Introduction

A10-year-old 19.3-kg (42.5-lb) neutered male Basset Hound (dog 1) was referred to North Carolina State University's veterinary teaching hospital because of a 3-week history of progressive signs of severe cervical pain. Conservative management for 3 weeks with strict rest and pain relief (carprofen, gabapentin, tramadol, and methocarbamol) had been initiated by the referring veterinarian, but the dog appeared to be in extreme pain and was subsequently referred for further management. At the time of the referral evaluation, the dog was being treated with gabapentin (7.8 mg/kg [3.5 mg/lb], PO, q 8 h), methocarbamol (26 mg/kg [11.8 mg/lb], PO, q 8 h), and carprofen (2.6 mg/kg [1.2 mg/lb], PO, q 12 h). The dog had low head carriage, and signs of severe pain were elicited on palpation of the neck, especially during dorsal extension. No lameness was identified. Hopping on the left thoracic limb was delayed, but the remainder of the neurologic and general physical examination findings were unremarkable. Subsequently, MRI was performed, which revealed left foraminal IVDP between C2 and C3 with mild C3 nerve root compression. Given the location of the disk protrusion, only mild spinal cord compression, and associated surgical risks, the owner declined surgical intervention. An ultrasound-guided paravertebral perineural glucocorticoid injection was performed.

Dog 1 was sedated with dexmedetomidine (5 μg/kg [2.27 μg/lb], IV) and methadone (0.3 mg/kg [0.14 mg/lb], IV) and positioned in right lateral recumbency with the neck and head in a neutral position (Figure 1). A pillow was placed under its neck to maintain a neutral alignment of the cervical vertebrae and to facilitate ultrasonographic visualization of the vertebral column. Hair on the dog's neck was clipped, and the skin was scrubbed; a sterile drape was applied. The left atlas wing and the transverse process of C2 were identified by palpation. A linear ultrasound probe^a in a sterile sleeve was positioned on the lateral aspect of the neck in a transverse plane over the transverse process of C2 with the cursor oriented ventrally. The nerve root of C3 appeared as a 2-mm-diameter hypoechoic nodular structure (honeycomb appearance) immediately superficial to the vertebral artery at the level of the C2-3 intervertebral foramen. Once the nerve root was identified, a 21-gauge, 100-mm nerve block needle^b was inserted in plane with the ultrasonographic beam to allow observation of the entire needle during insertion and injection. The needle was advanced in a ventrodorsal direction toward the nerve root, and the needle tip was placed immediately adjacent to the extraforaminal portion of the C3 nerve root (Figure 2). Color Doppler ultrasonography was used to identify the vertebral artery. Aspiration was performed to confirm a lack of vasculature puncture, and methylprednisolone acetate^c (1 mg/kg [0.45 mg/lb]) was injected perineurally under direct ultrasound guidance. The dog recovered uneventfully from sedation, and no complications related to the procedure were observed. The following day, the dog was discharged from the hospital with no immediate improvement of its signs of neck pain. At this time, the dog was being treated with gabapentin (15.5 mg/kg [7.0 mg/lb], PO, q 8 h), methocarbamol (38.9 mg/kg [17.7 mg/lb], PO, q 8 h), and carprofen (2.6 mg/kg, PO, q 12 h). Strict rest was continued at home.

View Full Size

Representative photograph of positioning of 1 of 4 dogs for ultrasound-guided cervical paravertebral perineural injection with methylprednisolone acetate (A) and illustrations of the anatomic landmarks for C3 nerve root identification and perineural injection (B) or C7 nerve root identification and perineural injection (C). A—For the procedure, the dog (dog 1) is sedated and positioned in right lateral recumbency with the neck and head in a neutral position. A pillow is placed under its neck to maintain a neutral alignment of the cervical vertebrae and to facilitate ultrasonographic visualization of the vertebral column. B—The palpable anatomic landmarks (green) and ultrasonographic landmarks for C3 nerve root identification are depicted. Paravertebral perineural injection at the C3 nerve root is performed from a ventral approach. C—The anatomic landmarks (green) and ultrasonographic landmarks for C7 nerve root identification and ultrasound probe and needle positions for injection are depicted. Nerves are shown in yellow and arteries in red. US = Ultrasound.

Citation: Journal of the American Veterinary Medical Association 258, 9; 10.2460/javma.258.9.999

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Representative transverse ultrasonographic image of the C3 nerve root (A) in a dog and schematic drawing of the anatomic landmarks (B) along with a transverse ultrasonographic image obtained during subsequent needle placement (arrow) for paravertebral perineural glucocorticoid injection (C) and schematic drawing of the injection technique (D). 1 = C3 nerve root. 2 = Vertebral artery. 3 = Transverse process of C2. 4 = Caudal articular process of C2. 5 = Spinous process of C2. 6 = Cervical muscles (cleidocervicalis, serratus ventralis, and omotransversarius muscles). 7 = Needle. 8 = Spread of injectate.

Citation: Journal of the American Veterinary Medical Association 258, 9; 10.2460/javma.258.9.999

• Download Figure
• Download figure as PowerPoint slide

Four weeks later, the dog was returned for a recheck examination and repeated paravertebral perineural glucocorticoid injection. The owner reported that the dog had signs of pain for 2 days after the first injection but had appeared very comfortable thereafter. The dosage of methocarbamol had been reduced to 26 mg/kg PO in the morning and 13 mg/kg (5.9 mg/lb) PO in the evening; dosages of the other medications remained the same. On examination, the dog had normal head carriage and no signs of cervical pain could be elicited on palpation of the neck. Similar to findings of the previous examination, hopping on the left thoracic limb was delayed. The dog was sedated, and an ultrasound-guided C2-3 left foraminal injection with methylprednisolone acetate^c was repeated as described. The dog recovered from sedation uneventfully and was discharged from the hospital the same day.

Three months later, dog 1 was presented because of recurrence of signs of cervical pain that started 1 week earlier. The dog was being treated with gabapentin (15.5 mg/kg, PO, q 8 h), methocarbamol (26 mg/kg, PO, in the morning and 13 mg/kg, PO, in the evening), and carprofen (2.6 mg/kg, PO, q 24 h). A third foraminal injection was performed with the previously described protocol. No complications occurred, and the dog was discharged from the hospital the same day.

All medications were discontinued about 1 month following the third paravertebral perineural glucocorticoid injection, and the dog had no further recurrence of signs of pain or lameness. Eleven months after the third paravertebral perineural glucocorticoid injection, the dog was returned to the hospital for a recheck examination. The owner reported that dog 1 was doing well at home with no recurrence of signs of pain. A 9-year-old 32.0-kg (70.4-lb) spayed female Belgian Malinois (dog 2) was brought to the neurology service of the veterinary teaching hospital for evaluation of a waxing and waning left forelimb lameness of 1 year's duration. The dog was previously managed by the hospital's orthopedic service. Signs of pain were initially localized to the dog's left shoulder region, and radiography and ultrasonography of the shoulder joint were performed with no notable findings identified. Medical management with gabapentin and grapiprant and a trial treatment with carprofen had been initially attempted. None of these medications seemed to improve the lameness, and the dog had received an intra-articular methylprednisolone acetate^c injection into the left glenohumeral joint with no improvement. On presentation to the neurology service, dog 2 was ambulatory with mild left forelimb lameness. No ataxia was noted, and postural reactions were normal. Signs of pain were elicited on lateral flexion of the neck. The remainder of the physical examination findings were unremarkable. The dog underwent MRI, which revealed multifocal degenerative disk disease with mild focal left-sided disk protrusion between C6 and C7 without associated spinal cord or nerve root compression. Surgical treatment was not indicated, and medical management was recommended. Although the degree of lameness and clinical signs did not seem to correlate with the minimal abnormalities identified by MRI (mild disk protrusion without spinal cord or nerve root compression), the owner elected to have the dog treated with a paravertebral perineural glucocorticoid injection.

Dog 2 was sedated with dexmedetomidine (5 μg/kg, IV) and methadone (0.3 mg/kg, IV) and positioned in right lateral recumbency with its neck and head in a neutral position. A pillow was placed under the neck to maintain a neutral alignment of the cervical portion of the vertebral column. The right thoracic limb including the scapula was retracted caudally to allow better exposure and ultrasonographic visualization of the caudal cervical portion of the vertebral column.

Paravertebral perineural injection of the C7 nerve root with methylprednisolone acetate^c (1 mg/kg) was performed with ultrasound guidance. Hair on the dog's neck was clipped, and the skin was scrubbed; a sterile drape was applied. The left ventral region of the lamina of the C6 transverse process was identified by palpation. The ultrasound probe in a sterile sleeve was positioned on the lateral aspect of the neck in a transverse plane over the transverse process of C6 with the cursor oriented dorsally (Figure 1). The nerve root of C7 appeared as a 3 × 2-mm hypoechoic oval structure with a hyperechoic rim (honeycomb appearance) immediately lateral to the vertebral artery at the level of the C6-7 intervertebral foramen (Figure 3). Once the nerve root was identified, a 21-gauge, 100-mm nerve block needle^c was inserted in plane with the ultrasound beam to allow observation of the entire needle during insertion and injection. The needle was advanced in a dorsoventral direction toward the nerve root, and the needle tip was placed immediately adjacent to the extraforaminal portion of the C7 nerve root. Aspiration was performed to confirm a lack of vasculature puncture, and methylprednisolone acetate^c (1 mg/kg) was injected perineurally under direct ultrasound guidance. Dog 2 recovered from sedation with no complications and was discharged from the hospital the same day. Gabapentin (12.5 mg/kg [5.7 mg/lb], PO, q 8 to 12 hours, as needed) was prescribed for pain.

View Full Size

Representative transverse ultrasonographic image of the C7 nerve root (A) in a dog and schematic drawing of the anatomic landmarks (B) along with a transverse ultrasonographic image obtained during subsequent needle placement (arrow) for paravertebral perineural glucocorticoid injection (C) and schematic drawing of the injection technique (D). In panel A, the arrows indicate the ventral lamina of the transverse process of C6, which is a palpable anatomic landmark used as the primary orienting ultrasonographic reference to identify the location of the C7 nerve root. In panel C, the nerve root of C7 (arrow) emerges from the intervertebral (C6-7) foramen and continues in a caudoventral direction lateral toward the vertebral artery. The nerve root appears as a hypoechoic oval structure with hyperechoic rim. 1 = C7 nerve root. 2 = Vertebral artery. 3 = Transverse process of C6. 4 = Ventral lamina of C6. 5 = Articular process of C6. 6 = Cervical musculature. 7 = Needle.

Citation: Journal of the American Veterinary Medical Association 258, 9; 10.2460/javma.258.9.999

• Download Figure
• Download figure as PowerPoint slide

In the first week after the injection, the owner reported only minimal improvement in the dog's signs of pain. After 1 week, the dog's lameness seemed mildly improved. One month later, the dog was returned for a repeated paravertebral perineural glucocorticoid injection because of progressive lameness. At this time, dog 2 was receiving gabapentin (12.5 mg/kg, PO, q 12 h). A second paravertebral perineural methylprednisolone acetate^c injection at the C7 nerve root was performed. No complications were observed, and the dog was discharged from the hospital. Treatment with gabapentin (12.5 mg/kg, PO, q 8 to 12 h, as needed) was prescribed for pain. After 1 week, the owner reported that the dog's degree of lameness had improved moderately. However, after a few days of improvement, the dog's signs of pain returned to the preinjection level. No additional paravertebral perineural glucocorticoid injections were performed given the lack of sustained response. Gabapentin was continued for pain control.

An 8-year-old 49.8-kg (109.8-lb) sexually intact male Labrador Retriever (dog 3) was evaluated at the veterinary teaching hospital because of a 3.5-week history of signs of neck pain. The dog was initially managed conservatively with strict rest, application of a fentanyl patch, and administration of gabapentin, methocarbamol, and prednisone. The medications seemed to help reduce the signs of pain only slightly, and the dog was subsequently referred for further management.

On presentation, dog 3 was tetraparetic and had a low head carriage. It had delayed proprioception in both pelvic limbs with intact spinal reflexes. Signs of pain in the caudal cervical region were identified on neck palpation, and the dog resisted extension and flexion of the neck. The dog underwent MRI, which revealed left foraminal IVDP between C6 and C7 with suspected focal spinal cord atrophy or mild compression. A small volume of extruded disk material at the level of the intervertebral foramen was suspected and interpreted as a possible acute or chronic disk protrusion. Continued medical management was initially recommended, and the dog was discharged from the hospital. The owner was instructed to provide the dog with strict rest and treatment with gabapentin (12 mg/kg [5.5 mg/lb], PO, q 12 h) and methocarbamol (10 mg/kg [4.5 mg/lb], PO, q 12 h).

Dog 3 continued to have signs of pain and was returned 3 weeks later for a paravertebral perineural glucocorticoid injection. The dog was sedated with dexmedetomidine (5 μg/kg, IV) and butorphanol (0.3 mg/kg, IV), and an ultrasound-guided left paravertebral perineural injection with methylprednisolone acetate^c (1 mg/kg) at the C7 nerve root was performed as described for dog 2. Dog 3 recovered without complications and was discharged from the hospital; the previously prescribed gabapentin and methocarbamol treatments were continued.

The dog was returned 4 weeks later for a repeated injection. The owner reported improvement after the first injection with only intermittent signs of pain. Mild lameness and signs of pain in the caudal cervical region were evident during the physical examination. A second C7 left paravertebral perineural glucocorticoid injection was performed as previously described. The dog was discharged from the hospital and administered gabapentin and methocarbamol as previously prescribed.

The dog's signs of neck pain improved markedly after the second paravertebral perineural glucocorticoid injection, and the dog was presented 1 month later for a repeated injection. Dog 3 had only mild left forelimb lameness and mild resistance to lateral neck flexion on physical examination. The dog was sedated, and a third paravertebral perineural glucocorticoid injection was administered, adhering to the previously described protocol. The dog recovered without complications and was discharged from the hospital the same day; the dog was to be treated with gabapentin (12 mg/kg, PO, q 12 h). The owner reported that after the third injection, the dog's signs of cervical pain dramatically improved along with complete resolution of the lameness.

A 7.5-year-old 9-kg (19.8-lb) spayed female Dachshund (dog 4) with a history of intermittent signs of neck pain of 2.5 years' duration was brought to the neurology service of the veterinary teaching hospital because of chronic signs of neck pain and limping on the right forelimb. The dog had been evaluated by the emergency service approximately 10 days earlier because of signs of neck pain and an ataxic gait. Treatment with gabapentin (11 mg/kg [5.0 mg/lb], PO, q 8 h), methocarbamol (28 mg/kg [12.7 mg/lb], PO, q 8 h), and grapiprant (2.2 mg/kg [1.0 mg/lb], PO, q 12 h) was initiated. For the first 2 to 3 days, dog 4 responded well to the medications and activity restriction, but then the signs of pain progressively worsened. During the preceding 2.5 years, the dog's intermittent signs of cervical neck pain had been managed medically with gabapentin and carprofen.

At the examination, the dog had right forelimb lameness and intermittent low head carriage with resistance to right lateral flexion of the neck and signs of pain in the caudal cervical region. The dog underwent MRI, which revealed right foraminal IVDP between C6 and C7 and multifocal cervical intervertebral disk degeneration with annulus fibrosus protrusion. The owner declined surgical options for treatment of the dog. An ultrasound-guided right paravertebral perineural injection at the C7 nerve root with methylprednisolone acetate^c (1 mg/kg) was performed, adhering to the technique described for dogs 2 and 3. No complications occurred.

The dog was discharged from the hospital the same day and was to be administered gabapentin (11 mg/kg, PO, q 12 h) and methocarbamol (28 mg/kg, PO, q 12 h) for 10 days. Instructions for strict activity restriction of the dog were given to the owner.

Dog 4 was returned for a second paravertebral perineural glucocorticoid injection 4 weeks after the first treatment. Apart from 1 episode of mild signs of neck pain 5 days after the first injection, the dog had responded well to treatment, and administration of analgesics at home had been discontinued. On presentation, the dog had a normal gait with no evident lameness and appeared to be comfortable during neck palpation and range-of-motion maneuvers. The dog was sedated, and a second paravertebral perineural glucocorticoid injection was administered as described. The procedure was uneventful, and the dog recovered from sedation with no complications. The dog was discharged from the hospital. One month following the second paravertebral perineural glucocorticoid injection, the dog had no signs of pain at home and no further analgesic treatment was needed.

Discussion

Intervertebral disk protrusion (or herniation) is a common neurologic disorder in dogs. The cervical portion of the vertebral column is affected in 12.9% to 25.4% of dogs with IVDP.^1,2 Cervical hyperesthesia alone or with uni- or bilateral thoracic limb lameness without notable neurologic deficits is a common clinical sign of IVDP.^3,4 For affected dogs with a sudden onset of signs of pain and no or minor neurologic deficits, conservative management is typically recommended and consists of administration of analgesics, anti-inflammatory drugs, and muscle relaxants and strict exercise restriction.^3,5 In 1 study⁵ of dogs with cervical intervertebral disk disease, medical management of 88 dogs had a success rate of 48.9%; 29 (33%) dogs had recurring signs after initial improvement, and 16 (18.1%) dogs had failure of conservative management. However, surgical intervention is not always indicated owing to an inability to achieve full decompression or the risk of causing possible cervical instability. Therefore, alternative approaches need to be considered when medical management alone fails and surgery is not possible.⁶ In humans with spinal pain attributable to cervical lesions, intraforaminal and perineural corticosteroid injections have been successfully performed.^7,8

To the authors' knowledge, the present report represents the first description of cervical nerve root paravertebral perineural glucocorticoid injections for management of signs of chronic pain related to cervical IVDP in dogs. All 4 dogs of the present report had an unsatisfactory response to traditional medical management and were not considered suitable surgical candidates.

Overall, the dogs of the present report had a favorable response to the paravertebral perineural methylprednisolone acetate injections. Dogs 1, 3, and 4 were treated successfully, and the signs of cervical pain and lameness resolved. Dog 2's condition did not improve following treatment. However, it is possible that the mild cervical disk protrusion was not the actual underlying cause of the dog's persistent forelimb lameness.

Although no studies of ultrasound-guided paravertebral perineural injections in veterinary species have been performed to the authors' knowledge, a good success rate was reported for 5 dogs with root signature signs treated with fluoroscopy-guided foraminal injection of methylprednisolone with or without bupivacaine.⁹ Three dogs had marked improvement within 1 week after the injection; for one dog, pain control appeared to be achieved within 2 weeks after the injection, and the other dog had only moderate improvement within 3 weeks.⁹

Variable responses to paravertebral perineural injections are also evident in humans. In humans with cervical disk disease and spondylosis, ultrasound-guided paravertebral perineural corticosteroid injections result in a 50% to 75% decrease in pain intensity in 53% to 78% of patients at 3 to 6 months after injection.^8,10,11 In a recent study¹² investigating 149 patients receiving cervical paravertebral perineural injections, 107 (72%) people had an initial improvement after the injection and 36 (24%) had no recurrence of signs at their last follow-up at least 6 weeks after the procedure. Another study¹³ involving MRI-guided perineural injections in 21 patients revealed only a limited success rate within the first week. Eleven patients had complete or strong improvement in pain severity 1 week after the procedure. Further improvement in pain severity was evident at the 6-month recheck with 6 (31.6%) patients reporting complete resolution and 7 (36.8%) patients having had strong improvement. Of the 21 patients, 3 (15.8%) did not have any improvement at the 6-month recheck.¹³

All dogs of the present report required at least 1 repeated injection approximately 4 weeks after the first injection to achieve apparently sustained pain control. In contrast, the condition of all dogs in the study⁹ involving fluoroscopy-guided injections improved after a single injection, and those dogs did not require any further interventions. A possible explanation for the different response was that all dogs of the present report had signs of pain for weeks to months and had already received extensive medical management, whereas the dogs in the other study were only under treatment for 1 to 4 weeks. One of these patients did not receive any pain management, and optimization of medical management was not attempted before the procedure in the other dogs. In humans, repeated paravertebral perineural injections are often required for long-lasting resolution of clinical signs associated with cervical foraminal disk protrusion. Among treated humans, recurrence of symptoms varies; in a recent study,¹² patients returned for repeated injections after a mean interval of 6 weeks (range, 0 to 38 weeks), whereas in another study,¹³ a single injection seemed to result in long-lasting pain control.

Paravertebral perineural nerve blocks guided by various imaging techniques have been used in human medicine. Injections can be performed with fluoroscopy, CT, MRI, or ultrasound guidance.^12,13,14 Jee et al¹⁵ reported no difference in efficacy between ultrasound-guided paravertebral perineural blocks and fluoroscopy-guided intraforaminal blocks in people. The ultrasound-guided technique seems to have many advantages. With ultrasound guidance, soft tissue structures, including the vasculature, can be accurately identified and it is possible to continuously visualize a needle during advancement. This allows accurate needle placement and therefore reduces the risk of intravascular injections or laceration of blood vessels. It also allows monitoring of the spread of deposited local anesthetic. Another advantage of the ultrasound-guided approach in veterinary medicine is that sedation of only a short duration is needed for the procedure, unlike CT- or fluoroscopy-guided interventions that require general anesthesia of patients. Furthermore, less advanced equipment is required, reducing costs, compared with the use of fluoroscopy or CT guidance. Other disadvantages of CT- and fluoroscopy-guided procedures include the greater radiation exposure for both patient and operator. In a prospective trial¹⁴ to compare ultrasound-guided injections with CT-guided injections in people, there was 100% accuracy in needle placement for the ultrasound-guided procedure. Another advantage of ultrasound guidance over CT guidance in that study¹⁴ was significantly faster needle placement (141 seconds vs 643 seconds) associated with the ultrasound-guided intervention.

The main complications documented with the transforaminal approach are inadvertent vertebral artery injections that result in spinal cord and brain stem infarction, most likely attributable to vasospasm or embolus formation.^16,17 In the cases described in the present report, extraforaminal paravertebral perineural injections at the nerve roots were performed to decrease the risk of iatrogenic vascular or spinal injury.

No obvious injection-related complications developed in the dogs of the present report. In humans, paravertebral perineural corticosteroid (with or without local anesthetic) injections are not associated with notable adverse effects or complications.^11,12,18 A prospective study¹⁸ that investigated immediate and long-term complications for 799 fluoroscopy-guided cervical nerve root blocks in humans identified 40 (5%) minor complications. Nineteen patients experienced vasovagal reactions including nausea, diaphoresis, or lightheadedness; 6 patients had signs of sympathetic blockade; 3 patients developed unusual pain; and 1 patient was suspected to have developed a hematoma.¹⁸ Minor local adverse effects, such as bruises, itchiness, or tenderness at the injection site, were noted in another human study¹¹ of cervical periradicular corticosteroid injections. More recently, cervical nerve root injections administered with a combination of fluoroscopy and ultrasound guidance were not associated with any adverse effects.¹² Although none of the dogs of the present report developed adverse effects, the case number was too low to adequately assess potential complications of the procedure. Also, most of the adverse effects, such as mild pain or lightheadedness, documented in reports of human studies are difficult to detect in veterinary patients and could have been easily missed.

Findings for the 4 dogs of the present report suggested that ultrasound-guided paravertebral perineural methylprednisolone acetate injections can be used as an effective treatment for signs of refractory pain associated with foraminal cervical IVDP in dogs when surgery is precluded. Additional studies are required to better assess the overall success rate and risks associated with this technique.