Veterinary Surgery, 18, 2, 123-1 29, 1989

Laminotomy of the Axis for Surgical Access to the Cervical Spinal Cord

A Case Report

JAMES M. FINGEROTH, DVM. DiplomateACVS, and DANIEL D. SMEAK, DvM. DiplomateACVS

A cranially hinged laminotomy of vertebra C2 was used to expose the cervical spinal cord of a dog with a meningioma in the region of the atlantoaxial articulation. By preserving the dorsal atlantoaxial ligament, the technique seemed to result in greater and more physiologic stability between the atlas and axis than dorsal laminectomy and prosthetic replacement of the dorsal atlantoaxial ligament. The procedure allowed a dorsal approach, avoiding injury to the verte- bral arteries and limited exposure, which are potential problems with hemilaminectomy of C1- C2. Further investigation is needed to evaluate long-term consequences of this procedure.

PACE-OCCUPYING LESIONS involving the first three S cervical spinal cord segments and nerve roots are oc- casionally diagnosed as the cause of neurologic distur- bances in dogs. These segments lie within the vertebral canal of the C, (atlas) and C2 (axis) vertebral bodies.’ Many of these lesions are neoplastic and. because the ra- tio of vertebral canal diameter to spinal cord diameter is large here]: such lesions may acquire a relatively large size before they induce neurologic deficits and are diag- nosed. Therefore, wide surgical exposure is indicated for attempted excision. Dorsal laminectomy at the Ci-C2 ar- ticulation may be contraindicated because it disrupts the dorsal atlantoaxial ligament. Although we are not aware of specific biomechanical studies in dogs that demon- strate the contribution of this ligament to overall atlan- toaxial stability, disruption of this structure usually ac- companies traumatic atlantoaxial luxation, suggesting an important stabilizing role in conjunction with the alar, apical, and transverse ligaments of the dens.’-’* Furthermore, prosthetic reconstruction of this ligament is usually the only treatment required for stabilization of atlantoaxial luxations, even though the odontoid liga- ments remain disrupted.(’.’ The dorsal atlantoaxial liga- ment helps prevent excessive flexion of the CI-C2 articu- lation.*-’

* A. De Lahunta. personal communication. March 1988.

Hemilaminectoniy at CI-C2 has been described. usu- ally in conjunction with the surgical repair of atlan- toaxial luxations.3.4 Access provided by this technique is limited, because only one side of the vertebral canal is exposed and the vertebral arteries and branches of the interarcuate arteries lie in close proximity. Overzealous extension of the hemilaminectomy may result in signifi- cant hemorrhage if one ofthese vessels is disrupted. Fur- thermore, lateral retraction of the paraspinal muscles is limited. making manipulation of instruments in this confined space difficult. especially in small patients. Bi- lateral hemilaminectomy is possible. but still provides limited overall access to the vertebral canal.

Ventral approaches such as decompressive “slots” provide access to the ventral aspect of the vertebral canal only. Although instruments may be directed toward the ventrolateral and foramina1 regions through this ap- proach, such manipulations are done with limited visual- ization. Other than for removal of extruded interverte- bra1 discs, or more rarely, odontoidectomy, current ven- tral approaches cannot be advocated for wide exposure of the cervical spinal cord.

In an attempt to achieve the spacious exposure offered by dorsal laminectomy with preservation of the dorsal atlantoaxial ligament and avoidance of major arteries, a *

laminotomy of the axis was devised. In this report, we describe the technique and its use in a representative clinical patient.

From the Department of Veterinary Clinical Sciences, The Ohio State University, Columbus, Ohio. Reprint requests: James M. Fingeroth. DVM, Animal Hospital of Pittsford, 2816 Monroe Ave. Rochester, NY 14618

123

124 LAMINOTOMY OF THE AXIS

Case Report and Surgical Technique

A 5'/2 year old, 8 kg, intact male miniature Schnauzer was referred for evaluation of weakness and episodic col- lapse. The dog had a 1 to 2 month history of stumbling, tiring easily, difficulty in climbing stairs, and inability to stand bipedally. Recently, there were acute episodes of collapse in all four limbs. The dog was ambulatory, albeit tetraparetic. The forelimbs appeared to be weaker than the hindlimbs. Proprioceptive and postural reaction deficits were present in all four limbs and were slightly worse on the right side. Cervical hyperpathia was not found. Results of a cisternal cerebrospinal fluid analysis were protein 58 mg/dl (normal, 10-30 mg/dl) and lym- phocytic pleocytosis (25 leukocyte count/pl; 8% mono- nuclear, 92% lymphocytes). Myelography was per- formed. The appearance of a large filling defect in the spinal cord at CI-Cz was consistent with either an in- tradural-extramedullary or an intramedullary lesion (Fig. 1).

Laminotomy of the axis was performed. The patient was positioned in ventral recumbency with straight alignment of the spine in the sagittal plane. A rigid vac- uum-type apparatus? was used to cradle the head with- out compression of the jugular veins. Adhesive tape was used cranial and caudal to the surgical field to provide additional cervical stabilization. The junction of the head and neck was slightly flexed. Inhalation anesthesia was administered with a volume-cycled, pressure-limited mechanical ventilator. Systemic arterial blood pressure was monitored via a catheter in the dorsal pedal artery.

After aseptic preparation, a midline incision was made from the external occipital protuberance to the spinous process ofthe fourth cervical vertebra. The median raphe between the paired splenius muscles and the dorsal cuta- neous branches of the cervical nerves was exposed and used as a guide in performing a midline dissection. The splenius muscles were divided over the spinous process of C2 and retracted laterally. The paraspinal epaxial mus- cles (biventer cervicis, complexus, and rectus capitis dor- salis major) were incised on each side of the spine of the axis and reflected from the spine and dorsal laminae with a periosteal elevator to the level of the C2-C3 articula- tions caudally and the ventral aspect of the C,-C2 inter- vertebral foramina cranially. The vertebral arteries coursed through the muscles just ventrolateral to this level, between the transverse foramina of C2 and C I . Anastomotic branches between the CI and C2 interarcu- ate arteries were avoided. Self-retaining retractors were used to facilitate exposure. At the cranial end of the dis- section, muscles were elevated from the dorsal atlan-

t Vac-Pac, Kmchner-Collinson, Aberdeen, MD

Fig. 1. Lateral and ventrodorsal myelograms of a 5% year old, 8 kg, intact male miniature Schnauzer with progressive tetraparesis. A large filling defect in the spinal cord is consistent with an intradural-extra- medullary or an intramedullary lesion.

toaxial ligament, but wide elevation of muscles from the dorsal arch of C1 was not performed. Caudally, the mus- cle elevation was continued until the multifidus muscle

FINGEROTH AND SMEAK 125

Fig. 2. Laminotomy of C2. The dorsal at- lantoaxial ligament is left intact. A high- speed pneumatic drill and small round burr are used to make lateral cuts in the verte- bral arch (top). The vertebral artery and its radicular branches are more easily avoided with the laminotomy than with standard hemilaminectomy. A caudal transverse os- teotomy is made with large bonecutting for- ceps (bottom). Placement of the bone cut- ters is aided by scoring the bone with the burr.

and nuchal ligament were exposed. These structures were left intact.

The laminotomy consisted of three cuts in the bone: a longitudinal one on each side of the vertebral arch and a transverse one at the caudal one fourth of the spinous process of the axis (Fig. 2). The dorsal atlantoaxial liga- ment and ligamentum flavum were left attached crani- ally to serve as a hinge. The lateral cuts were made first. Using a high-speed pneumatic drill and small round burr, a groove was cut on each side, just below the junc- tion ofthe lamina and the pedicle, from the C,-C2 inter- vertebral foramen to the caudal one fourth of the spine of the axis. The grooves were burred to a depth that left only a thin inner cortical wall. Irrigation with saline was performed frequently to clear the field of bone dust de- bris, and to prevent thermal injury to the bone or under- lying neural parenchyma. The transverse cut was made perpendicular to the lateral ones with sharp bone-cutting forceps, being careful to prevent the tips of the blades from entering ventral to the inner cortex of the lamina (Fig. 2, bottom). It was helpful to outline this cut by scor- ing the bone with the burr, but the burr was not used to cut through the entire depth of the outer cortical wall. Diploic hemorrhage from these cuts was minimal. lJsing hemostatic forceps as a handle, gentle craniodorsal trac- tion was applied to the spine (Fig. 3 ) . The remaining in- ner cortical bone in the lateral cuts was removed with small bone and dental curettes. As the laminae and spine were gradually levered forward, connecting filaments be- tween the endosteum of the neural canal and the dura mater were gently severed. Care was warranted because the dura mater began fanning out in the region of C, or the C,-C2 articulation to join the endosteum near the fo-

ramen magnum.' At the cranial extent of the lamino- tomy, the fascia between the atlas and axis, which formed a septum over each intervertebral foramen, was incised with a scalpel blade. The laminar flap was rotated 150" to 180" over the atlas. Some resistance was encoun- tered and the ligamentum flavum between C , and C2 was undercut, carefully preserving the dorsal atlantoaxial lig-

Fig 3 Hemostatic forceps are attached to the spinous process of C2 to elevate the laminotomy flap (top) Using the intact dorsal atlan- toaxial ligament as a pivot point, the flap is rotated 150" to 180" cra- niad and wrapped in a moistened surgical sponge After the neurosur- gical procedure is completed, the lamina is rotated back into its normal anatomic location Stabilization is provided by placement of two wire sutures (bottom) (A single wire was used in the patient described in the case report ) Small lateral gaps due to bone loss during the burring process are present

126 LAMINOTOMY OF THE AXIS

ament. The flap was wrapped with a saline-moistened sponge and retracted cranially by leaving the hemostatic forceps attached. The dura mater, C2 nerve roots, and the radicular (interarcuate) vessels were easily visualized and protected from injury.

There were no extradural lesions. A bulge was visible on the right side ofthe dural tube at the CI-Cz interspace. A dorsal durotomy was performed disclosing a large ( 1 .0 X 1.5 X 1.5 cm), firm, white, lobulated mass adhered to the spinal cord ventrolaterally in the region ofthe C2 dor- sal nerve rootlets. After rhizotomy, piotomy, and retrac- tion of the spinal accessory nerve, a cleavage plane was identified and 80 to 90% of the mass was resected. There was no gross evidence of direct spinal cord, vascular, or bony invasion by the mass. Because the tumor was grow- ing from the dura mater, the latter was partially resected. The durotomy incision was not closed. A free fat graft was placed over the spinal cord. The tumor was identi- fied histologically as a benign fibroblastic meningioma.

The laminotomy was closed by rotating the flap back into position, with the spinous process as a landmark. Interdigitation of cancellous bone on the surfaces of the transverse cut provided some stability (Fig. 3). Rigid fix- ation was accomplished with a 22 gauge wire suture in the spinous process (Fig. 3, bottom). The epaxial muscles were rejoined on midline with 2-0 polyglactin 9 10$ and remaining muscles, fascia, subcutaneous tissue, and skin were sutured routinely. Dexamethasone (0.25 mg orally every 12 hours) was administered for 3 days. A padded neck bandage was applied for 14 days.

There was nearly total resolution of the neurologic deficits by day 2, when the dog was released from the hospital. Three days later, clinical signs of head and neck discomfort were treated with aspirin (8 1 - 162 mg orally every 8 hours), and the dog improved within 48 hours.

Progress radiographs were made at week 2 (Fig. 4). The laminotomy flap was still anatomically reduced and the wire fixation appeared to be stable. Some widening of the osteotomy gap was interpreted as part of a normal bone healing process. Clinically the dog was reported by the owner to be near normal. It had resumed jumping on and off furniture and going up and down stairs. It was able to stand bipedally, and the owner no longer ob- served scuffing of the toes when it walked. No clinical signs of neck pain were reported. The dog was reexam- ined at month 6 as a long-term follow-up and to diagnose the cause of an acute onset of neck pain. No neural defi- cits were detected. An area of firm swelling and mild clin- ical signs of discomfort were found on physical examina- tion of the right side of the neck. A displaced fracture of the 'ght wing Of the Fig. 4. Radiographs made at week 2. The wire suture is visible at the

site of the laminotomy. Healing activity is inferred from the presence was of a widened gap at the caudal osteotomy. The fixation appears to be anatomic and stable. $ Vicryl. Ethicon, Somerville, NJ.

FINGEROTH AND SMEAK 127

Although the owner could not recall a specific traumatic event. the injury appeared to be unrelated to the surgery. The laminotomy site was completely healed and remod- eled. There was some ventral slippage of the bone flap. but minimal compromise of the vertebral canal lumen at that site (Fig. 5).

Discussion

Limited space and the close proximity of major vessels sometimes make it difficult to treat spinal cord lesions at the level of CI-C2 even with liberal hemilaminectomy. Dorsal laminectomy has been avoided because loss of the dorsal atlantoaxial ligament may cause instability. Be- cause tumors in this region are frequently benign and noninvasive, it is important to have a reliable method of obtaining wide surgical access to the neural canal in or- der to perform atraumatic excision and decompression. Laminotomy of C2 was used to meet these needs without sacrificing stability of the atlantoaxial joint.

As with all intraspinal surgery, it is important when positioning the patient to avoid pressure on the jugular veins, because increased jugular venous pressure may cause increased filling and pressure within the vertebral venous sinuses.’ When operating on the cranial cervical spinal cord it is especially advantageous to monitor vital signs closely, particularly respiration, heart rate. and sys- temic arterial blood pressure. Ventilatory support should be available because of potential complications associ- ated with manipulating the upper cervical spinal cord segments.

A high-speed pneumatic drill is preferred for the lateral cuts in the laminotomy because it allows precise control over direction and depth. A small burr is preferred, espe- cially in small patients, to minimize the resultant gap af- ter closure. Sharp bone cutting forceps were used for the transverse cut, to avoid bone loss resulting in a gap after closure or traction on the atlantoaxial joint as the wire suture was tightened. Bone cutting forceps provide good control over direction and depth. Large forceps should be used because smaller bone cutters may fracture the spinous process as the jaws initially compress but do not cut through the bone.

The site of the transverse cut should be chosen care- fully. Too far caudal a cut could disrupt the nuchal liga- ment and might not leave a sufficient segment of dorsal arch for placement of the stabilizing wire(s). Making the cut too far cranially shortens the lever arm, making rota- tion of the fragment difficult and decreasing the size of the opening.

Lesions cranial to the CI-C2 interspace are still prob- lematic. Although laminotomy of C2 affords excellent vi- sual and mechanical access to the caudal aspect of C, .

additional exposure is needed to enter the C, vertebral foramen. Laminotomy of C2 combined with hemilami- nectomy of C I preserves the dorsal atlantoaxial ligament, exposes the C, vertebral foramen, and provides better overall access to the area than CI-C2 hemilaminectomy alone. However, the choice of surgical approach should be based on interpretation of the clinical examination and neuroradiographic data. In some cases, hemilami- nectomy may be adequate or even preferred over lami- notomy of C2.

The decision to remove only 80 to 90%) of the tumor was based on balancing the risks of iatrogenic spinal cord and vascular injury as the ventral portion of the neural canal was probed and the risks of leaving behind what was presumed to be a slow-growing benign tumor. In an earlier study of spinal meningiomas, good results were obtained, even with only subtotal resection, as long as significant spinal cord decompression was achieved.” Ce- rebral meningiomas have responded to radiation ther- apy,” and it is likely that such therapy might also be effective for residual meningioma tissue left in the spine. Unfortunately, the upper cervical spine is difficult to irra- diate without exposing other viscera in the neck (e.g.. thyroid glands) to high doses of ionizing radiation. Therefore, regrowth of tumor and recurrence of clinical signs might be seen after subtotal resection alone, and the preferred treatment for such cases may be reoperation.

In the case presented, a fat graft was placed over the spinal cord before closure of the laminotomy. This was done to cover the defect in the dura mater, and need not be considered a normal part of the laminotomy proce- dure. A n alternative would have been to suture a dural replacement over the defect to achieve a water-tight seal. This is a technically demanding procedure requiring mi- croscopic guidance for best results. Because clinical problems related to cerebrospinal fluid leakage were not recognized in other dogs with spinal meningiomas,’ we elected to proceed in the manner described. It is possible, but unprovable, that the clinical signs of head and neck pain observed by the owner at home were related to cere- brospinal fluid leakage through the open dura.

Drilling holes in the spinous process for placement of the stabilizing wire was similar to that described during repair of atlantoaxial I u x a t i ~ n s . ” ~ . ~ . ~ . ’ It is important to drill the holes at the same levels of the spinous process to avoid subluxation of the flap as the wires are tightened. Because of the size, configuration, and presence of can- cellous bone in the edges of the laminae and spinous pro- cess (Fig. 3 ) , rotational instability is not a major problem even though there are thin gaps on either side along the lateral laminar cuts. However. two wires might be pre- ferred to eliminate any tendency for torsion of the lami- nar flap, or ventral slippage as occurred in this dog. Only

'1 28 LAMINOTOMY OF THE AXIS

Fig. 5. Lateral and ventrodorsal radio- graphs at month 6. The laminotorny flap has slipped ventrally, but appears to have mini- mal encroachment on the vertebral canal. Bony union and remodeling appear com- plete. Fixation with two wire sutures might have prevented this complication. There is an unrelated displaced fracture of the right wing of the atlas (arrow).

FINGEROTH AND SMEAK 129

a single wire was used for stabilization in this dog because of its small size and because of difficulty manipulating the drill between the retracted epaxial muscles and the bone.

References

I . Evans HE. Christensen GC. Miller 4 .4naiomj~ q/ /he Do,? 2nd ed. Philadelphia: WB Saunders Co, 1979:939-942.

2. Sorjonen DC. Shires PK. Atlantoaxial instability: a ventral surgical technique for decompression. tixation. and fusion. Vet Surg 1984: 10:22-29.

3. Geary JC, Oliver JE, Horlein BF. Atlantoaxial subluxation in the canine. J Small Anim Pract 1967;8:577-582.

4. Oliver JE, Lewis RE. Lesions of the atlas and axis in dogs. J Am Anim Hosp Assoc 1973;9:304-312.

5 . de Lahunta .A, Habel RE. eds. fpp/ /d l ~ ~ , / c ~ r . / ~ i u ~ j . , , I w r r / ~ , r i ~ ! . Phila- delphia: WB Saunders Co. 1986:73.

6. Cook JR. Oliver JE. Ailantoaxial luxation in the dog. Compend Cont Educ Pract Vet 1981:3:242-249.

7. Dvorak J. Panjabi MM Functional anatomy ofthe alar ligaments. Spine 1987: 12:183-189.

8. Sorjonen DC. Small animal cervical surgery. In: Oliver JE, Horlein BF. Mayhew lC. eds. I 'c/cvYnurl. N c w o l o g j ~ Philadelphia: WB Saunders. Co. 1987:442.

9. Fingeroth JM. Prata RG. Patnaik AK. Spinal meningiomas in dogs: 13 cases (1972-1987). J Am Vet Med Assoc 1987:191: 720-726.

10. Turrel JM. Fike JR. LeCouteur RA. ct al. Radiotherapy of brain tumors in dogs. J A m Vet Mcd Assoc 1984: I84:82-86.

1 I . Chambers J N . BettsCW. Oliver JE. The use ofnon-metallic suture material for stabilization of atlantoaxial subluxations. J Am Anim Hosp Assoc 1977: I3:602-604.

Abstract of Current Literature

COMPARISON OF OVER-THE-TOP AND TUNNEL TECHNIQUES FOR ANTE- RIOR CRUCIATE LIGAMENT REPLACEMENT Montgomery RD, Milton JL, Terry GC, McLeod WD, Madsen N Clinical Orthopaedics und Rclutcd Rcwurc,h 1988;23 I : 144- 153

The osseous tunnel technique of anterior cruciate ligament (ACL) reconstruction was com- pared with the over-the-top (OTT) technique in vivo. The ACL was replaced by a prosthesis in 35 dogs by one of three methods: double osseous tunnels. OTT. or a combined method using the unique Y shape of the prosthesis by placing one fork through a femoral osseous tunnel and the other fork over the top of the femoral condyle in the same stifle. Rupture occurred in all prostheses placed in double osseous tunnels and in one of- 12 prostheses placed by the OTT technique. When rupture occurred in only one fork ofprostheses placed in the combined method, all occurred in the tunnel-placed fork. If the results of the three groups are combined, rupture occurred in 16.7% of all OTT-placed prostheses V.S. 9 1.39'0 rupture of all osseous tunnel-placed prostheses. Objective assessment of the in v i w testing clearly indicates that the OTT technique is superior to the tunnel technique under the conditions of this study. The relationship between placement technique and rupture of other prostheses should also be tested in vim