 1.5 ⫾ 0.6 mA in the Morphologic Studies
continuous radiofrequency group (Fig 2C). To test this possibility, we then investigated how
Fig 1B illustrates the changes in excitatory postsynaptic pulsed radiofrequency and continuous radiofrequency
potentials slope amplitudes observed before and after affected neuronal survival when applied to dissociated
application of the 3 radiofrequency energy modes. For cortical monolayers, in which cell fate can be easily mon-
each experiment, results were expressed as percent of itored over a few days. To evaluate and compare the
baseline values and then averaged across experiments (n possible damaging effects of pulsed radiofrequency and
⫽ 6 to 7). As illustrated, application of radiofrequency continuous radiofrequency, different distances between
energy resulted in a reduction of excitatory postsynaptic cell cultures and the tip of the stimulating needle were
potential amplitude in all groups. However, whereas tested. Exposure to pulsed radiofrequency (38°C and
200 Modulatory Effect of Pulsed Radiofrequency

border of the lesion was relatively sharp. Glial cells lying

adjacent to the lesion site showed no morphologic signs
of degeneration, but the intensity of the GFAP immuno-
staining was more intense than in other glial cells else-
where in the monolayer (Fig 3E). Neurons, revealed by
the specific ␤-tubulin III staining, also preserved their
morphology next to the lesion site (Fig 3F).

Discussion
Neuropathic pain is usually considered a contraindica-
tion for the use of thermal lesioning by radiofrequen-
cy.15 It makes little sense to perform a neurodestructive
procedure in the presence of altered neural function,
risking aggravating neural pathology (ie, deafferenta-
tion pain, neural damage). However, pulsed radiofre-
quency, in which short bursts of radiofrequency energy
are applied to the nerve, is thought to be a safer alter-
native, because there is no clinical evidence of neural
damage, with less postprocedure soreness as experi-
enced after thermal radiofrequency lesioning.22 The
mechanism by which pulsed radiofrequency works re-
mains unclear.
In this study we took advantage of in vitro models to
compare more directly the acute effects of pulsed radio-
frequency and continuous radiofrequency. The main
finding of our present study is that exposure of neurons
to pulsed radiofrequency resulted in a transient inhibi-
tion of evoked excitatory transmission with full recovery
of synaptic activity within a few minutes, whereas con-
tinuous radiofrequency resulted in a lasting blockade
that did not recover during the next 15 to 30 minutes.
The amplitude of evoked synaptic responses in the hip-
pocampus is known to correlate with the number of ac-
tivated axonal fibers. The observation of a lasting reduc-
tion of synaptic transmission after continuous
radiofrequency application strongly suggests therefore a
reduced capacity of Schaffer collaterals to propagate ac-
tion potentials, an effect that might very well result from
Figure 3. Effects of pulsed and continuous radiofrequency on damage to the tissue.
cortical monolayers 24 hours after stimulation. (A, B) In control Temperature is unlikely to be the main factor respon-
groups, astrocytes (labeled with GFAP) form a confluent mono- sible for this lasting inhibition because similar changes
layer with neurons (labeled with tubulin) situated on the top.
(C-F) Application of continuous radiofrequency at a distance of were reached during pulsed radiofrequency and contin-
1000 ␮m resulted in a complete destruction of astrocytes (C) and uous radiofrequency, even though the voltage applied
neurons (D) on the monolayer under the needle. (E) and (F) are during continuous radiofrequency was only a fraction of
higher magnifications illustrating that juxtaposed neurons re- the voltage applied during the active phase of pulsed
mained preserved. (G, H) Application of pulsed radiofrequency
at the same distance of 1000 ␮m resulted in no destruction of radiofrequency. It would seem that it is the mode of
astrocytes (G) and neurons (H) on the monolayer under the radiofrequency energy administration and not the tem-
needle. Bar: (A-D) 100 ␮m; (E-H) 55 ␮m. perature that accounts for the difference.
This study showed that both procedures could produce
42°C) and continuous radiofrequency (42°C) did not af- cellular damage when cells are very close to the tip of the
fect glial cell or neuronal morphology when needle was needle (⬍ 500 ␮m). At larger distances (⬎ 2 mm), no
placed at a distance of 2000 ␮m or more from the cortical effect of either pulsed radiofrequency or continuous ra-
monolayer. However, at 1000 ␮m, neuronal architecture diofrequency was observed. This is thus consistent with
under the needle was destroyed with continuous radio- the interpretation that radiofrequency energy applica-
frequency but not with pulsed radiofrequency energy. tion produces only very localized effects. However, be-
At distances of 500 ␮m or less, we found a complete tween 500 and 2000 ␮m, our protocol clearly showed a
destruction of the monolayer in all treatment groups (Fig differential alteration of cell survival by pulsed radiofre-
3C to F). quency and continuous radiofrequency; continuous ra-
The denuded area had a quasicircular morphology (Fig diofrequency was more damaging than pulsed radiofre-
3C, D) corresponding to the shape of the needle, and the quency.
ORIGINAL REPORT/Cahana et al 201

This result was thus perfectly in line with the electro- eters that might influence clinical outcome. Further-
physiologic data, which suggested that prolonged alter- more, differences in tissue milieu (in vitro culture systems
ation of synaptic transmission observed with continuous versus intact nervous system), mode of energy applica-
radiofrequency could indeed have resulted from dam- tion (directly on neurons versus percutaneous on nerve
age to axonal and dendritic fibers. It is interesting also fibers), developmental age of subjects (adult humans
that the effects were observed only in the area directly versus young rats), and the state of neural substrates
under the tip of the needle, with a sharp limit with the (non-nociceptive versus nociceptive stimulus) render
surrounding tissue that remained perfectly preserved. clinical appreciation difficult. However having said that,
Thus radiofrequency energy application appeared to a differential biologic behavior does exist.
produce acute effects that decreased progressively as a Our data indicate that pulsed radiofrequency is less
function of distance and localization in relation to the tip damaging than continuous radiofrequency, even when
of the needle, with pulsed radiofrequency being less adjusting voltages to reach similar temperatures. This
damaging than continuous radiofrequency. These obser- suggests that probably we must abandon the traditional
vations could be compatible with the assumption that concept of thermal versus nonthermal radiofrequency
pulsed radiofrequency results in a neuromodulatory lesioning and regard the mode of application as the
rather than a neurodestructive effect when compared to main factor in lesion mechanism. If this is the case, we
continuous radiofrequency. might conclude that in our experiments pulsed radiofre-
Other studies suggested indirectly this neuromodula- quency showed an acute neuromodulatory effect (ie,
tory effect. High frequency stimulation might induce transient inhibition of evoked synaptic activity), whereas
long-term depression in the spinal cord16,17, and expo- continuous radiofrequency exhibited rather a neurode-
sure of dorsal root ganglion to pulsed radiofrequency structive effect (ie, lasting inhibition associated with tis-
currents activated c-fos protein expression in lamina I sue disruption). This might be of clinical relevance be-
and II neurons, not seen with continuous radiofrequen- cause pulsed radiofrequency might be a “softer” version
cy.9 Radiofrequency energy at subcytotoxic levels might of radiofrequency lesioning, appropriate when alter-
induce dramatic metabolic changes without morpho- ations of functional neural pathways exists, such as in
logic cell changes, and distinct “zones” can be identi- neuropathic pain states.
fied.7
Hyperthermic neuromodulation might involve an in- Conclusion
crease in cell membrane permeability, affecting impulse If pain chronifies through centralization, a neuroabla-
propagation and inducing protein synthesis such as heat tive procedure with no peripheral input remains sense-
shock protein 7012; however, there was no indication less.20 Pulsed radiofrequency was advocated as a nonde-
that temperature played a role in the effect of pulsed structive pain therapy, on the basis of the fact that
radiofrequency.20 Conversely, vanilloid receptors, which patients treated with pulsed radiofrequency did not
are heat gated ion channels, might be involved in the show clinical signs of nervous tissue destruction.22 How-
pain modulation as seen when exposing dorsal root gan- ever this evidence was indirect, and neural tissue injury
glion to pulsed radiofrequency energy.14 Another possi- might actually occur.23 The fact that fluctuating hyper-
bility suggested by experiments on fibroblast-like cells thermia by pulsed administration of radiofrequency en-
exposed to pulsed radiofrequency was an up-regulation ergy instead of continuous radiofrequency might give
of identified and unidentified genes (c-fos, c-myc, c-jun), rise to a neuromodulating and a nondestructing effect
all probably as a result of membrane perturbation with- remains speculative.
out rupture, a phenomenon shown in dielectrophoretic Our study suggests that this hypothesis might be true,
manipulation through pulsed radiofrequency applica- and that 38°C and 42°C pulsed radiofrequency treat-
tion in which functional modulation was associated ments provide a “stunning” transitory modulation of
without loss of cellular integrity.1 neural transmission with less disruptive morphologic
Presently, it is difficult to relate the above mentioned changes, as compared to continuous 42°C radiofre-
findings to clinical data. Lesion parameters (sensory and quency. The actual neuromodulatory mechanisms re-
motor stimulation thresholds), electrode position, lesion main to be elucidated and will definitely continue to fuel
duration, and local tissue properties might all be param- an already ongoing heated debate.2,4

References frequency lesions, transiently modulates excitatory synaptic

transmission in organotypic nervous tissue cultures. J Pain
1. Archer S, Li TT, Evans AT, Britland ST, Morgan H: Cell 3:25, 2002
reactions to dielectrophoretic manipulation. Biochem Bio-
physical Commun 257:687-698, 1999 4. Carr D: Reply to Dr Bogduk. Reg Anesth Pain Med 27:439-
440, 2002
2. Bogduk N: In defense of radiofrequency neurotomy. Reg
Anesth Pain Med 27:439-440, 2002 5. Dubuisson D: Treatment of occipital neuralgia by par-
tial posterior rhizotomy at C1-3. J Neurosurg 82:581-586,
3. Cahana A, Muller D: Pulsed radiofrequency but not radio- 1995
202 Modulatory Effect of Pulsed Radiofrequency

6. Ferrante FM, King LF, Roche EA, Kim PS, Aranda M, 18. Slappendel R, Crul BJ, Braak GJ, Geurts JW, Booij LH,
DeLaney LR, Mardini IA, Mannes AJ: Radiofrequency sacro- Voerman VF, De Boo T: The efficacy of radiofrequency le-
iliac joint denervation for sacroiliac syndrome. Reg Anesth sioning of the cervical dorsal root ganglion in a double
Pain Med 26:137-142, 2001 blinded randomized study: No difference between 40°c and
67°c treatments. Pain 73:159-163, 1997
7. Geurts JW, Roelof MvW: RF-DRG: A review of current con-
cepts. In: Geurts J, Roelof MvW (eds): Minimally Invasive 19. Sluijter ME: The role of radiofrequency in failed back
Procedures in the Treatment of Chronic Low Back Pain. Am- surgery patients. Curr Rev Pain 4:49-53, 2000
sterdam, The Netherlands, Thela Thesis, 2001, 90-114
20. Sluijter ME: Radiofrequency. Meggen (LU), Switzerland,
8. Geurts JW, Roelof MvW, Stolker RJ, Groen GJ: Efficacy of Flivo Press, 2001
radiofrequency procedures for the treatment of spinal pain:
A systematic review of randomized clinical trials. Reg Anesth 21. Sluijter ME, Cosman E, Rittman W, van Kleef M: The
Pain Med 26:394-400, 2001 effect of pulsed radiofrequency fields applied to the dorsal
root ganglion: A preliminary report. Pain Clin 11:109-117,
9. Higuchi Y, Nashold BS, Sluijter ME, Cosman E, Pearlstein 1998
RD: Exposure of dorsal root ganglion in rats to pulsed radio-
frequency activates dorsal horn lamina I and II neurons. Neu- 22. Sluijter ME, van Kleef M: Characteristics and mode of
rosurgery 50:850-856, 2002 action of radiofrequency lesions. Curr Rev Pain 2:143-150,
1998
10. Houpt JC, Conner ES, McFarland EW: Experimental study
of temperature distribution and thermal transport during 23. Stolker RJ, Groen GJ, Matthijssen MA, Hofstee-Hooft-
radiofrequency current therapy of the intervertebral disc. man TWA, Wolferen WJ, Teepen JL: Histopathological anal-
Spine 21:1808-1812, 1996 ysis of percutaneous partial radiofrequency rhizotomy: A
first report in humans, in Abstract book 8th World Congress
11. Lord SM, Barsley L, Bogduk N: Percutaneous radiofre- on Pain. Seattle, WA, IASP Press, 1996, p 463
quency neurotomy in the treatment of cervical zygapophy-
sial joint pain: A caution. Neurosurgery 36:732-739, 1995 24. Stoppini L, Buchs PA, Muller D: A simple method for
organotypic cultures of nervous tissue. J Neurosci Methods
12. Manzerra P, Brown IR: Expression of heat shock genes in 37:173-182, 1991
the rabbit spinal cord: Localization of constitutive and hy-
perthermia inducible mRNA species. J Neurosci Res 31:606- 25. van Kleef M, Barendse GA, Dingemans WA, Wingen C,
615, 1992 Lousberg R, Lange dS, Sluijter ME: The effects of producing
a radiofrequency lesion adjacent to the dorsal root ganglion
13. Munglani R: The longer term effect of pulsed radiofre- in patients with thoracic segmental pain. Clin J Pain 11:325-
quency for neuropathic pain. Pain 80:437-439, 1999 332, 1995
14. Nagy I, Rang H: Noxious heat activates all capsaicin sen-
26. van Kleef M, Barendse GA, Kessels A, Voets HM, Weber
sitive and also a sub-population of capsaicin insensitive dor-
WE, Lange dS: Randomized trial of radiofrequency lumbar
sal root ganglion neurons. Neuroscience 88:995-997, 1999
facet denervation for chronic low back pain. Spine 24:1937-
15. Patt RB, Cousins MJ: Techniques for neurolytic blockade. 1942, 1999
In: Cousins MJ, Bridenbaugh PO (eds): Neural Blockade, 3rd
27. van Suijlekom HA, van Kleef M, Barendse GA, Sluijter
edition. Philadelphia, PA, Lippincott Raven, 1998, 1007-1062
ME, Sjaastad O, Weber WE: Radiofrequency cervical zyg-
16. Randiae M, Jiang MC, Cerne R: Long term potentiation apophyseal joint neurotomy for cervicogenic headache: A
and long term depression of primary afferent neurotrans- prospective study of 15 patients. Funct Neurol 13:297-303,
mission in the rat spinal cord. J Neurosci 13:5228-5241, 1993 1998
17. Sandkühler J, Chen JG, Cheng G, Randiae M: Low fre- 28. Vutskits L, Djebbara-Hannas Z, Zhang H, Paccoud JP, Dur-
quency stimulation of afferent A␦-fibers induces long term bec P, Rugon G, Muller D, Kiss JZ: PSA-NCAM modulates
depression at primary afferent synapses with substantia ge- BDNF-dependent survival and differentiation of cortical
latinosa neurons in the rat. J Neurosci 17:6483-6491, 1997 neurons. Eur J Neurosci 13:1391-1402, 2001