Lumbar Disc Herniation

Abstract

Purpose of Review

Substantial advancements have been made in the cause, diagnosis, imaging, and treatment options available for patients with lumbar disc herniation (LDH). We examined the current evidence and highlight the concepts on the frontline of discovery in LDH.

Recent Findings

There are a myriad of novel etiologies of LDH detailed in recent literature including inflammatory factors and infectious microbes. In the clinical setting, recent data focuses on improvements in computer tomography as a diagnostic tool and non-traditional injection options including tumor necrosis alpha inhibitors and platelet-rich plasma. Operative treatment outcomes have focused on minimally invasive endoscopic approaches and demonstrated robust 5-year post-operative outcomes.

Summary

Advances in the molecular etiology of LDH will continue to drive novel treatment options. The role of endoscopic treatment for LDH will continue to evolve. Further research into10-year outcomes will be necessary as this surgical approach continues to gain widespread popularity.

Introduction

Nearly 80% of the population sustains an episode of low back pain (LBP) once during their lifetime [1]. Due to its high prevalence and significant contribution to disability, LBP incurs an annual cost exceeding $100 billion in the USA [1, 2]. Within the vast differential of LBP, the most common source is intervertebral degeneration leading to degenerative disc disease and lumbar disc herniation (LDH) [2]. Thus, an effective understanding of LDH, its origins, and how to appropriately treat LDH is of substantial importance.

Pathology of LDH

The intervertebral disc consists of an inner nucleus pulposus (NP) and an outer annulus fibrosus (AF). The central NP is a site of collagen secretion and contains numerous proteoglycans (PG), which facilitate water retention, creating hydrostatic pressure to resist axial compression of the spine [3, 4]. The NP is primarily composed of type II collagen, which accounts for 20% of its overall dry weight. In contrast, the AF functions to maintain the NP within the center of the disc with low amount of PG; 70% of its dry weight is comprised of primarily concentric type I collagen fibers [3, 5, 6]. In LDH, narrowing of the space available for the thecal sac can be due to protrusion of disc through an intact AF, extrusion of the NP through the AF though still maintaining continuity with the disc space, or complete loss of continuity with the disc space and sequestration of a free fragment.

Several changes in the biology of the intervertebral disc are thought to contribute to LDH. These include reduced water retention in the NP [5, 7, 8], increased percent of type I collagen within the NP and inner AF [9], degradation of collagen and extracellular matrix (ECM) materials [10], and upregulation of systems of degradation such as apoptosis, matrix metalloproteinase (MMP) expression, and inflammatory pathways [11].

Genetic Predisposition

A myriad of genes are involved in separate processes which predispose to LDH. It is estimated that the condition has approximately 75% heredity origin [12]. Genes that have been found to significantly increase risk of LDH include those encoding structural proteins, matrix metalloproteinases, apoptosis factors, growth factors, and single nucleotide polymorphisms in the vitamin D receptor gene resulting in inflammatory cytokine imbalance [11].

Dehydration

Dehydration is known to contribute to the pathogenesis of degenerative disc disease [7, 8]. Although a specific genetic polymorphism has not been associated with LDH, aquaporins have been implicated. A recent study examined the removed discs of 18 patients who underwent surgery for L4-5 or L5-S1 herniation, and found aquaporin-I (AQPI) to be linearly correlated with preoperative T2-weighted magnetic resonance imaging (MRI) signal intensity [13•]. This suggests that AQPI may have a role in this dehydration, which is known to contribute to degeneration. However, more research is necessary to better elucidate the role of aquaporins in the pathogenesis of LDH.

Axial Overloading

Not all disc herniation occurs in the context of degenerative disease. A subset of patients with LDH lack evidence of severe degenerative disc including proteoglycan and water loss in the NP [14]. In these cases, herniation occurs as a result of spinal overloading [15, 16]. A recent study of caprine intervertebral discs found that static overloading in particular, as compared to physiological loading and dynamic overloading, put the disc at risk for posterior herniation [17•]. The authors suggest that this may be the mechanism for the increased prevalence of lower back pain and herniation in younger individuals that live a sedentary and seated lifestyle.

Etiology of Symptoms

Several processes have been investigated in an attempt to understand how LDH produces its symptoms. These include inflammatory signaling, the presence and effects of Propionibacterium acnes, contributions of an acidic environment, and microstructural changes to the nerve root.

Inflammation

The role of inflammatory signaling in producing nerve pain in LDH has been well-established. The intervertebral disc is an immunoprivileged area. Consequently, the contents of the intervertebral disc, specifically the NP, are immunoreactive if they are found beyond their normal physiological boundaries. As the NP tissue extrudes into the epidural space, vascular endothelial cell changes trigger increased vascular permeability, vasodilation, the adhesion and migration of immune cells to the site, and inflammatory cytokine signaling [18]. Moreover, several inflammatory factors including COX-2, its upregulator follistatin-like protein 1 (FSTL1), and tumor necrosis factor alpha (TNF-α) have been shown to have significantly higher local and systemic concentrations in patients with LDH compared to non-herniated controls [19–22, 23•].

Propionibacterium acnes

Propionibacterium acnes is a Gram-positive, facultative anaerobic, fastidious bacterium, which may play a central role in LDH [24•]. In 2001, one study demonstrated that 53% patients with severe radiculopathy were found to have Gram-positive anaerobes. Nearly 84% of these patients were found to have P. acnes [25]. Recently, a significant portion of herniated discs (11%) have been shown to have a high P. acnes bacterial load (> 1000 CFU/mL) [26]. Aghazadeh et al. also found a high prevalence of P. acnes DNA and correlated it to vertebral edema as seen on preoperative MRI [27]. Chen et al. found that the bacterium was able to induce disc degeneration outright in rabbit models [28]. Furthermore, long-term antibiotics targeting anaerobes may improve lower back pain in those with disc herniation and vertebral edema [22]. Despite this, however, other studies have failed to correlate P. acnes prevalence with LDH pathology [29]. While it is likely that P. acnes may play a role in LDH symptomology and ultimately be a target for non-operative treatment, further research must be done to elucidate the measure of effect this bacteria has on symptomatic LDH.

Acidic Environment

Degenerated lumbar discs have a decreased pH (~ 1.0) compared to that of healthy intervertebral discs, which is likely due to decrease metabolite exchange [30]. The consequence of this acidity has been increased expression of multiple factors which inhibit cell proliferation and increase cell apoptosis of NP cells [31•, 32]. Recent data in has shown that these changes are blunted in a dose-dependent manner with the administration of ASIC blockers, and that in non-human models, ASIC blockade improved pain threshold [31•].

Clinical Presentation

Signs and Symptoms

The primary signs and symptoms of LDH are radicular pain, sensory abnormalities, and weakness in the distribution of one or more lumbosacral nerve roots [33, 34]. Focal paresis, restricted trunk flexion, and increases in leg pain with straining, coughing, and sneezing are also indicative [33, 34]. Patients frequently report increased pain when sitting, which is known to increase disc pressure by nearly 40% [35].

The affect dermatome varies based on level of herniation as well as herniation type. In paracentral herniations, the transversing nerve root is affected versus in far lateral herniations, the exiting nerve root is affected. For example, a paracentral herniation at L4-5 would cause L5 radiculopathy whereas a far lateral herniation at the same level would cause L4 radiculopathy.

Pain that is relieved with sitting for forward flexion are more consistent with lumbar spinal stenosis (LSS), as the latter motion increases disc pressure by 100–400% and would likely increase pain in isolated LDH [35]. Rainville et al. recently compared signs of LDH with LSS and found that LSS patients are more likely to have increased medical comorbidities, lower levels of disability and leg pain, abnormal Achilles reflexes, and pain primarily in the posterior knee [36].

Diagnostic Guidelines

In 2014, The LDH with Radiculopathy Work Group of the North American Spine Society’s (NASS) Evidence-Based Guideline Development Committee recommended manual muscle testing, sensory testing, and supine SLR test (and its crossed leg variant) as the gold standard for clinical diagnosis of LDH. Other tests such as the cough impulse test, hyperextension test, femoral nerve stretch test, lumbar range of motion, and absence of reflexes were not found to be as clinically helpful. A recent meta-analysis concluded that initial screening by the SLR test in conjunction with three of the following four symptoms in a nerve root distribution is sufficient for clinical diagnosis of LDH with radiculopathy: dermatomal pain, sensory deficits, reflex deficits, and/or motor weakness [37].

Factors for Increased Risk of Cauda Equina Syndrome

Cauda equina syndrome (CES) is a rare but devastating consequence of LDH. Krishnan et al. identified diabetes, acute onset of symptoms, L3-L4 involvement, sequestrated discs (fully separated NP from AF), superiorly migrated discs, posterior herniation, primary canal stenosis, and greater canal compromise as risk factors for CES [38•]. The presence of ≥ 4 of these factors produced a significantly higher chance of CES (sensitivity 74%, specificity 77%).

Imaging

Radiographs

Plain radiographs are the first-line imaging modality used in low back pain. For the treating primary care physician, radiographs should be obtained only after 6–12 weeks in the absence of neurologic compromise. Given that radiographs provide only a static understanding of the spine, we recommend that in addition to anteroposterior (AP) and lateral images, flexion and extension sequences are obtained to evaluate the role of instability in the patient’s symptoms. Findings suggestive of LDH in this modality include compensatory scoliosis, narrowed intervertebral space, and the presence of traction osteophytes.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is the gold standard for imaging to confirm suspected LDH with a diagnostic accuracy of 97% and high inter-observer reliability [39, 40]. MRI findings of increased T2-weighted signal from the posterior 10% of the disc diameter are highly suggestive of disc herniation [41]. However, given the significant resource use involved in this testing method, it is not indicated for all patients with LDH. Relative indications for MRI in the early period of LDH (< 6 weeks) include neurologic motor deficits and CES.

Diffusion tensor imaging (DTI) is a type of MRI that can be used to detect microstructural changes in the nerve roots in patients with LDH. Wu et al. assessed these changes and correlated them to disability score on the Oswestry Disability Index (ODI) and duration of sciatica symptoms [42]. Two microstructural parameters, low fractional anisotropy (more restricted osmosis in the tissue) and high apparent diffusion coefficient (decreased microstructural integrity of the nerve root), were associated with ODI score and symptom duration. This suggests that DTI may be able to be used to better understand the changes that occur in nerve roots due to compression in LDH, and differentiate patients between surgical and nonsurgical intervention.

Modic Changes

Modic changes are MRI signal changes representing categorized pathological changes occurring in vertebrae [43]. Modic type 1 changes are vascular developments in the vertebral body including inflammation and edema. Type 2 changes involve fatty replacement of the vertebral bone marrow. Type 3 changes, which are less common, are fractures or changes of the trabecular bone of the vertebrae. Modic type 1 changes have been shown to correlate significantly with the degree of degenerative disk disease [44].

Computed Tomography

While computed tomography (CT) was previously thought to be clinically inferior to MRI in LDH detection, advances including multidetector CT (MDCT) have brought the diagnostic level of CT to be nearly equal to that of MRI [45, 46]. The NASS Evidence-Based Guideline Development Committee recommends CT myelography as an appropriate diagnostic tool for confirming suspected LDH as an alternative to MRI [39]. There are several circumstances where CT myelography would be chosen as opposed to MRI including situations where MRI is unavailable or not possible (i.e., pacemakers or cochlear implants), and where patients would be excessively uncomfortable (claustrophobia or intractable back pain). However, given the invasive nature of this test, CT myelogram requires the assistance of a trained radiologist and is associated with risks including post-spinal headache (most common), radiation exposure, and meningeal infection.

Treatment

Non-operative

Non-operative management of symptomatic LDH is the treatment of choice for the majority of patients. Gugliotta et al. recently demonstrated equivalent medium- and long-term outcomes for conservative and surgical treatment of LDH [47]. However, other studies have also demonstrated improved 1- and 2-year outcomes in the surgical treated group when compared to conservatively managed patients. There has been no recent literature which clarifies the absolute non-operative versus operative criteria in LDH. Relative indications for acute surgery are discussed in the next section. Non-operative management should consist of a multimodal approach including anti-inflammatory medications, education, and physical therapy [48].

Local corticosteroid injections (CSI) are a commonly used technique for both the diagnosis and treatment of LDH. However, while interlaminar injection in acute LDH has shown a strong QALY benefit of $2050, the optimal patient profile for maximum benefit remains unknown [49•]. Historical success rates of these injections in acute LDH are 20–95% [50] and are thought to be due to reduction in local inflammatory cytokine concentration. When compared to placebo injections, historical data suggests CSI offer short-term improvements in leg pain and sensory deficits but may not reduce the rate of surgical intervention or provide substantial functional benefit [51]. Prior data has shown that transforaminal injections produce superior results compared to caudal or interlaminar approach [52]. In an effort to enhance the success of this treatment, recent literature has examined the role of patient position during epidural CSI injection. Altun et al. found that injection in the lateral decubitus position provided statistically more 6-month ODI improvement and 12-month numeric rating scale improvement compared to prone positioning [53•].

Kim et al. evaluated radiologic factors predicting successful outcomes in nerve injections [54]. Their study demonstrated that LDH smaller than 6.23 mm demonstrated significant clinical improvement with transforaminal epidural steroid injection. However, their study also found that LDH greater than 6.31 mm demonstrated significant surgical benefit. Thus, measurement of disc herniation size may be of statistical, but minimal clinical, significance in determining likelihood of injection efficacy.

In addition to patient position, optimum injection medication is still under debate. As aforementioned, TNF-α is a critical molecular mediator in the pathogenesis of radiculopathy. However, the results of TNF-α blockade as a clinical treatment modality have been mixed thus far [55]. Korhonen et al. found the TNF-α inhibitor infliximab to be effective in treating sciatica in patients with LDH [56]. More recent data has shown that subcutaneous injection and epidural administration of TNF alpha inhibitors have shown clinical benefit, while intradiscal injection has shown no benefit at low dose [57]. In a placebo-controlled randomized trial of transforaminal TNF-α inhibitor injection, Freeman et al. demonstrated a significant 3- to 6-month improvement in both worst leg pain and worst back pain scores with etanercept injections [55]. However, there was no dose-dependent response associated with this injection and no comparison population to CSI injection.

Much debate surrounds the use of non-traditional treatment modalities in the treatment of LDH. The combined effect of multi-modal non-traditional therapies including herbal supplementation, acupuncture, bee-venom pharmacopuncture, and spinal manipulation was investigated recently by Shin et al. [58]. Their study found that the long-term, 5-year, outcomes of this treatment method demonstrated broad improvements in VAS, ODI, and SF-36 QOL scores [58]. However, nearly 54% had a relapse of either low back or radicular pain necessitating intervention including repeated non-operative therapy, injections, or surgery. Additionally, placebo effect cannot be excluded due to lack of a control population and recent data demonstrating a 66% spontaneous resorption rate in LDH [59].

Another study evaluated the effect of traction therapy on short-term outcomes [60]. Both groups of patients using traction demonstrated improvement in 28-day outcomes, though there was no traction weight-dependent increase in improvement. The group allocated to traction equivalent to 10% of their body weight demonstrated better overall 28-day VAS scores compared to the 50% body weight group [60]. There was statistical decrease in opiate consumption in all groups and a reduction in the EIFEL disability scores for the 10% traction group. However, this study was limited by a small sample size of 17 patients and maximum of 2-week patient follow-up.

Traditional western formal physical therapy (PT) focused on exercise, core strengthening, and joint mobility are known to improve symptoms related to LDH [61]. Using the SPORT trial data, Thackeray et al. evaluated the profile of patients utilizing formal PT and their outcomes [62]. On multivariate analysis, those patients receiving early formal PT were more likely to have higher baseline ODI scores, prefer non-operative treatment, and have an associated neurological deficit [62]. Within the first 6 weeks, there was no difference in the outcomes between the two groups. There was less 1-year low back pain in the formal physical therapy group, but no difference in pain, ability, or cross-over to surgery at 1 year compared to the non-formal PT group. Given the current national opioid epidemic, perhaps the most important finding was significantly less opiate usage in the formal PT group, though they had higher use of muscle relaxants compared to the non-formal PT group.

The regenerative effects of mesenchymal stem cell (MSC) therapy and wound healing benefits of platelet-rich plasma (PRP) injection for LDH are increasingly reported [63, 64]. Several studies have demonstrated improvement in clinical scores including ODI, VAS, and JOA beginning 1 month post-PRP injection [64–67]. Additionally, these studies have reported no complications associated with MSC and PRP injections [64]. However, the number of patients with outcomes reported in the literature is still too low for widespread clinical implementation.

Operative

Operative management of LDH in several large studies has been previously associated with improved short-term benefits and conflicting value in the medium- to long-term range [68, 69]. This finding was echoed by a recent randomized Finnish study comparing non-operative treatment with microdiscectomy in LDH [70]. Their study found no difference in 2-year measurements such as ODI score HRQoL between the two groups, but did demonstrate significant improvement in leg pain at 6 weeks, and patient reported treatment satisfaction at 2 years. Additionally, the most novel finding of this study was a subgroup analysis, which demonstrated that microdiscectomy of L4-5 LDH resulted in superior patient reported outcomes compared to non-operative management including subjective work ability, ODI, and HRQOL scores [70]. From a financial perspective, a recent Swiss study demonstrated that the incremental cost per quality-adjusted-life-year with surgical treatment was approximately $73,245 [71].

Factors recently predicting successful outcome after discectomy include preoperative higher leg pain severity, better mental health status, shorter symptom duration, younger age, increased preoperative physical activity, and severe preoperative low back pain [72•, 73, 74]. Interestingly, the presence of motor deficit, vertebral level or side of herniation, gender, and type I modic changes were not found to affect postoperative outcomes.

Minimally Invasive Surgery

Minimally invasive approaches to spine surgery have been pioneered and increasingly utilized over the last 15 years. These approaches are associated with less soft tissue and bony trauma, lower acute care charges, decreased hospital length of stay, but also with a higher learning curve [75, 76]. As such, the outcomes of these procedures are increasingly reported.

There are several recognized percutaneous endoscopic approaches to LDHs: interlaminar, transforaminal, posterolateral, and transiliac [77, 78]. As an aggregate group, endoscopic discectomy is associated with decreased operative time and less blood loss with no increase in overall complications, reoperation rates, or wound infection when compared to open discectomy [79]. However, with respect to long-term patient-centered outcomes, a double-blind randomized control trial failed to detect a difference between open and endoscopic surgery in 325 patients [80].

Choi et al. evaluated the outcomes of 149 patients undergoing percutaneous endoscopic lumbar discectomy (PELD) for migrated disc herniations [81]. They found a 90% good or excellent outcome rate and improvement of 45 ODI points at 1-year follow-up. However, high-grade disc herniations with upward migration demonstrated a 13% rate of remnant disc fragment and 3% revision surgery rate.

Given the difficulty in utilizing the endoscopic approach with high iliac crests, Bai et al. applied a transiliac approach to endoscopic discectomy [77]. In comparison to the conventional transforaminal approach, there was no difference in 12-month outcomes, usage of fluoroscopy, or perioperative neurovascular injury. Given that each patient required a pre-operative CT scan to identify optimal starting point and risk of significant vascular injury, the results are promising but do not yet clearly establish endoscopic transiliac surgery as a superior option for the dysmorphic pelvis.

Interlaminar Approach

The interlaminar approach has the advantage of direct insertion under endoscopic guidance [78, 82]. However, compared to the aforementioned approaches, interlaminar resection requires nerve root and thecal sac retraction which may present a particular challenge in large LDH [78]. However, both Soman et al. and Tonusu et al. demonstrate significant improvement in outcome indices following LDH resection via this method [78, 82]. However, Tonusu et al. via the shoulder approach had three neurologic perioperative complications, and as such recommend that in large disc herniations with caudal extrusion access via the axillary approach should be obtained.

Long-term Follow-up

Limited long-term data is available with conflicting outcomes. Five-year data combining PELD with percutaneous lumbar foraminoplasty (PLF) demonstrates promising results with improvement in ODI and modified MacNab criteria. When combined with PLF in 134 patients, Li et al. demonstrated a 93% 5-year patient satisfaction rate [83]. Moreover, when evaluating the outcomes of full-endoscopic interlaminar discectomy (FELD), Tu et al. also demonstrated significant 5-year improvement in VAS, ODI, modified MacNab criteria, and 90% 5-year good or excellent patient satisfaction [84•]. However, their patient population had an 8.3% revision surgery rate, 4.2-day mean length of stay, and 6% dural tear rate. In contrast with the above positive findings, one study with 11-year follow-up of PELD patients found a 36% reoperation rate [85]. Additionally, re-herniation rates range from 5 to 15% [94, 95]. Risk factors for failure following PELD include obesity, age > 50, surgeon inexperience, and central disc herniation [86]. Hu et al. study suggests that surgeon inexperience can be modified by preoperative planning software which demonstrated an ability to reduce mean operative and fluoroscopic time [87].

Open Discectomy

As aforementioned, over the last decade, large studies including the SPORT and Maine trials have demonstrated the efficacy of open discectomy in LDH. Approaches for discectomy vary based on herniation type (paracentral versus far lateral). While the paracentral approach has robust utility LDH, it is associated with longer incisions, increased muscle stripping, and more difficulty in far lateral discectomy [88]. Given these difficulties, the Wiltse paraspinal approach between the multifidus and longissimus muscles is a well-recognized method of discectomy in far lateral herniation [89].

Much of the current data regarding open discectomy for LDH involves infection risk as outcome data has been well published over the last several decades. Predictors of infection following microdiscectomy include absence of prophylactic antibiotic dosing and duration of surgery > 68 min [90]. Interestingly, one study did not find that preoperative lumbar epidural CSI increased infection risk [91]. Kotil et al. evaluated 115 patients undergoing single-level discectomy with and without closed suction drains. With MRI post-operative days 1, 180, and 365, their data demonstrated significantly higher rates of epidural hematoma and fibrosis compared to the non-CSD group. However, there was no 12-month difference in ODI and VAS scores, or major post surgical complications [92]. More recently, Murphy et al. evaluated the effect of microscope use during open discectomy in 23,583 patients [93]. Their findings of increased operative time and equivalent perioperative complication rates in discectomy performed with and without microscope assistance led to recommendation for continued use of the microscope during decompression per surgeon preference [93].

Complications

There are several important complications associated with discectomy for LDH. The rate of dural tears following LDH ranges from 1 to 17% and is increased particularly with advanced age, obesity, and revision procedures [94]. The downstream effects of incidental durotomy include increased hospitalization costs by $4000 and wound dehiscence odds by 2.4 times [94]. Other complications include post-operative infection (1–5%), worsening of functional status (4%), and nerve root injury (0.2%) [92, 93]. Risk factors for recurrent herniation include pre-operative disc height index, trauma, older age, smoking, disc protrusion, disc sequestration, longer duration of sick leave, workers’ compensation, greater preoperative symptom severity, and diabetes [74, 95–100]. With a substantial increase in perioperative complications in revision LDH discectomy, minimizing risk factors for recurrence are critical; however, activity restriction has not demonstrated improvement in LDH recurrence rate [101].

Conclusion

LDH is a common etiology in low back pain. In the last 3 years, significant advances have been made in our understanding in the etiology of LDH including microstructural changes, molecular pathways, and microbial load. Additionally, over this same time period, minimally invasive approaches to LDH resection have demonstrated increasingly positive outcomes. As such, this approach remains a key area of investigation in the coming years in addition to better defining the absolute indications for maximal clinical benefit in surgical treatment of LDH.

Compliance with Ethical Standards

Conflict of Interest

Raj M. Amin and Nicholas S. Andrade declare that they have no conflict of interest.

Brian J. Neuman reports grants from Depuy Synthes, outside of the submitted work.

Human and Animal Rights and Informed Consent

This article does not contain any studies with human or animal subjects performed by any of the authors.