DIAGNOSTIC NEURORADIOLOGY

Lumbosacral transitional vertebra and S1 radiculopathy:the value of coronal MR imaging

Abraham Fourie Bezuidenhout & Jan Willem Lotz

Received: 25 February 2014 /Accepted: 31 March 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractIntroduction The association of a lumbosacral transitional ver-tebra with accelerated degeneration of the disc above has beendescribed. Lumbosacral transitional vertebrae have also beenreported as a cause of extraforaminal entrapment of the L5nerve root between the transverse segment of the transitionalvertebra and the sacral ala optimally demonstrated by coronalMRI. The association of the lumbosacral transitional vertebrapseudoarthroses and S1 nerve root entrapment due to degener-ative stenosis of the nerve root canal has never been described.Methods We present 12 patients with lumbosacral transitionalvertebrae that were referred for symptoms and signs of S1nerve root radiculopathy in which the sagittal and axial MRIsequences failed to identify a plausible cause for the patients’S1 nerve root symptoms. A coronal T1-weighted imaging(T1WI) MRI sequence was consequently added to theinvestigation.Results The coronal T1WI MRI sequence demonstrated hy-pertrophic degenerative stenosis of the S1 nerve root canal atthe level of the lumbosacral transitional vertebrapseudoarthrosis, with entrapment of the respective S1 nerveroot in all patients.Conclusion We emphasize the value of coronal T1WIMRI ofthe lumbosacral junction and sacrum if the cause for S1radicular symptoms was not identified on conventional sagit-tal and axial MRI sequences in patients with lumbosacraltransitional vertebrae.

Keywords LSTV (lumbosacral transitional vertebra) .

S1 radiculopathy

Introduction

A lumbosacral transitional vertebra (LSTV) is an anomalousvertebra with intermediate morphology resulting from eithersacralisation of the lowest lumbar segment or lumbarisation ofthe most superior sacral segment of the spine. LSTVare com-monly found in the general population with a reported prev-alence of 5–30 % [1]. In 1984, Castellvi et al. [2] formulated aradiographic classification system describing four types ofLSTV on the basis of morphologic characteristics. Type Irepresents an enlarged, dysplastic transverse process with aheight more than 19 mm. Type IIa has a unilateralpseudoarthrosis, and IIb bilateral pseudoarthroses. Type IIIahas complete fusion on one side, and IIIb complete fusion onboth sides. Type IV has a pseudoarthrosis on one side andcomplete fusion on the other side (Fig. 1) [3]. In a large cohortof patients with LSTV, the prevalence of the subtypes wastype I in 42 %, type II in 41 %, type III in 12 %, and 5 % hadLSTV type IV [1]. The association of LSTV with accelerateddegeneration of the disc above has been described extensively[2]. Elster [4] found the overall incidence of structural pathol-ogy no higher in patients with LSTV, but emphasized theproportion of prolapses in the interspace immediately abovethe transition. LSTV has been reported as a cause forextraforaminal entrapment of the L5 nerve root betweenthe transverse segment of the LSTV and the sacral alaoptimally demonstrated by coronal MRI [5]. Abe et al. [6]described a case of anterior decompression of foraminalstenosis below a LSTV. A report on the results from three-dimensional MRI reconstructions emphasized the value ofthe coronal plane in identifying extraforaminal L5 nerveroot compression [7].

In an extensive literature search, we were unable to finddocumentation of S1 nerve root entrapment due to degenera-tive stenosis of the nerve root canal in association with LSTVpseudoarthroses. This case series will demonstrate the utility

A. F. Bezuidenhout (*) : J. W. LotzDivision of Radiodiagnosis, Department of Medical Imaging andClinical Oncology, Faculty of Medicine and Health Sciences,Stellenbosch University, PO Box 19179, Tygerberg 7505, SouthAfricae-mail: fouriebez@yahoo.com

NeuroradiologyDOI 10.1007/s00234-014-1361-z

of coronal MR imaging in evaluating for S1 nerve root en-trapment associated with Castellvi type IIa, IIb and IV LSTVs.

Materials and methods

Institutional Health Research Ethics Committee approval wasobtained.

We present 12 patients with LSTV that were referred forsymptoms and signs of S1 nerve root radiculopathy during the

time period of 1 September 2012 to 31 January 2014. Allpatient imaging was performed at the same institution. Allpatients had prior radiographs consisting of AP and lateralviews of the lumbosacral spine. Vertebral levels were deter-mined by designating the last vertebral body with a rib ele-ment as T12, which is common practice at our institution. AFerguson view [2] (30° cranial angulation) of the lumbosacraljunction was added if there was any suggestion of a LSTVonthe initial AP view (Fig. 2a). CTstudies were not performed asroutine procedure.

MRI studies were performed on a 1.5 Tesla scanner. SagittalT1-weighted imaging (T1WI) (821/13 [repetition timemsec/echo time msec]), turbo T2WI (4500/115), axial T1WI(583/15) and turbo T2WI (4810/104) of the lumbar spine withslice thickness of 4 mm were performed as standard procedure.

A coronal T1WI sequence (642 ms/14 ms) with slicethickness of 3.5 mm, sectioned parallel to the cauda equinaat L3 to L5 vertebral body levels, was constructed on a sagittaltopogram. Coronal T1WI sequences were chosen as the au-thors believe that epidural and foraminal T1 fat hyperintensityprovides an ideal background for identifying masses and disccomponents in these spaces.

All imaging studies were read by a senior consultantradiologist.

Results

LSTVs were identified and classified on a Ferguson viewradiograph of the lumbosacral junction in all cases. In onecase, a CT referral to exclude unrelated pelvic pathologyafforded us an opportunity to reconstruct a coronal maximumintensity projection image. The combination of LSTV andnerve root canal stenosis was eloquently demonstrated on thisimage (Fig. 2b). Castellvi IIa was identified in seven, type IIbin four, and type IV in one case.

ConventionalMRI scans of the lumbar spine in sagittal andaxial planes demonstrated a variety of degenerative pathologyof the lower lumbar articulations. A definitive cause for S1nerve root symptoms could, however, not be identified on anyof the sequences. A coronal T1WI MRI sequence was conse-quently added to the investigation, and in this imaging plane,hypertrophic stenosis of the S1 nerve root canal at the levelof the pseudoarthrosis, with entrapment of the respective S1nerve root, was demonstrated in all patients.

The following four patients are representative of the find-ings encountered in the case series (Table 1).

Figure 3 shows the coronal T1WI MRI of a 42-year-oldfemale patient who was referred from an orthopaedic outpa-tient clinic with unexplained left S1 nerve root symptomatol-ogy. A Ferguson view radiograph demonstrated a Castellvi IIapseudoarthrosis with markedly sclerotic margins on the left.The coronal MRI identified marked hypertrophic stenosis

Fig. 1 Castellvi classification of LSTV. Type I represents an enlarged,dysplastic transverse process with a height more than 19 mm. Type IIahas a unilateral and IIb bilateral pseudoarthroses. Type IIIa has completefusion on one side, and IIIb complete fusion on both sides. Type IV has apseudoarthrosis on one side and complete fusion on the other. Courtesy ofDaniel M. Walz, M.D. [3]

Neuroradiology

with osteophyte formation within the S1 nerve root canalresulting in entrapment and compression of the S1 nerve root.

A 54-year-old female presented with symptoms and signs ofneurogenic claudication and bilateral S1 radiculopathy, moreprofound on the left. A Castellvi IIa pseudoarthrosis withdegenerative sclerotic marginal changes was present on theleft. Sagittal and axial MRI confirmed degenerative spinalcanal stenosis at the lower intervertebral disc levels, but failedto show a definitive cause for her severe left sided S1 nerve rootpain. A coronal T1WI MRI of the lumbosacral junction iden-tified entrapment of the left S1 nerve root by hypertrophicstenosis directly below the level of the pseudoarthrosis (Fig. 4).

A 62-year-old male with incapacitating bilateral S1radiculopathy presented with a Castellvi IIb LSTV with bilat-eral hypertrophic pseudoarthroses. The coronal T1WI MRIrevealed marked bilateral S1 nerve root canal stenosis. On theright, the S1 nerve root was compressed by an osteophyte

originating from the inferior margin of the pseudoarthrosis. Asmaller osteophyte compressed the left S1 root medially at alower level (Fig. 5).

A 49-year-old female, complaining of severe right-sided S1radiculopathy, was shown to have a Castellvi IIa LSTV. Thecoronal T1WI MRI identified an osteophyte originating me-dially from a transitional disc margin directly opposite thelevel of the pseudoarthrosis. The exiting S1 nerve root wasindented and angulated by extrinsic mass effect (Fig. 6).

Discussion

In this case series, we present 12 patients referred for MRI ofthe lumbar spine for investigation of S1 nerve root symptoms.An LSTV was identified on plain film radiography in allpatients. The sagittal and axial MRI sequences failed to

Fig. 2 a Ferguson viewradiograph demonstrating aCastellvi IIa LSTV with a right-sided pseudoarthrosis (whitearrow) and marked degenerativebony hypertrophy (black arrows).b Coronal CT maximum intensityprojection image confirming theFerguson view radiographfindings of a Castellvi IIa LSTVwith a right-sided pseudoarthrosis(black arrows) and associatedstenosis of the nerve root canal(white arrows)

Table 1 Summary of patient characteristics and clinical and imaging findings

Patient number Age (years), sex LSTV type(Castellvi classification)

S1 radiculopathy(left, right, bilateral or none)

S1 nerve root compression identifiedby coronal T1W1 MRI?

1 42, female IIa left Left Yes

2 56, female IIa left Bilateral (left more than right) Yes

3 57, female IIa right Right Yes

4 62, male IIb Bilateral Yes

5 69, female IIa right Right Yes

6 50, female IIa left Left Yes

7 75, male IIb Left Yes

8 51, male IIa left Left Yes

9 49, female IIa right Right Yes

10 40, female IIb Bilateral (left more than right) Yes

11 38, female IV (pseudoarthrosis on the left) Bilateral (left more than right) Yes

12 57, female IIb Bilateral Yes

Neuroradiology

identify a plausible cause for the patients’ S1 nerve rootsymptoms in all cases.

A coronal T1WI MRI sequence was additionally performedaccording to standardized technical parameters. S1 nerve rootcanal stenosis secondary to degenerative hypertrophy at or nearthe LSTV pseudoarthrosis was identified in all patients. It wasalso possible to demonstrate S1 nerve root entrapment and tomake a definitive imaging diagnosis for unexplained S1 nerveroot symptomatology.

Our case series highlights the importance of evaluating forS1 nerve root entrapment, which is often due to hypertrophic

degenerative stenosis associated with Castellvi type IIa, IIband IV LSTVs.

Conclusion

This case series emphasizes the value of coronal T1WI MRIof the lumbosacral junction and sacrum if the cause for S1radicular symptoms was not identified on conventional sagit-tal and axial MRI sequences in patients with LSTV.

Fig. 3 Coronal T1WI MRI identifying a Castellvi IIa LSTV with a left-sided pseudoarthrosis (black arrows) and associated entrapment of the S1nerve root due to hypertrophic stenosis and osteophyte formation (whitearrow)

Fig. 4 Coronal T1WI MRI identifying a Castellvi IIa LSTV with a left-sided pseudoarthrosis (black arrows) and associated entrapment of the S1nerve root due to hypertrophic stenosis (white arrows)

Fig. 5 Coronal T1WI MRI identifying a Castellvi IIb LSTV with bilat-eral pseudoarthroses (black arrows) and associated osteophytic compres-sion of the S1 nerve roots bilateral (white arrows)

Fig. 6 Coronal T1WI MRI identifying a Castellvi IIA LSTV with aright-sided pseudoarthrosis (black arrows). An osteophyte originatingmedially from a transitional disc margin directly opposite the level ofthe pseudoarthrosis indents and angulates the exiting S1 root (whitearrow)

Neuroradiology

Ethical standards This research project was approved by the appropriateethics committee and has therefore been performed in accordance with theethical standards laid down in the 1964 Declaration of Helsinki and its lateramendments. We declare that patient consent was waived for this study.

Conflict of interest We declare that we have no conflict of interest.

References

1. Nardo L, Alizai H, Virayavanich Wet al (2012) Lumbosacral transitionalvertebrae: association with low back pain. Radiology 265(2):497–503

2. Castellvi AE, Goldstein LA, ChanDP (1984) Lumbosacral transitionalvertebrae and their relationship with lumbar extradural defects. Spine9(5):493–495

3. Konin GP, Walz DM (2010) Lumbosacral transitional vertebrae: clas-sification, imaging findings, and clinical relevance. Am J Neuroradiol31(10):1778–1786

4. Elster AD (1989) Bertolotti’s syndrome revisited. Transitional verte-brae of the lumbar spine. Spine 14(12):1373–1377

5. Hashimoto M, Watanabe O, Hirano H (1996) Extraforaminal stenosisin the lumbosacral spine. Efficacy ofMR imaging in the coronal plane.Acta Radiol 37(5):610–613

6. Abe E, Sato K, Shimada Y et al (1997) Anterior decompression offoraminal stenosis below a lumbosacral transitional vertebra. A casereport. Spine 22:823–826

7. Byun WM, Kim JW, Lee JK (2012) Differentiation betweensymptomatic and asymptomatic extraforaminal stenosis in lum-bosacral transitional vertebra: role of three-dimensional magneticresonance lumbosacral radiculography. Korean J Radiol 13(4):403–411

Neuroradiology