| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Pain Research Center, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital and Harvard Medical School; and Department of Biostatistics, Department of Orthopaedic Surgery, Children’s Hospital and Harvard Medical School, Boston, Massachusetts
Address correspondence and reprint requests to Peter Gerner, MD, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115. Address e-mail to pgerner{at}partners.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We investigated whether doxepin is suitable for topical analgesia. Such a finding could make doxepin clinically useful in cutaneous analgesia during a variety of procedures, including venipuncture, IV cannulation, vaccination, circumcision, dermatological procedures, skin grafting, and for the management of chronic neuropathic pain. Similarly, we investigated the blocking properties of doxepin for spinal anesthesia. We hypothesized that doxepin produces long-lasting anesthesia in rats, specifically that (a) topical doxepin has significantly longer antinociceptive properties than control when assessed by pinprick, and (b) intrathecal doxepin has significantly longer antinociceptive properties than bupivacaine when assessed by pinch.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Doxepin HCl was purchased from Sigma Chemical CO (St. Louis, MO). Solutions of doxepin at 50-, 75-, and 100-mM concentrations were freshly prepared in a vehicle of 45% isopropyl alcohol, 45% water, and 10% glycerin, based on the solution used in the transdermal lidocaine study by Kissin et al. (10). The solution was titrated to a pH value of 8.7 with sodium hydroxide.
The technique of drug patch application has been described (11). Briefly, under sevoflurane inhaled anesthesia (Ultane, Abbott Laboratories, IL), the rats (n = 4–7 per group) were shaved, and a 2-cm2 test area on the lumbar area of their backs was covered with gauze held in place by a 6 x 7-cm transparent adhesive dressing (TegadermTM; 3M Health Care, St. Paul, MN). Care was taken not to injure the skin or break the skin barrier. Once the rats were fully recovered from the sevoflurane anesthesia, 0.3 mL of the doxepin solution at various concentrations was injected through the dressing with a 30-gauge needle, directed obliquely to avoid intradermal injection. The gauze was completely saturated by the solution, and no leakage was observed after withdrawal of the needle. The investigators were blinded to the study drugs. In pilot studies, a patch with doxepin at concentrations smaller than 50 mM (pH value of 8.7) demonstrated no reliable complete blockade (doxepin 50-mM patch at pH value of 5.5 in a similar vehicle did not cause any cutaneous blockade). Thus, we selected doxepin at concentrations of 50, 75, and 100 mM at a pH value of 8.7. The control group had only the vehicle solution applied.
After 2 h, the dressings were removed, and the test area was delineated with a marker. A blinded investigator evaluated the rats by both nocifensive (pain avoiding) reaction and cutaneous trunci muscle reflex (CTMR) in response to a blunt needle at the test area, which was compared to a contralateral control area. The nocifensive reaction consists of withdrawal and vocalization. CTMR is characterized by reflexive movement of the skin over the back, produced by twitches of the lateral thoracispinal muscles in response to local dorsal cutaneous stimulation (12). After observing the rat’s normal reaction at the control area, we applied three sets of six pinpricks (at a frequency of 0.5–1 Hz) to the test area and recorded the number of pinpricks to which the rat failed to respond. Six pinpricks were sufficient for consistent results without causing injury to the rat from excessive testing (12). The analgesic effect was scored quantitatively by the number of times the pinprick failed to elicit a nocifensive and CTMR response. The responses were graded from 0 to 6, with 6 being absence of response to all six pinpricks (complete analgesia); a response identical to the control was scored as 0. The final score was based on the mean number of responses of the three different trials. The pinprick testing was conducted at 2, 4, 6, 20, 30, and 60 h after application of the drug, ceasing when nocifensive behavior and CTMR indicated full recovery from the block.
For rats in the intrathecal catheter group, a PE-10 catheter was inserted during xylazine/ketamine anesthesia by the modified direct lumbar catheterization method (catheter through needle method) (13). The catheter was inserted about 2 cm beyond the tip of the guide cannula into the subarachnoid space to reach the level of the caudal ribs, which corresponds to the lumbar enlargement of the spinal cord. The needle was then carefully withdrawn, avoiding displacement of the catheter. The catheter was tunneled under the skin, exiting at the occipital region, and sutured in place.
For spinal blockade, the intrathecal catheter was injected with doxepin 60 µL at concentrations of 10, 20, and 50 mM (n = 8). These concentrations were based on the results of pilot studies that reported no reliable, complete block was achieved at a concentration <10 mM. Rats in the bupivacaine group (n = 8) were injected with 60 µL only at a concentration of 23 mM (0.75%) because a body of literature describing the blocking properties for the 0.25% and 0.5% concentrations of bupivacaine is already available. In our preliminary studies, some rats injected with the relatively large volume of 60 µL of bupivacaine in the usual bolus fashion died or had unacceptably high spinal block. Therefore, we chose to inject the drugs as a slow infusion (during at least 2 min) and not to increase the concentrations.
Neurobehavioral examination by a blinded investigator included motor function, proprioception, and nocifensive response immediately before injection and at various timed intervals after injection.
Motor function was evaluated by measuring the extensor postural thrust of the hindlimbs by holding the rat upright with the hindlimb extended so that the distal metatarsus and toes supported the rat’s weight and measuring the extensor thrust as the gram force applied to a digital platform balance (Ohaus Lopro; Fisher Scientific, Florham Park, NJ).
The reduction in this force, representing reduced extensor muscle contraction caused by motor blockade, was calculated as a percentage of the control force. Motor function score was graded with respect to the preinjection control value (range, 130–165 g) as follows: 0 (normal), 1 (mildly impaired; force between preinjection control value and 50% thereof), 2 (severely impaired; force between 50% of the preinjection control value and 20 g), and 3 (complete block, force <20 g, which is also referred to as weight of the flaccid limb).
Proprioception evaluation was based on the resting posture and postural reaction (tactile placing and hopping). The functional deficit was graded as 0 (normal), 1 (mildly impaired), 2 (severely impaired), and 3 (complete block). Keeping the rat in a normal resting posture with the toes flexed and the dorsi of the feet placed on the supporting surface, we evaluated tactile placing as the ability to reposition the toes. We evoked a hopping response by lifting the front half of the rat off the ground and then lifting one hindlimb at a time off the ground so that the rat moved laterally. This process normally evokes a prompt hopping with the weight-bearing limb in the direction of movement to avoid falling over. A predominant motor impairment causes a prompt but weaker than normal response. Conversely, with a predominant proprioceptive blockade, delayed hopping is followed by greater lateral hops to avoid falling over or, in the case of full blockade, no hopping at all.
Nocifensive reaction was evaluated by the withdrawal reflex or vocalization to pinch of a skinfold over the lateral metatarsus (cutaneous pain) and of the distal phalanx of the fifth toe (deep pain). Nocifensive reaction was graded on a scale of 0–3 and based on withdrawal reflex, escape behavior, and vocalization in the following manner: 0 (baseline or normal; brisk withdrawal reflex, normal escape behavior, and strong vocalization), 1 (mildly impaired), 2 (moderately impaired), and 3 (totally impaired nocifensive reaction). This neurobehavioral evaluation was modified after Thalhammer et al. (14).
The sample sizes in this study were aimed at 80% statistical power for detecting a median difference of 1 U in each of the neurobehavioral variables (proprioception, motor function, and nociception scores) between the intrathecal doxepin (n = eights rats per dose) and bupivacaine (n = eight rats) groups using nonparametric Mann-Whitney U-test comparisons with a two-tailed significance level of 0.05. Fewer rats were used in the topical doxepin experiment (n = 3–5 rats) because the dose-dependent antinociceptive effects (effect sizes) were expected to be larger than control, and thus, smaller sample sizes would provide adequate (80% power, ß = 0.2; two-tailed 
= 0.05) statistical power.
We primarily focused on comparing topical doxepin versus control and intrathecal doxepin versus bupivacaine with respect to median scores (and ranges) on the neurobehavioral tasks as determined by nonparametric analysis (Mann-Whitney U-test). The P values refer to the probabilities that observed group differences are caused by chance; a two-tailed level of 0.05 was used as the criterion for significance. Differences in duration were ascertained by comparing the slopes for the different groups during the whole time course of evaluation, where Greenhouse-Geisser F-test for interaction was used to assess differences in slope (duration of block) (15). Sample size and power calculations were conducted with the nQuery Advisor software package (version 5.0; Statistical Solutions, Boston, MA).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The antinociceptive effect of doxepin at the concentrations of 50 mM and 75 mM lasted for approximately 6 h. At 100 mM, the effect of doxepin was much more pronounced, with complete recovery occurring only at 60 h.
Two of the rats in the doxepin 100-mM group had erythema and intradermal hemorrhage after 24 h and a scar over the test area after 48 h that healed without sequelae. There were no skin changes in the other groups. No rats displayed any systemic neurobehavioral abnormalities at any time. In particular, no sedation was observed, and grooming patterns and exploratory behavior were normal. Nociception scores in control and topical doxepin groups are depicted in Table 1.
|
Doxepin injected intrathecally produced a dose-dependent duration of block. There were differences in duration as tested by repeated-measures analysis, with the Greenhouse-Geisser (conservative) F-test used for small samples to compare durations of block. Significant differences were detected for each of the 3 tasks between doxepin 50 mM and bupivacaine (all P < 0.0001). No differences were detected between doxepin 10 mM and bupivacaine (all P > 0.05), and a significant difference was detected between doxepin 20 mM and bupivacaine only for the nociception task (P = 0.002).
Neurobehavioral evaluations were not significantly different between doxepin 10 mM and 20 mM and bupivacaine 23 mM (Table 2). At each of the time points evaluated, median scores were not significantly different than those for bupivacaine, according to Mann-Whitney U-test comparisons, except for the nociception task at 150 and 180 min (P < 0.05). In addition, we used repeated-measures analysis of variance with a Greenhouse-Geisser F-test for the interaction between time and dose (slope test), and no differences were detected in the dose-response blockade from time 0 to 180 min between doxepin 10 mM and bupivacaine 23 mM for proprioception (F = 2.75; P = 0.10), motor function (F = 0.98; P = 0.38), or nociception (F = 2.73; P = 0.06). Similarly, no significant differences were found in the time-related changes in neurobehavioral tasks between doxepin 20 mM and bupivacaine 23 mM for proprioception (F = 0.24; P = 0.77) or motor function (F = 1.42; P = 0.25); however, there was a difference in the time-related change for nociception (F = 5.39; P = 0.002). With respect to the change in recovery over the 180-min time course, there were highly significant group differences between doxepin 50 mM and bupivacaine 23 mM for proprioception (F = 25.95; P < 0.0001), motor function (F = 17.54; P < 0.0001), and nociception (F = 24.42; P < 0.0001) variables. At various time points, the doxepin 50-mM group showed clear differences from bupivacaine 23 mM (P < 0.01), and the repeated-measures analysis indicated that the doxepin 50-mM group had a significantly longer block duration for all 3 neurobehavioral tasks (P < 0.01). However, unlike bupivacaine, the doxepin 50-mM group did not fully recover.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
None of the rats showed any neurobehavioral changes suggestive of systemic absorption. We assumed that the systemic absorption after a two-hour cutaneous application of doxepin patch was negligible, but topical doxepin has been reported to cause sedation and dry mouth (16). Although having information on blood concentrations after cutaneous doxepin application would be useful, we elected not to measure plasma levels because of the costs involved. The gathering of such data would be a mandatory part of any potential clinical trial.
Erythema of the skin was seen in only two rats in the 100-mM doxepin group. Possible origins include drug toxicity or irritation caused by the unphysiologically high pH value (because doxepin exists in a charged form and because the protein kinase A is relatively high, the pH value needed to be adjusted to 8.7, with the resultant possibility of skin irritation). Because these changes developed at least 48 hours after the drug application, it is also conceivable that it was due to repeated skin testing with the needle rather than the drug itself. Nonetheless, in the absence of contrary evidence, a drug-induced cutaneous toxicity is postulated.
In pilot studies, an unreliable block (usually partial block of the tail and legs) was found at the usual published intrathecal drug volumes of 10–20 µL (17). The need for a larger volume of doxepin (at the same dosage) than of bupivacaine or lidocaine could be due to a high logP (octanol-water coefficient ratio), causing a low hydrophilicity of the drug (the logP value of doxepin is 4.36. For comparison, bupivacaine’s logP value is 3.44).
To conduct studies that assess the therapeutic range of intrathecal application of doxepin versus bupivacaine or topical application of doxepin versus the Lidoderm 5% patch would have been clinically more meaningful. However, it would have required formally constructing dose-response curves to calculate the 50% effective dose, including histological analyses, which was beyond the scope of the study. Because it is more likely that doxepin finds future application via coinjection with other drugs rather than as a sole drug, we elected to first investigate the basic local anesthetic properties of doxepin while planning more detailed future studies combining doxepin with other adjuvant drugs.
In summary, doxepin, as a topically applied patch, provided effective cutaneous antinociception. Doxepin injected intrathecally at 20 mM is essentially equipotent to bupivacaine 23 mM (0.75%) in rats, whereas larger concentrations would most likely be neurotoxic.
|
Footnotes |
|---|
Supported, in part, by the National Institutes of Health grants nos. GM48090 to GKW and no. GM64051 to PG, Bethesda, Maryland.
Wang GK, Gerner P, inventors. The Brigham and Women’s Hospital, INC., assignee. Tricyclic antidepressants and their analogues as long acting local anesthetics and analgesics. US Patent 60/235 432. April 8, 2003.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
