bs_bs_banner
Pain Medicine 2014; 15: 196–205 Wiley Periodicals, Inc.
SPINE SECTION Review Article The mild® Procedure: A Systematic Review of the Current Literature D. Scott Kreiner, MD,* John MacVicar, MB ChB, MPainMed,† Belinda Duszynski, BS,‡ and Devi E. Nampiaparampil, MD§ *Ahwatukee Sport and Spine, Phoenix, Arizona; ‡ International Spine Intervention Society, San Rafael, California;
§ NYU School of Medicine, Anesthesiology and Rehabilitation Medicine, New York, New York, USA;
†
St Georges Medical Centre, Southern Rehabilitation Institute, Christchurch, New Zealand Reprint requests to: Belinda Duszynski, BS, International Spine Intervention Society, 161 Mitchell Boulevard, Suite 103, San Rafael, CA 94903, USA. Tel: 415-457-4747; Fax: 415-457-3495; E-mail: [email protected]. Disclosures: D. S. K.: None. J. M.: None. B. D.: None. D. N.: Expert witness for Kline Specter—one time; ProStrakan Medical Research Advisory Panel in June 2011—one time.
Abstract Objectives. This study’s objective was to determine if the literature supports use of the Minimally Invasive Lumbar Decompression (mild®) procedure (Vertos Medical, Aliso Viejo, CA, USA) to reduce pain and improve function in patients with symptomatic degenerative lumbar spinal stenosis. Design/Settings. The study was designed as an evidence-based review of available data. Studies were identified from PubMed, Embase, and the Cochrane Library. Articles were evaluated using the Grading of Recommendations Assessment, Development and Evaluation Working Group system. Results were compiled assessing short196
(4–6 weeks), medium- (3–6 months), and long-term (>1 year) outcomes. The primary outcomes evaluated were pain, measured by the visual analog scale (VAS), and function, measured by the Oswestry Disability Index (ODI). Secondary outcomes included pain and patient satisfaction, measured by the Zurich Claudication Questionnaire, adverse effects/ complications, and changes in utilization of co-interventions. Results. The literature search revealed one randomized controlled trial (RCT) and 12 other studies (seven prospective cohort, four retrospective, and one case series) that provided information on the use of mild® in patients with degenerative lumbar spinal stenosis. All studies showed statistically significant improvements in VAS and ODI scores at all time frames compared with preprocedure levels; the RCT showed improvement over controls. Categorical data were not provided; thus, the proportion of patients who experienced minimal clinically meaningful outcomes is unknown. Conclusion. The current body of evidence addressing mild® is of low quality. High-quality studies that are independent of industry funding and provide categorical data are needed to clarify the proportions of patients who benefit from mild® and the degree to which these patients benefit. Additional data at up to 2 years are needed to determine the overall utility of the procedure. Key Words. Percutaneous Decompression; Spinal Stenosis; Lumbar; Neurogenic Claudication; Intermittent Claudication; Interventional
Introduction Lumbar spinal stenosis, or narrowing of the central spinal canal, is associated with increasing age. The prevalence of acquired lumbar spinal stenosis is 4% in patients under the age of 40, but increases to 19.4% by the age of 60–69 [1]. Although narrowing of the spinal canal is not necessarily symptomatic, by the age of 70, approximately 10% of the population develops symptoms related to lumbar stenosis [2]. Narrowing of the canal causes a variable
The mild® Procedure: A Systematic Review clinical syndrome of gluteal pain, lower extremity pain, and fatigue, which may occur with or without back pain. Symptomatic lumbar spinal stenosis most typically is associated with exercise or positionally induced neurogenic claudication, and symptoms are typically relieved with forward flexion, sitting, and/or recumbency [3]. In degenerative lumbar spinal stenosis, the cause of narrowing in the canal is often multifactorial. Anteriorly, disc degeneration leads to disc bulging and disc-osteophyte complexes projecting posteriorly into the canal. Facet arthrosis and hypertrophy of the ligamentum flavum can narrow the lateral recess in the canal, and the canal may be further narrowed by degenerative spondylolisthesis. Surgical and nonsurgical treatment options exist for symptomatic lumbar spinal stenosis. Nonsurgical treatment options include analgesic medications, physical therapy [4], manipulation [5], and epidural steroid injections [6–8]. Of these conservative treatment options, a multiple injection regimen seems to be the most effective [8–10]. The most common surgical option for degenerative lumbar spinal stenosis is the laminectomy procedure. This procedure, first performed by Dr. Victor Alexander Haden Horsley in 1887, is known to be an effective treatment for degenerative lumbar spinal stenosis [11–13]. In 2004, an interspinous spacer, known as the X-STOP (Medtronic, Minneapolis, MN, USA), was described as a new option to treat degenerative lumbar spinal stenosis [14,15], although its true effectiveness has been challenged [16]. Most recently, a new technique named the mild® (Minimally Invasive Lumbar Decompression; Vertos Medical) procedure has been described [17] to treat degenerative lumbar spinal stenosis. This procedure involves inserting a cannula through a six-gauge portal and using tissue and bone sculptors to perform a minimal laminotomy and resect the hypertrophied ligamentum flavum in order to decompress the affected dural sac or nerve roots. This procedure is performed using fluoroscopic guidance with a contralateral oblique view epidurogram to visualize depth of the instruments to maintain safety. Members of the Standards Division of the International Spine Intervention Society (ISIS) performed a preliminary review of the literature on the mild® procedure in 2012. The initial motivation was to assist members of the ISIS Health Policy Division in deciding whether or not to recommend mild® for a Category I CPT Code. The current systematic review was performed in order to evaluate the state of the evidence on the procedure. Methods Search Strategy We conducted a comprehensive literature search in PubMed, Embase, and the Cochrane Library using the search terms: lumbar stenosis, percutaneous decompression, and mild® procedure. Only publications dealing
directly with the mild® procedure were retrieved. Studies describing other percutaneous decompression systems were excluded. We identified all relevant full-length articles in any language that were published through May 2013. The reference lists of the selected articles were searched. Study Selection Inclusion criteria were 1) study participants with symptomatic lower extremity claudication attributed to lumbar spinal stenosis and confirmed with computed tomography or magnetic resonance imaging, 2) subjects older than 18 years, 3) patients assigned to treatment with mild®, and 4) assessment of at least one predefined outcome measure. Individual studies may have excluded subjects based on additional criteria. Outcome Measures The primary outcome measure was lower extremity pain relief as measured using the visual analog scale (VAS) or numeric rating scale. Secondary outcomes included functional benefit measured using the Oswestry Disability Index (ODI), pain and patient satisfaction as measured by the Zurich Claudication Questionnaire (ZCQ), complications and adverse effects, and the utilization of co-interventions. These outcomes were assessed at various intervals including 30 days, 6 weeks, 3 months, 6 months, and 1 year or longer. Analysis The studies were then individually evaluated for quality using an instrument developed by the ISIS Standards Division to facilitate reliable assessment of studies of therapeutic effectiveness. Three authors reviewed all of the articles. Disagreements were resolved through discussion. A Microsoft® Excel spreadsheet and the GradePro software (GRADE Working Group) were then used to consolidate the data from the different studies into workable follow-up periods of short (4–6 weeks), medium (3–6 months), and long term (1 year or more). Results The literature search identified a single randomized controlled trial (RCT), seven prospective cohort studies, four retrospective cohort studies, a single case series, and a meta-analysis. For our purposes, the meta-analysis was discarded, but all references were considered for inclusion in this review. The technical and design features of the other studies are described in Table 1. Mekhail et al. stated that patients with symptomatic low back pain were included; however, elsewhere in the study they indicated that patients included experienced symptomatic lower extremity pain [18]. For this reason, we included the study of Mekhail in this review. Pain In this systematic review, we assessed pain with two outcome measures: the VAS score and the ZCQ overall 197
Kreiner et al.
Table 1
Article summary
Study
Design
Inclusion Criteria
N
F/U
Outcome Measures
Brown [20]
Prospective RCT
39
Prospective cohort
27
6 weeks, 12 weeks (patients allowed to crossover after 6 weeks) 6 months
VAS ODI ZCQ
Basu [25]
DLSS w/ LFH Neurogenic claudication ODI > 20 ≤grade I spondy DLSS w/ LFH Neurogenic claudication
Chopko and Caraway [21]
Prospective cohort
78
6 weeks
Chopko [19]
Prospective Cohort Prospective cohort
14
4–72 weeks
58
2-year data from Mekhail [18]
Chopko [29]
Deer and Kapural [17] Deer et al. [24]
Retrospective
Durkin et al. [23]
Retrospective
Mekhail et al. [26]
Prospective cohort
Mekhail et al. [18]
Prospective cohort
Lingreen and Grider [22]
Prospective cohort
Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx ≤5 mm spondy Symptomatic DLSS High risk for open spine surgery Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx
90 Symptomatic DLSS LF > 2.5 mm Walk > 10 ft Failed conservative tx
VAS ODI ZCQ VAS ODI ZCQ
VAS ODI VAS ODI ZCQ SF-12 not reported Complications
46
12 weeks, 6 months, and 1 year
VAS ODI ZCQ
50
1, 3, and 6 months
58
1 year
40
1 year
Retrospective
42
30 days
Wang et al. [30]
Retrospective
22
Variable
Wilkinson and Fourney [27]
Prospective Cohort
10
26 weeks (poststudy monitor to 18 months)
NRS ODI ZCQ VAS ODI ZCQ SF-12 PDI RMQ Walking Distance VAS VAS Walk > 15 minutes Stand > 15 minutes Time spent in specialty care Number of interventional procedures VAS VAS ODI ASG SF-12
Wong [28]
Case Series
17
1 year
Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx Symptomatic DLSS LF > 4.0 mm Walk > 10 ft Failed conservative tx
Symptomatic DLSS LF > 2.5 mm ≤3 mm spondy Walk > 10 ft Failure conservative tx
VAS ODI
ASG = Analgesic Severity Grade; DLSS = degenerative lumbar spinal stenosis; F/U = follow-up; LF = ligamentum flavum; LFH = ligamentum flavum hypertrophy; ODI = Oswestry Disability Index; PDI = pain disability index; RCT = randomized controlled trial; SF-12 = Short Form 12; VAS = visual analog scale; .ZCQ = Zurich Claudication Questionnaire.
198
The mild® Procedure: A Systematic Review
Table 2
Mean improvements in VAS scores Base
4–6 Weeks
12 Weeks
6 Months
1 Year
VAS
Imp (%)
VAS
Imp (%)
VAS
Imp (%)
VAS
Imp (%)
3.9
57
3.8 3.7
40 49
3.4
46 41 29
4.2 5.6
39 25
42
31 40
4.1 5.3
4.0
5.2 5.8
57 41
3.7 4.2
48 42
4.5 3.6
39 49
46
4.4
42
2.3 3.9
70 49
Study
Pats
VAS
Basu [25] Brown [20] Chopko and Caraway [21] Deer et al. [24] Durkin et al. [23] Lingreen and Grider [22] Mekhail et al. [26] Mekhail et al. [18] Wilkinson and Fourney [27] Wong [28] Weighted mean
27 21 78 46 50 42 58 40 10 17
9.1 6.3 7.3 6.9 7.5 9.6 7.4 7.1 7.3 7.6 7.6
3.7
49
3.1 4.3
4.5
41
4.1
VAS = visual analog scale.
symptom severity score. The scores are divided into time frames for consistency across studies. For this reason, we have opted to omit the study of Chopko [19] from our review because the data reported vary in duration of follow-up from 4 to 72 weeks (Table 2). Short Term Five studies assessed pain at 4–6 weeks using the VAS score and demonstrated a weighted mean improvement of 41% (4.5 points). The strongest data are provided by Brown’s RCT of 39 patients [20]. This study showed a statistically significant reduction in the VAS from 6.3 (95% confidence interval [CI] ± 0.7) to 3.8 (95% CI ± 1.3) compared with the control epidural steroid injection group, which had negligible improvement in the VAS over the same follow-up period [20]. Chopko performed a study of 75 patients, which showed a statistically significant decrease in the average VAS score from 7.3 (range 3–10) to 3.7 (range 0–10) [21]. This study also found a statistically significant reduction in the ZCQ overall symptom severity score from 3.69 (range 1.57–5) to 2.35 (range 1–4.57). It is unclear why the author chose to provide a score range instead of a CI or standard deviation (SD). Lingreen and Grider’s retrospective study of 42 patients also found a statistically significant improvement [22]. The VAS score decreased from 9.6 ± 0.42 to 5.8 ± 2.5. The authors did not report whether these ranges represent SDs or 95% CIs. Similarly, Durkin’s retrospective review of 50 mild® cases showed a VAS score change from 7.5 (95% CI ± 0.48) at baseline to 5.18 (95% CI ± 0.75) [23]. This study also reported improvements in the National Institutes of Health Patient Reported Outcomes Measurement Information System pain interference scores from 63.39 (95% CI ± 2.55) to 58.53 (95% CI ± 3.36).
retrospective study. The weighted mean improvement in pain was 46% (4.1 points) at 3 months and 42% (4.4 points) at 6 months. Brown’s study started as an RCT, but by 12 weeks all patients had crossed to the mild® group, relegating the study to providing cohort data [20]. This study showed that the VAS significantly decreased from 6.3 (95% CI ± 0.7) to an average of 3.4, although no SD or CI was provided for the 12-week data points. Deer’s 2012 study of 46 patients showed a statistically significant decrease in the VAS score from 6.9 (95% CI ± 0.6) to 4.2 (95% CI ± 1.0) at 12 weeks and 4.4 (95% CI ± 1.0) at 6 months [24]. Basu studied 27 patients and found that the VAS improved from 9.1 (95% CI ± 0.59) to 3.9 (95% CI ± 2.25) [25]. This study also assessed pain at 3–6 months using the ZCQ overall symptom severity score. The authors did not provide the quantitative data, but they reported that the pain and neuroischemic domains were significantly improved at P < 0.001. Mekhail et al. used VAS as a secondary outcome measure and reported scores at multiple time frames in his 2012 study [26]. The study provides graphical data on VAS improvements from 7.1 at baseline to 3.1 at 12-week follow-up and 3.7 at 6-month follow-up, although no SDs or 95% CIs were provided for this data set. Wilkinson et al. studied 10 patients and found that the VAS improved from 7.3 (±SD 1.5) to 4.3 at 12 weeks and to 4.2 (±SD 2.8) at 6 months [27]. Durkin reported VAS improvements from 7.5 (95% CI ± 0.48) at baseline to 5.31 (95% CI ± 0.95) at 3 months and 5.60 (95% CI ± 0.98) at 6 months. Long Term
Medium Term Six studies were included which provided medium-term results: five prospective cohort studies and a single
Five studies assessed pain at 12 months or greater using the VAS score. The weighted mean improvement in pain was 49% (3.9 points) at 1 year. 199
Kreiner et al. Mekhail et al.’s prospective observational study of 58 patients showed a reduction in the VAS from 7.4 (95% CI ± 0.5) to 4.5 (95% CI ± 0.8) (P < 0.0001) [18]. This study also evaluated pain using the ZCQ overall symptom severity score and found a significant improvement in the score from 4.54 to 2.88. SDs and CIs for the baseline measures were not provided, but the SD of the change in ZCQ was given as 0.34. Mekhail et al.’s 2012 study revealed a baseline VAS of 7.1 (95% CI ± 0.8), which improved to 3.6 (95% CI ± 0.9) at 1 year (P < 0.0001) [26]. Deer’s prospective cohort study also reported a statistically significant improvement in the VAS from 6.9 (95% CI ± 0.6) to 4.0 (95% CI ± 1.0) at 1 year (P < 0.0001) [24]. Wong et al. conducted a retrospective case series of 17 patients and found a VAS improvement from 7.6 to 2.3 [28]. SDs and CIs for the baseline measures were not provided, but the SD of the change in VAS was given as 1.5. Chopko’s 2013 study [29] reported 2-year data on Mekhail et al.’s [18] 58 patients, but 13 of the initial study patients were unavailable: three underwent lumbar spine surgery after 1 year, one died, and the remaining nine did not respond to multiple contacts or withdrew from the study. This author decided to only evaluate the cohort of remaining 45 patients independently. In these patients, there was a statistically significant improvement in the VAS from 7.2 (95% CI ± 0.6) to 4.8 (95% CI ± 0.8) (P < 0.0001) at 2-year follow-up, but it was not felt that this study provided valid information about the effectiveness of the procedure because of the exclusion of 22% of the initial patients. Function Ten studies (one RCT, five prospective cohorts, and four retrospective studies) assessed function after the mild® procedure. Nine of the 10 studies used the ODI as their primary functional outcome measure, while Mekhail used the Roland-Morris Disability Questionnaire (RMQ), standing time (ST), and walking distance (WD) (Table 3).
Table 3
Four studies assessed short-term functional improvement, revealing an early weighted mean improvement in the ODI of 16.5. Again, the strongest data are provided by Brown’s RCT [20]. This study showed an improvement in the ODI from 38.8 to 27.4, an 11.4-point improvement (95% CI ± 8.2), while there was no statistically significant improvement in the control epidural steroid injection group. Chopko and Caraway’s 2010 study also reported a statistically significant improvement in the ODI, which improved from 47.4 (range 16–84) to 29.5 (range 0–72) [21]. Durkin reported a baseline ODI of 40.58 (95% CI ± 4.5), which improved slightly to 32.53 (95% CI ± 4.7) (P 0.0002) [23]. Wilkinson also published early results, with a baseline ODI of 49.4 that improved to roughly 35 at 6 weeks [27]. These data are provided in graphical form with SD tics, but numerical data for all intermediate assessments are not provided. This change in ODI was noted not to be statistically significant (P < 0.05). Medium Term Five studies report functional data at the 3–6-month mark, showing weighted mean improvements in the ODI of 16.2 at 3 months and 15.4 at 6 months. Again, Brown’s study started as an RCT, but by 12 weeks all patients had crossed over to the mild® group [20]. This study was reported to show a statistically significant improvement in the ODI from 38.8 to 25.5 at week 12, although no SDs or CIs were provided for the 12-week data. Basu reported a statistically significant improvement in the ODI from 55.1 (95% CI ± 6.34) to 31.1 (95% CI ± 9.29) [25]. Deer reported a baseline ODI of 49.4 (95% CI ± 2.5), which improved to 35.1 (95% CI ± 5.6) at 12 weeks and 35.0 (95% CI ± 5.5) at 6 months, both of which were statistically significant (P < 0.01) [24]. Durkin’s retrospective review demonstrated a baseline ODI of 40.58 (95% CI ± 4.5), which improved to 32.22 (95% CI ± 5.1) (P = 0.0002) at 3 months and 29.24 (95% CI ± 6.8) (P = 0.0004) at 6 months [23]. Lastly, Wilkinson reported
Mean improvements in ODI scores Base
6 Weeks ODI
Study
Pats
ODI
Basu [25] Brown [20] Chopko and Caraway [21] Deer et al. [24] Durkin et al. [23] Lingreen and Grider [22] Mekhail et al. [26] Wilkinson and Fourney [27] Wong [28] Weighted mean
27 21 78 46 50 42 58 10 17
55.1 38.8 47.4 49.4 40.6
ODI = Oswestry Disability Index.
200
Short Term
48.6 49.4 48.4 47.0
12 Weeks
6 Months
1 Year
Imp
ODI
Imp
ODI
Imp
ODI
Imp
31.1
24.0
27.4 29.5
11.4 17.9
18.6
20.2 14.3 8.4
35.0 29.2
14.4 11.4
16.9
8.1
35.1 32.2
32.5
32.5
36.7
11.9
35.0
14.4
29.5
19.9
29.4
20.0
30.5
16.5
30.8
16.2
31.6
15.4
21.7 33.0
26.7 14.0
The mild® Procedure: A Systematic Review that the baseline ODI of 49.4 (95% CI ± 13.9) showed a statistically significant improvement to 29.4 (95% CI ± 19.8) at 26-week follow-up [27]. Long Term Three studies looked at long-term disability improvements, demonstrating a weighted mean improvement in the ODI of 14.0 at 1 year. In Mekhail et al.’s study, the baseline average ODI score of 48.6 (95% CI ± 3.8) decreased to a mean of 36.7 (95% CI ± 5.8) at 1-year follow-up, an improvement of 11.9 points (95% CI ± 4.7) [18]. Wong reported that, in his patients, the average baseline ODI of 48.4 improved to 21.7 at 1-year follow-up [28]. The mean ODI decreased 26.6 points from baseline to 1-year follow-up with a 95% CI of ±8.8 (−17.9 to −35.4). Chopko’s [29] 2-year extension of the Mekhail et al. study [18] showed improvements in the ODI with a baseline of 48.4 (95% CI ± 4.4) to 39.8 (95% CI ± 5.6) at 2 years. Again, these data are of questionable validity, as 13 patients were not available for follow-up and at least two proceeded to lumbar surgery after 1 year. Mekhail et al.’s 2012 study demonstrated an improvement in the RMQ from 14.3 (95% CI ± 2.1) to 6.6 (95% CI ± 2.0) (P < 0.0001) at 1 year [26]. This study also demonstrated an improvement in ST from 8 to 56 minutes (P < 0.0001) and an improvement in the WD from a baseline mean of 246–3,956 ft at 12 months (P < 0.0001). Complications Nearly every study assessed patients for complications including dural punctures or tears, nerve root injuries, bleeding, infections, and rehospitalization postprocedure. No substantial direct procedure-related complications were identified. The second of Chopko’s studies identified two patients with complications [19]. One with known lymphoma experienced a deep venous thrombosis and pulmonary embolism (PE) on post-op day one and a recurrent PE at 7 months. The second patient was obese and had experienced recurrent abdominal issues. This patient developed an incarcerated small bowel within 48 hours of the procedure. This study also reported that two patients died of complications related to systemic malignancies at 4 and 8 weeks postoperatively. Lingreen and Grider’s study noted that of the 42 patients, 20 had soreness at the site, four reported gluteal pain, one had bleeding at the puncture site, and one experienced “back spasms” [22].
Co-Interventions Chopko’s 2011 study evaluated changes in usage of pain medication preprocedure and postprocedure [19]. Of the 12 patients who were using opioid-based pain medication prior to the mild® procedure, six had either decreased or completely stopped their usage at a time interval between 4 and 72 weeks. Durkin’s retrospective review noted that 30% of patients were using opiate medications prior to the procedure [23]. Following the procedure, 55% of patients remained opioid free. Of the others, 12% no longer required opioids, 6% reduced opioid-usage, 10% used the same amount as before, but one subject (2%) required higher doses, and 10% required initiation of opioid therapy. Wilkinson and Fourney’s study also evaluated the use of pain medication [27]. Interestingly, preoperatively no patients were taking opioid-based analgesia. Following the procedure, the use of nonopioid analgesics decreased over the study period, although the use of opioid analgesics was introduced. The usage of opioid-based analgesics peaked at 1-week postprocedure and then decreased at the 6-week, 12-week, and seemingly the 26-week assessments. In addition, the authors continued their study past the original 6 months. They found that by 18 months post-mild®, 6 of 10 patients had undergone spine surgery. While the authors felt that surgical intervention represented a failure of the mild® procedure, no VAS or ODI measures were provided either before or after the surgery. For this reason, it is difficult to draw any conclusions from this finding. Wang et al. retrospectively reviewed the charts of 22 patients who had undergone the mild® procedure [30]. They compared the amount of time spent in specialty care and the number of interventional procedures performed before and after mild®. Two patients could not tolerate the mild® procedure; one was lost to follow-up. Twelve out of 22 patients (54.5%) including the patient lost to follow-up were discharged from the chronic pain clinic after mild®. They continued to be seen in the primary care clinic for residual symptoms. Ten patients continued their care in the chronic pain clinic. Of those 10 patients, seven received additional interventional procedures. Three were referred for surgical decompression. As a group, the patients received an average of 0.395 pain procedures per month prior to mild®, and as a group, they received 3 at all time frames, in all studies except for that of Wong [28], and were >4 in half of the studies. Mean final ODI was similarly >30 in the majority of studies, and a mean ODI improvement of >20 was reported in only two studies [25,28]. These figures suggest that minimally acceptable outcomes were not achieved for the average patient in the majority of studies, but the number of patients who benefited to a clinically significant degree from treatment, and the extent to which they might have benefited, is not known because continuous data (mean VAS and ODI scores) were provided before and after treatment, rather than the categorical data. The relevance of the short-term outcome measures is debatable. The mild® procedure cannot be repeated regularly like epidural steroid injections. However, it is not as invasive or risk laden as a laminectomy. A procedure that is being compared with a lumbar laminectomy should be effective for at least 6 months. The 6-week and 12-week data that have been collected are not useful in this regard. However, the 6- and 12-week data are useful in demonstrating an expected course of recovery. These data suggest that the mild® procedure is more effective than epidural steroid injections in the treatment of symptomatic lumbar stenosis. In the case of the mild® procedure, weighted mean VAS scores post treatment are 4.5 at 6 weeks, 4.1 at 12 weeks, 4.4 at 6 months, and 3.9 at 1 year. Mean weighted ODI scores are improved by 16.5 at 6 weeks, 16.2 at 12 weeks, 15.4 at 6 months, and 14.0 at 1 year. These VAS and ODI scores do not meet Carragee’s minimum acceptable outcomes. Opioid medication usage is not consistently documented in the available studies, and in the studies available the data are conflicting. In reviewing the body of evidence, the authors felt that there were no significant reasons to justify upgrading or downgrading the body of evidence based upon the Grading of Recommendations Assessment, Development and Evaluation Working Group system. The majority of the evidence is derived from observational studies with only a
The mild® Procedure: A Systematic Review single RCT and, with no justification to either upgrade or downgrade the quality, the current body of evidence addressing the mild® procedure is of low quality. Limitations Industry funding of various degrees was received to perform these studies. Most of this industry involvement was well-disclosed; however, the Chopko and Caraway study inaccurately reported that the lead author had no conflicts, but he later acknowledged he was a paid consultant for the manufacturer while the study was being conducted [21]. In addition, the Mekhail et al. article lacks an appropriate disclosure statement so it is unclear whether the author is paid directly by the manufacturer [18]. These issues raise the suspicion of bias toward the procedure. While it is expected that the manufacturer will provide the equipment gratis to enable the study to be done, the lead investigator in the majority of these studies was a trainer for the procedure and/or a paid investigator or consultant to the manufacturer. This degree of industry-funded involvement brings a substantial risk of bias to the relevant studies. Conclusion The current body of evidence addressing the mild® procedure is of low quality. The mild® procedure appears to be a relatively safe procedure for the treatment of symptomatic lumbar spinal stenosis. Outcome data suggest that the procedure provides statistically significant reductions in pain intensity and statistically significant improvements in function, but the data, which are continuous rather than categorical, suggest that these improvements do not meet some definitions of a minimally acceptable outcome. Future studies must provide categorical data that will clarify the proportions and characteristics of patients who are most likely to benefit from the mild® procedure and the degree to which these patients benefit. Additional studies, independent of industry funding and involvement, are needed. Data at time frames greater than 1 year, which are clearly lacking, and additional data at up to 2 years, would be beneficial in determining the overall utility of the procedure. Ideally, predictive factors could be identified to help differentiate which patients are most likely to have a positive or negative response to this procedure. References 1 Kalichman L, Cole R, Kim DH, et al. Spinal stenosis prevalence and association with symptoms: The Framingham Study. Spine J 2009;9(7):545–50. PubMed PMID: 19398386. 2 Ishimoto Y, Yoshimura N, Muraki S, et al. Prevalence of symptomatic lumbar spinal stenosis and its association with physical performance in a populationbased cohort in Japan: The Wakayama Spine Study. Osteoarthritis Cartilage 2012;20(10):1103–8. PubMed PMID: 22796511.
3 Watters WC 3rd, Baisden J, Gilbert TJ, et al. Degenerative lumbar spinal stenosis: An evidence-based clinical guideline for the diagnosis and treatment of degenerative lumbar spinal stenosis. Spine J 2008;8(2):305–10. PubMed PMID: 18082461. 4 Goren A, Yildiz N, Topuz O, Findikoglu G, Ardic F. Efficacy of exercise and ultrasound in patients with lumbar spinal stenosis: A prospective randomized controlled trial. Clin Rehabil 2010;24(7):623–31. PubMed PMID: 20530650. 5 Murphy DR, Hurwitz EL, Gregory AA, Clary R. A nonsurgical approach to the management of lumbar spinal stenosis: A prospective observational cohort study. BMC Musculoskelet Disord 2006;7:16. PubMed PMID: 16504078. Pubmed Central PMCID: 1397818. 6 Lee JW, Myung JS, Park KW, et al. Fluoroscopically guided caudal epidural steroid injection for management of degenerative lumbar spinal stenosis: Short-term and long-term results. Skeletal Radiol 2010;39(7):691–9. PubMed PMID: 20033148. 7 Manchikanti L, Cash KA, McManus CD, et al. Lumbar interlaminar epidural injections in central spinal stenosis: Preliminary results of a randomized, double-blind, active control trial. Pain Physician 2012;15(1):51–63. PubMed PMID: 22270738. 8 Manchikanti L, Cash KA, McManus CD, Pampati V, Abdi S. Preliminary results of a randomized, equivalence trial of fluoroscopic caudal epidural injections in managing chronic low back pain: Part 4–Spinal stenosis. Pain Physician 2008;11(6):833–48. PubMed PMID: 19057629. 9 Botwin K, Brown LA, Fishman M, Rao S. Fluoroscopically guided caudal epidural steroid injections in degenerative lumbar spine stenosis. Pain Physician 2007;10(4):547–58. PubMed PMID: 17660853. 10 Riew KD, Yin Y, Gilula L, et al. The effect of nerve-root injections on the need for operative treatment of lumbar radicular pain. A prospective, randomized, controlled, double-blind study. J Bone Joint Surg Am 2000;82-A(11):1589–93. PubMed PMID: 11097449. 11 Athiviraham A, Yen D. Is spinal stenosis better treated surgically or nonsurgically? Clin Orthop Relat Res 2007;458:90–3. PubMed PMID: 17308483. 12 Slatis P, Malmivaara A, Heliovaara M, et al. Long-term results of surgery for lumbar spinal stenosis: A randomised controlled trial. Eur Spine J 2011;20(7): 1174–81. PubMed PMID: 21240530. Pubmed Central PMCID: 3175822. 13 Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical versus nonoperative treatment for lumbar spinal 203
Kreiner et al. stenosis four-year results of the Spine Patient Outcomes Research Trial. Spine 2010;35(14):1329–38. PubMed PMID: 20453723. Pubmed Central PMCID: 3392200. 14 Zucherman JF, Hsu KY, Hartjen CA, et al. A prospective randomized multi-center study for the treatment of lumbar spinal stenosis with the X STOP interspinous implant: 1-year results. Eur Spine J 2004;13(1):22–31. PubMed PMID: 14685830. 15 Zucherman JF, Hsu KY, Hartjen CA, et al. A multicenter, prospective, randomized trial evaluating the X STOP interspinous process decompression system for the treatment of neurogenic intermittent claudication: Two-year follow-up results. Spine 2005;30(12):1351–8. PubMed PMID: 15959362. 16 Patil S, Burton M, Storey C, et al. Evaluation of interspinous process distraction device (X-STOP) in a representative patient cohort. World Neurosurg 2013; 80(102):213–7. PubMed PMID: 22484765. Epub 2012/04/10. Eng.
the MILD procedure. Pain Med 2013;14(5):650–6. PubMed PMID: 23489390. 24 Deer TR, Kim CK, Bowman RG 2nd, Ranson MT, Yee BS. Study of percutaneous lumbar decompression and treatment algorithm for patients suffering from neurogenic claudication. Pain Physician 2012;15(6): 451–60. PubMed PMID: 23159960. 25 Basu S. Mild procedure: Single-site prospective IRB study. Clin J Pain 2012;28(3):254–8. PubMed PMID: 21926907. 26 Mekhail N, Costandi S, Abraham B, Samuel SW. Functional and patient-reported outcomes in symptomatic lumbar spinal stenosis following percutaneous decompression. Pain Pract 2012;12(6): 417–25. PubMed PMID: 22651852. 27 Wilkinson JS, Fourney DR. Failure of percutaneous remodeling of the ligamentum flavum and lamina for neurogenic claudication. Neurosurgery 2012;71(1): 86–92. PubMed PMID: 22407072.
17 Deer TR, Kapural L. New image-guided ultra-minimally invasive lumbar decompression method: The mild procedure. Pain Physician 2010;13(1):35–41. PubMed PMID: 20119461.
28 Wong WH. mild Interlaminar decompression for the treatment of lumbar spinal stenosis: Procedure description and case series with 1-year follow-up. Clin J Pain 2012;28(6):534–8. PubMed PMID: 22673487.
18 Mekhail N, Vallejo R, Coleman MH, Benyamin RM. Long-term results of percutaneous lumbar decompression mild® for spinal stenosis. Pain Pract 2012;12(3):184–93. PubMed PMID: 21676166.
29 Chopko BW. Long-term results of percutaneous lumbar decompression for LSS: Two-year outcomes. Clin J Pain 2013;29(11):939–43. PubMed PMID: 23446067.
19 Chopko BW. A novel method for treatment of lumbar spinal stenosis in high-risk surgical candidates: Pilot study experience with percutaneous remodeling of ligamentum flavum and lamina. J Neurosurg Spine 2011;14(1):46–50. PubMed PMID: 21142460.
30 Wang JJ, Bowden K, Pang G, Cipta A. Decrease in health care resource utilization with MILD. Pain Med 2013;14(5):657–61. PubMed PMID: 23578021.
20 Brown LL. A double-blind, randomized, prospective study of epidural steroid injection vs. the mild® procedure in patients with symptomatic lumbar spinal stenosis. Pain Pract 2012;12(5):333–41. PubMed PMID: 22272730. 21 Chopko B, Caraway DL. MiDAS I (mild Decompression Alternative to Open Surgery): A preliminary report of a prospective, multi-center clinical study. Pain Physician 2010;13(4):369–78. PubMed PMID: 20648206. 22 Lingreen R, Grider JS. Retrospective review of patient self-reported improvement and post-procedure findings for mild (minimally invasive lumbar decompression). Pain Physician 2010;13(6):555–60. PubMed PMID: 21102968. 23 Durkin B, Romeiser J, Shroyer AL, et al. Report from a quality assurance program on patients undergoing 204
31 Tumialan LM, Marciano FF, Theodore N. Regarding: Long-term results of percutaneous lumbar decompression mild for spinal stenosis. Pain Pract 2012;12(3): 252; author reply -3. PubMed PMID: 22390527. 32 Hagg O, Fritzell P, Nordwall A. The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J 2003;12(1):12–20. PubMed PMID: 12592542. 33 van der Roer N, Ostelo RW, Bekkering GE, van Tulder MW, de Vet HC. Minimal clinically important change for pain intensity, functional status, and genera l health status in patients with nonspecific low back pain. Spine 2006;31(5):578–82. PubMed PMID: 16508555. 34 Giraudeau B, Rozenberg S, Valat JP. Assessment of the clinically relevant change in pain for patients with sciatica. Ann Rheum Dis 2004;63(9):1180–1. PubMed PMID: 15308536. Pubmed Central PMCID: PMC1755131.
The mild® Procedure: A Systematic Review 35 Carragee EJ, Lincoln T, Parmar VS, Alamin T. A gold standard evaluation of the “discogenic pain” diagnosis as determined by provocative discography. Spine 2006;31(18):2115–23. PubMed PMID: 16915099.
36 Carragee EJ, Cheng I. Minimum acceptable outcomes after lumbar spinal fusion. Spine J 2010;10(4):313–20. PubMed PMID: 20362247.
205
Pain Medicine 2014; 15: 196–205 Wiley Periodicals, Inc.
SPINE SECTION Review Article The mild® Procedure: A Systematic Review of the Current Literature D. Scott Kreiner, MD,* John MacVicar, MB ChB, MPainMed,† Belinda Duszynski, BS,‡ and Devi E. Nampiaparampil, MD§ *Ahwatukee Sport and Spine, Phoenix, Arizona; ‡ International Spine Intervention Society, San Rafael, California;
§ NYU School of Medicine, Anesthesiology and Rehabilitation Medicine, New York, New York, USA;
†
St Georges Medical Centre, Southern Rehabilitation Institute, Christchurch, New Zealand Reprint requests to: Belinda Duszynski, BS, International Spine Intervention Society, 161 Mitchell Boulevard, Suite 103, San Rafael, CA 94903, USA. Tel: 415-457-4747; Fax: 415-457-3495; E-mail: [email protected]. Disclosures: D. S. K.: None. J. M.: None. B. D.: None. D. N.: Expert witness for Kline Specter—one time; ProStrakan Medical Research Advisory Panel in June 2011—one time.
Abstract Objectives. This study’s objective was to determine if the literature supports use of the Minimally Invasive Lumbar Decompression (mild®) procedure (Vertos Medical, Aliso Viejo, CA, USA) to reduce pain and improve function in patients with symptomatic degenerative lumbar spinal stenosis. Design/Settings. The study was designed as an evidence-based review of available data. Studies were identified from PubMed, Embase, and the Cochrane Library. Articles were evaluated using the Grading of Recommendations Assessment, Development and Evaluation Working Group system. Results were compiled assessing short196
(4–6 weeks), medium- (3–6 months), and long-term (>1 year) outcomes. The primary outcomes evaluated were pain, measured by the visual analog scale (VAS), and function, measured by the Oswestry Disability Index (ODI). Secondary outcomes included pain and patient satisfaction, measured by the Zurich Claudication Questionnaire, adverse effects/ complications, and changes in utilization of co-interventions. Results. The literature search revealed one randomized controlled trial (RCT) and 12 other studies (seven prospective cohort, four retrospective, and one case series) that provided information on the use of mild® in patients with degenerative lumbar spinal stenosis. All studies showed statistically significant improvements in VAS and ODI scores at all time frames compared with preprocedure levels; the RCT showed improvement over controls. Categorical data were not provided; thus, the proportion of patients who experienced minimal clinically meaningful outcomes is unknown. Conclusion. The current body of evidence addressing mild® is of low quality. High-quality studies that are independent of industry funding and provide categorical data are needed to clarify the proportions of patients who benefit from mild® and the degree to which these patients benefit. Additional data at up to 2 years are needed to determine the overall utility of the procedure. Key Words. Percutaneous Decompression; Spinal Stenosis; Lumbar; Neurogenic Claudication; Intermittent Claudication; Interventional
Introduction Lumbar spinal stenosis, or narrowing of the central spinal canal, is associated with increasing age. The prevalence of acquired lumbar spinal stenosis is 4% in patients under the age of 40, but increases to 19.4% by the age of 60–69 [1]. Although narrowing of the spinal canal is not necessarily symptomatic, by the age of 70, approximately 10% of the population develops symptoms related to lumbar stenosis [2]. Narrowing of the canal causes a variable
The mild® Procedure: A Systematic Review clinical syndrome of gluteal pain, lower extremity pain, and fatigue, which may occur with or without back pain. Symptomatic lumbar spinal stenosis most typically is associated with exercise or positionally induced neurogenic claudication, and symptoms are typically relieved with forward flexion, sitting, and/or recumbency [3]. In degenerative lumbar spinal stenosis, the cause of narrowing in the canal is often multifactorial. Anteriorly, disc degeneration leads to disc bulging and disc-osteophyte complexes projecting posteriorly into the canal. Facet arthrosis and hypertrophy of the ligamentum flavum can narrow the lateral recess in the canal, and the canal may be further narrowed by degenerative spondylolisthesis. Surgical and nonsurgical treatment options exist for symptomatic lumbar spinal stenosis. Nonsurgical treatment options include analgesic medications, physical therapy [4], manipulation [5], and epidural steroid injections [6–8]. Of these conservative treatment options, a multiple injection regimen seems to be the most effective [8–10]. The most common surgical option for degenerative lumbar spinal stenosis is the laminectomy procedure. This procedure, first performed by Dr. Victor Alexander Haden Horsley in 1887, is known to be an effective treatment for degenerative lumbar spinal stenosis [11–13]. In 2004, an interspinous spacer, known as the X-STOP (Medtronic, Minneapolis, MN, USA), was described as a new option to treat degenerative lumbar spinal stenosis [14,15], although its true effectiveness has been challenged [16]. Most recently, a new technique named the mild® (Minimally Invasive Lumbar Decompression; Vertos Medical) procedure has been described [17] to treat degenerative lumbar spinal stenosis. This procedure involves inserting a cannula through a six-gauge portal and using tissue and bone sculptors to perform a minimal laminotomy and resect the hypertrophied ligamentum flavum in order to decompress the affected dural sac or nerve roots. This procedure is performed using fluoroscopic guidance with a contralateral oblique view epidurogram to visualize depth of the instruments to maintain safety. Members of the Standards Division of the International Spine Intervention Society (ISIS) performed a preliminary review of the literature on the mild® procedure in 2012. The initial motivation was to assist members of the ISIS Health Policy Division in deciding whether or not to recommend mild® for a Category I CPT Code. The current systematic review was performed in order to evaluate the state of the evidence on the procedure. Methods Search Strategy We conducted a comprehensive literature search in PubMed, Embase, and the Cochrane Library using the search terms: lumbar stenosis, percutaneous decompression, and mild® procedure. Only publications dealing
directly with the mild® procedure were retrieved. Studies describing other percutaneous decompression systems were excluded. We identified all relevant full-length articles in any language that were published through May 2013. The reference lists of the selected articles were searched. Study Selection Inclusion criteria were 1) study participants with symptomatic lower extremity claudication attributed to lumbar spinal stenosis and confirmed with computed tomography or magnetic resonance imaging, 2) subjects older than 18 years, 3) patients assigned to treatment with mild®, and 4) assessment of at least one predefined outcome measure. Individual studies may have excluded subjects based on additional criteria. Outcome Measures The primary outcome measure was lower extremity pain relief as measured using the visual analog scale (VAS) or numeric rating scale. Secondary outcomes included functional benefit measured using the Oswestry Disability Index (ODI), pain and patient satisfaction as measured by the Zurich Claudication Questionnaire (ZCQ), complications and adverse effects, and the utilization of co-interventions. These outcomes were assessed at various intervals including 30 days, 6 weeks, 3 months, 6 months, and 1 year or longer. Analysis The studies were then individually evaluated for quality using an instrument developed by the ISIS Standards Division to facilitate reliable assessment of studies of therapeutic effectiveness. Three authors reviewed all of the articles. Disagreements were resolved through discussion. A Microsoft® Excel spreadsheet and the GradePro software (GRADE Working Group) were then used to consolidate the data from the different studies into workable follow-up periods of short (4–6 weeks), medium (3–6 months), and long term (1 year or more). Results The literature search identified a single randomized controlled trial (RCT), seven prospective cohort studies, four retrospective cohort studies, a single case series, and a meta-analysis. For our purposes, the meta-analysis was discarded, but all references were considered for inclusion in this review. The technical and design features of the other studies are described in Table 1. Mekhail et al. stated that patients with symptomatic low back pain were included; however, elsewhere in the study they indicated that patients included experienced symptomatic lower extremity pain [18]. For this reason, we included the study of Mekhail in this review. Pain In this systematic review, we assessed pain with two outcome measures: the VAS score and the ZCQ overall 197
Kreiner et al.
Table 1
Article summary
Study
Design
Inclusion Criteria
N
F/U
Outcome Measures
Brown [20]
Prospective RCT
39
Prospective cohort
27
6 weeks, 12 weeks (patients allowed to crossover after 6 weeks) 6 months
VAS ODI ZCQ
Basu [25]
DLSS w/ LFH Neurogenic claudication ODI > 20 ≤grade I spondy DLSS w/ LFH Neurogenic claudication
Chopko and Caraway [21]
Prospective cohort
78
6 weeks
Chopko [19]
Prospective Cohort Prospective cohort
14
4–72 weeks
58
2-year data from Mekhail [18]
Chopko [29]
Deer and Kapural [17] Deer et al. [24]
Retrospective
Durkin et al. [23]
Retrospective
Mekhail et al. [26]
Prospective cohort
Mekhail et al. [18]
Prospective cohort
Lingreen and Grider [22]
Prospective cohort
Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx ≤5 mm spondy Symptomatic DLSS High risk for open spine surgery Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx
90 Symptomatic DLSS LF > 2.5 mm Walk > 10 ft Failed conservative tx
VAS ODI ZCQ VAS ODI ZCQ
VAS ODI VAS ODI ZCQ SF-12 not reported Complications
46
12 weeks, 6 months, and 1 year
VAS ODI ZCQ
50
1, 3, and 6 months
58
1 year
40
1 year
Retrospective
42
30 days
Wang et al. [30]
Retrospective
22
Variable
Wilkinson and Fourney [27]
Prospective Cohort
10
26 weeks (poststudy monitor to 18 months)
NRS ODI ZCQ VAS ODI ZCQ SF-12 PDI RMQ Walking Distance VAS VAS Walk > 15 minutes Stand > 15 minutes Time spent in specialty care Number of interventional procedures VAS VAS ODI ASG SF-12
Wong [28]
Case Series
17
1 year
Symptomatic DLSS LF > 2.55 mm Walk > 10 ft Failure conservative tx Symptomatic DLSS LF > 4.0 mm Walk > 10 ft Failed conservative tx
Symptomatic DLSS LF > 2.5 mm ≤3 mm spondy Walk > 10 ft Failure conservative tx
VAS ODI
ASG = Analgesic Severity Grade; DLSS = degenerative lumbar spinal stenosis; F/U = follow-up; LF = ligamentum flavum; LFH = ligamentum flavum hypertrophy; ODI = Oswestry Disability Index; PDI = pain disability index; RCT = randomized controlled trial; SF-12 = Short Form 12; VAS = visual analog scale; .ZCQ = Zurich Claudication Questionnaire.
198
The mild® Procedure: A Systematic Review
Table 2
Mean improvements in VAS scores Base
4–6 Weeks
12 Weeks
6 Months
1 Year
VAS
Imp (%)
VAS
Imp (%)
VAS
Imp (%)
VAS
Imp (%)
3.9
57
3.8 3.7
40 49
3.4
46 41 29
4.2 5.6
39 25
42
31 40
4.1 5.3
4.0
5.2 5.8
57 41
3.7 4.2
48 42
4.5 3.6
39 49
46
4.4
42
2.3 3.9
70 49
Study
Pats
VAS
Basu [25] Brown [20] Chopko and Caraway [21] Deer et al. [24] Durkin et al. [23] Lingreen and Grider [22] Mekhail et al. [26] Mekhail et al. [18] Wilkinson and Fourney [27] Wong [28] Weighted mean
27 21 78 46 50 42 58 40 10 17
9.1 6.3 7.3 6.9 7.5 9.6 7.4 7.1 7.3 7.6 7.6
3.7
49
3.1 4.3
4.5
41
4.1
VAS = visual analog scale.
symptom severity score. The scores are divided into time frames for consistency across studies. For this reason, we have opted to omit the study of Chopko [19] from our review because the data reported vary in duration of follow-up from 4 to 72 weeks (Table 2). Short Term Five studies assessed pain at 4–6 weeks using the VAS score and demonstrated a weighted mean improvement of 41% (4.5 points). The strongest data are provided by Brown’s RCT of 39 patients [20]. This study showed a statistically significant reduction in the VAS from 6.3 (95% confidence interval [CI] ± 0.7) to 3.8 (95% CI ± 1.3) compared with the control epidural steroid injection group, which had negligible improvement in the VAS over the same follow-up period [20]. Chopko performed a study of 75 patients, which showed a statistically significant decrease in the average VAS score from 7.3 (range 3–10) to 3.7 (range 0–10) [21]. This study also found a statistically significant reduction in the ZCQ overall symptom severity score from 3.69 (range 1.57–5) to 2.35 (range 1–4.57). It is unclear why the author chose to provide a score range instead of a CI or standard deviation (SD). Lingreen and Grider’s retrospective study of 42 patients also found a statistically significant improvement [22]. The VAS score decreased from 9.6 ± 0.42 to 5.8 ± 2.5. The authors did not report whether these ranges represent SDs or 95% CIs. Similarly, Durkin’s retrospective review of 50 mild® cases showed a VAS score change from 7.5 (95% CI ± 0.48) at baseline to 5.18 (95% CI ± 0.75) [23]. This study also reported improvements in the National Institutes of Health Patient Reported Outcomes Measurement Information System pain interference scores from 63.39 (95% CI ± 2.55) to 58.53 (95% CI ± 3.36).
retrospective study. The weighted mean improvement in pain was 46% (4.1 points) at 3 months and 42% (4.4 points) at 6 months. Brown’s study started as an RCT, but by 12 weeks all patients had crossed to the mild® group, relegating the study to providing cohort data [20]. This study showed that the VAS significantly decreased from 6.3 (95% CI ± 0.7) to an average of 3.4, although no SD or CI was provided for the 12-week data points. Deer’s 2012 study of 46 patients showed a statistically significant decrease in the VAS score from 6.9 (95% CI ± 0.6) to 4.2 (95% CI ± 1.0) at 12 weeks and 4.4 (95% CI ± 1.0) at 6 months [24]. Basu studied 27 patients and found that the VAS improved from 9.1 (95% CI ± 0.59) to 3.9 (95% CI ± 2.25) [25]. This study also assessed pain at 3–6 months using the ZCQ overall symptom severity score. The authors did not provide the quantitative data, but they reported that the pain and neuroischemic domains were significantly improved at P < 0.001. Mekhail et al. used VAS as a secondary outcome measure and reported scores at multiple time frames in his 2012 study [26]. The study provides graphical data on VAS improvements from 7.1 at baseline to 3.1 at 12-week follow-up and 3.7 at 6-month follow-up, although no SDs or 95% CIs were provided for this data set. Wilkinson et al. studied 10 patients and found that the VAS improved from 7.3 (±SD 1.5) to 4.3 at 12 weeks and to 4.2 (±SD 2.8) at 6 months [27]. Durkin reported VAS improvements from 7.5 (95% CI ± 0.48) at baseline to 5.31 (95% CI ± 0.95) at 3 months and 5.60 (95% CI ± 0.98) at 6 months. Long Term
Medium Term Six studies were included which provided medium-term results: five prospective cohort studies and a single
Five studies assessed pain at 12 months or greater using the VAS score. The weighted mean improvement in pain was 49% (3.9 points) at 1 year. 199
Kreiner et al. Mekhail et al.’s prospective observational study of 58 patients showed a reduction in the VAS from 7.4 (95% CI ± 0.5) to 4.5 (95% CI ± 0.8) (P < 0.0001) [18]. This study also evaluated pain using the ZCQ overall symptom severity score and found a significant improvement in the score from 4.54 to 2.88. SDs and CIs for the baseline measures were not provided, but the SD of the change in ZCQ was given as 0.34. Mekhail et al.’s 2012 study revealed a baseline VAS of 7.1 (95% CI ± 0.8), which improved to 3.6 (95% CI ± 0.9) at 1 year (P < 0.0001) [26]. Deer’s prospective cohort study also reported a statistically significant improvement in the VAS from 6.9 (95% CI ± 0.6) to 4.0 (95% CI ± 1.0) at 1 year (P < 0.0001) [24]. Wong et al. conducted a retrospective case series of 17 patients and found a VAS improvement from 7.6 to 2.3 [28]. SDs and CIs for the baseline measures were not provided, but the SD of the change in VAS was given as 1.5. Chopko’s 2013 study [29] reported 2-year data on Mekhail et al.’s [18] 58 patients, but 13 of the initial study patients were unavailable: three underwent lumbar spine surgery after 1 year, one died, and the remaining nine did not respond to multiple contacts or withdrew from the study. This author decided to only evaluate the cohort of remaining 45 patients independently. In these patients, there was a statistically significant improvement in the VAS from 7.2 (95% CI ± 0.6) to 4.8 (95% CI ± 0.8) (P < 0.0001) at 2-year follow-up, but it was not felt that this study provided valid information about the effectiveness of the procedure because of the exclusion of 22% of the initial patients. Function Ten studies (one RCT, five prospective cohorts, and four retrospective studies) assessed function after the mild® procedure. Nine of the 10 studies used the ODI as their primary functional outcome measure, while Mekhail used the Roland-Morris Disability Questionnaire (RMQ), standing time (ST), and walking distance (WD) (Table 3).
Table 3
Four studies assessed short-term functional improvement, revealing an early weighted mean improvement in the ODI of 16.5. Again, the strongest data are provided by Brown’s RCT [20]. This study showed an improvement in the ODI from 38.8 to 27.4, an 11.4-point improvement (95% CI ± 8.2), while there was no statistically significant improvement in the control epidural steroid injection group. Chopko and Caraway’s 2010 study also reported a statistically significant improvement in the ODI, which improved from 47.4 (range 16–84) to 29.5 (range 0–72) [21]. Durkin reported a baseline ODI of 40.58 (95% CI ± 4.5), which improved slightly to 32.53 (95% CI ± 4.7) (P 0.0002) [23]. Wilkinson also published early results, with a baseline ODI of 49.4 that improved to roughly 35 at 6 weeks [27]. These data are provided in graphical form with SD tics, but numerical data for all intermediate assessments are not provided. This change in ODI was noted not to be statistically significant (P < 0.05). Medium Term Five studies report functional data at the 3–6-month mark, showing weighted mean improvements in the ODI of 16.2 at 3 months and 15.4 at 6 months. Again, Brown’s study started as an RCT, but by 12 weeks all patients had crossed over to the mild® group [20]. This study was reported to show a statistically significant improvement in the ODI from 38.8 to 25.5 at week 12, although no SDs or CIs were provided for the 12-week data. Basu reported a statistically significant improvement in the ODI from 55.1 (95% CI ± 6.34) to 31.1 (95% CI ± 9.29) [25]. Deer reported a baseline ODI of 49.4 (95% CI ± 2.5), which improved to 35.1 (95% CI ± 5.6) at 12 weeks and 35.0 (95% CI ± 5.5) at 6 months, both of which were statistically significant (P < 0.01) [24]. Durkin’s retrospective review demonstrated a baseline ODI of 40.58 (95% CI ± 4.5), which improved to 32.22 (95% CI ± 5.1) (P = 0.0002) at 3 months and 29.24 (95% CI ± 6.8) (P = 0.0004) at 6 months [23]. Lastly, Wilkinson reported
Mean improvements in ODI scores Base
6 Weeks ODI
Study
Pats
ODI
Basu [25] Brown [20] Chopko and Caraway [21] Deer et al. [24] Durkin et al. [23] Lingreen and Grider [22] Mekhail et al. [26] Wilkinson and Fourney [27] Wong [28] Weighted mean
27 21 78 46 50 42 58 10 17
55.1 38.8 47.4 49.4 40.6
ODI = Oswestry Disability Index.
200
Short Term
48.6 49.4 48.4 47.0
12 Weeks
6 Months
1 Year
Imp
ODI
Imp
ODI
Imp
ODI
Imp
31.1
24.0
27.4 29.5
11.4 17.9
18.6
20.2 14.3 8.4
35.0 29.2
14.4 11.4
16.9
8.1
35.1 32.2
32.5
32.5
36.7
11.9
35.0
14.4
29.5
19.9
29.4
20.0
30.5
16.5
30.8
16.2
31.6
15.4
21.7 33.0
26.7 14.0
The mild® Procedure: A Systematic Review that the baseline ODI of 49.4 (95% CI ± 13.9) showed a statistically significant improvement to 29.4 (95% CI ± 19.8) at 26-week follow-up [27]. Long Term Three studies looked at long-term disability improvements, demonstrating a weighted mean improvement in the ODI of 14.0 at 1 year. In Mekhail et al.’s study, the baseline average ODI score of 48.6 (95% CI ± 3.8) decreased to a mean of 36.7 (95% CI ± 5.8) at 1-year follow-up, an improvement of 11.9 points (95% CI ± 4.7) [18]. Wong reported that, in his patients, the average baseline ODI of 48.4 improved to 21.7 at 1-year follow-up [28]. The mean ODI decreased 26.6 points from baseline to 1-year follow-up with a 95% CI of ±8.8 (−17.9 to −35.4). Chopko’s [29] 2-year extension of the Mekhail et al. study [18] showed improvements in the ODI with a baseline of 48.4 (95% CI ± 4.4) to 39.8 (95% CI ± 5.6) at 2 years. Again, these data are of questionable validity, as 13 patients were not available for follow-up and at least two proceeded to lumbar surgery after 1 year. Mekhail et al.’s 2012 study demonstrated an improvement in the RMQ from 14.3 (95% CI ± 2.1) to 6.6 (95% CI ± 2.0) (P < 0.0001) at 1 year [26]. This study also demonstrated an improvement in ST from 8 to 56 minutes (P < 0.0001) and an improvement in the WD from a baseline mean of 246–3,956 ft at 12 months (P < 0.0001). Complications Nearly every study assessed patients for complications including dural punctures or tears, nerve root injuries, bleeding, infections, and rehospitalization postprocedure. No substantial direct procedure-related complications were identified. The second of Chopko’s studies identified two patients with complications [19]. One with known lymphoma experienced a deep venous thrombosis and pulmonary embolism (PE) on post-op day one and a recurrent PE at 7 months. The second patient was obese and had experienced recurrent abdominal issues. This patient developed an incarcerated small bowel within 48 hours of the procedure. This study also reported that two patients died of complications related to systemic malignancies at 4 and 8 weeks postoperatively. Lingreen and Grider’s study noted that of the 42 patients, 20 had soreness at the site, four reported gluteal pain, one had bleeding at the puncture site, and one experienced “back spasms” [22].
Co-Interventions Chopko’s 2011 study evaluated changes in usage of pain medication preprocedure and postprocedure [19]. Of the 12 patients who were using opioid-based pain medication prior to the mild® procedure, six had either decreased or completely stopped their usage at a time interval between 4 and 72 weeks. Durkin’s retrospective review noted that 30% of patients were using opiate medications prior to the procedure [23]. Following the procedure, 55% of patients remained opioid free. Of the others, 12% no longer required opioids, 6% reduced opioid-usage, 10% used the same amount as before, but one subject (2%) required higher doses, and 10% required initiation of opioid therapy. Wilkinson and Fourney’s study also evaluated the use of pain medication [27]. Interestingly, preoperatively no patients were taking opioid-based analgesia. Following the procedure, the use of nonopioid analgesics decreased over the study period, although the use of opioid analgesics was introduced. The usage of opioid-based analgesics peaked at 1-week postprocedure and then decreased at the 6-week, 12-week, and seemingly the 26-week assessments. In addition, the authors continued their study past the original 6 months. They found that by 18 months post-mild®, 6 of 10 patients had undergone spine surgery. While the authors felt that surgical intervention represented a failure of the mild® procedure, no VAS or ODI measures were provided either before or after the surgery. For this reason, it is difficult to draw any conclusions from this finding. Wang et al. retrospectively reviewed the charts of 22 patients who had undergone the mild® procedure [30]. They compared the amount of time spent in specialty care and the number of interventional procedures performed before and after mild®. Two patients could not tolerate the mild® procedure; one was lost to follow-up. Twelve out of 22 patients (54.5%) including the patient lost to follow-up were discharged from the chronic pain clinic after mild®. They continued to be seen in the primary care clinic for residual symptoms. Ten patients continued their care in the chronic pain clinic. Of those 10 patients, seven received additional interventional procedures. Three were referred for surgical decompression. As a group, the patients received an average of 0.395 pain procedures per month prior to mild®, and as a group, they received 3 at all time frames, in all studies except for that of Wong [28], and were >4 in half of the studies. Mean final ODI was similarly >30 in the majority of studies, and a mean ODI improvement of >20 was reported in only two studies [25,28]. These figures suggest that minimally acceptable outcomes were not achieved for the average patient in the majority of studies, but the number of patients who benefited to a clinically significant degree from treatment, and the extent to which they might have benefited, is not known because continuous data (mean VAS and ODI scores) were provided before and after treatment, rather than the categorical data. The relevance of the short-term outcome measures is debatable. The mild® procedure cannot be repeated regularly like epidural steroid injections. However, it is not as invasive or risk laden as a laminectomy. A procedure that is being compared with a lumbar laminectomy should be effective for at least 6 months. The 6-week and 12-week data that have been collected are not useful in this regard. However, the 6- and 12-week data are useful in demonstrating an expected course of recovery. These data suggest that the mild® procedure is more effective than epidural steroid injections in the treatment of symptomatic lumbar stenosis. In the case of the mild® procedure, weighted mean VAS scores post treatment are 4.5 at 6 weeks, 4.1 at 12 weeks, 4.4 at 6 months, and 3.9 at 1 year. Mean weighted ODI scores are improved by 16.5 at 6 weeks, 16.2 at 12 weeks, 15.4 at 6 months, and 14.0 at 1 year. These VAS and ODI scores do not meet Carragee’s minimum acceptable outcomes. Opioid medication usage is not consistently documented in the available studies, and in the studies available the data are conflicting. In reviewing the body of evidence, the authors felt that there were no significant reasons to justify upgrading or downgrading the body of evidence based upon the Grading of Recommendations Assessment, Development and Evaluation Working Group system. The majority of the evidence is derived from observational studies with only a
The mild® Procedure: A Systematic Review single RCT and, with no justification to either upgrade or downgrade the quality, the current body of evidence addressing the mild® procedure is of low quality. Limitations Industry funding of various degrees was received to perform these studies. Most of this industry involvement was well-disclosed; however, the Chopko and Caraway study inaccurately reported that the lead author had no conflicts, but he later acknowledged he was a paid consultant for the manufacturer while the study was being conducted [21]. In addition, the Mekhail et al. article lacks an appropriate disclosure statement so it is unclear whether the author is paid directly by the manufacturer [18]. These issues raise the suspicion of bias toward the procedure. While it is expected that the manufacturer will provide the equipment gratis to enable the study to be done, the lead investigator in the majority of these studies was a trainer for the procedure and/or a paid investigator or consultant to the manufacturer. This degree of industry-funded involvement brings a substantial risk of bias to the relevant studies. Conclusion The current body of evidence addressing the mild® procedure is of low quality. The mild® procedure appears to be a relatively safe procedure for the treatment of symptomatic lumbar spinal stenosis. Outcome data suggest that the procedure provides statistically significant reductions in pain intensity and statistically significant improvements in function, but the data, which are continuous rather than categorical, suggest that these improvements do not meet some definitions of a minimally acceptable outcome. Future studies must provide categorical data that will clarify the proportions and characteristics of patients who are most likely to benefit from the mild® procedure and the degree to which these patients benefit. Additional studies, independent of industry funding and involvement, are needed. Data at time frames greater than 1 year, which are clearly lacking, and additional data at up to 2 years, would be beneficial in determining the overall utility of the procedure. Ideally, predictive factors could be identified to help differentiate which patients are most likely to have a positive or negative response to this procedure. References 1 Kalichman L, Cole R, Kim DH, et al. Spinal stenosis prevalence and association with symptoms: The Framingham Study. Spine J 2009;9(7):545–50. PubMed PMID: 19398386. 2 Ishimoto Y, Yoshimura N, Muraki S, et al. Prevalence of symptomatic lumbar spinal stenosis and its association with physical performance in a populationbased cohort in Japan: The Wakayama Spine Study. Osteoarthritis Cartilage 2012;20(10):1103–8. PubMed PMID: 22796511.
3 Watters WC 3rd, Baisden J, Gilbert TJ, et al. Degenerative lumbar spinal stenosis: An evidence-based clinical guideline for the diagnosis and treatment of degenerative lumbar spinal stenosis. Spine J 2008;8(2):305–10. PubMed PMID: 18082461. 4 Goren A, Yildiz N, Topuz O, Findikoglu G, Ardic F. Efficacy of exercise and ultrasound in patients with lumbar spinal stenosis: A prospective randomized controlled trial. Clin Rehabil 2010;24(7):623–31. PubMed PMID: 20530650. 5 Murphy DR, Hurwitz EL, Gregory AA, Clary R. A nonsurgical approach to the management of lumbar spinal stenosis: A prospective observational cohort study. BMC Musculoskelet Disord 2006;7:16. PubMed PMID: 16504078. Pubmed Central PMCID: 1397818. 6 Lee JW, Myung JS, Park KW, et al. Fluoroscopically guided caudal epidural steroid injection for management of degenerative lumbar spinal stenosis: Short-term and long-term results. Skeletal Radiol 2010;39(7):691–9. PubMed PMID: 20033148. 7 Manchikanti L, Cash KA, McManus CD, et al. Lumbar interlaminar epidural injections in central spinal stenosis: Preliminary results of a randomized, double-blind, active control trial. Pain Physician 2012;15(1):51–63. PubMed PMID: 22270738. 8 Manchikanti L, Cash KA, McManus CD, Pampati V, Abdi S. Preliminary results of a randomized, equivalence trial of fluoroscopic caudal epidural injections in managing chronic low back pain: Part 4–Spinal stenosis. Pain Physician 2008;11(6):833–48. PubMed PMID: 19057629. 9 Botwin K, Brown LA, Fishman M, Rao S. Fluoroscopically guided caudal epidural steroid injections in degenerative lumbar spine stenosis. Pain Physician 2007;10(4):547–58. PubMed PMID: 17660853. 10 Riew KD, Yin Y, Gilula L, et al. The effect of nerve-root injections on the need for operative treatment of lumbar radicular pain. A prospective, randomized, controlled, double-blind study. J Bone Joint Surg Am 2000;82-A(11):1589–93. PubMed PMID: 11097449. 11 Athiviraham A, Yen D. Is spinal stenosis better treated surgically or nonsurgically? Clin Orthop Relat Res 2007;458:90–3. PubMed PMID: 17308483. 12 Slatis P, Malmivaara A, Heliovaara M, et al. Long-term results of surgery for lumbar spinal stenosis: A randomised controlled trial. Eur Spine J 2011;20(7): 1174–81. PubMed PMID: 21240530. Pubmed Central PMCID: 3175822. 13 Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical versus nonoperative treatment for lumbar spinal 203
Kreiner et al. stenosis four-year results of the Spine Patient Outcomes Research Trial. Spine 2010;35(14):1329–38. PubMed PMID: 20453723. Pubmed Central PMCID: 3392200. 14 Zucherman JF, Hsu KY, Hartjen CA, et al. A prospective randomized multi-center study for the treatment of lumbar spinal stenosis with the X STOP interspinous implant: 1-year results. Eur Spine J 2004;13(1):22–31. PubMed PMID: 14685830. 15 Zucherman JF, Hsu KY, Hartjen CA, et al. A multicenter, prospective, randomized trial evaluating the X STOP interspinous process decompression system for the treatment of neurogenic intermittent claudication: Two-year follow-up results. Spine 2005;30(12):1351–8. PubMed PMID: 15959362. 16 Patil S, Burton M, Storey C, et al. Evaluation of interspinous process distraction device (X-STOP) in a representative patient cohort. World Neurosurg 2013; 80(102):213–7. PubMed PMID: 22484765. Epub 2012/04/10. Eng.
the MILD procedure. Pain Med 2013;14(5):650–6. PubMed PMID: 23489390. 24 Deer TR, Kim CK, Bowman RG 2nd, Ranson MT, Yee BS. Study of percutaneous lumbar decompression and treatment algorithm for patients suffering from neurogenic claudication. Pain Physician 2012;15(6): 451–60. PubMed PMID: 23159960. 25 Basu S. Mild procedure: Single-site prospective IRB study. Clin J Pain 2012;28(3):254–8. PubMed PMID: 21926907. 26 Mekhail N, Costandi S, Abraham B, Samuel SW. Functional and patient-reported outcomes in symptomatic lumbar spinal stenosis following percutaneous decompression. Pain Pract 2012;12(6): 417–25. PubMed PMID: 22651852. 27 Wilkinson JS, Fourney DR. Failure of percutaneous remodeling of the ligamentum flavum and lamina for neurogenic claudication. Neurosurgery 2012;71(1): 86–92. PubMed PMID: 22407072.
17 Deer TR, Kapural L. New image-guided ultra-minimally invasive lumbar decompression method: The mild procedure. Pain Physician 2010;13(1):35–41. PubMed PMID: 20119461.
28 Wong WH. mild Interlaminar decompression for the treatment of lumbar spinal stenosis: Procedure description and case series with 1-year follow-up. Clin J Pain 2012;28(6):534–8. PubMed PMID: 22673487.
18 Mekhail N, Vallejo R, Coleman MH, Benyamin RM. Long-term results of percutaneous lumbar decompression mild® for spinal stenosis. Pain Pract 2012;12(3):184–93. PubMed PMID: 21676166.
29 Chopko BW. Long-term results of percutaneous lumbar decompression for LSS: Two-year outcomes. Clin J Pain 2013;29(11):939–43. PubMed PMID: 23446067.
19 Chopko BW. A novel method for treatment of lumbar spinal stenosis in high-risk surgical candidates: Pilot study experience with percutaneous remodeling of ligamentum flavum and lamina. J Neurosurg Spine 2011;14(1):46–50. PubMed PMID: 21142460.
30 Wang JJ, Bowden K, Pang G, Cipta A. Decrease in health care resource utilization with MILD. Pain Med 2013;14(5):657–61. PubMed PMID: 23578021.
20 Brown LL. A double-blind, randomized, prospective study of epidural steroid injection vs. the mild® procedure in patients with symptomatic lumbar spinal stenosis. Pain Pract 2012;12(5):333–41. PubMed PMID: 22272730. 21 Chopko B, Caraway DL. MiDAS I (mild Decompression Alternative to Open Surgery): A preliminary report of a prospective, multi-center clinical study. Pain Physician 2010;13(4):369–78. PubMed PMID: 20648206. 22 Lingreen R, Grider JS. Retrospective review of patient self-reported improvement and post-procedure findings for mild (minimally invasive lumbar decompression). Pain Physician 2010;13(6):555–60. PubMed PMID: 21102968. 23 Durkin B, Romeiser J, Shroyer AL, et al. Report from a quality assurance program on patients undergoing 204
31 Tumialan LM, Marciano FF, Theodore N. Regarding: Long-term results of percutaneous lumbar decompression mild for spinal stenosis. Pain Pract 2012;12(3): 252; author reply -3. PubMed PMID: 22390527. 32 Hagg O, Fritzell P, Nordwall A. The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur Spine J 2003;12(1):12–20. PubMed PMID: 12592542. 33 van der Roer N, Ostelo RW, Bekkering GE, van Tulder MW, de Vet HC. Minimal clinically important change for pain intensity, functional status, and genera l health status in patients with nonspecific low back pain. Spine 2006;31(5):578–82. PubMed PMID: 16508555. 34 Giraudeau B, Rozenberg S, Valat JP. Assessment of the clinically relevant change in pain for patients with sciatica. Ann Rheum Dis 2004;63(9):1180–1. PubMed PMID: 15308536. Pubmed Central PMCID: PMC1755131.
The mild® Procedure: A Systematic Review 35 Carragee EJ, Lincoln T, Parmar VS, Alamin T. A gold standard evaluation of the “discogenic pain” diagnosis as determined by provocative discography. Spine 2006;31(18):2115–23. PubMed PMID: 16915099.
36 Carragee EJ, Cheng I. Minimum acceptable outcomes after lumbar spinal fusion. Spine J 2010;10(4):313–20. PubMed PMID: 20362247.
205
