Vol. - No. - - 2011 Journal of Pain and Symptom Management 1

Original Article

A Spicamycin Derivative (KRN5500) ProvidesNeuropathic Pain Relief in Patients WithAdvanced Cancer: A Placebo-Controlled,Proof-of-Concept TrialSharon M. Weinstein, MD, Amy P. Abernethy, MD, Susan E. Spruill, PStat, MS,Isadore M. Pike, MD, Andrea True Kelly, PhD, and Linda G. Jett, MSNPain Medicine and Palliative Care Program (S.M.W.), The Huntsman Cancer Institute, and

Department of Anesthesiology (S.M.W.), University of Utah, Salt Lake City, Utah; Duke Cancer Care

Research Program (A.P.A.), Duke Cancer Institute, Durham, North Carolina; Applied Statistics &

Consulting (S.E.S.), Spruce Pine, North Carolina; Izzy Pike MD Consulting (I.M.P.), Fairhope,

Alabama; ATK Clinical Consulting (A.T.K.), Charleston, South Carolina; and DARA BioSciences

(L.G.J.), Raleigh, North Carolina, USA

Abstract

Context. Neuropathic pain in patients with cancer can be difficult to treat

effectively.Objectives. The purpose of the study was to determine safety and efficacy of

KRN5500, a novel, spicamycin-derived, nonopioid analgesic agent, in patients withadvanced cancer and neuropathic pain of any etiology.

Methods. The study was a Phase 2a, multicenter, double-blind, placebo-controlled, dose escalation clinical trial. Patients with refractory neuropathic painand advanced cancer were randomly assigned 2:1 to receive a maximum of eightsingle escalating doses of KRN5500 or placebo, ranging from 0.6 to 2.2 mg/m2.The primary objective was safety and tolerability. The secondary objective wasefficacy, measured by change in average pain intensity on a 0e10 numeric ratingscale administered one week after the patient’s final dose.

Results. Nineteen patients received treatment (KRN5500 n¼ 12; placebon¼ 7). The most frequently reported adverse events were gastrointestinalsymptoms, which were more frequent and severe with KRN5500 than placebo; two(17%) KRN5500 patients discontinued the study because of nausea and vomiting.At study end point, KRN5500 exhibited a significant median decrease in painintensity from baseline of 24% compared with 0% for placebo (P¼ 0.03). Themedian for largest weekly reduction in target pain intensity was 29.5% forKRN5500 and 0% for placebo patients (P¼ 0.02).

Conclusion. This proof-of-concept study for KRN5500 in patients with advancedcancer and any type of neuropathic pain found gastrointestinal adverse events tobe the predominant safety concern. The results also provided the first indication

Address correspondence to: Linda G. Jett, MSN, DARABioSciences, 8601 Six Forks Road, Suite 160, Raleigh,NC 27615, USA. E-mail: ljett@darabio.com

Accepted for publication: May 5, 2011.

� 2011 U.S. Cancer Pain Relief CommitteePublished by Elsevier Inc. All rights reserved.

0885-3924/$ - see front matterdoi:10.1016/j.jpainsymman.2011.05.003

2 Vol. - No. - - 2011Weinstein et al.

of clinical and statistical efficacy in reducing pain intensity. J Pain SymptomManage 2011;-:-e-. � 2011 U.S. Cancer Pain Relief Committee. Published by ElsevierInc. All rights reserved.

Key Words

Neuropathic pain, cancer, KRN5500, spicamycin, allodynia

IntroductionNeuropathic pain is characterized by an ab-

normal hypersensitivity to innocuous and nox-ious stimuli and can persist long after thetissue damage and inflammation that inducedthe pain have resolved. The International Asso-ciation for the Study of Pain defines neuro-pathic pain as ‘‘pain initiated or caused bya primary lesion or dysfunction in the nervoussystem’’1 and more recently, as ‘‘pain arising asa direct consequence of a lesion or diseaseaffecting the somatosensory system’’ in anattempt to more effectively distinguish neuro-pathic pain from nociceptive (e.g., inflamma-tory pain) and musculoskeletal or other typesof pain that arise indirectly in the course of neu-rologic disorders.2

Eleven million patients worldwide are af-flicted by neuropathic pain.3 The etiology ofneuropathic pain includes diverse conditions,such as diabetic neuropathy, postherpetic neu-ralgia, trauma, HIV/AIDS, and cancer-relatedneuropathy secondary to tumors, surgery, ra-diotherapy, and chemotherapy, as well as toxiceffects from a number of medications.4 Morethan one pathophysiologic mechanism mayunderlie neuropathic pain symptoms in cancerpatients, making this population more difficultto treat effectively.5,6 In addition, some expertsin the field consider neuropathic pain moreresistant than nociceptive or inflammatorypain to standard analgesic treatments.7,8 Alarge prospective epidemiological study foundthat the proportion of neuropathic pain con-tributing to cancer pain was 33%, with opioidtherapy effective in providing 50% relief forthe neuropathic pain.9 Another large prospec-tive study surveyed routine clinical practice,based on the 1986 version of the World HealthOrganization cancer pain guideline, in a largepopulation of advanced cancer patients withneuropathic or mixed (neuropathic and noci-ceptive) pain. A combination of opioid and

nonopioid analgesics and adjuvant therapyconsisting of anticonvulsants, antidepressants,and corticosteroids was successful for bothtypes of pain, although the lack of controlledtreatment and comparison arms makes the re-sults difficult to generalize.10,11

KRN5500, a novel spicamycin derivative pro-duced by Streptomyces alanosinicus, was discov-ered in an effort to identify new agents thatinduced differentiation of myeloid leukemiacells. KRN5500 failed to establish significanttherapeutic efficacy as a potential antineo-plastic agent in Phase 1 trials conductedunder a cancer Investigational New Drugapplication.12e14 However, clinical observa-tions of a patient who experienced remissionof severe neuropathic pain while receivingKRN5500 for metastatic cancer sparked inter-est in this drug as a possible neuropathicpain treatment. For more than two decades,the patient, who was enrolled in a KRN5500cancer trial, had been treated for peripheralneuropathy and severe pain secondary to im-munoglobulin A monoclonal gammopathyand Raynaud’s disease. Surprisingly, althoughno pain response was anticipated, the patient’sneuropathic pain improved from the time ofhis first infusion with KRN5500. His progresswas well documented and later published asa case history. He remained free of this long-standing and severe neuropathic pain untilhis death from cancer two months after stop-ping study treatment, although over the sameperiod of time, there was no apparent effecton the patient’s upper quadrant pain, whichwas associated with liver metastases. The ab-dominal pain continued to worsen duringthe last months of his life, thus providing aninitial indication that KRN5500 may not beeffective for nociceptive pain.15

This serendipitous finding led to nonclinicalin vivo studies evaluating response to KRN5500in rodent pain models. In three standard rat

Vol. - No. - - 2011 3KRN5500 and Neuropathic Pain

models of nerve injury-induced neuropathicpain, single doses of KRN5500 reversed painhypersensitivity within two hours, with effectslasting up to six weeks. KRN5500 had no ef-fects in uninjured rats or in a rat model ofacute inflammatory pain, suggesting specificityfor pain that is neuropathic in nature and sup-porting earlier clinical observations of neuro-pathic but not nociceptive pain relief.16e18

KRN5500 inhibits acetylcholinesterase andfatty acid amide hydrolase enzymes, both tar-gets that can modulate aspects of neuropathicpain. KRN5500 lacks activity at 87 other evalu-ated G protein-coupled receptor, ion channel,and enzyme targets. The prolonged durationof action seen in the rat pain model studiesdoes not correlate with plasma pharmacokinet-ics (terminal half-life¼ 0.6e1.5 hours), sug-gesting that KRN5500 might have diseasemodifying activity through synergistic interac-tions between the two identified mechanismsor through an as yet unidentified mechanism.

MethodsPatients

This Phase 2a study was a multicenter,double-blind, randomized, placebo-controlleddose escalation trial designed to evaluate thesafety and efficacy of KRN5500 in patientswith advanced cancer and neuropathic painof any etiology (Fig. 1). Advanced cancer wasdefined as cancer in which there was no

Fig. 1. Schematic of dose escalatio

curative therapeutic option. There was no re-striction on therapies that might contributeto patients’ comfort or quality of life. Palliative(noncurative) chemotherapy was allowed ifcompatible with study drug in terms of dosingschedule and stable adverse event (AE) profile.Although the patient population comprisedpatients with advanced cancer, refractory neu-ropathic pain of any etiology was acceptablefor study entry. Refractory was defined as fail-ure to achieve adequate pain relief from atleast two commonly used treatments for neuro-pathic pain. The target neuropathic pain forthe study had to be characterized by at leasttwo of the following symptoms: burning pain,shooting/lancinating pain episodes, dysesthe-sias, or allodynia at an overall pain score of$4 on a 0e10 numeric rating scale (NRS), de-spite any pain therapies patients were receivingat the time. Patients were allowed to continuetheir usual pain treatments, including nonste-roidal anti-inflammatory, antidepressant, anti-convulsant, and opioid medications if thetreatments were thought to be contributingto partial pain relief for the target neuropathicpain or other pain conditions. Because eachpatient was likely to have more than one typeand location of pain, the neuropathic painthat was considered the primary complaintwas identified as the target neuropathic painfor clinical assessments and patient self-evaluation; study staff recorded the targetedpain area in each diary NRS page as reinforce-ment for patients in rating their daily ‘‘target’’

n and treatment decisions.

4 Vol. - No. - - 2011Weinstein et al.

pain. Eligibility was restricted to patients withno radiation therapy to the target pain sitewithin four weeks of screening and no majorsurgery within two weeks. Patients were re-quired to be at least five half-lives posttreat-ment with any other investigational drugs.Eligibility was not restricted by type of canceror projected life expectancy.

Study DesignThe primary objective of this trial was to eval-

uate the safety and tolerability of KRN5500for neuropathic pain in patients with advancedcancer. Secondary objectives were to assess theanalgesic activity and the dose-response rela-tionship of KRN5500. All investigative sites,the protocol, and study materials were ap-proved by the investigator’s local or a central-ized Institutional Review Board. The studywas independently monitored by i3Research,a contract research organization, and conduct-ed in accordance with Good Clinical Practiceguidelines. Safety was assessed on an ongoingbasis through monthly review of all availabledata by a safety committee comprising thestudy oncology adviser and medical monitor.

Patients were randomly assigned to receiveup to eight doses of test article (KRN5500)or placebo over a 10-week treatment periodfrom December 2006 to March 2009. The ran-domization schedule, generated by a third-party statistician, assigned 24 patients to oneof two treatment groups in a 2:1 ratio(KRN5500:placebo) using a block size of threeover all sites in an effort to obtain sufficientdata from at least 18 patients. The study wasnot powered for prospective efficacy out-comes; rather, information regarding esti-mates of efficacy was to be used to make aninformed decision regarding further develop-ment. As such, no formal sample size calcula-tions were carried out for this study. Thesample size of 18 was considered typical forsimilar proof-of-concept/feasibility studies.19

Patients were assigned to treatment in sequen-tial order based on the randomization schemeusing an automated web-based central ran-domization process accessed by site study per-sonnel on obtaining signed informed consentand verifying all eligibility criteria. Random-ized patients who discontinued study participa-tion before being dosed were replaced withnew patients assigned to the same treatment.

Clinical personnel, monitors, patients, andthe sponsor remained blinded throughoutthe study to treatment received.

Treatment ScheduleA maximum of eight single escalating doses

ranging from 0.6 to 2.2 mg/m2 were adminis-tered during weekly visits as an intravenous(IV) infusion completed in one hour or less,with a maximum of 5 mg established for anysingle dose. KRN5500 (6-[4-Deoxy-4-[(2E,4E)-tetradecadienoylglycyl]amino-L-glycero-ß-L-mannohepto-pyranosyl]amino-9H-purine) was for-mulated in a mixture of dehydrated alcohol;propylene glycol, NF; polysorbate 80, NF;N,N-dimethylacetamide (DMAC), monoetha-nolamine, NF; and normal saline and was ad-ministered intravenously. Placebo was normalsaline, indistinguishable visually from theKRN5500 IV solution. Blinding was furtherenhanced by covering IV bags with an opaquematerial.Patients were evaluated over a 14-week pe-

riod, inclusive of follow-up. At each clinic visit,the investigator evaluated whether to redosethe patient with the same dose, a lower dose,the next higher dose, or no dose, based on as-sessments of tolerability, efficacy, and the pa-tient’s desire for treatment. During eachclinic visit, diary 0e10 NRS pain scores were av-eraged for the six days after the previous dose,and the percent change from the previousweek’s average was calculated. This informa-tion was used clinically to determine the ap-propriate dose for that visit. For dosingpurposes, response to study drug was definedas at least a 20% decrease in target pain inten-sity from the previous week’s average NRSscores. Before dosing, patients were asked,‘‘Is your target pain still significant enough toyou that you’d like further treatment?’’ Fig. 1presents a schematic of dose escalationguidelines.During the course of treatment, if unaccept-

able treatment emergent AEs (TEAEs) oc-curred, the patient could receive the nextlower dose. If unacceptable TEAEs occurredat the minimum dose (0.6 mg/m2), the pa-tient was discontinued from the trial. Labora-tory values and patient-reported AEs wereconsidered when evaluating tolerability andadditional dosing. The patient’s overall physi-cal condition related to the patient’s cancer

Vol. - No. - - 2011 5KRN5500 and Neuropathic Pain

also was a factor in dosing. There were occa-sions when dosing was withheld because ofthe need for respite from therapy or to accom-modate increased treatment for underlyingconditions. Patients were encouraged to com-plete at least four treatment visits to allow forthe possibility of full dose escalation but wereinstructed that they were not expected to par-ticipate beyond that if they did not perceiveany benefit from study participation.

Safety, Tolerability, and Pain ResponseEvaluation of safety and tolerability of

KRN5500 was the primary objective of thisstudy. To assess safety and tolerability, AEs,clinical laboratory assessments, vital signs,and electrocardiogram data were collected ateach visit. TEAEs were defined as AEs with on-set after the date and time of the first treat-ment, or, if the event was present beforetreatment, worsening after study intervention.TEAEs were assessed for intensity (mild, mod-erate, or severe) and relatedness to treatmentby the investigator. Levels of intensity were de-fined as: milddthe symptom was barely notice-able to the patient, did not influenceperformance or functioning, and generallydid not require prescription drug treatment;moderatedthe symptom was of sufficient sever-ity to make the patient uncomfortable, perfor-mance of daily activities was influenced, andprescription of other treatment for the symp-tom may have been needed; severedthe symp-tom caused severe discomfort, sometimes ofsuch severity that the patient could not con-tinue in the study, and prescription or othertreatment for the symptom was likely neces-sary. Serious adverse events (SAEs) were de-fined per U.S. Food and Drug Administrationguidelines as any AE meeting one or more ofthe following criteria: 1) was fatal or life-threatening; 2) was permanently disabling;3) resulted in unplanned or prolongation ofhospitalization; 4) resulted in persistent orsignificant disability/incapacity; 5) resultedin birth defect/congenital anomaly; or 6) re-quired medical intervention to prevent any ofthese outcomes.20

The secondary objective of this study wasevaluation of efficacy. The efficacy end pointmeasurement was the average target pain in-tensity score over the previous 24 hours asmeasured by an 11-point NRS score, ranging

from 0 to 10, where 0 represented ‘‘no pain’’and 10 represented the ‘‘worst possible pain,’’collected by a clinician at each weekly clinicvisit. Secondary efficacy measurements in-cluded NRS scores recorded by patients eachevening in a diary. The mean change in scoresfor the six days after each dose was used fordosing decisions and secondary end pointanalyses. Other secondary efficacy assessmentsincluded examinations and questionnaires formeasuring dimensions of neuropathic painand quality of life. These included the Neuro-pathic Pain Questionnaire (NPQ), Brief PainInventory-Short Form (BPI-SF), allodyniaphysical examination, Karnofsky PerformanceStatus (KPS), and the 12-item Short FormHealth Survey (SF-12v2�).21e24 Patients alsorecorded all pain medication use in a diary.

Results were summarized by treatmentgroup. Statistical tests, when performed, weretwo sided, nonparametric tests (WilcoxonRank Sum Test), conducted at the a¼ 0.05 sig-nificance level. Data summaries and analysesof results were generated using SAS� version8 or later (SAS Institute, Cary, NC). No interimanalyses were performed. All statistical analyseswere prospectively planned and documentedin an analysis plan before unblinding the data.

ResultsDemographic and Baseline Characteristics

Nineteen male and female patients, aged18 years or older with advanced cancer and re-fractory neuropathic pain of any etiology, wereenrolled at nine centers across the continentalU.S. and Puerto Rico. Because of a randomiza-tion error at one site, a patient who was toreceive KRN5500 was given placebo, resultingin 12 patients receiving KRN5500 and sevenpatients receiving placebo. Data presentedare based on treatment received (Fig. 2).Chemotherapy-induced peripheral neuropa-thy was the most frequently recorded etiologyfor patients’ neuropathic pain, followed bycomplex regional pain syndrome-Type II andsurgery-related neuropathic pain. Investigatorsfrequently documented more than one etiol-ogy. Patients most often identified lower limbs,followed by upper limbs, back, and face as thesite of the target neuropathic pain for the pur-pose of assessing NRS scores, allodynia

Assessed for eligibility (n=25)

Excluded (n=6) Not meeting inclusion criteria (n=5) Declined to participate (n=0) Other reasons

Schedule conflict (n=1)

KRN5500 patients analyzed (n=12)

Excluded from analysis (n=1) Intention to treat (n=13) Modified efficacy (n=12)

Lost to follow-up (n=0) Discontinued KRN5500 (n=9)

Adverse event (n=3) Withdrawal of consent (n=3) Progression of disease (cancer) (n=2) Bed-rest ordered by oncologist (n=1)

Allocated to KRN5500 (n=13) Received KRN5500 (n=12)Did not receive KRN5500 (patient received

placebo in error) (n=1)

Lost to follow-up (n=0) Discontinued placebo (n=5)

Adverse event (n=1) Protocol violation (n=1) Withdrawal of consent (n=3)

Allocated to placebo (n=6) Received placebo (n=7)Did not receive allocated intervention (n=0)

Placebo patients analyzed (n=7)

Excluded from analysis (n=0) Intention to treat (n=6) Modified efficacy (n=7)

Randomized (n=19)

♦♦♦

♦♦

♦♦♦

♦♦

♦♦

♦♦♦♦

♦♦

Fig. 2. CONSORT diagram of patient disposition in the study.

6 Vol. - No. - - 2011Weinstein et al.

examination scores, and all other assessments.The same target neuropathic pain location,specified at the initiation of the trial, was tobe consistently assessed for efficacy measure-ments throughout the study. Baseline patientdemographics and neuropathic pain charac-teristics are summarized in Table 1.

Safety and TolerabilityExposure to Study Drug. Table 2 provides dataon exposure to study drug and reasons for dis-continuation. The median duration of expo-sure was 40 days for the KRN5500 treatmentgroup compared with 29 days for the placebotreatment group. On average, KRN5500 pa-tients took 5.3 doses whereas placebo patientstook 4.4 doses. The highest dose of KRN5500received was 2.2 mg/m2. Seven (58%) patientstreated with KRN5500 were exposed to at leastone dose at this level, and four (57%) placebopatients were escalated to at least one dose ofnormal saline at this level.

Given the serious nature of patients’ ill-nesses and the relatively long length of the

study (10 treatment weeks), the protocol andinformed consent form specified that patientswere not expected to remain in the study be-yond its perceived benefit to them. Of the 14patients who discontinued before receivingtheir eighth possible dose, nine (75%) weretreated with KRN550 and five (71%) weretreated with placebo.

Adverse Events. A summary of TEAEs is pre-sented in Table 3. All patients (100%) whoreceived KRN5500 had at least one TEAE, com-pared with 86% who received placebo. Eventsof greater severity were more frequent in theKRN5500 group compared with placebo,and more patients treated with KRN5500 dis-continued study participation because ofTEAEs (25%) compared with those treatedwith placebo (14%). The most frequently re-ported TEAEs were gastrointestinal (GI) in na-ture (primarily nausea and vomiting), whichoccurred more often and with greater severityin the KRN5500 group than in the placebogroup. GI events in the KRN5500 treatment

Table 1Baseline Patient Demographics and

Characteristics

KRN5500(n¼ 12)

Placebo(n¼ 7)

Variable Mean (SD) Mean (SD)

Age (years) 61 (12.0) 63 (13.4)

Body surface area (m2) 1.9 (0.21) 1.7 (0.35)

NRS 7.4 (1.62) 6.4 (1.90)

n (%) n (%)

GenderMale 6 (50) 4 (57)

RaceCaucasian 9 (75) 5 (71)Black 1 (8) 1 (14)Other 2 (17) 1 (14)

Etiology of neuropathic paina

Chemotherapy induced 10 (83) 7 (100)Complex regional painsyndrome-Type II

4 (33) 0

Complex regional painsyndrome-Type I

1 (8) 0

Postherpetic 1 (8) 0Cancer related 1 (8) 0Radiation induced 1 (8) 0Diabetic 1 (8) 0Surgery related 1 (8) 1 (14)Carpal tunnel syndrome 1 (8) 0

Target neuropathic painb

Lower extremity 7 (58) 5 (71)Upper extremity 3 (25) 1 (14)Trunk 1 (8) 1 (14)Face 1 (8) 0

SD¼ standard deviation; NRS¼ numeric rating scale.aMost patients had more than one etiology recorded.bTarget neuropathic pain identified as specific body area for all as-sessments; recorded on the NPQ and diary NRS forms for patientsto refer to.

Table 2Exposure to Study Drug and Discontinuation

MetricKRN5500(n¼ 12)

Placebo(n¼ 7)

Median number ofdoses (range)

5.5 (3, 8) 4.0 (2, 8)

Median days ofexposure (range)a

40 (14, 71) 29 (9, 64)

Median dose densities(mg/m2/week)b

(range)

1.27 (0.7, 1.5) 1.20 (0.6, 1.5)

Number (%) of patientsby dose levelc

0.6 mg/m2 12 (100) 7 (100)1.2 mg/m2 12 (100) 6 (86)1.8 mg/m2 10 (83) 6 (86)2.2 mg/m2 7 (58) 4 (57)

Patients completed $4dosing visits

10 (83) 6 (86)

Patients completed alldosing visits

3 (25) 2 (29)

Patients withdrawn 9 (75) 5 (71)Reason for early

withdrawalAEd 3 (25) 1 (14)Protocol violation 0 1 (14)Lost to follow-up 0 0Withdrawal of

consente3 (25) 3 (43)

Death 0 0Otherf 3 (25) 0

AE ¼ adverse event.aDuration of exposure¼ date of last dose� date of first doseþ 1 inthe specified interval.bDose density is calculated as average weekly dose, including restweeks.cThe number of unique patients receiving each dose level. The per-centage is calculated using the number of patients in each treat-ment group as the denominator.dOne KRN5500 SAE (seizures) and two KRN5500 AEs of nausea/vomiting led to withdrawal; one placebo SAE (stroke) led towithdrawal.eNo explanations were collected for ‘‘Withdrawal of Consent.’’ It ispossible that some patients withdrew consent because of lack ofbenefit (absence of pain relief).fReasons for ‘‘Other’’ included: two KRN5500 for progression of dis-ease; one KRN5500 attending oncologist recommended bed rest.

Vol. - No. - - 2011 7KRN5500 and Neuropathic Pain

group were considered related to study drugand required treatment with antiemetic medi-cations. After the first study dose, antiemeticscould be provided prophylactically if clinicallyindicated. Two patients receiving KRN5500withdrew from the study because of GIsymptoms.

A greater percentage of patients in theKRN5500 group (25%) experienced an SAEcompared with those in the placebo group(14%). All SAEs in the KRN5500 group oc-curred during the posttreatment follow-up pe-riod and were not considered related to studydrug. The following SAEs occurred in theKRN5500 group: 1) hospitalization for conges-tive heart failure and pneumonia 24 days afterthe last dose of study drug; 2) hospitalizationfor dehydration 29 days after the last dose of

study drug; and 3) hospitalization 18 days afterthe last dose of study drug for convulsions andsecondary memory loss (caused by the devel-opment of a new brain metastasis in a patientwith head and neck cancer). All events werefollowed until resolution. One patient receiv-ing placebo experienced a stroke after the sec-ond dose of placebo, which led to disabilityand discontinuation from the study. This SAEwas considered not related to treatment andremained unresolved at the end of thefollow-up period.

Other Safety Assessments. No significant labora-tory abnormalities related to study drug wereobserved in either treatment group. As a whole,

Table 3TEAEs and Organ System Most Affected

KRN5500(n¼ 12)

Placebo(n¼ 7)

TEAEs n (%) n (%)

Patients with any TEAE 12 (100) 6 (86)Patients with any TEAE by gender

Male 6 (100) 3 (75)Female 6 (100) 3 (100)

Patients with an SAEa 3 (25) 1 (14)Patients with an AE leading to

withdrawalb3 (25) 1 (14)

Patients with an AE outcome ofdeath

0 0

By organ system most affectedc

GI disorders 11 (92) 4 (57)Mild 4 (33) 2 (29)Moderate 3 (25) 2 (29)Severe 4 (33) 0Related 11 (92) 1 (14)

Nervous system disorders 6 (50) 2 (29)Infections and infestations 3 (25) 0Metabolism and nutritiondisorders

3 (25) 0

Musculoskeletal and connectivetissue disorders

3 (25) 2 (29)

Psychiatric disorders 3 (25) 0Skin and subcutaneous tissuedisorders

3 (25) 4 (57)

General disorders andadministration site conditions

2 (17) 0

TEAE¼ treatment emergent adverse event; SAE¼ serious adverseevent; AE¼ adverse event; GI¼ gastrointestinal.aAll SAEs were assessed as not related to study drug.bOne KRN5500 SAE (seizures) and two KRN5500 AEs of nausea/vomiting led to withdrawal; one placebo SAE (stroke) led towithdrawal.cEvents experienced by two or more patients in either treatmentgroup.

8 Vol. - No. - - 2011Weinstein et al.

no clinically relevant changes in vital signs ornegative changes in physical examinationswere observed in either treatment group. Elec-trocardiogram changes were within acceptablelimits for all parameters.

Efficacy AssessmentsAnalgesic Activity of KRN5500. Patients receiv-ing KRN5500 exhibited a statistically significantmedian decrease in average pain intensity of24% based on baseline in-clinic NRS scores,whereas the median change in patients receiv-ing placebo was zero (P¼ 0.03). There was noclear linear relationship between reduction inpain intensity and increasing dose for eithergroup (Table 4). The median for absolute unitdecrease on the 11-point NRS scale was twounits for KRN5500 and zero for placebo(Table 5). Fig. 3 provides individual patient

NRS data plotted to show baseline, weekly, andend point scores and depicts positive or nega-tive final outcome.The median greatest weekly reduction (i.e.,

best response) in target pain intensity was29.5% for KRN5500 patients, whereas the me-dian in placebo patients was 0% (P¼ 0.02)(Fig. 4). Clinic NRS data also were evaluatedfor clinically meaningful levels of pain reduc-tion achieved during the study (Table 5).There was little difference between treatmentgroups for patients that achieved $50% painreduction; 25% of KRN5500 vs. 14% of pla-cebo patients achieved this level. However,50% of KRN5500 vs. 14% of placebo patientsachieved $30% pain reduction. In addition,83% of KRN5500 patients reached $20%pain reduction compared with 29% of placebopatients who reached this level.Daily patient diary NRS ratings were highly

correlated with in-clinic ratings (r¼ 0.87;P¼ 0.007); however, there was no statistical dif-ference between groups, with the KRN5500group providing a median decrease from base-line of 16% compared with 0% for placebo(P¼ 0.07) (Fig. 5). Investigator allodynia test-ing yielded a 33% median decrease in painfor both mechanical and cold stimuli in theKRN5500 group compared with 0% and 8%decrease, respectively, in the placebo group.There was no median change (0%) in KPSscores for either treatment group (Table 5).

Additional Questionnaires and AssessmentsIn general, when a difference between treat-

ment groups was observed in patient self-ratedwritten questionnaires and assessments, theresults were mixed (Table 6). In the NPQ,placebo patients had more reduced scoresfor individual symptoms (seven of 10) thanKRN5500 patients (two of 10). However, thedifference in symptom improvement betweenthe two groups was generally <12%, exceptfor Pain because of weather change, which pro-vided the difference in favor of placebo with21% improvement compared with 3% forKRN5500 and Freezing pain, which showeda larger decrease in the KRN5500 group at33% compared with 0% in the placebo group.The BPI-SF also showed very little differencebetween treatment groups, with the only nota-ble difference in favor of placebo for Pain reliefin last 24 hours showing 32% improvement

Table 4Median Change From Baseline NRS in Target Pain Intensity and Median Best Response Across Dosesa

Dose

KRN5500 (mg/m2) Placebo (mg/m2)

0.6 1.2 1.8 2.2 Allb (range) 0.6 1.2 1.8 2.2 Allb (range)

n 1 2 3 6 12 1 0 2 4 7Endpointc change (%)

one week after final dose�25 �14 �30 �6 L24 (L100, 0) �20 d 0 0 0 (L20, 25)

Dose

KRN5500 (mg/m2) Placebo (mg/m2)

No dosed 0.6 1.2 1.8 2.2 Anye (range) No dosed 0.6 1.2 1.8 2.2 Anye (range)

n 3 4 1 2 2 12 2 3 1 0 1 7Bestf response (%) one

week after any dose�29 �24 �13 �32 �92 L29 (L100, 0) 13 0.0 �60 d �13 0.0 (L60, 25)

NRS¼ numeric rating scale.aBased on in-clinic NRS results.bIncludes all final doses for endpoint analysis.cP¼ 0.03, obtained from a Wilcoxon rank sum test.dSome patients achieved their best responses after the rest week, when no dose was given.eDenotes best response at any dose.fP¼ 0.02, obtained from a Wilcoxon rank sum test.

Vol. - No. - - 2011 9KRN5500 and Neuropathic Pain

compared with 6% in the KRN5500 group.The SF-12 quality-of-life instrument showedno difference between treatment groups, nordid two self-rated allodynia questions fortouch- and cold-induced pain.

Discussion and ConclusionsDiagnosis and treatment of neuropathic

pain in cancer patients remains challenging.There are often multiple pathophysiologicpain etiologies present in a single patient(neuropathic pain and nociceptive or inflam-matory pain), often with multiple origins. Forexample, neuropathic pain can be the resultof nerve lesions caused by direct tumor inva-sion or result from secondary causes, such

Table 5Changes From Baseline by Treatment Group for NRS A

Allodynia Examinations, and

Measure

Actual NRS unit change from baseline, median (range)

Clinic NRS: standard percentage reductionsNumber (%) of patients who achieved $50% reductionNumber (%) of patients who achieved $30% reductionNumber (%) of patients who achieved $20% reduction

Allodynia physical examination, median (range)Dynamic (mechanical)dtouch-induced painThermaldcold-induced pain

KPS, median (range)

NRS¼ numeric rating scale; KPS¼ Karnofsky Performance Status.

as fibrosis or myelopathy post-radiotherapy,chemotherapy, surgery, or combinationsthereof.25 Treatment of neuropathic pain incancer patients is even more complicatedgiven that epidemiological research is severelylacking.26 Additionally, most agents devel-oped for neuropathic pain have been studiedin noncancer patients, such as those withpostherpetic neuralgia or diabetic peripheralneuropathy. Successful results in those set-tings may not always apply to the complex pa-tient with advanced cancer and neuropathicpain.27

This Phase 2a randomized and controlledtrial using NRS scores as a measure of efficacyprovides the first demonstration of safety andefficacy for KRN5500 in a small cohort of

bsolute Units, Standard Percentage Reductions,Performance Status

KRN5500 (n¼ 12) Placebo (n¼ 7)

�2 (�7, 0) 0 (�1, 1)

3 (25) 1 (14)6 (50) 1 (14)

10 (83) 2 (29)

�33 (�100, 0) 0 (�100, 300)�33 (�100, 0) �8 (�100, 100)

0 (�10, 30) 0 (�10, 0)

0

1

2

3

4

5

6

7

8

9

10

0 1 2 3 4 5 6 7 8 9 10 11 12

Patients

Clin

ic N

RS

Baseline Endpoint Weekly Visits Pain Improved Pain Worsened

|-------------------------KRN5500--------------------------| |-----------PLACEBO---------------|1 2 3 4 5 6 7

Fig. 3. Absolute change in NRS scores for each patient by treatment group. (Total number of NRS weekly scoresper patient [up to 11] cannot be ascertained in this figure as duplicate scores are not depicted, with the exceptionthat a score [only 1] that overlays baseline and/or end point is visible.)

10 Vol. - No. - - 2011Weinstein et al.

patients with treatment-resistant neuropathicpain (of any etiology) and advanced cancer.The KRN5500 safety profile showed clinically

Fig. 4. Maximum reduction in neuropathic pain score by pagroup. (P¼ 0.02 obtained from a Wilcoxon rank sum test.)

significant GI symptoms emerging as the pre-dominant and only safety concern at doselevels that notably reduced pain intensity.

tient ordered from greatest to least within treatment

0

2

4

6

8

10

0 2 4 6 8 10

Diary NRS

Clin

ic N

RS

r = 0.87

Fig. 5. Correlation of weekly diary and clinic NRSscores for all subjects. (Diary scores are plotted asthe mean of the first six daily NRS scores aftereach weekly visit.)

Table 6Median Percent Change From Baseline in

Patient-Rated Questionnaires One Week AfterFinal Dose

KRN5500(n¼ 12)

Placebo(n¼ 7)

Patient self-ratedassessmentsa % change (range) % change (range)

NPQBurning pain �15 (�100, 0.0) �7 (�100, 67)Sensitive to

touch�8 (�100, 100) �17 (�100, 0)

Shooting pain 0 (�100, 200) �11 (�100, 20)Numbness �5 (�100, 50) �7 (�40, 11)Electric pain �14 (�100, 0) 20 (�100, 20)Tingling pain �14 (�100, 300) �23 (�100, 0)Squeezing pain 0 (�100, 300) 0 (�100, 20)Freezing pain �33 (�100, 0) 0 (�21, 0)Pain because

of touch�6 (�100, 700) �16 (�100, 100)

Pain becauseof weatherchanges

�3 (�100, 900) �21 (�100, 0)

BPI-SFPain at its worstin last 24 hours

0 (�100, 33) 0 (�25, 25)

Pain at its leastin last 24 hours

�16 (�100, 20) 0 (�40, 300)

Pain on theaverage

0 (�100, 100) 0 (�40, 13)

Pain right now 0 (�100, 67) 0 (�43, 50)Pain relief in

last 24 hours�6 (�50, 250) �33 (�100, 0)

Mean of painseverity

2 (�100, 29) 0 (�29, 12)

Mean of paininterference

�10 (�100, 131) 2 (�23, 51)

SF-12�

Physicalcomponentsummary

0 (�2, 10) �1 (�3, 2)

Mentalcomponentsummary

3 (�10, 11) 2 (�20, 8)

Self-ratedallodyniaquestionnaire

Touch-inducedpain

�30 (�100, 60) �17 (�100, 33)

Cold-inducedpain

�32 (�100, 29) �31 (�100, 0)

NPQ¼Neuropathic Pain Questionnaire; BPI-SF¼ Brief PainInventory-Short Form; SF-12�¼ Short Form Health Survey.aFor all assessments, decreases in scores indicate improvements,except for SF-12, where higher scores indicate better healthstatus.

Vol. - No. - - 2011 11KRN5500 and Neuropathic Pain

Study safety data demonstrate that drug-related GI symptoms of nausea, vomiting,and less frequently, diarrhea were a distressing,though temporary, adverse effect of KRN5500.Although these symptoms occurred in the pla-cebo group as well, the frequency, severity, andrelatedness to study drug indicate that this maybe a clinically significant problem associatedwith the use of KRN5500. No other safety con-cerns were identified.

Certain excipients, namely DMAC and mono-ethanolamine, part of the Phase 1 anticancerformulation that also was administered in thisstudy, are known to cause nausea and vomit-ing;28,29 in addition, patients receiving placebowere given normal saline without any of the for-mulation excipients used with KRN5500. Thus,it is not possible at the present time to deter-mine whether the GI effects are a result of theactive pharmaceutical ingredient, formulationexcipients, or a combination thereof. In an ef-fort to better understand and decrease theseside effects, reformulation efforts are underwayto replace emetogenic components with ingre-dients that are Generally Recognized as Safe(GRAS).

KRN5500 yielded a significant median de-crease in pain intensity from baseline whencompared with placebo. Efficacy was furthersupported by the significant difference inthe median best (largest) pain reduction perpatient. Diary NRS scores were highly

correlated with clinic NRS scores, and clinicaltesting of allodynia yielded a substantial me-dian pain decrease for both mechanical andcold stimuli in the KRN5500 group and nonein the placebo group.

12 Vol. - No. - - 2011Weinstein et al.

Study results also provided clinically mean-ingful decreases in absolute pain units ona standard clinical 0e10-point pain intensityrating scale, with 10 of 12 KRN5500 patientsachieving a two-point reduction comparedwith one of seven placebo patients (Fig. 3).Six of 12 KRN5500 patients achieved $30%reduction in pain, whereas one of seven pla-cebo patients achieved this level. In similarstudies, meeting a standard cutoff of 30%for pain reduction or an absolute reductionof two units on a standard 0e10-point painintensity rating scale is considered clinicallysignificant.30

Except for the diary NRS scores, none of thedata collected entirely by patient self-rated in-struments showed a trend for efficacy in eithergroup. Results from the NPQ and BPI-SF werevaried and did not provide clinically significantdifferences in individual symptoms or overalllevels of pain or relief. Small differences andmedian scores with wide ranges make theresults difficult to interpret. Aside from thesmall sample size, the authors have no clearexplanation for these inconsistencies otherthan the possibility that some tools weremore sensitive than others in detecting treat-ment differences in this particular group ofsubjects.

The NPQ divides the overall target painlevel into 12 components (symptoms). Withthe exception of burning pain, changes frombaseline on individual questions varied greatlyfrom week to week, suggesting that thecomponent questions were not a sensitivemeasure of treatment differences in this study.Because the pattern of symptoms is differentfor each patient, it may not be reasonable toexpect component scores to show the samepattern or degree of change as the overallpain rating. The NPQ is designed as a diagnos-tic tool for confirming neuropathic pain andhas not been validated as an instrument formeasuring efficacy through changes in symp-tom scores.31

The BPI-SF asks patients to rate pain in fourdifferent ways on the same 11-point NRS scale,as follows: ‘‘worst pain in the last 24 hours,’’‘‘least pain in the last 24 hours,’’ ‘‘pain on the av-erage,’’ and ‘‘pain you have right now.’’ None ofthese is exactly the same wording used for theclinic/diary NRS assessment. It is possible thatasking so many versions of the same question

led to fatigue and confusion in scoring. It alsois possible that several versions of the samequestion served to divide the level of pain intosmaller components, in much the same waythat the NPQ divides pain into symptomcomponents.A major limitation of the study was its small

sample size, making the study vulnerable tovariable outcomes. In particular, because pa-tients were allowed to escalate to an effectivedose, the number of patients in any particulardose level was too small to reliably estimatea dose-response relationship. Although it ispossible that positive results in this study repre-sent a Type I error (false positive outcome),32

there is strong evidence supporting the posi-tive analgesic outcome, including both clini-cally and statistically significant results, as wellas nonsignificant but positive trends.It is notable that, except for diary NRS scores,

each of the positive efficacy results is based ondata ascertained by clinician interview (askingthe patient to rate their pain) or clinician exam-ination with an interview component, such asthe allodynia examination, which combinedmechanical and cold stimulation with verbalcollection of NRS pain scores before and afterthe stimulation. The diary NRS question waswritten with exactly the same wording as the ver-bally administered clinic NRS question, whichlikely contributed to its positive correlationto primary efficacy end point results. None ofthe other data collected through patient-recorded outcomes provided meaningfultrends in either direction.Patients did choose a ‘‘target neuropathic

pain’’ site for focused assessment throughoutthe study, and although there may be variabilityin the degree to which patients and clinicianswere able to separate one site and kind ofpain from another, tests, such as the allodyniaphysical examination could only have pro-duced results by strictly focusing on the targetarea. The fact that pain was significantly re-duced is in itself a positive finding, even if it rep-resents a more nonspecific site or overall painrelief in some patients. Careful selection of di-agnostic and neurological tests, more thoroughtraining for patients, and appropriate patient-rated questionnaires will be important forfuture studies.Large placebo effect in pain trials often con-

tributes to difficulty in establishing significant

Vol. - No. - - 2011 13KRN5500 and Neuropathic Pain

differences between treatments.33 Placebo re-sponse, although present, was low in this study.The reasons for this are not known, but thefact that patients were more or less nearthe end of life, and were willing to enter intoa placebo-controlled trial, may be an indica-tion that pain was sufficiently severe to makeany sustained placebo response unlikely.Though baseline pain levels were not unusu-ally high, it is possible that baseline and contin-ued pain levels represented subjectively higherdiscomfort than would be the case in patientswho have not endured as much physical andemotional stress.

High study withdrawal rate is another factorthat has repeatedly contributed to failure inoncology symptom control studies.34 Becausethis study was considered lengthy for patientswith advanced cancer and a serious pain condi-tion, an unusual approach was used to proac-tively address an anticipated high withdrawalrate. In this study, over and above the usualstatement that participation is voluntary andpatients can withdraw at any time, patientswere explicitly told that they were not ex-pected to continue participation if the studywas not beneficial or if other more pressingneeds presented. Patients were asked to com-mit to at least four treatment visits if possible,to allow for the possibility of escalation to thehighest dose to assess for efficacy if it would oc-cur. This may have contributed to subjects re-maining on study at least long enough toreach an efficacious dose or to determinethat the study drug was not providing analgesia(no response). Most patients stayed in thestudy for the minimum period requested.The analysis plan called for data collectedone week after the last dose, no matter whenit occurred, to serve as the efficacy end point,which resulted in no missing end points andno data carried forward to fill in missingassessments.

Given the difficulty in reproducing positiveoutcomes in postmarketing and comparatortrials of analgesic medications (e.g., gabapen-tin), it is possible that pain studies may be par-ticularly susceptible to a Type II error (falsenegative outcome) related to some of the is-sues discussed above. As a result, potentiallybeneficial therapeutics may be abandonedtoo early in development because of negative

results or a positive trend not considered ro-bust enough. Given the burden of suffering as-sociated with severe pain, resultant greatclinical need for new pain therapies, and theoverwhelming odds against bringing new ther-apeutics to proof of concept, we believe that itis important to move forward with cautious op-timism to design a Phase 2b study. Future stud-ies of KRN5500 with modified formulationexcipients may shed additional light onwhether this investigational drug holds prom-ise as a tool for oncologists and pain specialistsworking with cancer patients or for those whomay have recovered from cancer only to findtheir quality of life seriously impaired by con-tinued neuropathic pain.

In conclusion, KRN5500 demonstratedsafety at dose levels providing first therapeuticevidence for treatment-resistant neuropathicpain in patients with advanced cancer, al-though causing transient GI side effectsthat were generally manageable. Despite therelatively small sample size, statistically signifi-cant and clinically meaningful improvementsin pain were observed in patients treatedwith KRN5500 when compared with placebo.With continued formulation development,better selection of diagnostic and efficacy as-sessment tools, and larger well-controlled tri-als, KRN5500 may hold promise as a safe andeffective treatment, especially for patientswith severe disabling neuropathic pain.

Disclosures and AcknowledgmentsThis study was funded by the sponsor,

DARA Therapeutics, a subsidiary of DARA Bio-Sciences (DARA). DARA personnel participatedwith industry advisers, investigators, and consul-tants in designing the study. i3Research, a con-tract research organization under contract withthe sponsor, conducted the study, includingsite monitoring, data management, and statisti-cal analysis.

Of the authors, Sharon M. Weinstein was aninvestigator; Amy P. Abernethy was an investiga-tor and has been a DARA consultant as well;Isadore M. Pike (oncologist), Susan E. Spruill(statistician), and Andrea True Kelly (clinicaladviser/medical writer) are DARA consultants.Linda G. Jett is an employee of DARA.

14 Vol. - No. - - 2011Weinstein et al.

The authors thank the patients, study coor-dinators, and enrolling investigators: GhassanAl-Jazayrly, Enser Cole, Thomas Cosgriff, TanyaDorff, Louis Rivera-Colon, Jack Saux, LeonardSender, and Jose Stable.

References1. Merskey H, Bogduk N. Classification of chronicpain. Seattle, WA: IASP Press, 1994.

2. Treede R-D, Jensen TS, Campbell JN, et al. Neu-ropathic pain: redefinition and a grading system forclinical and research purposes. Neurology 2008;70:1630e1635.

3. Olsen S. Providing an enhanced perspective onneuropathic pain. Consulting and marketing report.San Mateo, CA: WWMR, Inc., 2002. Available fromhttp://www.highbeam.com/publications/pr-newswire-p3672/aug-27-2002/2. Accessed September 28, 2011.

4. Pappagallo M, ed. The neurological basis ofpain. New York: McGraw-Hill, Co., 2005.

5. Baliki MN, Gehe PY, Apkarain AV, Chialvo DR.Beyond feeling: chronic pain hurts the brain, dis-rupting the default mode network dynamics.J Neurosci 2008;28:1398e1403.

6. Kuchinad A, Schweinhardt P, Wood P, Chizh BA,Bushnell MC. Accelerated brain gray-matter loss infibromyalgia patients: premature aging of the brain?J Neurosci 2007;27:4004e4007.

7. Arner S, Arner B. Differential effect of epiduralmorphine in the treatment of cancer-related pain.Acta Anaesthesiol Scand 1985;29:32e36.

8. Cherny NI, Thaler HT, Friedlander-Klar H, et al.Opioid responsiveness of cancer pain syndromescaused by neuropathic or nociceptive mechanisms:a combined analysis of controlled single-dose stud-ies. Neurology 1994;44:857e861.

9. Garcia de Paredes G, del Moral Gonzalez F,Martinez del Prado M, et al. First evidence of onco-logic neuropathic pain prevalence after screening8615 cancer patients. Results of the On study. AnnOncol 2011;22:924e930.

10. Grond S, Radbruch L, Neuser T, et al. Assess-ment and treatment of neuropathic cancer pain fol-lowing WHO guidelines. Pain 1999;79:15e20.

11. World Health Organization. Cancer pain relief.Geneva, Switzerland: World Health Organization,1986.

12. Gadgeel SM, Bionpally RR, Heilbrun LK, et al.A phase 1 clinical trial of spicamycin derivativeKRN5500 (NSC 650426) using a phase 1 acceleratedtitration ‘‘2B’’ design. Invest New Drugs 2003;21:63e74.

13. Supko J, Eder J, Ryan D, et al. Phase 1 clinicaltrial and pharmacokinetic study of the spicamycinanalog KRN5500 administered as a 1-hour

intravenous infusion for five consecutive days to pa-tients with refractory solid tumors. Clin Cancer Res2003;9:5178e5186.

14. Yamamoto N, Tamura T, Kamiya Y, et al. Phase 1and pharmacokinetic study of KRN5500, a spicamy-cin derivative, for patients with advanced solid tu-mors. Jpn J Clin Oncol 2003;33:302e308.

15. Borsook D, Edward A. Antineuropathic effectsof the antibiotic derivative Spicamycin, KRN5500.Pain Med 2004;5:104e108.

16. Abdi S, Vilassova N, Decosterd I, Feroz N,Borsook D. The effects of KRN5500, a spicamycinderivative, on neuropathic and nociceptive painmodels in rats. Anesth Analg 2000;91:955e999.

17. DiLorenzo L, Kobierski L, Moore KA,Borsook D. A water soluble synthetic Spicamycin de-rivative (San-Gly) decreases mechanical allodynia ina rodent model of neuropathic pain. Neurosci Lett2002;330:37e40.

18. Kobierski L, Abdi S, DiLorenzo L, Feroz N,Borsook D. A single intravenous injection ofKRN5500 (antibiotic spicamycin) produces long-term decreases in multiple sensory hypersensitivitiesin neuropathic pain. AnesthAnalg 2003;97:174e182.

19. Skare K, Sietsema W. From serendipity to sci-ence. Good Clin Pract J 2005;12:1e4.

20. U.S. Food and Drug Administration. Code ofFederal Regulations Title 21. IND safety reports.21CFR312.32.

21. Krause S, Backonja M. Development of a neuro-pathicpainquestionnaire.Clin JPain2003;19:306e314.

22. Cleeland C, Ryan K. Pain assessment: global useof the Brief Pain Inventory. Ann Acad Med 1994;23:129e138.

23. Karnofsky D, Abelman W, Craver L, Burchenal J.The use of nitrogen mustards in the palliative treat-ment of carcinoma. Cancer 1948;1:634e656.

24. Ware J, Kosinski M, Tuner-Bowker DM,Gandek B. Version 2 of the SF-12 health survey.Boston, MA: Quality Metric, Inc., 2002.

25. Hausheer F, Foley K. Cancer neuropathic pain:overview of current status and future objectives.Oncologist 2010;15(Suppl 2):1e2.

26. Lema M, Foley K, Hausheer F. Types and epide-miology of cancer-related neuropathic pain: the in-tersection of cancer pain and neuropathic pain.Oncologist 2010;15(Suppl 2):3e8.

27. Cleeland C, Farrar J, Hausheer F. Assessment ofcancer-related neuropathy and neuropathic pain.Oncologist 2010;15(Suppl 2):13e18.

28. Material Safety Data Sheets for monoethanol-amine. Available from http://www.sciencelab.com/msds.php?msdsId¼9922885. Accessed September11, 2011.

29. Material Safety Data Sheets for N, N-dimethyla-cetamide. Available from http://www.caledonlabs.

Vol. - No. - - 2011 15KRN5500 and Neuropathic Pain

com/upload/msds/3900-1e.pdf. Accessed September11, 2011.

30. Younger J, McCue R, Mackey S. Pain outcomes:a brief review of instruments and techniques. CurrPain Headache Rep 2009;13:39e43.

31. Backonja M, Glanzman R. Gabapentin dosingfor neuropathic pain: evidence from randomized,placebo-controlled clinical trials. Clin Ther 2003;25:81e104.

32. Dworkin R, Turk D, Katz N, et al. Evidence-based clinical trial design for chronic pain

pharmacotherapy: a blueprint for ACTION. Pain2011;152:S107eS115.

33. Dworkin R, Katz J, Gitlin M. Placebo response inclinical trials of depression and its implications forresearch on chronic neuropathic pain. Neurology2005;65(Suppl 4):S7eS19.

34. Rao R, Flynn P, Sloan J, et al. Efficacy of lamotri-gine in the management of chemotherapy-inducedperipheral neuropathy: a phase 3 randomized,double-blind, placebo-controlled trial, no1c3. Can-cer 2008;112:2801e2808.