Efficacy and Safety of Pulsed Radiofrequency Treatment in Chronic Pain Syndromes
A systematic review
Level-3 evidence, is it good enough?
Barreto J, Carrington M, Correia-Pinto J, Monteiro P, Oliveira L, Paula Pinto M, Pinto I, Pinto Faria A, Rey-Formoso J, Silvério T, Teixeira P, Soares J, Valente M –
Adviser: Luís Azevedo; Class 14
Abstract
Background
Pulsed radiofrequency (PRF) is a growing method for treating chronic pain, defined as an unpleasant sensory experience associated with actual or potential tissue damage with duration of more than six months.
Aim
Our primary aim is to systematically review the evidence about efficacy and safety of pulsed radiofrequency in patients with chronic pain syndromes.
Methods
Four different databases were searched (Pubmed, ISI Web of Knowledge, SCOPUS and Central), applying no limits and in all languages. We included articles, reporting studies with two comparative groups, that mention the methods and results used to test the effectiveness and safety of PRFin several chronic pain syndromes. After extensive analysis of each included article, wesynthesized the available evidence.No meta-analysis was conducted because the available primary data were insufficient and very heterogeneous.
Results
Our literature search for PRF yielded 78 citations. After reviewing abstracts (during a first selection stage) and analyzing the full text articles (second selection stage), we encountered 5 articles that matched our criteria.The most methodologically adequate study, given the defined quality criteria, was published by Van Zundert et al(2007)and reported a significant difference between a group treated with PRF and the control group, with an improvement in the VAS pain score at 3 months in 25% of sham intervention grouppatients and in 82% of PRF group. The other four included studies described the efficacy of PRF with varying magnitude of effect and significance. None of the included studies mentioned any noticeable complications or adverse event of PRF.
Conclusion
PRF is likely to be a safety and efficient treatment for chronic pain syndromes,but the current available evidence is insufficient and of limited quality (level 3 evidence). It is advisable to evaluate further this technology by conducting new and methodologically sound randomized controlled trials.
Key-words:chronic, pain,pulsed and radiofrequency.
Introduction
Chronic pain is an unpleasant sensory experience associated with actual or potential tissue damage with duration of more than six months. It is one of the most underestimated healthcare problems impacting quality of life. [1]There are different treatments for chronic pain, such as, relaxation techniques, physical therapy and electrical stimulation therapies. [2]
Pulsed radiofrequency is a recently developed treatment modality for pain syndromes such as chronic pain. It uses brief pulses of high voltage to produce controlled tissue destruction, reducing pain by modulating its transmission.[3]
The importance of this treatment consists on its apparent lack of side effects which is appealing for those who suffer from chronic pain which has detrimental effects on physical and mental health. [1]
Although it presents a growing popularity, its efficacy and safety still remains a controversial issue. It is important to notice that it is not uncommon for therapies with uncertain efficacies to become established before reliable experimentsare conducted. [4]
Thus, our primary aim is to systematically review the evidence about efficacy and safety of pulsed radiofrequency in patients with chronic pain syndromes.
METHODS
Study design
A systematic review and meta-analysis of primary studies focusing on the efficacy and safety of pulsed radiofrequency in the treatment of chronic pain syndromes was undertaken. The methodological approach included the development of selection criteria, definition of search strategies, quality assessment of the studies, data abstraction and statistical data analysis. [5]
Search strategy
The MEDLINE database, ISI Web of Knowledge, SCOPUS and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched for identification of primary studies focusing on efficacy and safety of pulsed radiofrequency in the treatment of chronic pain syndromes.
The queries used were: “pulsed[All Fields] AND radiofrequency[All Fields] AND (("pain"[MeSH Terms] OR "pain"[All Fields]) AND chronic[All Fields])” for MEDLINE, “TS=(Radiofrequency SAME pulsed) AND TS=pain AND TS=chronic” for ISI Web of knowledge, “TITLE-ABS-KEY(radiofrequency) AND TITLE-ABS-KEY(pulsed) AND TITLE-ABS-KEY(pain) AND TITLE-ABS-KEY(chronic)” for SCOPUS and “ #1 MeSH descriptor Pain explode all trees, #2 (Chronic): ti, ab, kw, #3 (Radiofrequency Near Pulsed): ti, ab, kw” for CENTRAL.
Selection criteria
Articles identified through the database searches were included in this review if the purpose of it was to study efficacy and safety of pulsed radiofrequency as a treatment for chronic pain syndromes. Those which did not mention the methods and results used in the study were excluded. Articles that had at least, two comparative groups - a group submitted to pulsed radiofrequency and a control group (experimental and observational studies – clinical trials, cohort studies and case-control studies) were included in the present review. Meta-analyses, case series, reviews, letters and case reports of single patients were excluded. Studies that were not conducted in humans were excluded as well. Studies were not excluded on the basis of quality assessment or language of publication.
Selection process
The studies selection process included two phases. In the first phase of selection, the titles and abstracts of the retrieved studies were screened for relevance by at least two reviewers. In the second phase, two reviewers independently analyzed the full-papers of articles identified as potentially relevant in the screening phase. Selection criteria were applied; exclusions were decided and recorded and disagreements were settled by a third reviewer.
Data abstraction and study quality assessment
We used a standardized method for collecting relevant variables in each primary article. Data was fetched about general characteristics of the articles (references, authors, date of publication and setting), methodological aspects (study design and limitations), criteria for quality assessment and quantitative and qualitative results on the efficacy and safety of PRF.
The quality assessment of the included articles was performed by using criteria based on CONSORT Statement (for clinical trials) [6] or STROBE (for observational studies). [7]
Statistical analysis
Summary presentation of general characteristics of included studies was performed applying appropriate descriptive statistics. For this descriptive analysis the statistical analysis software PASW 18th version was used. [8]
Primary studies analysis was performed with tables that evaluate quantitative results in terms of the Visual Analogue Scale (VAS) classification for each comparative group in all the studies. It was not possible to calculate pooled efficacy and safety measures, neither present forest plots due to the high heterogeneity in the scales used to measure pain in the different articles. Furthermore, it was not possible to present subgroup analysis and test for publication bias because there was few data available for this propose in each article.
RESULTS
Search for primary studies
In our search we identified a total of seventy-eight articles. The MEDLINE search identified forty-eight articles. Of those forty-eight articles, four were exclusive of the MEDLINE data base. Sixty-five articles were found through SCOPUS, including forty-four that were initially identified through the MEDLINE search. Of the sixty-five articles identified in SCOPUS data base, twenty were exclusive. The ISI Web of Knowledge search yielded thirty-two articles, of which twenty-three were previously identified. The ISI Web of Knowledge had eight exclusive articles. Three articles were found through CENTRAL and they had already been identified, so there were no exclusive articles in CENTRAL database.
Studies selection
In the selection process seventy-three studies were excluded(Figure 1).
There were two selection stages. In the first one, by reading all titles and abstracts (which were analyzedby two reviewers and, when there were a disagreement, it was used a third one), we excluded sixty-five articles. In the second stage, by reading full text articles (using two reviewers, as well), we excluded eight more articles, because studies did not meet the predefined selection criteria.
In the first selection stage, concerning the inclusion/exclusion argument there were no disagreements (0%) and 78 agreements (100%). In 74 cases (94,8%) there were agreements related to the reason of exclusion and, in 4 cases (5,2%), there were disagreements and was needed a third reviewer.
In the second selection stage, there were not disagreements (0%).
At the end of the process, we encountered 5 articles: 3 prospective and 2 retrospective studies.
Prospective controlled trials
The prospective controlled trials found in the databases are described and compared in general parameters on Table 1.
Van Zundert et al, in 2007, published an article about a controlled, randomized trial with twenty-three patients with cervical radicular syndrome. Eleven of those patients received PRF as treatment while twelve had sham treatment of the DRG (dorsal root ganglion). During three months, patients treated with PRF had good results: more than 50% improvement in GPE (global perceived effect scores) scores and 20 point reduction in pain measurement VAS (visual analogue scale) scores (in this case, the scale used was 0-100) (Table 3). There were only 23 patients included in this study, which is a limitation. Besides that, the two groups were different at baseline – sham treated group had a higher pre-treatment VAS score (76.2) than PRF group (55.7). There were a short-term follow up (3 months), which is also a handicap.[9]
The goal of Tekin et al (2007) study was to compare the effects of conventional radiofrequency and PRF denervation to medial branches of dorsal rami in the treatment of facet joint pain. In this prospective, randomized, double blinded controlled trial of 60 patients, all suffering from continuous low back pain, the patients were divided into three groups of twenty each. One control group, where local anesthetic was applied, a CRF (Continuous Radiofrequency) group where a 80ºC was applied and a PRF group, where 2 Hz PRF was applied. Pain relief was evaluated by visual analog scale (VAS) and Oswestry Disability Index (ODI) was used to evaluate disease related disabilityat pre-procedure, procedure, at 6 months and 1 year after the procedure. Reduction in analgesic usage, patients' satisfaction, and complications were assessed. Mean pre-procedural VAS and ODI scores were higher than post-procedural scores in all groups. Both VAS and ODI scores of PRF and CRF groups were lower than the score of the control group at the post-procedural evaluation. In spite of the decrease of the pain, score was maintained in the CRF group at 6 months and 1-year period, this decrease discontinued in the PRF group at the follow-up periods. The number of patients not using analgesics and patient satisfaction were highest in CRF group. The authors concluded that PRF and CRF are effective and safe alternatives in the treatment of facet joint pain but PRF is not as long lasting as CRF.[10]
Simopoulos et al, in 2008, published an article about a prospective, randomized, sham-controlled and double blinded trial, that included 76 patients with chronic lombo-sacral pain. The patients received two different types of treatment: 37 were treated with PRF and 39 with PRF followed by continuous radiofrequency (CRF). There was a 2 month follow up. After that time, there was success in both groups, but there wasn’t a statistically significant difference between them (Table 3). [11]
Study and methods / Patients and treatments / Results / Studylimitations / Number of patients in each group / Setting / Adverse reactions or eventsVan Zundert et al [9],
Prospective, randomised, sham-controlled, double blind trial / 23 patients with cervical radicular syndrome,
PRF of DRG compared with sham intervention / 3 months follow-up: PRF group with better outcome – GPE and VAS (20%) / Only 23 patients, short-term follow up (3 months), high pre-treatment VAS of sham treated group / 11 patients – PRF group; 12 patients – sham treated group / Netherland / -
Tekin I et al [10],
Prospective, randomized, double-blind, sham lesion controlled trial / 60 patients with continuous low back pain with no radicular syndrome. 80ºC CRF compared to 2Hz PRF and to local anesthetic. / VAS and ODI scores of PRF and CRF groups were lower than the control group. In the follow up period the scores were lower in the CRF group than those of the PRF. / Risk of ending the blinding process after 3 month follow-up. Few support studies.
Risk of false positive blocks / 20 patients- Control Group
20 patients- PRF Group
20 patients- CRF Group / Turkey / -
Simopoulos et al [11],
Prospective, randomised, sham-controlled, double blind trial / 76 patients with chronic lombo-sacral pain randomly assigned to one of two types of treatment: PRF of the DRG or PRF followed by continued radiofrequency. / 2 months follow up: success (2 groups); no statistically difference between the two treatments groups. / The efficacy of PRF is not tested on the comparative group / 37 - PRF
39 - PRF followed by continues radiofrequency / USA
Israel / -
Table 1: Prospective studies on pulsed radiofrequency
Retrospective trials
The retrospective studies identified are summarized and compared in general parameters on Table 2.
Cohen et al, in 2006, published a report on PRF, which included a retrospective (data analysis) study concerning 49 patients with post-surgical thoracic pain: 13 patients received PRF of DRG (Dorsal Root Ganglion), 15 patients received PRF of ICN (intercostals nerves) while 21 received conventional pharmacotherapy. At six weeks, 61.5% of the pulsed RF DRG group reported ≥ 50% pain relief vs. 27.3% in the Pharmacotherapy group and 21.4% in the ICN group. At 3-month follow-up, 53.8% in the DRG group continued to report ≥ 50% pain relief vs. 19.9% in the pharmacotherapy and 6.7% in the ICN groups (Table 3). The authors concluded that PRF of the DRG was a superior treatment to pharmacotherapy and PRF of the ICN in patients with CPTP (Chronic postsurgical thoracic pain). Prospective studies are needed to confirm these results and identify the best candidates for this treatment.[12]
Lindner et al published an article about a retrospective study, including 48 patients with low back pain and more than 50% relief following a diagnostic of medial branch block. 19 patients had been operated and 29 were in the non operated group. As results, a successful outcome was verified in 21/29 non operated patients and in 5/19 operated patients (Table 3).[13]
Study and methods / Patients and treatments / Results / Studylimitations / Number of patients in each group / Setting / Adverse reactions or eventsCohen et al [12], Retrospective study / 49 patients with post surgical thoracic pain,PRF of DRG and PRF of ICNcompared with pharmacotherapy / Both 6 weeks and 3 monthsfollow up: PRF of DRG group with better outcome / Retrospective, Only 49 patients, short-termmaximum follow up (3 months) / PRF of DRG (n=13)
PRF of ICN (n=15) Pharmacotherapy (n=21) / - / Only in the pharmacotherapy group, 7 of the 21 patients revealed adverse side effects
Lindner et al [13], Retrospective study / 48 patients with low back pain and >50% pain relief following a diagnostic medial branch block:
PRF treatment of the medial branch / 21/29 nonoperated patients and 5/19 operated patients, successful outcome. / The setup of our study does not permit a comparison with the results of continuous radiofrequency (CRF) for the same procedure / 19 patients operated
29 patients non-operated / Switzerland / -
Table 2: Retrospective studies on pulsed radiofrequency
Quantitative results
The efficiency of PRF was also measured through quantitative results, the visual analogue scale (VAS) for pain, for instance. A summary of these quantitative results for each included study is displayed on Table 3.
VASGroup 1 / VAS
Group 2 (PRF)
Article / Initially / Primary outcome / Secondary outcome / Initially / Primary outcome / Secondary outcome
Cohen et al, 2006 / Medical management group:
- / 6 weeks:
27.3% reported ≥
50% pain relief / 3 months: 19.9% reported ≥
50% pain relief / PRF group:
- / 6 weeks: 61.5% reported ≥
50% pain relief / 3 months:
53.8%
reported ≥
50% pain relief
Lindner et al, 2006 / Operated patients / 1 month:
- / 4 months: 5/19 reported ≥
60% pain relief / Nonoperated patients / 1 months:
- / 4 months: 21/29 reported ≥
60% pain relief
Van Zundert et al, 2007 / Sham group: 76,2 / 3 months: 3/12 (25%) improvement / 6 months: improvement / PRF group: 55,7 / 3 months: 9/11 (82%) improvement / 6 months: improvement
Tekin et al, 2007 / Control group:
Similar / 6 months: Higher / 1 year:
Similar / PRF group:
Similar / 6 months: Lower / 1 year:
Similar
Simopoulos et al, 2008 / PRF + CRF group:
7.1 ± 1.9 / 8 weeks:
2.3 ± 2.2 / - / PRF group:
7.8 ± 1.6 / 8 weeks:
3.5 ± 3.4 / -
Table 3: Quantitative results, using VAS (Visual Analogue Scale) for pain assessment
Quality assessmentoftheincludedarticles
For the quality evaluation of the included randomized controlled trials, we used the CONSORT Statement checklist.
In order to evaluate the included observational studies, we used the STROBE checklist for observational studies, as referred in the methods section.
In table 4, we present the results of that evaluation as the percentage of adequately accomplished items in each checklist for each included study.
Article / ChecklistevaluationVanZundert et al, 2007 [9] / 100%
Lindner et al, 2006 [13] / 93,75%
Cohen et al, 2006 [12] / 75%
Tekin et al, 2007 [10] / 56,25%
Simopoulos et al, 2008 [11] / 43,75%
Table 4: Evaluation of the included articles
The articles with the references [9], [10] and [11] were evaluated using CONSORT Statement checklist
The articles with the references [12] and [13] were evaluated using STROBE checklist
Side-effects and complications of PRF
None of the studies reviewed reported any noticeable complications of PRF application.
DISCUSSION
Discussion of results
Chronic pain is a common healthproblem that presents a major challenge to healthcare providers because of its complex natural history, unclear etiology, and poor response to therapy.PRF has been used to the treatment of various chronic painsyndromes almost over a decade. However, it continues to be a controversial treatment and its efficacy and safety are not adequately demonstrated. [4]