You are here

Article Text

Article menu

Download PDFPDF

Research paper
Segmental innervation in lumbosacral transitional vertebrae (LSTV): a comparative clinical and intraoperative EMG study
  1. Philipp Hinterdorfer1,
  2. Babak Parsaei1,
  3. Karl Stieglbauer2,
  4. Michael Sonnberger3,
  5. Johannes Fischer1,
  6. Gabriele Wurm1
  1. 1Neurosurgical Department Landesnervenklinik Wagner Jauregg, Linz, Austria
  2. 2Neurological Department, Landesnervenklinik Wagner Jauregg, Linz, Austria
  3. 3Institute of Neuroradiology Landesnervenklinik Wagner Jauregg, Linz, Austria
  1. Correspondence to Professor Gabriele Wurm, Neurosurgical Department, Landesnervenklinik Wagner Jauregg, Wagner Jauregg Weg 15, A-4020 Linz, Austria; gabriele.wurm{at}gespag.at

Abstract

Background Despite the high prevalence of lumbosacral transitional vertebrae (LSTV), little is known about the segmental innervation in this condition.

Methods The authors performed a prospective comparative clinical evaluation and an intraoperative electromyographic (EMG) investigation on patients with six lumbar vertebral bodies (6LVB) and on patients with five lumbar vertebrae (5LVB). First, clinical pain distribution in 80 patients (46 patients with 6LVB, 34 patients with 5LVB) with degenerative lumbar diseases were analysed between patient groups. Intraoperative EMG monitoring of five lower-limb muscles was performed. Compound muscle action potentials were obtained from 100 nerve roots of our 80 patients.

Results The EMG results compared fairly to the clinical findings: 40 CMAPs from 5LVB and 60 CMAPs from 6LVB patients were compared with each other within L3 to S levels. First, there was no difference between groups in the pattern of radicular pain and myotomal innervation at the level L3/4 and L4/5 (p=0.39–1.0). Second, the nerve root stimulated at the L5/6 level compares to the S1 root in 5LVB patients; the only difference was found in a coinnervation of the biceps femoris muscle that is less frequent in 6LVB patients (p=0.02). Third, the nerve root at the L6/S level corresponds to the S1 as well as to the S2 root in 5LVB patients.

Conclusion Intraoperative EMG monitoring of surgically decompressed nerve roots was found to be the ideal means of unequivocal determination of segmental innervation in LSTV patients.

  • Lumbosacral transitional vertebrae
  • segmental root innervation
  • intraoperative electromyography
  • evoked EMG mapping
  • compound muscle action potentials
  • spinal surgery
  • EMG
  • neurophysiology
  • neurosurgery

https://doi.org/10.1136/jnnp.2009.187633

Statistics from Altmetric.com

    Request Permissions

    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.

    Introduction

    Evolution, genetics, prevalence of lumbosacral transitional vertebrae

    The evolution of erect posture and bipedalism is a landmark in the morphology of human beings. The resilient S-curvature of the vertebral column is well suited for axial loading absorbing the weight of the body and transmitting it to the lower limbs.1 The key factor that maintains the erect posture is the sacrovertebral angle which produces the most abrupt change in direction in the vertebral column.2 The region around this sacrovertebral angle is prone to a high percentage of variations, anomalies and degenerative changes. Lumbosacral transitional vertebrae (LSTV), either showing an assimilation of the fifth lumbar vertebra to the sacrum (sacralisation, four lumbar vertebral bodies (LVB)) or showing transition of the first sacral vertebra to a lumbar configuration (lumbarisation, 6LVB), are the most common anomalies of the human lumbosacral region.3–5 The lumbosacral junction in LSTV patients is renamed according to the transition type; in case of sacralisation, it is called L4–S1, and in case of lumbarisation it is called L6–S.3

    Early hominids are cited as possessing 6LVB, and this lengthening of the spine was said to have facilitated the critical adoption of lumbar lordosis. A modal number of LVB in early and modern humans, nevertheless, seems to be more reasonable.2 Moreover, genetic factors are being held responsible for the development of LSTV.2 3 6–10 During embryogenesis, axial skeleton undergoes craniocaudal segmentation, resulting in so-called somites, and the identity of the somites is determined by different Hox-genes.6 10 It is assumed that changes in the axial pattern of lumbar and sacral vertebrae such as LSTV may result from mutations in the Hox-10 and Hox-11 genes.3 6 10

    Today's prevalence of LSTV reported in literature is about 12%, ranging from 4% to 37%.3–5 7 9 11–18 The condition of sacralisation has a mean prevalence of about 7.5%,3 and the condition of lumbarisation has a mean prevalence of 5.5%.3 In our population, lumbarisation is much more common and poses challenging questions concerning relation of clinical and radiological findings.

    Clinical implications of LSTV

    The clinical significance of LSTV is controversial with no consensus as to their relationship to pain or degenerative disc disease.4 11 13 14 19 Several authors have ascribed low back pain to LSTV,20–23 but the association between low back pain and LSTV has been a matter of debate3–5 9 15 16 21 23–39 since it was first described.20 The only consensus in literature seems to be today that disc herniation occurs more often at the level above the LSTV compared with patients without LSTV.3 5 16 21 23 25–27 40 This phenomenon may be due to a relative hypermobility and stress at the segment above the LSTV, and shows analogous features to advanced disc degeneration adjacent to block vertebrae or interbody fusion.3 5 27 37 41 On the other hand, due to restricted bending and rotational movements, LSTV seems to protect the segment below the LSTV from disc degeneration.5 16 19 26 37 40 In any case, the region around LSTV is a common region of pain and thus a common target of different therapeutic modalities.

    Therapeutic implications of LSTV

    The interpretation of clinical symptoms is commonly based on standard dermatomal and myotomal map.42–44 The most interesting question in LSTV is whether the function of lumbar nerve roots accompanying LSTV is equal or altered compared with patients with 5LVB. This question has major implications on clinical diagnosis and on surgical indication,3 11 13 17 26 45 46 especially when there is a discrepancy between clinical and radiographic findings.45 Previous conflicting studies on this topic mainly relied on anatomical dissection studies, animal studies, myelographic studies and palpation of muscle contraction or subjective description of sensations by the patient after root stimulation.26 43 47–53

    Caution in interpretation of clinical symptoms in LSTV patients, however, in case of symptomatic spinal diseases is of utmost importance in spinal infiltration techniques and in spinal surgery to avoid harmful patient management.3 16 18 31 45 Deciding which mode of therapy to employ depends very much upon the underlying anatomy and segmental innervation of nerve roots. Erroneous level designating of the affected nerve root is associated with the risk of operative intervention at the wrong level. Surgery at the wrong intervertebral level has been reported as a reason for failure of lumbar spine surgery18 31 54 requiring second surgery,31 especially in patients with multilevel degenerative diseases. Other possible sources of failed surgery can be conjoined nerve roots, closely adjacent nerve roots, and intra- and extradural anastomoses between nerve roots, which are also supposed to occur more often in LSTV patients.49 55

    Electrophysiological examinations thus may be particularly useful to gain greater insight in segmental innervations patterns in LSTV patients. However, to date, there has been no intraoperative electromyographic (EMG) comparative study considering segmental innervation in patients with 5LVB and patients with 6LVB.46 50 Our study was, therefore, undertaken in an attempt to prospectively assess the clinical relationship of nerve root symptoms between the presence of 5LVB and the presence of 6LVB and to verify it by means of intraoperatiove EMG monitoring. Mapping of segmental innervations by means of an objective method independent of the cooperation of the patient had the goal of increased systematisation of the segmental innervation of lower-limb muscles in the presence of LSTV. Our results are discussed in the light of the available literature. To the best of our knowledge, ours is the first report systematically and prospectively comparing direct nerve root stimulation in patients with 5LVB and patients with 6LVB.

    Patients and methods

    We prospectively compared 5LVB and 6LVB patients with nerve root symptoms caused by herniation of a lumbar disc and/or spinal canal stenosis requiring surgical decompression. From February 2001 to October 2008, 80 volunteers of a single surgeon, and thus unselected, surgical series, participated in the study. Thirty-four patients had 5LVB, and 46 patients had 6LVB. Ages ranged from 19 to 81 years (mean 49.7 years); we had 31 females and 49 males. Only patients who gave informed consent were included in the study. Patients with high-grade nerve palsy were excluded from the study, as EMG results were suspected to be unreliable in these patients. Clinical and electrophysiological data were prospectively identified in a database; data analysis was performed retrospectively by the first author, who had not been involved in patient investigation. The study was approved by the regional ethics committee; written informed consent was obtained from each patient.

    Patients were assessed in accordance with a uniform protocol. Radiographs of thoracic and lumbar vertebrae were performed in all cases to clearly define the T12 vertebra and to count the number of LVB. We determined the existence of 6LVB using Castellvi's criteria.25 The indication for decompressive surgery has been made on basis of clinical symptoms and MR and/or CT images of the lumbar spine. Clinical investigation compared distribution of pain radiating into the leg(s) of both patient groups according to standard dermatomal maps.56 Demographic and treatment-related data are presented in table 1.

    View this table:
    Table 1

    Study population

    Intraoperative monitoring requires a team approach; the interpretation of CMAPs is depending on the extent of pharmacological neuromuscular blockade. Good communication between the anaesthesia, the neurophysiology and the neurosurgical members of the team is an indispensable prerequisite for a reliable monitoring.57 Intubation was performed under short-acting muscle relaxants. Intramuscular needle electrodes were then positioned within the belly of five muscles in the anaesthetised patient (figure 1). At the surgically treated level(s), each nerve root was carefully decompressed by unroofing of the lateral recess, foraminotomy and sometimes medial facetectomy. This allowed direct visualisation and EMG stimulation of 100 decompressed nerve roots in 80 patients at various lumbar levels between L3 and the sacrum. Active stimulation protocol consisted of stimulus-evoked EMG, and recordings were monitored in the five limb muscles all at the same instant: the quadriceps femoris (Q), the biceps femoris (B), the tibialis anterior (T), the gastrocnemius (G) and the peroneus longus muscle (P) (figure 1); this pattern includes the crucial muscles for the myotomes L4 to S1 and allows coverage of the spinal segments L3 to S.56 The muscles were monitored simultaneously in order to obtain signals from single muscles or multiple signals from coinnervated muscles. All muscles were examined in all patients. The intraoperative stimulus was applied via a hand-held bipolar stimulator probe on the isolated nerve root. Bipolar stimulation provides a localised stimulation current, avoiding unwanted current spread to adjacent nerves. Current intensity started at 0.5 mA and was increased as necessary (up to 5 mA) using a monophasic square wave pulse. The stimulus-triggered EMG was performed by means of the Bravo Endeavor diagnostic device (Nicolet Biomedical), which has 16 channels and is equipped with two headboxes and an ink-jet writer.

    Figure 1
    Figure 1

    Intraoperative EMG setting. Compound motor action potentials were recorded via surface electrodes placed into the muscle belly of five lower-limb muscles: quadriceps femoris (1), biceps femoris (2), tibialis anterior (3), peroneus longus (4) and gastrocnemius (5).

    Successful monitoring could be performed under partial neuromuscular blockade with carefully titrated doses of short-acting agents in all patients. Communication and cooperation with the anaesthesiologist avoided underestimation of the degree of pharmacological neuromuscular blockage. All-or-nothing signals of compound muscle action potentials (CMAPs) were recorded and judged by both the neurologist and the neurosurgeon.

    For statistical analysis, all recorded CMAPs were entered into an Excel (Microsoft, Seattle, Washington) spreadsheet, and then listed in a contingency table. The results of the 6LVB patients were compared with the results of the 5LVB patients within segments. The SPSS program was used to study the relationship between categorical variables (SPSS, Chicago, Illinois). The Fisher exact test was used for testing; p values of less than 0.05 were considered to be statistically significant.

    Results

    Preoperative clinical findings

    Seventy-nine of our 80 patients showed pain in the dermatome of one or more spinal nerve roots (radiculopathies, RPs); one with 6LVB operated at L6/S had pain which was non-dermatomal in distribution. In sum, 100 RPs (41 in 5LVB, 59 in 6LVB) were compared between the group with 5LVB and that with 6LVB. The results have been compiled in table 2.

    View this table:
    Table 2

    Preoperative clinical findings: radiculopathy (RP)

    Patients with compressive pathologies at the L3/4 level showed 16 RPs according to the L4, L5 and/or S1 roots (11 in the 5LVB and five in the 6LVB group). In the 5LVB cohort, seven patients (64%) had RP L4, three patients (27%) had RP L5, and one patient (9%) had RP S1. In the 6LVB cohort, four patients (80%) had RP L4, and one patient (20%) had RP L5. There was no difference between groups in the pattern of radicular symptoms at the level L3/4 (p=1.0).

    At the L4/5 level, a total of 26 RPs were encountered (10 in the 5LVB and 16 in the 6LVB group). In the 5LVB cohort, seven patients (70%) had RP L5, two patients (20%) had RP L4, and one patient (10%) showed RP S1. In the 6LVB cohort, 13 patients (81%) had RP L5, two patients (13%) had RP L4, and one patient (6%) showed RP S1. p Values varied between 0.63 and 1.0; thus, there was no difference in distribution of radicular symptoms at this level.

    Among patients operated at the L5/S1 and L6/S level, respectively, 52 RPs were encountered (20 in the 5LVB group and 32 in the 6LVB group). Patients with 5LVB showed RP S1 in 18 cases (90%) and RP L5 in two cases (10%), while patients with 6LVB showed RP S1 in 22 cases (69%) and RP L5 in 10 cases (31%). Thus, we found the L6 nerve root in 6LVB patients to be highly similar to the S1 nerve root in 5LVB patients (p=0.1).

    Comparing segmental innervations in 6LVB patients operated at the segment L6/S with segmental innervations of 5LVB patients at S1/2 level was not possible, as we had no patients operated at S1/2. Patients assessed at L6/S showed a similar distribution as 5LVB patients assessed at L5/6; five of six patients showing RP at L6/S had RP S1 (83%) and one patient had RP L5 (17%). Comparing the L6/S level in 6LVB with L5/S1 in 5LVB showed that the nerve root emerging at L6/S was also similar to the S1 root emerging at L5/S1 (p=1.0). However, the pattern of our clinical investigation may have missed special innervations patterns of this nerve root; a further limitation for interpretation of these results may be that only seven patients needed operation at the L6/S level and no one at the S1/S2 level.

    Intraoperative electrophysiological findings

    Intraoperative monitoring of direct root-to-muscle innervations during scheduled lumbar spinal surgeries had no additional risks for patients. Our series revealed a good correlation between clinical and electrophysiological findings. Intraoperative evoked EMG was observed from all five muscles in every patient, and all CMAPs received were listed in tables. Nineteen of the 80 patients had multilevel procedures, so in sum, we examined 100 nerval roots in 80 patients (60 nerve roots in 46 patients with 6LVB, 40 nerve roots in 34 patients with 5LVB). Fifteen nerve roots were available for monitoring at L3/4, 26 at L4/5, 20 at L5/S1, 32 at L5/6 and seven at L6/S.

    We detected 15 CMAPs in patients stimulated at the level L3/4 (10 in the 5LVB and five in the 6LVB group), 26 CMAPs at the level L4/5 (10 in the 5LVB and 16 in the 6LVB group), 52 CMAPs at the level L5/S1/L5/6 (20 in the 5LVB and 32 in the 6LVB group) and seven CMAPs at the level L6/S (in the 6LVB group only) (table 3). Eleven different innervations patterns of muscles were found. Single muscle innervations were encountered in 44 stimulation cases (44%), coinnervation patterns in 56 stimulation cases (56%). The following innervations patterns were found.

    View this table:
    Table 3

    Intraoperative EMG findings

    Evoked EMG recording showed CMAPs of the root emerging at the L3/4 level in 15 cases (nine in the 5LVB and six in the 6LVB group). In the 5LVB group, five patients (50%) showed a response in the Q and one patient (10%) in the T only. Coinnervation of the Q and T was encountered in two patients (20%), coinnervation of T and P in one patient (10%), and coinnervation of three muscles (B,T and P) in one patient (10%). In the 6LVB group, direct root stimulation at the level provoked responses in the Q in four patients (80%) and a combination of Q and T in one patient (20%). No statistical difference could be found between groups (p=0.58–1.0) (table 3).

    Stimulation at the L4/5 level provoked responses in 26 cases among four muscles (T, G, Q, P). In the 5LVB cohort, coinnervation of T and P was encountered in seven patients (70%), whereas P showed innervations in two patients (20%) and G in one patient (10%). In the 6LVB cohort, coinnervation of T and P was observed in 10 patients (62%), innervations of P in two patients (13%) and innervation of T in two patients (13%); one patient showed innervations of Q (6%), and one of G and P (6%). p-Values between groups were between 0.39 and 1.0, and showed therefore no significant innervation differences (table 3).

    Comparison of 52 roots stimulated at the L5/S1 and L5/6 levels, respectively, showed single or coinnervation patterns with eight distribution patterns (G, P, B, G+P, T+P, B+G, Q+T, B+G+P). No patient showed innervations of Q or T only, and no patient had combined innervations of B,T and P. All other possibilities of single or multiple innervations of table 3 were encountered at that level: In the 5LVB group, the combination of B and G was recorded in nine patients (45%), in six patients (30%) CMAPs were recorded from the G, coinnervation of G and P was observed in two patients (10%), one patient showed CMAPs in the P (5%), one in Q and T (5%), and one further showed coinnervation in B, G and P (5%). In the 6LVB group, CMAPs were recorded in the G in eight patients (25%), in the P in seven patients (22%), in the G and P in six patients (19%), in T and P in six patients (19%) and in the B and G in four patients (12%), and one patient (3%) showed innervations of the B only. There were no statistical significances at the level L5/S1 versus L6/S (p=0.07–1.0) except that coinnervation of B and G occurred more often in the 5LVB group (p=0.02).

    CMAPs at the L6/S level were obtained in seven patients with 6LVB but could not be compared with any case operated at the S1/S2 level. Therefore, we compared the root emerging at L6/S with the root at L5/S1 in 5LVB. At the L6/S level, we found seven CMAPs, two (29%) in the B, two (29%) in the muscles B and G, two (29%) in the G and P and one (14%) in the G only. Comparing these results with the S1 nerve root in 5LVB patients showed no statistical difference (p=0.27–1.0), except for a trend (p=0.06) for innervations of B in the setting of 6LVB. The innervation of B without any coinnervation was detected at the level L5/6 only. Caution, however, is advised with this comparison, as no direct comparison to S1/S2 was available and as the setting of recording might have been insufficient to test the nerve root at that level.

    Discussion

    Segmental innervations of lower limbs

    Knowledge on segmental innervation is necessary as a basis of patient care and successful treatment decision-making. Today, the interpretation of clinical symptoms is based on dermatomal and myotomal maps,42–44 which are, however, still a matter of debate.17 26 56 58–61 The existence of ‘absolute innervation’ has been doubted in the literature43 49 55 59; there is growing awareness that dermatomes show variability in size and location, and especially pain and hypaesthesia have been shown to be overlapping between standard dermatomal maps.43 56 59 60 62 Concerning myotomal innervation, most muscles can be regarded as having a predominantly monosegmental innervation,17 63 64 but many studies have shown that the myotomal innervation pattern also allows considerable overlap.43 44 57–60 65 66 This implies that muscles and dermatomes may be innervated not only by axons of one spinal segment, but also partly by axons of adjacent levels. Moreover, a number of nerve root anomalies such as conjoined nerve roots or nerve root anastomoses have been documented in the lumbosacral region in clinical, radiological and anatomical series.55 67 Thus, clinical abnormalities may be attributable to either anatomical abnormalities or variations in segmental distribution of spinal nerve roots.

    Current ideas about specific root innervation of skeletal muscles and skin are mainly derived from dissection49 53 55 and animal47 62 studies, and from accumulating clinical experience.26 53 64 66 68 Considering the different approaches of studies, it is not surprising that there is considerable disagreement in the literature regarding the distribution of individual dermatomes and myotomes.58 61 65 Experimental study of the nervous tissue in life in man, however, is subject to considerable limitations.1 48 50 53 58–60 65 Therefore, it is not surprising that there are still conflicting reports on segmental innervations patterns, especially in patients with LSTV.17 26 59

    Anatomical dissection studies are limited by the fact that the functional supply of muscles cannot be established by dissection of cadaver specimens alone.49 53 55 69–71 Few of them deal with LSTV.69 70 Myelography studies49 51 on anomalies of lumbosacral nerve roots eventually also could not determine segmental distribution. Animal studies with stimulation of nerve roots were first reported in 1881,47 62 but care must be taken when animal models are used to study human conditions, as no model results in an identical situation to that found in man.1 The first adequate findings of root stimulation in man were published in 1929.48 Electrical stimulation and selective nerve root blocks have been used since then to investigate the innervation pattern of lumbar nerve roots.45 46 50 53 58 60 61 65 Opportunities, however, to investigate the effects of root stimulation of more than one root at the same individual at one time are rare.1 53 65 The variability of results regarding innervation of the lower limbs may be related in part to the patient population and in part to the method of investigation employed.46 58 59 65 68 72 The results of Phillips and Park, for example, who intraoperatively stimulated the ventral roots and recorded muscles from L2 to S2 in a study of 244 limbs before performing dorsal rhizotomy for spasticity,65 may have been influenced by spinal cord damage in their patients. Caution is advised with studies where patients are asked to describe the location of sensations or where the presence of muscular contractions is watched; such examinations of root stimulation by observation and palpation of contraction of lower-limb muscles and descriptions of patients' sensations are necessarily limited by subjective impressions.50 53 58 60 Studies on intraoperative EMG monitoring for safe spine fusion techniques73 do not report on segmental innervations patterns and have not been used up to now to include LSTV patients. Eickhoff and Helmke46 correlated neurological, EMG and radiological analysis of 12 patients with 6LVB; they found a good correlation between clinical and EMG findings; the series, however, was small and had no control group with 5LVB. The ongoing prospective study ‘Direct stimulation of spinal nerve roots to determine sensory and motor innervations patterns’ (ClinicalTrials.gov identifier: NCT00696501), which is still recruiting participants to develop a map of human innervation patterns during spine surgery, shows on the one hand that there is still significant uncertainty on this topic, even in patients with normal numbers of LVB.

    Radiology

    Accurate radiological numbering of vertebrae is a prerequisite for successful decision-making processes. Most authors use anterioposterior and lateral lumbosacral radiographs, and some combine it with a 30° angled cranially directed anterioposteror view that includes the thoracolumbar junction; the lumbar levels are then counted down from the T12 vertebra, defined as the vertebra from which the lowest rib originates.9 13 74 75 Common misleading features in correct counting of LVB are short 12th ribs and long transverse processes of the first LVB17; therefore, we recommended additional radiographs of the thoracic spine. Other authors base the diagnosis of LSTV on sagittal MRI images on the assumption that there are always seven cervical and 12 thoracic vertebrae; they use a cervicothoracic scout MRI in every patient and count caudally from C2.12 18 74 76 The height of the aorta bifurcation in relationship to the lumbar spine was found to be no valid landmark in case of LSVT.77 Hughes et al13 showed that the iliolumbar ligament on T1-weighted MRI images can be used as a landmark for LSTV as the ligament exclusively arises from the transverse process of L5. A further MRI method of determination of lumbosacral transition was proposed by Milicic et al,78 who obtained sagittal MR images of the sacrum and coccygeal bone in addition to the lumbar spine; by counting upwards from S5 they accurately identified LSTV. A consensus of determining the lumbosacral transition by radiological methods, nevertheless, is still lacking. Most authors still rely on the classification of Castellvi et al,25 which is based on form and orientation of the transverse processes on plain radiographs. We also performed radiographs of the thoracic and the lumbar spine to count the number of LVB and used the classification of Castellvi25 to determine the subtype of the LVB.

    Impact of segmental innervation in LSTV on clinical decision-making processes

    While there is little consensus on the clinical significance of LSTV, even less is known about segmental innervation in LSTV patients.3 26 46 52 The relationship between bony and nerval anomalies on segmental innervation of the lower limbs is not straightforward in the literature. Only scarce reports with small clinical series relying on subjective investigation methods are available in the literature, describing nerve root function in the case of LSTV (table 4). Precise knowledge of root innervation of the lower-limb musculature is, however, essential in diagnosis and treatment of various spinal diseases. Most radiculopathies occur at the lumbosacral level, most of them involving the L5 or S1 roots. As such, there is a particular need for objective criteria for those patients who might benefit from surgery, especially when there is a discrepancy between the clinical and the radiographic findings. The presence of LSTV can create confusion as to what level corresponds to which nerve root. The question of whether the function of lumbar nerve roots accompanying LSTV is equal or altered compared with patients with 5LVB has been discussed in conflicting studies for decades.17 26 50 52 Several authors have reported that the function of nerve roots is altered in the presence of LSTV.17 26 52 McCulloch et al suggested that the L5 nerve always originates in the ‘last mobile level’ of the lumbosacral region.17 The major shortcomings of this study, however, are that the results are based on only 11 cadavers with LSTV and a small clinical series of 15 LSTV patients undergoing electrical stimulation of L5 and S1 nerve roots during chemonucleolysis; furthermore, there was often conflict between the pattern of pain distribution and motor findings in their patient series; moreover, they mixed patients with 4LVB and patients with 6LVB in their series, which may explain the discrepancy with our results. Chang et al also revealed altered distribution of muscle weakness in case of compression of the nerve root emerging at the L5/6 level by herniated discs in patients with 6LVB compared with those with 5LVB26; this, however, was a retrospective study on 10 patients only. Seyfert examined in 1997 the dermatome gap for LSTV using the cremasteric reflex and described a variant position of the lumbosacral dermatomes in the presence of LSTV in males.52 The major shortcomings of this study are that only 10 of 50 patients had LSTV, only males were included, the number of LVB was counted on urograms, and thoracic radiographs were not available in all patients. Furthermore, this study relied on the patient's description and did not include motor responses. There were too few patients for statistical evaluation.

    View this table:
    Table 4

    Review of literature: studies in literature investigating the possibility that nerve root function may be altered in lumbosacral transitional vertebrae patients

    When these authors report a high incidence of altered segmental innervations in LSTV patients, one has to speculate that this might be due to either an insufficient number of patients or the fact that they included patients with 6LVB as well as those with 4LVB. Our findings and those of Kim et al50 differ from what has been suggested by above-mentioned authors. In the Kim et al's series, the L6 nerve root in 6LVB was found to be similar to S1 in 5LVB.50 In this small series on 12 patients with 6LVB with selective nerve root blocks and a rather subjective investigation method in five of the 12 patients, an overlap between L5 and S1 distribution was observed; a statistical analysis could not be performed because of the small number of included patients. Unfortunately, they also mixed patients with 4LVB and patients with 6LVB, but interestingly found altered nerve root function only in patients with 4LVB, which does not contradict our results.

    Above-mentioned studies (table 4) are open to certain criticisms. Subjective methods of mapping dermatomes and myotomes are unreliable. Second, mixing up patients with different numbers of LVB is difficult to accept. Finally, the low number of included patients precludes a statistical evaluation of the results. In fact, a correct diagnostic approach is fundamental in order to avoid errors in the approach to the affected nerve roots, thus guaranteeing proper management in patients with LSTV. In practice, the most serious consequence is that an operation is performed at an incorrect level. This can only be avoided by unequivocally designating the affected nerve root to the corresponding disc level. Given the high frequency of degenerative diseases involving spinal nerve roots in LSTV patients, a study that could define physiological function of the roots seemed to us to be of particular importance.

    EMG monitoring

    This is the first study in which stimulus-evoked electromyographic monitoring was used as an adjunct to clinical symptoms for prospective comparative evaluation of nerve root function in patients with 5LVB and with 6LVB. It was our aim not to rely on the subjective impressions of palpation of muscle contractions or on a description of pain and dysaesthesia by the patient, but to objectively compare CMAPs of stimulated nerve roots in a large patient series. Neither anatomical nor clinical nor radiological investigations have provided a gold standard to draw valid conclusions in segmental innervation in LSTV patients. Our results show the role of intraoperative electrophysiological monitoring by means of evoked EMG to answer critical questions about segmental innervation.

    No study systematically used evoked EMG for prospective comparison of segmental innervation of 5LVB and 6LVB patients, respectively, up to now. EMG monitoring is more sensitive than dermatomal SSEPs for detecting distinct innervation patterns.57 Direct root stimulation has the advantage of evaluating conduction in specific roots. We therefore decided to assess whether the presence of 6LVB affects segmental distribution in patients where decompressed nerve roots could easily be investigated by means of intraoperative evoked EMG. Multiple channels can be monitored simultaneously to observe stimulus-triggered CMAPs of multiple muscles at the same time. Bipolar stimulation provides a localised stimulation current, avoiding unwanted current spread to adjacent nerves.

    Our findings provide strong arguments that the function of the lumbosacral nerve roots is not significantly altered in patients with 6LVB, meaning that the L6 nerve root is equivalent to the S1 nerve root in patients with 5LVB. The only significant difference (p=0.02) was found in a coinnervation of the biceps femoris muscle that is less frequent in 6LVB patients; this difference in the rate of coinnervation reached statistical difference in the L5/6 segment (p=0.019). This higher incidence of coinnervation of muscles in patients with 5LVB is in contrast to the reported higher incidence of anatomical variations of nerve roots (conjoined nerve roots, anastomoses) in patients with LSTV.49 55 However, this coinnervation pattern has no major impact on decision-making processes for therapeutic techniques. Interestingly, the root emerging from L6/S not only primarily resembled the S1 root but also showed characteristics of the S2 nerve, with a tendency to innervation of the biceps femoris muscle (p=0.06). However, our results do not provide sufficient data to definitely classify this nerve root.

    Strengths and limitations of our study

    The major strength of this study is that prospective assessment of clinical and electrophysiological findings according to a uniform protocol has been performed on a large patient series which allowed statistical analysis of data between groups. Good correlation of clinical and EMG findings suggests reliable results. An added strength is that simultaneous recording of lower-limb muscles CMAPs directly evoked by stimulation of decompressed nerve roots helped us to overcome most shortcomings of previous studies; it provided the unique possibility to gain insight into distinct innervation patterns, thus improving the localisation power of EMG monitoring. Moreover, by placing the electrode close to the root with a threshold of <5 mA, responses could be elicitated by low amplitude current not expected to elicit activity in neighbouring nerves.58

    A number of factors may, however, have limited our results. First, our study does not provide class I evidence. Double-blinded clinical studies, however, are difficult to perform, as awareness of clinical information about the patients is generally considered to be a necessary aspect of physicians. Furthermore, muscles of asymptomatic healthy subjects are not available for prospective randomised studies; thus, grade I evidence is desirable but does not seem feasible on this topic. Second, compressive injury to motor axons can cause disruption of the normal interactions between nerve and muscles57 61; the failure to evoke a response therefore must not necessarily exclude muscle innervation, even though we excluded patients with severe paresis. Third, our series cannot yield sufficient data to permit statistical analysis of subgroups, for example within Castellvi's criteria, within ages or within other demographic parameters.25 Increasing the number of patients would allow statistical analysis between subgroups. Fourth, the LSTV group of patients with 4LVB (sacralisation) was not examined in our series. This group should be examined in a further series. Another criticism could be that the root emerging from L6/S could not be compared with the S2 root in 5LVB patients, as we had no patients with pathologies at S1/S2. Finally, our results from a personal surgical series do not directly address the question of prevalence of LSTV in our population, a question that might be addressed in a more conclusive fashion in a randomised series.

    Conclusion

    Our study provides new information with regard to the root innervation of dermatomes and of the muscles of the lower limbs in patients with anomalies of the number of LVB. We present the first prospective comparative study estimating the innervation pattern of lumbar nerve roots in patients with 6LVB using clinical signs and electrical stimulation via evoked EMG monitoring. The overall picture is that clinical pain distribution and the myotomal chart of electromyographically based patterns of muscle innervation in patients with five and in patients with 6LVB showed that the presence of lumbarisation does not significantly affect the segmental innervation. Thus, in general, patients matching the profile of 6LVB may be diagnosed and treated in the same way as those with 5LVB.

    Acknowledgments

    The authors wish to thank W Helga, Johannes Kepler University Linz, for her critical and suggestive comments on statistic evaluation of our series.

    References

    1. ↵
      1. Alini M,
      2. Eisenstein SM,
      3. Ito K,
      4. et al
      . Are animal models useful for studying human disc disorders/degeneration? Eur Spine J 2008;17:2–19.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Haeusler M,
      2. Martelli SA,
      3. Boeni T
      . Vertebrae numbers of the early hominid lumbar spine. J Hum Evol 2002;43:621–43.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Bron JL,
      2. van Royen BJ,
      3. Wuisman PI
      . The clinical significance of lumbosacral transitional anomalies. Acta Orthop Belg 2007;73:687–95.
      OpenUrlPubMed
    4. ↵
      1. Olofin MUE,
      2. Noronha C,
      3. Okanlawon A
      . Incidence of lumbosacral transitional vertebrae in low back pain patients. West African J Radiology 2001;8:1–6.
      OpenUrl
    5. ↵
      1. Vergauwen S,
      2. Parizel PM,
      3. van Breusegem L,
      4. et al
      . Distribution and incidence of degenerative spine changes in patients with a lumbo-sacral transitional vertebra. Eur Spine J 1997;6:168–72.
      OpenUrlCrossRefPubMed
    6. ↵
      1. Carapuço M,
      2. Nóvoa A,
      3. Bobola N,
      4. et al
      . Hox genes specify vertebral types in the presomitic mesoderm. Genes Dev 2005;19:2116–21.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Erken E,
      2. Ozer HT,
      3. Gulek B,
      4. et al
      . The association between cervical rib and sacralization. Spine 2002;27:1659–64.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Kessel M,
      2. Gruss P
      . Homeotic transformations of murine vertebrae and concomitant alteration of Hox codes induced by retinoic acid. Cell 1991;67:89–104.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Tini PG,
      2. Wieser C,
      3. Zinn WM
      . The transitional vertebra of the lumbosacral spine: its radiological classification, incidence, prevalence, and clinical significance. Rheumatol Rehabil 1977;16:180–5.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Wellik DM,
      2. Capecchi MR
      . Hox10 and Hox11 genes are required to globally pattern the mammalian skeleton. Science 2003;301:363–7.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Delport EG,
      2. Cucuzzella TR,
      3. Kim N,
      4. et al
      . Lumbosacral transitional vertebrae: incidence in a consecutive patient series. Pain Physician 2006;9:53–6.
      OpenUrlPubMed
    12. ↵
      1. Hahn PY,
      2. Strobel JJ,
      3. Hahn FJ
      . Verification of lumbosacral segments on MR images: identification of transitional vertebrae. Radiology 1992;182:580–1.
      OpenUrlPubMedWeb of Science
    13. ↵
      1. Hughes RJ,
      2. Saifuddin A
      . Imaging of lumbosacral transitional vertebrae. Clin Radiol 2004;59:984–91.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Hughes RJ,
      2. Saifuddin A
      . Numbering of lumbosacral transitional vertebrae on MRI: role of the iliolumbar ligaments. Am J Roentgenol 2006;59:984–91.
      OpenUrl
    15. ↵
      1. Hsieh CY,
      2. Vanderford JD,
      3. Moreau SR,
      4. et al
      . Lumbosacral transitional segments: classification, prevalence, and effect on disk height. J Manipulative Physiol Ther 2000;23:483–9.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Luoma K,
      2. Vehmas T,
      3. Raininko R,
      4. et al
      . Lumbosacral transitional vertebra: relation to disc degeneration and low back pain. Spine 2004;29:200–5.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. McCulloch JA,
      2. Waddell G
      . Variation of the lumbosacral myotomes with bony segmental anomalies. J Bone Joint Surg Br 1980;62:475–80.
      OpenUrlPubMed
    18. ↵
      1. Peh WC,
      2. Siu TH,
      3. Chan JH
      . Determining the lumbar vertebral segments on magnetic resonance imaging. Spine 1999;24:1852–5.
      OpenUrlCrossRefPubMed
    19. ↵
      1. MacLean JG,
      2. Tucker JK,
      3. Latham JB
      . Radiographic appearances in lumbar disc prolapse. J Bone Joint Surg 1990;72:917–20.
      OpenUrl
    20. ↵
      1. Bertolotti M
      . Contributo alla conoscenza dei vizi di differenzazione regionale del rachide con speciale riguardo all' assimiliazione sacrale della V lombare. Radiol Med 1917;4:113–44.
      OpenUrl
    21. ↵
      1. Elster AD
      . Bertolotti's syndrome revisited. Transitional vertebrae of the lumbar spine. Spine 1989;14:1373–7.
      OpenUrlPubMedWeb of Science
    22. ↵
      1. Magora A,
      2. Schwartz A
      . Relation between the low back pain syndrome and x-ray findings. 2. Transitional vertebra (mainly sacralization). Scand J Rehabil Med 1978;10:135–45.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Otani K,
      2. Konno S,
      3. Kikuchi S
      . Lumbosacral transitional vertebrae and nerve-root symptoms. J Bone Joint Surg Br 2001;83:1137–40.
      OpenUrlCrossRefPubMed
    24. ↵
      1. Batt ME,
      2. Skattum N,
      3. Chong BK,
      4. et al
      . Posterior element pain in a adolescent schoolgirl. Br J Sports Med 1996;30:356–8.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Castellvi AE,
      2. Goldstein LA,
      3. Chan DPK
      . Lumbosacral transitional vertebrae and their relationship with lumbar extradural defects. Spine 1984;9:493–5.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Chang HS,
      2. Nakagawa H
      . Altered function of lumbar nerve roots in patients with transitional lumbosacral vertebrae. Spine 2004:29:1632–5.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Connolly LP,
      2. Drubach LA,
      3. Connolly SA,
      4. et al
      . Young athletes with low back pain: skeletal scintigraphy of conditions other than pars interarticularis stress. Clin Nucl Med 2004;29:689–93.
      OpenUrlCrossRefPubMed
    28. ↵
      1. Dai L
      . Lumbosacral transitional vertebrae and low-back pain. Bull Hosp Jt Dis 1999;58:191–3.
      OpenUrlPubMed
    29. ↵
      1. Endo K,
      2. Ito K,
      3. Ichimaru K,
      4. et al
      . A case of severe low back pain associated with Richard disease (lumbosacral transitional vertebra). Minim Invasive Neurosurg 2004;47:253–5.
      OpenUrlCrossRefPubMed
    30. ↵
      1. Frymoyer JW,
      2. Newberg A,
      3. Pope MH,
      4. et al
      . Spine radiographs in patients with low-back pain an epidemiological study in men. J Bone Joint Surg Am 1984;66:1048–55.
      OpenUrlPubMed
    31. ↵
      1. Malanga GA,
      2. Cooke PM
      . Segmental anomaly leading to wrong level disc surgery in cauda equina syndrome. Pain Physician 2004;7:107–10.
      OpenUrlPubMed
    32. ↵
      1. Mann DC,
      2. Keene JS,
      3. Drummond DS
      . Unusual causes of back pain in athletes. J Spinal Disord 1991;4:337–43.
      OpenUrlPubMedWeb of Science
    33. ↵
      1. Peterson CK,
      2. Bolton J,
      3. Hsu W,
      4. et al
      . A cross-sectional study comparing pain and disability levels in patients with low back pain with and without transitional lumbosacral vertebrae. J Manipulative Physiol Ther 2005;28:570–4.
      OpenUrlCrossRefPubMed
    34. ↵
      1. Quinlan JF,
      2. Duke D,
      3. Eustace S
      . Bertolotti's syndrome. A cause of back pain in young people. J Bone Joint Surg Br 2006;88:1183–6.
      OpenUrlCrossRefPubMed
    35. ↵
      1. Santiago FR,
      2. Milena GL,
      3. Herrera RO,
      4. et al
      . Morphometry of the lower lumbar vertebrae in patients with and without low back pain. Eur Spine J 2001;10:228–33.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Steinberg EL,
      2. Luger E,
      3. Arbel R,
      4. et al
      . A comparative roentgenographic analysis of the lumbar spine in male army recruits with and without lower back pain. Clin Radiol 2003;58:985–9.
      OpenUrlCrossRefPubMed
    37. ↵
      1. Taskaynatan MA,
      2. Izci Y,
      3. Ozgul A,
      4. et al
      . Clinical significance of congenital malformations in young male population with prolonged low back pain. Spine 2005;30:210–13.
      OpenUrlCrossRef
    38. ↵
      1. Tertti MO,
      2. Salminen JJ,
      3. Paajanen HE,
      4. et al
      . Low-back pain and disk degeneration in children: a case–control MR imaging study. Radiology 1991;180:503–7.
      OpenUrlPubMedWeb of Science
    39. ↵
      1. van Tulder MW,
      2. Assendelft WJ,
      3. Koes BW,
      4. et al
      . Spinal radiographic findings and nonspecific low back pain A systematic review of observational studies. Spine 1997;22:427–34.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Aihara T,
      2. Takahashi K,
      3. Ogasawara A,
      4. et al
      . Intervertebral disc degeneration associated with lumbosacral transitional vertebrae: a clinical and anatomical study. J Bone Joint Surg Br 2005;87:687–91.
      OpenUrlCrossRefPubMed
    41. ↵
      1. Gay RE,
      2. Ilharreborde B,
      3. Zhao K,
      4. et al
      . Sagittal plane motion in the human lumbar spine: comparison of the in vitro quasistatic neutral zone and dynamic motion parameters. Clin Biomech (Bristol, Avon) 2006;21:914–19.
      OpenUrlCrossRefPubMed
    42. ↵
      1. Head H,
      2. Campbell AW,
      3. Kennedy PG
      . The pathology of Herpes Zoster and its bearing on sensory localisation. Rev Med Virol 1997;7:131–43.
      OpenUrlCrossRefPubMed
    43. ↵
      1. Falconer MA,
      2. Glasgow GL,
      3. Cole DS
      . Sensory disturbances occurring in sciatica due to intervertebral disc protrusions: some observations on the fifth lumbar and first sacral dermatomes. J Neurol Neurosurg Psychiatry 1947;10:72–84.
      OpenUrlFREE Full Text
    44. ↵
      1. Foerster O
      . The dermatomes in man. Brain 1933;56:1–39.
      OpenUrlFREE Full Text
    45. ↵
      1. Chiodo A,
      2. Haig AJ,
      3. Yamakawa KS,
      4. et al
      . Magnetic resonance imaging vs. electrodiagnostic root compromise in lumbar spinal stenosis: a masked controlled study. Am J Phys Med Rehabil 2008;87:789–97.
      OpenUrlCrossRefPubMed
    46. ↵
      1. Eickhoff W,
      2. Helmke K
      . Electromyographic and neurologic findings in radiologically demonstrated numerical variations in the lumbosacral transition. Neurochirurgia (Stuttg) 1986;29:189–92.
      OpenUrlPubMed
    47. ↵
      1. Ferrier D,
      2. Yeo GF
      . The functional relations of the motor roots of the brachial and lumbo-sacral plexuses. Proc Roy Soc 1881;32:12–20.
      OpenUrlCrossRef
    48. ↵
      1. Bumke O,
      2. Foerster O
      . Handbuch der neurologie, ergänzungsband 2. Berlin: Springer, 1929;960.
    49. ↵
      1. Kadish LJ,
      2. Simmons EH
      . Anomalies of the lumbosacral nerve roots. An anatomical investigation and myelographic study. J Bone Joint Surg Br 1984;66:411–16.
      OpenUrlPubMed
    50. ↵
      1. Kim YH,
      2. Lee PB,
      3. Lee CJ,
      4. et al
      . Dermatome variation of lumbosacral nerve roots in patients with transitional lumbosacral vertebrae. Anesth Analg 2008;106:1279–83.
      OpenUrlCrossRefPubMed
    51. ↵
      1. Postacchini F,
      2. Urso S,
      3. Ferro L
      . Lumbosacral nerve-root anomalies. J Bone Joint Surg Am 1982;64:721–29.
      OpenUrlPubMed
    52. ↵
      1. Seyfert S
      . Dermatome variations in patients with transitional vertebrae. J Neurol Neurosurg Psychiatry 1997;63:801–3.
      OpenUrlAbstract/FREE Full Text
    53. ↵
      1. Sharrard WJ
      . The segmental innervation of the lower limb muscles in man. Ann R Coll Surg Engl 1964;35:106–22.
      OpenUrlPubMed
    54. ↵
      1. Wiese M,
      2. Krämer J,
      3. Bernsmann K,
      4. et al
      . The related outcome and complication rate in primary lumbar microscopic disc surgery depending on the surgeon's experience: comparative studies. Spine J 2004;4:550–6.
      OpenUrlCrossRefPubMed
    55. ↵
      1. Chotigavanich C,
      2. Sawangnatra S
      . Anomalies of the lumbosacral nerve roots. An anatomic investigation. Clin Orthop Relat Res 1992;278:46–50.
      OpenUrlPubMed
    56. ↵
      1. Last RJ
      . Innervation of the limbs. J Bone Joint Surg Am 1949;31:452–64.
      OpenUrl
    57. ↵
      1. Holland NR
      . Intraoperative electromyography. J Clin Neurophysiol 2002;19:444–53.
      OpenUrlCrossRefPubMed
    58. ↵
      1. Liguori R,
      2. Krarup C,
      3. Trojaborg W
      . Determination of the segmental sensory and motor innervation of the lumbosacral spinal nerves. An electrophysiological study. Brain 1992;115:915–34.
      OpenUrlAbstract/FREE Full Text
    59. ↵
      1. Young A,
      2. Getty J,
      3. Jackson A,
      4. et al
      . Variations in the pattern of muscle innervation by the L5 and S1 nerve roots. Spine 1983;8:616–24.
      OpenUrlCrossRefPubMedWeb of Science
    60. ↵
      1. Wolff AP,
      2. Groen GJ,
      3. Crul BJ
      . Diagnostic lumbosacral segmental nerve blocks with local anesthetics: a prospective double-blind study on the variability and interpretation of segmental effects. Reg Anesth Pain Med 2001;26:147–55.
      OpenUrlCrossRefPubMedWeb of Science
    61. ↵
      1. Wolff AP,
      2. Groen GJ,
      3. Wilder-Smith OH,
      4. et al
      . Do diagnostic segmental nerve root blocks in chronic low back pain patients with radiation to the leg lack distinct sensory effects? A preliminary study. Br J Anaesth 2006;96:253–8. (orign. Wolff ap 2006).
      OpenUrlAbstract/FREE Full Text
    62. ↵
      1. Sherrington CS
      . Notes on the arrangement of some motor fibers in the lumbo-sacral plexus. J Physiol 1892;3:621–772.
      OpenUrl
    63. ↵
      1. Thage O
      . The myotomes L2–S2 in man. Acta Neurol Scand Suppl 1965;13:241–3.
      OpenUrlPubMed
    64. ↵
      1. Tsao BE,
      2. Levin KH,
      3. Bodner RA
      . Comparison of surgical and electrodiagnostic findings in single root lumbosacral radiculopathies. Muscle Nerve 2003;27:60–4.
      OpenUrlCrossRefPubMed
    65. ↵
      1. Phillips LH 2nd,
      2. Park TS
      . Electrophysiologic mapping of the segmental anatomy of the muscles of the lower extremity. Muscle Nerve 1991;14:1213–18.
      OpenUrlCrossRefPubMedWeb of Science
    66. ↵
      1. Voermans NC,
      2. Koetsveld AC,
      3. Zwarts MJ
      . Segmental overlap: foot drop in S1 radiculopathy. Acta Neurochir 2006;148:809–13.
      OpenUrlCrossRefPubMed
    67. ↵
      1. Davidson D,
      2. Rowan R,
      3. Reilly C
      . Lumbosacral nerve root anomaly associated with spondylolisthesis in an adolescent: a case report and review of the literature. Spine 2006;31:718–21.
      OpenUrlCrossRef
    68. ↵
      1. Johnson EW,
      2. Fletcher FR
      . Lumbosacral radioculopathy: review of 100 consecutive cases. Arch Phys Med Rehabil 1981;62:321–3.
      OpenUrlPubMed
    69. ↵
      1. Bardeen CR,
      2. Elting A
      . A statistical study of the variations in the formation and position of the lumbo-sacral plexus in man. Anat Anz 1901;19:124–135 209–38.
      OpenUrl
    70. ↵
      1. Eisler P
      , Der Plexus lumbosacralis des Menschen. Halle; M Niemeyer, 1892.
    71. ↵
      1. Kikuchi S,
      2. Hasue M,
      3. Nishiyama K,
      4. et al
      . Anatomic and clinical studies of radicular symptoms. Spine 1984;9:23–30.
      OpenUrlCrossRefPubMed
    72. ↵
      1. Shea PA,
      2. Woods WW,
      3. Werden DH
      . Electromyography in diagnosis of nerve root compression syndrome. Arch Neurol Psychiatry 1950;64:93–104.
      OpenUrlCrossRefPubMed
    73. ↵
      1. Bindal RK,
      2. Ghosh S
      . Intraoperative electromyography monitoring in minimally invasive transforaminal lumbar interbody fusion. J Neurosurg Spine 2007;6:126–32.
      OpenUrlCrossRefPubMed
    74. ↵
      1. Cimen M,
      2. Elden H
      . Numerical variations in human vertebral column: a case report. Okajimas Foilia Anat Jpn 1999;75:297–303.
      OpenUrl
    75. ↵
      1. Ebraheim NA,
      2. Xu R
      . Assessment of lumbosacral fusion mass by angled radiography. Technical notes. Spine 1998;23:842–3.
      OpenUrlCrossRefPubMed
    76. ↵
      1. O'Driscoll CM,
      2. Irwin A,
      3. Saifuddin A
      . Variations in morphology of the lumbosacral junction on sagittal MRI: correlation with plain radiography. Skeletal Radiol 1996;25:225–30.
      OpenUrlCrossRefPubMedWeb of Science
    77. ↵
      1. Chithriki M,
      2. Jaibaji M,
      3. Steele RD
      . The anatomical relationship of the aortic bifurcation to the lumbar vertebrae: a MRI study. Surg Radiol Anat 2002;24:308–12.
      OpenUrlCrossRefPubMed
    78. ↵
      1. Milicić G,
      2. Krolo I,
      3. Vrdoljak J,
      4. et al
      . Using magnetic resonance imaging to identify the lumbosacral segment in children. Coll Antropol 2006;30:55–8.
      OpenUrlPubMed

    Footnotes

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Ethics approval was provided by the Local Ethics Committee at the hospital Landesnervenklinik Wagner-Jauregg Linz, Austria.

    • Provenance and peer review Not commissioned; externally peer reviewed.