Dextromethorphan (DX) is a low-affinity, noncompetitive N-methyl-d-aspartate (NMDA) receptor antagonist used to treat a variety of painful conditions. A weak opioid often found in cough suppressants, DX has been shown to inhibit both wind-up and NMDA receptor-mediated nociceptive responses of secondary-order neurons within the spinal cord dorsal horn (1,2). In studies involving pain hypersensitivity from surgery for a wide range of operative procedures, DX has been shown to provide effective pain relief (preemptive analgesia) when administered before or during the surgical procedure (3–6). For neuropathic pain, clinical trials have been mixed (6–10). In patients taking opioids, NMDA antagonists may potentiate the effects of these analgesics (11,12), possibly by reducing the development of tolerance (13).
The noncompetitive NMDA antagonist ketamine has also been shown in clinical studies to attenuate pain hypersensitivity (14–16). This effect of ketamine on neuropathic pain seems to be more potent than that of DX (17). Advantages of using the standard IV formulation of ketamine for neuropathic pain treatment include its widespread availability in hospitals, the immediacy of response, and rapid determination of which patients will and will not respond to treatment. Disadvantages of ketamine include the frequent incidence of side effects, difficulty obtaining an oral formulation of the drug, and its abuse potential.
Although oral DX has a superior side-effect profile and tends to be better tolerated than ketamine (6), ascertaining the efficacy of DX in a particular patient often takes weeks because of the long dose titration period. Developing a method whereby clinicians could obtain rapid information regarding the potential effectiveness of oral DX in neuropathic pain patients would therefore save time and resources. Taking advantage of the rapid analgesic effects of IV ketamine and a mandatory waiting period at military facilities for initiating treatment with nonformulary drugs such as DX, we examined the relationship between the response to IV ketamine and oral DX in neuropathic pain patients. The purpose of this study was twofold: 1) to determine whether the analgesic response to IV ketamine might be used to predict the outcome of an oral DX treatment regimen, a rationale similar to that proposed for the IV lidocaine test (18–20); and 2) to provide a framework for more sophisticated future research projects in this area.
At Walter Reed Army Medical Center (WRAMC) and other United States military hospitals, DX capsules are nonformulary drugs requiring an average of 8–12 days for approval. Because Department of Defense personnel must fill their prescriptions at military facilities to avoid payment, this creates an obligatory delay from the time a prescription is written to the time it is filled. Military facilities are also not permitted to carry drug samples, which requires using an interim drug on formulary with similar pharmacological properties in the event that a physician wants to initiate immediate treatment. In the Pain Management Centers of WRAMC and Landstuhl Regional Army Medical Center, the anesthetic ketamine is a logical choice for pain patients prescribed DX.
Permission to conduct this study was obtained from the Department of Clinical Investigation at WRAMC who designated it as an “exempt” protocol, and all patients who gave their informed consent for the IV ketamine infusion. The data were obtained through retrospective chart analysis and clinic follow-up visits with 25 consecutive patients who received IV ketamine followed by oral DX. Before reviewing the data, a power analysis using the n Query 5 program indicated that a sample size of 25 patients had a 90% chance of detecting a Pearson correlation coefficient of 0.60. Inclusion criteria for the ketamine test included age ≥18 yr, a clinical diagnosis of chronic neuropathic pain ≥6 mo duration, and failure to respond to previous neuropathic medications. The exclusion criteria were unstable medical condition and active psychiatric illness.
Before the ketamine test, all patients completed 0–10 visual analog scales (VAS) for pain measurement. After inserting a 20-gauge IV catheter, an infusion of normal saline (NS) was begun, followed 5–10 min later by the injection of 0.1 mg/kg ketamine over 7 min. As per previously published guidelines for the IV lidocaine test (18), patients were blinded as to when the ketamine infusion was begun. Verbal pain scores on a 0–10 scale were obtained during the NS infusion, and again after the injection of ketamine.
After the ketamine infusion, computerized prescriptions for DX were entered into the pharmacy database on all patients, subject to approval by the anesthesia and pharmacy services. The starting dose was 0.5 mg/kg per os tid, titrating up to 1 mg/kg tid over 14 days as tolerated. In 20 patients, DX was started between 7–14 days after the ketamine test; in 4 patients, prescriptions were filled between 4–6 days postketamine infusion; in 1 patient, a delay in approval required waiting 20 days. Additional pain scores were obtained just before starting the DX, and again at the patients’ first follow-up visit 4–6 wk later. During the treatment period, no changes in medication regimens were made.
In addition to demographic data, the following clinical variables were recorded for analysis: primary diagnosis, duration of pain, whether or not the patient was regularly taking opioids (average daily dose of ≥30 mg morphine sulfate equivalents), preinfusion VAS pain score, post-NS verbal pain score, postketamine verbal pain score, pre-DX average daily pain score (obtained by telephone or e-mail), post-DX average daily VAS pain score, maximal DX dose, and side effects of both medications. A positive response to DX was considered to be ≥50% reduction in pain score from baseline. A receiver operating characteristic (ROC) curve was used to determine the optimal threshold to predict DX response from ketamine response. Based on the response to both treatment medications, all subjects were placed in 1 of 4 categories:
- Positive response to both ketamine and DX
- Negative response to both ketamine and DX
- Positive response to ketamine and negative response to DX
- Negative response to ketamine and positive response to DX
Sensitivity, specificity, positive and negative predictive values of the “ketamine infusion test” are given for ROC curve values of 0.5 (≥50% pain relief from ketamine designated as a positive response) and 0.67 (≥67% pain relief from ketamine designated as a positive response). For both of these values, the kappa measurement and observed agreements are noted. Correlation between ketamine and DX response was calculated using Spearman’s rank and Pearson correlation coefficients. Agreement between the ketamine and DX response was also estimated using the Altman-Bland method of differences (21). Categorical data were analyzed using McNemar’s χ2 tests. When indicated, two-tailed P values are given for all statistical analyses.
Twenty-six patients were offered an IV ketamine infusion before receiving DX, with only one patient refusing. The most common diagnoses in the 25 study patients were failed back surgery syndrome (FBSS) with a radicular component (n = 7), complex regional pain syndrome type I (n = 5), peripheral neuropathy (n = 4), central pain and postherpetic neuralgia (2 each). The mean age for the study patients was 46.0 (sd 15.9, range 21–78) yr, and the average duration of pain 3.8 (sd 3.0, range 1–10) yr. There were 10 men and 15 women. Ten patients were taking opioids regularly.
The mean preketamine infusion VAS pain score was 6.6 (sd 1.7, range 3–9). After NS, this decreased to 5.9 (sd 2.0, post-NS range 2–9, range of reduction 0–3); after the ketamine test, the mean pain score decreased to 3.0 (sd 2.3, range 0–6, with 6 patients reporting complete cessation of pain). The average VAS pain score pre-DX was 6.7 (sd 1.9, range 3–9); post-DX, this decreased to 4.0 (sd 2.1, range 0–8, with 1 patient reporting complete cessation of pain). The Pearson correlation coefficient between the responses to ketamine and DX was 0.67 (Spearman’s rank correlation coefficient = 0.68, P < 0.001). The average dose of DX in the 25 patients was 202 (range 90–360) mg/d.
Nineteen patients reported feeling confused or euphoric after ketamine, with one patient complaining of nausea. These side effects disappeared within 30 min in all subjects. Five patients taking DX reported nausea and/or vomiting; in 1 who experienced both nausea and sedation, this prevented increasing the medication beyond 90 mg/d. One patient developed a rash with DX after 16 days that forced her to stop taking the drug. This patient did not experience any pain relief with the drug and her post-DX average VAS score was recorded at 16 days. One patient experienced urinary retention at a DX dose of 240 mg/d, which necessitated reducing her dose to 180 mg/d. Another patient who experienced poor pain relief with DX reported hallucinations at 210 mg/d.
Of the 19 patients who experienced side effects with ketamine, 7 reported side effects with DX. In the eight patients who reported side effects with DX, all but one experienced side effects with ketamine. The incidence of side effects experienced was considerably more frequent for ketamine than DX (P = 0.003) The association between ketamine and DX-related side effects was not statistically significant (contingency coefficient = 0.18, P = 0.36).
When ≥50% pain relief was used as the criterion for a positive response to each drug, the results were as follows: 9 patients experienced a positive response to both ketamine and DX, 8 patients had a negative response to both, 5 patients responded to ketamine but failed to respond to DX, and 3 patients who did not respond to ketamine had ≥50% pain relief with DX. Based on these findings, the sensitivity of the IV ketamine test was 75% (95% confidence interval [CI] = 43%–95%), the specificity 62% (95% CI = 32%–86%), the positive predictive value 64% (95% CI = 35%–87%), and the negative predictive value 73% (95% CI = 39%–94%). The overall observed agreement was 68% (kappa measure of agreement 0.37, 95% CI = 0.01–0.72, P = 0.11). Analysis using the Altman-Bland method of differences suggested a weak agreement between the responses to the 2 drugs (Fig. 1).
Using an ROC curve to plot ketamine response, a ≥67% analgesic response to ketamine was found to have an observed agreement of 84%. Using this as the criterion for a positive response to ketamine, the results were as follows: 9 patients who responded to ketamine responded to DX, 12 patients responded to neither, 3 patients who failed to respond to ketamine obtained ≥50% pain relief with DX, and only 1 patient who had at least 67% pain relief with ketamine did not report good relief with DX. For the three most frequent diagnostic subgroups (FBSS with radiculopathy, complex regional pain syndrome, and peripheral neuropathy), six patients responded to both ketamine and DX, eight responded to neither, and only one patient had a discordant response (positive response to DX but not ketamine in one patient with FBSS). Based on these outcome measures, the sensitivity of the IV ketamine test was 75% (95% CI = 43%–95%), the specificity 92% (64%–100%), the positive predictive value 90% (56%–100%), and the negative predictive value 80% (95% CI = 52%–96%). The kappa measure of agreement was 0.67 (Table 1 and Figs. 2 and 3) (95% CI = 0.38–0.98, P = 0.001).
Ten of the 25 subjects experienced at least some pain reduction (≥1 point) after the NS infusion, with the mean decrease in VAS scores for these patients being 27% (1.8 points on the 0–10 VAS scale; sd 14%). Using ≥50% pain relief with DX and ≥67% relief with ketamine as response criteria, 6 of the placebo responders had a positive effect to both ketamine and DX, 1 responded to ketamine but not DX, 1 responded positively to DX but negatively to ketamine, and 2 patients did not obtain pain relief with either drug. Using these same outcome measures, placebo responders were more likely to respond to both ketamine (70% versus 20%, P = 0.03) and DX (70% versus 33%, P = 0.11) than nonplacebo responders. The demographic and clinical data of patients based on placebo response are presented in Table 2.
The mean preinfusion VAS pain score for the entire sample was 6.6, with the mean pre-DX score being 6.7. Whereas 3 patients did record slightly (≤1 point) lower pain scores just before starting their DX treatment after their ketamine test, 5 patients recorded increased pain scores. A one-time, small-dose ketamine infusion, therefore, seemed to have no measurable long-term (>1 wk) analgesic effect in our subjects.
There were 10 study patients taking daily opioid medications and 15 patients who were “opioid-naïve.” The mean dose in morphine equivalents for the patients receiving regular opioid therapy was 85.5 (sd 62.4, range 30–240) mg. Only 3 of the 15 patients in the “no opioid” group were using intermittent opioids, with all 3 taking short-acting oxycodone in doses <10 mg per day. Based on the ROC curve of ≥67% relief for ketamine and ≥50% for DX, 4 of the 10 patients taking opioids regularly responded to both ketamine and DX, with another 4 responding to neither. The remaining two patients responded to DX but not ketamine. For the 15 patients not receiving opioids, 5 patients responded to both drugs and 8 responded to neither. In the other two patients, one responded to ketamine but not DX, and the other to DX but not ketamine. The data indicate that opioid therapy does not reliably affect the analgesic response to NMDA receptor antagonists. The demographic and clinical data of patients based on opioid use are presented in Table 3.
The present study demonstrates that there is a statistically significant correlation between the analgesic response to IV ketamine and oral DX in patients with chronic neuropathic pain. This relationship seems to be independent of gender, age, the type of neuropathic pain, and coadministration with an opioid. Aside from the brief euphoric effects of IV ketamine, the side-effect profiles did not differ considerably between the drugs. Importantly, the positive and negative predictive values of the ketamine test in relation to the outcome of an oral DX regimen reached 90% and 80%, respectively, with a specificity of 92%. These results indicate that the IV ketamine test may be a useful tool for identifying patients suitable for a trial with oral DX.
Neuropathic pain behaviors such as hyperalgesia are reliably reversed in animal models of neuropathic pain after a single injection of an NMDA receptor antagonist such as MK-801, ketamine, or dextrorphan, the main metabolite of DX (22–26). The reversal of neuropathic pain behavior by a single injection of an NMDA receptor antagonist can last up to 72 hours in an animal model of neuropathic pain (22). These preclinical data indicate that NMDA receptors have a significant role in the maintenance of neuropathic pain. The results from the “IV ketamine test” in the present study are consistent with the findings from animal models of neuropathic pain. Collectively, both preclinical and clinical trials suggest that IV ketamine could be used as a quick indicator for responsiveness to NMDA receptor blockade.
Of note, although more than one-third of the study subjects showed positive responses to ketamine and DX, almost an equal number of study subjects responded negatively to both drugs. This finding is in agreement with preclinical data indicating a selective response of neuropathic pain behaviors to the inhibition of NMDA receptors with dextrorphan (26). For example, dextrorphan was shown to reverse hyperalgesic responses with little effect on allodynia in an animal model of neuropathic pain (26). This is an important issue in clinical neuropathic pain management for two reasons: first, preclinical data suggest that certain neuropathic pain signs and symptoms may not be mediated through an NMDA receptor mechanism. The lack of response to both ketamine and DX in some of our patients is consistent with previous clinical trials evaluating these drugs (17). Second, although quantitative sensory testing (QST) may be helpful in assessing the manifestations of neuropathic pain (hyperalgesia versus allodynia) in human subjects, this test is time-consuming and not readily available in most pain centers. Even with QST, it would still be unclear as to which patients would respond to ketamine or DX. Seen from this perspective, the IV ketamine test may provide a practical means of identifying patients who are likely to respond positively to an oral DX regimen.
An important caveat regarding our clinical experience is that the intensity of neuropathic pain was collectively reported as changes in VAS scores. Although valid and reliable for most clinical studies, changes in VAS scores cannot distinguish between different neuropathic signs and symptoms (e.g., hyperalgesia versus allodynia) in human subjects. It is thus conceivable that a patient with predominantly allodynic symptoms might not have had a favorable response to either ketamine or DX, as suggested by preclinical data (26). Future studies using QST in combination with IV ketamine and/or oral DX may provide additional information on this issue. Another caveat is that the subject size was relatively small in this study, although power analysis indicated the feasibility of reaching meaningful conclusions. Predictive values based on small study populations are limited in their generalizability. Even though this study involved more than one medical center, a prospective, multicenter study with a larger sample size might enhance our findings or even change our conclusions.
Another question that must be addressed is why some patients responded to one NMDA receptor antagonist but not the other. It is possible that the discrepancy in response to the two drugs may be entirely explained by the different routes of administration, variability in dosages, and the placebo response to one or both drugs. In fact, the increased response rate and incidence of side effects with ketamine may be at least partially explained by the stronger relative dose administered. If this is the case, then fine-tuning the IV ketamine test (i.e., using a smaller dose) might eliminate some or all of the inconsistencies. It is our belief, though, that there will never be 100% agreement between the clinical responses to two different medications. Although both ketamine and DX are often classified as NMDA antagonists, the reality is that they are in unrelated drug classes and contain different pharmacological properties. DX is the “d” isomer of the codeine analog of levorphanol. Unlike the “l” isomer, DX is a weak opioid with minimal addictive properties and a very low analgesic ceiling effect, properties that make it an ideal antitussive drug. Although ketamine also binds weakly to opioid receptors, its pharmacodynamic effects are much more diverse. Ketamine not only acts as an antagonist at NMDA receptors, but also blocks non-NMDA glutamate receptors, acts antagonistically at muscarinic cholinergic receptors, facilitates gamma-aminobutyric acidA signaling, and possesses local anesthetic and possibly neuroregenerative properties (27,28).1 Compared with DX, the psychomimetic effects of ketamine are more profound and realized at smaller doses. Because the underlying pain mechanisms cannot possibly be identified in a clinical trial such as this, the myriad of responses observed with both drugs are not surprising given the mechanistic heterogeneity.
It is interesting to note that increasing the cutoff response to ketamine from 50% to 67% did not reduce the number of patients who responded to both treatments, but better identified those who responded to neither. It is equally important to point out that using either cutoff, 3 of 12 (25%) DX responders failed to respond to ketamine. Had the IV ketamine test in its present form been used to guide treatment with DX, these 3 pain patients would never have been placed on a useful analgesic. This underscores the need for better research in this area.
It must also be emphasized that the present data do not provide information on the long-term outcome of DX treatment because the main focus of this study was to determine whether there is a correlation between the analgesic response to IV ketamine and oral DX in neuropathic pain patients. The outcome from a long-term DX treatment regimen (months to years) could be influenced by many different factors, including disease progression, treatment compliance, comorbidity, and use of coanalgesics; such a trial would require an alternative trial design that is beyond the scope of the present study. Nonetheless, the data from this pilot study provide preliminary evidence that an IV ketamine test might prove useful as a predictive measure for clinical response to an oral DX regimen in neuropathic pain treatment. Whether this test assumes the present form as outlined in this report or evolves into an even more accurate test is a question in dire need of an answer. It is anticipated that future studies on this topic would yield additional information on this relevant issue.
1. Dickenson AH, Sullivan AF, Stanfa LC, McQuay HJ. Dextromethorphan and levorphanol on dorsal horn nociceptive neurons in the rat. Neuropharmacology 1991;30:1303–8.
2. Tal M, Bennett GJ. Dextrorphan relieves neuropathic heat-evoked hyperalgesia in the rat. Neurosci Lett 1993;141:107–10.
3. Wu G-T, Yu J-C, Yeh C-C, et al. Preincisional dextromethorphan treatment decreases postoperative pain and opioid requirement after laparoscopic cholecystectomy. Anesth Analg 1999;88:1331–4.
4. Dawson GS, Seidman P, Ramadan HH. Improved postoperative pain control in pediatric adenotonsillectomy with dextromethorphan. Laryngoscope 2001;111:1223–6.
5. Moiniche S, Kehlet H, Berg J. A qualitative and quantitative systematic review of preemptive analgesia for postoperative pain relief. Anesthesiology 2002;96:725–41.
6. Weinbroum AA, Rudick V, Paret G, Ben-Abraham R. The role of dextromethorphan in pain control. Can J Anaesth 2000;47:585–96.
7. McQuay HJ, Carroll D, Jadad AR, et al. Dextromethorphan for the treatment of neuropathic pain: a double-blind randomised controlled trial with integral n-of-1 design. Pain 1994;59:127–33.
8. Nelson KA, Park KM, Robinovitz E, et al. High-dose oral dextromethorphan versus placebo in painful diabetic neuropathy and postherpetic neuralgia. Neurology 1997;48:1212–8.
9. Ben Abraham R, Marouani N, Kollender Y, et al. Dextromethorphan for phantom pain attenuation in cancer amputees: a double-blind crossover trial involving three patients. Clin J Pain 2002;18:282–5.
10. Sang CN, Booher S, Gilron I, et al. Dextromethorphan and memantine in painful diabetic neuropathy and postherpetic neuralgia. Anesthesiology 2002;96:1053–61.
11. Chapman V, Dickenson AH. The combination of NMDA antagonisms and morphine produces profound antinociception in the rat dorsal horn. Brain Res 1992;573:321–3.
12. Nadeson R, Tucker A, Bajunaki E, Goodchild CS. Potentiation by ketamine of fentanyl antinociception. I. An experimental study in rats showing that ketamine administered by non-spinal routes targets spinal cord antinociceptive systems. Br J Anaesth 2002;88:685–91.
13. Manning BH, Mao J, Frenk H, et al. Continuous co-administration of dextromethorphan or MK-801 with morphine: attenuation of morphine dependence and naloxone-reversible attenuation of morphine tolerance. Pain 1996;67:79–88.
14. Schmid RL, Sandler AN, Katz J. Use and efficacy of low-dose ketamine in the management of acute postoperative pain: a review of current techniques and outcomes. Pain 1999;82:111–25.
15. Enarson MC, Hays H, Woodroffe MA. Clinical experience with oral ketamine. J Pain Symptom Manage 1999;17:384–6.
16. Eide K, Stubhaug A, Oye I, Breivik H. Continuous subcutaneous administration of the N
-methyl-d-aspartic acid (NMDA) receptor antagonist ketamine in the treatment of post-herpetic neuralgia. Pain 1995;61:221–8.
17. Mao J. Translational pain research: bridging the gap between basic and clinical research. Pain 2002;97:183–7.
18. Galer BS, Harle J, Rowbotham MC. Response to intravenous lidocaine infusion predicts subsequent response to oral mexiletine: a prospective study. J Pain Symptom Manage 1996;12:161–7.
19. Trentin L, Visentin M. The predictive lidocaine test in the treatment of neuropathic pain. Minerva Anestesiol 2000;66:157–61 (in Italian).
20. Mao J, Chen LL. Systemic lidocaine for neuropathic pain relief. Pain 2000;87:7–17.
21. Ludbrook J. Statistical techniques for comparing measurers and methods of measurement: a critical review. Clin Exp Pharmacol Physiol 2002;29:527–36.
22. Mao J, Price DD, Mayer DJ, et al. Intrathecal MK 801 and local nerve anesthesia synergistically reduce nociceptive behaviors in rats with experimental peripheral mononeuropathy. Brain Res 1992;576:254–62.
23. Yamamoto T, Yaksh TL. Spinal pharmacology of thermal hyperalgesia induced by constriction injury of sciatic nerve: excitatory amino acid antagonists. Pain 1992;49:121–8.
24. Mao J, Price DD, Hayes RL, et al. Intrathecal treatment with dextrorphan and ketamine potently reduce pain-related behaviors in a rat model of painful peripheral mononeuropathy. Brain Res 1993;605:163–8.
25. Tal M, Bennett GJ. Dextrorphan relieves neuropathic heat-evoked hyperalgesia in the rat. Neurosci Lett 1993;151:107–10.
26. Tal M, Bennett GJ. Neuropathic pain sensations are differentially sensitive to dextrorphan. Neuroreport 1994;5:1438–40.
27. Kohrs R, Durieux ME. Ketamine: teaching an old dog new tricks. Anesth Analg 1998;87:1186–93.
28. Himmelseher S, Pfenninger E, Georgieff M. The effects of ketamine-isomers on neuronal injury and regeneration in rat hippocampal neurons. Anesth Analg 1996;83:505–12.