The Journal of Pain, Vol 15, No 12 (December), 2014: pp 1366-1376
`Available online at www.jpain.org and www.sciencedirect.com
`
`Lenalidomide for Complex Regional Pain Syndrome Type 1: Lack of
`Efficacy in a Phase II Randomized Study
`z
`Donald C. Manning,*,y
`{
`Guillermo Alexander,
`Alyse Cooper,
`Joseph C. Arezzo,
`jj
`Anne Louise Oaklander,**,yy
`zz
`R. Norman Harden,
`Srinivasa N. Raja,
`Richard Rauck,
`z
`and Robert Schwartzman
`*Department of Anesthesiology, University of Virginia Health System, Charlottesville, Virginia.
`y
`Adynxx Inc, San Francisco, California.
`z
`Department of Neurology, Drexel University, Philadelphia, Pennsylvania.
`x
`Department of Neuroscience and Neurology, Albert Einstein College of Medicine, Bronx, New York.
`{
`Celgene Corporation, Summit, New Jersey.
`jj
`Center for Pain Studies, Rehabilitation Institute of Chicago, Chicago, Illinois.
`yy
`**Departments of Neurology and
`Pathology (Neuropathology), Massachusetts General Hospital and Harvard
`Medical School, Boston, Massachusetts.
`zz
`Division of Pain Medicine, Johns Hopkins University, Baltimore, Maryland.
`xx
`Center for Clinical Research, Salem, North Carolina.
`
`xx
`
`x
`
`Abstract: Complex regional pain syndrome (CRPS) is a potentially debilitating chronic pain syndrome
`with a poorly understood but likely neuroimmune/multifactorial pathophysiology associated with
`axonal injury. Based on the potential contribution of proinflammatory cytokines to CRPS pathogenesis
`and prior research with thalidomide, we investigated lenalidomide, a thalidomide derivative, for CRPS
`treatment. We conducted a phase II, randomized, double-blind, placebo-controlled study to evaluate
`the efficacy of oral lenalidomide 10 mg once daily in consenting patients with unilateral or bilateral
`CRPS type 1. The study comprised 12 weeks of treatment followed by a long-term extension. The pri-
`mary efficacy outcome was reduced pain in the index limb, defined as $30% improvement from base-
`line using an 11-point numeric rating scale. One hundred eighty-four subjects enrolled. The primary
`endpoint was not met because equal proportions of treated (16.1%) and control (16.1%) subjects
`achieved the outcome; however, lenalidomide was well tolerated, with no evidence of neuropathy
`or major adverse effects. This study is the largest controlled, blinded clinical trial in subjects with chronic
`CRPS using the Budapest research criteria. It demonstrates the feasibility of conducting high-quality
`clinical trials in CRPS type 1 and provides considerations for designing future trials.
`Perspective: This article reports an adequately powered, controlled clinical trial in subjects with
`CRPS. Treatment and placebo were equally effective, but the study demonstrated that lenalidomide
`treatment is feasible in this population. The study provides examples to consider in designing future
`CRPS trials.
`ª 2014 by the American Pain Society
`
`Open access under CC BY-NC-ND license.
`
`Received February 19, 2014; Revised September 10, 2014; Accepted
`September 22, 2014.
`This study was sponsored by Celgene Corporation. Celgene developed
`the protocol in conjunction with the clinical investigators. Celgene was
`solely responsible for investigative site selection, data collection and stor-
`age methods and technology, clinical monitoring of protocol and Good
`Clinical Practice guidelines compliance, statistical analysis plan and
`execution, reporting of safety data to regulatory agencies, and the pro-
`vision of study drug and matching control. The authors received editorial
`support for manuscript preparation from Amy Zannikos, PharmD, of Pel-
`oton Advantage, LLC, funded by Celgene Corporation. The authors, how-
`ever, directed and are fully responsible for all content and editorial
`decisions for this manuscript. D.C.M. was an employee of Celgene Corpo-
`ration during the execution of the reported study, an owner of Celgene
`stock, and a former member of the Board of Directors for Reflex Sympa-
`thetic Dystrophy Syndrome Association-USA. G.A. received standard
`compensation for participating as an investigator in this study. J.C.A.
`received standard compensation for assisting with the design and anal-
`ysis of the electrophysiology data from this study. A.C. is an employee
`
`1366
`
`of Celgene Corporation. R.N.H. received standard compensation for
`participating as an investigator in this study and also had an unrestricted
`grant from Celgene at the time of this study. He also has received research
`support from the Reflex Sympathetic Dystrophy Syndrome Association.
`A.L.O. reports no conflicts of interest. S.N.R. received standard compen-
`sation for participating as an investigator in this study and reports a
`research grant from Medtronic Inc for studies on the mechanisms of spi-
`nal cord stimulation–induced analgesia. R.R. received standard compen-
`sation for participating as an investigator in this study. R.S. received
`standard compensation for participating as an investigator in this study.
`Trial registration: clinicaltrials.gov #NCT00109772.
`Address reprint requests to Donald C. Manning, MD, PhD, 16 Johnston
`Drive, Bloomsbury, NJ 08804. E-mail: donaldcmanning@gmail.com
`1526-5900/
`ª 2014 by the American Pain Society
`Open access under CC BY-NC-ND license.
`http://dx.doi.org/10.1016/j.jpain.2014.09.013
`
`IPR2018-00685
`Celgene Ex. 2016, Page 1
`
`

`

`Manning et al
`
`The Journal of Pain
`
`1367
`
`Key words: Complex regional pain syndrome, cytokines, immunomodulation, lenalidomide, thalido-
`mide, clinical trial.
`
`Complex regional pain syndrome (CRPS) is an often
`
`debilitating neuropathic/neuroimmune syndrome
`usually triggered by injury to a limb.7,13,34,42 Its
`clinical features can include pain symptoms such as
`allodynia and hyperalgesia; difficulty moving, including
`reduced range of motion; changes
`in limb color,
`temperature, appearance, and sweating; edema; and
`bone changes.15,17,25,38,40 CRPS affects more females
`than males (relative risk 2:1–4:1) and has an average
`age of onset of 37 to 60 years.3,8,33 Its pathophysiology
`is likely multifactorial,13 and far more patients have
`type 1, which often involves soft-tissue and small-fiber
`axonal injury,24 than type 2, which requires injury to a
`major nerve.15,22,24 Treatment guidelines based on
`limited available data and expert opinion recommend
`physical and occupational therapy, analgesic and other
`medications,
`interventional
`procedures,
`and
`therapy and support.15,27 Proposed
`psychological
`mechanisms include peripheral afferent and efferent as
`well as central neural involvement; clinical presentation
`suggests a neurogenic inflammation and/or activation
`of the immune system leading to an exaggerated
`inflammatory process. Regardless, there is increasing
`consensus that inflammation plays an important role in
`CRPS type 1.
`Increased levels of proinflammatory cytokines (tumor
`necrosis factor-a [TNF-a], interleukin [IL]-1b, IL-2, and
`IL-6) and anti-inflammatory cytokines (IL-4 and IL-10)
`have been identified in plasma and cerebrospinal fluid
`and within the affected limbs of CRPS patients compared
`with nonaffected controls.1,20,26,39 Increased levels of
`TNF-a and IL-6 have been identified in blister fluid from
`the skin of the affected limb when compared with the
`unaffected limb.16 The elevated levels of proinflamma-
`tory and anti-inflammatory cytokines found in CRPS sug-
`gest that immunomodulating agents might be useful
`because of their anti-inflammatory effects. Few well-
`controlled and adequately powered clinical treatment
`studies of anti-inflammatory agents have been per-
`formed, and existing study designs vary.9,14 Small pilot
`studies have also suggested possible therapeutic
`avenues, including a report of observable but variable
`resolution of CRPS in 2 patients treated with the TNF-a
`antagonist infliximab.21 Thalidomide is a relatively selec-
`tive inhibitor of TNF-a production by human monocytes,
`both in vitro and in vivo, and it exerts costimulatory
`effects on T-cell responses, including inhibiting proin-
`flammatory
`cytokines
`and
`stimulating
`anti-
`inflammatory cytokines.6 Following demonstration of
`near resolution of CRPS symptoms after thalidomide
`treatment in a woman with CRPS for 3 years,28 thalido-
`mide was studied in 42 patients with long-standing, pre-
`viously
`treated CRPS. Objective
`and subjective
`improvement was observed in 17% of patients, with
`14% experiencing at least modest pain relief.36 However,
`use of thalidomide is limited by its well-known neurotox-
`icity and teratogenicity. Teratogenicity is a particular
`concern in CRPS, because most patients are female.
`
`Lenalidomide is a thalidomide derivative created in a
`program to enhance the anti-inflammatory and immu-
`nomodulatory properties of thalidomide while reducing
`its toxicity. It is approved in the United States and Europe
`for treating multiple myeloma and anemia in some
`patients with myelodysplastic syndromes12,30 and in
`several other countries (eg, China, Japan) for treating
`multiple myeloma.4,37 Lenalidomide has approximately
`1,000 times more anti-inflammatory activity in vitro
`than thalidomide.6 Its pharmacologic activities include
`inhibiting secretion of the proinflammatory cytokines
`TNF-a, IL-1b, and IL-6, and increasing secretion of the
`anti-inflammatory cytokine IL-10 from stimulated mono-
`cytes.6,19 Lenalidomide’s safety and effect on reported
`pain in CRPS type 1 were studied in a preliminary phase
`IIa, multicenter, single-arm, open-label, 40-patient pilot
`study initiated in 2003.35 The good response rates ($2-
`point
`improvement
`from baseline on an 11-point
`numeric rating scale [NRS]) were 28.9% at week 12 and
`52.0% at week 52, and patient-rated pain and sleep
`scores also improved (P < .05).35 In addition, lenalido-
`mide was well tolerated, with few treatment-related
`serious adverse events (AEs) and no evidence of neuro-
`logic toxicity. These findings provided the rationale for
`the current phase IIb, multicenter, double-blind, pla-
`cebo-controlled study.
`
`Methods
`
`Subjects
`The inclusion criteria were age $18 years and CRPS
`type 1 (as per Budapest research criteria17) for $1 year
`with unilateral or bilateral involvement of a distal hand
`or foot, with or without proximal spread, plus a CRPS-
`related pain intensity score of $4 in the index limb. In
`cases where both limbs were affected, subjects selected
`their limb to be assessed (designated as the index
`limb). For subjects with single-limb involvement, that
`limb was the index limb. Because lenalidomide is struc-
`turally similar to thalidomide, a human teratogen, men
`and women with reproductive potential had to follow
`pregnancy prevention requirements,
`including coun-
`seling about preventing pregnancy. Subjects had to
`agree to use reliable forms of contraception, and women
`had to agree to periodic pregnancy testing. Potentially
`fertile women had to submit 2 negative pregnancy tests
`before randomization.
`Subjects were excluded for history of stroke or deep
`vein thrombosis in the past 5 years (based on experi-
`ence with lenalidomide in the oncology program);
`peripheral neuropathy; severe, progressive, or uncon-
`trolled renal, hepatic, hematologic, endocrine, pulmo-
`nary, cardiac, neurologic, or cerebral disease; any
`serious medical diagnoses or abnormal laboratory test
`results; or psychiatric
`illness
`that
`could prevent
`informed consent. Additional exclusions were prior
`treatment with lenalidomide or allergy to thalidomide;
`
`IPR2018-00685
`Celgene Ex. 2016, Page 2
`
`

`

`1368
`
`The Journal of Pain
`
`Lenalidomide in Complex Regional Pain Syndrome
`
`pregnancy or lactation; or use of concomitant medica-
`tions that increase the risk of deep vein thrombosis.
`Low-dose aspirin (81 mg) was required for subjects
`receiving sex hormones in any form for contraception
`or treatment of menopausal symptoms.
`
`Study Design
`This multicenter, phase IIb study was conducted from
`February 2005 to April 2008 at 24 study sites in the United
`States.
`It included a 2-week screening phase and a
`12-week double-blind treatment phase, followed by an
`optional extension phase that allowed treatment to
`continue as long as benefit was maintained (Fig 1). All
`subjects meeting CRPS-related morning and evening
`pain intensity numerical rating scale (PI-NRS) score re-
`quirements averaged over the 7-day period before
`randomization ($8 recorded scores, average score $4
`in affected limb) were randomized 1:1 to receive lenali-
`domide 10 mg orally once daily or placebo for up to
`12 weeks in the treatment phase. Concomitant therapy
`for CRPS pain (opioid and nonopioid analgesics, nonste-
`roidal anti-inflammatory drugs, anticonvulsants, antide-
`pressants, and other nondrug therapies) was permitted if
`the dose/regimen had been stable for $4 weeks before
`randomization and remained stable throughout the
`study. No new CRPS medications or nondrug therapies
`were allowed except for limited use of a rescue medica-
`tion (short-acting opioid for no more than 7 days) to
`treat pain flare, trauma, or invasive procedure. During
`the extension phase, subjects were permitted to initiate,
`reduce, increase, or withdraw from concomitant CRPS
`pain medications/nondrug therapies except for initiating
`experimental therapies.
`Randomization codes were generated by a Celgene
`statistician not involved in the study using a randomiza-
`tion program in the Statistical Analysis Software system.
`Randomization was performed centrally across all cen-
`ters using blocked randomization with a block size of
`4. Eligible patients were randomized in a 1:1 ratio to 1
`of the 2 treatment groups. Sites were provided with
`drug for 4 subjects at a time, and additional shipments
`were sent as required. Each qualifying subject was as-
`signed the next sequential number in the series of
`numbers assigned to the site. Code-break envelopes
`were provided in the boxes containing the drug, and
`site personnel were instructed to contact the Celgene
`Medical Monitor before breaking the code. The Celgene
`Clinical Supplies group worked with Xeremis, which
`packaged the drugs. All subjects completing the entire
`12-week double-blind phase were eligible for the exten-
`sion phase. Subjects first randomized to placebo were al-
`lowed to cross over to lenalidomide 10 mg/d and
`continue treatment in the extension phase.
`The study protocol, amendments, and informed con-
`sent form were approved by the institutional review
`board at each site, and the study was conducted in accor-
`dance with the International Conference on Harmonisa-
`tion Guidelines for Good Clinical Practice and the
`Declaration of Helsinki. All subjects provided written
`informed consent before initiating study procedures.
`
`Figure 1. Study design.
`
`Dose Selection
`The lenalidomide dose for continuous dosing in
`myelodysplastic syndrome starts at 10 mg/d and can be
`adjusted downward for renal function. For intermittent
`dosing, such as for multiple myeloma, the dose can start
`at 25 mg/d for a defined portion of the chemothera-
`peutic cycle. AEs in general were dose related. Prelimi-
`nary data from the open-label study in CRPS showed
`lenalidomide to be well tolerated with an expected AE
`profile. Subjects who responded to lenalidomide
`10 mg/d reported return of their CRPS if their dose was
`reduced to 5 mg/d because of AEs, and most chose to
`resume 10 mg/d to complete the study. On the basis of the
`requirement for continuous dosing as well as experience
`in other patient populations, lenalidomide 10 mg/d was
`selected as the study dose.
`
`Efficacy and Safety Assessments and
`Endpoints
`Subjects used an electronic diary to collect daily
`morning and evening efficacy assessment data,
`including the PI-NRS. The PI-NRS was presented twice
`per day at set consistent times (morning and evening)
`as a discrete 11-box Likert scale from 0 (no pain) to
`10 (worst pain imaginable) with the instructions
`‘‘Please rate your pain due to CRPS right now.’’ The pri-
`mary efficacy endpoint was the PI-NRS responder rate,
`with response defined as completion of the treatment
`phase with a $30% reduction from baseline (ie,
`improvement) in the 7-day averaged morning and eve-
`ning PI-NRS score for the week before the week 12
`assessment as compared with the 7-day averaged
`morning and evening PI-NRS scores during the baseline
`period. The secondary efficacy endpoints included
`change from baseline in the Short-Form McGill Pain
`Questionnaire total score and sensory and affective
`subscale scores, pain intensity ratings (combined morn-
`ing and evening assessments) using PI-NRS in the index
`limb, sleep rating (NRS), activity level rating (NRS), sub-
`ject’s assessment of CRPS signs and symptoms in the in-
`dex limb, allodynia NRS score, allodynia rating
`between the index limb and the normal limb at the
`end of week 12 (unilateral CRPS subjects only), and
`concomitant use of pain medications.
`Safety assessments included type, frequency, and
`severity of AEs; physical examination; vital signs; and
`laboratory test results. A battery of noninvasive nerve
`conduction studies were included as a safety measure
`to explore the possibility of
`lenalidomide-induced
`
`IPR2018-00685
`Celgene Ex. 2016, Page 3
`
`

`

`Manning et al
`
`peripheral neuropathy, and to evaluate whether the
`anti-inflammatory
`effects of
`lenalidomide
`could
`improve nerve conduction. The principal electrophysio-
`logic outcome was maximal nerve conduction velocity.
`Responses were recorded from sural sensory, median
`sensory, peroneal motor, and median motor nerves on
`2 occasions before the onset of dosing and 2 occasions
`at the end of the initial 12-week dosing period.
`
`Statistical Analysis
`When the responder definition of this study was
`applied to the data from the pilot study,35 the response
`rate was calculated to be 34%. Based on this, plus the
`19% response rate in patients with postherpetic neural-
`gia who received placebo in a published study,32
`response rates for the current study were hypothesized
`to be 15% for placebo and 35% for lenalidomide. Using
`these assumptions, a sample size of 83 subjects per group
`would have 80% power to detect a significant difference
`in PI-NRS score between groups using a 2-sided,
`continuity-corrected c2 test at the .05 significance level.
`The primary analysis used the modified intent-to-treat
`population (subjects who received at least 1 dose of study
`medication and had at least 1 postdose measurement).
`Differences in responder rates between groups were
`assessed using the Cochran-Mantel-Haenszel test. Sub-
`jects who discontinued the treatment phase prematurely,
`regardless of reason, were considered nonresponders. All
`comparisons used 2-tailed tests at the .05 level of signifi-
`cance. Summary statistics were provided for continuous
`outcome measures. Secondary endpoints were evaluated
`with an analysis of covariance model, which included
`terms of treatment, center, baseline score, concomitant
`CRPS medication change status, and treatment-by-
`baseline score. Analysis of covariance results presented
`here are for the observed cases subpopulation of the
`modified intent-to-treat population with no imputation
`for missing data, thus giving varying sample sizes,
`because not all data were available for all subjects.
`
`Results
`Among the 184 subjects randomized, 180 received $1
`dose of study medication (97.8%; the safety population)
`and had $1 postdose diary measurement (the modified
`intent-to-treat population). The 12-week, double-blind
`treatment phase was completed by 68 subjects who
`received lenalidomide (78.2%) and 79 who received
`placebo (84.9%; Fig 2). The primary reason for discontin-
`uation during this phase was AEs. These affected 14 sub-
`jects who received lenalidomide (16.1%) and 5 subjects
`who received placebo (5.4%). The mean duration of
`treatment in the double-blind treatment phase was
`similar between groups: 10.6 weeks with lenalidomide
`and 11.4 weeks with placebo. During the 12-week
`double-blind treatment phase, small numbers of subjects
`receiving lenalidomide (11; 12.6%) and receiving
`placebo (7; 7.5%) altered their concomitant CRPS medi-
`cation (including addition, discontinuation, or dosage
`change).
`
`The Journal of Pain
`
`1369
`
`One hundred forty-two subjects entered the extension
`phase (lenalidomide, n = 64; placebo, n = 78). Ninety-four
`subjects discontinued—among them, 45 receiving only
`lenalidomide (70.3%) and 49 receiving lenalidomide
`(62.8%) after crossing over from placebo. The primary
`reason for discontinuation among subjects receiving
`lenalidomide was lack of therapeutic effect (29.7%).
`Among placebo subjects who crossed over to lenalido-
`mide, it was AEs (24.4%). The mean duration of treat-
`ment in the extension phase was 43.8 weeks for
`subjects receiving lenalidomide and 46.2 weeks for pla-
`cebo subjects who crossed over to lenalidomide. The
`cumulative durations of treatment were 41.1 weeks
`and 49.9 weeks, respectively. The baseline demographics
`and characteristics of subjects are summarized in Table 1.
`The study was discontinued prematurely during the
`extension phase because unblinding of the data from
`the double-blind treatment phase revealed that the pri-
`mary efficacy endpoint had not been met. At week 12,
`16.1% of subjects receiving lenalidomide and 16.1%
`receiving placebo were considered responders based
`on changes in PI-NRS scores. Changes (morning assess-
`ments, evening assessments, and combined morning
`and evening assessments) from baseline to week 12 (7-
`day averaged scores) did not differ between groups
`(Table 2). In addition, there were no significant differ-
`ences between groups in change from baseline in the
`daily sleep assessment, Short-Form McGill Pain Question-
`naire, activity level rating, and allodynia NRS score at
`week 12 (Table 3; no adjustment for multiplicity was
`made, so P values are for reference only).
`To better understand the types of subjects in the
`responder and nonresponder categories, use of concom-
`itant drugs (Table 4), nondrug therapies (Table 5), and
`rescue medications (Table 6) was analyzed in relation to
`responder status for the lenalidomide and placebo treat-
`ment groups. There were no significant differences in
`these outcomes between the lenalidomide and placebo
`groups, with responders using only slightly fewer
`concomitant treatments, but this did not determine
`responder status. The vast majority of subjects either
`took no rescue medication or used the described
`opioid-rescue protocol. More subjects in the placebo
`group (n = 15) than lenalidomide-treated subjects
`(n = 3) used nonallowed rescue therapies.
`Given lenalidomide’s mechanism of action, we evalu-
`ated whether subjects with higher baseline scores for
`potential markers of inflammation or microvasculopathy
`(changes in skin temperature, color, or swelling) on the
`subject’s assessment of CRPS symptoms (Table 7) were
`more likely to respond. The placebo group had slightly
`lower assessment ratings for these symptom categories
`compared to the lenalidomide-treated group, but there
`were no appreciable differences between responders
`and nonresponders.
`
`Safety Outcomes
`During the double-blind treatment phase, at least 1 AE
`was reported by 81 subjects receiving lenalidomide
`(93.1%) and 76 subjects receiving placebo (81.7%).
`
`IPR2018-00685
`Celgene Ex. 2016, Page 4
`
`

`

`1370
`
`The Journal of Pain
`
`Lenalidomide in Complex Regional Pain Syndrome
`
`Figure 2. Subject disposition. *Defined as all subjects who took at least 1 dose of study drug.
`
`Cumulative data for the double-blind plus open-label
`treatment phases indicated that at least 1 AE was
`reported by 83 subjects receiving lenalidomide (95.4%)
`and 85 placebo subjects who crossed over to lenalido-
`mide (91.4%). The most common AEs are summarized
`in Table 8. In the double-blind phase, at least 1 serious
`AE was reported by 4 subjects receiving lenalidomide
`(4.6%)
`(ie, decreased potassium, pain,
`joint sprain,
`angioneurotic edema) and 6 receiving placebo (6.5%)
`(ie, increased blood pressure, neuropathic pain, vomit-
`ing, tremor, clostridium colitis, and respiratory failure);
`cumulative data indicated that $1 serious AE was
`reported in 12 subjects receiving lenalidomide (13.8%).
`This included, in addition to the above, joint sprain
`(n = 2), foot fracture, deep vein thrombosis, positive hu-
`man chorionic gonadotropin, false-positive pregnancy
`test, sick sinus syndrome, pregnancy (n = 2), neuropathic
`pain, spontaneous abortion, and arthritis. Sixteen pla-
`cebo subjects who crossed over
`to lenalidomide
`(17.2%) also reported accidental overdose, hip fracture,
`
`deep vein thrombosis (n = 2), bradycardia, thyroiditis, ad-
`renal insufficiency, cholecystitis, gallbladder obstruction,
`allergic alveolitis, hemolytic anemia, thrombotic throm-
`bocytopenic purpura, upper abdominal pain, vomiting,
`or endometriosis. Treatment-related AEs led to discon-
`tinuation in 18 subjects receiving lenalidomide (20.7%)
`and 5 receiving placebo (5.4%); cumulative data showed
`that treatment-related AEs led to discontinuation in 29
`subjects receiving lenalidomide (33.3%) and 29 placebo
`subjects who crossed over to lenalidomide (31.2%).
`Most AEs leading to discontinuation were grade 1
`(mild) or 2 (moderate). AEs led to dose reduction in 11
`subjects receiving lenalidomide (12.6%) and 2 receiving
`placebo (2.2%) during the double-blind phase. The
`most common AE leading to reduction of lenalidomide
`was rash (3.4% of subjects), and the 2 AEs in the placebo
`group were clostridium colitis and conjunctivitis. Few
`shifts in laboratory values from baseline were noted in
`the double-blind treatment phase and the extension
`phase; of those that were different, few were reported
`
`IPR2018-00685
`Celgene Ex. 2016, Page 5
`
`

`

`The Journal of Pain
`
`1371
`
`.27
`.28
`.26
`
`PVALUEy
`
`yNoadjustmentformultiplicitywasmade,becausethePvalueisforreferenceonly;thePvalueisfromtheanalysisofcovariancemodeladjustingforcenterandbaselinescore.
`*AdjustedmeanchangefrombaselineinPainIntensityRatings(observedcases)atweek12.Noimputationwasemployedforthesedata.Negativechangeindicatesimprovement.
`Abbreviations:M,mean;SD,standarddeviation.
`
`ADJUSTEDCHANGE*
`
`.51
`.36
`.43
`
`(n=87).4(1.5)
`(n=88).3(1.5)
`(n=88).4(1.5)
`
`CHANGE
`
`(n=87)6.7(2.3)
`(n=88)6.5(2.3)
`(n=88)6.6(2.3)
`
`(n=87)7.1(1.6)
`(n=88)6.9(1.7)
`(n=88)7.0(1.6)
`
`WEEK12
`
`BASELINE
`
`ADJUSTEDCHANGE*
`
`.78
`.63
`.71
`
`(n=80).7(1.7)
`(n=80).6(1.7)
`(n=80).7(1.7)
`
`CHANGE
`
`(n=80)6.6(2.1)
`(n=80)6.3(2.1)
`(n=80)6.5(2.1)
`
`(n=80)7.3(1.4)
`(n=80)6.9(1.5)
`(n=80)7.1(1.4)
`
`PM,M(SD)
`AM,M(SD)
`AM1PM,M(SD)
`
`WEEK12
`
`BASELINE
`
`PI-NRS
`
`PLACEBO
`
`LENALIDOMIDE
`
`Table2.CRPSPI-NRSSummaryofChangeFromBaselineatWeek12inObservedCases
`
`Manning et al
`Table 1. Subject Demographic and Baseline
`Characteristics
`
`CHARACTERISTIC
`
`Age, y, M (SD)
`Age distribution, n (%)
`#65 y
`>65 y
`Sex, n (%)
`Male
`Female
`Race/ethnicity, n (%)
`White
`Black
`Hispanic
`Asian/Pacific Islander
`Baseline CRPS PI-NRS
`score, M (SD)*
`Daily sleep assessment
`average score, M (SD)
`
`LENALIDOMIDE
`(N = 87)
`
`PLACEBO
`(N = 93)
`
`TOTAL
`(N = 180)
`
`43.9 (11.4)
`
`45.1 (11.1)
`
`44.5 (11.2)
`
`83 (95)
`4 (5)
`
`22 (25)
`65 (75)
`
`81 (93)
`3 (3)
`1 (1)
`2 (2)
`7.2 (1.4)
`(n = 85)
`5.4 (1.1)
`(n = 85)
`
`89 (96)
`4 (4)
`
`14 (15)
`79 (85)
`
`86 (93)
`4 (4)
`3 (3)
`0 (0)
`7.0 (1.7)
`(n = 92)
`5.5 (1.3)
`(n = 92)
`
`172 (96)
`8 (4)
`
`36 (20)
`144 (80)
`
`167 (93)
`7 (4)
`4 (2)
`2 (1)
`–
`
`–
`
`Abbreviations: M, mean; SD, standard deviation.
`*Combined morning and evening assessments, based on a 0–10 scale, with
`higher ratings indicating more severe pain.
`
`as AEs. Likewise, no notable shifts in thyroid function or
`vital sign parameters (diastolic and systolic blood pres-
`sure, pulse, temperature) were reported.
`The pattern and variance of the electrophysiologic
`measures in the current study were consistent with previ-
`ous reports for multicenter clinical trials of pain.2 There
`were no significant differences in mean changes from
`baseline to study termination for the 2 sensory or the 2
`motor nerve velocities. Lenalidomide, at the doses and
`for the duration tested, was not associated with slowing
`of velocity, reduction in amplitude, or exacerbation of
`peripheral neuropathy.
`
`Discussion
`This phase IIb study evaluated the efficacy and safety of
`oral lenalidomide 10 mg once daily in subjects with CRPS
`type 1 based on pilot data indicating that lenalidomide
`had acceptable safety and some evidence of efficacy in
`CRPS.35 Although the current outcomes show no differ-
`ences in efficacy between lenalidomide and placebo for
`the primary and secondary endpoints, subjects tolerated
`daily lenalidomide for up to 2 years without significant
`safety issues, demonstrating that lenalidomide can be
`well tolerated in a population other than cancer patients.
`Importantly, all AEs reversed after lenalidomide was dis-
`continued, and neurophysiologic surveillance revealed
`no evidence of treatment-induced peripheral neuropa-
`thy, in distinction to the neuropathy associated with
`thalidomide use.28
`Both the pilot study and the phase IIb study helped
`identify design elements to consider for future CRPS tri-
`als. For efficiency purposes, we reevaluated the assess-
`ment tools and general tolerability measures used in
`the open-label pilot study, acknowledging that the po-
`tential for false positives was high, but the small sample
`ensured that fewer subjects would be exposed to the
`
`IPR2018-00685
`Celgene Ex. 2016, Page 6
`
`

`

`1372
`
`The Journal of Pain
`
`Table 3. Summary of Selected Secondary
`Efficacy Assessments
`
`Lenalidomide in Complex Regional Pain Syndrome
`Table 5. Concomitant Nondrug Therapy
`(Double-Blind Phase)
`
`MEAN ADJUSTED CHANGE
`FROM BASELINE AT WEEK 12
`
`Daily sleep assessment
`
`SF-MPQ
`Total
`Sensory
`Affective
`Activity level rating (NRS)
`
`Allodynia NRS score
`
`LENALIDOMIDE
`
`PLACEBO
`
`P VALUE*
`
`(n = 85)
`.11
`(n = 85)
`6.36
`4.55
`1.77
`(n = 85)
`.17
`(n = 76)
`3.65
`
`(n = 92)
`.24
`(n = 92)
`3.94
`2.85
`1.15
`(n = 85)
`.19
`(n = 87)
`4.37
`
`.35
`
`.09
`.12
`.16
`.90
`
`.16
`
`LENALIDOMIDE, N
`
`PLACEBO, N
`
`TOTAL
`(N = 87)
`
`RESPONDERS
`(N = 14)
`
`TOTAL
`(N = 93)
`
`RESPONDERS
`(N = 15)
`
`72
`3
`5
`1
`2
`10
`0
`1
`
`0
`0
`1
`0
`0
`4
`0
`1
`
`82
`2
`1
`3
`0
`7
`2
`0
`
`0
`0
`
`2
`0
`1
`0
`0
`
`NONDRUG THERAPY
`
`None
`TENS
`Acupuncture
`Spinal cord stimulation
`Psychological treatment
`Physical therapy
`Chiropractic therapy
`Stellate ganglia blocks
`
`Abbreviation: SF-MPQ, Short-Form McGill Pain Questionnaire.
`*No adjustment for multiplicity was made, because the P value is for reference
`only; the P value is from the analysis of covariance model adjusting for center
`and baseline score.
`
`Abbreviation: TENS, transcutaneous electrical nerve stimulation.
`NOTE. The total will not equal the sample size as some subjects received >1
`concomitant drug therapy.
`
`experimental agent in this initial study. The open-label
`pilot study accurately predicted safety, but not efficacy,
`in the current, adequately powered, controlled study.
`The lenalidomide 10-mg/d dose was well tolerated in
`this study; however, some AEs were observed. Given
`
`Table 4. Concomitant Medication Use During
`the Study Treatment Phase
`
`TREATMENT
`GROUP AND
`OUTCOME
`STATUS
`
`CONCOMITANT MEDICATIONS
`
`TOTM AD MR OP LA AC NSAID BZ OTHER
`
`86
`2.9
`
`18
`.2
`
`17
`.2
`
`66
`.8
`
`10
`.7
`
`11
`.1
`
`2
`.1
`
`39
`.5
`
`9
`.6
`
`16
`.2
`
`3
`.2
`
`7
`.1
`
`0
`.0
`
`21
`.2
`
`2
`.1
`
`the lack of clear efficacy and the dose dependency of
`AEs with moderate to long-term administration of lena-
`lidomide, higher doses approaching 25 mg are not
`advised. The quantitative sensory testing and measure-
`ments of limb volume used in the pilot study, initially
`thought
`important
`to include, proved difficult
`to
`execute and were inconsistently applied in the pilot,
`and thus they were not employed in this study, which
`focused on pain endpoints. Streamlining the battery of
`assessments was judged to be critical for larger studies.
`In future studies of CRPS, the need for objective assess-
`ments in pain clinical research must be balanced by feasi-
`bility considerations.17
`The current study is the largest controlled study of
`people with CRPS and one of the first to employ the
`Budapest research criteria for defining CRPS. These pro-
`vide a useful tool for carefully selecting subjects, but
`this symptom-based definition does not address poten-
`tial heterogeneity in etiology or disease mechanisms,
`which may have affected study results. This study was
`
`Lenalidomide
`Total
`n
`M
`Responder
`14
`n
`2.6
`M
`Nonresponder
`n
`72
`M
`3.0
`Placebo
`Total
`n
`M
`Responder
`15
`n
`2.3
`M
`Nonresponder
`n
`75
`M
`2.7
`
`90
`2.7
`
`3
`.2
`
`15
`.2
`
`18
`.2
`
`2
`.1
`
`16
`.2
`
`2
`.1
`
`15
`.2
`
`24
`.2
`
`5
`.3
`
`19
`.1
`
`56
`.8
`
`9
`.1
`
`30
`.4
`
`69
`.8
`
`9
`.6
`
`60
`.8
`
`10
`.1
`
`2
`.1
`
`8
`.1
`
`43
`.5
`
`7
`.5
`
`36
`.5
`
`13
`.2
`
`17
`.2
`
`4
`.3
`
`13
`.2
`
`7
`.1
`
`7
`.1
`
`0
`.0
`
`7
`.1
`
`19
`.3
`
`15
`.2
`
`2
`.1
`
`13
`.2
`
`Abbreviations: TotM, total medications used, including multiple agents within a
`class; AD, antidepressants, including tricyclic agents; MR, muscle relaxants and
`baclofen; OP, opioids; LA, local anesthetic preparations; AC, anticonvulsants,
`including gabapentin and pregabalin; NSAID, nonsteroidal anti-inflammatory
`agents; BZ, benzodiazepines, including clonazepam; Other, alpha adrenergic
`agents, mexiletine, and acetaminophen; M, mean.
`NOTE. The n value reflects the incidence of a parti