Goldberg et al • Low Dose Ketamine for Treatment of CRPS 175
Pain Physician. 2005;8:175-179, ISSN 1533-3159
Prospective Case Series
Multi-Day Low Dose Ketamine Infusion for the Treatment of
Complex Regional Pain Syndrome
Michael E. Goldberg, MD, Richard Domsky, MD, Denise Scaringe, MD, Robert Hirsh, MD, Jessie Dotson, MSN,
Imran Sharaf, MD, Marc C. Torjman, PhD, and Robert J. Schwartzman, MD
Background: Complex regional pain the treatment of CRPS diagnosed by Interna- log pain scale of 0-10 and the affective com-
syndrome (CRPS) is characterized by pain tional Association for the Study of Pain (IASP) ponent using a verbal scale of 0-4.
that is out of proportion to the injury and is criteria in patients who have failed conserva- Results: There was a signiﬁcant reduc-
regional in distribution. A large body of litera- tive treatment. tion in pain intensity from initiation of infu-
ture supports a dynamic change in the physi- Design: Open label, prospective, pain sion (Day 1) to the 10th day, with a signiﬁcant
ology and structure of central pain projecting journal evaluation of a 10-day infusion of in- reduction in the percentage of patients expe-
neurons mediated through the N-methyl-D- travenous ketamine in the CRPS patient. riencing pain by Day 10 as well as a reduction
aspartate (NMDA) receptor. A critical factor Methods: Patients diagnosed with in the level of their “worst” pain. The nadirs
in central sensitization seems to be the re- CRPS by a single neurologist were assigned of pain were lower by Day 10 with a signiﬁ-
lease of the magnesium block on the NMDA to receive a 10-day outpatient infusion of ket- cant reduction in the incidence of “punishing
receptor with inﬂux of calcium and initia- amine supervised by an Anesthesiologist/ pain.” Moreover, there was a signiﬁcant im-
tion of intracellular cascades. Current litera- Pain Management Specialist. The infusion provement in the ability to initiate movement
ture supports the effectiveness of ketamine was administered in a short procedure unit by the 10th day.
in blocking central sensitization through its after each patient had been instructed on Conclusion: A four-hour ketamine infu-
effects on the NMDA receptor. Recent treat- how to complete a pain questionnaire. Moni- sion escalated from 40-80 mg over a 10-day
ment with anesthetic doses of ketamine in toring consisted of continuous ECG, pulse ox- period can result in a signiﬁcant reduction of
severely ill patients with generalized CRPS imetry, and non-invasive blood pressure ev- pain with increased mobility and a tendency
prompted our interest in a lower dose ther- ery 15 minutes. Patients made journal entries to decreased autonomic dysregulation.
apy. each day prior to the infusion of 40-80 mg of Keywords: Complex regional pain syn-
Objective: To report on the eﬃcacy of ketamine. The subjects were also asked to drome (CRPS), ketamine, neuropathic pain
low dose outpatient ketamine infusion for rate their pain intensity using a verbal ana-
Complex regional pain syndrome neurons mediated through the N-methyl- with consequent influx of calcium and
(CRPS) is characterized by pain that is D-aspartate (NMDA) receptor. The clini- consequent initiation of intercellular cas-
out of proportion to the injury and is re- cal elements include autonomic dysregu- cades, appears to be a critical factor in ini-
gional in distribution (1). It is primari- lation, spontaneous pain, evoked pain and tiation of central sensitization (6, 7). Ex-
ly caused by peripheral trauma although movement disorder, and in severe cases, perimental and clinical literature supports
approximately 10% occurs from lesions trophic changes. The incidence has not the effectiveness of ketamine in blocking
in central pain pathways. A large body been determined although it occurs more central sensitization by its effects on the
of literature now exists both from animal frequently in females (2). NMDA receptor (8). Ketamine is a drug
models and clinical experience (2-7) sup- Present evidence suggests that a per- that is rapidly distributed into the brain
porting a dynamic change in the physiolo- sistent nociceptive barrage maintains a and other highly perfused tissues, with
gy and structure of central pain projecting state of central sensitization in central about 10% of the drug bound to plasma.
pain projecting neurons (3, 4). The con- The bioavailability of ketamine is depen-
From: Department of Anesthesiology, Division of sequences of central and peripheral sen- dent on the route of administration, being
Pain Management, Cooper University Hospital, The sitization are a lower threshold to fire C as high as 93% for an intramuscular dose,
Robert Wood Johnson Medical School, Camden, NJ;
and Department of Neurology, Drexel University
and A delta nociceptors, a spread of cu- 20-50% for an intranasal dose, and ap-
College of Medicine, Philadelphia, PA taneous receptive fields of central project- proximately 20% for an oral dose.
Address Correspondence: ing neurons, a change in spinal cord and Recent treatment with anesthetic
Michael E. Goldberg, MD cortical pain maps, and spontaneous pain doses of ketamine for severely ill, gener-
One Cooper Plaza, Camden, NJ 08103
Disclaimer: There was no external funding in prepa- (5). There are clear changes in the central, alized CRPS patients has shown some ef-
ration of this manuscript. autonomic, and motor systems that evolve ficacy and prompted its use in less severe-
Conﬂict of Interest: None concomitantly with changes in pain path- ly ill patients by low dose infusion (9). The
Manuscript received on 11/13/2004
ways in CRPS (1). Central in this process results of this therapy are reported for 40
Revision submitted on 2/26/2005 is the NMDA receptor. The release of the patients with moderate to severe long-
Accepted for publication on 2/28/2005 magnesium block at the NMDA receptor standing CRPS.
Pain Physician Vol. 8, No. 2, 2005
176 Goldberg et al • Low Dose Ketamine for Treatment of CRPS
Table 1. Patient treatments prior to ketamine protocol
Physio- Anti- Anti- Sodium Channel Sympathetic Lidocaine
NSAID Spasmolytics Opioids
Proportion Therapy Depressants Convulsants Blocker Block Infusion
of Patients B A B A B A B A B A B A B A B A B A
(37) (37) (37) (37) (37) (37) (37) (37) (37) (37) (34) (34) (36) (36) (31) (31) (30) (30)
3% 8% 3% 8% 3% 9% 6% 3% 3%
< 10% 0% 0% 0% 0% 0% 0% 0% 0% 0%
(1) (3) (1) (3) (1) (3) (2) (1) (1)
43% 97% 13% 97% 16% 79% 24% 83% 35% 38% 88% 3% 64% 13% 97% 27% 97%
(16) (36) (5) (36) (6) (29) (9) (31) (13) (13) (30) (1) (23) (4) (30) (8) (29)
46% 57% 81% 8% 59% 3% 76% 43% 41% 78% 8% 61% 57%
30-50% 0% 0% 0% 0% 0%
(17) (21) (30) (3) (22) (1) (28) (16) (14) (28) (3) (19) (17)
8% 3% 22% 3% 5% 14% 14% 24% 22% 12% 6% 19% 25% 23% 3% 16% 3%
(3) (1) (8) (1) (2) (5) (5) (9) (8) (4) (2) (7) (9) (7) (1) (5) (1)
Patients reported initial and long-term quality of analgesia for their various therapies prior to entry into the intravenous Ketamine protocol.
B -Before indicates pain relief achieved for up to 8 weeks while on any one of the above therapies
A: After indicates chronic pain relief >8 weeks in duration. Three patients’ clinical records could not be located which explains n=37.
METHODS sessment of the patient’s ability to initiate RESULTS
After approval from the local Institu- movement of the effected extremity using Thirty-six female and four male pa-
tional Review Board, 40 American Society a 10-point scale. tients participated in the study. Mean de-
of Anesthesiologists, Physical Status Clas- Prior to ketamine infusion, subjects mographic data for age, weight, and height
sification I or II patients with a primary were admitted to a short procedure unit were 42 + 10 years, 156 + 45 lbs., and 65 +
diagnosis of CRPS I or II gave written in- and instructed on proper completion of 3.5 inches respectively. Compared to base-
formed consent to participate in this pro- a pain questionnaire. They were moni- line there were significant (p=0.001) re-
spective study. tored with continuous ECG, pulse oxim- ductions in pain intensity (7.54 + 1.93 vs.
The patients had a history of long- etry, and non-invasive blood pressure ev- 5.44 + 2.87) (Fig. 1) and in percentage of
standing or rapidly spreading CRPS, re- ery 15 minutes. Forty-80 mg of ketamine overall pain relief by the 10th day (43.61 ±
fractory to conventional therapy which was mixed in 500 cc of a normal saline so- 27.79) (Fig. 2). Analysis of each patient’s
included: a) Physical therapy; b) drug lution. All of the patients were started on a journal for levels of “worst daily pain” ex-
combinations of NSAIDS, tricyclic anti- 40 mg infusion lasting four hours, and the perienced revealed a significant reduction
depressants, anticonvulsants, and opioids; infusion was increased over a ten-day pe- (p<0.001) in this measure by the 10th day
c) sympatholysis either by intermittent riod to a maximum of 80 mg. Each patient of infusion (8.77 + 1.33 vs. 6.63 + 2.72)
superior cervical or paravertebral block, also received clonidine 0.1 mg orally prior (Fig. 1).
or five days intrapleural or epidural block. to the infusion to prevent a hypertensive Compared to the first day of treat-
Four patients had failed a therapeutic tri- response and possible muscle pain, as well ment, patients also had a lower “least dai-
al of dorsal column stimulation. The pa- as midazolam (2-4 mg) to relieve anxiety. ly pain” score (p=0.006) by the 10th infu-
tients referred for therapy were diagnosed Throughout the treatment period, pa- sion day (5.91 + 2.19 vs. 4.24 +2.75) (Fig.
to have persistent and/or progressive se- tients were monitored for side effects in- 1). In this population where pain was also
vere disease, and no known contraindica- cluding: hypertension, tachycardia, dys- described as burning, aching, and punish-
tions to ketamine, clonidine, or midazol- phoria, hallucinations, dreams, and head- ing, we found that a significant reduction
am. Prior to entering the ketamine proto- aches. The infusion was spread over a two- (p=0.007) in the incidence of “punishing
col, these patients had been treated for a week period excluding weekends. Patient pain” was achieved by the 10th day of infu-
period of three months to three years. journal entries were made each day pri- sion (1.61 ± 1.22 to 0.82 + 1.13) (Fig. 3).
The ketamine infusion was adminis- or to infusion. The subjects were asked to In addition, patients were asked to
tered on an outpatient basis under the su- rate the intensity of their pain using a ver- summarize their pain level over the pre-
pervision of an Anesthesiologist/Pain Spe- bal analog pain scale of 0-10 (0 = no pain, vious 24 hours as another measure over
cialist. The same neurologist (RJS) made 10 = worst pain possible) and the affec- time of treatment efficacy. By the 10th day
the diagnosis of CRPS based on the Inter- tive component of their pain using a ver- of infusion this pain measure had also
national Association for the Study of Pain bal scale of 0-4 (0 = none, 1 = mild, 2 = been reduced significantly (7.88 ± 2.02 vs.
(IASP) criteria. Patients were maintained moderate, 3 = severe). 5.5 ± 2.75) (Fig. 1) (p<0.001).
on their usual medications/treatments Pain data were analyzed using the Patients’ ability to initiate move-
and those were not altered during the Kruskall-Wallis test with p<0.05 consid- ment showed significant improvement
infusion period (Table 1). The baseline ered statistically significant. Data are pre- (p=0.012) by the 10th day of infusion (6.4
physical exam also included a general as- sented as mean ± standard deviation. + 2.6 vs. 4.4 + 3.2) (Fig. 4). A trend to-
Pain Physician Vol. 8, No. 2, 2005
Goldberg et al • Low Dose Ketamine for Treatment of CRPS 177
Day 1 Day 5 Day 6 Day 10
1.3 7.8 ± 7.9 ± 7.9*
1.9 2.1 1.9 6.6 ± ± 2.0 6.9* 6.8*
6.6 ± 6.5 ±
2.4 2.5 5.4 ± 2.7 5.9 ± ± 1.1 ± 2.4
5.3 ± 5.3 ± 5.5*
2.9 2.6 2.8 4.2 ± ± 2.8
Initial Worst pain Least pain Pain rating in past 24 hrs.
* indicates signiﬁcant difference from Day 1
Fig 1. Pain intensity (mean) at 1. initial, 2. worst pain, 3. least pain, and 4. pain rating in past 24 hrs.
wards a reduction in skin color changes
was noted by the 10th day of infusion al-
though this observation did not reach sta-
tistical significance. Overall, side effects
were minimal with 4/40 and 5/40 patients
reporting headaches and restlessness re-
spectively with infusion. There were no
episodes of desaturation (SpO2<93%)
and 3/40 patients experienced a 20% in-
crease over their baseline heart rate dur-
ing the infusion of ketamine. None of
these side effects required intervention.
No patient reported hallucinations or
nightmares over the duration of exposure
All of the patients expressed positive
feelings about the treatment, the quality
of their pain relief, and confirmed that
they would have no objection to repeating
this mode of therapy if necessary. Finally,
all of the changes recorded in the variables
Fig. 2. Percent pain relief from days 1 through 10 (mean ± SD) measured appeared to be progressive over
the days of infusion (Figs. 1-4).
Complex regional pain syndrome
is often described by patients as burn-
ing, throbbing, or aching pain, as well as
mechano- and thermal allodynia (1). The
syndrome is often debilitating and can re-
sult in complete disability. Multiple treat-
* indicates signiﬁcant difference from Day 1
ment modalities have been attempted in-
1.61 ± 1.22 cluding physical therapy, psychotherapy,
behavior modification, surgery, interven-
0.95* ± 1.15 0.84* ± 1.14 0.82* ± 1.13 tional pain therapies, and medications
(16, 17). All were reported to have some
degree of success, but with great variabil-
ity in the quality of the response. In the
most severe cases, the interventional treat-
Day 1 Day 5 Day 6 Day 10
ments are short lived and may not show
Fig 3. Mean pain intensity punishing type pain intensity positive effects.
Multiple studies have suggested that
Pain Physician Vol. 8, No. 2, 2005
178 Goldberg et al • Low Dose Ketamine for Treatment of CRPS
The authors wish to thank the ed-
itors of Pain Physician for peer review
and constructive criticism, which ulti-
mately improved the quality and un-
derstanding of the manuscript.
Michael E. Goldberg, MD
Chief, Department of Anesthesiology
Professor of Anesthesiology
UMDNJ – Robert Wood Johnson
* indicates signiﬁcant difference from Day 1 Medical School at Camden
Cooper University Hospital
One Cooper Plaza
Camden, NJ 08103
Days of treatment
Fig 4. Ability to initiate movement during the 10 days infusion
(mean ± SD) Richard Domsky, MD
Co-Director, Division of Pain
Management, Cooper University
the use of N-methyl-D-aspartate receptor Hospital, One Cooper Plaza
a decrease in their worst episodes of pain
antagonists can reduce the pain response Camden, NJ 08103
that we believe to be clinically significant.
in patients with neuropathic pain (18- Furthermore, patients reported that the Denise Scaringe, MD
21). These receptors are phosphorylated pain, when reduced, was much more tol- Attending Anesthesiologist
and their channel properties are altered, erable over a given 24-hour period. The Cooper University Hospital
thereby changing the physiology of cen- improvement was noted to be progres- One Cooper Plaza
tral pain projecting neurons (22, 23). sive over the infusion period and suggests Camden, NJ 08103
The goal of our treatment modality that continued treatment (longer than 10 Robert Hirsh, MD
was to expand on the technique previous- days) might produce a more significant Co-Director, Division of Pain
ly described by Kiefer et al (15). In that response. At the time of publication of Management, Cooper University
study, patients with severe CRPS who had this manuscript we report that four pa- Hospital, One Cooper Plaza
been resistant to conservative therapies tients (10%) had a return of “worst” and Camden, NJ 08103
successfully underwent high dose (coma “punishing” pain to pre-infusion levels Jessie Dotson, MSN
inducing) ketamine therapy in an ICU by two weeks post treatment. Twenty-five Advance Practice Nurse, Division of
setting (15). This procedure may have patients (62%) had at least a 70% reduc- Pain Management, Cooper University
significant risks and other difficulties re- tion of “worst” and “punishing” pain for Hospital, One Cooper Plaza
sulting from five days of immobilization, six weeks and were back to baseline pain Camden, NJ 08103
risk of nosocomial infection, need for in- levels by nine weeks post treatment. Eight
vasive monitoring, parenteral nutrition, patients (20%) had a >70% reduction Imran Sharaf, MD
endotracheal intubation, and mechanical in those same pain measures for 11-12 Resident in Anesthesiology
ventilation. Therefore, our rationale for weeks. Three patients remain CRPS free at Cooper University Hospital
the treatment of these less severely affect- 15 months following treatment. One Cooper Plaza
ed patients was to use a technique of low Camden, NJ 08103
dose ketamine administration, and a lon- CONCLUSION Marc C. Torjman, Ph.D.
ger infusion. The maximum dose used in The results of this study demonstrat- Director of the Division of Research
this study (20 mg/hr) was well below the ed clinically significant benefits of the Department of Anesthesiology
reported doses associated with psychomi- technique in this specific patient popu- Cooper University Hospital
metic effects (5, 24). lation. Although pain data showed some Education and Research Building,
The results indicate that the use of variability, the results are encouraging Suite 394, 401 Haddon Avenue,
an escalated infusion, from 40 mg over and point to the need for additional stud- Camden, NJ 08103
four hours to 80 mg over four hours/day ies (i.e., oral medication, longer thera- Robert J. Schwartzman, MD
for 10 days, can result in significant re- py, more specific therapies) with spe- Professor and Chairman, Department
duction of pain with increased mobility cific NMDA receptor antagonists in this of Neurology
and a tendency to decreased autonom- population. Further studies with specif- Drexel University College of Medicine
ic dysregulation. Our patients reported ic NMDA receptor antagonists would be 245 N. 15th Street Mail Stop 423
a significant increase in pain relief and beneficial in this population. Philadelphia, PA 19102
Pain Physician Vol. 8, No. 2, 2005
Goldberg et al • Low Dose Ketamine for Treatment of CRPS 179
REFERENCES gional pain syndromes: State of the sci- B, Scutte J. Analgesic eﬃcacy of low dose
ences. Anesth Analg 2002; 95:1812-1816. ketamine. Anesthesiology 1996; 85:304-
1. Janig W, Baron R. Complex regional pain 314.
syndrome: Mystery explained? Lancet 9. Schwartzman RJ, Popescu A. Reﬂex sym-
Neurol 2003; 2:687-697. pathetic dystrophy. Curr Rheumatol Rep 19. Oga K, Kojima T, Matsuura M, Nagashima
2002; 4:165-169. M, Kato J, Saeki S, Ogawa S. Effects of low-
2. Bell RF. Low-dose subcutaneous ketamine dose ketamine on neuropathic pain. An
infusion and morphine tolerance. Pain 10. Sheng M, Kim MJ. Postsynaptic signaling
and plasticity mechanisms. Science 2002; electroencephalogram-electrooculogram/
1999; 83:101-103. behavioral study. Psych Clin Neuro 2002;
3. Clark JL, Kalan GE. Effective treatment of 56:355-363.
severe cancer pain of the head using low- 11. Ji RR, Wolf CJ. Neuronal plasticity and sig-
nal transduction in nociceptive neurons: 20. Klepstad P, Borchgrevink P, Hval B, Flaat
dose ketamine in an opioid-tolerant pa- S, Kaasa S. Long-term treatment with ket-
tient. J Pain Sympt Manag 1995; 10:310- implications for the initiation and main-
tenance of pathologic pain. Neurobiol Dis amine in a 12-year-old girl with severe
314. neuropathic pain caused by a cervical spi-
4. Eide PK, Jorum E, Stubhaug A, Bremnes J, nal tumor. J Pediatr Hematol Oncol 2001;
Breivik H. Relief of post-herpetic neuralgia 12. Caroll RC, Zukin SR. NMDA-receptor traf-
with the N-methyl-D-aspartic acid recep- ﬁcking and targeting: implications for syn-
aptic transmission and plasticity. Trends 21. Nadeson R, Tucker A, Bajunaki E, Good-
tor antagonist ketamine: A double-blind, child CS. Potentiation by ketamine of fen-
cross-over comparison with morphine and Neurosci 2002; 25:571-577.
tanyl antinociception I. An experimental
placebo. Pain 1994; 58:347-354. 13. Choquet D, Triller A. The role of receptor
study in rats showing that ketamine ad-
5. Harbut RE, Correll GE. Successful treat- diffusion in the organization of the post
ministered by non-spinal routes targets
ment of a nine-year case of complex re- synaptic membrane. Nat Rev Neurosci
spinal cord antinociceptive system. Br J
gional pain syndrome type-i (reﬂex sym- 2003; 4:251-265.
Anaesth 2002; 88:685-691.
pathetic dystrophy) with intravenous ket- 14. Hirota K, Lambert DG. Ketamine: Its mech-
22. Schwartzman RJ. The autonomic nervous
amine-infusion therapy in a warfarin-anti- anism of action and unusual clinical uses.
system and pain. In Appenzella O (ed)
coagulated adult female patient. Am Acad Br J Anaesth 1996; 77:441-444.
Handbook of Clinical Neurology, Elsevier
Pain Med 2002; 3:147-155. 15. Kiefer RT, Rohr P, Unertl K, Altemeyer KH, Science, New York, 2001, pp 75:309-347.
6. Correll GE, Maleki JM, Gracely EJ, Muir JJ, Grothusen J, Schwartzman RJ. Recovery
23. Hall R, Murdoch J. Brain protection: physi-
Harbut RE. Subanesthetic ketamine infu- from intractable complex regional pain
ologic considerations. Part II: The pharma-
sion therapy: A retrospective analysis of a syndrome type I (RSD) under high dose in-
cology of brain protection. Can J Anaesth
novel therapeutic approach to complex re- travenous ketamine-midazolam sedation.
gional pain syndrome. Pain Med 2004; 5: Neurology 2002; 58A:475.
263-275. 24. Schmid RL, Sandler AN, Katz J. Use and ef-
16. Smith DJ, Bouchal RH, DeSanctis CA, Mon-
ﬁcacy of low-dose ketamine in the man-
7. Kvarnstrom A, Larlsten R, Quiding H, roe PJ, Amedro JB, Pliotti JM, Crisp T. Prop-
agement of acute post-operative pain:
Emanuelsson BM, Gordh T. The effective- erties of the interaction between ketamine
a review of current techniques and out-
ness of intravenous ketamine and lido- and opiate binding sites in vivo and in vi-
comes. Pain 1999; 82:111-125.
caine on peripheral neuropathic pain. Acta tro. Neuropharmacol 1987; 26:1253-1260.
Anaesthesiol Scand 2003; 47:868-877. 17. Schwartzman RJ, McLellan TL. Reﬂex sym-
8. Baron R, Fields HL, Janig W, Kitt C, Levine pathetic dystrophy. Arch Neurol 1987; 44:
JL. National Institute of Health Workshop: 555-561.
Reﬂex sympathetic dystrophy/complex re- 18. Kochs E, Scharein E, Molenberg O, Bromm
Pain Physician Vol. 8, No. 2, 2005