Commercial

• Type 1 CRPS
  • Continuing pain, allodynia, or hyperalgesia in which the pain is out of proportion to the initiating event
  • Evidence of edema, changes in skin blood flow, or abnormal sudomotor activity in the painful region (this criterion is satisfied by either a sign or a symptom)
  • No other condition that would account for the degree of pain and dysfunction

• Type 2 CRPS
  • Type 2 CRPS is diagnosed when, in addition to the above three criteria for CRPS 1, there is also an initiating noxious event or a cause of immobilization.

• Continuing pain, which is disproportionate to any inciting event
• At least one symptom reported in at least three of the following categories AND at least one sign* at time of evaluation in two or more of the following categories:
  • Sensory: reports/evidence of hyperalgesia and/or allodynia
  • Vasomotor: reports/evidence of temperature asymmetry and/or skin color changes and/or skin color asymmetry
  • Sudomotor/Edema: reports/evidence of edema and/or sweating changes and/or sweating asymmetry
• Motor/trophic: Reports/evidence of decreased range of motion (ROM) and/or motor dysfunction (weakness, tremor, dystonia) and/or trophic changes (hair, nail, skin)
• No other diagnosis that better explains the signs and symptoms

• Allodynia: perception of pain with an ordinarily nonpainful stimulus (e.g., air or clothing)
• Hyperalgesia: increased sensitivity to noxious stimulation
• Hyperesthesia: lowered pain threshold that permits increased pain from typically noxious stimulation
• Neuropathic: pertaining to injury to the peripheral or central nervous system
• Nociceptive: pertaining to damage to tissues due to thermal, chemical, mechanical, or other types of irritants and sensed by nociceptors
• Sudomotor: pertaining to stimulation of the sweat glands
• Vasomotor: pertaining to nerves supplying muscles in the walls of blood vessels

INTRAVENOUS REGIONAL SYMPATHETIC NERVE BLOCKADE (IVRB) IN GENERAL

IV regional sympathetic nerve blockade (IVRB) in CRPS is intended to confine and concentrate medication at the sympathetic nerve endings in a particular region to produce sympathetic denervation without interrupting motor activity (Brown et al., 1997). Frequent or chronic use of the various agents recommended in IVRB has never been supported in the scientific literature (Stanton-Hicks et al., 1998), although Harden (2001) suggests that a short trial of a limited number of nerve blocks with very clear goals and time limitations may be indicated for ethical reasons, and may be cost-effective if attempts at functional restoration are unsuccessful. In any event, the timing of IVRB should be guided by the need for functional restoration (Stanton-Hicks et al., 1998). For example, failure to progress in 2 to 4 weeks with a combination of reactivation, contrast baths, and desensitization to the next level of flexibility exercises, edema control, peripheral E-stimulation, and isometric strengthening should prompt consideration of introducing the most aggressive therapies, such as IV regional sympathetic blocks, psychotherapy, and/or pharmacotherapy (Stanton-Hicks et al., 1998). Few studies support the use of other agents in IVRB (atropine, bretylium, phentolamine, clonidine, guanethidine, parecoxib, reserpine, droperidol with heparin, lidocaine, methylprednisolone with lidocaine, ketorolac, ketanserin), alone or in combination. (Harden et al., 2013; O'Connell et al., 2013; Goebel et al., 2012; Abdi, 2014; Freedman et al., 2014).

SPECIFIC AGENTS PROPOSED FOR IVRB

INTRAVENOUS BRETYLIUM TOSYLATE IN IVRB
Bretylium tosylate had been promoted as an IVRB adrenergic blocking agent. However, bretylium has not been approved by the FDA for use in CRPS. Moreover, although it was once available for other indications, injectable bretylium is no longer available in the United States. Previously, IV bretylium in combination with lidocaine in IVRB significantly reduced pain compared with lidocaine alone in one randomized crossover, controlled study involving 12 patients (Hord et al., 1992). No other trials involving bretylium in CRPS were identified in this assessment.

INTRAVENOUS BUPIVACAINE IN IVRB
A search of the medical literature revealed a single low-quality study of bupivacaine in subjects diagnosed with RSD (Bonelli et al., 1983). The study found no difference between stellate ganglion block via bupivacaine and regional IV block via guanethidine. In 2012, a pilot study by Toshniwal et al. randomly assigned 33 individuals to either continuous stellate ganglion (CSG) block or continuous infraclavicular brachial plexus block (CIPB) for the management of CRPS 1 of the upper extremity. Bupivacaine, 0.125%, was infused for 1 week, after which the individuals were followed for 4 weeks. The CIPB group showed significant improvement compared with the CSG group during the first 12 hours after the infusions. After 12 hours, the pain score was similar between the groups. At 4 weeks, both groups showed significant improvement in edema and ROM scores, which led the authors to recommend a larger well-randomized, well-controlled clinical trial to confirm the findings. Therefore, evidence that IV bupivacaine improves the net health outcome is lacking. Furthermore, the FDA has not approved its use in CRPS.

INTRAVENOUS GUANETHIDINE IN IVRB
Guanethidine is a postadrenergic blocking agent that has been promoted for IVRB in CRPS. However, its use in CRPS has not been approved by the FDA. Guanethidine alone, during an attempt to achieve IVRB, produced no benefit over placebo in an RCT involving 16 patients (Jadad et al., 1995). In a crossover RCT, Ramamurthy and Hoffman (1995) found no significant differences from IV guanethidine alone or in combination with lidocaine. In a low-quality crossover study (Rocco et al., 1989) (12 patients), no difference in pain reduction between intravenously administered guanethidine plus lidocaine, reserpine plus lidocaine, or lidocaine alone was found. An RCT performed by Gschwind et al. (1995) in 71 patients undergoing fasciectomy for Dupuytren’s disease resulted in preoperative intravenous guanethidine showing no advantage over placebo in the prevention of CRPS. In addition, most guanethidine trials have failed to show improvement in efficacy over lidocaine or saline. O’Connell et al. (2013) found moderate-quality evidence that IVRB using guanethidine is not effective and may be associated with complications. A review by Rockett (2014) observes that intravenous regional anesthesia or intravenous regional block (IVRA or IVRB) with guanethidine is no longer recommended due to significant adverse effects. In summary, for intravenous guanethidine in CRPS, there is limited evidence suggesting that it does not improve the net health outcome. In addition, intravenous guanethidine has not received FDA approval for use in CRPS.

INTRAVENOUS KETAMINE IN IVRB
Ketamine, a drug well known for its ability to induce dissociation during anesthesia, has been proposed in IVRB in the treatment for CRPS. The agent has been reported to reduce continuous and evoked pain when given in prolonged and low doses in patients with injury of the peripheral and central nervous system (Correll et al., 2002; Eide et al., 1995). A search of the medical literature produced no RCTs involving the use of IV ketamine in CRPS. Existing studies investigating ketamine are of poor quality or address only non-CRPS pain. Therefore, IV ketamine in CRPS has not been shown to improve the net health outcome. In addition, IV ketamine has not been approved by the FDA for use in CRPS.

INTRAVENOUS KETANSERIN IN IVRB
Two RCT studies investigated IV ketanserin (a selective serotonin receptor antagonist with weak adrenergic receptor–blocking properties) against placebo to achieve regional sympathetic blockade. A study by Hanna and Peat (1989) found a significant improvement in pain in patients diagnosed with RSD who were receiving 10 mg of ketanserin in a single bolus. The other study involving ketanserin (Bounameaux et al., 1984) showed no difference in pain reduction with ketanserin compared with placebo. For IV ketanserin in CRPS, there is inadequate evidence upon which to make conclusions regarding health outcomes. IV ketanserin is not available outside investigational settings, and it has not been approved by the FDA for use in CRPS.

INTRAVENOUS LIDOCAINE IN IVRB
Lidocaine hydrochloride is used during IVRB as a local anesthetic and sometimes with other injectable medications such as a diluent solution providing local anesthesia. An RCT involving seven patients diagnosed with CRPS 1 (Price et al., 1998) showed that IV lidocaine administered to achieve an IVRB resulted in a slight improvement when compared with saline. The improvement was not significant, however. In a low-quality crossover study (Rocco et al.,1989) (12 patients), no difference in pain reduction between intravenously administered guanethidine plus lidocaine, reserpine plus lidocaine, or lidocaine alone was found.

A systematic review (Challapalli et al., 2005) of the analgesic effect of lidocaine and its oral analogues found that these agents were safe in controlled clinical trials for neuropathic pain, were better than placebo, and were as effective as other analgesics. However, these results are limited by the few trials with adequate information on safety and efficacy, and by the heterogeneity of the model. A retrospective study of IVRB with lidocaine and methylprednisolone for treatment of upper extremity CRPS in 168 individuals was conducted in Greece by Varitimidis et al. (2011). According to the authors, after a mean follow-up of 5 years, a complete absence of pain was reported by 92% of patients. They noted that treatment must start early to expect such good results. As with other nonrandomized, uncontrolled studies, the placebo effect cannot be discounted in this report. Eckmann et al. (2011) reported the results of a randomized, double-blinded, crossover study of lidocaine with ketamine in CRPS of the lower extremity in 10 adults where only one individual achieved significant improvement. In addition, Goebel et al. (2012) found evidence to be conflicting for lidocaine sympathetic ganglion blocks. Therefore, for IV lidocaine in CRPS, there is limited evidence that suggest that it does not improve the net health outcome. Lidocaine has not been approved by the FDA for use in CRPS.

INTRAVENOUS PARECOXIB IN IVRB
IV parecoxib is the only injectable COX-2 inhibitor. It is widely used in Europe under the brand name Dynastat. In a randomized, placebo-controlled, double-blind trial, neither pressure pain threshold nor quantitative sensory testing was improved with the drug (Breuer et al., 2014). The FDA declined to approve parecoxib for use in the United States in 2005. The lack of scientific evidence does not permit conclusions regarding net health outcomes.

INTRAVENOUS PHENOXYBENZAMINE IN IVRB
Because oral phenoxybenzamine and other oral α-adrenergic blocking agents have been used with varying degrees of success in CRPS, there has been speculation that phenoxybenzamine could be used as an IVRB agent in CRPS. The agent has not been approved by the FDA for use in​ CRPS (Malik et al., 1998). A search of the medical literature produced no RCTs involving the use of IV phenoxybenzamine in CRPS. Therefore, evidence upon which to base conclusions regarding the use of phenoxybenzamine in IVRB is lacking. The use of phenoxybenzamine in CRPS has not been shown to improve the net health outcome.

INTRAVENOUS PHENTOLAMINE IN IVRB
A high-quality study (Verdugo and Ochoa, 1994) (77 patients) showed that 30 minutes of 35 mg of IV phentolamine for the treatment of “reflex sympathetic dystrophy” was ineffective. Therefore, for IV phentolamine in CRPS, there is limited evidence upon which to base a conclusion (Kosharskyy et al., 2013). That evidence suggests that IV phentolamine does not improve the net health outcome. Therefore, IV phentolamine in CRPS has not been shown to be as beneficial as any established alternatives. According to Wheeler (2014), controlled clinical trials of IV phentolamine showed mixed results and are hampered by poor methodology. IV phentolamine is available outside investigational settings; however, it has not been approved by the FDA for use in CRPS.

INTRAVENOUS PHENYLEPHRINE IN IVRB
In a RCT by Verdugo and Ochoa (1994), phenylephrine given during IVRB for the treatment of “reflex sympathetic dystrophy” was ineffective. Accordingly, for IV phenylephrine in CRPS, there is limited evidence suggesting that it does not improve the net health outcome. Therefore, IV phenylephrine in CRPS has not been shown to be as beneficial as any established alternatives. IV phenylephrine is available outside investigational settings; however, it has not been approved by the FDA for use in CRPS.

INTRAVENOUS RESERPINE IN IVRB
IV reserpine combined with IV guanethidine to achieve regional sympathetic blockade gave no benefit over placebo in one RCT (Blanchard et al., 1990). For IV reserpine in CRPS, there is limited evidence suggesting that it does not improve the net health outcome. IV reserpine has not been approved for use in CRPS.

SYSTEMIC INTRAVENOUS AGENTS IN GENERAL

Casale et al. (2015) comprehensively reviewed currently available systemic drug treatments for CRPS. The authors indicate that over the past few decades, a large number of drugs have been proposed for systemic administration. However, due to the lack of RCTs, this approach to treatment remains empirical. Corticosteroids, although often the first agents to be employed, continue to be controversial. According to these authors, good results have been obtained using dexamethasone and mannitol. However, neither of these agents have been FDA approved for CRPS treatment. The NMDA receptor antagonists (ketamine, amantadine, memantine, dextromethorphan [DM], and methadone) need additional clinical trials to establish the most clinically useful cost/benefit regarding side effects and analgesic action. In addition, none of these drugs have FDA approval for CRPS. The α-adrenoceptor blockers including phenoxybenzamine, phentolamine, yohimbine, and clonidine have been used in CRPS with equivocal results. The use of propranolol has been abandoned. A calcium channel blocker, ziconotide, administered intrathecally has been used in CRPS. There is little compelling scientific evidence for sodium channel blocker antiepileptic agents in CRPS. Effective doses of IV lidocaine may be difficult to achieve due to development of adverse cardiovascular effects. Anti-TNF (e.g. infliximab), thalidomide, and lenalidomide need additional double-blind, controlled trials. Other agents are listed below.

SYSTEMIC INTRAVENOUS BISPHOSPHONATES
Alendronate, pamidronate, neridronate, etidronate, risedronate, ibandronate, and clodronate (also known as clodronate disodium, clodronic acid, or dichloromethylene bisphosphonate) are classed as bisphosphonates. These agents may be capable of relieving pain by modulation of nociceptive primary afferents in bone and pain-associated changes in the spinal cord (Schott, 1997 and 2001). Although bisphosphonates are promoted for systemic IV treatment of CRPS, their use in CRPS has not been approved by the FDA. One high-quality RCT (Varenna et al., 2000) found a significant improvement in pain reduction in patients receiving IV clodronate, compared to placebo. Another RCT (Adami et al., 1997) found that IV alendronate significantly improved pain, compared to placebo.

More recent studies of the use of bisphosphonates in CRPS have been performed. Verenna et al. (2013) performed a randomized, double-blind, placebo-controlled clinical trial of IV neridronate in 82 individuals with CRPS 1 in whom significant improvements in the Visual Analogue Scale (VAS) versus placebo were reported. However, this drug is approved only for treatment of osteogenesis imperfecta and only in Italy. O’Connell et al. (2013) found low-quality evidence that bisphosphonates may effectively reduce pain when compared with placebo, at least in the short term. They concluded that although the efficacy of bisphosphonates was promising, it has not been proven efficacious with confidence and warrants further investigation. Although the UK Guidelines found strong evidence of efficacy of IV alendronate, IV pamidronate, and IV clodronate in CRPS, none of these drugs are approved for use in either the UK or the US (Goebel et al., 2012). Trials of bisphosphonates in CRPS 1 have demonstrated the potential to reduce pain associated with bone loss. Most studies found improvement in pain symptoms and increased functionality. However, because of small sample sizes, more studies are needed to recommend bisphosphonates for CRPS treatment (Kosharskyy et al., 2013).

In a comprehensive review of CRPS, Borchers and Gershwin (2014) note that there is convincing evidence that bisphosphonates can significantly relieve spontaneous and stimulus-evoked pain and improve function in patients with early disease. According to Wheeler (2014), IV clodronate and alendronate have been shown to significantly improve pain, swelling, and ROM in individuals with acute CRPS. According to Varenna et al. (2014), studies show that bisphosphonates may be effective in the early stages of CRPS. However, according to Casale et al. (2015), only a few trials are available that would allow definitive conclusions.

A study by Eun Young et al. (2016) compared the effectiveness of IV pamidronate and oral prednisolone in 21 patients with CRPS subsequent to hemiplegic strokes. Patients were randomly assigned to pamidronate (n=11) or prednisolone (n=10). Subjective pain and hand edema (circumference of the middle finger and the wrist were measured at baseline and at 1, 2, and 4 weeks following the end of treatment. Both groups showed significant improvement in pain scores that were maintained to the 4-week period. Finger edema reduction was maintained at 4 weeks for those in the steroid group. Wrist edema reduction was maintained for 4 weeks in those on pamidronate. The authors noted the need for larger, controlled longer term studies to validate the findings. An editorial discussing the study indicated that for now, pamidronate appears to be the drug of choice in both posttraumatic and poststroke CRPS type 1 (Van Daele, 2016). Additionally, according to Frediani and Bertoldi (2015), IV clodronate has been used off-label for decades for CRPS type 1 with doses that range from 3 to 5 grams.

Therefore, for IV bisphosphonates in the treatment of CRPS, there is limited evidence that they improve the net health outcome.

SYSTEMIC INTRAVENOUS BUPIVACANE
No data were found for IV bupivacaine treatment of CRPS.

SYSTEMIC INTRAVENOUS DEXMEDETOMIDINE
Clinical investigations of dexmedetomidine for the treatment of CRPS are limited. according to Kosharskyy et al. (2013), further studies are needed to assess the efficacy of dexmedetomidine for treatment of chronic pain. Dexmedetomidine has not been approved by the FDA for treatment of CRPS.

SYSTEMIC INTRAVENOUS DIMETHYLSULFOXIDE (DMSO)
A number of free radicals have been proposed for use in CRPS including DMSO, but no convincing evidence is available. DMSO has been FDA approved only for interstitial cystitis (Casale et al., 2015).

SYSTEMIC INTRAVENOUS IMMUNOGLOBULIN G
In a double-blind placebo-controlled trial, ​IV​ immunoglobulin G (IVIG) was provided to a randomly assigned group of 13 patients who had refractory CRPS for 6 to 30 months. Pain levels were greater than 4 on a 0 to 10 rating scale. With IVIG treatment, the average pain intensity was 1.55 units lower than with placebo treatment (P=0.001). Despite these results, the researchers acknowledge the need for studies to determine best dose, duration of effect, and frequency of treatment (Goebel et al., 2010).

In their systematic reviews, both Cossins et al. (2013) and Goebel and Blaes (2012) found limited evidence for the efficacy of low dose IVIG. An opinion article suggested that immune mechanisms may be involved in long-standing pain, and IVIG may moderate pain by reducing immune activation (Gonzales et al., 2012). However, no data were presented to support the theory.

Like ketamine, the NINDS considers IVIG to be an emerging treatment for CRPS (NINDS, 2017). The authors cited a small study (Gobel et al., 2014, previously reported) and note that a larger study involving individuals with acute phase CRPS is planned. IVIG has not been FDA approved for treatment of CRPS.

In a review of data, Abdi (2015) continues to classify IVIG as an experimental approach to the treatment of CRPS. Gierthmuhlen et al. (2014) noted two aspects of IVIG in the treatment of CRPS: 1.) CRPS may have an autoimmunity component; and 2.) there have been observations of pain relief with IVIG in CRPS. Despite these points, the relevance of these findings remain unclear. Furthermore, per Gierthmuhlen et al. (2014) and Freedman et al. (2014), there is a scarcity of data; thus, further research is necessary.

In addition, IVIG has not received FDA approval for use in CRPS.

SYSTEMIC INTRAVENOUS KETAMINE
In a small observational study, ketamine was titrated from 10 mg/hr to a maximum dose of 40 mg/hr to achieve comfort without evidence of significant side effects. Infusions were provided to six patients with CRPS over a 5-day period. Patients were admitted to a monitored telemetry unit and maintained on their usual medication during the infusion period. Daily pain assessments and ketamine blood levels were collected. Minimal pain relief was observed on day 1, but significant pain relief (P<0.05) was observed by day 3, compared to baseline, and continued throughout the 5-day period. However, uniformity of relief was not achieved. On day 5, there was little or no change in the pain measure assessed as the worst pain experienced over the last 24 hours in 6 of the 16 patients. In this study, there was no follow-up after the 5-day period (Goldberg et al., 2010).

Twenty patients were infused with ketamine in anesthetic doses over 5 days. Significant pain relief was observed at 1, 3, and 6 months following treatment. Complete remission from CRPS was observed at 1 month in all patients, at 3 months in 17 patients, and at 6 months in 16 patients. Despite suggested benefit in relief of CRPS symptoms, the authors state that RCTs are needed to prove efficacy (Kiefer et al., 2008).

Sigtermans et al. (2009) enrolled 60 patients in a double-blind randomized placebo-controlled trial. Patients were admitted to a short-stay inpatient facility where they received either ketamine or saline infused over 4 days. Ketamine was titrated at regular intervals to a maximum of 30 mg/hr for a 70-kg patient. The infusion rate was increased when pain relief was insufficient. Liver function (daily) and blood pressure (three times daily) were measured. Pain scores over the 12-week monitoring period were significantly lower than those in patients receiving placebo (P<0.001); however, the significance was lost between the groups at week 12 (P=0.07). Patients receiving the drug experienced more psychomimetic side effects than controls (93% vs. 17%; P<0.001). In fact, patient and investigator guesses of administered treatment were correct in 74% and 88%, respectively, most likely due to psychomimetic side effects. Functional improvement did not occur in either group. The authors suggested that 4-day treatment with low-dose ketamine is safe, with psychomimetic side effects that were acceptable to most patients (Sigtermans et al., 2009).

Schwartzman et al. (2009) described a double-blind placebo-controlled outpatient trial that randomly assigned 19 patients with refractory CRPS into a ketamine group (n=9) or a placebo group (n=10). All subjects were infused with 100 mL of normal saline with or without ketamine for 4 hours (25 mL/h) daily for 10 days (5 days on, 2 days off, 5 days on). The maximum ketamine rate was 0.35 mg/kg/h not to exceed 25 mg/h. Subjects in both arms received clonidine and versed. Patients were seen at 2 weeks and then monthly for the following 3 months. Following treatment, the ketamine group showed a 21.4% reduction in pain scores (P< 0.01), while the placebo group demonstrated a nonsignificant 3.1% reduction (P>0.05). There was no change in patient activity in either the pre- or posttreatment phase. Side effects were described as nausea, headache, tiredness, or dysphoria in four of nine patients in the ketamine group and in two of 10 patients in the placebo group. The authors suggest that the lack of any psychomimetic side effects may be caused by the addition of midazolam and clonidine. This trial was conducted with multiday subanesthetic doses of ketamine, although the authors note (citing the study by Kiefer et al. [2009]) that in their experience only the 5-day IV ketamine regimen at anesthetic doses with midazolam and clonidine provide complete remission of CRPS symptoms lasting over 5 years. In fact, this trial was prematurely terminated because during the 2-year trial period the researchers found that doses of 50 mg/h (200 mg over a 4-hour period) provided much greater pain relief lasting for a longer period of time without complications. Therefore, they did not want to continue treating patients at the low ketamine dose (Schwartzman et al., 2009).

In a commentary on the previous two studies involving ketamine, Bell and Moore (2010) question the safety of repeated infusions of ketamine should pain return. Furthermore, they note that not only do safety questions remain, but optimal ketamine treatment regimens for chronic pain have yet to be established (Bell and Moore, 2010).

A retrospective study (without a control group) of ketamine infusions for refractory chronic pain was conducted at the University of Chicago Medical Center (Patil et al., 2013). Forty-nine patients underwent 369 infusions. Of these patients, 18 (37%) had a diagnosis of CRPS, eight had refractory headaches, and seven severe back pain. All patients reported significant reduction in VAS scores and reportedly, up to half of the patients experienced pain relief for up to 3 weeks.

Cossins et al. (2013) found moderate evidence for low-dose ketamine, but, according to the author, this evidence becomes compromised by the findings by Noppers et al. (2011), reporting liver failure after prolonged or repeated treatment. O’Connell et al. (2013) found low-quality evidence that IV ketamine may effectively reduce pain. Moreover, it is associated with many side effects, and, in the studies cited, the effect was not sustained beyond 4 to 11 weeks posttreatment.

The NINDS considers ketamine an emerging treatment for CRPS. The authors noted that investigators are using low doses of ketamine to either reduce substantially, or eliminate, the chronic pain of CRPS. “In certain clinical settings, ketamine has been shown to be useful in treating pain that does not respond well to other treatments.” It should be noted that the information on its website does not represent endorsement of the treatments or an official position by the NINDS (2022).

According to Kosharskyy et al. (2013) in a literature review, there are obstacles to the use of ketamine for chronic pain, including low oral bioavailability, a lack of an easily available formulation for chronic delivery, concerns over psychomimetic side effects, and mixed efficacy in clinical trials. The authors concluded that the use of ketamine infusions for the treatment of CRPS shows promise, but further prospective, randomized, double-blind placebo-controlled studies of anesthetic and subanesthetic doses of ketamine are needed.

In a systematic literature review specific to ketamine treatment for CRPS, Azari et al. (2012) concluded that the drug, while promising, needs well-designed clinical trials to determine the optimum dose, route, and timing of administration. The efficacy, safety, and long-term benefit also need to be demonstrated.

Wheeler (2014) indicates that while the rationale for ketamine for CRPS seems reasonable, no studies have shown benefit using objective outcome measures with double-blind randomized controlled methodology. In addition, several research questions remain to be settled. Birklein et al. (2015) reported that that currently the only medical treatment that is effective against chronic CRPS pain is ketamine infusions; however, the patient's function is not improved and the infusions have side effects.

An UpToDate review (Abdi, 2015) continues to classify ketamine as an experimental approach to treatment of CRPS. Additionally, researchers indicate that overall evidence for IV ketamine treatment in CRPS is limited and adverse effects restrict its usage. Further studies are necessary to assess efficacy and risk-benefit ratio (Casale et al. 2015; Gierthmuhlen et al., 2014). Ketamine has not been FDA approved for treatment of CRPS.

An additional study from Drexel University College of Medicine noted that poor responders to ketamine treatment had a lower body mass index (BMI) than responders. The researchers investigated the mechanisms underlying lower BMI that characterizes CRPS patients who respond poorly to IV ketamine therapy. They also sought to identify potential biomarkers for predicting response. Regulation of proopiomelanocortin (POMC) expression is crucial in normal body weight homeostasis. The authors found that although ketamine treatment did not alter POMC expression, poor responders had higher levels of POMC mRNA than responders. There was a positive correlation between the pretreatment levels of miR-34a to BMI and response to therapy. Larger studies are required to confirm the findings (Shenoda et al., 2016).

The above authors also investigated treatment-induced circulating microRNA as potential biomarkers. According to the authors, differences in miRNA signature in responders and poor responders before and after therapy indicate the prognostic value of miRNA response to IV ketamine (Douglas et al., 2015).

An observational longitudinal cohort study was conducted to assess the efficacy of subanesthetic ketamine infusions on children with chronic pain. Sixty-three individuals who received infusions were included. Of these, 23 had CRPS (37%). Ketamine infusions at doses of 0.1 to 0.3 mg/kg/h lasted for 4 to 8 hours per day up to a maximum of 16 hours over a maximum of 3 consecutive days. Pain scores after each infusion were significantly reduced from pretreatment scores (P<0.001). Greater reductions occurred in patients with CRPS than in patients with other chronic pain syndromes. Ketamine infusions did not change oral morphine intake compared to baseline doses. No follow-up except after treatment days was reported. Effect on function was not reported. The authors note the need for further studies to assess the optimum dose and long-term effect of ketamine on CRPS in children and adolescents (Sheehy et al., 2015).

A study to assess the effects of long-term ketamine treatment on cognitive function was reported by Kim et al. in 2016. Thirty CRPS patients were divided into a long-term frequent ketamine treatment group (n=14) and a non–long-term group (n=16). The participants completed questionnaires including demographic and clinical characteristics and variables affecting cognitive function. They also performed neuropsychological tests. Patients who received long-term ketamine treatment showed impairment in cognitive function—specifically in executive function compared with the non–long-term individuals.

Ketamine in anesthetic dosage was administered to five patients with CRPS of a mean of 8 years duration. The drug was administered over a 10-day period and patients received one to five doses. Pain reduction began on the fourth and fifth day of treatment. No improvement was noted in function. Pain reduction lasted 1.5 to 2.5 months following treatment, then relapsed to baseline level. The authors observed the short-term analgesic effect of the therapy but also noted the lack of effect on movement or function of the affected limbs (Puchalski and Zyluk, 2016).

Wheeler (2015) observed that the optimal dosing and duration of infusions of ketamine is unknown. These and other questions have yet to be answered. In addition, studies to date have not validated the benefit of ketamine using objective outcome parameters with double-blind randomized controlled methodology.

SYSTEMIC INTRAVENOUS LIDOCAINE
A study of lidocaine in CRPS, used as a continuous IV infusion in escalating doses, was reported. The treatments were administered to 49 CRPS patients over 5 days. Pain parameters were assessed at 1, 3, and 6 months following therapy. The researchers reported significant reduction in pain that lasted an average of 3 months (Schwartzman et al., 2009).

When IV lidocaine was compared with lidocaine used in a lumbar sympathetic block, IV lidocaine failed to produce significant changes in spontaneous and evoked pain intensity measurement compared to pretreatment values (Meier et al 2009).

There are some uncontrolled studies of lidocaine in neuropathic pain, but few address IV lidocaine in CRPS (Mackey S 2007; Carroll and Younger 2010; Challapalli et al 2005). One recent small study found benefit from lidocaine infusions for CRPS (Schwartzmann et al 2009), while another found no significant changes in pain resulting from IV lidocaine use (Meier et al., 2009). Additionally, there is no consensus for this treatment, and long-term response is unknown.

According to Harden et al. (2013), lidocaine infusions have fallen out of favor and are lacking evidence of efficacy. Kosharskyy et al. (2013) performed a comprehensive review and noted that the best results were reported by Schwartzman et al. (2009), but this study was small and nonrandomized. They therefore concluded that additional studies were needed to confirm the results. O’Connell et al. (2013) and Gierthmuhlen et al. (2014) found low-quality evidence (based on one study) that high-dose IV lidocaine may have a small effect on pain when compared with diphenhydramine. Casale et al. (2015) notes that systemic lidocaine is only therapeutic in cases of cold-induced allodynia. However, its channel blocking activity is unspecific and effective doses may be difficult to achieve due to adverse cardiovascular effects.

Lidocaine is not FDA approved for IV infusion for treatment of CRPS.

SYSTEMIC INTRAVENOUS MAGNESIUM
In a pilot study, IV magnesium was provided to eight patients, while two patients received normal saline as placebo. The placebo results were not analyzed or reported. At follow-up of 12 weeks, pain was significantly reduced from baseline. Impairment level and quality of life also improved, while there was no difference in skin sensitivity or functional limitations (Collins et al., 2009). This is a small study whose results need to be confirmed in a large, well-designed trial. In a 2015 review, Casale et al. indicated that IV magnesium can be considered a potential nonpharmacological supplement because it has fewer side effects than other treatments; however, further studies are necessary to identify the best route of administration. No additional studies of magnesium for treatment of CRPS have been performed. The drug is not FDA approved for CRPS.

SYSTEMIC INTRAVENOUS MANNITOL
It has been proposed that free radical scavengers may have a role in curtailing the CRPS disease process. Perez et al. (2008) conducted a randomized, placebo-controlled, double-blind trial involving 41 patients that compared IV mannitol in normal saline, infused in 4 hours over 5 days, with equal volumes of normal saline as placebo. Both groups received physical therapy. Pain was monitored using the VAS during the trial. Impairment and disability levels and quality of life were assessed up to 9 weeks. At the end of the study, no significant differences were found between mannitol and placebo treatment.

A small study compared treatment strategies in four groups of CRPS type I individuals (Lee et al., 2012). Group A was administered an NSAID (n=10); group B, oral gabapentin (n=12); group C, IV mannitol 10% and a steroid (n=11); group D, IV mannitol, steroid, and oral gabapentin (n=26). The best results occurred in group D, which showed recovery of grip strength and improvement in pain level and finger ROM. This is a small study that needs replication. Moderate evidence of nonefficacy was found in the UK review (Goebel et al., 2012). O’Connell et al. (2013) found very low-quality evidence that IV mannitol is not effective.

IV mannitol has not been FDA approved for treatment of CRPS.

SYSTEMIC INTRAVENOUS OPIOIDS
According to Freedman et al. (2014), opioids for chronic benign pain syndromes remain controversial secondary to potential for abuse, diversion, and overdoses leading to death. Recent reviews found in Casale et al. (2015) reported that opioids are not recommended as systemic treatment for CRPS. Although it is generally recognized that opioids have little or no effect on chronic nerve pain, there have been some reports of pain reduction and improved quality of life in individuals with chronic nerve pain. However, no controlled studies have been performed to demonstrate this finding.

INTRATHECAL AGENTS

INTRATHECAL BACLOFEN
A search of the literature for studies of intrathecal baclofen (ITB) in CRPS revealed one small randomized crossover trial of seven patients (van Hilten et al., 2000), suggesting that either continuous therapy or bolus injections of the agent might be effective in the treatment of dystonia in RSD. Therefore, there is limited evidence suggesting that ITB in CRPS improves the net health outcome. However, ITB has not been approved by the FDA for use in CRPS.

ITB was evaluated for treatment of dystonia in CRPS. Thirty-eight patients met the responder criteria, and 36 of those received a pump for continuous ITB administration in a dose-escalation study. Dose effect of baclofen on dystonia severity was shown in 31 patients with doses up to 450 mcg/day. One patient did not respond, and three patients dropped out. Thirty-six patients entered an open-label study. Intention-to-treat analysis at 12 months found improvement in dystonia, pain, disability, and quality of life. Eighty-nine adverse events occurred in 26 patients. Of these, 19 were directly related to baclofen, 52 to pump defects, and 18 to unspecified problems. The pump was explanted in six patients during the follow-up phase. Although CRPS symptoms improved and ITB remained efficacious over a period of 1 year, the authors noted the problems with high complication rate and suggested that methods to improve patient selection and pump integrity are warranted (Van Rijn et al., 2009).

A study to assess the effect of varying the infusion rate on the efficacy and safety of ITB delivery was reported (van der Plas et al., 2011). Individuals with CRPS-related dystonia who had no beneficial response to a minimum dose of baclofen, or who were intolerant of dose escalation because of side effects, were randomly assigned to slower infusion rate delivery (SIRD) or four times faster infusion rate delivery (FIRD). There were no significant differences between the FIRD and the SIRD groups for the primary outcomes of dystonia and pain. The same researchers in a later open study found significant improvement in global intense pain, sharp pain, dull pain, and deep pain during the first 6 months in a study of individuals with CRPS-related dystonia receiving titrated doses of ITB. After this period, pain scores leveled off despite further improvement of dystonia and continued ITB titration (van der Plas et al., 2013).

ITB is advised for individuals who display a dystonic component to CRPS. Harden et al. (2013) reported level 3 evidence for ITB in dystonic CRPS. According to Harden et al. (2013), if oral baclofen is effective but poorly tolerated, administration by intrathecal pump is a treatment option. Although O’Connell et al. (2013) concluded that ITB infusion needs further study, Goebel et al. (2012) recommends using ITB if other options have failed.

INTRATHECAL CLONIDINE
Like epidural clonidine, intrathecal clonidine infusion requires additional investigation (Tran et al., 2010). A small study showed a 30% or greater reduction in pain 2 hours after injection in 10 out of 22 individuals. However, the change in pain report did not correlate with the percentage change in areas of hyperalgesia (P= 0.09) or allodynia (P=0.24). The authors note that analgesia does not parallel antihyperalgesia with these treatments (Rauck et al., 2015.) No other recent studies were found.

INTRATHECAL GLYCINE
Intrathecal glycine was investigated as a potential therapy for pain and movement disorders in a randomized double-blind placebo-controlled trial in 19 patients. Over a 4-week period, no significant differences were found between the treatment and control groups (Munts et al., 2009).

INTRATHECAL METHYLPREDNISONE
A double-blind, randomized, placebo-controlled parallel-group trial was reported using a single intrathecal administration of methylprednisolone. The trial was stopped prematurely at 6 weeks when no difference was found between the treatment and control group (Munts et al., 2010; Gierthmuhlen et al., 2014). A 2014 review article (Rijsdijk et al., 2014) comments that the reports of the efficacy of intrathecal methylprednisolone are contradictory and that its safety is debated.

INTRATHECAL OPIOIDS
Patients with intractable malignant and nonmalignant pain were treated with intrathecal opioid infusion with an implantable pump. At follow-up of 3 years, reduction of nonmalignant pain on the VAS was reduced by more than 50% in 71.3% of patients. According to the authors, the treatment should be provided only in specialized centers (Koulousakis et al., 2007). According to Singh et al. (2019), a morphine pump should be carefully considered for chronic pain of nonmalignant origin. Other studies recommend intrathecal opioid administration as an option for nonmalignant pain, including CRPS, which can significantly improve quality of life in selected patients (Koulousakis et al., 2007; Singh, Nelson 2006). A Consensus Guideline (Deer et al., 2010) published an algorithm for intrathecal therapies that advised opioids and/or ziconotide be first-line treatments for CRPS. According to the NINDS, intrathecal infusion requires lower doses than that required for oral administration, therefore decreasing side effects and increasing effectiveness.

INTRATHECAL OPIOIDS IN COMBINATION WITH BUPIVACAINE
In small retrospective studies, bupivacaine was found to be stable in an implantable infusion system (Hildebrand et al., 2001; Goucke et al., 2010). However, clinical outcomes were mixed. In retrospective studies to assess the efficacy of bupivacaine when added to intrathecal opioids for treatment of nonmalignant pain, a significant improvement in paid reduction, disability and quality of life scores was reported (Kumar et al., 2009). In a larger retrospective study (n=126), a significant reduction in opioid dose escalation was demonstrated when bupivacaine combined with opioids was compared with opioids alone (Veizi et al., 2011). It should be noted that of the 143 subjects enrolled in these trials, only 12 were treated for CRPS. The only randomized trial of opioids alone or opioids with bupivacaine found the addition of bupivacaine to intrathecal opioids failed to produce improvement in pain control, and the researchers concluded that intrathecal mixtures of opioids and bupivacaine are not efficacious in the treatment of chronic nonmalignant pain. There is a need for additional large randomized trials to define the use of intrathecal bupivacaine, including long-term toxicity and neuropathology, particularly when used with other agents. Also unknown is the type of pain for which intrathecal bupivacaine is most efficacious (Deer et al., 2002; Kumar et al., 2009; Veizi et al., 2011; Mironer et al., 2002). Bupivacaine is not approved by the FDA for intrathecal administration.

INTRATHECAL OPIOIDS IN COMBINATION WITH LIDOCAINE
There is a lack of scientific evidence of positive impact on health outcomes with intrathecal lidocaine/opioid treatment in CRPS. Lidocaine is not FDA approved for intrathecal infusion.

INTRATHECAL ZICONOTIDE (PRIALT®)
A randomized, placebo-controlled trial with gradually increasing dosage over a 3-week period showed a statistically significant improvement in the Visual Analogue Scale of Pain Intensity (VASPI) with ziconotide (Prialt®) versus placebo. In addition, an open-label, multicenter study enrolled 644 patients with chronic pain in an evaluation of ziconotide (Prialt®) long-term use. Of patients with VASPI scores of 50 mm or greater at baseline who completed 1 month of therapy, 32.7% had greater than 30% improvement in VASPI scores. Also recorded was the pain impact on daily life, which significantly differed between baseline and month two.

INTRAMUSCULAR AGENTS

INTRAMUSCULAR BOTULINUM
A search of the literature for RCTs related to intramuscular botulinum in CRPS produced one record: a small observational study of 11 patients with CRPS type 1 affecting one upper extremity. The study suggests that botulinum A toxin may provide analgesic effects for several neuropathic pain states, including CRPS type 1, spinal cord injury pain, postherpetic neuralgia, and pain associated with brachial plexopathies (Argoff, 2002). A retrospective study of intramuscular botulinum toxin in 37 CRPS patients with spasm/dystonia in the neck and/or upper limb girdle muscles was reported (Kharkar et al., 2011). Ninety-seven percent of patients had significant pain relief with mean pain score decreases of 43%. These results were reported only in the short-term, up to 4 weeks. As noted by the authors, this is a study that lacks a control group, so the placebo effect cannot be discounted. Also, according to the UK Guidelines, evidence for regional botulinum toxin for CRPS-related dystonia is poor (Goebel et al., 2012). A comprehensive review (O’Connell et al., 2013), reported very low-quality evidence that sympathetic block using botulinum toxin A with a local anesthetic may effectively increase the duration of analgesia compared to local anesthetic alone by approximately 2 months. However the authors were unclear regarding the level of pain relief that might be achieved by this intervention. In 2014, Brown et al. conducted a review to summarize the highest quality literature pertaining to application of botulinum toxin in neuropathic pain conditions, including CRPS. They found level U (insufficient) evidence in CRPS.

Therefore, for intramuscular botulinum toxin A in CRPS, there is limited evidence that it improves the net health outcome. Intramuscular botulinum has not been approved by the FDA for use in CRPS.

INTRAMUSCULAR KETAMINE
In a review of topical and peripheral ketamine, Sawynok (2014) noted that peripheral administration of ketamine by localized injection produced some alterations in sensory thresholds in experimental trials in CRPS subjects, but many variables remained unchanged. Intramuscular ketamine is not FDA approved for CRPS treatment. The lack of scientific evidence does not permit conclusions regarding net health outcomes.

INTRAMUSCULAR MAGNESIUM SULFATE
A double-blind, randomized placebo-controlled crossover study was performed to assess the effect and safety of intramuscular magnesium sulphate (IMMG) versus placebo in CRPS patients with dystonia (Van der Plas et al., 2013). Thirty patients were administered IMMG or placebo over 3 weeks. The primary outcome was the difference in change in the Burke-Fahn-Marsden dystonia scores between both interventions. Only 22 patients were available for analysis, which revealed no differences between IMMG and placebo. According to the investigators, there is insufficient support for new studies evaluating the efficacy of other routes of magnesium sulphate administration in CRPS-related dystonia. IMMG has not been approved by the FDA for use in CRPS.

EPIDURAL AGENTS

EPIDURAL CLONIDINE
Epidural infusion with an α2-adrenergic agonist such as clonidine has been proposed for the relief of purportedly “sympathetically maintained pain,” such as RSD (Rauck et al., 1993). This method of treatment was classified as a sympathetic blockade in the review by Forouzanfar et al. (2002). A single high-quality RCT (Rauck et al., 1993) in 27 patients found that epidural clonidine in patients diagnosed with RSD reduced pain significantly more than placebo. According to Harden et al. (2013), the effectiveness of epidural analgesia for treatment of CRPS has been demonstrated in several studies. The value of the treatment is to allow a more rigorous physical therapy program (Harden et al., 2013). van Elijs et al. (2011) suggests that epidural infusions “can be tried” for individuals refractory to other conventional treatment. The main limitation to continuous infusions is the high infection rate of indwelling lines which, according to Harden et al. (2013), needs to be defined by further prospective study on infusion techniques in CRPS patients. O’Connell et al. (2013) found very low quality evidence that clonidine provides relief of pain that is refractory to sympathetic blockade (one study), and recommends further study. According to Freedman et al. (2014), there is no convincing evidence for epidural clonidine to treat sympathetic pain in CRPS. Therefore, for epidural clonidine in CRPS, there is limited evidence that it improves the net health outcome. However, epidural clonidine has not been approved by the FDA for use in CRPS.

INTRATHORACIC MEDICATIONS

Intrathoracic (i.e., intrapleural or intrapulmonary) administration of nonopioid medication for CRPS that does not have a direct goal of chemical ablation of sympathetic ganglia or fibers, does not appear to be used.

In summary, despite optimism for the use of various parenterally administered nonopioid drugs for CRPS, there are no completed clinical trials on the efficacy of these treatments with consistently applied evidence-based medicine criteria. Most reports of trials in CRPS represent small anecdotal clinical studies with few experimental findings. Unquestionably, a consensus definition of CRPS, with standardized diagnostic criteria, is needed. Practical agreement about the minimal clinical criteria (signs and symptoms) that define CRPS does exist among neurologists, anesthesiologists, and others. Without a universally accepted definition and diagnostic criteria for CRPS, it is difficult to accurately identify CRPS patients, to select patients for clinical trials, to validate experimental human and animal model systems for research, and to formulate testable hypotheses (NINDS, 2001). Finally, none of the nonopioid agents proposed for the treatment of CRPS, and in particular none of the parenterally administered agents, has been approved by the FDA.

PLASMAPHERESIS

Plasmapheresis is a method of withdrawing blood and separating it into plasma and cells, and transfusing the cells back into the bloodstream. It is used to remove antibodies when treating autoimmune conditions. Although it is hypothesized that there is an autoimmune component to CRPS, no outcome data were found specific to CRPS treatment with plasmapheresis. The lack of scientific evidence does not permit conclusions regarding net health outcomes.

PLASMA EXCHANGE

Plasma exchange is an extracorporeal therapy that extracts the patient’s whole blood which is then separated into plasma and blood cells. The plasma is removed and replaced with another solution such as human albumin in saline or a specifically prepared donor plasma. The reconstituted plasma substitute along with the blood cells is then returned to the patient.

A study from Drexel University College of Medicine described a retrospective case series of patients with CRPS and a clinical presentation suggestive of a small fiber neuropathy (SFN) who were treated with plasma exchange (PE). The use of PE was based on a hypothesis proposing an autoimmune etiology for CRPS. Thirty-three patients who had not responded, or responded poorly, to their current treatment received a series of PEs (mean, 7.2 treatments) over a 2- to 3-week period. Three patients showed no improvement in pain with the treatment, and one patient reported short-term relief that returned to pretreatment level within 3 days. The remaining patients reported significant pain reduction that was either maintained with immune-modulating therapies of weekly PE (n=15) or oral immune-modulating therapy (n=8) or they slowly returned to pretreatment pain levels with no further therapy (n=6). The authors conceded that the study, because of its retrospective, nonrandomized, and uncontrolled nature, is limited. Moreover, because the patients described were only those with SFN, the results cannot be extrapolated to the entire CRPS population. They concluded that randomized, placebo-controlled studies may be required to confirm the results (Aradillas et al., 2015).