디스크 파열 통증.. 요추 2-3번, 요추 4-5번 추간판 .. 동척추신경(체성, 자율신경)지배?

[문형철]

The Spine Journal

Volume 22, Issue 3, March 2022, Pages 472-482

Basic Science

Origin, branching pattern, foraminal and intraspinal distribution of the human lumbar sinuvertebral nerves

Author links open overlay panelMarcus C. Breemer MD a, Martijn J.A. Malessy MD, PhD a, Robbert G.E. Notenboom PhD b

Show more

Add to Mendeley

Share

Cite

https://doi.org/10.1016/j.spinee.2021.10.021Get rights and content

Under a Creative Commons license

Open access

Highlights

• •

  The root origins of high (L2) and low (L5) lumbar human sinuvertebral nerves (SVNs) were investigated in cadavers.

• •

  L2 SVNs are formed by dual somatic and autonomic roots in 90% of cases.

• •

  At L5, 40% of SVNs arise from somatic and autonomic roots; other SVNs are autonomic.

• •

  L5 discogenic low-back pain (LBP) is mediated both segmentally and nonsegmentally in 40% of cases.

• •

  Only targeting the nonsegmental discogenic LBP pathway may provide incomplete pain reduction.

Abstract

BACKGROUND CONTEXT

The lumbar sinuvertebral nerve (SVN) innervates the outer posterior intervertebral disc (IVD); it is thought to mediate discogenic low-back pain (LBP). Controversy, however, exists on its origins at higher (L1–L2) versus lower (L3–L5) lumbar levels. Additionally, lack of knowledge regarding its foraminal and intraspinal branching patterns and extensions may lead to iatrogenic damage.

PURPOSE

To systematically describe the origins of the L2 and L5 SVNs, their morphological variation in the intervertebral foramen (IVF) and intraspinal distribution.

STUDY DESIGN

Dissection-based study of 20 SVNs with histological confirmation in five embalmed human cadavers.

METHODS

The origin, branching pattern and distribution of the L2 and L5 SVNs was investigated bilaterally in five human cadavers using dorsal and anterolateral dissection approaches. Parameters studied included somatic and/or autonomic SVN root contributions, foraminal SVN morphology and course, diameter, branching point, intraspinal distribution and IVD innervation pattern. Nerve tissue was confirmed by immunostaining for neurofilament and S100 proteins.

RESULTS

The SVN and its origins was identified in all except one IVF at L2 and in all foramina at L5. At L2, the SVN arose in nearly 90% of sides from both somatic and autonomic roots and at L5 in 40% of sides. The remaining SVNs were formed by purely autonomic roots. The SVN arose from significantly more roots at L2 than L5 (3.1 ± 0.3 vs. 1.9 ± 0.3, respectively; p=.022). Four different SVN morphologies could be discerned in the L2 IVF: single filament (22%), multiple (parallel or diverging) filament (33%), immediate splitting (22%) and plexiform (22%) types, whereas the L5 SVN consisted of single (90%) and multiple (10%) filament types. SVN filaments were significantly thicker at L2 than L5 (0.48 ± 0.06 mm vs. 0.33 ± 0.02 mm, respectively; p=.043). Ascending SVN filaments coursed roughly parallel to the exiting spinal nerve root trajectory at L2 and L5. Branching of the SVN into ascending and descending branches occurred mostly intraspinal both at L2 and L5. Spinal canal distribution was also similar for L2 and L5 SVNs. Lumbar posterior IVDs were innervated by the descending branch of the parent SVN and ascending branch of the subjacent SVN.

CONCLUSIONS

The SVN at L2 originates from both somatic and autonomic roots in 90% of cases and at L5 in 40% of cases. The remaining SVNs are purely autonomic. In the IVF, the L2 SVN is morphologically heterogeneous, but generally consists of numerous filaments, whereas at L5 90% contains a single SVN filament. The L2 SVN is formed by more roots and is thicker than the L5 SVN. Intraspinal SVN distribution is confined to its level of origin; lumbar posterior IVDs are innervated by corresponding and subjacent SVNs (ie, two spinal levels).

주요 내용

• •시신에서 높은 (L2) 및 낮은 (L5) 요추 동척추 신경 (SVN)의 근원을 조사했습니다.
• •L2 SVN은 90%의 경우 체세포와 자율 신경의 이중 뿌리에 의해 형성됩니다.
• •L5에서 SVN의 40%는 체세포와 자율 신경의 뿌리에서 발생하며, 나머지 SVN은 자율 신경에 속합니다.
• •L5 디스크성 요통(LBP)은 40%의 경우 세그멘탈 및 비세그멘탈 경로를 통해 매개됩니다.
• •비세그멘탈 디스크성 LBP 경로만을 표적으로 삼는 것은 통증 감소가 불완전할 수 있습니다.

초록

배경 및 배경

요추 동척추 신경(SVN)은

후방 외측 추간판(IVD)에 신경 분포를 하고 있으며,

추간판성 요통(LBP)을 매개하는 것으로 생각됩니다.

그러나,

요추 상부(L1~L2)와 하부(L3~L5)에서

그 기원이 어디인지에 대해서는 논란이 있습니다.

또한,

신경의 foraminal 및 척추 내 분지 패턴과 확장 범위에 대한 지식 부족은

의료적 손상을 초래할 수 있습니다.

foraminal and intraspinal branching patterns and extensions

목적

L2 및 L5 SVN의 기원을 체계적으로 기술하고,

추간공(IVF) 내에서의 형태학적 변이 및 척추 내 분포를 조사하는 것입니다.

연구 설계

5구의 부검된 인간 시체에서

20개의 SVN을 해부학적 방법으로 조사하고 조직학적 확인을 진행했습니다.

방법

L2 및 L5 SVN의 기원과 분지 패턴, 분포를

5구의 인간 시체에서 등측 및 전측측 해부학적 접근법을 통해 양측적으로 조사했습니다.

연구 대상 파라미터에는

체성 및/또는 자율 신경 SVN 뿌리 기여도,

포라멘 내 SVN 형태 및 경로, 직경, 분지점, 척추 내 분포 및 추간판(IVD)

신경 분포 패턴이 포함되었습니다.

신경 조직은

신경 필라멘트 및 S100 단백질에 대한 면역 염색으로 확인되었습니다.

결과

SVN 및 그 기원은 L2의 모든 IVF 중 하나를 제외하고 모든 IVF에서, L5의 모든 포라멘에서 확인되었습니다.

L2에서는 SVN이 양측의 약 90%에서 체성 및 자율 신경근에서 발생했으며, L5에서는 40%에서 발생했습니다.

나머지 SVN은 순수한 자율 신경근에서 형성되었습니다.

SVN은 L2에서 L5보다 유의미하게 더 많은 신경근에서 발생했습니다 (3.1 ± 0.3 vs. 1.9 ± 0.3; p=.022). L2 IVF에서는 4가지 다른 SVN 형태가 구분되었습니다: 단일 필라멘트(22%), 다중(평행 또는 분기) 필라멘트(33%), 즉시 분열(22%) 및 신경망형(22%) 유형이었으며, L5 SVN은 단일(90%) 및 다중(10%) 필라멘트 유형으로 구성되었습니다. SVN 필라멘트는 L2에서 L5보다 유의미하게 두꺼웠습니다(0.48 ± 0.06 mm vs. 0.33 ± 0.02 mm; p=0.043). 상승 SVN 필라멘트는 L2와 L5에서 모두 출구 척추 신경근의 경로와 대략 평행하게 진행되었습니다. SVN의 상승 및 하강 분지는 L2와 L5 모두에서 주로 척추 내부에 발생했습니다. 척추관 분포도 L2와 L5 SVN에서 유사했습니다. 요추 후방 IVD는 부모 SVN의 하강 분지와 하부 SVN의 상승 분지에 의해 신경 분포를 받았습니다.

결론

L2의 SVN은 90%의 경우 체성 및 자율 신경근에서 기원하며, L5에서는 40%의 경우에서 그렇습니다. 나머지 SVN은 순수하게 자율 신경성입니다. IVF에서 L2 SVN은 형태학적으로 이질적이지만 일반적으로 수많은 섬유로 구성되어 있으며, L5에서는 90%가 단일 SVN 섬유를 포함합니다. L2 SVN은 더 많은 뿌리로 구성되어 있으며 L5 SVN보다 두껍습니다. 척추 내 SVN 분포는 그 기원의 수준에 국한되며, 요추 후방 추간판은 해당 및 하부 SVN에 의해 신경 분포를 받습니다(즉, 두 척추 수준).

CLINICAL SIGNIFICANCE

Our findings indicate that L5 discogenic LBP may be mediated both segmentally and nonsegmentally in 40% of cases and nonsegmentally in 60% of cases. Failure of lower lumbar discogenic pain treatment may be the result of only interrupting the nonsegmental pathway, but not the segmental one as well. Relating SVN anatomy to microsurgical spinal approaches may prevent iatrogenic damage to the SVN and the formation of postsurgical back pain.

임상적 의미

우리의 연구 결과는 L5 디스크성 요통이 40%의 경우 세그멘탈 및 비세그멘탈 경로를 통해 매개될 수 있으며, 60%의 경우 비세그멘탈 경로를 통해 매개될 수 있음을 나타냅니다. 하부 요추 디스크성 통증 치료의 실패는 비세그멘탈 경로만 차단하고 세그멘탈 경로는 차단하지 않았기 때문일 수 있습니다. SVN 해부학과 미세수술적 척추 접근법을 연관시키는 것은 SVN의 의학적 손상과 수술 후 요통 형성을 예방할 수 있습니다.

• Previous article
• Next article

Keywords

Afferent pathways

Anatomy

Discogenic low-back pain

Innervation

Intervertebral disc

Intervertebral foramen

Recurrent meningeal nerve

Spinal surgery

Introduction

The sinuvertebral or recurrent meningeal nerve of Luschka (ramus meningeus nervi spinalis) [1] is located bilaterally on every vertebral level, innervating the posterior intervertebral disc (IVD) and posterior longitudinal ligament (PLL), vertebral body and pedicles, and associated soft tissues in the intervertebral foramen (IVF) and anterior spinal canal. In humans, the sinuvertebral nerve (SVN) has been described at cervical [2], [3], [4], [5], [6], [7], thoracic [2,3,6,8,9], lumbar [3,6,[8], [9], [10], [11], [12], [13], [14]] and sacral [2,6,9,11] levels, but in greatest detail in the lumbar region (for a recent review, see [15]).

In the classic pattern, the SVN receives a somatic contribution from the spinal nerve or its ventral ramus and an autonomic contribution from the gray ramus communicans, which unite to form a single SVN proper that continues medially in the IVF [8,12]. The SVN is located ventral to the dorsal root ganglion (DRG) in the ventral portion of the IVF, and generally travels alongside the spinal branch of the lumbar artery. Lateral to the PLL, the SVN splits into an ascending and a lesser descending branch, which may or may not have multisegmental ramifications and interconnect with sub- and superjacent SVNs [3,6,9].

Clinically, the SVN is important because it is the most likely mediator of discogenic low-back pain (LBP). In Western countries, chronic LBP is the leading cause of disability, with a large individual and societal burden; its costs are estimated to be 1%-2% of the gross national product [16,17]. The intervertebral disc is the primary pain generator in approximately 25% of chronic LBP cases [18]. The exact role of the SVN in the generation of (discogenic) LBP is not yet fully determined. Surgery in the lumbosacral spinal region frequently leads to LBP [19], which may be caused by unintentional SVN damage during exploration.

Murine studies indicate that lower lumbar discogenic pain is mediated by two separate pain pathways: segmentally via the somatic SVN root to its corresponding level DRG neurons, and nonsegmentally via the autonomic SVN root, with nociceptive fibers ascending through gray rami communicantes and the sympathetic trunk to L1–L2 DRG neurons [20], [21], [22], [23], [24]. It is not known whether L5 LBP in humans is similarly transmitted nonsegmentally to L1–L2 DRGs. Remarkably, the murine model is used to justify several therapies, which presumably target the nonsegmental pathway, frequently with disappointing results [25,26].

Thus far, cadaver studies on the origin of the human lumbar SVNs have produced conflicting results, with some studies finding exclusively autonomic SVN origins at all lumbar levels [6] and others indicating a classic dual root SVN origin at higher lumbar (L1–L2) levels, but a purely autonomic one at lower lumbar (L3–L5) levels [13]. More recently, Zhao et al. [14] reported both origins, but their study does not indicate if a dual root lumbar origin co-exists, nor if the somatic and/or autonomic root contributions differ per lumbar level.

Precise knowledge of lumbar SVN origins is, however, crucial to unraveling discogenic LBP pathways and in treating it. We therefore studied the lumbar anatomy of the SVN and specifically focused on its somatic and/or autonomic origins at higher (L2) vs lower (L5) lumbar levels. Additionally, we provide a detailed description of its foraminal and intraspinal branching pattern and distribution at these lumbar levels.

소개

Luschka의 동척추 신경 또는 재발성 뇌막 신경 (ramus meningeus nervi spinalis) [1]은 모든 척추 수준에 양측으로 위치하며, 후방 추간판 (IVD) 및 후방 종방 인대 (PLL), 척추체 및 척추 경, 그리고 추간공 (IVF) 및 전방 척추관 내의 관련 연조직에 신경 분포를 합니다. 인간에서 동척추 신경(SVN)은 경추[2], [3], [4], [5], [6], [7], 흉추[2,3,6,8,9], 요추[3,6,[8], [9], [10], [11], [12], [13], [14]] 및 천추부 [2,6,9,11] 수준에서 보고되었지만, 요추부에서 가장 상세히 기술되었습니다(최근 리뷰는 [15] 참조).

전형적인 패턴에서 SVN은 척추 신경 또는 그 전방 가지로부터 체성 신경 기여를 받고, 회색 연결 가지로부터 자율 신경 기여를 받아 단일 SVN 본체를 형성하며, 이는 IVF 내측으로 계속됩니다 [8,12]. SVN은 IVF의 복측 부분에 위치한 DRG의 복측에 위치하며, 일반적으로 요추 동맥의 척추 분지와 함께 진행됩니다. PLL의 측방에서 SVN은 상승 분지와 하강 분지로 분열되며, 이 분지들은 다분절 분지화를 가질 수 있으며, 상하위 SVN과 연결될 수 있습니다 [3,6,9].

임상적으로 SVN은 디스크성 요통(LBP)의 가장 가능성이 높은 매개체이기 때문에 중요합니다. 서구 국가에서 만성 요통은 장애의 주요 원인이며, 개인 및 사회에 큰 부담을 초래하며, 그 비용은 국내 총생산의 1%-2%로 추정됩니다 [16,17]. 만성 요통 사례의 약 25%에서 추간판이 주요 통증 발생원입니다 [18]. SVN이 (디스크성) LBP 발생에 미치는 정확한 역할은 아직 완전히 규명되지 않았습니다. 요추-천추 척추 부위 수술은 자주 LBP를 유발하며 [19], 이는 탐색 과정에서 SVN의 우발적 손상으로 인해 발생할 수 있습니다.

쥐 실험 결과, 하부 요추 추간판성 통증은 두 개의 별도 통증 경로를 통해 매개됩니다: 체성 SVN 신경근을 통해 해당 수준 DRG 신경세포로 전달되는 세그멘탈 경로와, 자율신경 SVN 신경근을 통해 회색 연결 가지를 거쳐 교감신경 간을 통해 L1–L2 DRG 신경세포로 상승하는 비세그멘탈 경로 [20], [21], [22], [23], [24]. 인간에서 L5 요추 통증이 유사하게 비분절적으로 L1–L2 DRG로 전달되는지 여부는 알려져 있지 않습니다. 주목할 점은 쥐 모델이 비분절적 경로를 표적으로 삼는 것으로 추정되는 여러 치료법을 정당화하는 데 사용되었으며, 이는 종종 실망스러운 결과를 초래했습니다 [25,26].

현재까지 인간 요추 SVN의 기원에 대한 시체 연구는 상반된 결과를 보여주었습니다. 일부 연구는 모든 요추 수준에서 순수 자율 신경 SVN 기원을 발견했으며 [6], 다른 연구는 상위 요추 수준(L1–L2)에서 고전적인 이중 뿌리 SVN 기원을, 하위 요추 수준(L3–L5)에서 순수 자율 신경 기원을 보고했습니다 [13]. 최근 Zhao 등 [14]은 두 가지 기원을 모두 보고했지만, 그들의 연구는 이중 뿌리 요추 기원이 공존하는지, 또는 체성 및/또는 자율 신경 뿌리 기여도가 요추 수준에 따라 다른지 밝히지 않았습니다.

요추 SVN의 기원에 대한 정확한 지식은 디스크성 요통의 경로를 규명하고 치료하는 데 필수적입니다. 따라서 우리는 SVN의 요추 해부학을 연구했으며, 특히 상부(L2)와 하부(L5) 요추 수준에서 체성 및/또는 자율 신경 기원에 초점을 맞췄습니다. 또한 이 요추 수준에서 SVN의 포르멘탈 및 척추 내 분지 패턴과 분포에 대한 상세한 설명을 제공합니다.

Materials and methods

Anatomic dissection

Microanatomic dissections of the lumbar SVNs were performed in 5 embalmed human Caucasian cadavers (2 male, 3 female; mean age, 70.0 ± 8.5 years [range, 37–85 years]) obtained from body donation. All donors signed Donor Consent forms in accordance with institutional guidelines and the Dutch Burial and Cremation Act. None of the cadavers had externally visible lumbar abnormalities. Embalmment was within 36 hours after death by femoral artery perfusion with a 2%–4% formaldehyde fixation solution.

Dissections were performed using a Zeiss Universal S2 operating microscope (Carl Zeiss AG, Oberkochen, Germany) at 4–40x magnification. The L2 and L5 SVNs were dissected bilaterally. Photographs were taken with a Sony Cybershot DSC-H50 (Tokyo, Japan) or a Canon EOS 10D (Tokyo, Japan) camera.

Dorsal and anterolateral approaches were used. In the dorsal approach, after removal of the skin and paravertebral muscles, laminectomies, facetectomies and pediculectomies were performed bilaterally from L1–S1. The anterior L2 and L5 spinal canal and IVF was exposed by resection of the spinal cord and dura mater, and retraction of the dural root sleeve and DRG. The SVN was meticulously dissected and freed from the surrounding adipose tissue and vasculature to reveal its ramifications and to note its anatomical relationships. The ascending and descending SVN branches were isolated from the extensive anterior internal vertebral venous plexus and traced as far as possible cranially, caudally and medially. Dural branches were not traced further. Careful dissection at high magnification was performed laterally in the IVF to trace the somatic and autonomic SVN origins.

An anterolateral approach to the lateral IVF was used to further trace the SVN roots. After removal of the viscera and identification of the sympathetic trunk and rami communicantes, the psoas major muscle was resected from its vertebral origins to trace the communicating rami to their junction with the ventral spinal ramus, just lateral to the IVF. Transforaminal ligaments [27] were identified, and where possible, conserved.

SVN roots were counted and designated somatic if they macroscopically originated from the ventral ramus, spinal nerve or DRG, autonomic if they arose from the gray ramus communicans, and typed as junctional if they exited at the ramus communicans-spinal nerve junction. No macroscopic distinction between somatic and autonomic could be made in the latter. The distances between the somatic SVN root exit and the distal DRG pole were measured, and between the autonomic SVN root exit and the gray ramus communicans-spinal nerve junction (Fig. 1). The distal DRG pole was determined using its lateral convexity margin and gray coloration relative to the spinal nerve. Other measurements included thickest (foraminal) SVN filament diameter, angle between the longitudinal axis of the SVN and the transverse anatomical plane, and the SVN branching point; distance of SVN splitting into an ascending and descending branch, relative to the superjacent medial pedicle border (ie, central canal, subarticular or foraminal zone branching) (Fig. 1). Foraminal SVN distribution types and IVD (SVN) innervation patterns were noted. When multiple SVN filaments were present, the diameter of the thickest filament was measured. Filament thickness and distances were measured with a half-millimeter ruler; values for filament thickness were determined at 10x magnification and rounded to the nearest 0.1 mm. Angles were measured with a protractor. Axial lumbar zonal terminology adhered to the nomenclature by Fardon et al. [28]; intraspinal denotes both the central canal and subarticular zones.

1. Download: Download high-res image (246KB)
2. Download: Download full-size image

Fig. 1. Schematic illustration of the parameters measured. Dorsal view of anterior vertebral canal after removal of posterior arches, zygapophyseal joints and spinal cord with dura, and clockwise rotation of the spinal roots-dorsal root ganglion (DRG)-spinal nerve complex. Measurements included: (a) distance between the somatic root exit of the sinuvertebral nerve (SVN) and the distal DRG pole, (b) distance between the autonomic SVN root exit and the gray ramus communicans-spinal nerve junction, (c) related acute angle between the longitudinal axis of SVN and the transverse anatomical plane, and (d) branching point as defined by the distance of SVN splitting into an ascending and descending branch, relative to the superjacent medial pedicle border. Foraminal indicates foraminal zone (shaded dark red) and intraspinal denotes subarticular and central canal zones (shaded light red). ar, autonomic SVN root; drg, dorsal root ganglion; grc, gray ramus communicans; ivd, intervertebral disc; ped, pedicle; ped', subjacent pedicle; pll, posterior longitudinal ligament; svn, sinuvertebral nerve; sr, somatic SVN root; sla, spinal branch of the lumbar artery. (Color version of figure is available online.)

Tissue processing and histology

Samples from all dissected SVNs were embedded in paraffin, cut into 7-μm thick serial sections and mounted on poly-L-lysine-coated glass slides (VWR International, Radnor, PA). Tissue sections were dewaxed and either stained with hematoxylin and eosin (HE) or immunostained with a monoclonal mouse anti-neurofilament protein antibody (1:400; clone 2F11; M0762; Dako Denmark A/S, Glostrup, Denmark) (staining of axons) or a polyclonal rabbit anti-S100 serum (1:20,000; Z0311; Dako Denmark A/S) (staining of Schwann cells) to confirm‎ their neural tissue content.

Antibody binding was detected using the avidin-biotin complex (ABC) method [29] with a Vectastain Elite ABC Kit (PK-6100; Vector Laboratories, Inc., Burlingame, CA) and 3,3’-diaminobenzidine tetrahydrochloride (0.05% in 50 mM Tris-HCl, pH 7.0, containing 0.01% H2O2; Sigma-Aldrich, St. Louis, MO) as chromogen. Endogenous peroxidase activity was quenched with 3% H2O2 in methanol for 20 minutes. Primary antisera were diluted in phosphate-buffered saline (PBS; pH 7.4) containing 1% bovine serum albumin (BSA) and applied overnight at room temperature. Second layer antisera included biotinylated rabbit anti-mouse (1:200; E0354; Dako Denmark A/S) and biotinylated swine anti-rabbit (1:400; E0353; Dako Denmark A/S) IgG serum and were applied in PBS/1% BSA for 1 hour each. After all incubations, the sections were rinsed three times in PBS for 5 minutes. Sections were lightly counterstained with hematoxylin, dehydrated, mounted, and coverslipped.

Except for one L2 sample (left side), which appeared to be a vascular structure surrounded by connective tissue strands and was therefore excluded from our analyses, all other resection tissue (n=19) was neural tissue, containing both myelinated and nonmyelinated nerve fibers (Fig. 2).

1. Download: Download high-res image (2MB)
2. Download: Download full-size image

Fig. 2. Histological characterization of L2 and L5 sinuvertebral nerves. Tissue samples from microdissected sinuvertebral nerves (SVNs) were verified for nerve tissue using hematoxylin and eosin (HE) staining and immunostaining for neurofilament (neuronal marker) and S100 (Schwann cell marker) proteins. (A,B) Dorsal view of L2 (A) and L5 (B) SVNs in the anterior intervertebral foramen (IVF) and vertebral canal after removal of the posterior arches, zygapophyseal joints and spinal cord with dura, pedicle reduction, and retraction of the dorsal root ganglion (DRG). (A) Plexiform type SVN; ascending branch (left side). (B) Single filament type SVN; SVN proper (left side). (C–H) Tissue sections taken from the level indicated by the box in A (C–E) and B (F–H) and stained with HE (C,F), and for S100 protein (D,G) and neurofilament protein (E,H). drg, dorsal root ganglion (retracted laterally); ivf, intervertebral foramen; ped, pedicle; pll, posterior longitudinal ligament; svn, sinuvertebral nerve; sla, spinal branch of the lumbar artery. Scale bar: 2 mm (A and B); 50 µm (C and F); 25 µm (D,E,G and H).

Statistical analysis

Statistical analyses were performed with SPSS version 25 (IBM, Armonk, NY) using an independent samples t test (parametric data) or a Mann-Whitney U test (nonparametric data). A Kruskal-Wallis test (nonparametric data) with post hoc Bonferroni correction was used for multiple-group comparisons. Parametric assumptions were tested using the Shapiro-Wilk test of normality, Z-scores and Levene's test of equality of variances. Because left- and right-sided data per level (number of roots, filament diameter and branching point) did not differ significantly, pooled data were analyzed for inter- or intra-level differences. Metric data are presented as mean ± SEM values, categorical data by frequency and percentage. p Values <.05 were considered significant. Scatter plots were computed using GraphPad Prism version 8.4.2 (GraphPad Software, Inc., San Diego, CA).

ResultsL2 SVN origins

At L2, a SVN and its origins was identified in all IVFs. No L2 SVN was formed purely by somatic or autonomic roots. Six out of 9 SVNs (66.7%; 3 right sides, 3 left sides) arose from both somatic and autonomic roots (Fig. 3A, E). Two SVNs (22.2%; 1 right side, 1 left side) arose from somatic, autonomic and junctional roots (Fig. 3C), and 1 SVN (11.1%; 1 right side) was formed by somatic and junctional roots (summarized in Table 1). The SVN was composed of up to 4 roots, on average by 3.1 ± 0.3 roots (range, 2–4 roots; n=9 SVNs) (Table 2). The somatic contribution usually arose from the spinal nerve at a distance ranging from 1.0 to 10.0 mm (mean, 3.67 ± 0.89 mm; n=9) to the distal DRG. Once, a somatic root exited 3.0 mm proximal to the distal DRG (ie, from the DRG itself) (right side). The autonomic root arose from the gray ramus communicans at a distance ranging from 2.0–8.0 mm (mean, 4.75 ± 0.73 mm; n=8) from the ramus communicans-spinal nerve junction. No SVN origins from the paravertebral sympathetic ganglia or sympathetic trunk were found.

1. Download: Download high-res image (3MB)
2. Download: Download full-size image

Fig. 3. Origin and course of the L2 and L5 sinuvertebral nerve. (A,B) Ventrolateral, (C,D) dorsolateral and (E,F) dorsal views of L2 (A,C,E) and L5 (B,D,F). (A) Dual (somatic and autonomic) L2 sinuvertebral nerve (SVN) origin. Ventrolateral view of the L2 intervertebral foramen (IVF) after resection of the psoas major muscle; the gray ramus communicans is retracted. Single filament type SVN (arrow) formed by two somatic roots and one autonomic root. (B) Dual (somatic and autonomic) L5 SVN origin, found in 40% of sides. Ventrolateral view of the L5 IVF after resection of the psoas major muscle; outer somatic SVN roots pulled apart and gray ramus communicans partly retracted. Single filament type SVN (arrow) formed by three somatic roots and one autonomic root. (C) Classic dual (somatic and autonomic) L2 SVN origin, found in 88.9% of sides. Dorsolateral view of the right L2 IVF from within the spinal canal with the dorsal root ganglion (DRG) and somatic SVN root retracted with monofilament suture, and the spinal branch of the lumbar artery reflected onto the DRG. Immediate splitting type SVN (arrows). After root convergence into SVN proper, immediate splitting into an ascending and descending branch. Note additional junctional root emerging from the ramus communicans-spinal nerve junction. (D) Pure autonomic L5 SVN origin, found in 60% of sides. Dorsolateral view of the left L5 IVF from within the spinal canal with the DRG retracted. The autonomic root continues as a single filament type SVN (arrow). (E) Foraminal and intraspinal distribution of the L2 SVN. Dorsal view of the L2–L3 anterior epidural space. Immediate splitting (left side) and plexiform (right side) type SVNs (arrows). The SVN courses in the superior or middle anterior portion of the IVF, ramifies around the spinal branch of the lumbar artery and forms a plexus lateral to the posterior longitudinal ligament (PLL), innervating both the PLL and sub-/superjacent intervertebral discs (IVDs). (F) Foraminal and intraspinal distribution of the L5 SVN. Dorsal view of the L5–S1 anterior epidural space. Single filament type SVNs (left and right side; arrows). Bilateral intraspinal branching points, located deep to the anterior internal vertebral venous plexus. Asterisk in A, C and D indicates IVD nerve branches from ventral ramus of spinal nerve, SVN, and gray ramus communicans, respectively. ar, autonomic SVN root; dr, dorsal ramus of spinal nerve; drg, dorsal root ganglion (retracted laterally); grc, gray ramus communicans; iv, intervertebral vein; ivd, intervertebral disc; jr, junctional SVN root; lp, lumbar plexus; ped, pedicle; ped', subjacent pedicle; pll, posterior longitudinal ligament; sr, somatic SVN root; sla, spinal branch of the lumbar artery; st, sympathetic trunk; vr, ventral ramus of spinal nerve; vvp, anterior internal vertebral venous plexus; wrc, white ramus communicans. Double sideways chevrons indicate cranial direction. Scale bar=5 mm.

Table 1. Sinuvertebral nerve root origins (%) at L2 and L5

Root origin(s)L2L5(n=9)(n=10)

Empty Cell

⁎

Roots were typed as junctional if they exited at the gray ramus communicans-spinal nerve junction. np, not present.

Table 2. Sinuvertebral nerve characteristics at L2 and L5

ParameterL2L5Mean ± SEM(Range) nMean ± SEM(Range) np Value*

Empty Cell

Empty Cell

NA, not applicable; np, not present.

⁎

p Values <.05 were considered significant; independent samples t test (branching point) or Mann-Whitney U test (number of roots and root types, diameter, ascending filament angle).

†

The number of somatic vs autonomic roots at L2 or L5 did not differ (p=1.000 and p=.227, respectively; Kruskal-Wallis test with post hoc Bonferroni correction).

‡

Roots were typed as junctional if they exited at the gray ramus communicans-spinal nerve junction.

§

Relative to superjacent medial pedicle border.

L2 foraminal SVN morphology

Four SVN types were found in the IVF after its roots conjoined (summarized in Fig. 4): single filament on 2 of 9 sides (22.2%; 1 right side, 1 left side) (Fig. 4A; cf. Fig. 3A), multiple filament (up to 3; parallel or diverging) on 3 of 9 sides (33.3%; 2 right sides, 1 left side) (Fig. 4B), immediate splitting after union of roots on 2 of 9 sides (22.2%; 1 right side, 1 left side) (Fig. 4C; cf. Fig. 3C, E, left side), and plexiform configurations on 2 of 9 sides (22.2%; 1 right side, 1 left side) (Fig. 4D; cf. Fig. 3E, right side).

1. Download: Download high-res image (593KB)
2. Download: Download full-size image

Fig. 4. Foraminal sinuvertebral nerve types and frequencies (%) at L2 and L5. Dorsal view of anterior vertebral canal after removal of posterior arches, zygapophyseal joints and spinal cord with dura, and removal of the spinal roots-dorsal root ganglion-spinal nerve complex. Four distinct foraminal sinuvertebral nerve (SVN) types are recognized: (A) single filament, (B) multiple filament (parallel or diverging), (C) immediate splitting, and (D) plexiform types. At L2, all four SVN types (A–D) are found, whereas at L5 only single (A) or multiple filament (B) type SVNs are present. Note that the descending SVN branch (closed arrowhead) innervates the corresponding (L2/3 or L5/S1) intervertebral disc (IVD), whereas the ascending branch (open arrowhead) innervates the superjacent IVD. ivd, intervertebral disc; ped, pedicle; ped', subjacent pedicle; pll, posterior longitudinal ligament; svn, sinuvertebral nerve.

Most commonly (6 of 9 sides), the SVN was located in the superior portion of the IVF (66.7%; 4 right sides, 2 left sides) and ran predominantly superior to the spinal branch of the lumbar artery. All SVNs and their branches were located ventral to the DRG; none were found in the dorsal IVF. The mean diameter of the thickest SVN filament was 0.48 ± 0.06 mm (range, 0.2–0.8 mm; n=9) (Table 2; Fig. 5A). Ascending foraminal SVN filaments (5 of 9 sides, 55.6%; 3 right sides, 2 left sides) coursed roughly parallel but recurrent to the exiting spinal nerve root trajectory at a mean angle of 36.0°±4.8° (range, 20°–45°; n=5) relative to the transverse anatomical plane (Table 2). Descending L2 SVN filaments (3 of 9 sides, 33.3%; 3 right sides) coursed at a mean angle of −16.7°±1.7° (range, −20° to −15°; n=3) (Table 2). Ascending and descending filaments included single filament, multiple filament (parallel and diverging) and immediate splitting SVN types.

1. Download: Download high-res image (121KB)
2. Download: Download full-size image

Fig. 5. Scatter plots of (A) sinuvertebral nerve filament thickness (diameter of the thickest filament) and (B) foraminal or intraspinal branching (distance of sinuvertebral nerve splitting into an ascending and descending branch, relative to the superjacent medial pedicle border). Negative values in B indicate foraminal sinuvertebral (SVN) branching; ie, splitting lateral to the medial superjacent pedicle border. Open squares, left-sided L2 SVN; closed squares, right-sided L2 SVN; open triangles, left-sided L5 SVN; closed triangles, right-sided L5 SVN. *p Value <.05; ns, not significant.

L2 SVN branches to the posterolateral L2/3 IVD were found on 1 of 9 sides (11.1%; right side) (Fig. 3C). Three other sources innervated the posterolateral L2/3 IVD: the gray ramus communicans on 3 of 9 sides (33.3%; 1 right side, 2 left sides), the ventral spinal ramus on 3 of 9 sides (33.3%; 1 right side, 2 left sides) (Fig. 3A) and the spinal nerve on 1 of 9 sides (11.1%; left side). The posterolateral L2 vertebral body and its periosteum received L2 SVN branches on 2 of 9 sides (22.2%; 2 left sides) and from two other sources: the gray ramus communicans on 3 of 9 sides (33.3%; 1 right side, 2 left sides) and the spinal nerve on 1 of 9 sides (11.1%; left side). Of these, one side (11.1%; left side) contained branches from all three sources.

L2 SVN spinal canal distribution

Branching of the L2 SVN into ascending, descending and transverse branches occurred in the foraminal zone on 3 of 7 sides (42.9%; 1 right side, 2 left sides) and intraspinal on 4 of 7 sides (57.1%; 3 right sides, 1 left side) (Fig. 5B). On average, branching occurred 0.57 ± 1.63 mm (range, −6.0 to 5.0 mm; n=7 SVNs) intraspinal to the superjacent medial pedicle border (Table 2; Fig. 5B). These SVN branches formed a plexus lateral to the PLL, ran perivascular to ramifications of the spinal branch of the lumbar artery and either intertwined with or coursed mostly ventral to the anterior internal vertebral venous plexus (Fig. 3E). Ascending, transverse, and descending SVN branches disappeared under the PLL (Fig. 3E), and occasionally entered the basivertebral foramen or into separate small paramedian foramina, together with the basivertebral vein and branches of the spinal branch of the lumbar artery. Small non SVN-derived nerve fibers (cf. [30]) were not found in the IVF or spinal canal.

When present, the ascending SVN branch was generally directed superiorly (8 of 9 sides; 88.9%, 5 right sides, 3 left sides), running to the posterior aspect of the L1/2 IVD on 3 of 9 sides (33.3%; 1 right side, 2 left sides). It could not be traced beyond this level. The posterior surface of the L2/3 IVD received ramifications from the descending SVN branch on 5 of 9 sides (55.6%; 3 right sides, 2 left sides) (Fig. 3E, C). Sometimes a descending branch coursed between the lateral PLL expansion and L2/3 IVD (Fig. 3E, right side, C), but this could not be traced further caudally. Once, an intersegmental connection between the L2 descending and L3 ascending branch of the SVN was observed. In one specimen, the L2 ascending branch ran at both sides to the L2 basivertebral foramen and the descending L2 branch to the L3 basivertebral foramen. Ascending and descending SVN branches appeared roughly equal in thickness (cf. [12]), although in one specimen the ascending branch seemed markedly thicker.

The cranio- and caudolateral PLL extensions overlaying the IVD received inputs from the parent and superjacent SVN and the parent and subjacent SVN, respectively. The PLL overlaying the vertebral body was innervated by the parent SVN.

L5 SVN origins

At L5, a SVN and its origins was identified in all IVFs. It had purely autonomic origins on 6 of 10 sides (60%; 2 right sides, 4 left sides), usually arising from 1 root (Fig. 3D). On 3 of 10 sides (30%; 2 right sides, 1 left side), the SVN arose from both somatic and autonomic roots (Fig. 3B); these sides contained more roots than pure autonomic origins: up to 4 (2, 3 and 4, respectively). One side (10%; right side) contained somatic and junctional roots. No SVN had a purely somatic origin at L5. The SVN origins at L5 are summarized in Table 1. On average, the L5 SVN was composed of 1.9 ± 0.3 roots (range, 1–4 roots; n=10 SVNs) (Table 2). The somatic contribution usually arose from the spinal nerve at a distance ranging from 2.0 to 5.0 mm (mean, 2.71 ± 0.42 mm; n=7) to the distal DRG. The autonomic root arose from the gray ramus communicans at a distance ranging from 1.0–6.0 mm (mean, 3.41 ± 0.41 mm; n=11) from the ramus communicans-spinal nerve junction. No SVN origins from the paravertebral sympathetic ganglia or sympathetic trunk were found.

L5 foraminal SVN morphology

A single filament SVN type was found on 9 of 10 sides (90%; 4 right sides, 5 left sides) (Figs. 3B, D, F and 4A), one of which split and reunited while still in the IVF. Except for a multiple filament type SVN (Fig. 4B) on 1 of 10 sides (10%; right side), no other SVN types were found. The SVN ran ventral to the DRG, was usually located superior to the spinal branch of the lumbar artery, and ran in the middle IVF portion on 7 of 10 sides (70%; 4 right sides, 3 left sides). No SVN or its branches was found in the dorsal IVF. The mean diameter of the thickest SVN filament was 0.33 ± 0.02 mm (range, 0.2–0.4 mm; n=10) (Table 2; Fig. 5A). All L5 SVN filaments followed an ascending foraminal course (roughly parallel but recurrent to the exiting spinal nerve root trajectory) at a mean angle of 22.0°±3.5° (range, 10°–35°; n=10) relative to the transverse anatomical plane (Table 2).

L5 SVN spinal canal distribution

Branching of the L5 SVN into ascending, descending and transverse branches occurred intraspinal on 6 of 10 sides (60%; 4 right sides, 2 left sides) and at the foraminal zone-intraspinal border on 2 of 10 sides (20%; 1 right side, 1 left side). No branching was observed on 2 of 10 sides (20%; 2 left sides). On average, branching occurred 4.63 ± 1.22 mm (range, 0.0–10.0 mm; n=8) intraspinal to the superjacent medial pedicle border (Table 2; Fig. 5B).

No differences in intraspinal SVN distribution were found compared with L2. Transverse branches were traced over the midline multiple times. The ascending branch of the SVN was traced to the posterior L4/5 IVD on 4 of 10 sides (40%; 3 right sides, 1 left side); they could not be traced further than this level with confidence. Ramifications from the descending SVN branch over the L5/S1 IVD were seen on 6 of 10 sides (60%; 4 right sides, 2 left sides) (Fig. 3F). Intersegmental SVN communications were not found. Only once were posterolateral L5/S1 IVD nerve branches seen, originating from the adjacent gray ramus communicans (Fig. 3D). (Peri)osteal L5 nerve branches were found posterolaterally and in the lateral recess, originating from the L5 SVN on 5 of 10 sides (50%; 3 right sides, 2 left sides) or from the DRG on 1 of 10 sides (10%; left side). Osteal branches entered the L5 vertebra via small foramina, sometimes alongside a small arterial branch.

Discussion

In the present study, we examined the anatomy of the SVN at high (L2) and low (L5) lumbar levels, with special focus on its somatic and autonomic origins. This knowledge is important for unraveling the ‘wiring diagram’ of discogenic LBP, and hence, to improve targeting of treatment modalities. Furthermore, it may contribute to preventing iatrogenic damage to the SVN during (transforaminal) surgical approaches to the lumbar spine. Our data show that a dual root origin is common at L2 and that at L5 there is autonomic root predominance but not exclusivity (Table 1). The finding that about half of L5 SVNs are formed by a dual root origin suggests that lower lumbar discogenic pain may not only be transmitted by nonsegmental neural pathways as is widely accepted in clinical practice [31], but also segmentally.

Origin of the SVN

In our study, nearly 90% of SVNs at L2 arose from both somatic and autonomic roots, confirm‎ing findings of most authors [3,8,[11], [12], [13]]. At L5, our results confirm‎ an autonomic root predominance, but not exclusivity, as 40% of L5 SVNs contained an additional somatic component. Macroscopically, the SVN at L5 is thus not always purely autonomic as previously reported [6,13].

These discrepancies in previously reported SVN origins might be explained by differences in methodology or dissection techniques to study the SVN; notably lower lumbar SVN roots are difficult to visualize. In human fetuses, Groen et al. [6] using an acetylcholinesterase whole-mount method found that except for the cervical region, all SVNs are exclusively derived from gray rami communicantes. Expression‎ of acetylcholinesterase in somatic SVN roots, however, might well be a later event during fetal development than in autonomic roots. Furthermore, while we have found a similar plexiform arrangement in the ramus communicans’ termination area as Groen et al. [6], and also what appeared macroscopically as “crossing” fiber bundles to the dorsal ramus, the latter did not contribute to the SVN (cf. [32]). Misidentification by others of these “crossing” fibers as originating from the spinal nerve or ventral ramus has, however, been put forward by Groen et al. [6] as a possible explanation for their finding of an exclusive autonomic SVN origin. Groen et al. [6] also contend that the posterolateral retraction of the DRG-dorsal-ventral rami complex, exposing the anterior IVF, might give the impression that the SVN roots originate from the spinal nerve, but that these are actually autonomic in origin. While individual root identification is difficult, we have clearly demonstrated somatic contributions to the SVN from the ventral ramus, spinal nerve and DRG both at L2 and L5 (Fig. 3).

In agreement with previous studies [13], the SVN arose from significantly more roots at L2 than L5 (Table 2). The number of somatic vs autonomic roots at L2 did not differ (p=1.000, post hoc Bonferroni correction), nor did it at L5 (post hoc p=.227) (Table 2).

Foraminal SVN morphology

Four different foraminal SVN types could be discerned: single filament, multiple filament, immediate splitting and plexiform types (Fig. 4). Configurations with numerous SVN filaments (multiple filament, immediate splitting and plexiform types) are common at L2 (77.8% of sides), whereas at L5, 90% of sides contained a single SVN filament. Lazorthes et al. [10] described two SVN types: single nerve and multiple independent or interconnecting nerves (corresponding to our multiple filament and plexiform types) without reporting regional variation.

Our results indicate that the SVN is significantly thicker at L2 than L5 (Table 2; Fig. 5A). Since only the thickest filament was measured for all SVN types, we performed a subgroup analysis for single filament SVN types, which confirmed this finding. The difference both in the number of SVN roots and the SVN diameter at L2 vs L5, suggests that level-related differences might exist in SVN nerve fiber count and/or composition (small nociceptive C fibers vs larger proprioceptive Aα fibers). It might also indicate that the L5/S1 IVD is innervated more sparsely than higher lumbar IVDs.

The course of the SVN in the IVF matches existing literature: ventral to the DRG and roughly parallel but recurrent to the exiting spinal nerve root trajectory. SVN filaments generally run superior to the spinal branch of the lumbar artery at L2 and L5, occasionally forming a perivascular nerve plexus. This plexus is different from the previously reported nerve plexus around the radicular branch of the lumbar artery, that arises from the gray ramus communicans at the point of SVN origin [6].

Posterolateral IVD branches from the adjacent SVN were uncommon (1 of 9 sides at L2, none at L5). Posterolateral IVD nerve branches derived predominantly from the adjacent ramus communicans and spinal ventral ramus as previously described by Bogduk et al. [12] and Zhao et al. [14]. Zhao et al. [14] classified these branches as SVNs (type I, “SVN deputy branches”), however, the present study did not, in accordance with Bogduk et al. [12]. More numerous posterolateral IVD nerve branches were found at L2/3 than L5/S1; this too could signify a sparser lower lumbar innervation but might also represent dissection artefacts, as this disc region is difficult to dissect with neural structures in situ. The lateral IVD received branches from the adjacent ramus communicans. SVN branches to the anterior longitudinal ligament (cf. [33]) were not found.

SVN distribution in the spinal canal

The present findings that the intraspinal SVN distribution does not differ at L2 and L5 and that the SVN does not extend beyond its level of origin are in accordance with Luschka's original description [8], Lazorthes et al. [10], and Bogduk et al. [12]. Although branching of the SVN into ascending, descending and transverse branches has been described [6,12], its precise branching point has not been further detailed. Branching occurred intraspinal in the subarticular or central canal zone, and generally more medially at L5 than L2 (Table 2; Fig. 5B), irrespective of SVN type.

The posterior L2/3 and L5/S1 IVDs were innervated ipsilaterally by the descending branch of the parent SVN and ascending branch of the subjacent SVN, that is by two spinal levels. This is in accordance with Lazorthes et al. [10] and Bogduk et al. [12], but in contrast to Groen et al. [6], who found that lumbar posterior IVDs were innervated by the parent, sub- and superjacent SVNs, that is by three levels. Such an arrangement could only be valid if a long descending SVN branch extended to the IVD one level lower, but these were not found in the present study. In one specimen, a L2 descending SVN branch ran to the L3 basivertebral foramen, but this did not extend further caudally, although it could theoretically give off thin branches to the L3/4 IVD. Descending SVN branches passing between the lateral PLL extension and IVD were not traced further, but descending SVN branches from superjacent levels were not found at L2 and L5, neither free lying nor appearing from between the superjacent lateral extension of the PLL and its associated IVD.

Clinical implications

Our results indicate that 60% of L5 SVNs are formed purely by autonomic roots. In the absence of a somatic root contribution, discogenic pain transmission can occur via the subjacent SVN, as an IVD is innervated by two spinal levels (this study). Alternatively, it might be mediated nonsegmentally through the previously described autonomic inflow diversion [20], [21], [22], [23], [24], whereby nociceptive fibers from lower lumbar IVDs ascend via the autonomic SVN root through gray rami communicantes and the sympathetic trunk to L1–L2 DRG neurons. So far, however, this neural pathway has only been described in rats and awaits further human study.

Remarkably, current therapies for discogenic LBP specifically aim to target this nonsegmental pathway by interrupting pain transmission through nerve blocks, transection, or radiofrequency lesioning of the L2 DRG, spinal nerve or ramus communicans [25,26,31,[34], [35], [36], [37]]. These interventions are not always effective [25,26]. Since we have found that 40% of SVNs at L5 are formed also by somatic roots, which are presumed to contain segmental nociceptive fibers, interruption of the nonsegmental pathway alone is expected to provide only partial pain reduction and may explain some of the disappointing results in the aforementioned studies. Interindividual “wiring diagram” variations may thus call for different therapeutic approaches to pain reduction in patients with discogenic LBP.

Although the precise role of the SVN in discogenic LBP has not been fully established yet, data from immunohistochemical studies have clearly revealed that healthy and painful lumbar posterior IVDs are innervated by nociceptive fibers from the SVN, indicating a function in pain perception (reviewed in [15,24,38]). Whether these nociceptive fibers indeed mediate painful stimuli both segmentally and nonsegmentally awaits further (functional) studies. Although performing experiments in humans limits research, with the emergence of novel dyes, such as NeuroVue dyes, histological methods for ex vivo axon tracing have also become available for human postmortem tissue (reviewed in [39]).

The spectrum of transforaminal ligaments found in our study matches previous observations [27]. Amonoo-Kuofi et al. [40] have reported that the SVN is constantly transmitted into the IVF through the compartment created by the deep anterior intraforaminal ligament. In the present study, the SVN proper was not transmitted through a compartment created by trans- or intraforaminal ligaments, although the compartment formed by a superior transforaminal ligament did transmit an autonomic SVN root once at L5. The inconstancy of these ligaments indicates they cannot be used reliably as landmarks for the location of the SVN at the foraminal-extraforaminal border.

The anatomy described in this study also has implications for Kambin's triangular safe zone (described in [41]) and Harms’ transforaminal corridor [42], used in (minimally invasive and endoscopic) transforaminal spinal approaches (reviewed in [43]). As the SVN generally courses through the superior part of these triangular safe zones, it effectively truncates them into quadrangular safe zones, if a nerve-sparing approach is sought. Unfamiliarity with these neural relations may lead to iatrogenic SVN damage.

Our results are based on a relatively limited sample size (n=5 cadavers), as is not uncommon in this type of labor-intensive anatomical research [3,6,9,[12], [13], [14]]. Nevertheless, our findings are relevant to spinal surgeons who strive to keep the SVN intact during minimally invasive (foraminal) approaches. Future studies will further determine the full extent of foraminal SVN subtypes and distributions.

Conclusions

At L2, the SVN arises in nearly 90% of sides from both somatic and autonomic roots and at L5 in 40% of sides. The remaining SVNs are formed by purely autonomic roots. Lower lumbar discogenic pain is presumably mediated segmentally via the somatic SVN root and nonsegmentally through the autonomic SVN root; targeting only the nonsegmental pathway may provide incomplete pain reduction. In the IVF, the L2 SVN generally consists of numerous filaments, whereas at L5 90% contains a single SVN filament. Relating SVN anatomy to microsurgical spinal approaches may prevent iatrogenic damage to the SVN and the formation of postsurgical back pain.

Declarations of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We thank Ingrid Hegeman-Kleinn for technical assistance with the histology.

This research was supported with funding from the PainLess foundation and the Schumacher-Kramer Stichting.

References

1. [1]

   FIPAT

   Terminologia anatomica

   Federative international programme for anatomical terminology (2nd ed) (2019)

   Available at:

   https://fipat.library.dal.ca/ta2

   Accessed May 31, 2021

   Google Scholar

2. [2]

   A. Hovelacque

   Le nerf sinu-vertébral

   Ann Anat Pathol Anat Norm Med Chir, 2 (1925), pp. 435-443

   Google Scholar

3. [3]

   MA Edgar, S. Nundy

   Innervation of the spinal dura mater

   J Neurol Neurosurg Psychiatry, 29 (6) (1966), pp. 530-534, 10.1136/jnnp.29.6.530

   Google Scholar

4. [4]

   N Bogduk, M Windsor, A. Inglis

   The innervation of the cervical intervertebral discs

   Spine (Phila Pa 1976), 13 (1) (1988), pp. 2-8, 10.1097/00007632-198801000-00002

   View in ScopusGoogle Scholar

5. [5]

   XQ Chen, S Bo, SZ. Zhong

   Nerves accompanying the vertebral artery and their clinical relevance

   Spine (Phila Pa 1976), 13 (12) (1988), pp. 1360-1364, 10.1097/00007632-198812000-00006

   View in ScopusGoogle Scholar

6. [6]

   GJ Groen, B Baljet, J. Drukker

   Nerves and nerve plexuses of the human vertebral column

   Am J Anat, 188 (3) (1990), pp. 282-296, 10.1002/aja.1001880307

   View in ScopusGoogle Scholar

7. [7]

   C Rennie, MR Haffajee, MA. Ebrahim

   The sinuvertebral nerves at the craniovertebral junction: a microdissection study

   Clin Anat, 26 (3) (2013), pp. 357-366, 10.1002/ca.22105

   View in ScopusGoogle Scholar