Treatment Options for Failed Back Surgery Syndrome: An Umbrella Systematic Review of Systematic Reviews on the Effectiveness of Therapeutic Interventions

Abstract

Background

Failed back surgery syndrome (FBSS) is a common and incapacitating condition affecting patients with previous spine surgery in whom treatment approach can be challenging. This study aimed to summarize existing secondary studies and up-to-date randomized clinical trials (RCTs) that assess the effectiveness of available treatment options for FBSS.

Methods

Systematic searches were carried out in five databases (PubMed, Cochrane, Scielo, Epistemonikos, and Google scholar) for all systematic reviews on the effectiveness of treatment options for FBSS published after 2012. Outcomes of interest were pain levels measured through visual analog scale or numeric rating scale, Oswestry Disability Index, and quality of life. Methodological and risk of bias assessments were performed with the AMSTAR-2 tool for systematic reviews and the Joanna Briggs Institute checklist for RCT. Prospective PROSPERO registration: CRD42022307609.

Results

Fifteen studies, seven systematic reviews, and eight RCTs met the inclusion criteria and fulfilled the methodological quality assessment. Of the 15 included studies, 8 were on neurostimulation, 4 on adhesiolysis, 4 on epidural or intrathecal injections, and 3 on other treatment modalities. The risk of bias was low in seven studies, moderate in five, and high in three.

Conclusions

Based on this systematic overview and the considerable heterogeneity among studies, the FBSS therapeutic approach must be individualized. FBSS treatment should start with conservative management, considering the implementation of neurostimulation, a technique with the most robust evidence of effective results, in cases of refractory axial or neuropathic pain. As the last resource, in light of the evidence found, more invasive procedures or new surgical interventions are indicated.

Introduction

The failed back surgery syndrome (FBSS), also known as persistent spinal pain syndrome or postoperative spine syndrome, is a condition afflicting a heterogeneous group of patients who suffer from persistent or recurring low back pain after spine surgeries with or without referred or radicular symptoms affecting their quality of life and mental health^1-3). Following primary spinal surgical interventions, the symptoms usually improve, but in about 10%-40% of the cases, the pain may continue or reappear quickly, even after successful and well-indicated procedures^2-4). The steady increment in the lifetime prevalence of chronic low back pain (LBP) (60%-85%) goes hand in hand with the yearly increasing rates of spine surgeries performed to treat the different associated conditions, contributing to the rise of FBSS frequency^2-4).

The multifactorial nature of this syndrome represents a challenge for the treating physicians, often conflicting with the therapeutic options given the paucity of high-quality evidence related to the effectiveness and safety of different treatment modalities that can be offered to patients suffering from FBSS^2,3). Existing systematic reviews often focus only on one treatment modality and frequently overlap due to the limited number of primary studies.

Umbrella systematic reviews of systematic reviews and meta-analyses attempt to synthesize existing studies into a practical and comprehensible document⁵⁾, merging more exhaustive knowledge than that contained in single primary or secondary studies⁶⁾, assessing the effectiveness and safety of several treatments for the same condition^7,8). An umbrella review centers on a condition and the often competing available interventions for its treatment, providing an accessible summary of evidence compiled in systematic reviews and considering their pros and cons^5,8).

This study was designed to be of clinical relevance and aims to investigate the following research questions: (1) In patients with failed back surgery syndrome, what are the treatment options and their clinical effectiveness? (2) What are the optimal treatment options for individualized patients?

Materials and Methods

The systematic review (SR) was conducted according to the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analysis) statement⁹⁾. The protocol for this review was registered a priori on the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD42022307609. The protocol experienced some adjustments during the development of the study.

Search strategy

Two reviewers conducted an independent literature search of existing systematic reviews and meta-analyses relating to treatment options for FBSS. The search was performed on February 20, 2022, and was restricted to 10 years (2012-2022). The search was conducted through PubMed (National Library of Medicine, NCBI), Cochrane Database of Systematic Reviews, Scielo (Scientific Electronic Library Online), Epistemonikos database, and Google scholar. To minimize the risk of missing relevant primary studies, on March 10, 2022, and January 24, 2023, the same reviewers performed a second and third complementary search for clinical trials published after 2019; in the updating searches, some new SRs were retrieved and included. No language restrictions were applied to any of the searches.

The electronic search strategy was designed following the Peer Review of Electronic Search Strategy (PRESS) guidelines¹⁰⁾ and combining medical subject headings and text words that were related to FBSS and treatment options: “failed back surgery syndrome,” “low back pain,” “chronic pain,” “disease management,” “therapeutics,” “surgical procedures,” “non-surgical interventions,” “pain management,” “complementary therapies,” “efficacy,” “effectiveness,” “complications,” “quality of life,” “systematic review,” “meta-analysis,” and the term “randomized controlled trial,” for the second and third search (supplemental material).

Eligibility criteria

All SRs and meta-analyses published before January 2023 assessing any treatment option for FBSS in adult patients were considered. The studies should have analyzed at least one of the outcomes, such as efficacy or effectiveness (pain level and functional assessment), complications or adverse outcomes, and quality of life. In the case of SR that included mixed populations (i.e., with FBSS and other diagnoses), only data from patients with FBSS were considered for this review. For the primary studies of the complementary searches, randomized controlled clinical trials evaluating or comparing the effectiveness and safety of treatment options for FBSS in adult patients were included.

Exclusion criteria

Narrative reviews, consensus without a well-documented SR, letters to editors, opinion articles, and animal, preclinical, and observational primary studies were excluded. Abstracts, thesis, protocols, and proceedings were also excluded. Moreover, studies published in languages different from English or Spanish were excluded.

Selection procedures

Two reviewers independently removed duplicates and screened the retrieved references for suitability based on their title and abstract; all preselected articles were read in full text, assessing their eligibility based on the inclusion and exclusion criteria. The references were screened and selected using Mendeley Desktop (Version 1.19.8, Ⓒ 2008-2020 Mendeley Ltd.). In the case of overlapping SR, only the most up-to-date and comprehensive one was included to avoid double-counting studies. Likewise, the most updated and complete outcomes of interest were selected for inclusion when the same population was contemplated in more than one RCT. Any disagreement between the two reviewers was first resolved by discussion. A third senior author was consulted if a consensus could not be reached.

Data extraction

Three authors independently extracted the data into two tables, one for SR and one for RCT. For the SR, the reviewers extracted the following data: first author, year, country of origin, aims, population and total sample, intervention(s), comparator(s), search strategy, observation window, number, type, and quality of primary studies included, heterogeneity, follow-up duration, outcomes assessed, and conclusions and recommendations. The pooled rates for all outcomes of interest were also included. For the RCTs, the data obtained from the studies were as follows: first author, year, aim, recruiting period, particularities of design, population (inclusion and exclusion criteria), intervention, comparator, sample, age and sex, follow-up duration, lost patients, outcomes of interest, and conclusions. Based on this data, a descriptive analysis of each manuscript was conducted to identify the effectiveness and safety of treatments and interventions to improve patient-reported outcomes. The authors of the included studies were not contacted for further or missing information.

Quality Assessment and primary overlap

The methodological quality of the included systematic reviews and meta-analyses was evaluated using the AMSTAR-2 (A Measurement Tool to Assess Systematic Reviews - 2)¹¹⁾, and the clinical trials were evaluated using the JBI critical appraisal checklist for randomized controlled trials (RCTs)¹²⁾ by two reviewers, independently. Disagreements were settled by consensus or solved by a third senior author. Based on the AMSTAR-2 critical domains, the confidence in the SR methodology (overall quality) was rated as high, moderate, low, or critically low. Reviews that were rated “critically low” were excluded. The total score of the JBI appraisal tool permitted the grading of the overall risk of bias (RoB) into high (up to 49%), moderate (50%-69%), and low (>70%). Clinical trials of low quality (high RoB) were excluded from the analysis. The primary overlap between studies was assessed by calculating the corrected covered area (CCA) according to the guideline by Hennessy and Johnson (2020)¹³⁾ for the overall overlap across included systematic reviews. CCA values of less than 5% indicated minimal overlap, while those greater than 15% indicated high overlap.

Results

Search results and study selection

Fig. 1 summarizes the study selection process presented in a flowchart (PRISMA). Seven SRs and eight RCTs met all inclusion criteria and were included in this study. Three different searches were performed: the first retrieved systematic reviews, the second retrieved RCTs not included in the SRs due to the observation window, and the third updated the publications since February 2022. The searches returned 486 records, and after excluding duplicates and screening by title and abstract, there were 66 papers read in full text, 44 of which were excluded. Ten SRs were evaluated and rated by AMSTAR-2; three were excluded for critically low quality. Twelve RCTs were evaluated with the JBI checklist; four were excluded due to a high RoB (Fig. 1).

Figure 1.

PRISMA flow diagram for all searches.

Of the seven SRs included, two were assessed as high quality (low RoB)^14,15), two as intermediate quality (moderate RoB)^16,17), and three as low quality (high RoB)^18-20). The most commonly noted omissions or unclear factors related to the clear statement of a protocol established before the development of the review, the list of excluded studies with justifications, the explanation of the selection of the primary studies' design, the detailed description of each included study, and the report of the sources of funding. Three SRs opted for a meta-analytical approach^15,16,18); two presented a high risk of publication bias (Fig. 2). Of the eight RCTs included, five had a low RoB (high quality)^21-25) and three had a moderate RoB (intermediate quality)^26-28). The main flaw observed in most of the studies was related to blinding; some RCTs failed to perform proper blinding, while in others, it could not be possible given the characteristics of the interventions (Fig. 3). There were no disagreements between the first and second reviewers. The CCA for the neurostimulation intervention (four reviews)^16,17,19,20) was 2% indicating minimal overlap across the studies. For adhesiolysis (four systematic reviews)^14,15,17,18), the CCA was 7.4% indicating intermediate overlapping of the primary studies of those included reviews. The CCA was 0% for epidural injections among the two included reviews^17,20). Of the 74 individual studies across the 7 systematic reviews, 66 were unique, 7 were repeated in 2 reviews, and 1 primary study was repeated in 3 of the 7 SRs.

Figure 2.

RoB of the systematic reviews and meta-analysis evaluated with AMSTAR-2.

Figure 3.

RoB of the randomized controlled trials evaluated with JBI critical appraisal checklist for RCT.

Characteristics of the included studies

Table 1, 2 show characteristics of the included SRs and RCTs. Of the 15 documents included, 4 SRs and 4 RCTs were on neurostimulation^{16,17,19,20,22,25-27)}, 4 SRs on adhesiolysis^14,15,17,18), 2 SRs and 2 RCTs on epidural/intrathecal injections^17,20,23,24), and 1 SR and 2 RCTs on other treatment modalities such as revision surgery, extended-release gabapentin, and balneotherapy^17,21,28). Most of the studies were published in the last 5 years, but two SRs, one on neurostimulation published in 2016¹⁹⁾ and one on adhesiolysis from 2012¹⁵⁾. The quality of the primary studies included in most SRs was mainly assessed as high or moderate, with few low-quality observational studies. Although evaluated as of high quality, one of the systematic reviews reported all the primary studies as of high RoB¹⁴⁾.

Table 1.

Characteristics of Each Systematic Review (n=7) Included in the Analysis.

Authors, year	Population	Included studies	Intervention	Comparator	Quality of primary studies
Brito-Garcia et al. 2019(14)	Chronic low back and/or radicular pain	9 studies: 2 RCTs 7 observational	Adhesiolysis	Epidural injections (EI) with steroids and/or anesthetics	High risk of bias
Cho et al. 2017(17)	Patients with FBS	22 studies: 6 RCTs 4 SRs 12 observational	•SCS •Epidural adhesiolysis •EI •Revision surgeries	Control or placebo/sham or other interventions	Low or moderate risk of bias
Geudeke et al. 2021(18)	Low back and/or leg pain after spinal surgery	9 studies: 2 RCTs 7 observational	Mechanical adhesiolysis with or without targeted drug	EI	Low, moderate, and high risk of bias
Grider et al. 2016(19)	Chronic spinal pain after surgery without CRPS	6 RCTs	SCS with implantable pulse generator	•HF10 •Reoperation •CMM •Sham •Manual vs. adaptive •Burst vs. Sham	Low and moderate risk of bias
Helm et al. 2012(15)	Post-lumbar surgery syndrome and spinal stenosis*	4 studies*: 3 RCTs 1 Observational	Adhesiolysis with normal or hypertonic saline	EI or physical therapy	Low and moderate risk of bias High risk of bias (observational study)
Kurt et al. 2022(16)	Patients with FBSS	11 studies: 4 RCTs 7 no randomized trials	SCS	•Oral medications •Nerve blocks •Epidural corticosteroids •Epidural adhesiolysis •Neurotomies Physiotherapy •Psychologic rehabilitative therapy •Chiropractic care •Acupuncture	Low, moderate, and high risk of bias
Papalia et al. 2022(20)	LBP after surgery	15 studies: 13 RCTs 2 observational	•SCS •Epidural steroid injections •Transcranial magnetic stimulation	Usual care/CMM	Low, moderate, and high risk of bias

FBSS: failed back surgery syndrome; RCT: randomized controlled trial; SCS: spinal cord stimulation; EI: epidural injections; NA: not available; CRPS: complex regional pain syndrome; HF10: 10 kHz high-frequency therapy; HFSCS: high-frequency spinal cord stimulation; CMM: conventional medical management; LBP: low back pain

*Only studies and data from FBSS patients were considered.

Table 2.

Characteristics of Each RCT (n=8) Included in the Analysis.

Authors, year	Population	Participants	Intervention	Comparator	Follow-up
Duse et al. 2019(26)	FBSS with chronic low back and/or leg pain	28	PB-SCS waveform programs: •Burst •1 kHz	Between them (crossover)	Daily for 7 days
Gewandter et al. 2019(21)	FBSS with NRS ≥4 and PainDETECT ≥12	32	Extended-release gabapentin (600-mg tablet)	Placebo	6, 14, and 16 weeks
Hara et al. 2022(22)	LBP after decompressive or fusion procedure for degenerative lumbar spine disease, refractory to OMM	50	SCS-burst stimulation	Placebo	12 months
Hashemi et al. 2019(23)	Low back pain due to FBSS	50	Epidural, caudal dexmedetomidine injection with triamcinolone and bupivacaine	Epidural, caudal injection of triamcinolone and bupivacaine	4 weeks
Hayek et al. 2023(24)	Chronic and persistent LBP after lumbar spine surgery or vertebral compression fracture	36	Continuous infusion of intrathecal bupivacaine and fentanyl	Continuous intrathecal infusion of normal saline	6 and 12 months
Rigoard et al. 2021(27)	FBSS refractory to OMM	100	SCS monocolumn programming (monogroup)	Multicolumn programming (multigroup)	1, 3, 6, and 12 months
Rigoard et al. 2021(25)	PSPS-T2 with failed OMM and previously implanted with SCS	14	SCS+PNfS	SCS	1, 3, 6, and 12 months
Yolgösteren/Külekçioˇglu. 2021(28)	PSPS after spine surgery	40	Aquatic exercises (balneotherapy).	Non-aquatic exercise	2 weeks and 1 to 6 months

FBSS: failed back surgery syndrome; SCS: spinal cord stimulation; LBP: low back pain; PB-SCS: paresthesia-based spinal cord stimulation; PSPS-T2: persistent spinal pain syndrome type 2; PNfS: peripheral nerve field stimulation; OMM: optimal medical management; NRS: numeric rating scale; NA: not available

The most frequent comparative outcomes reported in the SR and RCT were as follows: pain measurement using a visual analog scale or numeric pain rating scale, disability evaluation with the Oswestry Disability Index (ODI), quality of life assessment utilizing Short Form 36 (SF-36), and the European Quality of Life 5 Dimension Questionnaire (EQ-5D).

Neurostimulation

Four SRs and four RCTs evaluated the efficacy of spinal cord stimulation^{16,17,19,20,22,25-27)} (Table 3 and supplemental material). The SRs included RCTs that compared spinal cord stimulation (SCS) to other treatment modalities like reoperation and conventional medical management, showing better rates of pain relief ≥50% in the short and long term in patients treated with SCS^17,19). The SR by Cho et al. also explored observational studies concerning different modalities of SCS wherein the rates of pain relief ≥50% fluctuated between 36% for SCS with octopolar lead and 70% for high frequency (HF)-SCS¹⁷⁾. In the SRs by Grider et al.¹⁹⁾ and Papalia et al.²⁰⁾, HF-SCS showed significantly better rates of pain relief and functional improvement than low-frequency SCS. One SR evaluated the efficacy of combining SCS with optimized medical management (OMM) compared to OMM alone, showing better outcomes related to pain level with the SCS plus OMM²⁰⁾. The SR by Kurt et al.¹⁶⁾ showed a sustained and significant effect of SCS on pain level and ODI score at 3, 6, and 12 months compared to different treatment modalities. Duse et al.²⁶⁾ found that any SCS program (paresthesia-based, burst, or 1 kHz) significantly reduced the pain numeric rating scale. The preference among patients was still for the paresthesia-based SCS²⁶⁾. The SR by Papalia et al.²⁰⁾ and RCT by Hara et al.²²⁾ showed that burst stimulation, compared to placebo and other neurostimulation modalities, causes a nonsignificant improvement in pain intensity. In two RCTs by Rigoard et al.^25,27), monocolumn SCS was compared to multicolumn SCS and SCS plus peripheral nerve field stimulation (PNfS) was compared to SCS alone. At 6 months of follow-up, the authors found no significant difference in outcomes between the mono- and multicolumn SCSs²⁵⁾. However, adding PNfS to SCS showed a significantly better decrease in back pain than SCS alone at 3 months of follow-up²⁷⁾.

Table 3.

Outcomes Reported for Neurostimulation.

Authors	Outcome	SCS	Comparator
Cho et al. 2017(17)	Pain relief in legs ≥ 50%	48%	9% (medical treatment)
		66% (octopolar lead)	-
	Back pain relief ≥ 50%	47%	12% (reoperation)
		55% (unipolar)	-
		36% (octopolar lead)	-
		48% (tripolar electrode)	-
		42.1% (multicolumn)	-
		70% (HF-SCS)	-
	No abandonment of treatment/good pain relief	62% (radiofrequency)	-
	ODI decrease	From 54 to 39 (p<0.001) (HF-SCS)	-
Grider et al. 2016(19)	Outcome	SCS	Comparator
	Pain relief ≥ 50%	55%	80% (HF10)
		52%	10% (reoperation)
		48%	18% (CMM)
Kurt et al. 2022(16)	Outcome	SCS (6 months)	SCS (12 months)
	Low back pain (NRS)	2.25 (CI 95% 0.40-4.11)	3.16 (CI 95% 1.58-4.75)
	Leg pain (NRS) Overall pain (NRS)	3.25 (CI 95% 1.70-4.80) 2.81 (CI 95% 1.70-3.93)	4.01 (CI 95% 3.68-4.34) 2.68 (CI 95% 1.58-3.77)
	ODI	9.69 (CI 95% 4.41-14.98)	13.92 (CI 95% 2.94-24.91)
	SF-36 (physical component score)	5.08 (CI 95% 3.50-6.66)
	SF-36 (mental component score)	3.52 (CI 95% 0.04-7.01)	0.28 (CI 95% 0.09-0.46)
Papalia et al. 2022(20)	Outcome	SCS	Comparator
	Success	92% (HF-SCS)	84% (LF-SCS)
	VAS reduction	77% (HF-SCS)	64% (LF-SCS)
	Back pain relief ≥ 50%	33.9%	1.7% (CMM)
	Back pain relief	87.5% (Burst)	54.9% (HF-SCS)
Duse et al. 2019(26)	Outcome	SCS	Comparator
	NRS score	3.8 (PB-SCS)	9.0 (pre-SCS)
		4.3 (Burst)
		4.4 (1 kHz)
	Preference	50% (PB-SCS)	-
		21% (Burst)	-
		14% (1 kHz)	-
Hara et al. 2022(22)	Outcome	SCS (burst)	Comparator (placebo)
	ODI	34.0 (30.0-38.1)	35.4 (31.3-39.4)
	Leg pain (NRS)	5.9 (5.3-6.4)	6.1 (5.6-6.6)
	Back pain (NRS)	5.7 (5.2-6.2)	6.1 (5.6-6.6)
	EQ-5D-5L	0.48 (0.39-0.56)	0.44 (0.35-0.53)
Rigoard et al. 2021(25)	Outcome	SCS (Mono-group)	Comparator (SCS multi-group)
	Back VAS	46.3±28.7	44.0±30.0
	Back pain VAS decrease relative to baseline	35.5%	41.3%
	Leg pain VAS	31.7±27.2	27.7±26.1
	Leg pain VAS decrease relative to baseline	56.8%	64.2%
	Global pain VAS	40.3±26.3	39.3±27.4
	Health-related quality of life (EQ5D)	0.55±0.26	0.59±0.22
Rigoard et al. 2021(27)	Outcome	SCS+PNfS	Comparator (SCS-only)
	% decrease back pain surface	－80.2%±21.3%	13.2%±94.8%
	% decrease back pain paresthesia coverage	16.05%±16.16%	－0.94%±2.2%
	% decrease back pain VAS	－68.8%±19.9%	4.0%±15.0%
	% decrease ODI score	－31.5%±34.1%	5.0%±29.7%
	EQ-5D-3L index	0.23±0.33	0.02±0.17

SCS: spinal cord stimulation; HF-SCS: high-frequency spinal cord stimulation; LF: low frequency; CMM: conventional medical management; NRS: numeric rating scale; PB-SCS: paresthesia-based spinal cord stimulation; VAS: visual analog scale; ODI: Oswestry Disability Index; EQ-5D: European Quality of Life 5 Dimension questionnaire; SF-36: Short Form-36; PNfS: peripheral nerve field stimulation

Adhesiolysis

Four SRs evaluated the efficacy of adhesiolysis^14,15,17,18); one also reported evidence of safety¹⁴⁾ (Table 4 and supplemental material). Brito-Garcia et al.¹⁴⁾, Cho et al.¹⁷⁾, and Helm et al.¹⁵⁾ included data from RCTs and observational studies comparing percutaneous adhesiolysis to epidural injections (EI) that showed the superiority of the former in the improvement of pain measured with different scores and function up to 24 months of follow-up with a low rate of adverse events. Geudeke et al.¹⁸⁾ assessed the short- and long-term outcomes of mechanical endoscopic adhesiolysis (EA) with and without drugs finding that both options significantly improved pain and function at 6 months. However, the effect of the mechanical EA with drugs persisted for up to 12 months of follow-up.

Table 4.

Outcomes Reported for Adhesiolysis.

Authors	Outcome	PA	Comparator (EI)
Brito-Garcia et al. 2019(14)	Pain relief of >50% (3, 6, and 12 months)	87%	35%
		83%	23.6%
		72.6%	12.3%
	ODI score (3, 6, 12, 18, and 24 months)	15.2	20.2
		15.2	22.3
		15.8	23.3
		14.6	23.3
		13.9	23.2
	Back pain improvement 2 weeks and 6 months	65.4%	53.2%
		55.8%	33.9%
	Leg pain improvement (2 weeks and 6 months)	65.4%	64.5%
		53.8%	30.6%
	Disability improvement (2 weeks and 6 months)	63.5%	50%
		59.6%	29%
Cho et al. 2017(17)	Outcome	Epidural adhesiolysis	Comparator (EI)
	Pain relief of >50% and ODI reduction of >40% (12 and 24 months)	73%	12%
		82%	5%
	MacNab scale (patient satisfaction excellent or good)	50%	5.26%
	Outcome	EA (mechanical+drugs)	Comparator (EI)
Geudeke et al. 2021(18)	VAS (basal, 6 and 12 months)	7.5	6.5
		5.65	5.09
		5.73	6.45
	VAS reduction of >50% (6 and 12 months)	56%	-
		48%	-
	ODI (basal, 6 and 12 months)	54.2	59.4
		40.3	49.2
		45.1	53.7
	Outcome	EA (mechanical+drugs)	EA (mechanical alone)
	VAS reduction of >50% (6 and 12 months)	~40%	38.9%
		~22%-33%	62%
Helm et al. 2012(15)	Outcome	Adhesiolysis (mechanical plus drugs)	Comparator (Caudal EI)
	NRS reduction of >50% (3 and 12 months)	90%	35%
		73%	12%
	>40% improvement in ODI (12 months)	77%	13%
	Outcome	Adhesiolysis	Comparator (Physical therapy)
	VAS reduction (leg) (basal, 3 and 12 months)	7.2	-
		2.4	-
		2.8	-

PA: percutaneous adhesiolysis; EA: endoscopic adhesiolysis; EI: epidural injections; NRS: numeric rating scale; VAS: visual analog scale; ODI: Oswestry Disability Index 2.0

Epidural/intrathecal injections

Two SRs and two RCTs reported the efficacy of epidural and intrathecal injections^17,23) (Table 5 and supplemental material). The SRs considered data from four RCTs related to EI¹⁷⁾. One explored the analgesic effect of transforaminal injections (no mention of the specific solutions), showing positive but not significant results at 1 month of follow-up that decreased after the 3rd month of the intervention. In the second study, caudal injection with lidocaine and lidocaine plus betamethasone induced ≥50% sustained improvement in pain and function in about 50% of the patients for 2 years of follow-up but without significant difference among the injected solutions¹⁷⁾. One of the RTCs included in the SR by Papalia et al.²⁰⁾ explored the effect of caudal injection with local anesthetic alone than local anesthetic plus betamethasone with no significant difference in pain relief and disability reduction. The other RCT compared epidurogram plus a solution of local anesthetic and betamethasone to adhesiolysis plus the same solution finding significantly better results with the latter²⁰⁾. The RCT by Hashemi et al.²³⁾ compared two solutions of triamcinolone and bupivacaine with and without dexmedetomidine for EI; there was a significant difference in the pain level and quality of life among both solutions favoring the addition of dexmedetomidine. Hayek et al.²⁴⁾ compared the effect of intrathecal infusion of fentanyl and bupivacaine to saline solution, showing a significant improvement in pain level at rest and activity and ODI score with fentanyl plus bupivacaine.

Table 5.

Outcomes Reported for Epidural/intrathecal Injections.

Authors	Outcome	Caudal injection with lidocaine	Comparator (caudal injection with lidocaine+betamethasone)
Cho et al. 2017(17)	Pain relief and ODI reduction of ≥50% (24 months)	47%	58%
Papalia et al. 2022(20)	Outcome	Epidurogram + injection (lidocaine+betamethasone)	Comparator (adhesiolysis+injection: lidocaine+betamethasone)
	Pain relief and functional status improvement	12%	73%
	Outcome	Caudal EI (lidocaine)	Comparator (caudal EI: lidocaine+betamethasone)
	Pain relief and disability reduction	53%	59%
Helm et al. 2012(15)	Outcome	EI (dexmedetomidine, triamcinolone, and bupivacaine)	Comparator (EI: triamcinolone and bupivacaine)
	VAS (p=0.001)	2.56±0.1	5.21±0.1
	SF-36 (p=0.022)	74.43±0.72	72.08±0.34
Hayek et al. 2023(24)	Outcome	Intrathecal fentanyl/bupivacaine	Intrathecal saline solution
	NRS activity (p=0.016)	4.26±3.03	5.42±2.92
	NRS rest (p=0.006)	3.97±3.02	5.16±2.92
	ODI (<0.001)	20.38±8.42	27.07±9.76

NRS: numeric rating scale; VAS: visual analog scale; SF-36: Short Form-36; ODI: Oswestry Disability Index

Other treatment modalities

One SR and two RCTs explored other treatment modalities for FBSS^17,21,28) (Supplemental material). The SR, by Cho et al.¹⁷⁾, reported data from observational studies regarding different surgical modalities of reoperation. Spine stabilization in reintervention after lumbar discectomy was a surgical modality with a higher percentage of satisfactory outcomes, while instrumented spinal fusion and lumbosacral decompression reported lower rates of successful results. Total disc replacement did not show competitive results. The RCT from Gewandter et al.²¹⁾ evaluated extended-release gabapentin versus placebo without significant differences in the efficacy outcomes. The RCT by Yolgösteren and Külekçioğlu²⁸⁾ showed significant differences in favor of the practice of aquatic exercise compared to non-aquatic exercise regarding pain levels and functional status from the first 2 weeks up to 6 months into treatment.

Discussion

Umbrella reviews are studies typically assessing systematic reviews and seeking to inform on a specific clinical condition or topic. They aim at reviewing and synthesizing compelling evidence into tabular synthesis with a narrative commentary in an accessible and usable document^8,29). This study investigated the treatment options for FBSS, and their clinical effectiveness, enabling coverage of a wide range of therapeutic alternatives. Our literature search retrieved extensive evidence of the efficacy and effectiveness of treatment modalities such as neurostimulation, adhesiolysis, and epidural and intrathecal injections, which seemed to be the most commonly used interventions in clinical practice to manage this syndrome.

As the population ages, the prevalence of degenerative pathologies and the consequent surgical treatment procedures are expected to increase in the next few years. The management of FBSS remains challenging due to its complex clinical presentation and lack of gold-standard treatment and clinical guidelines. Among the included studies, the population of interest fulfilled different criteria regarding age, time suffering from low back and leg pain after spine surgery, pain level, and previous treatments. Considering the diversity of definitions for the condition, individualization of management then becomes more relevant.

Neurostimulation is one of the most effective semi-invasive measures in patients with a predominance of neuropathic pain and FBSS^{16,17,19,20,22,25-27)}. As far as our study identified, different neuromodulation techniques, such as conventional tonic stimulation, bursts, high frequency, and stimulation of the dorsal root ganglion and the peripheral nervous system, have shown favorable results, with reductions of up to 50% in back and leg pain and decreased need for analgesics, and improvement in the ODI score for periods of 3-24 months. In the included studies, SCS was more effective than OMM, reoperation, and nerve blocks; HF-SCS was more effective than other SCS modalities; and the addition of PNfS to SCS showed better results than SCS alone. However, most primary studies were controlled open trials comparing SCS to placebo or conventional medical management; caution in interpreting this data is necessary due to potential bias. Patient selection is also crucial to obtain optimal results; the candidate for this type of procedure should meet the following criteria: failed conservative management, no clear indication for reoperation, and absence of local infection.

The evidence on percutaneous or EA, alone or combined with medications, came from four reviews of primary studies showing favorable results and superiority regarding pain relief and disability score compared to EIs^14,15,17,18). The conclusions are contradictory among the reviews; while three of the reviews consider adhesiolysis as a promising therapy for FBSS patients, Brito et al.¹⁴⁾ suggest the evidence to be insufficient to recommend the intervention for the management of these patients. Mechanical adhesiolysis with and without drugs had similar effects; however, adhesiolysis with drugs showed a more lasting effect¹⁸⁾.

Regarding EI, two SRs and one RCT showed positive results related to pain relief and functional status improvement of EI with the combination of a local anesthetic with steroids^17,20,23). For intrathecal injections, the combination of anesthetics also provided positive results in pain level and disability than saline solution²⁴⁾. Transforaminal injections did not show a significant effect¹⁷⁾. Overall, the reported effect of these injections is short-termed relief that could benefit FBSS patients combined with other therapies or for symptom mitigation before any procedure.

The challenge when determining the best course of treatment arises in that the semi-invasive treatment options, like neurostimulation in all its modalities, have not been compared to the invasive approaches such as adhesiolysis and epidural or intrathecal injections. All those treatment options have been shown to be effective for pain relief; however, it remains unknown whether or not neurostimulation is more effective, and the choice of any of those treatments should be made based on previous management and pain and disability levels.

Other therapeutic approaches had scarce evidence; Cho et al.¹⁷⁾ presented some fair results for reoperation and recommended that particular indications, such as diagnosis of a clear mechanical cause for the pain susceptible to improvement through a new surgical intervention and failed conservative management or minimally invasive procedures after 6 months or more. Neuromodulatory drugs such as extended-release gabapentin showed positive but not significant results in relieving and preventing radicular pain in patients with FBSS than placebo²¹⁾. Alternative treatments such as balneotherapy and thermal aquatic exercise therapy were reported to improve pain severity, low back pain, and quality of life than exercise alone. These therapies can be complementary to invasive or minimally invasive procedures²⁸⁾.

Based on the evidence compiled in this study, we propose the following sequential strategy for FBSS management.

Initial approach using conservative and semi-invasive strategies should extend for 6 months as follows:

•Conservative options such as physiotherapy, hydrotherapy, and neuromodulatory drugs for a period of 6-12 weeks. Incorporate epidural injections/infiltrations (evidence of satisfactory responses and short-term improvements) in case of acute and severe pain.

•Move to neurostimulation techniques (best available evidence and satisfactory results in follow-up periods of around 3-6 months) in the presence of refractory axial or persistent neuropathic pain.

Adhesiolysis (treatment modality with contradictory results) could be performed in case of no adequate response and persistence of symptoms.

As a last resource, after treatment failure, a new surgical procedure (low success rates) could be considered.

This study was performed under a prespecified protocol to summarize data from secondary studies about the effectiveness of the different interventions for FBSS, a topic of interest whose treatment is still highly controversial and affected by potential biases. Umbrella reviews have disadvantages from combining reviews from primary studies with various treatments and outcomes that increase heterogeneity⁸⁾; this review was no stranger to that drawback. There was considerable heterogeneity among primary studies of the included reviews regarding the intervention, comparator, duration of treatments, and follow-ups. Generalizable conclusions and solid recommendations cannot be stated. Although heterogeneity of the data was a common denominator among all the included reviews, most of the SRs and their primary studies showed a low or moderate RoB (high or moderate methodological quality), providing some confidence in the validity of reported results. The current SR comprehensively analyzed available treatments for patients with FBSS; however, it might be subject to publication bias, given the possibility of excluding primary studies not contemplated by the included reviews. Performing different searches to capture as much updated evidence as was available was a strategy to mitigate this bias.

Conclusion

The heterogeneity of the studies included in this review limits the quality of new recommendations; however, with the results of our study, we hope to highlight the importance of an individualized and systematic approach considering multiple psychosocial and pathophysiological mechanisms that determine the rational use of the different treatment strategies. FBSS treatment should always start with conservative management, considering the implementation of neurostimulation, a technique with the most robust evidence of effective results, in cases of refractory axial or neuropathic pain. As the last resource, in light of the evidence found, more invasive procedures or new surgical interventions are indicated.

Conflicts of Interest: The authors declare that there are no relevant conflicts of interest.

Sources of Funding: None.

Author Contributions: HG and SA conceived the study; HG and AMH designed the study; HG, CJ, and AMH acquired and analyzed data; HG, SA, AC, and SF interpreted the data; HG, AC, and AMH drafted the manuscript; SA and CJ worked on tables (data extraction); and all authors revised the manuscript and approved the final version for publication.

Ethical Approval: Ethical approval was waived by the ethics committee due to the study design: a systematic review.

Informed Consent: Not required.