Case Report: Extraforaminal endoscopic lumbar discectomy in two dogs with far-lateral intervertebral disc extrusions
Authors: Colin J. Driver, Kerrie Morrison, Michaela Sojak, Heidi Thatcher, Jeremy Rose
Published: 2026

ABSTRACT
Far-lateral intervertebral disc extrusions (IVDE) represent an uncommon subset of intervertebral disc disease in dogs, in which disc material is herniated dorsolaterally or laterally to compress the exiting nerve root in an extraforaminal location. While minimally invasive spinal surgery techniques have been described in dogs, to the authors knowledge, full-endoscopic discectomy has not previously been reported for extraforaminal IVDE. This case report describes the clinical presentation, surgical technique, and short-term outcomes of extraforaminal endoscopic lumbar discectomy (EELD) in two dogs. Two middle-aged crossbreed dogs were presented with subacute onset of vocalization, pelvic limb lameness presumed to be a nerve root signature, and marked lumbar hyperaesthesia without other abnormalities on neurological examination. MRI or CT confirmed L5–6 extraforaminal disc extrusion with L5 spinal nerve root compression in both cases. Medical management failed to resolve clinical signs. Using a uniportal approach, a 6.3 mm diameter working-channel endoscope was fluoroscopically docked onto the transverse process of the vertebra caudal to the affected disc, creating a muscle-sparing lateral corridor. Extruded disc material was visualized, released from its fibrous capsule, and removed piecemeal without the need for osteotomy. Surgical times were 25 and 30 min. Post-operative CT confirmed effective decompression in both dogs. Post-operative recovery was rapid, with low pain scores and no requirement for opioid analgesia. Dogs were discharged within 24 h, and by 3 weeks both were free of lameness and spinal pain. At 12 weeks, caregiver-reported outcomes demonstrated sustained improvement, including a substantial reduction in Canine Brief Pain Inventory scores. EELD provided effective decompression through a minimally invasive corridor with excellent early outcomes. This first description in dogs supports further investigation of EELD as an alternative to conventional open surgical approaches for extraforaminal IVDE.

Introduction

Intervertebral disc extrusion (IVDE) is a common cause of paresis and paralysis in dogs, which relates to the popularity of predisposed chondrodystrophic dog breeds (1, 2). Surgical decompression of the spinal cord and/or spinal nerve roots is often recommended when pain is persistent despite appropriate medical management (3).

Minimally invasive spinal surgery (MISS) has been described in dogs (both ex-vivo and in small case series) with the use of small incisions and muscle sparing retractors for microscopic (4–6) or endoscope-assisted (7–13) foraminotomy and discectomy (“mini-hemilaminectomy”). The term “full-endoscopic” has been proposed for the myriad of techniques utilizing a working-channel endoscope in humans, which are introduced to the spine through a single small endoscopic portal (14). In dogs, dorsolateral endoscopic thoracolumbar foraminotomy (15), dorsal endoscopic lumbosacral laminectomy (16) and lateral endoscopic lumbosacral foraminotomy (17) have so far been described.

Typically, IVDE occur on mid-line or para-median within the vertebral canal to cause spinal cord injury; however, a less common subset of patients will suffer foraminal or extraforaminal (often termed “far-lateral”) IVDE to cause spinal nerve root compression (18–20). Foraminal and extraforaminal IVDE most commonly occur at L5-6 and L6-7, causing pelvic limb pain or lameness (20). Open surgical options include hemilaminectomy, foraminotomy, corpectomy, a combination of foraminotomy and pediculectomy, or discectomy without bone resection (19, 20). Although direct comparison remains to be studied, these open procedures potentially require longer skin incisions, greater muscle dissection and therefore blood loss and pain, in comparison to endoscopic techniques.

Full-endoscopic approaches to far-lateral disc extrusions have not yet been described in dogs; here we present two cases of endoscopic discectomy for lumbar IVDE in an extraforaminal location (extraforaminal endoscopic lumbar discectomy, EELD).

Case descriptions

. History, clinical evaluation, and diagnostic imaging findings

Two middle-aged crossbreed (Cocker Spaniel crossed with Poodle) dogs (case 1, a 5 years 3 months of age, female neutered, weighing 11.9 kg; case 2, a 9 years 5 months of age, also female neutered, weighing 6.8 kg) were referred for a subacute onset of behavioral changes suggestive of pain and pelvic limb lameness presumed to be a nerve root signature.

Case 1 had a 2-week history of intermittent vocalization, difficulty jumping and climbing stairs, with left pelvic limb lameness of variable severity, which was interpreted as a nerve root signature. These signs had not significantly improved following treatment with 0.1 mg/kg meloxicam once daily and 10 mg/kg gabapentin 3 times daily, orally, given for the 2 weeks prior to referral. The dog had a recent history of urinary tract infection that had resolved with broad spectrum antibiotics. On examination, there was moderate left pelvic limb lameness. There was vocalization on extension of the left hip (hip rotation was tolerated) and on firm dorsal pressure on the caudal lumbar spine. The remainder of the neurologic and orthopedic examinations were unremarkable. The caregiver described significant loss of quality of life due to inactivity, unwillingness to exercise and intermittent vocalization.

Case 2 had a 3-week history of intermittent vocalization, lethargy, a mildly kyphotic posture and mild right pelvic limb lameness, which was interpreted as a nerve root signature. There was only slight improvement with oral analgesic medication given during the 3-week period prior to referral; 15 mg/kg paracetamol 3 times daily, meloxicam 0.1 mg/kg once daily and 10 mg/kg gabapentin 3 times daily. Orthogonal radiographs were performed at the referring veterinary practice which revealed a well circumscribed mineralized radiopacity, adjacent to the L5-6 intervertebral disc and foramen on the right side (Figure 1). On examination, there was persistent mild right pelvic limb lameness and similar to case 1, vocalization was elicited on right hip extension (with hip rotation tolerated) and on dorsal lumbar spinal palpation. There were no neurologic deficits. The caregiver also described significant loss of quality of life due to loss of function and signs suggestive of pain.

Pre-operative diagnostic imaging of cases 1 (A, B) and 2 (C, D). (A) Transverse T2-weighted transverse MRI image at the level of the L5-6 intervertebral disc. A far-lateral intervertebral disc extrusion (arrowed) causes extra-foraminal compression of the exiting L5 nerve root with loss of normal foraminal fat signal. (B) Dorsal fat-suppressed short tau inversion recovery (STIR) image demonstrated peri-neural increased signal (arrow). (C) Transverse CT image, at the level of the L5-6 intervertebral disc, reconstructed in a bone window (width 1500 HU; level 300 HU) with 1.5 mm slice thickness. A far-lateral intervertebral disc extrusion (arrowed) causes extra-foraminal compression of the exiting L5 nerve root. There is also a poorly delineated and patchy hyperattenuating lesion in the subjacent psoas muscle, later confirmed in surgery as a mixture of extruded disc material and blood (asterisk). The mean Hounsfield units within regions of interest of the extrusion (arrowed) and extrusion/haematoma (asterisk) were 413 and 185, respectively (normal vertebral bone 481). (D) Ventrodorsal radiograph of the pelvis and caudal lumbar spine demonstrating the focal radiopacity representing a superimposition of the extruded disc material and haematoma lateral to the intervertebral foramen.

In both cases, the examination was consistent with spinal pain and the lameness was assumed to be the result of radicular pain, given the clinical syndrome was subacute in onset, included asymmetric lameness and vocalization which are atypical for osteoarthritis of the hip. The anatomic localization was to the L4–S1 spinal nerve roots. Differential diagnoses were considered to include intervertebral disc disease, degenerative lumbosacral stenosis, discospondylitis, peripheral nerve sheath or paraspinal neoplasia, immune-mediated radiculopathy and foraminal stenosis secondary to congenital facet joint dysplasia with secondary hyperostosis. The caregivers were invited to complete the canine brief pain inventory (CBPI) questionnaire (21), utilizing a quick-reference code to access an online form, that could be completed at home in private (Microsoft Forms, Microsoft, WA, USA). CBPI is a validated caregiver-reported outcome measure for the treatment of osteoarthritis in dogs (22, 23). Scores were 79 and 76 for cases 1 and 2, respectively. In both cases, quality of life was recorded as poor.

In both cases, blood was collected from the jugular vein for hematology and serum biochemistry profiles, which did not reveal abnormalities. Multiplanar advanced imaging of the caudal lumbar and lumbosacral spine was performed under general anesthesia; case 1 underwent high-field MRI (1.5 T; Magnetom Flow.Ace, Siemens, Germany) and case 2 CT (Somatom Scope, Siemens, Germany). The choice of imaging modality was based on the preference of the attending clinician. In both cases, a 21 G cephalic cannula was placed and 0.2–0.3 mg/kg methadone and 3 mcg/kg dexmedetomidine was administered intravenously as a premedicant, prior to the induction of anesthesia with 1–2 mg/kg alfaxalone to effect. Anesthesia was maintained with isoflurane.

In case 1, the MRI revealed left-sided dorsolateral extrusion of the L5-6 intervertebral disc (Figure 1); the extruded disc material was hypointense on T2-weighted images and was predominantly dorsolateral to the disc in an extraforaminal location, with a very small amount of material in the ventral aspect of the intervertebral foramen. The foraminal L5 spinal nerve root was compressed and deviated dorsally. Adjacent to the extruded disc material and within the paraspinal muscle, there was a focal area of STIR hyperintensity relative to the muscle. A three-dimensional T2-weighted SPACE (Sampling Perfection with Application optimized Contrast using different flip angle Evolution) sequence helped appreciate the course of the extraforaminal nerve root lateral to the disc herniation. The diagnosis was a left-sided far-lateral L5-6 intervertebral disc extrusion.

In case 2, the CT confirmed the presence of hyperattenuating material adjacent to the dorsolateral aspect of the L5-6 intervertebral disc, with assumed compression of the L5 extraforaminal nerve root given the impingement of its normal anatomic course ventral to the transverse process of L6 (Figure 1). There was also a poorly circumscribed region of mixed hyper- and iso-attenuating material in the psoas muscle medial to the L6 transverse process, presumed to relate to a muscular haematoma mixed with the extruded disc material. The diagnosis was a right-sided far-lateral L5-6 intervertebral disc extrusion.

Both dogs were recovered from general anesthesia. Given the distressing nature of the vocalization, sudden loss of acceptable quality of life, and failure to improve with medical management, we discussed surgical decompression as a potential option. Given our experience with lateral endoscopic lumbosacral foraminotomy (17), we proposed a minimally invasive endoscopic procedure, and consent was given to proceed with EELD.

Surgical procedures

In both cases, the EELD procedure was scheduled for the following day. The anesthetic protocol was determined by a diplomate of the European College of Veterinary Anesthesia and Analgesia, which is an institutional norm. The anesthetic premedicant and induction protocols were identical to those used for the diagnostic imaging anesthetic protocol, described above. In both cases, a choice for epidural injection of 1 mg/kg bupivacaine at L7-S1, was made. Prophylactic antibiotics were given intravenously at induction (cefuroxime 20 mg/kg). Additional perioperative analgesia was pre-determined to be provided with a continuous rate infusion of ketamine (5 mcg/kg/min).

The following surgical instruments and equipment were supplied by Arthrex (Naples, FL, USA) for the duration of the study (17): a spine-specific, working channel endoscope of 6.3 mm diameter and 130 mm working length, with 30° angled optics. It was introduced through a 7 mm diameter, 125 mm working length cannula, with elevator tip. Cautery was provided with the Synergy high frequency (4 MHz) electrosurgery unit and a 28 cm length flexible tip radiofrequency ablation probe. The Synergy camera, light cable and UHD4 console unit and monitor were also used. Other equipment included five serial muscle dilators (diameters 2.5, 4.1, 5.1 and 6mm), a 2.5 × 260 mm cup forcep and a 2.5 × 260 mm nerve hook probe. A spinal burr system was not required. Fluid irrigation was passive, utilizing gravity (3 L fluid bags elevated 0.5 m above the patient level).

Dogs were positioned on a radiolucent carbon fiber orthopedic operating table in a prone position. The surgeon stood on the same side as the disc herniation. A C-arm image intensifier (Fujifilm FDR Cross) and scope tower were positioned opposite. The surgical instruments and electrocautery unit were positioned at the foot of the patient table.

Following aseptic patient preparation and draping, the C-arm was orientated dorsoventral to the caudal lumbar spine, using its guidance laser. Local radiation safety rules concerning fluoroscopy were obeyed, including the use of lead gowns, personal- and ring-dosimeters. Fluoroscopy (voltage 55–75 kV, collimation 14 cm, fluoroscopic mode at 25 frames per second) was first used to identify the location of the skin incision and to “dock” the endoscopic trochar adjacent to the transverse process of the caudal vertebra, caudal to the disc extrusion (Figure 2). Dorsoventral x-ray projections allowed the superimposition of a 0.9 mm diameter stainless steel k-wire held (at a safe distance) over the following landmarks, which were marked on the skin with a sterile pen: the sagittal mid-line (aligned to the dorsal spinous processes) and the transverse orientation of the affected intervertebral disc. The distance along the transverse line to make the initial incision had already been determined from the pre-operative scans; the aim was for the incision to allow a k-wire to be advanced (by hand) at an angle of 30° to vertical, to firmly dock at the transverse process of the vertebral caudal to the intervertebral disc (Figure 2).

Docking the spinal endoscope cannula with fluoroscopy. A 1.1 mm k-wire is held over the operative field and x-ray projections are used to confirm the level of the transverse process of L6 (A). Following a stab incision in the skin and fascia, the k-wire is advanced through the muscle to the base of the transverse process, with an angle of 30° to vertical, and with a slight caudolateral to craniomedial orientation (B). The correct location is check in lateral (C) before serial muscle dilators are used to facilitate insertion of the endoscopic cannula caudal to the intervertebral foramen (D).

An 8 mm stab incision was made using a no. 11 blade, parasagittal to the vertebral column. The incision was extended deep to the subcutaneous tissue through the lumbodorsal fascia. The k-wire was then inserted through the para-spinal muscle using the 30° angled trajectory, until firm resistance from the transverse process was felt. The correct craniocaudal location of the pin was confirmed with a dorsoventral x-ray projection. The C-arm was then re-orientated perpendicular to the patient and a lateral x-ray projection was used to confirm the k-wire was docked onto the base of the transverse process just caudal to the affected intervertebral disc. The muscle dilators were sequentially inserted over the k-wire and lastly the 7 mm diameter endoscopic trochar was gently inserted, before the dilators removal. The elevator tip of the trochar was orientated cranially. Intermittent dorsoventral x-ray projections were made during the process to confirm the final “docked” location of the trochar (Figure 2).

The endoscope was introduced into the trochar with active fluid irrigation. The remaining muscle at the base of the trochar was gently elevated from the base of the transverse process with the trochar. Minor soft-tissue bleeding was managed with the bipolar cautery. The elevator tip of the trochar was then gently “wanded” cranially to expose the dorsolateral aspect of the intervertebral disc, with the assumption the exiting spinal nerve root was protected. In both cases, the white-yellow colored extruded disc material could then be appreciated and was found to be partially firm and encapsulated in a thin layer of fibrous tissue (Figure 3, case 1). The fibrous layer was cauterized and penetrated with the radiofrequency probe, before the extruded disc material could slowly be removed piecemeal using the nerve hook. Following this process, the exiting L5 nerve root was more easily appreciated and appeared decompressed (Figure 3). In case 2, the lesion ventrolateral to the disc was explored using the radiofrequency probe, allowing partial evacuation of a mixture of extruded disc material and blood. The endoscope and trochar were then slowly removed from the 8 mm surgical wound and the skin was closed with a single nylon suture in a cruciate pattern. The total procedure times (from incision to closure) were 25 and 30 min for cases 1 and 2, respectively.

Extraforaminal endoscopic lumbar discectomy procedure in case 1. The location of the transverse process of L6 (star) is confirmed with the bipolar cautery tip (A). With the elevator tip of the cannula orientated cranially, the cannula is wanded to elevate soft tissue and expose the encapsulated disc material (cross) lateral to the intervertebral disc (B). The capsule is penetrated with the nerve hook and extruded disc material (arrowed) is released and can be removed piecemeal (C). With the exiting spinal nerve root of L5 now entering the endoscopic view (asterisk), removal of disc material is completed (D) and a dissector is used to explore the lateral aspect of the disc (arrow) whilst protecting the nerve root (asterisk, E). Upon completion of the procedure, the nerve root has returned to its normal anatomic orientation (F).

Prior to recovery from general anesthesia, a post-operative CT was performed to confirm procedure efficacy (Figure 4).

Pre vs. post-operative imaging from cases 1 (A1, A2) and 2 (B1, B2). (A1) and (B1) are labeled as in Figure 1. In case 1, the post-operative CT (A2) revealed removal of the extruded disc material (thick arrow) and only subtle signs of soft-tissue disruption related to the procedure (a gas bubble and fluid in the subcutaneous fat, thin arrows). In case 2, the post-operative CT (B2) extraforaminal extruded disc is removed (thick arrow), signs of soft-tissue disruption are limited to a small amount of fluid in the subcutaneous fat (thin arrow).

Recovery and follow-up

Both dogs recovered uneventfully from general anesthesia. The short form of the Glasgow Composite Pain Scale (GCPS) was recorded every 4 h during hospitalization (24) by suitably qualified caregivers (registered veterinary nurses) who were not blinded to the procedure, but who were experienced in GCPS-SF assessments in dogs undergoing a range of surgical procedures in a multi-disciplinary specialty referral setting. The highest pain scores recorded for cases 1 and 2 were 2 and 3, respectively. As the prompt for the administration of additional opioid analgesia in our institution is a score of 6, no opioids were given during hospitalization. The morning after the procedure (after the expected duration of activity of the epidural bupivacaine had lapsed), both dogs appeared bright and willing to ambulate. Physical examination was performed by a board-certified neurologist (CD), which confirmed the absence of neurologic deficits and neurogenic lameness. Palpation in the region of the surgical wound was also tolerated. The GCPS-SF was repeated unto the point of discharge, which occurred the following morning (less than 24 h after the EELD procedure), with instructions for 10 days of lead-restricted exercise and continued treatment with a non-steroidal anti-inflammatory drug. After 10 days, resumption of normal activity was allowed (this was an institutional protocol for endoscopic procedures).

Follow-up was obtained by physical examination at 3 weeks post-operative and by telephone interview at 12 weeks. At 3 weeks, both dogs were considered neurologically normal, with no residual gait abnormality, evidence of pain or need for continued oral analgesics. At 12 weeks, the resolution of major clinical signs was reportedly sustained. Caregiver CBPI scores covering the 3-month period post-operative had improved from 79 to 14, and 76 to 8, for cases 1 and 2 respectively. Quality of life was scored as very good in both cases.

Discussion

In this case report we present the first description of EELD for the management of extraforaminal IVDE in two dogs. EELD is a minimally invasive, uniportal full-endoscopic surgical technique that resulted in clinician and caregiver reported marked improvement in clinical signs with restoration of normal quality of life. The procedure minimized muscle trauma, spared vertebral bone, and alongside the use of locoregional anesthesia, could be performed without the need for opioid medication, lengthy hospitalization or prolonged exercise restriction.

Both dogs had typical clinical signs and imaging findings previously reported after far-lateral lumbar IVDE (18, 20); notably, pelvic limb lameness was associated with hyperaesthesia on hip extension, lumbar spine palpation and intermittent painful vocalization. In both cases, the patellar reflex was found to be intact, despite the L5 nerve root making the largest contribution to the femoral nerve (25). This might relate to the lack of involvement of the L4 and L6 spinal nerve roots and the potential preservation of L5 nerve root function, given the perceived lack of paresis in the muscles innervated by the femoral nerve.

In case one, the extraforaminal located extruded disc material was associated with the previously reported finding of hyperintense signal on dorsal STIR sequences (20). Both dogs were Cocker Spaniels crossed with Poodles (Cockapoo); outside of Dachshund and related crossbreeds, the Cocker Spaniel was previously commonly presented for far-lateral lumbar IVDE (20). The breed predisposition to IVDE likely relates to chondrodystrophy (26).

In both cases, the choice for surgical management was based on the severity and nature of the clinical signs which failed to improve with medication, despite the lack of neurologic deficits. The episodes of spontaneous vocalization were particularly distressing for the caregiver. This is consistent with the existing literature where the presence of neurologic deficits did not inform the choice of medical vs. surgical treatment (20). Outcomes for far-lateral lumbar IVDE have been reported as excellent or good in 95% of surgically managed and 90% of medically managed cases (20). Given this rate of successful medical management, the dogs in this study may have gone on to recover without surgery with time, albeit with a risk of recurrent clinical signs. The rate of good to excellent outcome in humans following surgery for foraminal or extraforaminal disc herniations ranges from 68 to 91%, and surgical management is often favored over conservative treatment (reported success rate as 10%−71%), but this may relate to the prevalence of neurologic deficits (27).

In human neurosurgery, several open approaches for far-lateral discectomy have been described. These include laminotomy with medial facetectomy, complete facetectomy and an intertransverse' approach (28), which whilst being amenable to microendoscopic discectomy (29), still requires a small osteotomy and does not address foraminal herniations (28). The more laterally orientated “transforaminal” approach for full-endoscopic discectomy is recognized as an efficient and minimally invasive technique for managing lumbar disc herniation (30–32), and has been shown to offer faster recovery with shorter length of hospital stay and reduced blood loss vs. open surgery (33). A modification of the transforaminal technique was described for targeting extraforaminal disc herniations in humans, without the need for performing a foraminotomy (34). We felt a similar modification was likely to be helpful in dogs, given that the endoscopic sheath cannot be safely targeted to the intervertebral disc through the intervertebral foramen due to their relative sizes.

Our purpose was to develop EELD to avoid the need for osteotomy, with a safe approach that would spare the exiting spinal nerve root from iatrogenic trauma. We therefore chose to target the transverse process of the vertebra just caudal to the dorsolateral and lateral aspect of the affected intervertebral disc. This was successful in exposing the partially encapsulated extruded disc material without visibly damaging the nerve root or causing post-operative neurologic deficits. Larger or foraminal-limited IVDE may require a more targeted approach to the accessory process, which is situated craniodorsal to the foramen, for “endoscopic mini-hemilaminectomy” (15), which is effectively a foraminotomy that incorporates some resection of the adjacent vertebral pedicles if wider access to extruded disc material is needed. Similar to human surgery, where access to foraminal IVDE with an extraforaminal approach may be facilitated by the addition of foraminotomy (35), there is no reason why the extraforaminal and foraminal approaches could not be combined in dogs if required.

Prophylactic intervertebral disc fenestration refers to removing disc material in situ, with the goal of preventing future extrusion; this procedure has previously been recommended to be performed at the time of decompressive surgery at the level of the affected disc (3). A potential disadvantage to endoscopic discectomy would be a reduced ability to fenestrate adequately, however, the authors have unpublished experience with annulotomy for radiofrequency cauterization of the nucleus pulposus following endoscopic mini-hemilaminectomy for thoracolumbar IVDE. This technique has not yet been studied as a comparable method to reduce the risk of recurrence. As prophylactic fenestration was not performed in the two cases presented in this study, a potential risk of future recurrence exists.

In this study we utilized a caregiver reported outcome measure (CBPI) for objective assessment of change in quality of life, pain and the effect of pain on function (21–23). Despite reporting a subjectively normal comfort and function at 12-weeks, caregivers did not record all assessments as “zero” (scores 14 and 8 for cases 1 and 2, respectively). It may have been more accurate to utilize a clinician-based assessment for pre-operative and follow-up pain, such as the LS Pain scale, which is valid and reliable in the assessment and monitoring of lumbosacral pain in dogs (36).

In conclusion, EELD was a safe and effective technique for the management of extra-foraminal IVDE in this small case series. Similar to lateral endoscopic lumbosacral foraminotomy (17), there was a small surgical wound and a short period of hospitalization with no need for post-operative opioids. The caregivers were satisfied with the rate and extent of clinical improvement seen. The main limitation to this small case series is the limited ability to determine operative risk and success rates in a larger population of dogs and to make comparison with open procedures. This is due to its nature as a case report, with small sample size, lack of control group, heterogeneous imaging modality utilization, short follow-up, non-blinded clinical assessment and reliance on caregiver-reported outcome by telephone at 12-week follow up. The need for specific instrumentation and surgical facilities potentially poses a challenge to the widespread adoption of this technique.