Background The acupuncture point BL23 is located in the region of the posterior ramus of the second lumbar spinal nerve (L2) and has historically been used to treat conditions such as lower back pain, pollakiuria, erectile dysfunction, dysmenorrhoea, tinnitus, and vertigo. Some of these treatment effects have been hypothesised to be mediated by the sympathetic nervous system. It was recently discovered that the posterior ramus of the spinal nerve (PRSN) at L2 forms not two but three branches.
Objective To examine the relationship between the acupuncture point BL23 and the L2 PRSN in order to consider the pathways possibly affected by BL23 acupuncture.
Methods Acupuncture needles were inserted through the skin at BL23 to a depth of 3 cm a total of 13 times in eight donor cadavers (seven right-sided, six left-sided). Leaving the needle in place, ventral dissection was performed to determine the PRSN anatomy between the L1 and L3 spinal segments. In four cadavers, the relationship between the L2 spinal nerve and sympathetic branches was additionally evaluated. Following dissection, three-dimensional (3D) data were acquired using a photo scanner and 3D structural images were created using 3D computer graphics software. One additional (female) cadaver was studied without insertion of an acupuncture needle (due to significant scoliosis).
Results The L2 PRSN was divided into medial, intermediate and lateral branches. The needle inserted at BL23 came to lie in the region of the intermediate or lateral branches in all cases. Rami communicantes were found between the L2 spinal nerve and sympathetic trunk with fibres going on to supply the superior hypogastric plexus.
Conclusions Our findings suggest that acupuncture needles inserted at BL23 come into close proximity with the intermediate or lateral branch of the L2 PRSN, which could result in stimulation of both the somatic and sympathetic nervous systems.
Statistics from Altmetric.com
It has been shown that stimulation of certain acupuncture points can effect biological changes through the nervous and endocrine systems. Although locations of acupuncture points relative to anatomical structures have been previously reported,1–5 the structures adjacent to every acupuncture point require further elucidation to avoid medical malpractice, for example, pneumothorax, vascular or nerve injury. BL23, called Shenshu in Chinese and Jinyu in Japanese, has been used for the treatment of lower back pain, pollakiuria, erectile dysfunction, dysmenorrhoea, tinnitus, and vertigo. Previous reports have confirmed that acupuncture at BL23 affects kidney function, bone metabolism, cerebral function, and immune function,6–12 and is at least partly mediated by the sympathetic nervous system.6–8 ,13
BL23 is located in the lumbar region at the level of the inferior border of the second lumbar vertebral (L2) spinous process and 1.5 B cun, defined as half the width of four fingers at the dorsal crease of the proximal interphalangeal joint of the middle finger, lateral to the posterior median line.14 Therefore, BL23 is expected to overlie the region of the posterior ramus of the L2 spinal nerve (SN).
The root of the posterior ramus of the SN (PRSN) was traditionally believed to divide into two branches: medial and lateral.15 However, in 1982 Bogduk16 reported that PRSN formed not two but three branches: medial, lateral, and intermediate. We have also demonstrated that PRSN in the thoracic and lumbar segments forms three branches.17–19 In addition, it has been reported that four rami communicantes can be identified at level L2, but only one ramus communicans exists at level L3, L4, and L5 in two out of six studied cadavers,20 and that lumbar sympathetic ganglia are observed most frequently at the level of L2–L3.21 In this study, we aim to examine the relationship between BL23 and the branches of PRSN and to investigate the neural pathways that could be affected by acupuncture at BL23.
A total of nine cadavers (six male, three female) donated to Aichi Medical University between 2012 and 2014 were included. Before death, the donors signed documents agreeing to body donation and their use in clinical studies. The format of the document is within the expectations of the Japanese law ‘Act on Body Donation for Medical and Dental Education’. The average age of the donors was 84.6±5.2 years.
All nine donors were fixed in 10% formaldehyde solution. In eight of the nine donors, a 20 G needle was inserted perpendicularly at BL23 to a depth of 3 cm.22 A total of 13 needle insertions were studied, seven of which were right-sided and six left-sided. The selection of laterality was in accordance with the research schedule of Aichi Medical University. The ninth (female) cadaver was excluded in view of the presence of a significant degree of scoliosis. The point of needle insertion on the skin was 1.5 cun lateral to the lower border of the tip of the spinous process of L2, which was identified using Jacoby's line. Subsequently the thoracolumbar vertebral column and surrounding muscle were dissected from each cadaver, as described below. Detailed images of the PRSN and its relation to the needle tip were obtained using a ventral approach as described in our previous study.18 The relationships between the L2 SN and the sympathetic chain were additionally examined in four of the nine cadavers.
The abdominal contents were first removed. Next the vertebral bodies were separated from their pedicles and lifted from the spinal canal to expose the neural structures of the spinal canal and their relation to the pedicles, intervertebral foramen, and posterior structures of the spine, thereby allowing detailed inspection of the dural sac (with the meninges covering the ventral side of the spinal cord), the SNs and their branches. Half of each vertebral body was removed to visualise the main root of the SN on the side ipsilateral to the needle. Using this approach, each SN and the main ramification of its posterior ramus could be easily identified. The ventral approach to the SN revealed the origin and major branching of the PRSN without injuring it. We followed the SN through the intervertebral foramen to facilitate easy identification. Following examination, the vertebral arch pedicles were removed along with their corresponding transverse processes. Nerves, including the ventral rami, were then photographed and the ramifications of the posterior rami were tracked to the erector spinae divisions and skin. Each PRSN was ultimately dissected from its origin with each branch followed as far as possible.
Three-dimensional analysis of PRSN
Three-dimensional (3D) data were acquired using a photo scanner (Artec Eva/MHT, Nagoya, Japan) and Artec Studio 9 software (Data Design Group, Nagoya, Japan) in cadavers 1–4. Data were further processed using the 3D computer graphic software LEIOS (Data Design Group, Nagoya, Japan) for trimming and Light Wave 3D (New Tek Inc, San Antonio, Texas, USA) for compilation.
The photo scanner provides a 3D image of the structure of the exposed surface of the PRSN. After acquiring the surface structure, the PRSN were isolated from the whole view using the LEIOS software by erasing polygons related to structures distinct from the PRSN, representing the surrounding tissue. Several measurements were required to record 3D images of PRSN structure for each segment. Next, lumbar vertebrae were further dissected and scanned. The origin of the investigated area, tentatively defined as the upper surface of the vertebral body of L4, was used to calibrate the zero point of the photo scanner. The x-axis, y-axis, and z-axis were assigned as transverse, vertical, and sagittal, respectively. Data for each vertebral segment were compiled into a single layer of the whole region.
To interpret segmental data from the photo scanner, photographs were also taken and sketches made representing the location of the SN and PRSN within the vertebral structure of each segment. The remaining cadaver was used to study the anatomy of the PRSN without insertion of an acupuncture needle.
The structure of the PRSN was visualised in 3D on a computer using graphic design software as previously described.18 Fragments from the photo scanner data (in LEIOS software) were transferred to Light Wave 3D software. Data fragments from each cadaver representing PRSN dissections and lumbar vertebrae were digitally interconnected to recreate the whole anatomy of the PRSN in the lumbar segments while referring to the images and sketches made using Light Wave 3D software. The 3D data of the PRSN from four cadavers were averaged to merge the same points on isolated structures. The shapes of the exposed ventral surface structures were captured using a laser scanner. Notably, however, the surface of the most distant side of the structure could not be captured as the shape of the soft nerve was deformed during complete dissection. Therefore, the structure of the further (distant) side of the nerve was drawn by hand using Light Wave 3D software with reference to photographs and sketches.
Using a ventral approach, it was noted that there were rami communicantes between the L2 SN and the sympathetic trunk in all cadavers, which were continuous with the superior hypogastric plexus (figure 1). Distal to the SN trunk, each SN divided into anterior and posterior rami (figure 2A). We identified three major branches originating from the PRSN: a medial, a lateral, and an intermediate branch (figure 2B). In five cadavers, we observed that the intermediate and lateral branches ran parallel with one another for a total length of 4.7±1.7 mm (figures 2B and 3A). The lateral branch entered the iliocostalis muscle, while the intermediate branch entered the longissimus (figure 3).
Acupuncture at BL23
With respect to the 13 needle insertions (seven right-sided, six left-sided) at BL23 in eight cadavers, in all cases the tip of the acupuncture needle reached the area supplied by the intermediate or lateral branch of the PRSN at L2 (figure 3A, B). In 11 of 13 needle insertions (five right-sided, six left-sided), the needle tip was found in the vicinity of the intermediate branch. In the remaining two insertions (both right-sided), the needle tip lay in the vicinity of the lateral branch. In all cases, the needle tip was consistently found to lie between a point 2 mm medial to the intermediate branch and the lateral branch of the L2 PRSN per se. As shown in figure 3C, D, the intermediate branch of the L2 PRSN ran close to the lateral branch, which subdivided into two branches (labelled A and B). Figure 4 shows a 3D image created from a subset of four cadavers. In two cadavers, the needle tip reached the periphery of the region supplied by the intermediate branch. In the other two, the needle tip reached the periphery of the region supplied by the lateral branch or directly came into contact with the lateral branch.
In this study, we confirmed that the L2 PRSN is divided into medial, intermediate, and lateral branches and demonstrated that acupuncture needles inserted at BL23 came to lie posterior and inferior to the branching point, within the regions supplied by the intermediate and lateral branches of the L2 PRSN. To our knowledge, this is the first description of the anatomical relationship between BL23 and the posterior ramus of the L2 SN.
Traditional acupuncture must be rationalised in order to be integrated with modern biomedicine.1 Although the risk of nerve injury is negligible due to the thinness of the acupuncture needle, in some cases lower back pain has been reported to deteriorate following acupuncture treatment.23 ,24 Consequently, it is important to understand the anatomical structures underlying traditional acupuncture points. We observed that BL23 has a close anatomical relationship with the L2 SN. In the present study, the needle tip was always found to lie in the region supplied by the intermediate or lateral branches of the L2 PRSN in the area behind the transverse process of the L3 vertebra. Consequently, given that the needle tip following acupuncture at BL23 is in close proximity with the neurilemma of the intermediate or lateral branch of the L2 PRSN, an effect of acupuncture treatment may be anticipated.
Our study has some limitations. Unlike in live humans, we were unable to rely on the objective or subjective response of the subjects to needling (de qi sensation) in the case of cadavers, which would otherwise have helped guide needle placement. During formaldehyde-fixation, the soft tissues of the cadavers become hardened, therefore needle insertion is not as easy as in clinical practice. Furthermore, the tissue shrinkage that is inevitable during fixation and dehydration, plus the relatively advanced age of the subjects, may have produced anatomical deformities that could limit the ability to extrapolate our findings to living and/or younger people, particularly with respect to the depth of needling. Possible differences in positioning of the needle tip should be addressed in future research with live human patients. Nevertheless, in our experience, cadaveric study is applicable to clinical use.25 ,26
In the present study, the intermediate branch of the L2 PRSN was always close to the lateral branch. Bogduk reported that the length of the common stems between the lateral and intermediate branches of L2 was 3.8±2.0 mm.16 Herein we observed that the intermediate and lateral branches are contiguous.
Our previous report demonstrated that the L2 SN has the largest diameter of all the SNs between T3 and L5.17 Higuchi reported that the upper lumbar segments participate in the innervation of a more extensive area spreading across the lower lumbar region and suggested that L1–L2 is the centre of non-segmental lumbar spinal innervation.27 We speculate that acupuncture stimulation at BL23 may modulate somatic nociceptive fibres passing through the posterior ramus of the L2 SN. Furthermore, through the spinal reflex arc, it may also influence efferent pathways such as the sympathetic fibres passing through the L2 SN and the communicating branches between the SN and the sympathetic trunk.20 ,28 In the present study, we identified a large communicating branch that went on to supply the hypogastric plexus. Although this was a purely anatomical as opposed to a functional experiment, we propose that acupuncture at BL23 can stimulate the sensory afferents of the intermediate and lateral branches of the L2 PRSN resulting in stimulation of both the sympathetic and somatic nervous systems (figure 5). Moreover, we suggest that BL23 may be a particularly good location in the dorsal region at which to needle in order to influence the sympathetic nervous system.
We determined that acupuncture needles inserted at BL23 repeatedly came to lie in the region supplied by the intermediate and lateral branches of the L2 PRSN during 13 needle insertions in a total of eight cadavers. We identified a ramus communicans between the L2 SN and the sympathetic trunk with sympathetic fibres going on to supply the superior hypogastric plexus. Therefore, we propose that acupuncture at BL23 can stimulate the sensory afferents of the intermediate and lateral branches of L2 PRSN, resulting in stimulation of the somatic nervous system and potentially also the sympathetic nervous system, including the hypogastric plexus.
Contributors TS, HS and TN conceived the study. KU and HS designed the study. KU, TS and TY acquired the data. KU, TS, MN, SH and HS analysed the data. KU, TS, MN and HS interpreted the data. KU, TS, MN, SH, HS and TN wrote the first draft of the paper. All contributors revised the paper, approved the final version, and agreed to be accountable for all aspects of the work.
Funding This study was approved and supported by Japan Science Promotion Society (JSPS) KAKENHI Grant-in-Aid for Scientific Research © Number 25460259.
Competing interests None declared.
Patient consent Obtained.
Ethics approval The ethics committee of Aichi Medical University.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement We all attended the cadaver study, and attended the data acquisition. After we produced the manuscript, we all shared and read it, and agreed that the manuscript will be published.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.