- Email: [email protected]
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture
The Journal of Pain, Vol 10, No 8 (August), 2009: pp 870-875 Available online at www.sciencedirect.com
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture Seiji Ohtori, Masaomi Yamashita, Gen Inoue, Kazuyo Yamauchi, Munetaka Suzuki, Sumihisa Orita, Yawara Eguchi, Nobuyasu Ochiai, Shunji Kishida, Masashi Takaso, and Kazuhisa Takahashi Department of Orthopedic Surgery, Graduate School of Medicine, Chiba University, Chiba, Japan.
Abstract: Elderly patients with osteoporosis sometimes experience lumbar vertebral fracture and may feel diffuse nonlocalized pain in the back, the lateral portion of the trunk, and the area surrounding the iliac crest. The pattern of sensory innervation of vertebral bodies remains unclear. Some sensory nerves from the L2 and L5 vertebral bodies may enter the paravertebral sympathetic trunks and reach the L2 dorsal root ganglion. Our randomized controlled study was to clarify the effect of L2 spinal nerve block on low back pain originating from acute osteoporotic lumbar vertebral fracture. Patients with low back pain originating from acute L3 or L4 osteoporotic vertebral fractures received a spinal nerve root block (L2 block group, n = 30) or subcutaneous injection (control, n = 30). Both groups received 1.5 mL of 1% lidocaine. The visual analog scale score, Roland Morris Disability Questionnaire, and Short Form questionnaire were examined before and after treatment. In both groups, spinal nerve blocks were significantly effective in alleviating low back pain (P < .05). One hour, 1 week, and 2 weeks after treatment, the visual analog scale score improved more in the L2 block group than in the control group (P < .05). From 1 month to 4 months after treatment, there were no significant differences in the pain scores between groups (P > .05). We conclude that L2 spinal nerve block for acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks, but it had no long-term effects on pain and social function. Perspective: L2 spinal nerve block treatment for L3 or L4 osteoporotic vertebral body fracture was effective. This results suggest that the L2 dorsal root ganglion may innervate the L3 and L4 vertebral bodies in humans. L2 spinal nerve block for lumbar osteoporotic vertebral fracture may be a useful strategy to treat acute low back pain. ª 2009 by the American Pain Society Key words: L2 spinal nerve block, low back pain, osteoporosis, vertebral fracture, randomized controlled study.
M
any studies have investigated lumbar vertebral bodies and intervertebral discs as possible sources of low back pain. Human studies have described the existence of sensory nerve endings in the lumbar vertebral body and in the annulus fibrosus.1,5,7,8 It is believed that such nerve endings originate from the sinuvertebral nerves branching from the ventral ramus of the spinal nerve and the ramus communicans of the corresponding level in humans.3,4
Received September 11, 2008; Revised January 17, 2009; Accepted March 4, 2009. Address reprint requests to Dr Seiji Ohtori, Department of Orthopedic Surgery, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8677, Japan. E-mail: [email protected] 1526-5900/$36.00 ª 2009 by the American Pain Society doi:10.1016/j.jpain.2009.03.002
870
Some elderly patients with osteoporosis experience lumbar vertebral fractures. Whereas some patients feel pain at the fracture site, others experience pain at other sites such as the lateral portion of the trunk and the area surrounding the iliac crest. Thus far, the underlying causes of pain at these secondary sites are largely unknown. Recent studies have revealed that the dorsal portion of the L5/6 intervertebral disc is multisegmentally innervated by dorsal root ganglia (DRGs) from T13 to L6 levels in rats.11,12 Some rat sensory nerve fibers from the L5/6 intervertebral disc pass to upper DRGs via paravertebral sympathetic trunks.11,12 Ohtori et al 10,16 performed an L2 spinal nerve block based on animal sensory innervation and reported that the block was effective in patients with discogenic low back pain.10,16 Based on data from intervertebral discs, the same authors have reported
Ohtori et al that the rat sensory nerve fibers in the L2 and L5 vertebral bodies are derived from the T11-L3 and T13-L6 DRGs, respectively. Most sensory nerves from the L2 and L5 vertebral bodies enter the paravertebral sympathetic trunks and reach the L2 DRG.14,15 However, the effect of L2 spinal nerve block on acute lumbar osteoporotic vertebral fractures in humans is unknown. This study is a randomized controlled trial. The purpose was to clarify the effect of L2 spinal nerve block on low back pain originating from acute osteoporotic lumbar vertebral fracture.
Methods The protocols for human procedures in these experiments received approval from the ethics committees of our institution.
Patients Sixty postmenopausal women ages 50 to 85 years (69.6 6 7.5 years, mean 6 SD) with acute vertebral fractures of the L3 or L4 vertebra were examined at our hospitals between April 2004 and December 2007. Informed consent was obtained from each of the participants. All were diagnosed as having osteoporosis, based on Japanese criteria.17 All patients had low back pain. The backgrounds of the subjects are shown in Table 1. To examine lumbar vertebral fractures, radiography and computed tomography (CT) (all patients) were performed before treatment. The type of fracture was classified as fracture of anterior column only or fracture of anterior and middle column. Other spinal diseases (such as spinal canal stenosis, infection, and tumors) were excluded. Follow-up radiographs in flexion and extension positions were obtained 1, 2, 3, and 4 months after treatment to examine stability of the fracture site and bone union and to exclude any new vertebral fractures. We define instability (-) as bone union at follow up periods. Microdensitometry was used for evaluation of bone mineral density (BMD) in the metacarpals of the hands of all patients before treatment. Osteoporosis was diagnosed in less than 70% of young adult men. Low back pain was assessed with a visual analogue scale (VAS) before and after 1 hour, 1 and 2 weeks, and 1, 2, and 4 months of treatment; with the Roland Morris Disability Questionnaire (RDQ) before and after 1, 2, and 4 months of treatment; and with the Short Form (SF)-36 questionnaire (SF-36) before and after 1, 2, and 4 months of treatment.
Treatment The patients were divided randomly into 2 groups (Fig 1). Patients with low back pain originating from an acute L3 or L4 vertebral fracture received a spinal nerve root block (L2 nerve block group, n = 30) or subcutaneous injection (control, n = 30). Both groups received 1.5 mL of 1% lidocaine solution. In the L2 nerve block group, on the predominantly painful side, a 22-guage spinal nerve block needle was advanced obliquely to the corresponding spinal nerve under fluoroscopic control. Contrast
871 Table 1.
Profiles of Subjects TOTAL
Sex Female Age 55 64 65 74 75 85 Height (cm) Mean 6 SD Min max Body weight (kg) Mean 6 SD Min max Body mass index Mean 6 SD Min max Postmenopausal duration (Y) Mean 6 SD Min max History of oophorectomy Yes No Radiography before treatment Performed Not performed CT before treatment Performed Not performed Radiography after treatment Performed Not performed
60 PATIENT N (%)
60 (100.0) 7 (24.0) 13 (44.0) 10 (34.0) 149.8 6 7.5 138.0 165 51.0 6 8.2 35 69.5 23.5 6 4.5 16.5 30.5 23.0 6 8.5 5 35 0 60 (100.0) 60 (100.0) 0 60 (100.0) 0 60 (100.0) 0
medium (Iotorolan, 0.5 mL; Schering AG, Berlin, Germany) was then injected to confirm the position of the spinal nerve. Unilateral lidocaine administration was then performed. Nonsteroidal anti-inflammatory drugs and soft lumbar support belts were given to control low back pain in all patients. A period of bed rest was allowed according to the degree of low back pain. The maximum rest period was 7 days and the average period was 3 6 1.2 days.
Statistical Analysis Data are presented as mean 6 SD. The Wilcoxon signed rank test was used, and a value of P < .05 was accepted as indicating significance.
Results The VAS scores before blocks ranged from 60 to 100 (70 6 25) in the L2 spinal nerve block group and from 55 to 100 (72 6 22) in the control group. There was no significant difference between groups (P > .05). In both groups, treatment was significantly effective in attenuating low back pain 1 hour after injection (P < .05) (Fig 1). Average VAS scores were 35 6 23 (1 hour), 43 6 26 (1 week), and 45 6 19 (2 weeks) in the L2 spinal nerve block group and 56 6 23 (1 hour), 57 6 21 (1 week), and 55 6 20 (2 weeks) in the control group. There were significant differences between the groups 1 hour and
872
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture Roland Morris Disability Questionnaire
L3 or L4 osteoporotic vertebral fracture (n=60)
21.0 18.0
L2 root block Control
15.0
Randomised (n=30 each)
12.0 9.0 6.0
Spinal nerve root block (L2 block group, n=30)
Subcutaneous injection (Control, n=30).
Follow up to 4 months (L2 block group, n=30)
Follow up to 4 months (Control, n=30)
3.0 0.0
Before treatment
1M
2M
3M
4M
Figure 3. RDQ scores for low back pain after lidocaine injection in the L2 spinal nerve and control groups.
Radiographic and CT Examination Analysed (L2 block group, n=30)
Analysed (Control, n=30)
Figure 1. Consort diagram of the current randomized controlled trial study. 1 and 2 weeks after the injections (P < .05) (Fig 1). The average VAS scores gradually decreased from 1 to 4 months after treatment. The average VAS scores were not significantly different between the groups 1 to 4 months after the injections (P > .05) (Fig 2). The RDQ scores before block ranged from 10 to 18 (14 6 5.4) in the L2 spinal nerve block group and from 9 to 18 (15 6 4.0) in the control group. There was no significant difference between groups (P > .05). The average RDQ scores decreased slowly, similar to the VAS scores, from 1 to 4 months. The average RDQ scores were not significantly different between groups 1 to 4 months after injection (P > .05) (Fig 3). Table 2 shows 8 items of the SF-36. As observed with VAS and RDQ scores, each score improved slowly over 4 months. However, the average SF-36 scores were not significantly different between groups 1 to 4 months after injection (P > .05) (Table 2). VAS score 100.0 L2 root block
80.0
Control
* p<0.05
60.0
*
40.0
*
*
20.0
0.0 Before treatment
1 hour 1 week 2 week after after after treatment treatment treatment
1M
2M
3M
4M
Figure 2. VAS scores for low back pain after lidocaine injection in the L2 spinal nerve and control groups.
Details are shown in Table 3. Comparing the L2 spinal nerve block group and the control group, there were significant differences in the ratio of type of fracture at injury, of bone union, and of instability of fracture site at follow-up periods (P > .05). Furthermore, any new vertebral fractures were seen in follow-up periods.
Discussion A Pain Pathway From the L3 or L4 Vertebral Body in Humans This study demonstrated that L2 spinal nerve block to an acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks. This result suggested that the L2 dorsal root ganglion may innervate the L3 and L4 vertebral bodies in human. We have reported that sensory nerve fibers in the L2 and L5 vertebral bodies are derived from the T11-L3 and T13-L6 DRGs in rats, respectively.14,15 Some sensory nerves from the L2 and L5 vertebral bodies enter the paravertebral sympathetic trunks and reach the DRGs at multisegmental levels. In DRGs, most sensory nerve fibers from the L2 DRG innervated L2 and L5 vertebral bodies. We studied L3 or L4 vertebral fractures in humans because, if we used L2 spinal nerve block for an acute L2 vertebral body fracture, we were concerned about direct infiltration of lidocaine to the L2 vertebral body. We also excluded L5 vertebral body fracture because it is usually rare in clinical practice. We did not examine sensory innervation of the L3 or L4 vertebral body in rats. However, the result that L2 spinal nerve block of acute L3 or L4 vertebral body fracture was effective is consistent with the hypothesis that the L2 DRG innervates the L3 and L4 vertebral bodies in humans.
Pain From Acute L3 or L4 Vertebral Body Fracture In patients with severe back pain and markedly reduced disc height, proliferation of blood vessels and accompanying nerve fibers have been observed in the
Ohtori et al Table 2.
873
Short Form-36
8 ITEMS (Physical functioning) PF Before 1 mo 2 mo 4 mo (Physical role) PR Before 1 mo 2 mo 4 mo (Bodily pain) BP Before 1 mo 2 mo 4 mo (Health perception) GH Before 1 mo 2 mo 4 mo (Vitality) VT Before 1 mo 2 mo 4 mo (Social functioning) SF Before 1 mo 2 mo 4 mo (Emotional role) ER Before 1 mo 2 mo 4 mo (Mental health) MH Before 1 mo 2 mo 4 mo
Table 3.
L2 BLOCK GROUP
CONTROL GROUP
MEAN 6 SD
MEAN 6 SD
WILCOXON SIGNED RANK TEST P VALUE
9.5 6 9.5 37.5 6 14.0 39.0 6 13.0 41.5 6 12.0
10.5 6 15.0 36.3 6 14.5 39.5 6 13.0 40.1 6 15.0
NS NS NS NS
10.0 6 15.0 37.35 6 11.0 38.7 6 12.3 42.3 6 12.0
15.0 6 12.3 32.7 6 13.6 36.5 6 15.3 42.0 6 12.0
NS NS NS NS
15.6 6 8.6 40.0 6 8.0 40.1 6 6.0 49.6 6 12.6
12.3 6 8.5 40.2 6 6.2 36.8 6 7.4 50.2 6 13.2
NS NS NS NS
15.0 6 11.0 46.3 6 12.3 50.2 6 16.0 52.0 6 13.0
16.3 6 9.0 43.0 6 10.2 45.2 6 10.2 49.0 6 9.6
NS NS NS NS
20.0 6 11.0 48.5 6 9.3 40.5 6 6.8 52.0 6 11.0
18.5 6 6.9 44.5 6 15.2 45.0 6 9.6 49.8 6 11.2
NS P = 0.045 NS NS
20.3 6 9.2 42.3 6 5.6 44.2 6 10.2 45.2 6 10.0
18.5 6 8.5 36.5 6 6.5 45.2 6 10.0 18.0 6 10.0
NS P = 0.045 NS NS
20.3 6 10.0 35.6 6 11.0 44.1 6 10.6 47.0 6 10.2
21.0 6 9.0 36.0 6 11.0 45.2 6 11.7 48.9 6 9.9
NS NS NS NS
16.0 6 7.5 45.0 6 9.4 50.1 6 8.5 50.4 6 9.6
18.5 6 10.0 50.2 6 8.3 50.1 6 10.0 52.0 6 9.6
NS NS NS NS
endplate region and the underlying vertebral bodies. Many of these nerves have been shown to be immunoreactive to substance P (SP) and calcitonin gene-related peptide (CGRP), and SP- and CGRP-immunoreactive nociceptors not in association with blood vessels have been observed.1 Both peptidergic nerves, represented by immunoreactivity to neuropeptide Y and vasoactive intestinal polypeptide, and noradrenergic nerves as represented by tyrosine hydroxylase immunoreactivity, have been observed in the bone and periosteum of vertebral bodies. These nerves were most abundant in the vertebral bone marrow and periosteum.5 In rats, vertebral bodies are mainly innervated by CGRP-IR DRG neurons.14 CGRP-IR neurons may play a role in pain sensation via peptidergic DRG neurons related to inflammatory pain. Thus, the vertebral body is innervated by several peptidergic nerves related to pain transmission.
Radiographic and CT Examination
CT befor treatment Type of fracture Anterior column only Both anterior and middle column Radiography before and after treatment Union or instability (-) Before 1 mo 2 mo 3 mo 4 mo
NO. OF PATIENTS L2 BLOCK GROUP
NO. OF PATIENTS CONTROL GROUP
18 12
20 10
0 5 8 16 20
0 6 10 15 22
Recently some authors have reported a fracture pain model in tibia and femur in rats.6 Females and males showed highly robust pain behaviors that were maximal at day 1 after femur fracture and returned gradually to normal unfractured levels at days 14 to 21 after fracture. It has been reported that tibia fracture upregulated nerve growth factor (NGF) or TNF-a expression in hind paw skin and tibia bone along with sciatic nerve neuropeptide content.18,19 Nociceptive sensitization, enhanced spinal cord Fos expression, osteopenia, and enhanced cytokine content of hind paw skin on the side of the fracture occurred. Anti-NGF and anti–TNF-a treatment reduced neuropeptide levels in sciatic nerve and reduced nociceptive sensitization.18,19 These results demonstrated that inflammation from the bone fracture site was important to pain transmission. Generally, NGFdependent neurons contain neuropeptides such as SP and CGRP.2 Thus, SP and CGRP immunoreactive sensory nerve fibers in vertebral body may transmit pain sensation via NGF and TNF-a pathways. Although the duration of efficacy of lidocaine is about 2 hours, we often see longer effects after spinal nerve root block to a spinal disorder, and some patients experience relief from pain for 2 or 3 weeks.9 This phenomenon is usually explained as breaking a vicious cycle of pain by exerting a positive effect on memory. It has been reported that acid-sensing ion channel 3 (ASIC3), a depolarizing sodium channel gated by protons during tissue acidosis, is specifically expressed in sensory neurons. The channel has been associated with cardiac ischemic and inflammatory pain. Lidocaine decreased ASIC3 expression in DRG neurons and pain associated with disc herniation in rat for 1 week.13 In visceral pain, DRG neurons innervating visceral organs are small and medium-sized neurons that express ASIC3.20 Primary afferent activation by mechanical stimulation from visceral organs through ASIC3 may produce action potentials, which in turn results in pain transmission in DRG neurons.20 We did not examine the AISC3-positive neurons innervating the vertebral body in animals and humans. However, if the vertebral body is similarly innervated by ASIC3-positive DRG neurons, ASIC3 may be important for transmitting pain from the vertebral body. In the
874
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture
current study, the effective lidocaine relief period was 2 weeks. We speculate that lidocaine decreased gated sodium channels, such as ASIC3, in DRG neurons activated by vertebral bone fracture for 2 weeks.
Limitations of the Study Although this is a randomized controlled trial, the number of patients was small. Unilateral 1.5 mL lidocaine solution administration was performed for the L2 nerve root block group, and the nerve block was effective. However, we cannot exclude the possibility of leakage of lidocaine into the L3 or L4 vertebral body in the L2 nerve root block group. In most cases in the current study, the origin of low back pain was thought to be from a vertebral fracture. However, other factors such as paravertebral muscles, facet degenerations, and spondylosis must be associated with a fracture for the devel-
opment and maintenance of osteoporosis-related low back pain, even in acute phases. In fact, facet joints and back muscles at lower levels have been reported to be innervated by L2 dorsal root ganglia.21,22 Second, because placebo effects are known to be greater in major treatments such as spinal nerve block than in minor injections or medication, it is difficult to compare spinal nerve block with subcutaneous injection. Finally, to evaluate placebo effects, baseline treatment group without any injection should be used as a control without spinal nerve and subcutaneous injection. Further study is needed to clarify these points. In conclusion, L2 spinal nerve block to acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks. If humans have sensory innervation from the lumbar vertebral body to the DRG similar to that in rats, L2 spinal nerve block to lumbar osteoporotic vertebral fracture may be a useful strategy to treat acute low back pain.
References
tion of the dorsal portion of the lumbar intervertebral disc in rats. Spine 24:2295-2299, 1999
1. Ahmed M, Bjurholm A, Kreicbergs A, Schultzberg M: Neuropeptide Y, tyrosine hydroxylase and vasoactive intestinal polypeptide-immunoreactive nerve fibers in the vertebral bodies, discs, dura mater, and spinal ligaments of the rat lumbar spine. Spine 18:268-273, 1993
12. Ohtori S, Takahashi K, Chiba T, Yamagata M, Sameda H, Moriya H: Sensory innervation of the dorsal portion of the lumbar intervertebral discs in rats. Spine 26:946-950, 2001
2. Averill S, McMahon SB, Clary DO, Reichardt LF, Priestley JV: Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur J Neurosci 7:1484-1494, 1995 3. Bogduk N, Tynan W, Wilson AS: The nerve supply to the human lumbar intervertebral discs. J Anat 132:39-56, 1981 4. Bogduk N: The innervation of the lumbar spine. Spine 8: 286-293, 1983 5. Brown MF, Hukkanen MV, McCarthy ID, Redfern DR, Batten JJ, Crock HV, Hughes SP, Polak JM: Sensory and sympathetic innervation of the vertebral endplate in patients with degenerative disc disease. J Bone Joint Surg [Br] 79: 147-153, 1997 6. Freeman KT, Koewler NJ, Jimenez-Andrade JM, Buus RJ, Herrera MB, Martin CD, Ghilardi JR, Kuskowski MA, Mantyh PW: A fracture pain model in the rat: Adaptation of a closed femur fracture model to study skeletal pain. Anesthesiology 108:473-483, 2008 7. Hirsch C, Ingelmark B, Miller M: The anatomical basis for low back pain: Studies on the presence of sensory nerve endings in ligamentous, capsular, and intervertebral disc structures in the human lumbar spine. Acta Orthop Scand 33: 1-17, 1963 8. Jackson HC, Winkelmann RK, Bickel WH: Nerve endings in the human lumbar spinal column and related structures. J Bone Joint Surg [Am] 48:1272-1281, 1996 9. Kikuchi S, Hasue M, Nishiyama K, Ito T: Anatomic and clinical studies of radicular symptoms. Spine 9:23-30, 1984 10. Nakamura S, Takahashi K, Takahashi Y, Yamagata M, Moriya H: The afferent pathways of discogenic low back pain. J Bone Joint Surg [Br] 78:606-612, 1996 11. Ohtori S, Takahashi Y, Takahashi K, Yamagata M, Chiba T, Tanaka K, Hirayama J, Moriya H: Sensory innerva-
13. Ohtori S, Inoue G, Koshi T, Ito T, Doya H, Saito T, Moriya H, Takahashi K: Up-regulation of acid-sensing ion channel 3 in dorsal root ganglion neurons following application of nucleus pulposus on nerve root in rats. Spine 31: 2048-2052, 2006 14. Ohtori S, Inoue G, Koshi T, Ito T, Yamashita M, Yamauchi K, Suzuki M, Doya H, Moriya H, Takahashi Y, Takahashi K: Characteristics of sensory dorsal root ganglia neurons innervating the lumbar vertebral body in rats. J Pain 8:483-488, 2007 15. Ohtori S, Inoue G, Koshi T, Ito T, Watanabe T, Yamashita M, Yamauchi K, Suzuki M, Doya H, Moriya H, Takahashi Y, Takahashi K: Sensory innervation of lumbar vertebral bodies in rats. Spine 32:1498-14502, 2007 16. Ohtori S, Sameda H, Murata Y, Hanaoka E, Nakamura S, Takahashi Y, Yamagata M, Takahashi K: Involvement of a sensory afferent pathway in human discogenic low back pain: Comparison of upper and lower spinal nerve block effects on discogenic low back pain. Chiba Med J 84: 121-126, 2008 17. Orimo H, Hayashi Y, Fukunaga M, Sone T, Fujiwara S, Shiraki M, Kushida K, Miyamoto S, Soen S, Nishimura J, Oh-Hashi Y, Hosoi T, Gorai I, Tanaka H, Igai T, Kishimoto H: Diagnostic criteria for primary osteoporosis: Year 2000 revision. J Bone Miner Metab 19:331-337, 2001 18. Sabsovich I, Wei T, Guo TZ, Zhao R, Shi X, Li X, Yeomans DC, Klyukinov M, Kingery WS, Clark JD: Effect of anti-NGF antibodies in a rat tibia fracture model of complex regional pain syndrome type I. Pain 15:47-60, 2008 19. Sabsovich I, Guo TZ, Wei T, Zhao R, Li X, Clark DJ, Geis C, Sommer C, Kingery WS: TNF signaling contributes to the development of nociceptive sensitization in a tibia fracture model of complex regional pain syndrome type I. Pain 2007 [Epub ahead of print]. 20. Sakurai J, Obata K, Ozaki N, Tokunaga A, Kobayashi K, Yamanaka H, Dai Y, Kondo T, Miyoshi K, Sugiura Y, Matsumoto T, Miwa H, Noguchi K: Activation of
Ohtori et al extracellular signal-regulated protein kinase in sensory neurons after noxious gastric distention and its involvement in acute visceral pain in rats. Gastroenterology 134: 1094-1103, 2008 21. Suseki K, Takahashi Y, Takahashi K, Chiba T, Tanaka K, Morinaga T, Nakamura S, Moriya H: Innervation of the
875 lumbar facet joints: Origins and functions. Spine 22: 477-485, 1997 22. Taguchi T, John V, Hoheisel U, Mense S: Neuroanatomical pathway of nociception originating in a low back muscle (multifidus) in the rat. Neurosci Lett 427:22-27, 2007
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture Seiji Ohtori, Masaomi Yamashita, Gen Inoue, Kazuyo Yamauchi, Munetaka Suzuki, Sumihisa Orita, Yawara Eguchi, Nobuyasu Ochiai, Shunji Kishida, Masashi Takaso, and Kazuhisa Takahashi Department of Orthopedic Surgery, Graduate School of Medicine, Chiba University, Chiba, Japan.
Abstract: Elderly patients with osteoporosis sometimes experience lumbar vertebral fracture and may feel diffuse nonlocalized pain in the back, the lateral portion of the trunk, and the area surrounding the iliac crest. The pattern of sensory innervation of vertebral bodies remains unclear. Some sensory nerves from the L2 and L5 vertebral bodies may enter the paravertebral sympathetic trunks and reach the L2 dorsal root ganglion. Our randomized controlled study was to clarify the effect of L2 spinal nerve block on low back pain originating from acute osteoporotic lumbar vertebral fracture. Patients with low back pain originating from acute L3 or L4 osteoporotic vertebral fractures received a spinal nerve root block (L2 block group, n = 30) or subcutaneous injection (control, n = 30). Both groups received 1.5 mL of 1% lidocaine. The visual analog scale score, Roland Morris Disability Questionnaire, and Short Form questionnaire were examined before and after treatment. In both groups, spinal nerve blocks were significantly effective in alleviating low back pain (P < .05). One hour, 1 week, and 2 weeks after treatment, the visual analog scale score improved more in the L2 block group than in the control group (P < .05). From 1 month to 4 months after treatment, there were no significant differences in the pain scores between groups (P > .05). We conclude that L2 spinal nerve block for acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks, but it had no long-term effects on pain and social function. Perspective: L2 spinal nerve block treatment for L3 or L4 osteoporotic vertebral body fracture was effective. This results suggest that the L2 dorsal root ganglion may innervate the L3 and L4 vertebral bodies in humans. L2 spinal nerve block for lumbar osteoporotic vertebral fracture may be a useful strategy to treat acute low back pain. ª 2009 by the American Pain Society Key words: L2 spinal nerve block, low back pain, osteoporosis, vertebral fracture, randomized controlled study.
M
any studies have investigated lumbar vertebral bodies and intervertebral discs as possible sources of low back pain. Human studies have described the existence of sensory nerve endings in the lumbar vertebral body and in the annulus fibrosus.1,5,7,8 It is believed that such nerve endings originate from the sinuvertebral nerves branching from the ventral ramus of the spinal nerve and the ramus communicans of the corresponding level in humans.3,4
Received September 11, 2008; Revised January 17, 2009; Accepted March 4, 2009. Address reprint requests to Dr Seiji Ohtori, Department of Orthopedic Surgery, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8677, Japan. E-mail: [email protected] 1526-5900/$36.00 ª 2009 by the American Pain Society doi:10.1016/j.jpain.2009.03.002
870
Some elderly patients with osteoporosis experience lumbar vertebral fractures. Whereas some patients feel pain at the fracture site, others experience pain at other sites such as the lateral portion of the trunk and the area surrounding the iliac crest. Thus far, the underlying causes of pain at these secondary sites are largely unknown. Recent studies have revealed that the dorsal portion of the L5/6 intervertebral disc is multisegmentally innervated by dorsal root ganglia (DRGs) from T13 to L6 levels in rats.11,12 Some rat sensory nerve fibers from the L5/6 intervertebral disc pass to upper DRGs via paravertebral sympathetic trunks.11,12 Ohtori et al 10,16 performed an L2 spinal nerve block based on animal sensory innervation and reported that the block was effective in patients with discogenic low back pain.10,16 Based on data from intervertebral discs, the same authors have reported
Ohtori et al that the rat sensory nerve fibers in the L2 and L5 vertebral bodies are derived from the T11-L3 and T13-L6 DRGs, respectively. Most sensory nerves from the L2 and L5 vertebral bodies enter the paravertebral sympathetic trunks and reach the L2 DRG.14,15 However, the effect of L2 spinal nerve block on acute lumbar osteoporotic vertebral fractures in humans is unknown. This study is a randomized controlled trial. The purpose was to clarify the effect of L2 spinal nerve block on low back pain originating from acute osteoporotic lumbar vertebral fracture.
Methods The protocols for human procedures in these experiments received approval from the ethics committees of our institution.
Patients Sixty postmenopausal women ages 50 to 85 years (69.6 6 7.5 years, mean 6 SD) with acute vertebral fractures of the L3 or L4 vertebra were examined at our hospitals between April 2004 and December 2007. Informed consent was obtained from each of the participants. All were diagnosed as having osteoporosis, based on Japanese criteria.17 All patients had low back pain. The backgrounds of the subjects are shown in Table 1. To examine lumbar vertebral fractures, radiography and computed tomography (CT) (all patients) were performed before treatment. The type of fracture was classified as fracture of anterior column only or fracture of anterior and middle column. Other spinal diseases (such as spinal canal stenosis, infection, and tumors) were excluded. Follow-up radiographs in flexion and extension positions were obtained 1, 2, 3, and 4 months after treatment to examine stability of the fracture site and bone union and to exclude any new vertebral fractures. We define instability (-) as bone union at follow up periods. Microdensitometry was used for evaluation of bone mineral density (BMD) in the metacarpals of the hands of all patients before treatment. Osteoporosis was diagnosed in less than 70% of young adult men. Low back pain was assessed with a visual analogue scale (VAS) before and after 1 hour, 1 and 2 weeks, and 1, 2, and 4 months of treatment; with the Roland Morris Disability Questionnaire (RDQ) before and after 1, 2, and 4 months of treatment; and with the Short Form (SF)-36 questionnaire (SF-36) before and after 1, 2, and 4 months of treatment.
Treatment The patients were divided randomly into 2 groups (Fig 1). Patients with low back pain originating from an acute L3 or L4 vertebral fracture received a spinal nerve root block (L2 nerve block group, n = 30) or subcutaneous injection (control, n = 30). Both groups received 1.5 mL of 1% lidocaine solution. In the L2 nerve block group, on the predominantly painful side, a 22-guage spinal nerve block needle was advanced obliquely to the corresponding spinal nerve under fluoroscopic control. Contrast
871 Table 1.
Profiles of Subjects TOTAL
Sex Female Age 55 64 65 74 75 85 Height (cm) Mean 6 SD Min max Body weight (kg) Mean 6 SD Min max Body mass index Mean 6 SD Min max Postmenopausal duration (Y) Mean 6 SD Min max History of oophorectomy Yes No Radiography before treatment Performed Not performed CT before treatment Performed Not performed Radiography after treatment Performed Not performed
60 PATIENT N (%)
60 (100.0) 7 (24.0) 13 (44.0) 10 (34.0) 149.8 6 7.5 138.0 165 51.0 6 8.2 35 69.5 23.5 6 4.5 16.5 30.5 23.0 6 8.5 5 35 0 60 (100.0) 60 (100.0) 0 60 (100.0) 0 60 (100.0) 0
medium (Iotorolan, 0.5 mL; Schering AG, Berlin, Germany) was then injected to confirm the position of the spinal nerve. Unilateral lidocaine administration was then performed. Nonsteroidal anti-inflammatory drugs and soft lumbar support belts were given to control low back pain in all patients. A period of bed rest was allowed according to the degree of low back pain. The maximum rest period was 7 days and the average period was 3 6 1.2 days.
Statistical Analysis Data are presented as mean 6 SD. The Wilcoxon signed rank test was used, and a value of P < .05 was accepted as indicating significance.
Results The VAS scores before blocks ranged from 60 to 100 (70 6 25) in the L2 spinal nerve block group and from 55 to 100 (72 6 22) in the control group. There was no significant difference between groups (P > .05). In both groups, treatment was significantly effective in attenuating low back pain 1 hour after injection (P < .05) (Fig 1). Average VAS scores were 35 6 23 (1 hour), 43 6 26 (1 week), and 45 6 19 (2 weeks) in the L2 spinal nerve block group and 56 6 23 (1 hour), 57 6 21 (1 week), and 55 6 20 (2 weeks) in the control group. There were significant differences between the groups 1 hour and
872
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture Roland Morris Disability Questionnaire
L3 or L4 osteoporotic vertebral fracture (n=60)
21.0 18.0
L2 root block Control
15.0
Randomised (n=30 each)
12.0 9.0 6.0
Spinal nerve root block (L2 block group, n=30)
Subcutaneous injection (Control, n=30).
Follow up to 4 months (L2 block group, n=30)
Follow up to 4 months (Control, n=30)
3.0 0.0
Before treatment
1M
2M
3M
4M
Figure 3. RDQ scores for low back pain after lidocaine injection in the L2 spinal nerve and control groups.
Radiographic and CT Examination Analysed (L2 block group, n=30)
Analysed (Control, n=30)
Figure 1. Consort diagram of the current randomized controlled trial study. 1 and 2 weeks after the injections (P < .05) (Fig 1). The average VAS scores gradually decreased from 1 to 4 months after treatment. The average VAS scores were not significantly different between the groups 1 to 4 months after the injections (P > .05) (Fig 2). The RDQ scores before block ranged from 10 to 18 (14 6 5.4) in the L2 spinal nerve block group and from 9 to 18 (15 6 4.0) in the control group. There was no significant difference between groups (P > .05). The average RDQ scores decreased slowly, similar to the VAS scores, from 1 to 4 months. The average RDQ scores were not significantly different between groups 1 to 4 months after injection (P > .05) (Fig 3). Table 2 shows 8 items of the SF-36. As observed with VAS and RDQ scores, each score improved slowly over 4 months. However, the average SF-36 scores were not significantly different between groups 1 to 4 months after injection (P > .05) (Table 2). VAS score 100.0 L2 root block
80.0
Control
* p<0.05
60.0
*
40.0
*
*
20.0
0.0 Before treatment
1 hour 1 week 2 week after after after treatment treatment treatment
1M
2M
3M
4M
Figure 2. VAS scores for low back pain after lidocaine injection in the L2 spinal nerve and control groups.
Details are shown in Table 3. Comparing the L2 spinal nerve block group and the control group, there were significant differences in the ratio of type of fracture at injury, of bone union, and of instability of fracture site at follow-up periods (P > .05). Furthermore, any new vertebral fractures were seen in follow-up periods.
Discussion A Pain Pathway From the L3 or L4 Vertebral Body in Humans This study demonstrated that L2 spinal nerve block to an acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks. This result suggested that the L2 dorsal root ganglion may innervate the L3 and L4 vertebral bodies in human. We have reported that sensory nerve fibers in the L2 and L5 vertebral bodies are derived from the T11-L3 and T13-L6 DRGs in rats, respectively.14,15 Some sensory nerves from the L2 and L5 vertebral bodies enter the paravertebral sympathetic trunks and reach the DRGs at multisegmental levels. In DRGs, most sensory nerve fibers from the L2 DRG innervated L2 and L5 vertebral bodies. We studied L3 or L4 vertebral fractures in humans because, if we used L2 spinal nerve block for an acute L2 vertebral body fracture, we were concerned about direct infiltration of lidocaine to the L2 vertebral body. We also excluded L5 vertebral body fracture because it is usually rare in clinical practice. We did not examine sensory innervation of the L3 or L4 vertebral body in rats. However, the result that L2 spinal nerve block of acute L3 or L4 vertebral body fracture was effective is consistent with the hypothesis that the L2 DRG innervates the L3 and L4 vertebral bodies in humans.
Pain From Acute L3 or L4 Vertebral Body Fracture In patients with severe back pain and markedly reduced disc height, proliferation of blood vessels and accompanying nerve fibers have been observed in the
Ohtori et al Table 2.
873
Short Form-36
8 ITEMS (Physical functioning) PF Before 1 mo 2 mo 4 mo (Physical role) PR Before 1 mo 2 mo 4 mo (Bodily pain) BP Before 1 mo 2 mo 4 mo (Health perception) GH Before 1 mo 2 mo 4 mo (Vitality) VT Before 1 mo 2 mo 4 mo (Social functioning) SF Before 1 mo 2 mo 4 mo (Emotional role) ER Before 1 mo 2 mo 4 mo (Mental health) MH Before 1 mo 2 mo 4 mo
Table 3.
L2 BLOCK GROUP
CONTROL GROUP
MEAN 6 SD
MEAN 6 SD
WILCOXON SIGNED RANK TEST P VALUE
9.5 6 9.5 37.5 6 14.0 39.0 6 13.0 41.5 6 12.0
10.5 6 15.0 36.3 6 14.5 39.5 6 13.0 40.1 6 15.0
NS NS NS NS
10.0 6 15.0 37.35 6 11.0 38.7 6 12.3 42.3 6 12.0
15.0 6 12.3 32.7 6 13.6 36.5 6 15.3 42.0 6 12.0
NS NS NS NS
15.6 6 8.6 40.0 6 8.0 40.1 6 6.0 49.6 6 12.6
12.3 6 8.5 40.2 6 6.2 36.8 6 7.4 50.2 6 13.2
NS NS NS NS
15.0 6 11.0 46.3 6 12.3 50.2 6 16.0 52.0 6 13.0
16.3 6 9.0 43.0 6 10.2 45.2 6 10.2 49.0 6 9.6
NS NS NS NS
20.0 6 11.0 48.5 6 9.3 40.5 6 6.8 52.0 6 11.0
18.5 6 6.9 44.5 6 15.2 45.0 6 9.6 49.8 6 11.2
NS P = 0.045 NS NS
20.3 6 9.2 42.3 6 5.6 44.2 6 10.2 45.2 6 10.0
18.5 6 8.5 36.5 6 6.5 45.2 6 10.0 18.0 6 10.0
NS P = 0.045 NS NS
20.3 6 10.0 35.6 6 11.0 44.1 6 10.6 47.0 6 10.2
21.0 6 9.0 36.0 6 11.0 45.2 6 11.7 48.9 6 9.9
NS NS NS NS
16.0 6 7.5 45.0 6 9.4 50.1 6 8.5 50.4 6 9.6
18.5 6 10.0 50.2 6 8.3 50.1 6 10.0 52.0 6 9.6
NS NS NS NS
endplate region and the underlying vertebral bodies. Many of these nerves have been shown to be immunoreactive to substance P (SP) and calcitonin gene-related peptide (CGRP), and SP- and CGRP-immunoreactive nociceptors not in association with blood vessels have been observed.1 Both peptidergic nerves, represented by immunoreactivity to neuropeptide Y and vasoactive intestinal polypeptide, and noradrenergic nerves as represented by tyrosine hydroxylase immunoreactivity, have been observed in the bone and periosteum of vertebral bodies. These nerves were most abundant in the vertebral bone marrow and periosteum.5 In rats, vertebral bodies are mainly innervated by CGRP-IR DRG neurons.14 CGRP-IR neurons may play a role in pain sensation via peptidergic DRG neurons related to inflammatory pain. Thus, the vertebral body is innervated by several peptidergic nerves related to pain transmission.
Radiographic and CT Examination
CT befor treatment Type of fracture Anterior column only Both anterior and middle column Radiography before and after treatment Union or instability (-) Before 1 mo 2 mo 3 mo 4 mo
NO. OF PATIENTS L2 BLOCK GROUP
NO. OF PATIENTS CONTROL GROUP
18 12
20 10
0 5 8 16 20
0 6 10 15 22
Recently some authors have reported a fracture pain model in tibia and femur in rats.6 Females and males showed highly robust pain behaviors that were maximal at day 1 after femur fracture and returned gradually to normal unfractured levels at days 14 to 21 after fracture. It has been reported that tibia fracture upregulated nerve growth factor (NGF) or TNF-a expression in hind paw skin and tibia bone along with sciatic nerve neuropeptide content.18,19 Nociceptive sensitization, enhanced spinal cord Fos expression, osteopenia, and enhanced cytokine content of hind paw skin on the side of the fracture occurred. Anti-NGF and anti–TNF-a treatment reduced neuropeptide levels in sciatic nerve and reduced nociceptive sensitization.18,19 These results demonstrated that inflammation from the bone fracture site was important to pain transmission. Generally, NGFdependent neurons contain neuropeptides such as SP and CGRP.2 Thus, SP and CGRP immunoreactive sensory nerve fibers in vertebral body may transmit pain sensation via NGF and TNF-a pathways. Although the duration of efficacy of lidocaine is about 2 hours, we often see longer effects after spinal nerve root block to a spinal disorder, and some patients experience relief from pain for 2 or 3 weeks.9 This phenomenon is usually explained as breaking a vicious cycle of pain by exerting a positive effect on memory. It has been reported that acid-sensing ion channel 3 (ASIC3), a depolarizing sodium channel gated by protons during tissue acidosis, is specifically expressed in sensory neurons. The channel has been associated with cardiac ischemic and inflammatory pain. Lidocaine decreased ASIC3 expression in DRG neurons and pain associated with disc herniation in rat for 1 week.13 In visceral pain, DRG neurons innervating visceral organs are small and medium-sized neurons that express ASIC3.20 Primary afferent activation by mechanical stimulation from visceral organs through ASIC3 may produce action potentials, which in turn results in pain transmission in DRG neurons.20 We did not examine the AISC3-positive neurons innervating the vertebral body in animals and humans. However, if the vertebral body is similarly innervated by ASIC3-positive DRG neurons, ASIC3 may be important for transmitting pain from the vertebral body. In the
874
L2 Spinal Nerve–Block Effects on Acute Low Back Pain From Osteoporotic Vertebral Fracture
current study, the effective lidocaine relief period was 2 weeks. We speculate that lidocaine decreased gated sodium channels, such as ASIC3, in DRG neurons activated by vertebral bone fracture for 2 weeks.
Limitations of the Study Although this is a randomized controlled trial, the number of patients was small. Unilateral 1.5 mL lidocaine solution administration was performed for the L2 nerve root block group, and the nerve block was effective. However, we cannot exclude the possibility of leakage of lidocaine into the L3 or L4 vertebral body in the L2 nerve root block group. In most cases in the current study, the origin of low back pain was thought to be from a vertebral fracture. However, other factors such as paravertebral muscles, facet degenerations, and spondylosis must be associated with a fracture for the devel-
opment and maintenance of osteoporosis-related low back pain, even in acute phases. In fact, facet joints and back muscles at lower levels have been reported to be innervated by L2 dorsal root ganglia.21,22 Second, because placebo effects are known to be greater in major treatments such as spinal nerve block than in minor injections or medication, it is difficult to compare spinal nerve block with subcutaneous injection. Finally, to evaluate placebo effects, baseline treatment group without any injection should be used as a control without spinal nerve and subcutaneous injection. Further study is needed to clarify these points. In conclusion, L2 spinal nerve block to acute L3 or L4 osteoporotic vertebral body fracture was effective for 2 weeks. If humans have sensory innervation from the lumbar vertebral body to the DRG similar to that in rats, L2 spinal nerve block to lumbar osteoporotic vertebral fracture may be a useful strategy to treat acute low back pain.
References
tion of the dorsal portion of the lumbar intervertebral disc in rats. Spine 24:2295-2299, 1999
1. Ahmed M, Bjurholm A, Kreicbergs A, Schultzberg M: Neuropeptide Y, tyrosine hydroxylase and vasoactive intestinal polypeptide-immunoreactive nerve fibers in the vertebral bodies, discs, dura mater, and spinal ligaments of the rat lumbar spine. Spine 18:268-273, 1993
12. Ohtori S, Takahashi K, Chiba T, Yamagata M, Sameda H, Moriya H: Sensory innervation of the dorsal portion of the lumbar intervertebral discs in rats. Spine 26:946-950, 2001
2. Averill S, McMahon SB, Clary DO, Reichardt LF, Priestley JV: Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur J Neurosci 7:1484-1494, 1995 3. Bogduk N, Tynan W, Wilson AS: The nerve supply to the human lumbar intervertebral discs. J Anat 132:39-56, 1981 4. Bogduk N: The innervation of the lumbar spine. Spine 8: 286-293, 1983 5. Brown MF, Hukkanen MV, McCarthy ID, Redfern DR, Batten JJ, Crock HV, Hughes SP, Polak JM: Sensory and sympathetic innervation of the vertebral endplate in patients with degenerative disc disease. J Bone Joint Surg [Br] 79: 147-153, 1997 6. Freeman KT, Koewler NJ, Jimenez-Andrade JM, Buus RJ, Herrera MB, Martin CD, Ghilardi JR, Kuskowski MA, Mantyh PW: A fracture pain model in the rat: Adaptation of a closed femur fracture model to study skeletal pain. Anesthesiology 108:473-483, 2008 7. Hirsch C, Ingelmark B, Miller M: The anatomical basis for low back pain: Studies on the presence of sensory nerve endings in ligamentous, capsular, and intervertebral disc structures in the human lumbar spine. Acta Orthop Scand 33: 1-17, 1963 8. Jackson HC, Winkelmann RK, Bickel WH: Nerve endings in the human lumbar spinal column and related structures. J Bone Joint Surg [Am] 48:1272-1281, 1996 9. Kikuchi S, Hasue M, Nishiyama K, Ito T: Anatomic and clinical studies of radicular symptoms. Spine 9:23-30, 1984 10. Nakamura S, Takahashi K, Takahashi Y, Yamagata M, Moriya H: The afferent pathways of discogenic low back pain. J Bone Joint Surg [Br] 78:606-612, 1996 11. Ohtori S, Takahashi Y, Takahashi K, Yamagata M, Chiba T, Tanaka K, Hirayama J, Moriya H: Sensory innerva-
13. Ohtori S, Inoue G, Koshi T, Ito T, Doya H, Saito T, Moriya H, Takahashi K: Up-regulation of acid-sensing ion channel 3 in dorsal root ganglion neurons following application of nucleus pulposus on nerve root in rats. Spine 31: 2048-2052, 2006 14. Ohtori S, Inoue G, Koshi T, Ito T, Yamashita M, Yamauchi K, Suzuki M, Doya H, Moriya H, Takahashi Y, Takahashi K: Characteristics of sensory dorsal root ganglia neurons innervating the lumbar vertebral body in rats. J Pain 8:483-488, 2007 15. Ohtori S, Inoue G, Koshi T, Ito T, Watanabe T, Yamashita M, Yamauchi K, Suzuki M, Doya H, Moriya H, Takahashi Y, Takahashi K: Sensory innervation of lumbar vertebral bodies in rats. Spine 32:1498-14502, 2007 16. Ohtori S, Sameda H, Murata Y, Hanaoka E, Nakamura S, Takahashi Y, Yamagata M, Takahashi K: Involvement of a sensory afferent pathway in human discogenic low back pain: Comparison of upper and lower spinal nerve block effects on discogenic low back pain. Chiba Med J 84: 121-126, 2008 17. Orimo H, Hayashi Y, Fukunaga M, Sone T, Fujiwara S, Shiraki M, Kushida K, Miyamoto S, Soen S, Nishimura J, Oh-Hashi Y, Hosoi T, Gorai I, Tanaka H, Igai T, Kishimoto H: Diagnostic criteria for primary osteoporosis: Year 2000 revision. J Bone Miner Metab 19:331-337, 2001 18. Sabsovich I, Wei T, Guo TZ, Zhao R, Shi X, Li X, Yeomans DC, Klyukinov M, Kingery WS, Clark JD: Effect of anti-NGF antibodies in a rat tibia fracture model of complex regional pain syndrome type I. Pain 15:47-60, 2008 19. Sabsovich I, Guo TZ, Wei T, Zhao R, Li X, Clark DJ, Geis C, Sommer C, Kingery WS: TNF signaling contributes to the development of nociceptive sensitization in a tibia fracture model of complex regional pain syndrome type I. Pain 2007 [Epub ahead of print]. 20. Sakurai J, Obata K, Ozaki N, Tokunaga A, Kobayashi K, Yamanaka H, Dai Y, Kondo T, Miyoshi K, Sugiura Y, Matsumoto T, Miwa H, Noguchi K: Activation of
Ohtori et al extracellular signal-regulated protein kinase in sensory neurons after noxious gastric distention and its involvement in acute visceral pain in rats. Gastroenterology 134: 1094-1103, 2008 21. Suseki K, Takahashi Y, Takahashi K, Chiba T, Tanaka K, Morinaga T, Nakamura S, Moriya H: Innervation of the
875 lumbar facet joints: Origins and functions. Spine 22: 477-485, 1997 22. Taguchi T, John V, Hoheisel U, Mense S: Neuroanatomical pathway of nociception originating in a low back muscle (multifidus) in the rat. Neurosci Lett 427:22-27, 2007