aKDbEKztFhJvLhVQt6fdh0AWZ0iO-trial_annotation_document

Randomized Placebo-Controlled Study Evaluating Lateral BranchRadiofrequency Denervation for Sacroiliac Joint PainSteven P. Cohen, M.D. [Associate Professor]*, Robert W. Hurley, M.D., Ph.D. [AssistantProfessor]**, Chester C. Buckenmaier III, M.D. [Associate Professor]#, Connie Kurihara, R.N.[Research Assistant]##, Benny Morlando, R.N. [Research Assistant]##, and AnthonyDragovich, M.D. [Assistant Professor]*#*Dept. of Anesthesiology & Critical Care Medicine, Johns Hopkins School of Medicine, Baltimore, Maryland,and Walter Reed Army Medical Center, Washington, DC**Dept. of Anesthesiology & Critical Care Medicine, Johns Hopkins School of Medicine#Dept. of Anesthesiology, Uniformed Services University of the Health Sciences, Bethesda, Maryland, andWalter Reed Army Medical Center##Dept. of Surgery, Walter Reed Army Medical Center*#Uniformed Services University of the Health Sciences, and Chief, Pain Management Center, Fort Bragg,North CarolinaAbstractBackground—Sacroiliac joint pain is a challenging condition accounting for approximately 20%of cases of chronic low back pain. Currently, there are no effective long-term treatment options forsacroiliac joint pain.Methods—A randomized, placebo-controlled study was conducted in 28 patients with injection-diagnosed sacroiliac joint pain. Fourteen patients received L4-5 primary dorsal rami and S1-3 lateralbranch radiofrequency denervation using cooling-probe technology following a local anestheticblock, and 14 patients received the local anesthetic block followed by placebo denervation. Patientswho failed to respond to placebo injections crossed over and were treated with radiofrequencydenervation using conventional technology.Results—One, 3 and 6-months post-procedure, 11 (79%), 9 (64%) and 8 (57%) of radiofrequencytreated patients experienced ≥ 50% pain relief and significant functional improvement. In contrast,only 2 (14%) patients in the placebo group experienced significant improvement at their 1-monthfollow-up, and none experienced benefit 3-months post-procedure. In the crossover group (n=11), 7(64%), 6 (55%) and 4 (36%) patients experienced improvement 1, 3 and 6-months post-procedure.Address Correspondence to: Steven P. Cohen, M.D., Johns Hopkins Pain Management Division, 550 North Broadway, Suite 301,Baltimore, MD 21029, Phone: 410-955-1818, Fax: 410-614-7597, E-mail: scohen40@jhmi.edu.The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting theviews of the Dept. of the Army or the Dept. of Defense.Conflict of Interest: Steven Cohen lectured on sacroiliac joint pain at 2 symposiums sponsored by Baylis Medical: American Societyof International Pain Physicians 9th Annual Meeting, June 25, 2007, Washington, DC; and International Spinal Interventional Society15th Annual Meeting, July 20, 2007, Baltimore, Maryland. Disposable equipment (e.g. RF tubing and needles) supplied by BaylisMedical, Montreal, Quebec, Canada.Trial Registration: clinicaltrials.gov identifier: NCT00373724Summary Statement: This randomized controlled study evaluating sacroiliac joint radiofrequency denervation provides preliminaryevidence that the procedure may provide intermediate-term pain relief and functional improvement in carefully selected patients.NIH Public AccessAuthor ManuscriptAnesthesiology. Author manuscript; available in PMC 2009 August 1.Published in final edited form as:Anesthesiology. 2008 August ; 109(2): 279–288. doi:10.1097/ALN.0b013e31817f4c7c.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

One year after treatment, only 2 (14%) patients in the treatment group continued to demonstratepersistent pain relief.Conclusions—These results provide preliminary evidence that L4 and L5 primary dorsal rami andS1-3 lateral branch radiofrequency denervation may provide intermediate-term pain relief andfunctional benefit in selected patients with suspected sacroiliac joint pain. Larger studies are neededto confirm our results, and determine the optimal candidates and treatment parameters for this poorlyunderstood disorder.IntroductionSacroiliac joint pain is a challenging condition estimated to account for between 15% and 20%of chronic axial low back pain cases.1,2 Presently, there is no reliably effective treatment forsacroiliac pain. In randomized studies evaluating peri- and intra-articular corticosteroidinjections in patients suspected of having sacroiliac joint pain, the results are divided as towhether or not they afford any long-term benefit.3–7 Studies evaluating conservative therapiesare flawed by the lack of adequate control subjects and inappropriate diagnostic work-ups.1In the past several years, radiofrequency denervation has emerged as a promising treatmentalternative for refractory cases of sacroiliac joint pain.8 The concept of disrupting the nervesupply to pain-generating spinal structures was extrapolated from over 30 years of experienceusing radiofrequency lesioning for zygapophsial (facet) joint pain.9 In 4 studies evaluatingdifferent variants of lower lumbar primary dorsal rami and sacral lateral branch radiofrequencydenervation, all reported success rates ranging between 67% and 89%.10–13 However noneof these studies were controlled, which raises questions regarding their validity andapplicability. In order to determine whether sacroiliac joint denervation is a viable treatmentfor patients suffering from chronic, intractable, injection-diagnosed sacroiliac joint pain, weconducted a placebo-controlled study evaluating L4 and L5 primary dorsal rami and S1-3lateral branch radiofrequency lesioning.Materials and MethodsPermission to conduct this study was granted by the internal review boards at Johns HopkinsMedical Institutions, Baltimore, Maryland, and Walter Reed Army Medical Center,Washington, District of Columbia, and all study participants who provided informed consent.The standardized protocol was performed at both institutions, with recruitment and allprocedures occurring between May 2005 and August 2006. A two-tailed power analysisdetermined a sample size of 14 in each group had 80% power (beta of 0.2) to detect a 2-pointdifference in the 0–10 numeric rating scale (NRS) between groups with a significance level(alpha) of 0.05.All procedures were done in an outpatient setting using local anesthesia, and for radiofrequencydenervation, intravenous sedation. Subjects were recruited from the regular pain clinicpopulations at the participating institutions. Inclusion criteria included age > 18 years; axiallow back or buttock pain ≥ 6 months in duration; tenderness overlying the sacroiliac joint(s);failure to respond to conservative therapy (e.g. physical therapy and pharmacotherapy),including long-term (> 2 months) pain relief with sacroiliac joint corticosteroid injections; and≥ 75% pain relief as calculated from a 6-hour post-block pain diary following a singlediagnostic sacroiliac joint injection. Exclusion criteria were focal neurological signs orsymptoms; radiological evidence of a symptomatic herniated disc; spondyloarthropathy;untreated coagulopathy; and unstable medical (e.g. unstable angina) or psychiatric illness (e.g.untreated depression) that might preclude an optimal treatment response. Prior to enrollment,all patients underwent magnetic resonance imaging to rule out other possible sources for theirback pain. Six patients underwent previous diagnostic spinal procedures, three in each group.Cohen et al. Page 2Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

These included four discograms and four medial branch (facet joint nerve) blocks, all of whichwere negative. Throughout the recruitment phase, 62 patients were excluded for a variety ofreasons, of which the most common was failure to achieve ≥ 75% documented pain relief fromthe diagnostic sacroiliac joint block (n=38; fig. 1).Screening Sacroiliac Joint InjectionsSacroiliac joint injections were performed using 22-gauge spinal needles inserted into thebottom one-third of the joint using fluoroscopic guidance in either a slightly oblique or antero-posterior view. Correct placement was ascertained in all cases by a sacroiliac joint arthrogram.Following confirmation of joint penetration, a 3 ml solution containing 2 ml of bupivacaine0.5% and 1 ml of 40 mg/ml of depo-methylprednisolone (Pharmacia and Upjohn, Kalamazoo,MI) was administered. After the injection, patients were instructed to engage in normalactivities and fill out 0–10 numerical rating scale (NRS) pain diaries every half hour over theensuing 6 hours. Only those patients who experienced ≥ 75% pain relief for at least 3 hourswhile performing their normal activities of daily living, but whose pain returned to near baselinewithin 2 months, were eligible for enrollment.Randomization and Primary TreatmentThe treatment of all subjects was done by a physician not involved in randomization. Studypatients were randomized in a 1:1 ratio to receive either true or placebo denervation. A researchnurse not involved in patient care performed randomization in blocks of 4 via pre-sealedenvelopes at each institution. Under sterile conditions with the patient positioned prone, a C-arm intensifier was used to optimize visualization of the target sites. For blockade and lesioningof the L4 and L5 dorsal rami, 22-gauge SMK-C10 (Radionics, Burlington, MA) cannulae with5-mm active tips were inserted parallel to the course of the nerve until bone was contacted justsuperior and medial to the junction between the superior border of the transverse and superiorarticular processes for procedures done at L4, and at the junction of the ala and articular processof the sacrum for L5 procedures, similar to previously published studies.14,15 Since it is notpossible to discern electrostimulation between the various branches of the L4 primary dorsalramus (it is the lateral branch that may innervate the sacroiliac joint), the targeted nerve at thislevel is referred to as the parent branch. At each level, placement of the electrode in closeproximity to the nerve was confirmed using electrostimulation at 50 Hz, with concordantsensation achieved at ≤ 0.5 V. Prior to lesioning, the absence of leg contractions was verifiedwith stimulation at 2 Hz up to 2 V. After satisfactory electrode placement, 0.5 ml of lidocaine2% was injected through each cannula to reduce thermal pain and ensure blinding. Theradiofrequency probe was then reinserted and a 90 second, 80° C lesion was made using aradiofrequency generator set to the lowest audible volume to blend in with ambient noise(Electrothermal 20S Spine System, Smith and Nephew, Andover, MA or Radionics RF LesionGenerator System, Model RFG-3C, Radionics, Valleylab, Boulder, CO).For S1-3 lateral branch procedures, 17-gauge 75-mm cooled electrodes with 4 mm active tips(Baylis Medical, Montreal, Quebec, Canada) were inserted between 3 and 5 mm from the lateralborder of the foramina at pre-designated positions. For right-sided S1 and S2 procedures, theseapproximately corresponded to the 1:00, 3:00 and 5:30 o’clock positions on the face of a clock;on the left, the target sites were at 6:30, 9:00 and 11:00 (fig. 2). At S3, needles were placed at1:30 and 4:30 on the right side, and 7:30 and 10:30 on the left. In 10 patients in whom the S4foramen was located level with or just below the inferior portion of the sacroiliac joint, oneupper lesion was also done at S4. Sensory stimulation was performed at each level only for thefirst needle placement, revealing concordant sensation at ≤ 0.5 volts. Prior to lesioning, 0.5 mlof lidocaine 2% was administered per spinal level. In order to ensure that anesthetic spread toadjacent foramina did not impede sensory testing, electrodes were placed and stimulated atcontiguous levels before denervation commenced. Once the needles were properly positioned,Cohen et al. Page 3Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

monopolar electrodes were sequentially inserted into the cannulae and 2.5- minute lesions weremade using a water-cooled radiofrequency heating system (Pain Management SInergy System,Baylis Medical) and generator (PMG-115-TD, V2.0A, Baylis Medical). Using cooling-probetechnology, the tissue temperature immediately adjacent to the cooled electrode is maintainedat 60° C, while the target tissue is heated to 75° C, resulting in a lesion diameter ranging between8 and 10 mm (fig. 3). For safety reasons, this aggressive lesioning precludes using cooling-probe technology for lumbar primary dorsal rami.In the control group, electrodes were similarly positioned, electrostimulation was performedin an identical manner, and 0.5 ml of lidocaine 2% was administered, but no current wasadministered. The average time it took to perform the radiofrequency and placebo procedureswere comparable (a mean of 61 minutes vs. 54 minutes, respectively, for the first 5 people ineach group).Outcome Measures, Crossover Group Treatment, and Follow-UpA physician unaware of the patient’s study group assignment obtained all outcome data duringscheduled follow-up visits. Between the procedure and first follow-up, no contact waspermitted between any patient and investigator(s) except for emergencies. All patients wereseen in the treating clinic 1-month post-procedure. If a patient obtained a positive globalperceived effect (GPE) and significant (≥ 50%) pain relief obviating the need for furthertherapy, he or she was re-evaluated 3 and 6-months post-treatment. Abridged follow-upinterviews were done by e-mail or telephone every 2 months after the 6-month follow-up inpatients who reported persistent relief in order to determine the duration of benefit. Patientswho did not obtain adequate symptomatic improvement were unblinded at follow-up. For thosewho obtained significant relief 1-month post-procedure, unblinding was done 3-months aftertreatment.Patients in the initial radiofrequency denervation (cooled electrode) group who failed to obtaina positive outcome were recorded as a treatment failure and offered alternative treatment. Allplacebo patients who failed to achieve a positive outcome were offered the opportunity tocrossover and receive sacroiliac joint denervation using conventional (non-cooled) technologyin an open-label parallel arm. The reason for using conventional equipment was based onavailability (i.e. cooled equipment was not ordered for patients in whom the outcome and hencetreatment plan was not known beforehand). In these patients, 22 gauge SMK-C10 (Radionics)cannulas with 5-mm active tips were placed in an identical fashion to that described for thetreatment group. Once concordant sensory stimulation was obtained and 1% lidocaineadministered at each level, the monopolar radiofrequency probe was reinserted and a 90 second,80° C lesion was made using a different generator (Electrothermal 20S Spine System, Smithand Nephew, or Radionics RF Lesion Generator System, Model RFG-3C, Radionics). Datafrom the crossover group was analyzed separately from that of the initial experimental group.The primary outcome measure was a 0–10 NRS pain score, which reflected the average painexperienced by the patient for 10 days prior to follow-up. Secondary outcome measuresincluded Oswestry disability index (ODI version 2.0, MODEMS, Des Plaine, IL, reflecting the10 days prior to follow-up) score, reduction in analgesic medications (defined as a 20%reduction in opioid use or complete cessation of a non-opioid analgesic),16 GPE, and acomposite successful outcome. A positive GPE was defined as an affirmative response to thefollowing 3 questions:1. My pain has improved/ worsened/ stayed the same since my last visit;2. The treatment I received improved/ did not improve my ability to perform dailyactivities;Cohen et al. Page 4Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

3. I am satisfied/ not satisfied with the treatment I received and would recommend it toothers.The composite binary variable “successful outcome” was predefined prior to initiation of thestudy as a ≥ 50% reduction in numerical pain score, a positive GPE, and either a 10-pointdecrease in ODI or a 4-point decrease coupled with a reduction in medication usage.17,18Statistical MeasuresStatistical analyses were performed using STATA version 10.0 (Statcorp, College Station,TX). The Shapiro-Wilk W test for normal data was performed on continuous outcomemeasures. The distribution of categorical variables in each group was compared using Fisherexact test. Continuous variables are reported as mean and standard deviation or median andinterquartile range. Categorical data are reported by number of subjects and percentage.Comparisons between the initial radiofrequency treatment group and the placebo group weremade with unpaired t-tests or Mann-Whitney U test. Since the continuous data in each grouphad a normal distribution, comparisons between and within the initial radiofrequency andcrossover treatment groups were made with two-way analysis of variance (ANOVA). Formultiple significance testing, post-hoc Bonferroni correction was used. Because baseline ODIdifferences were a potential confounding factor, an adjusted multiple linear and logisticregression analysis was performed for each continuous and categorical outcome measure,respectively.ResultsDemographicsData were analyzed on 28 patients. Demographic (including active duty status) and clinicalcharacteristics were balanced between the radiofrequency denervation treatment and controlgroup. Two patients, one each in the control and treatment groups, received bilateralprocedures. Thirteen patients were taking opioids, and 24 were on non-opioid analgesics. Therewere no differences with regard to treatment location with the exception of military duty status,which was not present in those subjects treated at Johns Hopkins. Pre-procedure NRS scoresdid not appear to differ between the radiofrequency treatment and the placebo group (6.5 ± 1.9and 6.1 ± 1.8, respectively). However, preprocedural ODI scores did differ between thetreatment and placebo arms (37.1 ± 10.6 and 47.9 ± 9.3, respectively, table 1).Three patients in the placebo group declined to crossover to the radiofrequency denervationtreatment group. Among these 3 patients, one elected not to receive the true procedure becausethe placebo treatment was “too painful”, and 2 sought alternative care. Based on thedemographic and clinical characteristics, there appeared to be no difference between thesepatients and those who elected to cross-over.Primary Outcome MeasureA significant difference in the primary outcome, NRS pain score, was detected between thetreatment and placebo groups at follow-up (table 2). One month after the procedure, thetreatment group had significantly lower NRS scores than the placebo group (2.4 ± 2.0; range0–8 vs. 6.3 ± 2.4; range 2–10, p<0.001, respectively). In the placebo group, only 2 patients at1 month and no patients at 3 months reported a positive outcome. The primary outcomeremained significantly different between the two groups when baseline ODI scores wereanalyzed as a covariate (coefficient of variation −3.8, 95% CL −5.8 to −1.8; p<0.001). At 3and 6-month follow-up, 8 and 4 patients in the treatment group, respectively, reported NRSpain scores ≤ 2.Cohen et al. Page 5Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

In a within-groups analysis, subjects who received radiofrequency treatment reportedsignificantly lower NRS scores at 1, 3 and 6 months post-procedure compared to baseline scores(p<0.001). Patients’ pain scores were reduced by 60%, 60%, 57% at 1, 3 and 6 months,respectively. In contrast, the 1-month NRS scores of subjects who received the placebotreatment were unchanged from baseline (6.4 ± 1.9 and 6.3 ± 2.4, respectively; p>0.9). Nofurther within groups analysis was performed because of insufficient patients remaining in theplacebo group at 3 (n=2) and 6-month (n=0) time points.Eleven subjects in the placebo arm crossed over to the radiofrequency treatment; 9 crossedover at 1-month and 2 at 3-months. In the crossover phase of this trial, the placebo group’sNRS scores after conventional radiofrequency treatment did not significantly differ from thoseof the initial radiofrequency group (3.6 ± 2.6 vs. 2.4 ± 2.0, respectively). Similar to the initialtreatment group, the placebo crossover group experienced a significant decrease in NRS scores1 (44%), 3 (67%) and 6 (52%) months after denervation when compared to baseline (p<0.001,table 2).Secondary Outcome MeasuresOswestry Disability Index—A significant difference in ODI was detected betweentreatment and placebo groups. One month after the procedure, the treatment group had lowerODI scores than the placebo group (20.9 ± 10.9; range 4–38 vs. 43.6 ± 14.0; range 16–70,respectively; p<0.03). In a within group analysis, subjects who received radiofrequencytreatment reported significantly lower ODI scores at 1, 3 and 6 months when compared to theirbaseline scores (p<0.001, table 3). Subjects’ ODI scores were reduced by 44%, 50%, 39% at1, 3 and 6 months, respectively. In contrast, the mean 1 month ODI score of subjects whoreceived the placebo treatment was unchanged from baseline (43.6 vs. 47.9 ± 9.3; range 28–59, respectively).In the crossover phase, the placebo group’s ODI scores after radiofrequency treatment werereduced by 28%, 59% and 49% at 1-, 3- and 6-months post-procedure, respectively. ODI scoresin the crossover group did not significantly differ from those of the initial treatment group 3or 6 months after the procedure. However, the initial radiofrequency treatment group hadsignificantly lower ODI scores at 1 month compared to the placebo/crossover group (20.9 vs.34.3 ± 16.3; range 4–58, respectively; p<0.03). The difference between baseline and post-procedure ODI scores in the crossover group was statistically significant 3 and 6-months afterthe conventional radiofrequency procedure (p<0.02), but not 1-month following denervation.Global Perceived Effect (GPE)—Subjects who received radiofrequency treatmentreported a significantly higher proportion of positive GPE responses at 1 month compared tosubjects who received placebo treatment (93% vs. 21%, respectively; p<0.001). The percentageof subjects in the treatment group (n=14) with a positive GPE was 93% (n=13), 71% (n=10)and 50% (n=7) at 1, 3 and 6 months, respectively (table 4). In the crossover phase, thepercentage of subjects who underwent conventional denervation (n=11) with a positive GPEwas 72% (n=8), 64% (n=7) and 46% (n=5) at 1, 3 and 6-months post-procedure, respectively.The crossover group’s GPE proportion after treatment did not significantly differ from thoseof the initial radiofrequency group.Medication Reduction—The radiofrequency treatment group had a significantly higherproportion of patients able to reduce their analgesic medications following the procedure at 1month compared to subjects who received placebo treatment (77%; 10/14 vs. 8%; 1/11,respectively; p<0.001). The percentage of subjects in the radiofrequency group who were ableto reduce their analgesic intake was 77% (n=10), 64% (n=9) and 36% (n=5) at 1, 3 and 6months, respectively (table 5). In the crossover phase (n=11), the percentage of subjects whoCohen et al. Page 6Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

reported a decrease in medication requirements were 73% (n=8), 46% (n=5) and 27% (n=3)at 1, 3 and 6 months, respectively. The reduction in analgesic intake between the original andcrossover radiofrequency groups was not statistically different.Percent Successful Treatment—The proportion of subjects who experienced a “positiveoutcome” was significantly higher in the denervation group than control group (p<0.001). Thissuccess rate persisted at 3- and 6-month follow-up visits (fig. 4). In contrast, only 2 (14.3%)subjects in the placebo group experienced a positive composite outcome 1-month post-procedure, and none obtained relief exceeding 3 months.Placebo subjects who crossed over to conventional radiofrequency treatment (n=11)experienced slightly lower success rates than the original treatment group (7 (64%) versus 11(79%) at 1 month, 6 (55%) versus 9 (64%) at 3 months, and 4 (36%) versus 8 (57%) at 6 months,respectively). However, the proportion of successful procedures in the crossover group wasnot statistically different than in the initial radiofrequency treatment group.Duration of Pain Relief—Subjects in the treatment group had a mean duration of pain reliefof 5.8 (± 4.2; range 0–12) months vs. 0.7 (±1.6; range 0–1) months in the placebo group. Themean duration of relief in the radiofrequency crossover group did not significantly differ fromthat of the initial treatment group (4.6 ± 4.6; range 0–12 vs. 5.8 ± 4.2 months, respectively).Two patients each in the cooled and conventional radiofrequency groups continued toexperience significant pain relief 1 year after treatment. Among patients with a successfuloutcome at any time point, the mean duration of pain relief was 7.9 ± 4.7 months.Adequacy of Blinding—A disinterested observer querying patients before discharge fromtheir procedure assessed the adequacy of blinding. In the 14 patients in the radiofrequencygroup, 9 thought they received denervation, 2 thought they received placebo treatment and 3were unable to guess which group they were randomized to despite prodding. In the 14 placebopatients, 8 believed they received radiofrequency denervation, 3 felt they received the placebotreatment and 3 were unsure which group they were randomized. In the radiofrequency group,the 11 successful outcomes at 1-month were comprised of 8 patients who thought they receiveddenervation, one who felt he received placebo treatment, and 2 patients who were unsure whichgroup they were allocated to. Both successful outcomes in the placebo-arm at 1-month thoughtthey received denervation.Complications—A majority of patients reported temporary worsening pain typically lastingbetween 5 and 10 days after the procedure, which was attributed to both procedure-related painand/or temporary neuritis, the latter which may be attenuated by preemptive corticosteroidadministration.19 However, there were no serious complications reported for either the 14placebo or 25 radiofrequency treatments. In the radiofrequency treatment group, one patientreported transient non-painful buttock paresthesias that resolved without therapy.DiscussionThe results of this placebo-controlled study provide preliminary evidence that radiofrequencydenervation of the L4 and L5 primary dorsal rami and S1-3 lateral branches may providesignificant pain relief and functional improvement in carefully selected patients with suspectedsacroiliac joint pain. At 1-, 3- and 6-months post-procedure, 79%, 64% and 57% of patients,respectively, obtained ≥ 50% pain relief and clinically relevant functional improvement.The high success rate in this study may be partially explained by the combination of stringentinclusion criteria employed and several innovations over previously described sacroiliac jointdenervation techniques. First, rather than targeting individual nerves, this techniqueCohen et al. Page 7Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

endeavored to lesion a continuous volume of tissue lateral to the S1-3 foramina. The rationalefor this approach is based on a recent cadaveric study demonstrating a complex arcade of smallnerve fibers anastamosing with multiple primary dorsal rami around each foramina.12 Whileindividual branch location was shown to vary from level to level and specimen to specimen,they all course through a finite volume of tissue between the lateral edge of the foramen andjoint. By placing electrodes strategically around the foramen, this finite volume of tissue canbe heated to neuroablative temperatures, thus severing all nociceptive input converging on theprimary dorsal ramus. If single lesions had been used as in previously published studies,10–12 some of the afferent input from the sacroiliac joint would likely have remained intact.Creating strip lesions has been previously advocated for sacroiliac joint lesioning,13 but weredescribed using smaller electrodes. Since there is a direct correlation between lesion size andelectrode diameter,20 the use of small electrodes increases the likelihood of inadvertentlysparing neural input. The probability of 3 geometrically-configured lesions failing to coalescewas further reduced by the use of a water-cooled electrode. Internal cooling enhances lesionsize by removing the constraint of high temperature charring in tissue adjacent to the electrode,thus allowing effective ionic heating at a greater distance.21This study was not powered or designed to detect a difference between outcomes or durationof benefit in patients who underwent denervation with the 17-gauge water-cooled system andthose who were treated with the conventional 22-gauge needles, but the slightly higher successrate in the former group (albeit in a non-randomized comparison) despite a lower inferredplacebo response is consistent with pre-clinical and clinical data supporting larger lesions forradiofrequency denervation.22 Typically, reported success rates in open-label studies tend tobe higher than in controlled studies using similar techniques. This issue needs to be examinedin a subsequent randomized trial to determine whether lesion size is an important factor in thesuccess after radiofrequency denervation.Since we elected for ethical reasons to treat our placebo-controlled patients with conventionaldenervation at their 1-month follow-up, one can only speculate about any long-term differencesbetween the treatment and placebo arms. The rationale for this decision was based on pilot dataexamined before embarking on this study that determined the chances of someone obtaininglong-term benefit if none was experienced 1-month postprocedure to be exceedingly low.Finally, our main inclusion criterion of ≥ 75% pain relief after a single diagnostic SI jointinjection was stricter than that used in some prior studies.10–13 This relatively high inclusionthreshold may have contributed to our high success rate. Thus, caution must be heeded whenextrapolating these results to conditions wherein less rigorous selection criteria are employed.In a prevalence study conducted in 43 patients with low back pain below L5-S1, Schwarzer etal.23 found that 30% obtained ≥ 75% pain relief following low-volume sacroiliac jointinfiltration. Since the intent of this trial was to examine the therapeutic benefit of this technique,the use of strict inclusion criteria was deemed justified in order to limit the number of patientswho did not have true sacroiliac joint-related pain (i.e. “false-positives”), thereby enhancingthe internal validity of the trial. Once the beneficial effects of treatment are established,subsequent trials can be conducted under less rigorous conditions in order to better assessexternal validity.Five of 14 (36%) patients in the treatment arm and 5 of 11 (45%) in the open-label crossovergroup failed to obtain significant improvement 3-months after the procedure. There are severalexplanations for this including a short-lived placebo-response to the diagnostic block but notthe definitive treatment, the high false-positive rate associated with single sacroiliac jointblocks,24 and the fact that the L4 thru S3 primary dorsal rami do not supply all the innervationto the sacroiliac joint.8 In the first two scenarios, the use of double confirmatory diagnosticsacroiliac blocks done with 2 different local anesthetics might reduce the failure rate. In theCohen et al. Page 8Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

latter case, performing prognostic lateral branch blocks might screen out those patients whosepain emanates from a part(s) of the sacroiliac joint not innervated by the targeted dorsal ramibranches. In the two studies whereby both sacroiliac joint and lateral branch blocks were usedto screen radiofrequency treatment candidates, the authors reported identical 9-month successrates of 89%.10,13One disappointing finding is that the high success rate and more aggressive lesion size realizedwith cooled radiofrequency did not translate into a longer duration of pain relief. Similar tostudies conducted with conventional radiofrequency technology,1,9,25 the duration of benefitseems to be constrained by nerve regeneration to between 6 months and one year. Future studiesshould address whether refinements in technique (e.g. creating bipolar lesions), and/orselection criteria (e.g. examining pain referral patterns, and the use of controlled sacroiliacjoint or prognostic lateral branch blocks) can influence the success rate or duration of painrelief, and what the long-term consequences of repeat denervation(s) are.One criticism that might be made levied against this study is our decision to target five levelsfor lesioning. The innervation of the sacroiliac joint is a subject of great contention. Whereassome experts have cited contributions to the superior aspect of the joint from as high as L4,1,26 other investigators have failed to confirm these findings.27 Branches derived from the L4and L5 dorsal rami may ostensibly innervate not only the sacroiliac joint and surroundingligaments, but also paraspinal muscles, the L5-S1 zygapophysial joint, and the inferior pole ofthe L4–5 zygapophysial joints as well.9 Although screening sacroiliac joint blocks wereperformed on all our patients, the specificity of diagnostic spinal injections is inherently low.28,29 In particular, uncontrolled sacroiliac joint blocks are associated with a high false-positiverate.24,30 Whether a less aggressive lesioning scheme targeting fewer levels would yieldsimilar results is something that should be addressed in future clinical trials.There are several flaws in this study that need to be addressed. First, although the power analysisdesigned to detect significant differences between the groups was borne out by positive results,the small number of patients enrolled in this study had the unintended consequence of creatingtwo groups of patients with potentially clinically significant differences in several variablesknown to influence outcome, including baseline functionality (i.e. ODI scores), prior spinesurgery, and disability or worker’s compensation cases.15,31,32 Recruiting more patients ina bi-center pilot study to redress these inequities when a beneficial effect for the studiedtreatment has purportedly been proved would undermine the goodwill of subjects who werepaid nothing for their participation. Large multi-center studies, which are needed to confirmour preliminary results, should be adequately powered to address these issues.The small numbers of patients enrolled also leaves unresolved questions regarding the safetyof cooled radiofrequency. Fourteen patients is an insufficient number to detect the small butclinically significant risk of a neurological complication, which may be magnified by the moreambitious lesioning scheme used here. Caution should thus be heeded until large numbers ofpatients are safely treated by multiple clinicians.A second flaw revolves around our testing of blinding adequacy. The effectiveness of blindingin this study was evaluated shortly after the conclusion of the procedure, when the effects ofthe local anesthetic were still active. A more valid indicator of the adequacy of blinding mighthave been to query patients several days after the procedure, when the cues of actual treatment(e.g. procedure-related pain) were more manifest.In summary, the results of this placebo-controlled study provide preliminary support for theuse of radiofrequency denervation to treat presumptive sacroiliac joint pain. Larger, multi-center studies with long-term follow-up and comprehensive outcome measures are needed toCohen et al. Page 9Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

confirm our findings, further establish safety, and determine how best to identify candidatesfor this treatment.AcknowledgementsFunded in part by the John P. Murtha Neuroscience and Pain Institute, Johnstown, Pennsylvania and the Army RegionalAnesthesia & Pain Medicine Initiative, Washington, DC (SPC and CCB), and National Institutes of Health grant #MH075884 (RWH), Bethesda, Maryland.References1. Cohen SP. Sacroiliac joint pain: A comprehensive review of anatomy, diagnosis, and treatment. AnesthAnalg 2005;101:1440–1453. [PubMed: 16244008]2. Dreyfuss P, Dreyer SJ, Cole A, Mayo K. Sacroiliac joint pain. J Am Acad Orthop Surg 2004;12:255–265. [PubMed: 15473677]3. Luukkainen R, Nissila M, Asikainen E, Sanila M, Lehtinen K, Alanaatu A, Kautiainen H. Periarticularcorticosteroid treatment of the sacroiliac joint in patients with seronegative spondyloarthropathy. ClinExp Rheumatol 1999;17:88–90. [PubMed: 10084038]4. Luukkainen R, Wennerstrand PV, Kautiainen HH, Sanila MT, Asikainen EL. Efficacy of periarticularcorticosteroid treatment of the sacroiliac joint in non-spondyloarthropathic patients with chronic lowback pain in the region of the sacroiliac joint. Clin Exp Rheumatol 2002;20:52–54. [PubMed:11892709]5. Maugars Y, Mathis C, Vilon P, Prost A. Corticosteroid injection of the sacroiliac joint in patients withseronegative spondylarthropathy. Arthritis Rheum 1992;35:564–568. [PubMed: 1575790]6. Hanly JG, Mitchell M, MacMillan L, Mosher D, Sutton E. Efficacy of sacroiliac corticosteroidinjections in patients with inflammatory spondyloarthropathy: results of a 6 month controlled study.J Rheumatol 2000;27:719–722. [PubMed: 10743815]7. Fischer T, Biedermann T, Hermann KG, Diekmann F, Braun J, Hamm B, Bollow M. Sacroiliitis inchildren with spondyloarthropathy: therapeutic effect of CT-guided intra-articular corticosteroidinjection. Rofo 2003;175:814–821. [PubMed: 12811695](in German)8. Cohen SP. Epidemics, evolution and sacroiliac joint pain. Reg Anesth Pain Med 2007;32:3–6.[PubMed: 17196485]9. Cohen SP, Raja SN. Pathogenesis, diagnosis and treatment of lumbar zygapophysial (facet) joint pain.Anesthesiology 2007;106:591–614. [PubMed: 17325518]10. Cohen SP, Abdi S. Lateral branch blocks as a treatment for sacroiliac joint pain: a pilot study. RegAnesth Pain Med 2003;28:113–119. [PubMed: 12677621]11. Buijs EJ, Kamphuis ET, Groen GJ. Radiofrequency treatment of sacroiliac joint-related pain aimedat the first three sacral dorsal rami: a minimal approach. Pain Clinic 2004;16:139–146.12. Yin W, Willard F, Carreiro J, Dreyfuss P. Sensory stimulation-guided sacroiliac joint radiofrequencyneurotomy: technique based on neuroanatomy of the dorsal sacral plexus. Spine 2003;28:2419–2425.[PubMed: 14560094]13. Burnham RS, Yasui Y. An alternate method of radiofrequency neurotomy of the sacroiliac joint: Apilot study of the effect on pain, function and satisfaction. Reg Anesth Pain Med 2007;32:12–19.[PubMed: 17196487]14. van Kleef M, Barendse GA, Kessels A, Voets HM, Weber WE, de Lange S. Randomized trial ofradiofrequency lumbar facet denervation for chronic low back pain. Spine 1999;24:1937–1942.[PubMed: 10515020]15. Cohen SP, Hurley RW, Christo PJ, Winkley J, Mohiuddin MM, Stojanovic MP. Clinical predictorsof success and failure for lumbar facet radiofrequency denervation. Clin J Pain 2007;23:45–52.[PubMed: 17277644]16. Cohen SP, Wenzell D, Hurley RW, Kurihara C, Buckenmaier CC 3rd, Griffith S, Larkin TM, DahlE, Morlando BJ. Intradiscal etanercept as a treatment for discogenic low back pain and sciatica: adouble-blind, placebo-controlled, dose-response pilot study. Anesthesiology 2007;107:99–105.[PubMed: 17585221]Cohen et al. Page 10Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

17. Hagg O, Fritzell P, Nordwall A. Swedish Lumbar Spine Study Group: The clinical importance ofchanges in outcome scores after treatment for chronic low back pain. Eur Spine J 2003;12:12–20.[PubMed: 12592542]18. Meade TW, Dyer S, Browne W, Townsend J, Frank AO. Low back pain of mechanical origin:randomised comparison of chiropractic and hospital outpatient treatment. BMJ 1990;300:1431–1437. [PubMed: 2143092]19. Dobrogowski J, Wrzosek A, Wordliczek J. Radiofrequency denervation with or without addition ofpentoxifylline or methylprednisolone for chronic lumbar zygapophysial joint pain. Pharmacol Rep2005;57:475–480. [PubMed: 16129914]20. Bogduk N, Macintosh J, Marsland A. Technical limitations to the efficacy of radiofrequencyneurotomy for spinal pain. Neurosurgery 1987;20:529–534. [PubMed: 2953988]21. de Baere T, Denys A, Wood BJ, Lassau N, Kardache M, Vilgrain V, Menu Y, Roche A.Radiofrequency liver ablation: experimental comparative study of water-cooled versus expandablesystems. AJR Am J Roentgenol 2001;176:1213–1215. [PubMed: 11312184]22. Walach H, Sadaghiani C, Dehm C, Bierman D. The therapeutic effect of clinical trials: understandingplacebo response rates in clinical trials--a secondary analysis. BMC Med Res Methodol 2005;5:26.[PubMed: 16109176]23. Schwarzer AC, Aprill CN, Bogduk N. The sacroiliac joint in chronic low back pain. Spine1995;20:31–37. [PubMed: 7709277]24. Maigne JY, Aivaliklis A, Pfefer F. Results of sacroiliac joint double block and value of sacroiliacpain provocation tests in 54 patients with low back pain. Spine 1996;21:1889–1892. [PubMed:8875721]25. Lord SM, Barnsley L, Wallis BJ, McDonald GJ, Bogduk N. Percutaneous radio-frequency neurotomyfor chronic cervical zygapophyseal-joint pain. N Engl J Med 1996;335:1721–1726. [PubMed:8929263]26. Bernard, TN.; Cassidy, JD. The sacroiliac syndrome. In: Frymoyer, JW., editor. Pathophysiology,diagnosis and management, The Adult Spine: Principles and Practice. New York: Raven; 1991. p.2107-2130.27. Grob KR, Neuhuber WL, Kissling RO. Innervation of the sacroiliac joint in humans. Zeitschrift furRheumatologie 1995;54:117–122. [PubMed: 7793158](in German)28. North RB, Kidd DH, Zahurak M, Piantadosi S. Specificity of diagnostic nerve blocks: a prospective,randomized study of sciatica due to lumbosacral spine disease. Pain 1996;65:77–85. [PubMed:8826493]29. Cohen SP, Hurley RW. The ability of diagnostic spinal injections to predict surgical outcomes. AnesthAnalg 2007;105:1756–1775. [PubMed: 18042881]30. Irwin RW, Watson T, Minick RP, Ambrosius WT. Age, body mass index, and gender differences insacroiliac joint pathology. Am J Phys Med Rehabil 2007;86:37–44. [PubMed: 17304687]31. Underwood MR, Morton V, Farrin A. UK BEAM Trial Team: Do baseline characteristics predictresponse to treatment for low back pain? Secondary analysis of the UK BEAM dataset. Rheumatology2007;46:1297–1302. [PubMed: 17522096]32. Vaccaro AR, Ring D, Scuderi G, Cohen DS, Garfin SR. Predictors of outcome in patients with chronicback pain and low-grade spondylolisthesis. Spine 1997;22:2030–2034. [PubMed: 9306535]Cohen et al. Page 11Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Figure 1.CONSORT chart showing progression of subjects in study arms.Footnotes: SIJ- sacroiliac joint, RF- radiofrequency, N- number of patientsCohen et al. Page 12Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Figure 2.Schematic diagram illustrating:A. Target points for right-sided conventional (L4 and L5) and cooled (S1–3) radiofrequencydenervation at the junction of the L5 superior articular and transverse processes (L4 primarydorsal ramus), the sacral ala (L5 primary dorsal ramus), and S1-3 foramina (lateral branches).B. Anticipated lesions at each of the target points.Footnotes: L4 and L5- 4th and 5th lumbar spinal levels, respectively. S1–3- First, second andthird sacral spinal levels, respectively.Cohen et al. Page 13Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Figure 3.Adjacent photographs demonstrating the difference in lesion size between cooled (A) andconventional (B) radiofrequency probes in chicken meat. Each small line represents a distanceof 1 mm.Cohen et al. Page 14Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

Figure 4.Bar graph demonstrating the percentage of patients with a successful treatment outcome atvarious time points. A positive outcome is defined as a ≥ 50% reduction in numerical painscore, a positive global perceived effect, and either a 10-point decrease in Oswestry disabilityindex score or a 4-point decrease coupled with a reduction in medication usage.Footnote: TX- treatmentCohen et al. Page 15Anesthesiology. Author manuscript; available in PMC 2009 August 1.NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptCohen et al. Page 16Table 1Demographic and Clinical Characteristics of Study Patients Placebo (n=14) Lateral Branch Denervation(n=14)Sex  Male (n=11) 6 (43%) 5 (36%)  Female (n=17) 8 (57%) 9 (64%)Age (SD, range) 51.8 (13.1; 31–74) 51.9 (13.6; 27–75)Active Duty (n=6) 3 (21%) 3 (21%)Opioid Use (n=13) and Dosage in Morphine 7 (50%) 6 (43%)Equivalents per Day (mean, SD, range) 46.4 (43.1, 7.5–130) 60 (50.0, 7.5–150)Worker’s Compensation, Disability, or Military Medical BoardClaim (n=9) 3 (21%) 6 (43%)*Failed Back Surgery Syndrome (n=6)** 4 (29%) 2 (14%)Baseline Numerical Rating Scale Score (SD, 6.5 (1.9; 3.5–10) 6.1 (1.8; 3–8)range) Median (interquartile range) 6 (5.5–7) 6 (5–8)Baseline Oswestry Disability Index (SD, 47.9 (9.3; 28–58) 37.1 (10.6; 18–49)range) 50.5 (44–56) 41 (26–46)Median (interquartile range)Continuous data listed as the mean and (standard deviation, range) and median and (interquartile range, 25%–75%), categorical data as number and(percentage).*Includes 3 active duty soldiers undergoing a medical board.**Includes 5 patients with spinal fusion and 1 status post-laminectomy.Anesthesiology. Author manuscript; available in PMC 2009 August 1.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptCohen et al. Page 17Table 2Numerical Rating Pain Scores Stratified by Treatment Group and Time PointTime point Placebo (n=14) Lateral BranchDenervation (n=14) Lateral BranchDenervation Crossover(n=11)Baseline  Mean (SD, range) 6.5 (1.9, 3.5–10) 6.1 (1.8, 3–8) 6.3 (2.4, 2–10)  Median (interquartile range) 6 (5.5–7) 6 (5–8) 6 (4–7)One month *, ** *, **  Mean (SD, range) 6.3 (2.4, 2–10) 2.4 (2.0, 0–8) 3.6 (2.6, 0–10)  Median (interquartile range) 7 (4–7) 2 (1–3) 3 (2–5)Three months (n=2) ** **  Mean (SD, range) 6 (0, 6–6) 2.4 (2.3, 0–7) 2.1 (2.4, 0–7)  Median (interquartile range) 6 (6–6) 1.5 (1–4.5) 1.5 (0.5–3)Six months ** **  Mean (SD, range) (no data) 2.6 (2.2, 0–7) 3.1 (2.1, 0–6)  Median (interquartile range) 2 (1.5–2.5) 3.5 (1.5–4)*P<0.05 as compared to placebo group**P<0.05 as compared to baseline of the respective groupAnesthesiology. Author manuscript; available in PMC 2009 August 1.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptCohen et al. Page 18Table 3Oswestry Disability Index Score (%) Stratified by Treatment Group and Time PointTime point Placebo Lateral BranchDenervation (n=14) Lateral BranchDenervation Crossover(n=11)Baseline (n=14)  Mean (SD, range) 47.9 (9.3, 28–59) 37.1 (10.6, 18–49) 43.6 (14, 16–70)  Median (interquartile range) 50.5 (44–56) 41, (26–46) 41 (34–56)One month (n=14) *, **, ***  Mean (SD, range) 43.6 (14, 16–70) 20.9 (10.9, 4–38) 34.3 (16.2, 4–58)  Median (interquartile range) 41 (34–56) 19 (14–29) 33 (24–46)Three months (n=2) *** ***  Mean (SD, range) 24 (8.5, 18–30) 18.5 (11.6, 0–36) 19.4 (18.1, 0–44)  Median (interquartile range) 24 (18–30) 20 (9.5–27) 16 (4–44)Six months *** ***  Mean (SD, range) No data 22.6 (10.6, 7–40) 24.3 (21.0, 0–56)  Median (interquartile range) 20 (16–24) 20 (8–42)*P<0.05 as compared to placebo group**P<0.05 as compared to lateral branch denervation crossover group***P<0.05 as compared to baseline of the respective groupAnesthesiology. Author manuscript; available in PMC 2009 August 1.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptCohen et al. Page 19Table 4Percent Positive Global Perceived Effect Stratified by Treatment Group and Time PointTime point Placebo Lateral Branch Denervation(n=14) Lateral Branch DenervationCrossover (n=11)One month (n=14) *  Percent (95% Cl) 21 (2–45) 93 (78–100) 72 (41–100)Three months (n=2)  Percent (95% Cl) 0 83 (59–100) 86 (51–100)Six months  Percent (95% Cl) No data 89 (63–100) 89 (63–100)*P<0.05 as compared to placebo groupAnesthesiology. Author manuscript; available in PMC 2009 August 1.

NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptCohen et al. Page 20Table 5Positive Percent Medication Reduction Stratified by Treatment Group and Time PointTime point Placebo Lateral Branch Denervation(n=14) Lateral Branch DenervationCrossover (n=11)One month (n=14) * *  Percent (95% Cl) 8 (0–25) 77 (52–100) 78 (44–100)Three months (n=2)  Percent (95% Cl) 0 82 (55–100) 82 (55–100)Six months  Percent (95% Cl) No data 67 (28–100) 60 (0–100)*P<0.05 as compared to placebo groupAnesthesiology. Author manuscript; available in PMC 2009 August 1.