Introduction

Cervical spondylomyelopathy is a complex, progressive disease that results in cervical spinal cord compression in large and giant breed dogs.¹ Generally, the pathogenesis can be divided into disc-associated, osseous-associated, and dynamic compressions.¹ Dorsal cervical laminectomy is the surgical procedure of choice in dogs with static dorsal or dorsolateral cervical spinal cord compressions.² Postoperative seromas can occur after a dorsal (posterior) approach to the cervical spinal cord in dogs and humans.^2–4 In humans, these seromas are associated with the use of recombinant human bone morphogenic protein-2 and demineralized bone matrix allografts and commonly occur subfascially.⁴ In dogs, seromas are mostly subcutaneous and only one case of subfacial seroma has been described in a rottweiler.^2,3,5

Although Doberman pinschers are commonly presented with disc-associated cervical spondylomyelopathy, this case report describes a postoperative seroma after treatment of osseus-associated spondylomyelopathy.⁶ The computed tomographic (CT) appearance, minimally invasive treatment, and outcome are also discussed.

Case Report

A 14 mo old, 30 kg, spayed female Doberman pinscher was examined for difficulty in rising, a wide based stance, progressive pelvic limb gait abnormalities, and cervical pain of 2 mo duration. The clinical signs were unresponsive to previous treatment with mavacoxib^a (2 mg/kg, per os, q 2 wk during the initial mo and then 2 mg/kg monthly). Physical examination at the time of presentation was unremarkable. Neurological examination revealed pelvic limb ataxia and cervical spinal hyperesthesia. Spinal reflexes and cranial nerve examination were normal. Based on these findings, the dog was suspected of pathology localized to either the C1–C5 or the C6–T2 spinal cord segments. Complete blood count, biochemistry, and buccal mucosal bleeding time were all within normal limits

The dog was premedicated for CT by injection of methadone^b (0.1 mg/kg, intramuscular) and dexmedetomidine^c (5 μg/kg, intramuscular). Anesthesia was induced with propofol^d (6 mg/kg, IV, to effect) and after endotracheal intubation maintained with isoflurane^e vaporized in 100% oxygen. The CT was performed (120 kVp and 150 mA) using a four slice helical CT scanner^f to further assess the cervical and cranial thoracic vertebral column. The images were reconstructed using bone and soft tissue algorithms. Initial CT findings indicated bony proliferation of both caudal articular processes including part of the lamina of C6 and both cranial articular processes and part of the lamina of C7. These findings resulted in bilateral lateral and dorsolateral spinal cord compression that was more pronounced on the left side (Figure 1). Post-contrast images after bolus administration of iodinated contrast agent^g (2 ml/kg, IV) did not reveal any additional abnormalities. A diagnosis of bilateral vertebral lamina and articular process proliferation at the level of C6–C7 with secondary compressive myelopathy was made.

View Figure

• Download as Powerpoint
• Download Figure

The dog was admitted to surgery and anesthetized the next day as previously mentioned. Additionally, a constant rate infusion of fentanyl^h (5 μg/kg/h, IV) and a single dose of cefazolinⁱ (15 mg/kg, IV) were administered. The dorsal neck area was clipped and prepared routinely for surgery. The dog was placed in sternal recumbency and, using sandbags, the cervical spine was positioned parallel to the surgical table. A standard dorsal approach to the cervical spine was performed with the skin incision extending from C2 to T3. A limited dorsal laminectomy was performed using an electric-powered burr^j. The laminectomy was centered over the left interarcuate space between C6 and C7 extending cranially and caudally for a distance equal to half the length of each vertebral lamina. The medial border of the laminectomy was located to the left of the base of the spinous process extending laterally to incorporate the medial third of the articular process. The deep border extended to the floor of the vertebral canal (Figure 2). An autologous free fat graft was placed in the laminectomy defect and the central tendinous raphe with its attached muscles was sutured in three layers using simple continuous suture patterns of 2/0 polydioxanone^k. The subcutaneous tissue and skin (intradermal) were closed routinely using 3/0 poliglecaprone 25^l.

View Figure

• Download as Powerpoint
• Download Figure

Postoperative analgesia consisted of methadone^b (0.1 mg/kg, IV, q 4 hr) and gabapentin^m (13 mg/kg, q 8 hr).

The first day postoperatively, the dog showed neurologic deterioration in the form of mild ambulatory tetraparesis. The dog still exhibited spinal hyperesthesia on manipulation of the neck and pelvic limb ataxia. Conscious proprioception was delayed in both pelvic limbs and spinal reflexes were unremarkable. Two days postoperatively, a 2.5 × 7 × 15 cm fluctuant bulge developed in the subcutaneous tissue under the incision line. The dog's neurological status progressively deteriorated, ending up in a non-ambulatory tetraparetic state, with delayed conscious proprioception in all four limbs and decreased withdrawal reflexes in both thoracic limbs by the fourth postoperative day.

The CT examination was repeated and revealed a mild amount of subcutaneous and intermuscular emphysema together with epidural gas located in and around the surgical site. Two well-defined fluid-filled pockets were visible along the trajectory of the surgical approach (Figure 3A, B). The first one was flat in shape (1 × 5 × 16.7 cm) and located subcutaneously, dorsally, and to the left of the midline. The second pocket was bilobed in shape (1.8 × 5.8 × 9.7 cm), located dorsal to the spinous processes of C5–C7, and communicated with the laminectomy site. No communication could be identified between the two pockets. The fluid in both pockets had identical characteristics and was hypoattenuating (mean of 18 Hounsfield units [HU]) and not enhancing after the administration of iodinated contrast, although rim enhancement was present. At the level of the laminectomy site, the spinal cord was moderately compressed, dorsolaterally from the left. On the postcontrast images, there was a patchy increased attenuation dorsally and laterally at the level of the laminectomy, between the spinal cord and fluid pocket that communicated with the laminectomy site. The diagnosis of a subcutaneous and a subfascial seroma with secondary compressive myelopathy was made.

View Figure

• Download as Powerpoint
• Download Figure

An ultrasonographic examinationⁿ of the dorsal and left lateral neck region was performed using a linear array transducer^o operated at 5 megahertz to facilitate safe collection of fluid for cytological analysis. Two well-delineated encapsulated pockets of anechoic fluid matching the CT findings were easily identified. A fine-needle aspirate from the fluid of both fluid pockets was performed. The retrieved fluid was serohaemorrhagic in nature (haematocrit 7% for both) and the cytological findings were consistent with seroma. Aerobic bacteriological culture results were negative.

The dorsal neck region was routinely prepared for surgical drainage of the subfascial seroma. Under ultrasound guidance, a 4.7 mm diameter introducer needle connected to a multifenestrated wound catheter^p was introduced percutaneously to the left of the midline at the craniolateral aspect of the surgical wound, through the long axis of the subfascial seroma and exited the skin caudolaterally on the right side (Figure 4A). The introducer needle was pulled through and out of the skin making sure that the fenestrated part was well localized in the subfascial seroma. The surplus of the wound catheter was cut at the entrance hole and then advanced slightly by traction until the tip of the tube was pulled into the subcutaneous tissue. The entrance hole at the skin entry site was sutured with one simple interrupted suture of 3/0 polyamide^q. The wound catheter was secured to the skin at the exit site with a Chinese fingertrap suture pattern using the same suture material. The introducer needle was cut off from the drain to connect it to a continuous closed suction drainage system^p. Serosanguineous fluid immediately filled the tubing of the wound catheter (Figure 4B). A medical pet shirt^r was placed to prevent patient interference with the drainage system. No attempt was made to drain the subcutaneous seroma.

View Figure

• Download as Powerpoint
• Download Figure

To assess the effect of seroma drainage on spinal cord compression CT of the cervical spine was performed immediately after placement of the wound catheter. The subfascial seroma could no longer be identified on the precontrast images and the spinal cord resumed its normal shape and position. After iodinated contrast administration, local enhancement persisted in the perispinal region at the location of the laminectomy site. The rim enhancement of the subfascial seroma was still visible dorsal to the spinous process of C6 and the subcutaneous seroma was still present.

Within the first 24 hr after wound catheter placement, 52 mL serosanguineous fluid was drained. The dog regained ambulatory function, although still tetraparetic, but cervical spinal hyperesthesia was no longer present. Spinal reflexes in both thoracic limbs were still decreased and conscious proprioception was delayed in the left, but not the right, pelvic limb. The daily amount of fluid drained decreased to 8 mL on the second day and 2 mL on the third day after wound catheter placement. Repeat cytology on those days did not reveal any changes. The wound catheter was subsequently removed.

During the following days, the dog improved neurologically with only mild proprioceptive deficits and slight residual ataxia in both pelvic limbs remaining after 6 days. The dog was discharged from the hospital with instructions for strict cage confinement for the next 3–4 wk.

Five wk after minimally invasive drainage, the dog still showed mild proprioceptive deficits and ataxia in both pelvic limbs. Spinal reflexes in all limbs were unremarkable. There were no signs of cervical spinal hyperesthesia and the subcutaneous seroma was not palpable anymore. The owner reported that the residual ataxia disappeared during the following 4 wk and neurological examination at 9 wk and 1 yr was unremarkable.

Discussion

Vertebral canal stenosis can occur due to osseous changes related to proliferation of the vertebral lamina, articular process, or articular process and pedicle. The current patient had compression of the spinal cord that was mostly from proliferation of the articular process and lamina on the left side and for this reason the slot created for the dorsal laminectomy was limited. The authors' aim was to limit the extent of the approach in an attempt to decrease the postoperative incidence of restrictive fibrosis.

Early postoperative neurologic deterioration is common after dorsal laminectomy in dogs.^3,8 However, progressive worsening of the neurological symptoms and persistent spinal hyperesthesia postoperatively should prompt the surgeon to repeat diagnostic imaging and potentially revise the surgery.^8,9 The approach for a dorsal cervical laminectomy requires extensive soft tissue dissection and seroma formation is a well-known postoperative complication.^3,5,10 Other early postoperative complications include hematoma formation, ischemic injury, inflammation, infection, vertebral fracture, and vertebral luxation/subluxation.² A seroma develops when inflammatory mediator-induced capillary leakage and lymphatic injury lead to the accumulation of serum in surgical dead space.^4,11 The incidence of seroma formation increases with traumatic surgical technique, poor hemostasis, improper anatomical tissue apposition, the presence of foreign debris and irritants, high activity levels postoperatively, and highly mobile regions.^11,12 A small seroma resolves spontaneously within 1–2 wk with exercise restriction that can be combined with warm compress application or therapeutic ultrasound.¹² The untreated subcutaneous seroma in the current patient resolved and was impalpable at the 5 wk follow-up examination. Larger seromas should be drained either via single aspiration followed by compression bandaging or via intermittent aspiration every few days for a period of about 2 wk.¹¹ Rapid refilling of a seroma, within the first 24 hr after drainage, or development of clinical signs as a result of the seroma necessitate surgical drainage.^11,13 The location of the subfacial seroma made the application of a pressure bandage impossible. Intermittent aspiration was also discarded as a treatment option because of the risk of persistent or recurrent spinal cord compression when the seroma would (partially) refill in between aspirations. In addition, there is a risk of introducing bacteria with multiple needle punctures.¹¹ Theoretically, the wound catheter can also be contaminated with external flora when it is introduced through the skin, but daily cytological evaluation of the drained fluid did not reveal any signs of infection.¹⁴ In humans, the use of closed suction wound catheters after lumbar decompressive surgery to prevent postoperative hematoma formation did not increase the risk for postoperative wound infections and may even decrease the incidence of laminectomy membrane formation.^15,16 When compared to passive drainage systems, the closed systems have the benefits of decreasing the risk of infection and skin maceration by drainage fluid, as well as allowing the surgeon to accurately document the amount and consistency of fluid ^17,18 In addition to this, collapse of the dead space during negative pressure seals leaking lymphatics and reduces fluid production.¹⁹ A study in people has failed to identify any neurological deterioration related to wound catheter placement in patients with single-level lumbar laminectomy.²⁰

While the overall diagnostic sensitivity of magnetic resonance imaging (MRI) is superior in identifying soft tissue disorders, it is inferior to CT when evaluating bony pathology.²¹ Because of its low sensitivity for soft tissue pathology, non-contrast-enhanced CT should not be used as a stand-alone diagnostic test.²¹ Sensitivity increases with the administration of IV or subarachnoid (CT myelogram) contrast.²² Advantages of CT compared to low-field MRI include a fast acquisition time, its cost effectiveness, the ability to reconstruct the images in different planes, and its superior image quality.²¹ For these reasons, the authors chose CT over MRI for a first line diagnostic screening. A MRI would have been performed if CT could not identify cervical spinal cord pathology. The MRI (high field) findings of a subfascial seroma after dorsal laminectomy for stenotic cervical myelopathy in a dog have been described previously, but to the authors' knowledge no such report exists describing the CT appearance.⁵ The MRI appearance of a the subfascial seroma appeared as several pockets of high signal intensity on T2-weighted (T2W) images and low signal intensity on T1-weighted (T1W) images, dissecting between the fascial planes and resulting in dorsal spinal cord compression.⁵

On CT examination, the subfascial and the subcutaneous seroma had a similar appearance, both showing a homogenous attenuation (fluid) with a thin rim enhancement (hemorrhage) after contrast administration.²³ These features are not pathognomonic for a seroma and aspiration of the fluid is still advised to differentiate seroma from hematoma or abscess.²³

Certain findings like a thick irregular wall of soft tissue density, that enhances postcontrast administration, and a mixed soft tissue/fluid/gas density filling the lesion during CT evaluation can be more indicative of abscess. ²⁴ The attenuation of a hematoma on the other hand depends on the aggregation of globin molecules and thus can vary depending its age.²⁵ Immediately after hemorrhage, a hematoma has a density of 40–60 HU slowly progressing to 80–100 HU with a hypodense halo as the clot starts contracting.²⁵ Within a few wk, the density of hematomas increase by 0.7–1.5 HU per day from the periphery to the center.²⁵

A postprocedural CT demonstrated the effectiveness of the seroma drainage and documented immediate decompression of the spinal cord.

Conclusion

CT is a good diagnostic modality to evaluate spinal cord compression secondary to subfascial seroma formation after limited dorsal laminectomy. Minimally invasive ultrasound-guided placement of a closed-suction wound catheter can be successfully used to manage a subfascial seroma.