Skip to main content
Radiology Case Reports logoLink to Radiology Case Reports
. 2026 Sep 26;21(12):6298–6304. doi: 10.1016/j.radcr.2026.09.002

Conjoined lumbar nerve roots: A potential cause of lumbosciatica and an incidental finding—A three-case report

Cherif Mohamadou Aidara a, Magatte Gaye Sakho b, Tahibou Balia Diallo a, Ibrahima Faye a,⁎, Sokhna Astou Gawane Thiam c, Abdoulaye Ndoye Diop d, Sokhna Ba Diop d
PMCID: PMC13635636  PMID: 42835319

Abstract

Congenital anomalies of the lumbar nerve roots are uncommon within the broader spectrum of congenital vertebromedullary malformations. They are typically identified in adulthood. Often underrecognized, they may lead to inadvertent nerve injury during surgery. These anomalies can be discovered incidentally or during lumbosciatica imaging investigation. In our region, CT scan is the only available cross-sectional imaging modality. Although it has lower resolution than MRI, it enabled the detection of this anomaly in three patients. In one case, the anomaly was revealed by lumbosciatica, whereas in the other two cases, it was an incidental finding. In this article, we highlight the key imaging features necessary for accurate recognition of this underreported anomaly. Its prevalence is likely underestimated in preoperative imaging. It should be systematically considered among the differential diagnoses of enlarged nerve roots to ensure it is not overlooked. Its presence should be explicitly reported in the preoperative assessment, as failure to recognize the anomaly may result in nerve injury during surgery. Classification systems for lumbar nerve root anomalies have evolved substantially over the past decades. Among the various classification systems proposed, that of Neidre and MacNab (NM), subsequently amended on two occasions, remains one of the most widely used. More recently, Nikolenko et al. proposed a unifying classification system aimed at reconciling and integrating the various existing approaches to spinal nerve root anomalies.

Keywords: Conjoined lumbar nerve roots, Lumbosciatica, Incidental finding, Computed tomography, Classification

Introduction

Vertebromedullary malformations are complex and most commonly encountered in the pediatric population. Those involving the lumbar nerve roots are rare, scarce, likely underrecognized, and typically identified in adulthood. They are often discovered intraoperatively when imaging fails to establish the diagnosis. Cross-sectional imaging—computed tomography (CT) or magnetic resonance imaging (MRI)—is an essential step in the evaluation of lumboradicular pain. Its aims include, among others, identifying the affected nerve root(s) and determining the underlying cause(s) of radicular impairment. While extrinsic factors affecting the nerve root — such as intervertebral discal, ligamentous or facet joint abnormalities — are the most common causes, intrinsic factors related to the nerve root itself are rare. Among these, conjoined nerve roots are an uncommon etiology. In this paper, we discuss the clinical manifestations and imaging features of conjoined nerve roots. We focus particularly on radioanatomical correlations and the Neidre and MacNab (NM) classification as a practical framework for characterizing these anomalies.

Case 1

Mr. I.D., a 35-year-old man with no significant past medical history, working as a market porter, was referred for computed tomography (CT) imaging to evaluate bilateral L5 lumbosciatica. The pain was consistent, intermittent, and exertion-related (load carrying) and progressively resolved with rest. The symptoms had been evolving for several months without significant improvement despite medical treatment. Neurological examination revealed no deficit, with preserved motor and sensory function and intact reflexes. Imaging revealed a diffuse and moderate L5–S1 disc bulge, along with a conjoined origin of the L5 and S1 nerve roots. The foraminal course of the L5 nerve roots was confined to the discoarticular corridor (Fig. 1). The lesion was classified as type 1b according to the NM classification.

Fig. 1.

Fig 1 dummy alt text

(A) Coronal view showing the bilateral common emergence called corner sign (star) of the L5 (white arrowhead) and S1 (red arrowhead) nerves and the fat crescent sign (arrow). (B) Parasagittal view showing the L5 nerve root attached to the S1 nerve root below (sagittal shoulder sign). While the L4 nerve root remains unaffected. (C) axial view demonstrating the entire foraminal course of the L5 nerve roots (parallel sign). (D) oblique coronal view showing the conjoint emergence of the L5 (white arrowhead) and S1 (red arrowhead) nerve roots on the left.

Initial management was conservative, consisting of avoidance of heavy lifting, nonsteroidal anti-inflammatory and analgesic therapy, followed by epidural steroid injection. Surgical intervention could be considered in the event of failure of conservative treatment.

Case 2

Mr. M.L.D., a 41-year-old man, was referred for whole-spine computed tomography (CT) to evaluate paresthesia affecting all four limbs. There was no motor deficit, and the deep tendon reflexes were preserved. The patient had no significant medical history or fever, and his general condition was otherwise unremarkable.

Imaging revealed a moderately narrowed lumbar spinal canal with right-sided nerve root conflict at L3. Bilateral conjoined origin of the S1 and S2 nerve roots was an incidental finding, unrelated to the patient’s clinical presentation (Fig. 2). The anomaly was classified as type 1b (NM).

Fig. 2.

Fig 2 dummy alt text

(A) Axial view showing a bilateral corner sign (arrow). (B) Axial view showing a parallel sign on the left S1 nerve (arrowhead).

Following CT examination of the spine, no definite organic cause for the patient’s symptoms was identified. Brain MRI and spinal MRI were requested but were unavailable in our department.

Case 3

Mr. D.M.S., a 74-year-old man, was referred for lumbar computed tomography (CT) as part of the diagnostic workup for non-febrile bilateral paraparesis, with muscle strength graded at 3–4/5, evolving over six months. The sensorimotor reflexes were significantly impaired. There was no sphincter dysfunction.

Imaging revealed a biconcave lumbar vertebral compression without evidence of thoracic or lumbar spinal cord compression. A left-sided conjoined L5–S1 nerve root was incidentally identified, with no sign of compression (Fig. 3). The anomaly was classified as type 1b (NM).

Fig. 3.

Fig 3 dummy alt text

(A) Axial view showing a corner sign on the left (arrow). (B) Axial view showing the left S1 nerve root along almost its entire foraminal course, corresponding to the parallel sign (arrow).

Following the CT examination, the cause of the paraparesis remained unclear. MRI was requested but was not available in our region.

Comments

Clinical features

Common low back pain is a frequent condition that may require cross-sectional imaging when symptoms are resistant to medical treatment or when clear radiculopathy is present. Once the underlying causes of pain—such as disc herniation, facet joint osteoarthritis, or lumbar spinal stenosis—have been identified, an appropriate therapeutic strategy can be planned objectively. A conjoined nerve root may itself be responsible for radicular pain. It may also be associated with other congenital abnormalities, including facet aplasia, spina bifida, or spondylolisthesis. In most cases, conjoined nerve roots arise from a common dural sleeve, sometimes resulting in distortion of the normal radicular anatomy. From a radiologic perspective, this entity should be considered whenever an enlarged or unusually positioned nerve root is identified, ensuring its inclusion in the differential diagnosis [[1], [2], [3]].

Imaging features

Cross-sectional imaging, including computed tomography (CT) and magnetic resonance imaging (MRI), is now widely available for establishing the preoperative diagnosis. CT may provide valuable diagnostic information, particularly when MRI is unavailable. However, MRI remains the most sensitive and accurate imaging technique for evaluating conjoined nerve roots [[4], [5], [6]]. Prior to the advent of cross-sectional imaging, myelography using water-soluble contrast agents was able to demonstrate this anomaly [7]. The reported incidence of lumbar conjoined nerve roots varies considerably. Cadaveric studies have described a prevalence of up to 14%, whereas myelography detects these anomalies in approximately 4% of cases, and MRI studies report a prevalence of about 6% [6,8,9]. Artico et al. reported a frequency of 0.25% among 1200 lumbar CT and MRI examinations [4]. We found a frequency of 0.21% among 1438 lumbar and thoracolumbar CT examinations performed between January 1, 2023, and December 31, 2025. These findings suggest that this anomaly may be underdiagnosed on imaging.

Conjoined nerve roots are most often discovered intraoperatively. Failure to recognize this anomaly preoperatively may result in unsuccessful surgery or iatrogenic nerve injury [3,5,8,9].

Radiological anatomy review

At each lumbar intervertebral foramen, a corresponding spinal nerve exits the spinal canal [[10], [11], [12]]. Normally, each lumbar nerve root arises from the dural sac above the corresponding intervertebral foramen and follows a smooth lateral course within its meningeal sheath, without focal outpouching of the dural sac. Consequently, the dural sac narrows smoothly at each nerve root origin, without any visible focal outpouching. In contrast, the common origin of two nerve roots produces a focal outpouching of the dural sac and corresponds to a type 1b anomaly in the NM classification system. This outpouching at the site of conjoined nerve root emergence is called “corner sign” (Figs. 1A, 2A, 3A) [6,13].

The nerve root courses within the vertebral canal before extending laterally and exiting through the intervertebral foramen. Normally, the exiting lumbar nerve root courses laterally from the dural sac within its meningeal sheath, with no intervening neural structure as illustrated in Fig. 1D. In the presence of a conjoined nerve root, particularly a type 1B anomaly, a nerve structure may be identified on the sagittal plane (Fig. 1B). Kang et al. described this finding as the “sagittal shoulder sign” [14]. This sign is also demonstrated on the oblique reconstruction shown in Fig. 1D Conversely, the apparent absence of a nerve root at its expected foraminal level may reflect the common emergence of two nerve roots through the same foramen. This anomaly corresponds to type 2 of the NM classification.

From its dural origin to its exit through the intervertebral foramen, the nerve root follows an oblique course inferolaterally. Within the foramen, it lies closely along the inferior and medial aspects of the pedicle. This anatomical relationship is particularly well demonstrated on coronal CT images (Figs. 4A and 4C). Several authors highlighted the usefulness of coronal MRI sequences for depicting the foraminal course of the nerve root [4,15]. These anatomical features have two important imaging consequences:

  • −

    First, axial images obtained at the disc level do not allow complete visualization of the nerve root along its entire foraminal course (Fig. 4B). Consequently, in type 1B conjoined nerve roots, the nerve may be visualized along a substantial portion of its foraminal course, producing the so-called “parallel sign” (Figs. 1C and 3B).

  • −

    Second, there is normally little or no significant fat between the pedicle and the nerve root. The low position of the nerve root at its foraminal emergence therefore results in a characteristic crescent-shaped area of epidural fat around the root, giving rise to the so-called “fat crescent sign” (Fig. 1D). The extension of the fat beneath the pedicle is relatively easy to recognize.

Fig. 4.

Fig 4 dummy alt text

(A) Coronal view showing a normal subpedicular arrangement of independent L5 (blue) and S1 (yellow) nerve roots. (B) Axial disc view showing the partially visible L5 nerve roots. (C) Oblique reconstruction through the dural sac and L5 nerve root, demonstrating the normal recess between the detached L5 nerve root and the dural sac (yellow circle), with no abnormal outpouching or associated structural abnormality. Note the oblique course of the nerve root beneath and close to the pedicle (red circle).

Under normal circumstances, the nerve root is relatively protected beneath the pedicle and is less exposed to mechanical constraints from the disc and facet joints. In conjoined nerve roots, the low position of the root may increase its susceptibility to dynamic mechanical stress, particularly during heavy lifting. This mechanism may have contributed to the symptoms observed in Case 1.

The corner sign was bilateral in Cases 1 and 2, reflecting the bilateral nature of the anomaly, and unilateral in Case 3. The fat crescent sign was identified in one patient and the sagittal shoulder sign in one patient. The parallel sign was observed in all cases.

Our cases share several similarities with those reported by Artico et al., including involvement of the lumbosacral nerve roots and the use of CT in one of their cases [4]. In two of our cases, the conjoined nerve root was an incidental finding. We did not identify any associated osseous abnormalities in our patients such as spina bifida and partial neural arch defects,

Careful analysis of CT examination may be sufficient to establish the diagnosis when characteristic imaging features are clearly demonstrated. However, when available, MRI should be performed to provide a more comprehensive assessment. In our low-resource country, where MRI is not yet available, meticulous evaluation of CT images is therefore essential for detecting these anomalies and establishing the most accurate possible diagnosis.

In summary, several imaging findings should raise suspicion for a conjoined nerve root, notably the parallel sign, corner sign, fat crescent sign, and sagittal shoulder sign, all of which can be identified on both CT and MRI [9,13,14,16].

Once identified, the anomaly should be classified according to the appropriate classification system.

Classification systems

Several classification systems have been proposed for lumbosacral nerve root anomalies, including those of Cannon et al. (1962), Postacchini et al. (1982), Neidre and MacNab (1983), Kadish and Simmons (1984) [[17], [18], [19]]. The NM classification subsequently became one of the most widely cited and used classification systems and was later amended by Burke et al., who introduced a fourth category for confluent nerve roots. Yu et al. subsequently proposed a further amendment describing an additional variant of confluent nerve roots [20,21]. Nikolenko et al. proposed a more recent unifying classification in 2021 based on anatomical location and the type of deformity. This latter system distinguishes four major categories: intradural anastomotic anomalies, intradural spacing anomalies, extradural anastomotic anomalies, and extradural spacing anomalies. However, because this framework relies in part on previously established classification systems to characterize individual anomalies, we believe that its autonomy is somewhat limited [22]. Figure 5 illustrates the NM classification (Types I–III), including the Type IV amendment proposed by Burke et al., which is briefly reviewed below.

Fig. 5.

Fig 5 dummy alt text

Illustration of the Neidre and MacNab classification, including the Type 4 variant introduced by Burke et al [20].

Neidre and Macnab classification [1,6,20]

  • •
    Type 1: Conjoined nerve roots
    • ○
      Type 1a: Two nerve roots emerge from a common dural sleeve.
    • ○
      Type 1b: Two nerve roots arise in close proximity within the dural sac but do not share a common sleeve.
  • •

    Type 2: Redundant nerve roots

    Two distinct nerve roots exit through a single neural foramen.
    • ○
      Type 2a: One neural foramen remains devoid of a nerve root.
    • ○
      Type 2b: All foramina contain one or more nerve roots, with an additional root present in one foramen.
  • •

    Type 3: Anastomotic nerve roots

    Communication (anastomosis) between adjacent nerve roots through a vertical or transverse connecting segment.

  • •

    Type 4: Confluent nerve roots

    Two adjacent nerve root contributions merge within the dural sac to form a single nerve root exiting through the neural foramen.

Conjoined nerve roots are more frequently encountered at the S1 level than at L5, and their preoperative radiologic diagnosis remains uncommon [4,8]. This underscores the importance of considering this entity in routine radiologic practice as a potential cause of an enlarged nerve root, alongside other differential diagnoses such as Tarlov cysts, nerve root edema or irritation, and schwannomas.

The literature suggests that this anomaly is frequently underrecognized. Radiologists should therefore be familiar with its characteristic imaging features, particularly during the preoperative assessment of patients undergoing lumbar disc surgery.

The successive modifications proposed to existing classification systems reflect the ongoing difficulty in establishing a universally accepted consensus. Such a consensus would nevertheless be valuable for ensuring greater consistency in the description, reporting, and scientific communication of conjoined nerve root anomalies.

Conclusion

Conjoined nerve roots are congenital anatomical variants of the lumbosacral nerve roots that are often diagnosed later in adulthood. They may be symptomatic because their anomalous configuration reduces the available space for nerve root passage and expansion, or they may coexist with more common degenerative conditions, such as disc herniation. This frequently underdiagnosed anomaly may occur in isolation or in association with other, more readily recognizable abnormalities that may divert the radiologist’s attention. Careful evaluation of the nerve roots, particularly on multiplanar CT and MRI, is therefore essential. Recognition and explicit reporting of this condition are important to ensure that the surgeon is adequately informed and to reduce the risk of intraoperative nerve root injury.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used OpenAI’s Chatgpt to assist in translating content. The authors take full responsibility for the entire content of the publication.

Patient consent

Written informed consent for the publication of these cases reports was obtained from the patients.

Footnotes

Competing Interests: The authors have declared that no competing interests exist.

References

  • 1.Kuroki H., Nagai T. Conjoined nerve root in a patient with lumbar disc herniation accompanied by a lumbosacral spine anomaly: a case report. J Med Case Rep. 2023;17(1):82. doi: 10.1186/s13256-022-03749-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Takagi Y., Yamada H., Ebara H., Hayashi H., Inatani H., Toyooka K., et al. Conjoined lumbosacral nerve root: a case report. J Med Case Rep. 2024;18:91. doi: 10.1186/s13256-024-04415-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Yoshioka S., Sairyo K., Sakai T., Yasui N. Congenital absence of lumbosacral articular facet joint associated with conjoined nerve root: a case report. J Orthopaed Traumatol. 2010;11:183–187. doi: 10.1007/s10195-010-0100-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Artico M., Carloia S., Piacentini M., Ferretti G., Dazzi M., Franchitto S., et al. Conjoined lumbosacral nerve roots: observations on three cases and review of the literature. Neurocirugía. 2006;17:54–59. doi: 10.1016/S1130-1473(06)70370-0. [DOI] [PubMed] [Google Scholar]
  • 5.Böttcher J., Petrovitch A., Sörös P., Malich A., Hussein S., Kaiser W.A. Conjoined lumbosacral nerve roots: current aspects of diagnosis. Eur Spine J. 2004;13:147–151. doi: 10.1007/s00586-003-0634-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Neidre A., MacNab I. Anomalies of the lumbosacral nerve roots. Review of 16 cases and classification. Spine. 1983;8(3):294–299. doi: 10.1097/00007632-198304000-00010. [DOI] [PubMed] [Google Scholar]
  • 7.McCormick C.C. Developmental asymmetry of roots of the cauda equina at metrizamide myelography: report of seven cases with a review of the literature. Clin Radiol. 1982;33:427–434. doi: 10.1016/s0009-9260(82)80312-7. [DOI] [PubMed] [Google Scholar]
  • 8.Lotan R., Al-Rasshdi A., Yee A., Finkelstein F. Clinical features of conjoined lumbosacral nerve roots versus lumbar intervertebral disc herniations. Eur Spine J. 2010;19:1094–1098. doi: 10.1007/s00586-010-1329-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Oh C.H., Park J.S., Choi W.S., Choi E., Ji G.Y. Radiological anatomical consideration of conjoined nerve root with a case review. Anat Cell Biol. 2013;46(4):291–295. doi: 10.5115/acb.2013.46.4.291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Bergamaschi J.P.M., Chaves B.J.M., Campos C.A.R., Da Fonseca M.G., Morale V., Tirapelli L.F., et al. Intervertebral foramen anatomy of the lumbar spine: an overview. Coluna/Columna. 2024;23(4) doi: 10.1590/S1808-185120242304291413. [DOI] [Google Scholar]
  • 11.Bogduk N. 6th ed. Elsevier; Philadelphia: 2023. Clinical and radiological anatomy of the lumbar spine; pp. 117–130. [Google Scholar]
  • 12.Gkasdaris G., Kapetanakis S. Clinical anatomy and significance of the lumbar intervertebral foramen: a review. J Anat Soc India. 2015;64(2):166–173. doi: 10.1016/j.jasi.2015.10.003. [DOI] [Google Scholar]
  • 13.Song S.J., Lee J.W., Choi J.Y., Honh S.H., Kim N.R., Kim K.J., et al. Imaging features suggestive of a conjoined nerve root on routine axial MRI. Skelet Radiol. 2008;37:133–138. doi: 10.1007/s00256-007-0403-6. [DOI] [PubMed] [Google Scholar]
  • 14.Kang C.H., Shin M.J., Kim S.M., Lee S.H., Kim H.K., Ryu J.A., et al. Conjoined lumbosacral nerve roots compromised by disk herniation: sagittal shoulder sign for the preoperative diagnosis. Skeletal Radiol. 2008;37(3):225–231. doi: 10.1007/s00256-007-0421-4. [DOI] [PubMed] [Google Scholar]
  • 15.Gomez J.G., Dickey J.W., Bachow T.B. Conjoined lumbosacral nerve roots. Acta Neurochir (Wien) 1993;120:155–158. doi: 10.1007/BF02112035. [DOI] [PubMed] [Google Scholar]
  • 16.Akhaddar A. Atlas of sciatica: etiologies, diagnosis, and management. Springer; Cham: 2023. Lumbosacral conjoined nerve roots; pp. 701–707. [DOI] [Google Scholar]
  • 17.Cannon B.W., Hunter S.E., Picaza J.A. Nerve root anomalies in lumbar disc surgery. J Neurosurg. 1962;19:208–214. doi: 10.3171/jns.1962.19.3.0208. [DOI] [PubMed] [Google Scholar]
  • 18.Postacchini F., Urso S., Ferro L. Lumbosacral nerve-root anomalies. J Bone Joint Surg. 1982;64A:721–729. [PubMed] [Google Scholar]
  • 19.Kadish L.J., Simmons E.H. Anomalies of the lumbosacral nerve roots. An anatomical investigation and myelographic study. J Bone Joint Surg Br. 1984;66(3):411–416. doi: 10.1302/0301-620X.66B3.6725353. [DOI] [PubMed] [Google Scholar]
  • 20.Burke S.M., Safain M.G., Kryzanski J., Riesenburger R.I. Nerve root anomalies: implications for transforaminal lumbar interbody fusion surgery and a review of the Neidre and Macnab classification system. JNS Focus. 2013;35(2):E9. doi: 10.3171/2013.2.FOCUS1349. [DOI] [PubMed] [Google Scholar]
  • 21.Yu C., Zhengqi C., Xiuchun Y. An amendment to the Neidre and MacNab Classification System for lumbosacral nerve root anomaly and its implication in percutaneous. Endoscopic lumbar discectomy. World Neurosurg. 2018;111:16–21. doi: 10.1016/j.wneu.2017.12.004. [DOI] [PubMed] [Google Scholar]
  • 22.Nikolenko V.N., Shkarubo A.N., Zmeeva E., Sinelnikov M.Y. Proposed unifying classification criteria for spinal nerve root variations. Ochsner J. 2021;21:123–125. doi: 10.31486/toj.21.0014. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Radiology Case Reports are provided here courtesy of Elsevier

RESOURCES