| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Chronic pain after thoracotomy is common, although its basis and therapy have not been well characterized. In this study we characterize the allodynic responses (mechanical and cold) as well as the histopathologic changes after thoracotomy and rib retraction in rats. The antinociceptive effect of systemic and intrathecal analgesics was also evaluated. Male Sprague-Dawley rats were anesthetized and the right 4th and 5th ribs surgically exposed. The pleura was opened between the ribs and a retractor placed under both ribs and opened 8 mm. Retraction was maintained for 5, 30, or 60 min. Control animals had pleural incision only. Beginning Day 2 postsurgery, animals were tested for mechanical allodynia using calibrated von Frey filaments and cold allodynia using acetone applied to the incision site. Two weeks after surgery, animals were tested for reduction of allodynia with intraperitoneal and intrathecal injections of analgesics. Intercostal nerve histology was examined at 14 days postsurgery. Allodynia developed in 50% of the animals with 60 min retraction but in only 11% and 10% of animals when the retraction time was 5 and 30 min, respectively, and in none of the control animals. Allodynic animals showed extensive axon loss in the intercostal nerves of the retracted ribs. Allodynia appeared by Day 10 in the rib-retraction model and lasted at least 40 days. Systemic morphine sulfate (50% effective dose [ED50], 1.06 mg/kg) and gabapentin (ED50, 24.2 mg/kg), as well as intrathecal morphine (ED50, 1.19 nmol), gabapentin (ED50, 13.8 nmol), clonidine (ED50, 72.7 nmol), and neostigmine (ED50, 0.54 nmol) reduced allodynia. Rib-retraction in rats for 60 min produces allodynia that lasts more than 1 mo, and this allodynia is reduced by morphine, gabapentin, clonidine, and neostigmine. This new model may be useful for quantifying the efficacy of techniques to reduce the frequency and severity of long-term postthoracotomy pain. IMPLICATIONS: A new model of persistent postthoracotomy pain has been developed using thoracotomy and rib-retraction for 60 min in rats. Pathologic changes in nerves can be demonstrated, persistent mechanical and cold allodynia evolve, and responses to systemic and intrathecal analgesic drugs can be quantified.
Chronic postthoracotomy pain (CPTP) is defined as "pain that recurs or persists along a thoracotomy incision at least two months following the surgical procedure" (1). Typically, the pain is described as a continuous dysesthesia with burning and aching in the general area of the thoracotomy incision. The incidence of persistent pain 1 yr after thoracotomy is approximately 50%, which makes this phenomenon among the most frequent complications of thoracotomy (2,3). Although muscle-sparing surgical approaches have been investigated, the frequency of chronic pain assessed 1 yr after operation was similar between muscle-sparing and standard lateral thoracotomy procedures (4). Many patients do not seek help for CPTP and even when identified, drug treatment is often inadequate (5). Because the site of the pain is usually along the distribution of the periincisional intercostal nerves, they have been implicated in this neuropathic pain syndrome (6,7). A clinical study has shown that rib retraction alone consistently caused total conduction block in the intercostal nerves on both sides of the retractor (8). Intercostal nerves that were one rib away from the retractor in each direction exhibited approximately a 50% conduction block. The exact etiology of CPTP has not been determined. A chronic pain syndrome in the thoracic skin region can be produced in rats by inducing a chronic constriction injury (CCI) to the intercostal nerves with chromic gut sutures (9). This experimental approach is analogous to the neuropathic pain model originally developed by Bennett and Xie (10) for inducing neuropathy of the sciatic nerve. However, it is unknown whether this method of inducing chronic thoracic pain is applicable to the study of CPTP. Hence, the development of an animal model with similar surgical trauma is important to further understand the mechanism of CPTP and guide development of better treatment modalities. We hypothesized that rib-retraction after thoracotomy in rats can produce long-term neuropathic pain and serve as a model of CPTP. Our study was designed to determine the time course of the development of mechanical and cold allodynia and to correlate allodynia with histological findings in the intercostal nerves of the retracted ribs. The effect of analgesic compounds, such as morphine, gabapentin, clonidine, and neostigmine, on reducing the allodynia elicited in this model was also evaluated.
After Institutional Animal Care and Use Committee approval, male Sprague-Dawley rats (300 g; Sasco SD, Charles River, Wilmington, MA) were briefly anesthetized with isoflurane and an endotracheal catheter (16-gauge, 51-mm long Teflon IV catheter; Terumo Medical, Somerset, NJ) was placed using tongue retraction. The catheter was connected to a Y-connector attached to tubing from a small animal ventilator (model 683; Harvard Instruments, Holliston, MA). An isoflurane vaporizer (SurgiVet) was connected to the intake of the ventilator to deliver a tidal volume of 2 mL at a rate of 80/min with the output concentration maintained between 1.0%1.5% isoflurane in oxygen. A 3-cm incision was made in the skin of the lateral chest wall between the right 4th and 5th ribs. The deep and superficial muscles covering the ribs were retracted to expose the intercostal muscle. Using iris scissors, a 1.5-cm incision was made in the intercostal muscle and pleura above the 5th rib, taking care to keep the blades of the scissors as superficial as possible to avoid damaging the lung. The blunt tines of a small self-retaining retractor (model SU-3146; Mueller, McGaw Park, IL) were coated with lubricant (Surgilube) and carefully placed under the 4th and 5th ribs. The retractor was opened to its 3rd position producing a rib separation of 8 mm. The retractor was left in place for 5 (n = 18), 30 (n = 10), or 60 (n = 10) min. One group of control animals (n = 12) had a pleura incision only, which was left open to air. During this time, the wound opening was covered by gauze soaked in normal saline. After the retraction period, the retractor was returned to the closed position and removed. A 4-cm long piece of 0.047 in. diameter silicone rubber tubing was placed 1 cm inside the pleural space for later removal of air. The deep muscles covering the ribs were sutured closed with 20 silk, with a secure muscle apposition around the tubing. Air was aspirated from the pleural cavity with a 5-mL syringe attached to the tubing to restore normal intrapleural pressure. The tubing was then removed from the pleural space and the exit point immediately sealed with veterinary adhesive (VetBond; 3M, St. Paul, MN). The Y-connector was disconnected from the endotracheal tube, with verification that the animal had spontaneous ventilation. The superficial muscles covering the ribs were then apposed with 20 silk sutures, the skin was closed with 40 nylon sutures, and the endotracheal catheter was removed after the animal was spontaneously moving.
Starting on the first day postsurgery all animals were tested daily for 40 days for mechanical allodynia using calibrated von Frey filaments applied to the dorsal skin (T45 dermatome) around the incision site (Fig. 1a) (9). The skin over the right dorsal upper back was lightly shaved on the afternoon before each test. Each animal was placed in a plastic cage and allowed to accommodate for 15 min. Subsequently, the skin just above the incision was probed with von Frey filaments (range, 0.415.1 g) using the up-down method to calculate the threshold force (11). Scratching of the dorsal right upper back skin by the hindpaw within 6 s of the application of a filament was interpreted as a positive aversive response in the up-down method. Rats not responding to even the highest force filament were assigned a value of 15.1 g and rats responding to even the lowest filament were assigned a value of 0.3 g. Animals were classified as having mechanical allodynia if the withdrawal threshold was
On days 1428 after thoracotomy, animals demonstrating mechanical allodynia were tested at 3- to 4-day intervals with randomly assigned doses of either morphine or gabapentin injected intraperitoneally in a volume of 0.5 mL. Baseline mechanical and cold testing was performed between 89 AM and the animals then injected with drug while being lightly restrained in a cloth towel. Animals were re-tested at 30 min (morphine) or at 60 min (gabapentin) for changes in mechanical and cold allodynia. A group of allodynic animals (n = 14, evaluated at 2 wk postsurgery) were implanted, under 1.5% isoflurane anesthesia, with intrathecal catheters for bolus drug injection. The catheter (polyethylene, 0.6-mm outer diameter; B-D) was inserted through the cisterna magna and advanced to the upper thoracic spinal cord level (12). The catheter tip was positioned slightly caudal to the T45 spinal region innervating the area of rib retraction, consistent with intrathecal catheter placement for such a condition in humans. The catheter was secured to the dorsal neck musculature, and all skin margins were closed, leaving 3 cm of external catheter (occluded by an internal stylet) above the skull for bolus drug injections (1214). Any animal that exhibited neurological impairment after catheter placement was killed. Seven days after catheter implant, animals were tested for mechanical and cold allodynia and then injected with 8 µL of drug followed by 8 µL of normal saline to flush the catheter dead space. Animals were re-tested at 30 min (or at 60 min for gabapentin) for changes in mechanical and cold allodynia. Animals were re-injected at 3-to 4-day intervals with a different drug in a random fashion. The drugs used in the study were morphine sulfate (Eli Lilly, Indianapolis, IN), gabapentin, clonidine hydrochloride, neostigmine methylsulfate, MK-801 and CNQX (all Sigma Chemical, St. Louis, MO). All drugs were diluted in preservative-free 0.9% sodium chloride for injection. The doses of morphine, gabapentin, clonidine, and neostigmine used in the present study, both systemically and intrathecally, had been used previously in another rat study and had been shown not to produce sedation in rats, as evidenced by the animal remaining on a rotating rod for 180 s at 10 rpm (15). The maximum intrathecal MK-801 dose (30 nmol) is below the level that produces motor dysfunction in rats (16). CNQX (30 nmol) also does not produce motor dysfunction (17). At 14 days postsurgery, 6 animals (n = 3 60 min rib retraction and allodynia and n = 3 60 min rib retraction without allodynia) were perfused intracardially with saline wash followed by a dilute fixative (1% paraformaldehyde, 1.25% glutaraldehyde) and then concentrated fixative (4% paraformaldehyde, 5% glutaraldehyde). The 4th and 5th intercostal nerves were carefully dissected out and cut into 3 pieces. The distal segment was cut from the largest cutaneous branch distal to its bifurcation. The middle segment included the bifurcation of the lateral cutaneous branch, whereas the proximal segment was 8 mm proximal to the bifurcation. The nerve sections were stained with aqueous 2% osmium tetroxide and embedded in Epon-Araldite for 1-µm transverse sectioning and histologic analysis. Intercostal nerves not subjected to retraction (pleural incision only) were used as histological controls. The investigator (JMK) examining the histological slides was blinded as to the group allocation. The prevalence of mechanical allodynia as a function of rib-retraction time, including pleural incision control, was compared using Fishers exact test. The withdrawal force thresholds and the scratches/min for each animal were averaged over days 1040 and then compared to presurgery baseline value with Wilcoxons signed rank test. Statistical power was not sufficient to compare presurgical baseline allodynia to the allodynia on each day postsurgery over the 40-day observation period. For drug studies, withdrawal force thresholds (mechanical allodynia) and scratches/min (cold allodynia) pre- and postdrug injection values were compared with the Wilcoxons signed rank test. For dose-response curves of mechanical allodynia, the force in grams was converted to a maximum possible effect (%MPE) by the formula [(force threshold postinjection force threshold preinjection)/(15.1 force threshold preinjection)] x 100. For dose-response curves of cold allodynia, the number of scratches/min was converted to a %MPE by the formula [(scratches/min preinjection scratches/min postinjection)/(scratches/min preinjection)] x 100. The 50% effective dose (ED50) and confidence intervals (CI) of each drug to reduce allodynia was calculated using probit analysis (SPSS software; SPSS, Chicago, IL). All data are displayed as mean ± SE.
Before surgery, the median force threshold to von Frey filament stimulation of the dorsal upper back was 15.1 g. With rib retraction, mechanical allodynia was not seen prior to Day 2. By Day 10 mechanical allodynia postsurgery was either well established or it did not occur at all in an animal, even with continued assessment until 40 days postsurgery. The most frequent mechanical allodynia (comparison made at Day 40 postthoracotomy) was 50% (threshold force, 1.8 g), and this occurred after rib-retraction duration of 60 min. This was different from the incidence of allodynia with pleural incision animals (0%). Retraction times of 5 and 30 min produced mechanical allodynia in 11% and 10% of the animals, respectively (not significantly different from pleural incision animals). The time course of the development of mechanical allodynia is shown in Figure 1b. Withdrawal force over 1040 days was significantly less than presurgery baseline (Table 1). Before surgery, there were 1.3 ± 1.0 scratches/min in response to acetone application. In the 60 min rib-retraction animals evidencing allodynia (50% of animals), this increased to 9.4 ± 0.8 scratches/min at 40 days postthoracotomy (Fig. 1c). The number of scratches/min over 1040 days was significantly more than presurgery baseline (Table 1).
Rib-retraction animals with allodynia did not show any difference in body weight gain (16.7 ± 4.2 g/wk) as compared with pleural incision control animals (18.0 ± 2.1 g/wk). Allodynic animals did not evidence spontaneous vocalization or licking of the incision site. These animals did occasionally scratch the incision region without stimulation, but the frequency was <2 scratches per 15 min period for all animals and was thus too infrequent to quantify in this study. Animals with mechanical allodynia on the right dorsal upper back also demonstrated force thresholds below presurgery baseline (<15 g) on the corresponding region on the left side (measured at 13 days postsurgery). However, the contralateral thresholds were on average 4.3 times larger than the ipsilateral side, and so we measured only the ipsilateral side for the remainder of the allodynia time course experiments and all drug injection experiments. Cold allodynia was not different from baseline on the contralateral side on 13 days postsurgery. The cross-sectional nerve histology was examined at a point of bifurcation midway along the course of the 4th and 5th intercostal nerves and at locations proximal and distal to this site. The pleural incision control 4th intercostal nerve (Fig. 2a) and all of the 5th intercostal nerves had a normal appearance in the size and distribution of myelinated axons. Abnormalities observed in the experimental 4th intercostal nerves depended on whether animals exhibited allodynia after rib retraction. In those animals exhibiting allodynia, the general appearance was extensive Wallerian degeneration with only a few intact fibers (<10% of pleural incision control) and some remyelinated axons with a small diameter and thin myelin sheath (Fig. 2b). The 4th intercostal nerve of one rat with allodynia showed total degeneration (i.e., no spared fibers) and extensive proliferation of endoneural cells, but no remyelination was evident at 14 days postsurgery (Fig. 2c). The 4th intercostal nerves from 2 of the 3 animals without allodynia were almost normal in appearance (>95% fibers intact) with only an occasional profile of degeneration or myelin sheath disruption (Fig. 2d). In one nonallodynic rat, some degeneration was observed with swollen remnants of the myelin sheath and some newly regenerated myelinated axons. However, even in this animal, most myelinated axons appeared normal (>90% fibers intact).
Intraperitoneal injections of morphine sulfate or gabapentin reduced mechanical allodynia in the rib-retraction model. The dose-response curves are shown in Figure 3a, and the ED50 and CI for all of the drug testing are summarized in Table 2. Morphine sulfate was tested in doses from 0.5 to 3 mg/kg and had an ED50 of 1.06 mg/kg for mechanical stimulation. Gabapentin was administered over the range 12.537.5 mg/kg, and the mechanical ED50 was 24.2 mg/kg. Cold allodynia was also reduced by intraperitoneal morphine, with an ED50 of 1.98 mg/kg; however, gabapentin never achieved 50% MPE over the dose range tested (Fig. 3b). Intraperitoneal saline did not reduce allodynia (mechanical % MPE, 2.4%; cold % MPE, 0.2%).
Intrathecal injections of morphine, gabapentin, clonidine, and neostigmine also reduced mechanical allodynia (Fig. 4a). Morphine sulfate was tested in doses from 0.55.0 nmol (0.191.9 µg salt) with an ED50 of 1.19 nmol for mechanical stimulation. Gabapentin in doses from 560 nmol (0.8310 µg) yielded an ED50 of 13.8 nmol. Doses of clonidine from 37.593.8 nmol (1025 µg salt) produced an ED50 of 72.7 nmol. Neostigmine in doses from 0.333.3 nmol (0.11.0 µg salt) had an ED50 of 0.54 nmol. MK-801 was tested over the range 1030 nmol (3.410.2 µg salt), with an ED50 of 23.5 nmol, although the dose-response curve did not reach full efficacy at the largest dose. CNQX did not reduce allodynia even at the 30-nmol dose. Cold allodynia was effected similarly by the same drugs given intrathecally (Fig. 4b), with intrathecal ED50 values as follows: morphine 0.92 nmol, gabapentin 16.6 nmol, clonidine 58.5 nmol, neostigmine 0.39 nmol, and MK-801 21.4 nmol. Intrathecal saline did not reduce allodynia (mechanical % MPE, 0.4%; cold MPE, 6.7%).
This new animal model of CPTP produces a nerve pressure or stretch injury, as can occur in patients during thoracotomy with rib-retraction. A 60-minute rib-retraction between the 4th and 5th ribs produced a 50% incidence of long-term allodynia in rats, which is a similar rate to that observed in humans (2,3,18). The intercostal incision that is used extensively for thoracotomies has now become increasingly popular for various minimally invasive approaches to the heart, with rib-retraction typically lasting for a few hours. The combination of skin and muscle incisions and pleura opening followed by a 5 or 30 minute rib-retraction produced only a 10%11% incidence of allodynia, suggesting that more prolonged pressure or stretch of the intercostal nerves is a key factor producing the syndrome of CPTP. The primary sensory modality used to evaluate pain in this study was mechanical stimulation, chosen to reflect the increased discomfort that patients with CPTP experience with constant chest movement. We did not evaluate nerve conduction immediately after rib-retraction as in the clinical study of Rogers et al., (8) because of the short length (1.5 cm) of the exposed intercostal nerve region making it difficult to accurately measure conduction velocity in rats. The mechanical and thermal allodynia in the rib-retraction model is similar to the hypersensitivity seen in the CCI model of Nara et al. (9). Ligation of intercostal nerves yielded allodynia in 70% of rats in the CCI model. In addition, both the mechanical and cold allodynia lasted for at least 27 days, which is similar to our rib-retraction model producing allodynia for at least 40 days. However, intercostal nerve ligation is not routinely performed during thoracotomy procedures, which limits extrapolation of that model to the clinical scenario. The nerve injury attributable to long-duration rib-retraction can be a result of pressure on the nerve transmitted from the retractor tines (causing direct axonal compressional injury or ischemic injury) or the stretching of the intercostal nerve at both edges of the retractor as a result of the displacement of the tissue containing the nerve. Because the 5th intercostal nerve was normal in all animals it is likely that in our model compression was more important than stretch as the cause of the nerve lesion. Animals with near normal intercostal nerves correspond to Sunderland Grade 1 injury (neuropraxia) in which there is minimal or no discernible histopathology of nerve structure (19,20). Clinically, there can still be loss of function that persists for hours or days (20), and this is consistent with the findings of Rogers et al. (8) that 100% of patients had loss of intercostal nerve conduction just above the retractor. Animals with almost total or complete fiber degeneration in the intercostal nerves correspond to Sunderland Grade 2 injury, in which axon continuity is disrupted with Wallerian degeneration and increased numbers of Schwann cells, followed by regeneration. Some characteristics of a Sunderland Grade 3 injury, specifically edema, are also seen (Fig. 2c). Allodynia in the rib-retraction model was associated with the loss of nearly all (>90%) myelinated fibers at 14 days postsurgery. This finding is similar to the CCI sciatic nerve model in which most myelinated fibers disappear during the time course of allodynia (21,22). It is reasonable to question why only 50% of the 60 min rib-retraction animals develop allodynia. Neuropathic pain models producing overt injury to the nerve related to nerve ligation have a more frequent incidence of allodynia (10). In the rib-retraction model, the compression of the nerve is transmitted indirectly via the muscle, connective tissue, and rib adjacent to the nerve. Therefore, there can be variability in the degree of injury to the nerve itself. It is also likely that axon recovery is not possible with a more severe injury such as nerve ligation, but with an indirect injury of limited duration (60 min rib-retraction) there is a greater probability that permanent structural damage to the axon will be avoided. Recently there is a trend to use opioids for the clinical management of neuropathic pain (23). Morphine administered either systemically or intrathecally was effective in reducing allodynia in both this model and the CCI sciatic model in rats (2). It should be noted, however, that we only evaluated single bolus injections and the findings may not be extrapolated to chronic therapy of allodynia produced by rib retraction. Gabapentin, which is often used as a medication for neuropathic pain (24), was effective both systemically and intrathecally in reducing mechanical allodynia and was effective intrathecally in reducing cold allodynia. Two additional drugs, clonidine (25) and neostigmine (26), that have shown efficacy for modulating neuropathic pain in both animal models and clinically with intrathecal administration also demonstrated good efficacy in the rib-retraction model. Glutamate antagonists appear less effective intrathecally in this model, with the N-methyl-D-aspartic acid antagonist MK-801 showing some antiallodynic effect but never achieving full efficacy. At doses that produce antinociception in other animal models (27) the AMPA antagonist CNQX was ineffective in attenuating allodynia after rib retraction. In retrospect, the ED50 values would be more accurate if more doses less than 50% MPE had been tested. This study involved administration of therapeutic drugs after allodynia had been established in this new animal model of CPTP. A more effective strategy might use preventive measures to reduce or eliminate long-term postthoracotomy pain. Some clinical studies have emphasized preemptive analgesia, although the effectiveness of this approach varies greatly among investigators (28,29). Two groups of investigators report that preemptive thoracic epidural analgesia reduced long-term pain after thoracotomy (30,31), but another group did not find long-term reduction of pain (32). Another group compared presurgical systemic morphine and diclofenac and intercostal nerve blocks with the same regime postsurgery and found no difference in postthoracotomy pain 1 year later (33). The main value of this rat model of rib-retraction induced allodynia may be in testing and optimizing new techniques for preventing CPTP syndrome before clinical trials. In summary, we characterize a new model that induces long-term allodynia after thoracotomy and rib-retraction and demonstrates correlation of histologic abnormality with measured allodynia. The feasibility of using the model to test interventions during surgery to reduce risk of nerve injury, as well as strategies to reduce the frequency and severity of persistent pain syndrome after thoracotomy and rib-retraction, are future goals.
Supported, in part, by University Anesthesiologists.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|