Ò
PAIN 155 (2014) 356–366
www.elsevier.com/locate/pain
Analgesic treatment with pregabalin does not prevent persistent pain after peripheral nerve injury in the rat Fang Yang a, John Whang a, William T. Derry b,c, Daniel Vardeh b,d, Joachim Scholz a,e,⇑ a
Department of Anesthesiology, Columbia University College of Physicians and Surgeons, New York, NY, USA Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA c Department of Radiology, David Geffen School of Medicine at the University of California Los Angeles, Los Angeles, CA, USA d Department of Neurology, Massachusetts General Hospital, Boston, MA, USA e Department of Pharmacology, Columbia University College of Physicians and Surgeons, New York, NY, USA b
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 11 October 2013 Accepted 21 October 2013
Keywords: Postoperative pain Neuropathic pain Preventive analgesia Pregabalin
a b s t r a c t Reducing the risk of chronic postoperative pain through preventive analgesia is an attractive therapeutic concept. Because peripheral nerve lesions are a major cause of chronic pain after surgery, we tested in rats whether analgesic treatment with pregabalin (PGB) has the capacity to mitigate the development of persistent neuropathic pain-like behavior. Starting on the day of spared nerve injury or 1 week later, we treated rats with a continuous intrathecal infusion of PGB (300 or 900 lg/24 hours) or vehicle for up to 28 days. Rats receiving early PGB treatment had almost normal withdrawal thresholds for punctate mechanical stimuli and were clearly less sensitive to pinprick or cold stimulation. The responses to punctate mechanical and cold stimulation were still reduced for a brief period after the infusion was terminated, but the difference from vehicle-treated rats was minor. Essentially, the analgesic effect of PGB was limited to the duration of the infusion, whether analgesia started at the time of surgery or with a delay of 1 week, independently of the length of the treatment. PGB did not suppress the activation of spinal microglia, indicating that analgesia alone does not eliminate certain pain mechanisms even if they depend, at least partially, on nociceptive input. Unexpectedly, intrathecal infusion of PGB did not inhibit the nerve injury-induced accumulation of its binding target, the voltage-gated calcium channel subunit a2d1, at primary afferent terminals in the spinal cord. Interference with the synaptic trafficking of a2d1 is not required to achieve analgesia with PGB. Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
1. Introduction Peripheral nerve injury is a major risk factor for chronic pain after surgery. Persistent pain occurs in 10%–50% of patients undergoing procedures such as thoracotomy, breast surgery, or inguinal hernia repair, which expose local nerves to possible damage from pressure or transection [33]. In as many as two-thirds of patients with chronic postoperative pain, the pain is of probable or definite neuropathic origin [27]. Neuropathic pain following surgery involves multiple mechanisms. Injured and adjacent unaffected nerve fibers develop spontaneous activity and become more sensitive to mechanical or thermal stimulation [9,68]. This rise in input leads to enhanced synaptic efficacy in central nociceptive pathways, a process termed central sensitization [36]. Nociceptive ⇑ Corresponding author. Address: Columbia University Medical Center, Department of Anesthesiology, 630 West 168th Street, P&S Box 46, New York, NY 10032, USA. Tel.: +1 212 305 1274; fax: +1 212 304 6539. E-mail address: [email protected] (J. Scholz).
transmission may increase further as spinal inhibition is reduced and descending modulation from the brainstem shifts toward facilitation [17,26,56,65]. Additional sources of nociceptive input include protracted wound healing or sustained inflammation, which may occur after the implantation of synthetic material, for example, a prosthetic mesh for hernia repair [9,58]. The anatomy of an operation site often limits the options for modifying surgical techniques in order to minimize the risk of nerve injury. Pharmacological strategies that have been developed to reduce the incidence, duration, or intensity of postoperative pain can be divided into preemptive and preventive approaches. Preemptive analgesia aims to curb the acute postoperative pain by starting pain treatment ahead of the surgery. A variety of medications have been clinically tested for this purpose, with mixed results [4,45]. The concept of preventive analgesia is based on the hypothesis that the risk of chronic pain can be decreased by blocking pain mechanisms (such as central sensitization) that depend on nociceptive input. However, evaluating the efficacy of preventive analgesia in a clinical trial is difficult, because multiple
0304-3959/$36.00 Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2013.10.024
Ò
F. Yang et al. / PAIN 155 (2014) 356–366
variables, including preoperative pain, surgical technique, and anesthetic regimen, need to be controlled to ensure comparability of the treatment groups [18]. In addition, it is unknown when the transition from acute to chronic postoperative pain occurs, or whether there is a distinct transition period at all [58]. We employed a rat model of nerve injury-induced chronic pain to test whether analgesia can achieve a ‘‘disease-modifying’’ effect beyond the immediate attenuation of pain-like behavior. We treated the rats with pregabalin (PGB), a first-line recommended analgesic for neuropathic pain. PGB binds to the a2d1 subunit and, with less affinity, the a2d2 subunit of voltage-gated calcium channels (VGCC) [24]. PGB interferes with the synaptic targeting of VGCCs [5,30] and, as a result, reduces glutamate and neuropeptide release [35,66]. PGB is considered a prime candidate for the prevention of postoperative pain because of its efficacy against neuropathic pain, even though not every patient responds to it equally well [18,25]. The drug does not impede motor or somatosensory functions other than pain, an important advantage over sodium channel blockers, which also possess the capacity to inhibit activity-dependent changes in nociceptive transmission [4,64]. And it is better tolerated than antagonists of the N-methyl-d-aspartate (NMDA)-type glutamate receptor, which have been used to decrease central sensitization [22,71]. 2. Methods 2.1. Animals We used male Sprague-Dawley rats (Charles River Laboratories) aged 2–3 months for all experiments. Animal procedures adhered to the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain and were approved by the Subcommittee on Research Animal Care of Massachusetts General Hospital and the Institutional Animal Care and Use Committee of Columbia University Medical Center. 2.2. Peripheral nerve injury Surgery for spared nerve injury (SNI) was performed on animals anesthetized using 3% isoflurane inhalation for induction and 2% isoflurane during maintenance. We ligated the tibial and the common peroneal nerves with nylon (5–0) and transected them distally, leaving intact the third branch of the sciatic nerve, the sural nerve [19]. 2.3. Drug treatment PGB was provided by Pfizer. We dissolved the drug in phosphate-buffered saline (PBS), pH 7.4, and employed subcutaneously implanted osmotic pumps (Alzet, Cupertino, CA) for continuous intrathecal delivery of 300 or 900 lg PGB/24 hours at an infusion rate of 1 or 10 lL/h, respectively. PBS served as vehicle control. The osmotic pumps were implanted either at the time of nerve injury or 7 days later. They were connected to a polyurethane catheter (Alzet), which we inserted into the lumbar subarachnoid space with the catheter tip positioned on the dorsal surface of the spinal cord at segmental level L3. We verified the integrity of the catheter by dissection following completion of the experiments. Seven animals with neurological deficits or signs of infection after the catheter implantation were excluded from the experiments. 2.4. Behavioral testing Investigators were blind to the treatment in all experiments. Behavioral assessments were performed in the morning to avoid interference of circadian differences in animal activity.
357
Somatosensory function was examined after habituating the rats to the testing environment. We obtained 2 baseline measurements in the week preceding the surgery. Following SNI, we tested neuropathic pain-like behavior at defined intervals during the infusion of PGB or vehicle and for up to 12 days after the treatment was terminated. The animals were placed on an elevated wire grid and stimulated on the plantar surface of the hind paw, in the territory of the ‘‘spared’’ sural nerve. We used calibrated von Frey monofilaments of logarithmically increasing force (range 0.0174 to 34.7 g) to determine the withdrawal threshold for punctate mechanical stimulation. The threshold was defined as the lowest force that provoked a brisk paw withdrawal at least twice in 10 applications. To test for mechanical hyperalgesia, we measured the withdrawal duration after pricking the plantar surface of the hind paw with a medium-sized safety pin. The response to cold stimulation was tested by applying a drop of acetone, which produces a cool sensation on the skin upon evaporation. The acetone drop was dispensed from a syringe without touching the skin. We recorded behavioral responses to the acetone application over 1 minute and measured the cumulative time the animal spent licking, shaking, or lifting the paw [19,55]. Motor performance was evaluated in a group of uninjured rats 7 days after the intrathecal drug treatment started. We assessed walking and spontaneous exploratory behavior in an open field. To test the righting reflex, we turned the rats on their backs and observed how they regained a normal upright position through coordinated twisting of the body. We considered the reflex intact if the rats righted themselves promptly and successfully [10]. To evaluate the hopping response, the examiner supported the trunk of the animal, fixed one hind leg in his hand and moved the body laterally. We recorded the ability of the rat to hop with the weight-bearing limb in the direction of the movement. Righting and hopping were scored as 1 if intact or 0 if compromised. The stepping reflex was evoked by drawing the dorsum of the hind paw across the edge of a table while supporting the animal’s trunk. A normal reflex consists of a flexion and upward movement of the hind leg and repositioning of the paw with toes spread on the surface of the table. We rated the performance as 0 if the repositioning failed, 1 for severe impairment, 2 for slight impairment, and 3 if the repositioning was completed without deficit [62]. 2.5. Immunohistochemistry We deeply anesthetized the rats and perfused them transcardially with PBS, followed by a phosphate-buffered solution of 4% paraformaldehyde. The L4 segment of the spinal cord was dissected, postfixed for 2 hours, cryoprotected overnight in 20% sucrose, and embedded in Tissue-Tek (Sakura Finetek, Torrance, CA). Transverse sections through the spinal cord were cut at a thickness of 10 lm on a cryostat. We blocked unspecific protein binding sites by incubating the sections for 1 hour in PBS containing 0.5% bovine serum albumin (Sigma-Aldrich, St. Louis, MO), 1% blocking reagent (Roche Applied Science, Madison, WI) and 0.1% Triton X-100 (Sigma-Aldrich). The sections were incubated overnight at 4°C with primary antibodies directed against the VGCC subunit a2d1 (1:500, produced in mouse; Sigma-Aldrich), Cd11b (1:500, mouse; AbD Serotec, Raleigh, NC) or ionized calcium-binding adaptor molecule 1 (1:500, rabbit; Wako Chemicals, Richmond, VA). The immunostaining was completed by incubating the sections for 1 or 2 hours at room temperature with species-specific secondary antibodies that were conjugated with Alexa Fluor dyes (1:500; Life Technologies, Grand Island, NY). Immunodetection of a2d1 was enhanced by warming the slides in Tris buffer (pH 8.0) for 1 hour at 70°C before the blocking of unspecific protein binding sites [67]. We used the antifading reagent ProLong Gold (Life Technologies) to mount the stained sections.
Ò
358
F. Yang et al. / PAIN 155 (2014) 356–366
To analyze the intensity of immunofluorescent staining for
a2d1, we scanned spinal cord sections (optical thickness 1.0 lm) of 5–7 animals in each experimental condition on an A1 Confocal System (Nikon), using a 10 CFI Plan Fluor objective (numerical aperture 0.30). A rectangular region of interest (150 435 lm) was marked at the boundary between the medial and central thirds of the dorsal horn to capture a2d1 expression in the projection territories of the injured tibial and common peroneal nerves [46,72]. We used ImageJ software to measure the fluorescence intensity in this region [53]. Fluorescence intensity and area under the intensity curve were normalized to the contralateral side [5]. 2.6. Western blotting Western blots were performed as previously described, with slight modifications [55]. After euthanizing the rats, we dissected the ipsilateral L4 dorsal horns and homogenized the tissue with a motorized pestle. The protein extraction buffer (pH 7.5) contained 150 mM sodium chloride, 50 mM Tris, 1 mM ethylenediaminetetraacetic acid (EDTA), 2% sodium dodecyl sulfate, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, Complete Mini (without EDTA) protease inhibitors (Roche Applied Science) and Phosphatase Inhibitor Cocktails 1 and 2 (Sigma-Aldrich). We employed NuPAGE Novex 4–12% Bis-Tris gels for the electrophoresis (Life Technologies) and transferred the protein to an Immobilon-FL polyvinylidene fluoride (PVDF) membrane (EMD Millipore, Darmstadt, Germany). We blocked unspecific protein binding sites with Odyssey buffer (Li-Cor, Lincoln, NE) for 1 hour at room temperature and incubated the PVDF membranes overnight at 4°C with primary antibodies for a2d1 (1:1000, mouse; Sigma-Aldrich), ionized calcium-binding adaptor molecule 1 (1:500, rabbit; Wako Chemicals) or phosphorylated p38 mitogen-activated protein kinase (1:300, rabbit; Cell Signaling Technology, Beverly, MA), diluted in blocking buffer containing 0.1% Tween 20. Protein–antibody complexes were detected by incubating the membranes for 1 hour at room temperature with species-specific IRDye-conjugated secondary antibodies (1:20,000; Li-Cor) in a blocking buffer that contained 0.1% Tween 20 and 0.02% sodium dodecyl sulfate. The intensity of fluorescent bands was measured on an Odyssey Infrared Imaging System (Li-Cor). We immunostained for tubulin b3 (1:10,000; Sigma-Aldrich) to determine sample loading and normalize the fluorescent signal intensity of immunoreactive bands. 2.7. Statistics We used Prism 5 (GraphPad Software, La Jolla, CA) for statistical analysis. To compare somatosensory responses between treatment groups, we employed a two-way analysis of variance (ANOVA) followed by Sidak’s test for multiple comparisons. Righting and hopping scores were analyzed with the v2 test, and scores for paw placement with one-way ANOVA. An unpaired student’s t test was used to assess differences in the immunofluorescent a2d1 staining of tissue sections or the Western blot results. Data are presented as means ± SEM. Differences were considered significant if P < 0.05. 3. Results 3.1. Short-term treatment with PGB Nerve injury led to a decrease in the withdrawal threshold for punctate mechanical stimulation with von Frey filaments, an exaggerated response to pinprick, and a pain-like response to the normally nonpainful cold sensation produced by acetone (Fig. 1). Intrathecal infusion of PGB, starting on the day of the surgery, effectively reduced both mechanically and thermally evoked responses.
One week after SNI, the withdrawal threshold for stimulation with von Frey filaments was 6.6 ± 2.0 g in rats receiving 300 lg PGB/24 hours compared with 0.4 ± 0.1 g in vehicle-treated rats. The withdrawal duration after pinprick was 2.9 ± 1.0 seconds compared with 12.1 ± 2.1 seconds, and the withdrawal response after the application of acetone lasted 8.3 ± 2.6 seconds, substantially shorter than the average withdrawal duration of 12.6 ± 2.1 seconds in vehicle-treated rats (Fig. 1). The infusion of 900 lg PGB/24 hours produced a similar attenuation of pain-like behavior: 1 week after SNI, the withdrawal threshold after punctate mechanical stimulation was 7.3 ± 2.3 g, the withdrawal duration after pinprick was 3.6 ± 0.9 seconds, and acetone-evoked responses lasted 5.8 ± 2.0 seconds (Fig. 1). The difference between PGB and vehicle treatment was significant for all tested manifestations of neuropathic pain-like behavior (P < 0.001 for the mechanical stimulations, P < 0.05 for the acetone test, determined in a two-way ANOVA covering the treatment period of 7 days). Pain reduction did not differ between the low and high doses of intrathecal PGB. The analgesic effect of PGB was only partially sustained beyond the week-long infusion. Withdrawal responses evoked by von Frey filaments or acetone remained attenuated for 6 days in rats previously treated with 900 lg PGB/24 hours (P < 0.05), but this difference was very small compared with the reduction of pain-like behavior during the infusion (Fig. 1). 3.2. Long-term analgesia To test whether longer treatment would have a greater influence on the course of neuropathic pain, we administered PGB for 28 days. We used only the lower dose of 300 lg PGB/24 hours in these experiments, because it produced the same acute analgesic effect as 900 lg PGB/24 hours and did not interfere with movement coordination, whereas rats receiving 900 lg PGB/24 hours scored lower in the righting and hopping tasks (P < 0.05; Fig. 2). Stepping, spontaneous walking, and exploratory behavior were unaffected by either dose of PGB. We conducted two experiments, one in which we started longterm treatment at the time of nerve injury, and one in which the treatment was delayed by 7 days. The first experiment was designed to preempt the nerve injury-induced increase in pain sensitivity. With the second approach, we modeled a situation in which analgesic therapy begins after neuropathic pain has developed. Vehicle-treated rats exhibited neuropathic pain-like behavior (low withdrawal thresholds to punctate mechanical stimuli, increased withdrawal responses to pinprick or acetone) for 47 days after SNI, the maximum length of our experiments. Mechanically induced and cold-induced pain-like behavior was reduced by PGB in both treatment models. After 4 weeks of PGB infusion with early onset, the withdrawal threshold to punctate mechanical stimulation was 5.3 ± 2.0 g compared with 0.9 ± 0.7 g in vehicle controls (P < 0.001 for the treatment effect over 28 days, determined in a two-way ANOVA; Fig. 3). The withdrawal duration after pinprick was 2.1 ± 1.5 seconds compared with 5.5 ± 1.5 seconds (P < 0.01), and the response to cold stimulation lasted for 1.0 ± 0.4 seconds compared with 8.4 ± 2.7 seconds (P < 0.001; Fig. 3). The analgesic effect of PGB was noticeably weaker when the intrathecal infusion started with a delay. Rats in which the PGB infusion began 7 days after SNI, when neuropathic pain-like behavior was already established, showed a withdrawal threshold of 1.0 ± 0.5 g to punctate mechanical stimulation after 28 days of treatment. This threshold was close to that of 0.8 ± 0.2 g in vehicle-treated rats. However, the difference between rats receiving PGB and those receiving a vehicle infusion was still clear and significant when compared over the entire treatment period of
Ò
359
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10
**
*
3
Vehicle ipsi PGB 300 µg/24 h ipsi PGB 900 µg/24 h ipsi
1
Vehicle contra PGB 300 µg/24 h contra PGB 900 µg/24 h contra
0.3 Treatment
0.1 B1 B2
5
SNI
10
15
Days after surgery
Pinprick Withdrawal duration (s)
14 12 10 8 6
***
4 2
***
0 B1 B2
5
10
15
Acetone
Withdrawal duration (s)
18 16 14 12 10 8
#
6 4 2 0 B1 B2
5
10
15
Fig. 1. Short-term pregabalin (PGB) treatment. Continuous intrathecal infusion of PGB (300 or 900 lg/24 hours) or vehicle started at the time of spared nerve injury (SNI) and lasted 7 days (shaded area). Pain-like behavior was tested for 15 days after surgery. We used calibrated von Frey filaments to determine the withdrawal threshold for punctate mechanical stimulation. The withdrawal duration after pinprick was recorded to assess the reaction to mechanical stimulation above the pain threshold. The response to cold was tested with evaporating acetone and measured as the time the animal spent licking, shaking, or lifting the paw. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle treatment in all 3 tests during treatment (P < 0.001 for the stimulation with von Frey filaments or pinprick, P < 0.05 for the response to acetone). For 6 days after termination of the treatment, rats that had received an infusion of 900 lg PGB/24 hours remained less sensitive to the stimulation with von Frey filaments or acetone when compared with vehicle-treated rats (P < 0.05). Asterisks and hashes indicate the results of Sidak’s follow-up tests. ⁄ P < 0.05, ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001 for the comparison of PGB with vehicle treatment; #P < 0.05 for the comparison of 300 lg with 900 lg PGB/24 hours. N = 8–9. B1 and B2, baseline assessments.
28 days (P < 0.01; Fig. 4). An initial increase in the withdrawal threshold of vehicle-treated rats, from 0.6 ± 0.2 g on day 7 to 1.9 ± 0.7 g on day 14 after SNI, may have been a prolonged effect of the inhalation anesthesia that we administered during the second surgery for catheter placement and osmotic pump implantation. Withdrawal responses after pinprick lasted 0.5 ± 0.0 seconds at the end of the 4-week treatment with PGB compared with 2.0 ± 0.7 seconds in vehicle-treated rats (P < 0.01). Responses to cold stimulation were recorded over 3.7 ± 1.3 seconds compared with 4.6 ± 1.4 seconds (P < 0.05; Fig. 4). The rats tolerated the continuous infusion of 300 lg PGB/24 hours well; we did not observe adverse effects. Long-term PGB treatment did not have a sustained analgesic effect beyond the infusion, irrespective of the timing of treatment
onset. Neuropathic pain-like behavior after 28 days of PGB treatment was indistinguishable from the responses to test stimuli in rats that received a vehicle infusion (Figs. 3 and 4). 3.3. VGCC subunit a2d1 in the dorsal horn Following nerve injury, somatosensory neurons upregulate the expression of VGCC subunit a2d1. Trafficking of the subunit to the central terminals of primary afferents, rather than local induction, is responsible for a subsequent increase in a2d1 in the superficial dorsal horn of the spinal cord [5,38,41]. To determine the effect of PGB on the presynaptic enrichment of a2d1, we immunostained transverse cryosections through the L4 segment of the spinal cord for the VGCC subunit. We measured the intensity of
Ò
360
F. Yang et al. / PAIN 155 (2014) 356–366
Righting
Hopping
Stepping 3
*
60
*
40
Score
80
2 1
20 0
4
4
h
h
le hi c
µg /2
µg /2
0 90 B
PG
PG
PG
B
B
90
30
0
0
µg /2
µg /2
4
4
h
h
le hi c 30 B PG
PG
B
90
30
0
0
Ve
4 µg /2
µg /2 0
Ve B PG
h
h 4
hi c
le
0 Ve
Successful completion (%)
100
Fig. 2. Coordinated movement tasks. Uninjured rats were tested following intrathecal infusion of 300 or 900 lg PGB/24 hours or vehicle for 7 days. The 2 bar graphs on the left show the percentage of rats successfully completing the righting or hopping tasks. Successful stepping after a passive extension of the ankle joint was scored as 3, whereas a score of 0 reflects task failure. ⁄P < 0.05, determined in a v2 test. N = 10–12. PGB, pregabalin.
the fluorescent staining in the medial two-thirds of the dorsal horn, which receive most of their afferent input from the tibial and common peroneal nerves [46,72], the two branches of the sciatic nerve that are ligated and cut during SNI (Fig. 5A). Seven days after SNI, the integrated fluorescence intensity in the ipsilateral dorsal horn of untreated rats was 177.6% ± 15.8% (relative to the contralateral dorsal horn), compared with 104.8% ± 10.5% in sham-operated animals (P < 0.01). The intensity was highest, 145.0% ± 8.5% (relative to the maximum fluorescence intensity in the contralateral dorsal horn), in lamina II, at a distance of 100 lm from the outer contour of the dorsal horn (Fig. 5B). In lamina III, at 200 lm from the contour, we measured an intensity of 73.7% ± 11.0%. The corresponding values in sham-operated animals were 87.8% ± 4.9% (P < 0.001) and 47.0% ± 4.8% (not significant) (Fig. 5B). The nerve injury-induced change in a2d1 immunostaining was limited to the ipsilateral dorsal horn; the fluorescence intensity in the contralateral dorsal horn of rats after SNI and rats after sham surgery was not different. Intrathecal treatment with PGB did not reduce the increase in a2d1 immunoreactivity (Fig. 5C). Seven days after SNI, the integrated fluorescence intensity in the ipsilateral dorsal horn of rats receiving a continuous infusion of 300 lg PGB/24 hours was 182.3% ± 21.4% (relative to the contralateral dorsal horn) compared with 178.8% ± 7.6% in vehicle-treated animals. The signal intensities in laminas II and III were 156.1% ± 18.7% and 76.1% ± 9.5% (relative to the maximum intensity in the contralateral dorsal horn) compared with 167.0% ± 13.1% and 84.2% ± 7.9%, respectively, in vehicle-treated animals (Fig. 5D). None of these differences were significant, nor did the fluorescence intensity in the contralateral dorsal horn differ between rats treated with PGB and rats receiving a vehicle infusion. Western blot analysis confirmed an increase in a2d1 protein to 137.1% ± 6.6% in the ipsilateral dorsal horn 7 days after SNI compared with 100.0% ± 14.38% after sham surgery (P < 0.05). Continuous infusion of 300 lg PGB/24 hours did not reduce a2d1 protein in the dorsal horn (Fig. 6). 3.4. Microglial activation Primary afferent input drives important mechanisms of neuropathic pain, for example, central sensitization in the dorsal horn or the activation of spinobulbospinal circuits involved in descending facilitation [36,50,65]. Increases in nociceptive and non-nociceptive input also contribute to the activation of spinal microglia [28,55,62,70], a process that is crucial for the development of pain after nerve injury [57]. We tested whether analgesia provided by a
continuous infusion of 300 lg PGB/24 hours, beginning at the time of SNI, is sufficient to suppress microglial activation in the dorsal horn of the spinal cord. Microglia in the ipsilateral dorsal horn reacted strongly to SNI. Seven days after nerve injury, microglial cells showed markedly enhanced immunostaining for Cd11b (also known as integrin aM) and ionized calcium-binding adapter molecule 1 (Iba1, also known as allograft inflammatory factor 1). Cd11b or Iba1 immunostaining in rats treated with PGB did not differ from that in rats receiving a vehicle infusion (Fig. 7A). Western blot analysis showed a 5-fold increase in Iba1 protein at 7 days after SNI, to 496.3% ± 85.37% compared with 100.0% ± 13.27% after sham surgery (P < 0.001). PGB treatment did not lower Iba1 expression in the ipsilateral dorsal horn (Fig. 7B). We also examined the phosphorylation of p38 mitogen-activated protein kinase (MAPK), an enzyme prominently involved in microglial signaling pathways that regulate the synthesis of inflammatory mediators such as nitric oxide synthase 2 or interleukin 1b. The level of phosphorylated (active) p38 MAPK in rats receiving PGB was 87.4% ± 8.2% compared with 100.0% ± 6.4% in vehicle-treated rats (not significant; Fig. 7B), indicating that the inflammatory response to SNI in the dorsal horn was essentially unchanged. 4. Discussion By reducing afferent input or inhibiting signal transmission at nociceptive synapses, analgesics may prevent activity-dependent pain mechanisms, such as central sensitization, and have a sustained effect beyond the immediate attenuation of pain. At least this is the assumption behind the concept of preventive analgesia [18]. The main findings of our study challenge this concept. PGB, a first-line recommended medication for nerve injury-induced pain, produced analgesia only as long as the drug was administrated, without blocking the emergence of persistent pain. Somewhat unexpectedly, we also found that PGB achieved its analgesic effect without lowering the concentration of its binding target, the VGCC subunit a2d1, at the central terminals of primary afferents. Short-term experiments consisted of intrathecal PGB infusion for 1 week to provide analgesia during the development of painlike behavior after SNI [19]. Long-term treatment for 4 weeks was designed to cover the period during which central sensitization establishes, spinal inhibition decreases, descending pain modulation shifts toward facilitation, and microglia in the dorsal horn is activated [11,17,36,55,56]. As an inhibitor of neurotransmitter release, PGB was expected to interfere with these mechanisms of
Ò
361
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10
**
*
3 1
Vehicle ipsi PGB 300 µg/24 h ipsi
0.3 0.1
Vehicle contra PGB 300 µg/24 h contra
Treatment 0.03 B1B2
5
10
15
20
25
30
35
40
Days after surgery
SNI
Pinprick
Withdrawal duration (s)
14 12 10 8 6 4
*
2 0 B1B2
5
10
15
20
25
30
35
40
20
25
30
35
40
Withdrawal duration (s)
Acetone 20 18 16 14 12 10 8 6 4 2 0
* B1B2
5
10
15
Fig. 3. Long-term pregabalin (PGB) treatment with early onset. Intrathecal infusion of 300 lg PGB/24 hours or vehicle started at the time of spared nerve injury (SNI) and continued for 28 days (shaded area). Pain-like withdrawal responses to punctate mechanical stimulation (von Frey filaments), normally painful mechanical stimulation (pinprick), or cold stimulation (acetone) were tested for 40 days after surgery. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle during treatment in all 3 tests (P < 0.001 for the response to stimulation with von Frey filaments or acetone, P < 0.01 for the response to pinprick). However, PGB did not have an effect on pain-like behavior beyond termination of the infusion. ⁄P < 0.05, ⁄⁄P < 0.01 for PGB compared with vehicle, determined in Sidak’s follow-up test. N = 5–7. B1 and B2, baseline assessments.
neuropathic pain because their induction involves the enhanced activity of somatosensory afferents [68]. PGB effectively reduced pain after SNI. The decrease in pain-like behavior was smaller, though still significant, when the PGB infusion was delayed by 1 week, suggesting that the drug should be given as early as possible [35]. However, PGB failed to modify the chronic course of neuropathic pain, even when it was administered continuously for 4 weeks. Once the infusion ended, the rats exhibited sharply reduced withdrawal thresholds for mechanical stimulation and an increased sensitivity to pain evoked by pinprick or acetone evaporation. These behavioral changes, which are equivalent to the common features of chronic postoperative pain in humans— mechanical allodynia, mechanical hyperalgesia, and cold allodynia [1,69]—were indistinguishable from those in rats that received a vehicle infusion. Consistent with clinical experience in the longterm use of a2d1 ligands [52], the analgesic effect of PGB was
stable over time, indicating that tolerance was not responsible for the lack of pain prevention. Mainly sodium channel blockers have been evaluated for preventive analgesia after nerve injury in previous experimental studies. Lidocaine, bupivacaine, or mepivacaine were directly applied to the nerve or a dorsal root ganglion (DRG), or were administered in a single systemic or intrathecal injection. The compounds rarely attenuated neuropathic pain-like behavior for >1 week [64]. Longer or even complete prevention of pain hypersensitivity was only observed after ‘‘bathing’’ the sciatic nerve in 2% lidocaine or administering 10% lidocaine directly onto spinal nerve roots [8,60]. Such concentrated applications risk nerve damage. Slow-release formulations of sodium channel blockers provide a safer approach to obtaining sustained analgesia. Repeated injections of saxitoxin liposomes blocked sciatic nerve conduction for 18 days after SNI and delayed mechanical hypersensitivity in rats for 4 weeks [59].
Ò
362
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10 3 Vehicle ipsi PGB 300 µg/24 h ipsi
1 0.3
Vehicle contra PGB 300 µg/24 h contra
Treatment
0.1 B1 B2
5
10
15
20
25
30
35
40
45
50
30
35
40
45
50
30
35
40
45
50
Days after surgery
SNI
Pinprick Withdrawal duration (s)
4 3 2 1 0 B1 B2
5
10
15
20
25
Acetone Withdrawal duration (s)
14 12 10 8 6 4 2 0 B1 B2
5
10
15
20
25
Fig. 4. Long-term pregabalin (PGB) treatment with late onset. Intrathecal infusion of 300 lg PGB/24 hours or vehicle started 7 days after spared nerve injury (SNI) and continued for 28 days (shaded area). Pain-like withdrawal responses to punctate mechanical stimulation (von Frey filaments), normally painful mechanical stimulation (pinprick), and cold stimulation (acetone) were tested for 47 days after surgery. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle during treatment in all 3 tests (P < 0.01 for the stimulation with von Frey filaments or pinprick, P < 0.05 for the response to acetone). Differences after termination of the treatment were not significant. N = 5–8. B1 and B2, baseline assessments.
However, the rats exhibited heat hyperalgesia soon after the block, suggesting differential prevention of mechanically evoked pain. Timing and efficacy of the conduction block are crucial. Early and complete suppression of afferent activity with a faster-acting bupivacaine depot or continuous infusion of tetrodotoxin directly at the nerve lesion site shortened the duration of the blockade necessary to accomplish sustained analgesia [14,73]. Gradual release of sodium channel blockers from liposomes or microspheres may inhibit afferent activity partially or late and therefore require repeated administration to achieve pain prevention [59,63]. Notably, even repeated injections of saxitoxin liposomes did not alter the upregulation of a2d1 in somatosensory neurons [59]. Only nonselective sodium channel inhibitors are available for clinical use. When applied to mixed sensory and motor nerves, paralysis for the duration of the block is an inevitable risk [4,64]. Prolonged application of sodium channel blockers is therefore not feasible in humans, which explains the interest in analgesics
that might prevent pain without compromising other nerve functions. A neurokinin 1 receptor antagonist [12], high-dose remifentanil [42] and topical clonidine applied at the nerve lesion site [37] partially protect against pain hypersensitivity in the rat. The translational potential of some of these treatment approaches is uncertain, considering the clinical failure of neurokinin 1 receptor antagonists as analgesics. By contrast, the a2d1 ligands PGB and gabapentin are established analgesics for neuropathic pain [3] and produce preemptive analgesia after surgery [44,76]. However, clinical trials on pain prevention have yielded inconclusive results [16,34]. Differences between surgical procedures, anesthetic regimens, outcome measures, and lack of placebo control hamper the interpretation of clinical evidence [18]. Animal models allow studying the pathophysiology and treatment of postoperative pain under standardized conditions. We used this advantage to compare short-term and long-term infusions of PGB in a rat model of postoperative pain caused by nerve
Ò
363
F. Yang et al. / PAIN 155 (2014) 356–366
B
25 0
co
ip
140
m
Sh
am
###
160
Sh a
180
tra
0
50
i
Sham
75
ps
20
100
on
40
125
Ic
Ipsi Contra
60
150
Ii
80
175
SN
120 100
***
200
SN
140
si
Sham
160
225
nt ra
II III
## 180
Area under the curve (%)
Ipsilateral
Fluorescence intensity (%)
A
120 100 80 60 40 20
SNI
0 0
50 25
e SN
ve
hi
cl
hi ve SN
SN
100
I+
I+
140 120
nt co
B
co
nt
ip cl
e
180 160
ra
0
si
0
75
ip
SNI + vehicle
100
B
20
125
PG
40
150
I+
Ipsi Contra
60
175
ra
80
**
PG
100
200
I+
140 120
***
225
SN
180 160
si
SNI + vehicle
Dorsoventral distance (μm)
D Fluorescence intensity (%)
C
50 100 150 200 250 300 350 400 450
Area under the curve (%)
SNI
80 60 40
SNI + PGB
20
SNI + PGB
0 0
50 100 150 200 250 300 350 400 450
Dorsoventral distance (μm) Fig. 5. Immunostaining for the voltage-gated calcium channel subunit a2d1 in the dorsal horn. (A) Representative transverse cryosections, immunostained for a2d1, through the dorsal horns of untreated rats 7 days after sham surgery or spared nerve injury (SNI). Rectangles indicate the regions of interest (150 435 lm) in which we measured the fluorescence intensity. We placed the regions over the boundary between the medial and central thirds of the dorsal horn, where the axons of somatosensory fibers from the injured tibial and common peroneal nerves terminate [46,72]. (B) Fluorescence intensity profiles of the dorsal horn, normalized to the maximum value (100%) of the contralateral side. We calculated the mean intensity for each condition at distances of 100 and 200 lm from the outer contour of the dorsal horns to obtain representative values for a2d1 in laminas II and III, respectively. The areas under the intensity profiles are compared in the bar graph. (C) Representative dorsal horn sections of rats that received an intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI. Treatment began at the day of nerve injury. (D) Normalized fluorescence intensity profiles of the dorsal horn after SNI and vehicle or 300 lg PGB/24 hours treatment, respectively. Fluorescence intensities of the contralateral dorsal horns did not differ between conditions. Scale bars, 200 lm. N = 5. ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001 for the comparison of ipsilateral with contralateral dorsal horns; ##P < 0.01, ###P < 0.001 for the comparison of SNI with sham surgery, determined in unpaired t tests. PGB, pregabalin.
injury. We found that even PGB treatment for 4 weeks did not prevent the resurgence of high-intensity pain-like behavior once the infusion ended. Sustained analgesia for up to 7 days following intrathecal injection or infusion of gabapentin has been reported in rats and mice after spinal nerve ligation, but it is unclear whether true long-term prevention was achieved [15,47,48]. Pain hypersensitivity spontaneously decreased following the ligation and gabapentin may simply have accelerated this recovery [47,48]. In a recent clinical trial, the intrathecal infusion of gabapentin failed to reduce chronic pain [54]. However, few of the subjects who received gabapentin at a dose of 30 mg/24 hours, which
would be equivalent to the intrathecal infusion of PGB in our study, had neuropathic pain. Mixed treatment response in a heterogeneous trial population and gabapentin’s pharmacokinetic properties, rather than the intrathecal route of administration, may therefore explain the lack of a significant analgesic effect. PGB remains in the central nervous system longer than gabapentin and is the more potent analgesic [7]. Although afferent activity drives important neuropathic pain mechanisms, our findings indicate that some mechanisms, including the reaction of spinal glia to peripheral nerve lesion, operate at least partially independently. Continuous analgesia by PGB did not
Ò
F. Yang et al. / PAIN 155 (2014) 356–366
PG
cl
I+
hi ve I+ SN
v SN ehic I + le PG B
B
Tubulin β 3 e
am SN I
Sh
SN
Sh
I
Tubulin β 3
α2δ1
I+
α2δ1
140 120 100 80 60 40 20 0
SN
α2δ1 protein (%)
*
SN
160 140 120 100 80 60 40 20 0
am
α2δ1 protein (%)
364
Fig. 6. Western blot analysis of voltage-gated calcium channel subunit a2d1 in the dorsal horn. Quantification of a2d1 protein levels in the ipsilateral dorsal horn of untreated rats 7 days after sham surgery or spared nerve injury (SNI), and in rats that received an intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI, beginning at the time of nerve injury. The integrated fluorescent signal of labeled bands is expressed as the percentage relative to sham surgery for the comparison with untreated rats after SNI, or relative to vehicle treatment for the comparison with PGB. N = 5. ⁄P < 0.05, determined in an unpaired t test. PGB, pregabalin.
SNI + vehicle
SNI + PGB
CD11b
Iba1 protein (%)
A
B
***
600 500 400 300 200
Iba1
100
Tubulin β 3
100 80 60 40 20
80
40
Phospho p38
20
Tubulin β 3 hi
SN
I+
ve
h SN icle I+ PG B
0
ve I+
Tubulin β 3
60
0
SN
Iba1
100
cl SN e I+ PG SN B I+ ve SN hicl I+ e PG B
Iba1 protein (%)
I 120
Phosphorylated p38 Mapk protein (%)
120
Iba1
SN
am Sh
I SN
Sh
am
0
Fig. 7. Microglial activation after spared nerve injury (SNI). (A) Immunostaining for Cd11b or ionized calcium-binding adapter molecule 1 (Iba1). Representative images show the ipsilateral dorsal horn of rats treated with intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI, beginning at the time of nerve injury. (B) Quantification of Iba1 and phosphorylated p38 mitogen-activated protein kinase (MAPK) by Western blotting. Bar graphs show the integrated fluorescent signal of labeled bands expressed as the percentage relative to sham surgery for the comparison with untreated rats after SNI, or relative to vehicle treatment for the comparison with PGB. N = 4–6. PGB, pregabalin.
stop microglial activation after SNI. Although nociceptive and nonnociceptive input modulates glial cell proliferation and signaling pathways [28,62,70], nerve conduction blocks or interventions that target primarily nociceptive signaling may be insufficient to fully suspend glial responses. Inhibiting action potentials in injured peripheral axons attenuates the mobilization of spinal glia [59,74], but resident microglia, invading monocytes, and astroglia react to multiple chemokines such as Ccl2 or adenosine triphosphate, which are released not only from primary somatosensory neurons [57]. Complete interruption of afferent input by dorsal root transection is required to abolish the typical pattern of microglial activation in the dorsal horn [40,55]. On the other hand, our results demonstrate that analgesia can be achieved independently of microglial activity. The binding target for PGB, the VGCC subunit a2d1, is expressed in DRG neurons, most prominently in nociceptive neurons. Nerve injury leads to marked upregulation and synaptic trafficking of
a2d1 [5,38,41,49], which acts as a chaperone for the pore-forming VGCC subunit a1 and stabilizes its surface expression [13,31]. High VGCC density at axon terminals improves the coupling of calcium entry to transmitter exocytosis and is an important mediator of activity-dependent pain mechanisms [30,31]. Knocking a2d1 down after spinal nerve ligation decreases pain-like behavior [38], whereas a2d1 overexpression in uninjured mice results in abnormal pain sensitivity similar to neuropathic allodynia and hyperalgesia [39]. Continuous intrathecal treatment with PGB had no impact on a2d1 accumulation in the superficial dorsal horn. We did not evaluate a2d1 expression in rats receiving 900 lg PGB/ 24 hours, but the results after the infusion of 300 lg PGB/24 hours demonstrate clearly that the drug produced analgesia without disrupting a2d1 trafficking. Our findings are consistent with previous reports showing that intrathecal administration of gabapentin or PGB reduces neuropathic pain-like behavior despite incomplete (
PAIN 155 (2014) 356–366
www.elsevier.com/locate/pain
Analgesic treatment with pregabalin does not prevent persistent pain after peripheral nerve injury in the rat Fang Yang a, John Whang a, William T. Derry b,c, Daniel Vardeh b,d, Joachim Scholz a,e,⇑ a
Department of Anesthesiology, Columbia University College of Physicians and Surgeons, New York, NY, USA Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital, Boston, MA, USA c Department of Radiology, David Geffen School of Medicine at the University of California Los Angeles, Los Angeles, CA, USA d Department of Neurology, Massachusetts General Hospital, Boston, MA, USA e Department of Pharmacology, Columbia University College of Physicians and Surgeons, New York, NY, USA b
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
a r t i c l e
i n f o
Article history: Received 26 August 2013 Received in revised form 11 October 2013 Accepted 21 October 2013
Keywords: Postoperative pain Neuropathic pain Preventive analgesia Pregabalin
a b s t r a c t Reducing the risk of chronic postoperative pain through preventive analgesia is an attractive therapeutic concept. Because peripheral nerve lesions are a major cause of chronic pain after surgery, we tested in rats whether analgesic treatment with pregabalin (PGB) has the capacity to mitigate the development of persistent neuropathic pain-like behavior. Starting on the day of spared nerve injury or 1 week later, we treated rats with a continuous intrathecal infusion of PGB (300 or 900 lg/24 hours) or vehicle for up to 28 days. Rats receiving early PGB treatment had almost normal withdrawal thresholds for punctate mechanical stimuli and were clearly less sensitive to pinprick or cold stimulation. The responses to punctate mechanical and cold stimulation were still reduced for a brief period after the infusion was terminated, but the difference from vehicle-treated rats was minor. Essentially, the analgesic effect of PGB was limited to the duration of the infusion, whether analgesia started at the time of surgery or with a delay of 1 week, independently of the length of the treatment. PGB did not suppress the activation of spinal microglia, indicating that analgesia alone does not eliminate certain pain mechanisms even if they depend, at least partially, on nociceptive input. Unexpectedly, intrathecal infusion of PGB did not inhibit the nerve injury-induced accumulation of its binding target, the voltage-gated calcium channel subunit a2d1, at primary afferent terminals in the spinal cord. Interference with the synaptic trafficking of a2d1 is not required to achieve analgesia with PGB. Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
1. Introduction Peripheral nerve injury is a major risk factor for chronic pain after surgery. Persistent pain occurs in 10%–50% of patients undergoing procedures such as thoracotomy, breast surgery, or inguinal hernia repair, which expose local nerves to possible damage from pressure or transection [33]. In as many as two-thirds of patients with chronic postoperative pain, the pain is of probable or definite neuropathic origin [27]. Neuropathic pain following surgery involves multiple mechanisms. Injured and adjacent unaffected nerve fibers develop spontaneous activity and become more sensitive to mechanical or thermal stimulation [9,68]. This rise in input leads to enhanced synaptic efficacy in central nociceptive pathways, a process termed central sensitization [36]. Nociceptive ⇑ Corresponding author. Address: Columbia University Medical Center, Department of Anesthesiology, 630 West 168th Street, P&S Box 46, New York, NY 10032, USA. Tel.: +1 212 305 1274; fax: +1 212 304 6539. E-mail address: [email protected] (J. Scholz).
transmission may increase further as spinal inhibition is reduced and descending modulation from the brainstem shifts toward facilitation [17,26,56,65]. Additional sources of nociceptive input include protracted wound healing or sustained inflammation, which may occur after the implantation of synthetic material, for example, a prosthetic mesh for hernia repair [9,58]. The anatomy of an operation site often limits the options for modifying surgical techniques in order to minimize the risk of nerve injury. Pharmacological strategies that have been developed to reduce the incidence, duration, or intensity of postoperative pain can be divided into preemptive and preventive approaches. Preemptive analgesia aims to curb the acute postoperative pain by starting pain treatment ahead of the surgery. A variety of medications have been clinically tested for this purpose, with mixed results [4,45]. The concept of preventive analgesia is based on the hypothesis that the risk of chronic pain can be decreased by blocking pain mechanisms (such as central sensitization) that depend on nociceptive input. However, evaluating the efficacy of preventive analgesia in a clinical trial is difficult, because multiple
0304-3959/$36.00 Ó 2013 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2013.10.024
Ò
F. Yang et al. / PAIN 155 (2014) 356–366
variables, including preoperative pain, surgical technique, and anesthetic regimen, need to be controlled to ensure comparability of the treatment groups [18]. In addition, it is unknown when the transition from acute to chronic postoperative pain occurs, or whether there is a distinct transition period at all [58]. We employed a rat model of nerve injury-induced chronic pain to test whether analgesia can achieve a ‘‘disease-modifying’’ effect beyond the immediate attenuation of pain-like behavior. We treated the rats with pregabalin (PGB), a first-line recommended analgesic for neuropathic pain. PGB binds to the a2d1 subunit and, with less affinity, the a2d2 subunit of voltage-gated calcium channels (VGCC) [24]. PGB interferes with the synaptic targeting of VGCCs [5,30] and, as a result, reduces glutamate and neuropeptide release [35,66]. PGB is considered a prime candidate for the prevention of postoperative pain because of its efficacy against neuropathic pain, even though not every patient responds to it equally well [18,25]. The drug does not impede motor or somatosensory functions other than pain, an important advantage over sodium channel blockers, which also possess the capacity to inhibit activity-dependent changes in nociceptive transmission [4,64]. And it is better tolerated than antagonists of the N-methyl-d-aspartate (NMDA)-type glutamate receptor, which have been used to decrease central sensitization [22,71]. 2. Methods 2.1. Animals We used male Sprague-Dawley rats (Charles River Laboratories) aged 2–3 months for all experiments. Animal procedures adhered to the guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain and were approved by the Subcommittee on Research Animal Care of Massachusetts General Hospital and the Institutional Animal Care and Use Committee of Columbia University Medical Center. 2.2. Peripheral nerve injury Surgery for spared nerve injury (SNI) was performed on animals anesthetized using 3% isoflurane inhalation for induction and 2% isoflurane during maintenance. We ligated the tibial and the common peroneal nerves with nylon (5–0) and transected them distally, leaving intact the third branch of the sciatic nerve, the sural nerve [19]. 2.3. Drug treatment PGB was provided by Pfizer. We dissolved the drug in phosphate-buffered saline (PBS), pH 7.4, and employed subcutaneously implanted osmotic pumps (Alzet, Cupertino, CA) for continuous intrathecal delivery of 300 or 900 lg PGB/24 hours at an infusion rate of 1 or 10 lL/h, respectively. PBS served as vehicle control. The osmotic pumps were implanted either at the time of nerve injury or 7 days later. They were connected to a polyurethane catheter (Alzet), which we inserted into the lumbar subarachnoid space with the catheter tip positioned on the dorsal surface of the spinal cord at segmental level L3. We verified the integrity of the catheter by dissection following completion of the experiments. Seven animals with neurological deficits or signs of infection after the catheter implantation were excluded from the experiments. 2.4. Behavioral testing Investigators were blind to the treatment in all experiments. Behavioral assessments were performed in the morning to avoid interference of circadian differences in animal activity.
357
Somatosensory function was examined after habituating the rats to the testing environment. We obtained 2 baseline measurements in the week preceding the surgery. Following SNI, we tested neuropathic pain-like behavior at defined intervals during the infusion of PGB or vehicle and for up to 12 days after the treatment was terminated. The animals were placed on an elevated wire grid and stimulated on the plantar surface of the hind paw, in the territory of the ‘‘spared’’ sural nerve. We used calibrated von Frey monofilaments of logarithmically increasing force (range 0.0174 to 34.7 g) to determine the withdrawal threshold for punctate mechanical stimulation. The threshold was defined as the lowest force that provoked a brisk paw withdrawal at least twice in 10 applications. To test for mechanical hyperalgesia, we measured the withdrawal duration after pricking the plantar surface of the hind paw with a medium-sized safety pin. The response to cold stimulation was tested by applying a drop of acetone, which produces a cool sensation on the skin upon evaporation. The acetone drop was dispensed from a syringe without touching the skin. We recorded behavioral responses to the acetone application over 1 minute and measured the cumulative time the animal spent licking, shaking, or lifting the paw [19,55]. Motor performance was evaluated in a group of uninjured rats 7 days after the intrathecal drug treatment started. We assessed walking and spontaneous exploratory behavior in an open field. To test the righting reflex, we turned the rats on their backs and observed how they regained a normal upright position through coordinated twisting of the body. We considered the reflex intact if the rats righted themselves promptly and successfully [10]. To evaluate the hopping response, the examiner supported the trunk of the animal, fixed one hind leg in his hand and moved the body laterally. We recorded the ability of the rat to hop with the weight-bearing limb in the direction of the movement. Righting and hopping were scored as 1 if intact or 0 if compromised. The stepping reflex was evoked by drawing the dorsum of the hind paw across the edge of a table while supporting the animal’s trunk. A normal reflex consists of a flexion and upward movement of the hind leg and repositioning of the paw with toes spread on the surface of the table. We rated the performance as 0 if the repositioning failed, 1 for severe impairment, 2 for slight impairment, and 3 if the repositioning was completed without deficit [62]. 2.5. Immunohistochemistry We deeply anesthetized the rats and perfused them transcardially with PBS, followed by a phosphate-buffered solution of 4% paraformaldehyde. The L4 segment of the spinal cord was dissected, postfixed for 2 hours, cryoprotected overnight in 20% sucrose, and embedded in Tissue-Tek (Sakura Finetek, Torrance, CA). Transverse sections through the spinal cord were cut at a thickness of 10 lm on a cryostat. We blocked unspecific protein binding sites by incubating the sections for 1 hour in PBS containing 0.5% bovine serum albumin (Sigma-Aldrich, St. Louis, MO), 1% blocking reagent (Roche Applied Science, Madison, WI) and 0.1% Triton X-100 (Sigma-Aldrich). The sections were incubated overnight at 4°C with primary antibodies directed against the VGCC subunit a2d1 (1:500, produced in mouse; Sigma-Aldrich), Cd11b (1:500, mouse; AbD Serotec, Raleigh, NC) or ionized calcium-binding adaptor molecule 1 (1:500, rabbit; Wako Chemicals, Richmond, VA). The immunostaining was completed by incubating the sections for 1 or 2 hours at room temperature with species-specific secondary antibodies that were conjugated with Alexa Fluor dyes (1:500; Life Technologies, Grand Island, NY). Immunodetection of a2d1 was enhanced by warming the slides in Tris buffer (pH 8.0) for 1 hour at 70°C before the blocking of unspecific protein binding sites [67]. We used the antifading reagent ProLong Gold (Life Technologies) to mount the stained sections.
Ò
358
F. Yang et al. / PAIN 155 (2014) 356–366
To analyze the intensity of immunofluorescent staining for
a2d1, we scanned spinal cord sections (optical thickness 1.0 lm) of 5–7 animals in each experimental condition on an A1 Confocal System (Nikon), using a 10 CFI Plan Fluor objective (numerical aperture 0.30). A rectangular region of interest (150 435 lm) was marked at the boundary between the medial and central thirds of the dorsal horn to capture a2d1 expression in the projection territories of the injured tibial and common peroneal nerves [46,72]. We used ImageJ software to measure the fluorescence intensity in this region [53]. Fluorescence intensity and area under the intensity curve were normalized to the contralateral side [5]. 2.6. Western blotting Western blots were performed as previously described, with slight modifications [55]. After euthanizing the rats, we dissected the ipsilateral L4 dorsal horns and homogenized the tissue with a motorized pestle. The protein extraction buffer (pH 7.5) contained 150 mM sodium chloride, 50 mM Tris, 1 mM ethylenediaminetetraacetic acid (EDTA), 2% sodium dodecyl sulfate, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, Complete Mini (without EDTA) protease inhibitors (Roche Applied Science) and Phosphatase Inhibitor Cocktails 1 and 2 (Sigma-Aldrich). We employed NuPAGE Novex 4–12% Bis-Tris gels for the electrophoresis (Life Technologies) and transferred the protein to an Immobilon-FL polyvinylidene fluoride (PVDF) membrane (EMD Millipore, Darmstadt, Germany). We blocked unspecific protein binding sites with Odyssey buffer (Li-Cor, Lincoln, NE) for 1 hour at room temperature and incubated the PVDF membranes overnight at 4°C with primary antibodies for a2d1 (1:1000, mouse; Sigma-Aldrich), ionized calcium-binding adaptor molecule 1 (1:500, rabbit; Wako Chemicals) or phosphorylated p38 mitogen-activated protein kinase (1:300, rabbit; Cell Signaling Technology, Beverly, MA), diluted in blocking buffer containing 0.1% Tween 20. Protein–antibody complexes were detected by incubating the membranes for 1 hour at room temperature with species-specific IRDye-conjugated secondary antibodies (1:20,000; Li-Cor) in a blocking buffer that contained 0.1% Tween 20 and 0.02% sodium dodecyl sulfate. The intensity of fluorescent bands was measured on an Odyssey Infrared Imaging System (Li-Cor). We immunostained for tubulin b3 (1:10,000; Sigma-Aldrich) to determine sample loading and normalize the fluorescent signal intensity of immunoreactive bands. 2.7. Statistics We used Prism 5 (GraphPad Software, La Jolla, CA) for statistical analysis. To compare somatosensory responses between treatment groups, we employed a two-way analysis of variance (ANOVA) followed by Sidak’s test for multiple comparisons. Righting and hopping scores were analyzed with the v2 test, and scores for paw placement with one-way ANOVA. An unpaired student’s t test was used to assess differences in the immunofluorescent a2d1 staining of tissue sections or the Western blot results. Data are presented as means ± SEM. Differences were considered significant if P < 0.05. 3. Results 3.1. Short-term treatment with PGB Nerve injury led to a decrease in the withdrawal threshold for punctate mechanical stimulation with von Frey filaments, an exaggerated response to pinprick, and a pain-like response to the normally nonpainful cold sensation produced by acetone (Fig. 1). Intrathecal infusion of PGB, starting on the day of the surgery, effectively reduced both mechanically and thermally evoked responses.
One week after SNI, the withdrawal threshold for stimulation with von Frey filaments was 6.6 ± 2.0 g in rats receiving 300 lg PGB/24 hours compared with 0.4 ± 0.1 g in vehicle-treated rats. The withdrawal duration after pinprick was 2.9 ± 1.0 seconds compared with 12.1 ± 2.1 seconds, and the withdrawal response after the application of acetone lasted 8.3 ± 2.6 seconds, substantially shorter than the average withdrawal duration of 12.6 ± 2.1 seconds in vehicle-treated rats (Fig. 1). The infusion of 900 lg PGB/24 hours produced a similar attenuation of pain-like behavior: 1 week after SNI, the withdrawal threshold after punctate mechanical stimulation was 7.3 ± 2.3 g, the withdrawal duration after pinprick was 3.6 ± 0.9 seconds, and acetone-evoked responses lasted 5.8 ± 2.0 seconds (Fig. 1). The difference between PGB and vehicle treatment was significant for all tested manifestations of neuropathic pain-like behavior (P < 0.001 for the mechanical stimulations, P < 0.05 for the acetone test, determined in a two-way ANOVA covering the treatment period of 7 days). Pain reduction did not differ between the low and high doses of intrathecal PGB. The analgesic effect of PGB was only partially sustained beyond the week-long infusion. Withdrawal responses evoked by von Frey filaments or acetone remained attenuated for 6 days in rats previously treated with 900 lg PGB/24 hours (P < 0.05), but this difference was very small compared with the reduction of pain-like behavior during the infusion (Fig. 1). 3.2. Long-term analgesia To test whether longer treatment would have a greater influence on the course of neuropathic pain, we administered PGB for 28 days. We used only the lower dose of 300 lg PGB/24 hours in these experiments, because it produced the same acute analgesic effect as 900 lg PGB/24 hours and did not interfere with movement coordination, whereas rats receiving 900 lg PGB/24 hours scored lower in the righting and hopping tasks (P < 0.05; Fig. 2). Stepping, spontaneous walking, and exploratory behavior were unaffected by either dose of PGB. We conducted two experiments, one in which we started longterm treatment at the time of nerve injury, and one in which the treatment was delayed by 7 days. The first experiment was designed to preempt the nerve injury-induced increase in pain sensitivity. With the second approach, we modeled a situation in which analgesic therapy begins after neuropathic pain has developed. Vehicle-treated rats exhibited neuropathic pain-like behavior (low withdrawal thresholds to punctate mechanical stimuli, increased withdrawal responses to pinprick or acetone) for 47 days after SNI, the maximum length of our experiments. Mechanically induced and cold-induced pain-like behavior was reduced by PGB in both treatment models. After 4 weeks of PGB infusion with early onset, the withdrawal threshold to punctate mechanical stimulation was 5.3 ± 2.0 g compared with 0.9 ± 0.7 g in vehicle controls (P < 0.001 for the treatment effect over 28 days, determined in a two-way ANOVA; Fig. 3). The withdrawal duration after pinprick was 2.1 ± 1.5 seconds compared with 5.5 ± 1.5 seconds (P < 0.01), and the response to cold stimulation lasted for 1.0 ± 0.4 seconds compared with 8.4 ± 2.7 seconds (P < 0.001; Fig. 3). The analgesic effect of PGB was noticeably weaker when the intrathecal infusion started with a delay. Rats in which the PGB infusion began 7 days after SNI, when neuropathic pain-like behavior was already established, showed a withdrawal threshold of 1.0 ± 0.5 g to punctate mechanical stimulation after 28 days of treatment. This threshold was close to that of 0.8 ± 0.2 g in vehicle-treated rats. However, the difference between rats receiving PGB and those receiving a vehicle infusion was still clear and significant when compared over the entire treatment period of
Ò
359
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10
**
*
3
Vehicle ipsi PGB 300 µg/24 h ipsi PGB 900 µg/24 h ipsi
1
Vehicle contra PGB 300 µg/24 h contra PGB 900 µg/24 h contra
0.3 Treatment
0.1 B1 B2
5
SNI
10
15
Days after surgery
Pinprick Withdrawal duration (s)
14 12 10 8 6
***
4 2
***
0 B1 B2
5
10
15
Acetone
Withdrawal duration (s)
18 16 14 12 10 8
#
6 4 2 0 B1 B2
5
10
15
Fig. 1. Short-term pregabalin (PGB) treatment. Continuous intrathecal infusion of PGB (300 or 900 lg/24 hours) or vehicle started at the time of spared nerve injury (SNI) and lasted 7 days (shaded area). Pain-like behavior was tested for 15 days after surgery. We used calibrated von Frey filaments to determine the withdrawal threshold for punctate mechanical stimulation. The withdrawal duration after pinprick was recorded to assess the reaction to mechanical stimulation above the pain threshold. The response to cold was tested with evaporating acetone and measured as the time the animal spent licking, shaking, or lifting the paw. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle treatment in all 3 tests during treatment (P < 0.001 for the stimulation with von Frey filaments or pinprick, P < 0.05 for the response to acetone). For 6 days after termination of the treatment, rats that had received an infusion of 900 lg PGB/24 hours remained less sensitive to the stimulation with von Frey filaments or acetone when compared with vehicle-treated rats (P < 0.05). Asterisks and hashes indicate the results of Sidak’s follow-up tests. ⁄ P < 0.05, ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001 for the comparison of PGB with vehicle treatment; #P < 0.05 for the comparison of 300 lg with 900 lg PGB/24 hours. N = 8–9. B1 and B2, baseline assessments.
28 days (P < 0.01; Fig. 4). An initial increase in the withdrawal threshold of vehicle-treated rats, from 0.6 ± 0.2 g on day 7 to 1.9 ± 0.7 g on day 14 after SNI, may have been a prolonged effect of the inhalation anesthesia that we administered during the second surgery for catheter placement and osmotic pump implantation. Withdrawal responses after pinprick lasted 0.5 ± 0.0 seconds at the end of the 4-week treatment with PGB compared with 2.0 ± 0.7 seconds in vehicle-treated rats (P < 0.01). Responses to cold stimulation were recorded over 3.7 ± 1.3 seconds compared with 4.6 ± 1.4 seconds (P < 0.05; Fig. 4). The rats tolerated the continuous infusion of 300 lg PGB/24 hours well; we did not observe adverse effects. Long-term PGB treatment did not have a sustained analgesic effect beyond the infusion, irrespective of the timing of treatment
onset. Neuropathic pain-like behavior after 28 days of PGB treatment was indistinguishable from the responses to test stimuli in rats that received a vehicle infusion (Figs. 3 and 4). 3.3. VGCC subunit a2d1 in the dorsal horn Following nerve injury, somatosensory neurons upregulate the expression of VGCC subunit a2d1. Trafficking of the subunit to the central terminals of primary afferents, rather than local induction, is responsible for a subsequent increase in a2d1 in the superficial dorsal horn of the spinal cord [5,38,41]. To determine the effect of PGB on the presynaptic enrichment of a2d1, we immunostained transverse cryosections through the L4 segment of the spinal cord for the VGCC subunit. We measured the intensity of
Ò
360
F. Yang et al. / PAIN 155 (2014) 356–366
Righting
Hopping
Stepping 3
*
60
*
40
Score
80
2 1
20 0
4
4
h
h
le hi c
µg /2
µg /2
0 90 B
PG
PG
PG
B
B
90
30
0
0
µg /2
µg /2
4
4
h
h
le hi c 30 B PG
PG
B
90
30
0
0
Ve
4 µg /2
µg /2 0
Ve B PG
h
h 4
hi c
le
0 Ve
Successful completion (%)
100
Fig. 2. Coordinated movement tasks. Uninjured rats were tested following intrathecal infusion of 300 or 900 lg PGB/24 hours or vehicle for 7 days. The 2 bar graphs on the left show the percentage of rats successfully completing the righting or hopping tasks. Successful stepping after a passive extension of the ankle joint was scored as 3, whereas a score of 0 reflects task failure. ⁄P < 0.05, determined in a v2 test. N = 10–12. PGB, pregabalin.
the fluorescent staining in the medial two-thirds of the dorsal horn, which receive most of their afferent input from the tibial and common peroneal nerves [46,72], the two branches of the sciatic nerve that are ligated and cut during SNI (Fig. 5A). Seven days after SNI, the integrated fluorescence intensity in the ipsilateral dorsal horn of untreated rats was 177.6% ± 15.8% (relative to the contralateral dorsal horn), compared with 104.8% ± 10.5% in sham-operated animals (P < 0.01). The intensity was highest, 145.0% ± 8.5% (relative to the maximum fluorescence intensity in the contralateral dorsal horn), in lamina II, at a distance of 100 lm from the outer contour of the dorsal horn (Fig. 5B). In lamina III, at 200 lm from the contour, we measured an intensity of 73.7% ± 11.0%. The corresponding values in sham-operated animals were 87.8% ± 4.9% (P < 0.001) and 47.0% ± 4.8% (not significant) (Fig. 5B). The nerve injury-induced change in a2d1 immunostaining was limited to the ipsilateral dorsal horn; the fluorescence intensity in the contralateral dorsal horn of rats after SNI and rats after sham surgery was not different. Intrathecal treatment with PGB did not reduce the increase in a2d1 immunoreactivity (Fig. 5C). Seven days after SNI, the integrated fluorescence intensity in the ipsilateral dorsal horn of rats receiving a continuous infusion of 300 lg PGB/24 hours was 182.3% ± 21.4% (relative to the contralateral dorsal horn) compared with 178.8% ± 7.6% in vehicle-treated animals. The signal intensities in laminas II and III were 156.1% ± 18.7% and 76.1% ± 9.5% (relative to the maximum intensity in the contralateral dorsal horn) compared with 167.0% ± 13.1% and 84.2% ± 7.9%, respectively, in vehicle-treated animals (Fig. 5D). None of these differences were significant, nor did the fluorescence intensity in the contralateral dorsal horn differ between rats treated with PGB and rats receiving a vehicle infusion. Western blot analysis confirmed an increase in a2d1 protein to 137.1% ± 6.6% in the ipsilateral dorsal horn 7 days after SNI compared with 100.0% ± 14.38% after sham surgery (P < 0.05). Continuous infusion of 300 lg PGB/24 hours did not reduce a2d1 protein in the dorsal horn (Fig. 6). 3.4. Microglial activation Primary afferent input drives important mechanisms of neuropathic pain, for example, central sensitization in the dorsal horn or the activation of spinobulbospinal circuits involved in descending facilitation [36,50,65]. Increases in nociceptive and non-nociceptive input also contribute to the activation of spinal microglia [28,55,62,70], a process that is crucial for the development of pain after nerve injury [57]. We tested whether analgesia provided by a
continuous infusion of 300 lg PGB/24 hours, beginning at the time of SNI, is sufficient to suppress microglial activation in the dorsal horn of the spinal cord. Microglia in the ipsilateral dorsal horn reacted strongly to SNI. Seven days after nerve injury, microglial cells showed markedly enhanced immunostaining for Cd11b (also known as integrin aM) and ionized calcium-binding adapter molecule 1 (Iba1, also known as allograft inflammatory factor 1). Cd11b or Iba1 immunostaining in rats treated with PGB did not differ from that in rats receiving a vehicle infusion (Fig. 7A). Western blot analysis showed a 5-fold increase in Iba1 protein at 7 days after SNI, to 496.3% ± 85.37% compared with 100.0% ± 13.27% after sham surgery (P < 0.001). PGB treatment did not lower Iba1 expression in the ipsilateral dorsal horn (Fig. 7B). We also examined the phosphorylation of p38 mitogen-activated protein kinase (MAPK), an enzyme prominently involved in microglial signaling pathways that regulate the synthesis of inflammatory mediators such as nitric oxide synthase 2 or interleukin 1b. The level of phosphorylated (active) p38 MAPK in rats receiving PGB was 87.4% ± 8.2% compared with 100.0% ± 6.4% in vehicle-treated rats (not significant; Fig. 7B), indicating that the inflammatory response to SNI in the dorsal horn was essentially unchanged. 4. Discussion By reducing afferent input or inhibiting signal transmission at nociceptive synapses, analgesics may prevent activity-dependent pain mechanisms, such as central sensitization, and have a sustained effect beyond the immediate attenuation of pain. At least this is the assumption behind the concept of preventive analgesia [18]. The main findings of our study challenge this concept. PGB, a first-line recommended medication for nerve injury-induced pain, produced analgesia only as long as the drug was administrated, without blocking the emergence of persistent pain. Somewhat unexpectedly, we also found that PGB achieved its analgesic effect without lowering the concentration of its binding target, the VGCC subunit a2d1, at the central terminals of primary afferents. Short-term experiments consisted of intrathecal PGB infusion for 1 week to provide analgesia during the development of painlike behavior after SNI [19]. Long-term treatment for 4 weeks was designed to cover the period during which central sensitization establishes, spinal inhibition decreases, descending pain modulation shifts toward facilitation, and microglia in the dorsal horn is activated [11,17,36,55,56]. As an inhibitor of neurotransmitter release, PGB was expected to interfere with these mechanisms of
Ò
361
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10
**
*
3 1
Vehicle ipsi PGB 300 µg/24 h ipsi
0.3 0.1
Vehicle contra PGB 300 µg/24 h contra
Treatment 0.03 B1B2
5
10
15
20
25
30
35
40
Days after surgery
SNI
Pinprick
Withdrawal duration (s)
14 12 10 8 6 4
*
2 0 B1B2
5
10
15
20
25
30
35
40
20
25
30
35
40
Withdrawal duration (s)
Acetone 20 18 16 14 12 10 8 6 4 2 0
* B1B2
5
10
15
Fig. 3. Long-term pregabalin (PGB) treatment with early onset. Intrathecal infusion of 300 lg PGB/24 hours or vehicle started at the time of spared nerve injury (SNI) and continued for 28 days (shaded area). Pain-like withdrawal responses to punctate mechanical stimulation (von Frey filaments), normally painful mechanical stimulation (pinprick), or cold stimulation (acetone) were tested for 40 days after surgery. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle during treatment in all 3 tests (P < 0.001 for the response to stimulation with von Frey filaments or acetone, P < 0.01 for the response to pinprick). However, PGB did not have an effect on pain-like behavior beyond termination of the infusion. ⁄P < 0.05, ⁄⁄P < 0.01 for PGB compared with vehicle, determined in Sidak’s follow-up test. N = 5–7. B1 and B2, baseline assessments.
neuropathic pain because their induction involves the enhanced activity of somatosensory afferents [68]. PGB effectively reduced pain after SNI. The decrease in pain-like behavior was smaller, though still significant, when the PGB infusion was delayed by 1 week, suggesting that the drug should be given as early as possible [35]. However, PGB failed to modify the chronic course of neuropathic pain, even when it was administered continuously for 4 weeks. Once the infusion ended, the rats exhibited sharply reduced withdrawal thresholds for mechanical stimulation and an increased sensitivity to pain evoked by pinprick or acetone evaporation. These behavioral changes, which are equivalent to the common features of chronic postoperative pain in humans— mechanical allodynia, mechanical hyperalgesia, and cold allodynia [1,69]—were indistinguishable from those in rats that received a vehicle infusion. Consistent with clinical experience in the longterm use of a2d1 ligands [52], the analgesic effect of PGB was
stable over time, indicating that tolerance was not responsible for the lack of pain prevention. Mainly sodium channel blockers have been evaluated for preventive analgesia after nerve injury in previous experimental studies. Lidocaine, bupivacaine, or mepivacaine were directly applied to the nerve or a dorsal root ganglion (DRG), or were administered in a single systemic or intrathecal injection. The compounds rarely attenuated neuropathic pain-like behavior for >1 week [64]. Longer or even complete prevention of pain hypersensitivity was only observed after ‘‘bathing’’ the sciatic nerve in 2% lidocaine or administering 10% lidocaine directly onto spinal nerve roots [8,60]. Such concentrated applications risk nerve damage. Slow-release formulations of sodium channel blockers provide a safer approach to obtaining sustained analgesia. Repeated injections of saxitoxin liposomes blocked sciatic nerve conduction for 18 days after SNI and delayed mechanical hypersensitivity in rats for 4 weeks [59].
Ò
362
F. Yang et al. / PAIN 155 (2014) 356–366
Von Frey filaments Withdrawal threshold (g)
30 10 3 Vehicle ipsi PGB 300 µg/24 h ipsi
1 0.3
Vehicle contra PGB 300 µg/24 h contra
Treatment
0.1 B1 B2
5
10
15
20
25
30
35
40
45
50
30
35
40
45
50
30
35
40
45
50
Days after surgery
SNI
Pinprick Withdrawal duration (s)
4 3 2 1 0 B1 B2
5
10
15
20
25
Acetone Withdrawal duration (s)
14 12 10 8 6 4 2 0 B1 B2
5
10
15
20
25
Fig. 4. Long-term pregabalin (PGB) treatment with late onset. Intrathecal infusion of 300 lg PGB/24 hours or vehicle started 7 days after spared nerve injury (SNI) and continued for 28 days (shaded area). Pain-like withdrawal responses to punctate mechanical stimulation (von Frey filaments), normally painful mechanical stimulation (pinprick), and cold stimulation (acetone) were tested for 47 days after surgery. A two-way analysis of variance (ANOVA) showed significant differences between PGB and vehicle during treatment in all 3 tests (P < 0.01 for the stimulation with von Frey filaments or pinprick, P < 0.05 for the response to acetone). Differences after termination of the treatment were not significant. N = 5–8. B1 and B2, baseline assessments.
However, the rats exhibited heat hyperalgesia soon after the block, suggesting differential prevention of mechanically evoked pain. Timing and efficacy of the conduction block are crucial. Early and complete suppression of afferent activity with a faster-acting bupivacaine depot or continuous infusion of tetrodotoxin directly at the nerve lesion site shortened the duration of the blockade necessary to accomplish sustained analgesia [14,73]. Gradual release of sodium channel blockers from liposomes or microspheres may inhibit afferent activity partially or late and therefore require repeated administration to achieve pain prevention [59,63]. Notably, even repeated injections of saxitoxin liposomes did not alter the upregulation of a2d1 in somatosensory neurons [59]. Only nonselective sodium channel inhibitors are available for clinical use. When applied to mixed sensory and motor nerves, paralysis for the duration of the block is an inevitable risk [4,64]. Prolonged application of sodium channel blockers is therefore not feasible in humans, which explains the interest in analgesics
that might prevent pain without compromising other nerve functions. A neurokinin 1 receptor antagonist [12], high-dose remifentanil [42] and topical clonidine applied at the nerve lesion site [37] partially protect against pain hypersensitivity in the rat. The translational potential of some of these treatment approaches is uncertain, considering the clinical failure of neurokinin 1 receptor antagonists as analgesics. By contrast, the a2d1 ligands PGB and gabapentin are established analgesics for neuropathic pain [3] and produce preemptive analgesia after surgery [44,76]. However, clinical trials on pain prevention have yielded inconclusive results [16,34]. Differences between surgical procedures, anesthetic regimens, outcome measures, and lack of placebo control hamper the interpretation of clinical evidence [18]. Animal models allow studying the pathophysiology and treatment of postoperative pain under standardized conditions. We used this advantage to compare short-term and long-term infusions of PGB in a rat model of postoperative pain caused by nerve
Ò
363
F. Yang et al. / PAIN 155 (2014) 356–366
B
25 0
co
ip
140
m
Sh
am
###
160
Sh a
180
tra
0
50
i
Sham
75
ps
20
100
on
40
125
Ic
Ipsi Contra
60
150
Ii
80
175
SN
120 100
***
200
SN
140
si
Sham
160
225
nt ra
II III
## 180
Area under the curve (%)
Ipsilateral
Fluorescence intensity (%)
A
120 100 80 60 40 20
SNI
0 0
50 25
e SN
ve
hi
cl
hi ve SN
SN
100
I+
I+
140 120
nt co
B
co
nt
ip cl
e
180 160
ra
0
si
0
75
ip
SNI + vehicle
100
B
20
125
PG
40
150
I+
Ipsi Contra
60
175
ra
80
**
PG
100
200
I+
140 120
***
225
SN
180 160
si
SNI + vehicle
Dorsoventral distance (μm)
D Fluorescence intensity (%)
C
50 100 150 200 250 300 350 400 450
Area under the curve (%)
SNI
80 60 40
SNI + PGB
20
SNI + PGB
0 0
50 100 150 200 250 300 350 400 450
Dorsoventral distance (μm) Fig. 5. Immunostaining for the voltage-gated calcium channel subunit a2d1 in the dorsal horn. (A) Representative transverse cryosections, immunostained for a2d1, through the dorsal horns of untreated rats 7 days after sham surgery or spared nerve injury (SNI). Rectangles indicate the regions of interest (150 435 lm) in which we measured the fluorescence intensity. We placed the regions over the boundary between the medial and central thirds of the dorsal horn, where the axons of somatosensory fibers from the injured tibial and common peroneal nerves terminate [46,72]. (B) Fluorescence intensity profiles of the dorsal horn, normalized to the maximum value (100%) of the contralateral side. We calculated the mean intensity for each condition at distances of 100 and 200 lm from the outer contour of the dorsal horns to obtain representative values for a2d1 in laminas II and III, respectively. The areas under the intensity profiles are compared in the bar graph. (C) Representative dorsal horn sections of rats that received an intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI. Treatment began at the day of nerve injury. (D) Normalized fluorescence intensity profiles of the dorsal horn after SNI and vehicle or 300 lg PGB/24 hours treatment, respectively. Fluorescence intensities of the contralateral dorsal horns did not differ between conditions. Scale bars, 200 lm. N = 5. ⁄⁄P < 0.01, ⁄⁄⁄P < 0.001 for the comparison of ipsilateral with contralateral dorsal horns; ##P < 0.01, ###P < 0.001 for the comparison of SNI with sham surgery, determined in unpaired t tests. PGB, pregabalin.
injury. We found that even PGB treatment for 4 weeks did not prevent the resurgence of high-intensity pain-like behavior once the infusion ended. Sustained analgesia for up to 7 days following intrathecal injection or infusion of gabapentin has been reported in rats and mice after spinal nerve ligation, but it is unclear whether true long-term prevention was achieved [15,47,48]. Pain hypersensitivity spontaneously decreased following the ligation and gabapentin may simply have accelerated this recovery [47,48]. In a recent clinical trial, the intrathecal infusion of gabapentin failed to reduce chronic pain [54]. However, few of the subjects who received gabapentin at a dose of 30 mg/24 hours, which
would be equivalent to the intrathecal infusion of PGB in our study, had neuropathic pain. Mixed treatment response in a heterogeneous trial population and gabapentin’s pharmacokinetic properties, rather than the intrathecal route of administration, may therefore explain the lack of a significant analgesic effect. PGB remains in the central nervous system longer than gabapentin and is the more potent analgesic [7]. Although afferent activity drives important neuropathic pain mechanisms, our findings indicate that some mechanisms, including the reaction of spinal glia to peripheral nerve lesion, operate at least partially independently. Continuous analgesia by PGB did not
Ò
F. Yang et al. / PAIN 155 (2014) 356–366
PG
cl
I+
hi ve I+ SN
v SN ehic I + le PG B
B
Tubulin β 3 e
am SN I
Sh
SN
Sh
I
Tubulin β 3
α2δ1
I+
α2δ1
140 120 100 80 60 40 20 0
SN
α2δ1 protein (%)
*
SN
160 140 120 100 80 60 40 20 0
am
α2δ1 protein (%)
364
Fig. 6. Western blot analysis of voltage-gated calcium channel subunit a2d1 in the dorsal horn. Quantification of a2d1 protein levels in the ipsilateral dorsal horn of untreated rats 7 days after sham surgery or spared nerve injury (SNI), and in rats that received an intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI, beginning at the time of nerve injury. The integrated fluorescent signal of labeled bands is expressed as the percentage relative to sham surgery for the comparison with untreated rats after SNI, or relative to vehicle treatment for the comparison with PGB. N = 5. ⁄P < 0.05, determined in an unpaired t test. PGB, pregabalin.
SNI + vehicle
SNI + PGB
CD11b
Iba1 protein (%)
A
B
***
600 500 400 300 200
Iba1
100
Tubulin β 3
100 80 60 40 20
80
40
Phospho p38
20
Tubulin β 3 hi
SN
I+
ve
h SN icle I+ PG B
0
ve I+
Tubulin β 3
60
0
SN
Iba1
100
cl SN e I+ PG SN B I+ ve SN hicl I+ e PG B
Iba1 protein (%)
I 120
Phosphorylated p38 Mapk protein (%)
120
Iba1
SN
am Sh
I SN
Sh
am
0
Fig. 7. Microglial activation after spared nerve injury (SNI). (A) Immunostaining for Cd11b or ionized calcium-binding adapter molecule 1 (Iba1). Representative images show the ipsilateral dorsal horn of rats treated with intrathecal infusion of vehicle or 300 lg PGB/24 hours for 7 days after SNI, beginning at the time of nerve injury. (B) Quantification of Iba1 and phosphorylated p38 mitogen-activated protein kinase (MAPK) by Western blotting. Bar graphs show the integrated fluorescent signal of labeled bands expressed as the percentage relative to sham surgery for the comparison with untreated rats after SNI, or relative to vehicle treatment for the comparison with PGB. N = 4–6. PGB, pregabalin.
stop microglial activation after SNI. Although nociceptive and nonnociceptive input modulates glial cell proliferation and signaling pathways [28,62,70], nerve conduction blocks or interventions that target primarily nociceptive signaling may be insufficient to fully suspend glial responses. Inhibiting action potentials in injured peripheral axons attenuates the mobilization of spinal glia [59,74], but resident microglia, invading monocytes, and astroglia react to multiple chemokines such as Ccl2 or adenosine triphosphate, which are released not only from primary somatosensory neurons [57]. Complete interruption of afferent input by dorsal root transection is required to abolish the typical pattern of microglial activation in the dorsal horn [40,55]. On the other hand, our results demonstrate that analgesia can be achieved independently of microglial activity. The binding target for PGB, the VGCC subunit a2d1, is expressed in DRG neurons, most prominently in nociceptive neurons. Nerve injury leads to marked upregulation and synaptic trafficking of
a2d1 [5,38,41,49], which acts as a chaperone for the pore-forming VGCC subunit a1 and stabilizes its surface expression [13,31]. High VGCC density at axon terminals improves the coupling of calcium entry to transmitter exocytosis and is an important mediator of activity-dependent pain mechanisms [30,31]. Knocking a2d1 down after spinal nerve ligation decreases pain-like behavior [38], whereas a2d1 overexpression in uninjured mice results in abnormal pain sensitivity similar to neuropathic allodynia and hyperalgesia [39]. Continuous intrathecal treatment with PGB had no impact on a2d1 accumulation in the superficial dorsal horn. We did not evaluate a2d1 expression in rats receiving 900 lg PGB/ 24 hours, but the results after the infusion of 300 lg PGB/24 hours demonstrate clearly that the drug produced analgesia without disrupting a2d1 trafficking. Our findings are consistent with previous reports showing that intrathecal administration of gabapentin or PGB reduces neuropathic pain-like behavior despite incomplete (
