INTRODUCTION
Herniated lumbar discs (HLDs) could be a cause of chronic low back pain and affects 1%–3% of the population each year [1]. HLD has a higher incidence among people ages 30 to 50, with a higher prevalence in males than females [2].
Disc herniation is broadly defined as a “localized or focal displacement of disc material beyond the limits of the intervertebral disc space [3].” It is most frequently observed in the lumbar region of the spine, with around 95% of disc herniations occurring at the L4–5 or L5–S1 level [4,5]. The symptoms associated with HLD can vary. There are asymptomatic disc herniations, which can be present in 20%–40% of the adult population [6-9]. Some of the asymptomatic HLD can progress and become symptomatic [10]. If the extruded disc material irritates or compresses the lumbar nerves, the patient may endorse severe neurologic symptoms such as weakness, pain, and sensory changes [11-13]. In extreme cases, HLD can put pressure on the cauda equina, causing low back pain and bladder and bowel dysfunctions [5]. While HLD is a common occurrence in the population, there is a paucity of standard treatment paradigms for managing HLD.
Commonly, the first-line treatments for HLD with radiculopathy are nonoperative measures unless there is a neurological deficit, including weakness in a muscle group or cauda equina syndrome [14,15]. Studies have recommended conservative treatments for 6 weeks to up to 2 months after the onset of symptoms before considering surgery [4,16]. More than 85% of patients with acute HLD with radiculopathy will have resolution of their symptoms over time [17]. Spontaneous reabsorption of disc herniation has also been found to occur in more than half of symptomatic HLD cases managed by conservative treatments [18,19].
However, for certain pathological conditions related to HLD where symptoms persist and fail to improve with nonoperative measures, surgery may be considered. Many studies have advocated for surgery for HLD with radiculopathy within 8 weeks of symptom onset, suggesting that this may achieve optimal cost-effectiveness and patient satisfaction [20,21] Surgery frequently results in the immediate resolution of radicular pain [22-24]. For patients who met the indications for surgery, those who underwent surgery had better improvement postoperatively than patients managed conservatively [22,23]. Nonetheless, the literature has not reached an agreement on whether operative or nonoperative treatments have superior long-term outcomes [23,25].
To help pinpoint the most appropriate treatments for HLD, many professional associations across the globe have dedicated efforts to compiling guidelines that include surgical and nonsurgical treatments [11,25-35]. In this study, we aimed to investigate the differences in the recommendations given by these guidelines and provide insight into the current views of surgical and nonsurgical HLD treatments.
METHODS
A systematic literature search was conducted using PubMed, Web of Science, Scopus, Guidelines International Network, and Turning Research into Practice databases, as well as the websites of medical specialty societies. Search terms and filters are listed in Supplementary Table 1. PubMed generated 4 initial results, Web of Science generated 71 results, Scopus generated 143 results, Guidelines International Network generated 2 results, and Turning Research into Practice database generated 56 results, totaling 276 articles. Additionally, 2 medical society websites contained HLD guidelines. The selection process followed the PRISMA (Preferred Reporting Items for Systematic Review Meta-analyses) guideline. Inclusion criteria included the most recent version of guidelines published by a professional association relating to HLD in English with full-text availability. Non-HLD-specific guidelines and all other article types were excluded. Guidelines were defined as sets of recommendations on HLD treatment and were either titled or labeled/identified as guidelines by their authors. Twelve items satisfied the inclusion criteria and were analyzed as illustrated in Fig. 1. Four were issued by non-United States (US) associations: Japanese Orthopaedic Association (JOA), Ontario Protocol for Traffic Injury Management Collaboration (OPTIMa), Polish Society of Spinal Surgery (PSSS), and International Society for the Advancement of Spine Surgery (ISASS). The remaining 8 were from US-based associa-tions. The guidelines by medical societies included 3 pain societies: American Society of Pain and Neuroscience (ASPN), American College of Physicians/American Pain Society (ACP/APS), and American Society of Interventional Pain Physicians (ASIPP), 2 surgical societies: American Association of Neurological Surgeons (AANS) and American Academy of Orthopaedic Surgeons (AAOS), and 1 multidisciplinary medical society: North American Spine Society (NASS). Two government-issued guidelines came from Washington State Department of Labor & Industries (L&I) and Department of Veterans Affairs/Department of Defense (VA/DoD). Each guideline was independently reviewed by 3 authors (HJ, AML, FGR), and consensus was reached on recommendation classifications.
1. Primary Audience and Recommendation Classification
Treatment recommendation criteria employed by the guidelines are shown in Table 1 and Supplementary Table 2. The guidelines could be divided into 2 groups based on the target audience: clinicians or patients.
An evidence-based approach guided the treatment evaluation process for clinician-focused guidelines: ASPN, ACP/APS, ASIPP, L&I, VA/DoD, NASS, JOA, OPTIMa, PSSS, and ISASS guidelines [11,25-31,34,35]. ACP/APS, VA/DoD, JOA, and PSSS used a form of the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system, despite deriving their criteria from different sources [11,25,27,35]. The GRADE system was developed to provide a structured method for assessing evidence quality and strength that supported the confidence of recommendation [36,37].
Besides the GRADE approach, ACP/APS also adopted the United States Preventive Services Task Force criteria, which was referred to by ASPN when classifying evidence quality [11,30]. ASIPP used methodology from Part 1 of the Update of Comprehensive Evidence-Based Guidelines for Interventional Techniques of Chronic Spinal Pain [31], NASS formulated a standard grading system assessing the supporting literature followed by member voting if evidence was inadequate [29], L&I conducted a literature review [34], OPTIMa applied the Appraisal of Guidelines for Research and Evaluation II instrument to assess the methodological quality of guidelines [26], ISASS compiled randomized controlled trials, retrospective cohort studies, and meta-analyses to inform treatments for HLD [28]. Similar to the guidelines that rated treatments using the GRADE system, ASPN, ASIPP, and NASS stratified their treatment recommendations based on levels of evidence [29-31].
The guidelines’ determination of evidence quality could be classified into 3 categories: “good,” “fair,” and “limited.” “Good” required well-designed, well-conducted studies in representative populations that directly assessed effects on health outcomes. “Fair” required sufficient evidence, but the strength of evidence was limited by the study design (e.g., number, quality, size, or consistency of study results), clinical relevance, or indirect assessment of health outcome. “Limited” indicated evidence that had indirect clinical relevance and non-generalizable findings (e.g., lack of randomized controlled trials) or inconsistent findings although evidence towards one recommendation direction outweighed the other.
Treatment recommendation strength was determined based on evidence quality and direction. The categories were presented using the grades “A,” “B,” “C,” “D,” and “I.” “A” indicated “good” quality of evidence recommending the treatment, implying a high certainty that the net benefit of the treatment was substan-tial, and the treatment recommendation was, therefore, unlikely to be changed by future studies. “B” indicated a “fair” quality of evidence recommending the treatment, therefore there was a high certainty that the net benefit was moderate or moderate certainty that the net benefit was substantial, and there existed the possibility that future studies might trigger reevaluation of the treatment recommendation. “C” indicated “limited” quality of evidence recommending the treatment, thus there was limited confidence in treatment efficacy or some certainty that the net benefit was small and only applicable to selected patients depending on individual circumstances. “C” recommendations could be altered as future evidence arises. When recommendation strength could not be differentiated from the treatment recommendation description, an “R” was assigned. “D” indicated evidence recommending against the treatment, and there was some certainty that the treatment had no net benefit or that harms outweighed benefits. “I” was assigned if guidelines stated that evidence was insufficient to assess benefits and harms, therefore no recommendation for or against the treatment could be made. This could be due to a lack of evidence, poor evidence quality (e.g., flawed study design), or conflicting evidence.
On the other hand, L&I, OPTIMa, and ISASS did not specify the quality of evidence or treatment recommendation strengths despite using evidence-based approaches to develop their guidelines [26,28,34]. The treatments assessed in these guidelines were determined to be “R” (recommended), “D” (not recommended), or “I” (of insufficient evidence) based on the guideline’s wording/descriptions of the treatments.
The AANS and AAOS guidelines were not designed for clinical use but for patient education [32,33]. The 2 surgical society guidelines were web pages that could be accessed from their websites. AANS did not disclose the method of decision-making used when drafting their recommendations [32], while a single boardcertified orthopedic surgeon wrote the AAOS guideline, which was later peer-reviewed [33]. Though lacking a formal evidence-based assessment in their treatment recommendation process, patient-focused guidelines seemed to contain less medical jargon and facilitated patient comprehension despite compromised scientific scrutiny [38]. The AANS and AAOS guidelines do not describe their recommendations by evidence strength, but only on the direction of recommendation [32,33]. Thus, the treatments were considered “R” (recommended) or “D” (not recommended).
2. Categorization of Treatments
Treatments were categorized as noninvasive, invasive nonsurgical, and surgical treatments in Tables 2–4 and Figs. 2–4. Noninvasive treatments involve no insertion through the skin or into a body opening. Invasive, nonsurgical treatments involve the insertion of instruments through the skin, such as needles or electrodes, which could inflict patient discomfort or require local anesthesia. Invasive, nonsurgical treatments are typically performed in an office or outpatient setting without general anesthesia. Surgical treatments involve incisions and local or general anesthesia, are performed in operating rooms with high sterilization requirements, and are done by specialized physicians or surgeons such as pain physicians, neurosurgeons, and orthopedic surgeons.
3. Recommendation Direction Analysis
Recommendation direction was consistent if all guidelines that assessed the treatment provided the same general opinion in describing it as recommended, not recommended, or of insufficient evidential support (i.e., no recommendations made), corresponding to “R,” “D,” and “I” in Tables 2–4. Recommendation grades of “A,” “B,” “C,” and “R” all indicated recommendations for treatments. Provisional consensus in recommendation direction described less than 5 guidelines recommending the treatment in the same direction, while definite consensus in recommendation direction described 5 or more guidelines recommending the treatment in the same direction.
4. Treatment Recommendation Analysis
Treatment recommendation strength and direction were summarized in Figs. 2–4. Treatments assessed by only one guideline were excluded from this analysis. Recommendation grades were quantified with “A” corresponding to “+3,” “B” corresponding to “+2,” “C” corresponding to “+1,” “I” corresponding to “0,” and “D” corresponding to “-1.” Due to how “R” could encompass “A,” “B,” and “C,” the average of “+2” was taken as a numerical representation of “R.” The individual recommendation grade assigned by each guideline for the same treatment was summed and then divided by the number of guidelines that assessed the treatment. The weighted recommendation scores were graphed, with positive versus negative representing overall recommendation direction and magnitude of treatment recommendation capturing recommendation strength.
RESULTS
The recommended noninvasive, invasive nonsurgical, and surgical treatments for HLD are shown in Tables 2–4. The guidelines by ACP/APS, VA/DoD, and OPTIMa did not provide any surgical recommendations [11,26,35]. Guidelines by the PSSS and ISASS did not offer any nonsurgical recommendations [27,28].
In the evaluation of treatment efficacy, pain management was a common theme that ran across the guidelines. The pain societies, ASPN, ACP/APS, and ASIPP, assessed treatments by their net benefit in relieving pain associated with HLD [11,30,31]. Other than pain societies, L&I and PSSS also estimated treatment efficacy by the magnitude of pain reduction [27,34]. Besides pain improvement, ACP/APS included back-specific functional status and other outcomes in their grading criteria [11]. VA/DoD took a holistic approach by looking at anticipated benefits and harms, values and preferences, resource use, equity, acceptability, and feasibility [35]. Additionally, ISASS’s grading criteria encompassed incision size, soft tissue damage, hospital stay for surgical interventions, symptom relief, and functional improvements [28].
1. Grading Variations Between Guidelines due to Different Pathologies
The PSSS guideline examined differences in treatment efficacy for primary HLD versus recurrent HLD and found fusion to be a grade “I” treatment for primary HLD in contrast to a grade “C” treatment for recurrent HLD [27]. A grade “I” was consistent with the ratings of NASS and JOA [25,29]. While a grade “C” was consistent with the recommendation made by L&I for recurrent HLD [34].
2. Consensus in Recommendation Direction
Of the nonsurgical treatments examined in the guidelines, regenerative therapies, including prolotherapy and platelet-rich plasma, allogeneic mesenchymal stem cells, bone marrow-de-rived mesenchymal stem cells, and adipose tissue-derived mesenchymal stem cells, were only mentioned in the ASPN guideline [30]. Similarly, intradiscal high-pressure saline injection was only assessed by NASS [29], while basivertebral nerve ablation was only assessed by ASPN [30]. Therefore, no consistency analysis could be performed on these treatments.
Noninvasive interventions have largely been preferred as first-line treatments [15]. Provisional consensus was seen in the recommendation direction for back schools, which was recommended by 3 guidelines. Back schools were rated “B” by VA/DoD [35], “C” by ACP/APS [11], and recommended by OPTIMa [26]. On the other hand, cognitive behavioral therapy and nonsteroidal antiinflammatory drugs (NSAIDs) achieved more definite consensus in the recommendation direction, which were recommended by 5 and 9 guidelines, respectively. Cognitive behavioral therapy was rated “A” by VA/DoD [35], “B” by ACP/APS [11], and recommended by ASIPP [31], L&I, and OPTIMa [26,34]. For oral medications, NSAIDs were rated “A” by VA/DoD [35], “B” by ASPN and ACP/APS [11,30], “C” by JOA [25], and recommended by ASIPP, AANS, AAOS, L&I, and OPTIMa [26,31-34].
Among the invasive, nonsurgical strategies for pain management in patients with HLD, definite consensus in recommendation directions was seen for short-term epidural steroid injections. Three types of epidural steroid injections were examined in the guidelines: transforaminal, interlaminar, and caudal epidural steroid injections. Short-term transforaminal epidural steroid injection was recommended by 8 guidelines, rated “A” by ASIPP and NASS [29,31], “B” by VA/DoD and JOA [25,35], and recommended by ASPN, AANS, AAOS, and L&I [30,32-34]. Short-term interlaminar epidural steroid injection was recommended by 8 guidelines, rated “A” by ASIPP [31], “B” by VA/DoD and JOA [25,35], “C” by NASS [29], and recommended by ASPN, AANS, AAOS, and L&I [30,32-34]. Lastly, short-term caudal epidural steroid injection was recommended by 7 guidelines, rated “A” by ASIPP [31], “B” by VA/DoD and JOA [25,35], and recommended by ASPN, AANS, AAOS, and L&I [30-34].
Of the surgical treatments explicitly mentioned in the guidelines, aggressive discectomy and medial facetectomy were only assessed by a single guideline, thus no conclusion on consistency could be drawn. Multiple discectomy procedures earned definite consensus in recommendation directions. A total of 8 guidelines recommended open discectomy and microdiscectomy, and 5 guidelines recommended microendoscopic discectomy. Open discectomy and microdiscectomy were recommended by AANS, AAOS, L&I, and ISASS [28,32-34]. When rating open discectomy and microdiscectomy, NASS and JOA gave a grade “B” recommendation [25,29], while PSSS assigned a grade “A [27]”. In addition, ASPN gave a grade “B” rating to microdiscectomy and recommended open discectomy [30]. Endoscopic discectomy was divided into percutaneous endoscopic discectomy and microendoscopic discectomy when determining recommendations. Percutaneous endoscopic discectomy and microendoscopic discectomy differ in the instruments used, level of invasion, and local versus general anesthesia [39]. Microendoscopic discectomy was recommended by AANS, AAOS, and ISASS [28,32,33], and was rated “B” by ASPN and JOA [25,30].
3. Inconsistency in Recommendation Direction
Recommendations for most noninvasive measures lacked consistency in recommendation direction among guidelines. Physical therapy was recommended by 8 out of 10 guidelines but was considered lacking sufficient evidence by 2 guidelines. Physical therapy was recommended by ASPN, ASIPP, AANS, AAOS, L&I, and OPTIMa [26,30-34], rated “B” by ACP/APS and VA/DoD [11,35], and “I” by NASS and JOA [25,29]. Similarly, clinician-guided exercise was positively recommended by 6 out of 9 guidelines, recommended against HLD treatment by 1 guideline, and deemed insufficient evidence by 2 guidelines. Exercised was recommended by ASIPP, AANS, AAOS, L&I, and OPTIMa [26,31-34], rated “B” by ACP/APS [11], “D” by VA/DoD [35], and “I” by NASS and JOA [25,29]. Bedrest was recommended by 1 guideline, AAOS [33], but recommend against by 2 guidelines, AANS and VA/DoD [32,35]. Spinal manipulation was recommended by 4 out of 6 guidelines, recommended against HLD treatment by 1 guideline, and received a weak recommendation due to almost insufficient evidence by 1 guideline. Spinal manipulation was rated “B” by VA/DoD [35], “C/I” by NASS [29]. recommended by ASPN, L&I, and OPTIMa [26,30,34], but not recommended by ACP/APS [11]. Traction was recommended by 3 out of 8 guidelines, ASIPP, AANS, and L&I [31,32,34], rated “D” by one guideline, ACP/APS [11], and “I” by 4 out of 8 guidelines, VA/DoD [35], NASS, JOA, and OPTIMa [25,26,29,35], Bracing was recommended by ASIPP [31], but rated “I” by 3 out of 4 guidelines, VA/DoD, NASS, and JOA [25,29,35]. Transcutaneous electrical nerve stimulation is a noninvasive method that delivers electrical stimulation to the skin for pain relief [40]. It was rated “I” by 1 guideline, NASS [29], but not recommended by 3 out of 4 guidelines, ACP/APS, VA/DoD, and OPTIMa [11,26,35].
Though consistency was observed in the recommendation direction of short-term epidural steroid injections, long-term epidural steroid injections were rated “D” by 1 out of 3 guidelines, VA/DoD [35], and “I” by the other 2 guidelines, NASS and JOA [25,29]. Intra-articular lumbar facet joint injections have also been used for pain alleviation and were rated “C” by 2 out of 3 guidelines, ASPN and ASIPP [30,31], and “D” by 1 guideline, VA/DoD [35]. Additionally, various minimally invasive treatments have been developed for chronic pain reduction. Percutaneous peripheral nerve stimulation is another minimally invasive therapy with an analgesic effect achieved by delivering electrical currents to the affected nerve [41]. It was rated “C” by 2 out of 3 guidelines, ASPN and VA/DoD [30,35], and “I” by 1 guideline, ACP/APS [11]. Radiofrequency ablation was rated “A” by 2 out of 3 guidelines, ASPN and ASIPP [30,31]. and “I” by 1 guideline, VA/DoD [35]. Acupuncture was rated “B” by 2 out of 4 guidelines, ACP/APS and VA/DoD [11,35], “I” by 1 guideline, NASS [29], and recommended by 1 guideline, OPTIMa [26].
Despite consistency in the recommendation of common discectomy techniques, tubular discectomy, a variation of microscopic discectomy, was recommended for the treatment of HLD by 6 out of 7 guidelines, but was determined to have insufficient evidence by 1 guideline. It was recommended by AANS, AAOS, and ISASS [28,32,33], rated “A” by PSSS [27], “B” by ASPN and JOA [25,30], but “I” by NASS [29]. Percutaneous endoscopic discectomy was recommended by 7 out of 8 guidelines and was of insufficient evidence according to 1 guideline. Percutaneous endoscopic discectomy was recommended by AANS, AAOS, and ISASS [28,32,33]. It was given a grade of “B” by ASPN and JOA [25,30], and “C” by ASIPP and PSSS [27,31]. However, percutaneous endoscopic discectomy was rated “I” by NASS [29]. Other surgical interventions that treated HLD included nucleoplasty, herniotomy/sequestrectomy, and fusion. Nucleoplasty was recommended by 1 out of 3 guidelines and determined insufficient evidence by 2 guidelines. It was rated “B/C” by ASIPP [31], and “I” by L&I and NASS [29,34]. Herniotomy/sequestrectomy was rated “B” for decompression by NASS [29], “I” by PSSS [27], and “D” by JOA [25]. Fusion was recommended for recurrent HLD by 2 guidelines and deemed of insufficient evidence for primary HLD by 3 guidelines. It was recommended by L&I for recurrent HLD with multiple failed attempts of decompression, rated “C” for recurrent HLD by PSSS [27,34], and determined “I” for primary HLD by NASS, JOA, and PSSS [25,27,29]. Spinal cord stimulation, peripheral nerve field stimulation, and percutaneous adhesiolysis were pain treatments requiring surgical procedures and assessed by 2 guidelines. Spinal cord stimulation was rated “B” by ASIPP and “I” by NASS [29,31]. Peripheral nerve field stimulation was rated “C” by ASPN and “I” by NASS [29,30]. Percutaneous adhesiolysis was rated “B” by ASIPP and “I” by PSSS [27,31]. In addition, by heating the posterior annulus and destroying afferent nociceptors, intradiscal electrothermal annuloplasty has been used to treat discogenic low back pain and was rated “B” by 1 guideline, ASIPP [31], and “I” by 2 guidelines, NASS and PSSS [27,29].
4. Overall Treatment Recommendation
Treatments evaluated by only 1 guideline, namely regenerative therapy, basivertebral nerve ablation, intradiscal high-pressure saline injection, aggressive discectomy, and medial facetectomy were excluded from Figs. 2–4 since additional evidence is likely needed to fully appraise treatment efficacy.
Among noninvasive treatments, cognitive behavioral therapy and nonsteroidal anti-inflammatory drugs received the highest positive recommendation scores of 2.2 and 2.0, respectively, indicating strong recommendation for the treatment among guidelines. Moderate recommendations were found for back schools, physical therapy, spinal manipulation, and exercise, scoring 1.7, 1.6, 1.3, and 1.2, respectively. Traction and bracing were weakly recommended for HLD treatment, scoring 0.63 and 0.50, respectively. Bedrest received a recommendation score of 0, suggesting the potential for future guidelines and additional evidence to further evaluate the efficacy of bedrest in HLD treatment. Transcutaneous electrical nerve stimulation was the only noninvasive treatment assessed that was overall not recommended for the treatment of HLD, which had a negative recommendation score of -0.75.
Short-term transforaminal, caudal, and interlaminar epidural steroid injections were invasive, nonsurgical measures strongly recommended for pain management in HLD, scoring 2.3, 2.1, and 2.0, respectively. Radiofrequency ablation also received a high positive recommendation score of 2.0, suggesting a relatively strong recommendation for the use of radiofrequency ablation in reducing pain associated with HLD. Acupuncture was moderately recommended for HLD treatment, scoring 1.5. Percutaneous peripheral nerve stimulation and intra-articular lumbar facet joint injections were weakly recommended, scoring 0.7 and 0.3, respectively. On the other hand, long-term epidural steroid injections were not recommended for HLD treatment, scoring -0.3.
For surgical measures, open discectomy, microdiscectomy, and microendoscopic discectomy were all strongly recommended for HLD treatment, with recommendation scores of 2.1, 2.1, and 2.0, respectively. Tubular discectomy was also recommended with a moderately high score of 1.9. Percutaneous endoscopic discectomy was moderately recommended for HLD treatment, scoring 1.5. Although fusion was moderately recommended for recurrent HLD with a score of 1.5, it scored 0 for primary HLD, indicating a need for further investigation to assess its efficacy in primary HLD. Moderate to weak recommendations were observed for spinal cord stimulation and percutaneous adhesiolysis, both scoring 1.0. Intradiscal electrothermal annuloplasty was weakly recommended for HLD treatment, scoring 0.7. Both nucleoplasty and peripheral nerve field stimulation received a score of 0.5, suggesting weak recommendations for their application in HLD treatment. Sequestrectomy/herniotomy received a small positive recommendation score of 0.3.
DISCUSSION
1. Recommendations for Noninvasive Treatments
Noninvasive treatments should generally be thoroughly explored before surgery is indicated. Oral NSAIDs were recommended by ASPN, ACP/APS, ASIPP, AANS, AAOS, L&I, VA/DoD, JOA, and OPTIMa [11,25,26,30-35]. The VA/DoD guideline also cited serotonin-norepinephrine reuptake inhibitors (SNRI), particularly duloxetine, for their potential to improve function and reduce low back pain [35]. Such therapeutic effect has been proposed to extend to other SNRIs as well [35].
Nonpharmacological interventions have been cited in the literature to affect pain. Cognitive behavioral therapy, modified to address chronic pain, has shown fair to good evidence of benefit [11,26,31,34,35]. Though HLD could cause inflammatory and neuropathic pain, its psychosomatic aspect might respond to cognitive behavioral therapy [31].
Movement remains an important part of HLD management. AANS and VA/DoD explicitly recommended against bedrest for this reason [32,35]. Light exercise and physical therapy were recommended by many sources including the AANS and other professional associations [32]. The JOA and NASS, however, reported insufficient high-quality evidence to make that assertion [25,29].
2. Recommendations for Invasive Nonsurgical Treatments
Epidural steroid injections carry 3 optional approaches: interlaminar (midline or parasagittal), transforaminal, or caudal. The ideal approach can depend on the patient’s surgical history and can affect the specificity in targeting the ventrolateral space where the nerve root lies. Upon reviewing 36 randomized controlled trials, the ASPN concluded high-quality evidence supporting the efficacy of all 3 approaches in treating disc disease [30]. The ASPN based its recommendations on the work of ASIPP and naturally agreed [30,31]. The 2014 NASS guideline agreed with ASPN and ASIPP in short-term transforaminal epidural steroid injections, whereas insufficient data was found to support their long-term efficacy [29-31]. Nevertheless, though not explicitly commenting on long-term transforaminal epidural steroid injections, the 2013 ASIPP guideline cited a literature review by Benny & Azari that reported evidence for their long-term effectiveness [31,42].
Based on a meta-analysis of randomized controlled trials, the JOA noted that no trials compared epidural steroid injections to no treatment [25]. All comparisons were made based on varying injection approaches and steroid presence/concentration [25]. The JOA only moderately recommended epidural steroid injections perhaps due to a stronger emphasis on potential complications and patient preference [25]. Some sources cited leg pain/radiculopathy as an indication for epidural steroid injections, including the NASS guideline [29].
Radiofrequency and basivertebral nerve ablations both use radio waves to heat the target area. Radiofrequency ablation in the spine has been used to damage nerves at a joint that is causing pain, while basivertebral nerve ablation targets and ablates the basivertebral nerve within the vertebral body [43,44]. The ASPN analyzed 14 randomized controlled trials that showed the efficacy of radiofrequency and basivertebral nerve ablations [30]. However, this was in the context of facet joint pain and vertebral endplate damage respectively rather than HLD [30].
The benefit of short-term epidural steroid injections in pain relief has largely been established, whereas intra-articular lumbar facet joint injections have shown limited supporting results. Guidelines expressed concerns about complications associated with intra-articular lumbar facet joint injections, including infection, bleeding, and more rarely, neurological injuries [30,31].
3. Recommendations for Surgical Treatments
Discectomy is an established procedure for treating primary HLD with radiculopathy after unresponsiveness to nonsurgical care or for cases presenting severe symptoms [29,45]. Open microscopic discectomy is the most well-established method to remove herniated discs, but others have been developed to reduce pain and complications such as microendoscopic discectomy, percutaneous discectomy, and tubular discectomy [30]. An additional component called annular closure has also been introduced to the procedure to prevent reherniation and showed effectiveness in a 2019 meta-analysis and a 2023 randomized controlled trial [45,46]. In comparing different approaches, all 3 show similar therapeutic outcomes. The endoscopic discectomy approach is noninferior to the well-established microdiscectomy [47], but ASPN, NASS, JOA, and other groups have not yet determined one approach superior to the rest [25,29,30]. This might be difficult to parse given the differing surgeon experience and unique patient characteristics. Additionally, these methods might be investigated more heavily by groups outside the US.
Minimally invasive percutaneous nucleoplasty, which uses radiofrequency to evaporate the nucleus pulposus, has shown safety in 2 studies [48,49]. Nonetheless, only 3 guidelines commented on this procedure, with NASS and L&I determining nucleoplasty as lacking sufficient support for its implementation [29,34].
Nerve stimulators received limited attention, and no agreement emerged on their recommendation grades. For each stimulation therapy, at least one guideline reported insufficient evidence, suggesting the potential for further investigation.
4. Temporal Changes in Treatment Recommendation
The 12 guidelines reviewed ranged from the ACP/APS guideline in 2007 to the AANS guideline in 2024, through which constant updates were added to the literature [11,25-35]. We compared guidelines based on their publication year, dividing them into 2 groups. Guidelines by 3 societies, ACP/APS, ASIPP, and NASS, belonged to the pre-2015 cohort [11,29,31], while guidelines by 9 associations, PSSS, OPTIMa, ISASS, VA/DoD, L&I, JOA, ASPN, AAOS, and AANS, belonged to the post-2015 cohort [25-28,30,32-35].
Regenerative therapy, which explores the potential of biological materials in restoring disc integrity and reducing inflammation and pain [50], is an active field of research that was not addressed in pre-2015 guidelines. Only one post-2015 society, ASPN (2022), assessed regenerative therapy, noting insufficient evidence [30]. Therefore, as research in regenerative therapy progresses, future guidelines could investigate its efficacy.
Percutaneous peripheral nerve stimulation began to gain attention after 2015. Initially rated as lacking sufficient evidence by ACP/APS (2007) [11], it was weakly recommended by VA/DoD (2017) and ASPN (2022), suggesting growing research and clinical application [30,35].
Temporal changes were also observed among surgical measures, especially minimally invasive procedures. Intradiscal electrothermal annuloplasty was moderately recommended by ASIPP (2013) [31], but was labeled as insufficient evidence by NASS (2014) and PSSS (2016) [27,29]. Thus, as more evidence became available, the confidence in the net benefit of intradiscal electrothermal annuloplasty began to diminish. Percutaneous endoscopic discectomy was either weakly recommended or considered lacking evidential support before 2015 by ASIPP (2013) and NASS (2014), respectively [29,31]. However, overall increase in recommendation strength and consistency in recommendation direction were shown among post-2015 guidelines. Similarly for tubular discectomy, NASS (2014) was unable to assign a recommendation grade due to insufficient evidence [29], but all post-2015 guidelines that evaluated tubular discectomy recommended it for HLD treatment despite slight variations in recommendation strength. The increase in recommendation confidence for percutaneous endoscopic discectomy and tubular discectomy suggested recent advances in the development and utility of minimally invasive techniques.
5. Limitations
This review is limited to the treatments explicitly mentioned in the selected guidelines. Several associations—AANS, AAOS, L&I, OPTIMa, and ISASS—did not recommend treatments based on a grading system [26,28,32-34]. Thus, recommendations were inferred from descriptions and explanations of effectiveness, and the distinction between grades “A”, “B”, and “C” could not be made. Figs. 2–4 might include errors due to the generalized nature of the “R” grade, rendering an accurate quantification of “R” impossible. In addition, Figs. 2–4 did not account for guideline quality, which could vary due to different guideline development methods.
The AANS guideline only broadly recommended discectomy without specifying approaches [32]. An assumption was made when compiling Table 4 that discectomy implied common discectomy techniques, including open-, micro-, and endoscopic discectomy, as well as their subtypes and variations.
Inconsistencies in treatment recommendations could also be attributed to cultural differences. For example, the JOA grading criteria emphasized treatments relevant to the clinical practice in Japan while omitting other uncommonly used treatments [25]. Considering how certain treatments have different prevalence in different countries, country-specific guidelines could have special value in addressing country-specific medical needs and preferences.
Finally, only English-language guidelines were included, introducing language and accessibility bias. The clinical practice of HLD treatment in non-English-speaking countries could be underrepresented. The guidelines written in non-English languages could provide additional insights into HLD treatment in international settings. Therefore, omitting guidelines in non-English languages might limit the global applicability of this review.
CONCLUSION
The treatment and management of HLD remains a complex paradigm due to the numerous treatment options for patients. Variations in treatment recommendations across treatment guidelines could potentially be a result of the complex etiology of pain in HLD. Higher evidential confidence in recommendation estimates due to adequate support from the literature was reported for established nonsurgical treatments such as oral nonsteroidal anti-inflammatory drugs, cognitive behavioral therapy, short-term epidural steroid injections, and surgical measures like open, micro-, and endoscopic discectomy. Fusion, while lacking sufficient evidence for treating primary HLD, was moderately recommended for treating recurrent HLD. Most nonsurgical treatments and newer surgical innovations were subject to lesser certainty in net benefit estimates. To address the discrepancies in guideline recommendations, especially for emerging and noninvasive treatments, future research could include and evaluate some of the newer methods for HLD treatments.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement
Print
