Low-Dose Intravenous Immunoglobulin Treatment for Long-Standing Complex Regional Pain Syndrome

Complex regional pain syndrome (CRPS) is a rare chronic pain condition (population prevalence, less than 1 in 2000) arising after trauma to distal limbs (1, 2). The diagnosis is clinical, based on the assessment of sensory, motor, and autonomic abnormalities in the affected limb (3). Most patients improve spontaneously; however, the 15% who still have symptoms 1 year after onset have a quality of life among the worst of patients with any medical condition, and their prognosis is poor (4, 5). Treatment with analgesic drugs, such as antidepressants, or with anticonvulsants rarely is effective (6). Recommended therapy is multidisciplinary care; however, many patients will not achieve pain relief (7). After a chance observation, we conducted a prospective open study and a small randomized crossover trial in which low-dose intravenous immunoglobulin (IVIg) substantially reduced pain in this patient group. In both studies, 25% of patients had profound pain relief of greater than 50% (8, 9). The phase 3 LIPS (Low-Dose Immunoglobulin in Long-Standing Complex Regional Pain Syndrome) randomized controlled trial (RCT) was conducted to confirm the efficacy of repeated-dose treatment with low-dose IVIg compared with placebo in a large group of patients with long-standing CRPS. The primary outcome was pain intensity measured daily for 6 weeks after infusion and was compared between the IVIg and placebo groups. Methods Design Overview In this parallel-group trial, patients with CRPS were randomly assigned in 1:1 allocation to receive either 2 infusions of IVIg, 0.5 g/kg of body weight, or placebo. All patients were offered an open-label extension of 2 IVIg infusions. Providers, researchers, and outcome assessors were blinded to treatment assignment. The East of England Welwyn committee gave ethics approval (reference: 12/EE/0164). Patients were provided with informational leaflets about the trial, and those interested in participating provided written informed consent. The study protocol has been published (10) and is available in Supplement 1. Supplement 1. Study Protocol Setting and Participants The study recruited patients from 7 secondary and tertiary care pain treatment centers in the United Kingdom; potential participants were identified from the internal databases of these study centers or were referred to the centers as new patients. To enhance recruitment, the study was promoted regularly throughout the United Kingdom in pain medicine journals, through letters to each specialist pain clinic, on social network sites, and within CRPS patient organizations. Eligible participants were nonpregnant adults with moderate or severe CRPS (on the basis of the Budapest research criteria [3]). The CRPS severity cutoff was concealed and determined by a mean pain intensity score of 5 points or higher on an 11-point (0- to 10-point) numeric rating scale (NRS) recorded in the first 7 daily pain diary entries during screening, with no single entry below 4 points. A pain intensity score of 4 points is considered a cut point between mild and moderate pain (11). The Budapest research criteria, used by all the recruitment centers, require the presence of at least 1 regional sign in at least 2 of the following 4 diagnostic categories: sensory abnormalities, such as allodynia; swelling or sweating; color or temperature changes; and motor or trophic changes. The criteria also require the report of symptoms in all 4 categories. Patients with CRPS type I (without nerve injury) or II (with nerve injury) were eligible to participate. Eligible patients had CRPS for 1 to 5 years and no other pain that, in the study physician's opinion, might interfere with their personal assessment of CRPS pain changes. Before enrollment, patients received tricyclic antidepressants, gabapentinoids, and mild and strong opioids, as well as specialized pain physiotherapy, unless these treatments were refused or contraindicated. Patients with an implanted spinal cord stimulator were eligible if they met pain intensity criteria with the stimulator turned on. Participants continued their usual exercise and medication regimens. Further details regarding inclusion and exclusion criteria are provided in the study protocol (Supplement 1) (10). After giving consent and being screened for eligibility, suitable patients completed a screening diary for 7 days, then were contacted by telephone to ascertain their diary values; those who met the pain eligibility criteria were randomly assigned to a study group 10 to 21 days after screening (that is, on day 0). Randomization and Interventions Participants were individually randomly assigned (1:1) to receive IVIg or placebo by site staff via an independent online system, using block randomization with randomly varying block sizes, stratified by study center. The intervention was blinded by preparation of the IVIg (0.5 g/kg) or placebo solution (0.1% albumin in normal saline) in bottles of identical appearance. Upon notification, nonblinded dispensing site pharmacists removed the bottle label indicating the trial group before dispensing the agents. All other study site staff, the trial manager or site monitor, the statistician, and the chief investigator remained blinded to the patients' treatment assignments until database lock. No participants required emergency unblinding. Blinded infusions were scheduled on days 1 and 22 after randomization. A predetermined time window around the infusion days provided flexibility (first infusion, up to 5 working days; second infusion, day 221 day). The primary outcome period, days 6 to 42 after randomization, remained fixed and was thus independent of the actual infusion dates. Patients who completed the blinded phase were offered open-label IVIg on days 43 and 64 after randomization. The dosages prescribed were within normal, weight-determined clinical limits (0.5 g/kg) for low-dose treatment. Outcomes and Follow-up The primary outcome measure was the average 24-hour pain intensity score on an 11-point NRS, with 0 designating no pain and 10 pain as bad as you can imagine. This outcome was measured daily from days 6 to 42, with this interval prespecified to exclude the period of early, unspecific, temporary pain increases, such as headaches (8). Secondary outcomes were pain interference, measured with the interference subscale of the Brief Pain Inventory (11-point NRS from 0 to 10 points), with higher scores indicating more interference (12), and quality of life (measured by the 5-level EuroQol 5-dimensional questionnaire [EQ-5D-5L]), where higher scores indicate a better quality of life (13). All other outcomes were exploratory. Paper diaries documenting the participants' average 24-hour pain intensity were self-administered from days 1 to 42 after randomization (an example diary is provided in Supplement 2). Patients who decided to have 2 open infusions after the blinded phase completed additional 24-hour diaries until day 84, then completed weekly pain diaries for 9 more weeks. The other patients completed weekly diaries for 3 weeks. Site staff contacted participants twice after each infusion to confirm their adherence to keeping the pain diaries and to document any adverse events. Supplement 2. Web Appendix At screening and day 43, patients completed questionnaires assessing their multidimensional pain experience. Multidimensional assessment tools were used, including those for pain interference, quality of life, and other factors, in line with consensus recommendations for pain trials (14); details are provided in Supplement 2 and the study protocol (Supplement 1) (10). At these 2 time points, we also measured skin temperature of both the CRPS-affected and contralateral limbs (see protocol in Supplement 1). At the screening visit, blood was drawn for safety analysis (full blood count and assessment of serum immunoglobulin, creatinine, urea, and electrolyte levels) and, if applicable, pregnancy tests were administered to determine each patient's eligibility. Reasons for withdrawal from randomly assigned treatment were reported at days 22 and 43 after randomization. Patients recorded adverse events and reactions in their diaries, and these notes were transcribed on days 22 and 43 after randomization. In addition, study nurses queried patients for adverse events by asking open-ended questions during scheduled telephone calls on days 2 and 5 after each infusion. A study physician rated the severity and causality of each event on categorical scales. Adverse events related to open-label infusions, reported from 43 to 85 days after randomization, were tabulated separately. Serious adverse events were monitored for 21 days after the final dose of IVIg (or placebo) or until resolution. Statistical Analysis The sample size was based on the following assumptions from a pilot study (8): 122 participants were needed to detect a clinically meaningful difference, on a group level (15), in a pain score of 1.2 points on the NRS, determined by using a 2-sample t test assuming 5% statistical significance, 85% power, and a common SD of 2.2 (as in the previous study). Assuming 10% loss to follow-up and 5% nonadherence increased this number to 152 participants. We intended to collect 37 measurements of pain intensity (the primary outcome) per participant and to analyze the outcome by using a mixed-effects regression model. Therefore, the sample size was reduced on the basis of these extra measurements. From the pilot study (8), the correlation between a patient's measures was assumed to be 0.7; hence, the multiplying factor was [1 + (371)0.7]/37= 0.71. Therefore, the total sample size required was calculated as 1520.71= 108 participants (54 participants per study group). All statistical analyses were conducted by using Stata, version 14 (StataCorp). The primary outcome was analyzed by using a random intercept mixed model (Stata: mixed) to establish any difference in pain scores between the IVIg and placebo gro