Introduction

A dermoid sinus is characterized by the incomplete separation of skin from the neural tube during embryonic development that can open into tracts on the skin along dorsal midline and is associated with Rhodesian ridgebacks.^1–20 The dermoid sinus is a tube-like tract lined by hair follicles, sweat, and sebaceous glands extending into the subcutaneous tissue or through the spinal canal to the dura mater.^{9,11–13,16,18,19} The sinus can enlarge as a result of infection causing abscessation, and dogs commonly present with a cutaneous lesion or signs of infection.^1,6,13 Surgical excision is recommended because of the potential for connection to the dura mater, risk of infection, or inflammation of the spinal cord, and it is typically curative.^{1,4–8,11–13}

A sinus tract may extend to the spinal canal, so a dorsal laminectomy may be required to remove all abnormal tissue.^12,15,16 In rare cases, dogs have multiple vertebral and spinal malformations associated with neurological deficits.^5,12–14 These can include neural tube defects, dermoid sinuses, and occult spina bifida, with spina bifida being related to dermoid sinuses that reach the spinal canal.^5,12,15,20

While dermoid sinuses are described in dogs, there is no consensus on optimal imaging although a number of relatively expensive or invasive methods have been proposed, including computed tomography (CT), MRI, fistulography, and myelography.^{1–5,7–10,12–15,17} Each technique has its advantages and disadvantages with complete characterization of the lesion not being guaranteed.^{1–5,7–10,12,14,15,17}

Our objective was to document two cases of dermoid sinuses in which single-phase contrast CT alone was used to successfully identify and characterize the extent of bony and soft tissue lesions associated with dermoid sinuses prior to surgical exploration. This case report also documents a dermoid sinus in an atypical location in a previously unreported Rhodesian ridgeback mix.

Case Reports

Case 1

A 6 mo old male castrated bloodhound–Rhodesian ridgeback was referred for a suspect thoracolumbar dermoid sinus. When the dog was adopted at 4 mo of age, the owner noted a soft tissue mass along the dorsum, which did not respond to empirical antibiotics. At 6 mo of age, the dog was referred to the North Carolina State College of Veterinary Medicine for further evaluation. On examination, there was a nonpainful, 2 cm subcutaneous soft tissue mass at the thoracolumbar junction with a palpable stalk at the cranial aspect that extended into deeper tissues. Neurological and general physical exam were otherwise unremarkable.

Diagnostic imaging was recommended to assess the extent of the mass and guide surgical planning. A complete blood cell count and blood chemistry performed 1 mo prior to referral were unremarkable. Food and water were withheld for 12 hr prior to anesthesia. Maropitant citrate^a (1 mg/kg IV) was administered preoperatively. After premedication with hydromorphone^b (0.1 mg/kg intramuscularly [IM]) and dexmedetomidine hydrochloride^c (3 mcg/kg IM), anesthesia was induced with propofol^d (6 mg/kg IV). Anesthesia was maintained with isoflurane^e in oxygen through an endotracheal tube with the assistance of mechanical ventilation. Intravenous fluids using lactated ringers solution^f (5 mL/kg/hr) was administered throughout anesthesia. A CT scan of the thoracolumbar spine was performed using a 64 slice multidetector CT scanner^g. A dose of 2 mL/kg (700 mgI/kg) of nonionic iodinated contrast medium^h was administered intravenously with a power injectorⁱ at 3 mL/sec into the cephalic vein. Images were acquired as a transverse multislice dataset before and after contrast administration using the following technical parameters: 130 kV, 190 mAs/slice, pitch 0.65, slice thickness 1.0 mm. The multislice data set was reconstructed into 3.0mm and 1.0mm transverse sequences during pre- and postcontrast administration. Sagittal and dorsal plane sequences were reconstructed post contrast.

The CT study showed a pedunculated soft tissue attenuating subcutaneous mass (4 cm W × 3.3 cm L × 1.8 cm H) in the thoracolumbar region, with a fluid-attenuating center. A thin stalk tapered cranioventrally through subcutaneous tissues and the epaxial musculature, extending through a tubular defect in the spinous process of T12 and terminating within the vertebral canal (Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

Surgical excision was performed immediately after the CT scan. At the start of surgery, a fentanyl^j constant rate of infusion (3 mcg/kg/hr) was initiated and cefazolin^k (22 mg/kg IV) was administered, followed by a dose every 90 min during surgery. The dog was placed in sternal recumbency. The thoracolumbar area of the dorsum was aseptically prepped, and a pinpoint skin defect noted (Figure 2A). A 2 cm elliptical skin incision was made around the mass and extended cranially and caudally by 3 cm. The subcutaneous tissues and epaxial muscles around the stalk identified cranially were dissected as it tapered toward T12 (Figure 2B). A hemostatic clip^l was placed on the stalk at the dorsal lamina of the vertebrae, and it was transected. After transection, hair follicles were identified in the lumen, so a partial dorsal laminectomy was performed with a pneumatic burr^m, and the stalk was dissected to the vertebral canal and transected (Figure 2C). No additional hair follicles were noted. The epaxial fascia and subcutaneous tissue were apposed using 3-0 polydioxanoneⁿ in a simple continuous pattern. An intradermal pattern was performed using 3-0 poliglecaprone^o and skin sutures were placed in a cruciate pattern using 3-0 nylon^p. The sinus tract was submitted for histopathology in formalin (Figure 2D). No neurological signs were noted after surgery. The dog was discharged the day after surgery with carprofen^q (2.2 mg/kg per os q 12 hr) and a fentanyl transdermal patch^r for 3 days postoperatively. Postoperatively, the dog was confined to a crate and limited to leash walks for 2 wk.

View Figure

• Download as Powerpoint
• Download Figure

Histolopathology revealed a deep extension of a tract of dermal collagen with adnexal structures and a dilated tubular structure lined by keratinized stratified epithelium filled with hair shafts and keratin debris. This was consistent with a dermoid sinus.

A year after discharge the client reported that the dog developed a small, self-limiting seroma at the surgical site. Patient had otherwise done well.

Case 2

A 2.5 yr old male castrated Rhodesian ridgeback was evaluated for a dorsal cervical dermoid sinus, which was noted during an exam at 4 mo of age. No lesions were detected on cervical radiographs and the mass was surgically removed, at which time a stalk into deeper tissues was noted. Histopathology was consistent with a dermoid sinus and the mass grew back within 3 mo. The dog was referred to the North Carolina State College of Veterinary Medicine for further evaluation. On examination, there was a nonpainful, 3 cm subcutaneous soft tissue mass on the dorsal midline of the cranial cervical region, 5 cm caudal to the occipital protuberance. The neurological and physical examinations were otherwise unremarkable.

Single-phase contrast CT and surgery were recommended. A complete blood count and blood chemistry were unremarkable. Food and water were withheld for 12 hr prior to anesthesia. After premedication with hydromorphone^b (0.1 mg/kg IM) and dexmedetomidine^c (3 mcg/kg IM), anesthesia was induced with midazolam^r (0.2 mg/kg IV) and propofol^d (2 mg/kg IV). Anesthesia was maintained with isoflurane^e in oxygen via an endotracheal tube using mechanical ventilation. Intravenous fluids with lactated ringers solution^f were administered during anesthesia. The dog was placed in sternal recumbency and a CT scan of the cervical spine was performed using a 64 slice multidetector CT scanner^g as described for case 1.

The CT identified a mass (3.5 cm W × 2 cm L × 1.8 cm H) in the craniodorsal cervical subcutaneous tissues with a thin soft tissue attenuating rim and a fluid-attenuating center. A thin tract extended cranioventrally through the subcutaneous tissues and epaxial musculature, terminating between C1 and C2. A midline defect was noted in the cranial aspect of the first cervical vertebrae arch, consistent with spina bifida, but was not associated with the dermoid sinus tract further caudally (Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

Surgical excision was performed. The dog was placed in sternal recumbency with the cervical region ventroflexed. At the start of surgery, a fentanyl^j constant rate of infusion (3 mcg/kg/hr) was initiated and cefazolin^k (22 mg/kg IV) was administered, followed by a dose every 90 min during surgery. The skin over the cervical thoracic vertebrae was aseptically prepped; a 5 cm elliptical skin incision was made to include the surgical scar. The stalk was identified cranially, and the subcutaneous tissues were dissected from around the mass and stalk through the dorsal epaxial musculature. The stalk extended through the biventer cervicis and rectus capitus muscles and continued to the dorsal laminae of the first cervical vertebrae and terminated in the dorsal atlanto-axial ligament between the first and second cervical vertebrae. A metal hemostatic clip^l was placed at the level of the atlanto-axial ligament and the stalk transected. The epaxial musculature and subcutaneous tissue were apposed using 3-0 polydioxanoneⁿ in a simple continuous pattern. An intradermal pattern was performed using 3-0 poliglecaprone^o. The skin was apposed using stainless steel skin staples. Thick, tan fluid from within the sinus was submitted for aerobic and anaerobic culture. The sinus tract was submitted for histopathology in formalin. Postoperatively, no neurological signs were noted. The dog was discharged the day after surgery with fentanyl^k transdermal patch 3 days postoperatively and 1 wk of clavamox^s (13.75 mg/kg per os q 12 hr). The dog was confined to a crate and limited to leash walks for 2 wk.

The fluid culture was negative for aerobic and anaerobic growth. Histopathology revealed a disrupted wall of a cyst characterized by a thin layer of stratified squamous epithelium that keratinized through a granular cell layer. Around the sinus there was a thick layer of mature, dense collagen that paralleled the wall. This was consistent with a dermoid sinus.

Owner reported that the dog was clinically normal at 10 mo postop, at which time, the dog was lost to follow-up.

Discussion

Despite the atypical breed and thoracolumbar location in case 1, a dermoid sinus was suspected in both cases. Physical examination and history suggested deeper extension of the sinus toward the spinal canal in both dogs, although neither had neurological abnormalities. Veterinary literature suggests myelography or fistulography are indicated to characterize dermoid sinus extension, especially when spinal cord involvement is suspected.^{3–5,7–10,13,15,17–19} However, we were hesitant to use these techniques because of the potential morbidity of contrast myelography and the risk of potential myelitis after infusing contrast into the contaminated cutaneous exit point of a dermoid sinus.¹ Instead, a rapid and noninvasive imaging plan was elected using single-phase contrast CT scan. In both dogs, pre- and postcontrast CT images provided accurate and detailed images of the location, vertebral anomalies, and extension of the sinus.

Different diagnostic modalities are reported in veterinary and human medical literature to assess dermoid sinuses, which include radiographs, CT, MRI, fistulograms, and myelograms.¹⁹ Although radiographs can be done, they can be unremarkable even if a dermoid sinus is present (as in case 2) or show soft tissue swelling or vertebral abnormalities.^{1,2,5–8,10,13–15,18,19} Myelography or fistulography have been used to show the extent of the tract, but it has been suggested that injecting contrast material or probing the tract should be avoided due to risk of inducing mengingitis or injurying underlying structures.¹ Additionally, fistulography may not show the extent of the sinus tract if debris is present.¹³

The diagnostics most helpful in diagnosing and defining the extent of the dermoid sinus is MRI and CT because they delineate the tract and show cross-sectional images, although the extent of the pathology may still not be visualized until surgery.^1,3,4,10,17 In these cases, a CT scan was used, although MRI is historically more common in human and veterinary medicine.^{1–3,9,10,14,17–19}

MRI is noninvasive but has limited availability in veterinary medicine and a relatively high cost.¹⁷ In earlier reports in which standard MRI images were used to assess spinal abnormalities, it allowed the detection of extra-spinal portion of the sinus tract, intramedullary tumors and cord tethering, although the images are not reliable in identifying the intraspinal extension of the tract, extramedullary tumors or termination of the tract.^1,3,17 In some earlier reports, the tract had the potential to be missed on MRI if it was out of the imaging plane and images could be difficult to interpret with an infection or small tract.^1,3,9 In some cases, the extent of the dermoid sinus was unable to be determined on MRI images.^10,12,18,19 False positives have also been reported with MRI, with a blood vessel being mistaken for a deep tract.^9,14,17 Although new MRI technology may have negated some of the limitations of using this modality for imaging of dermoid sinuses, this historical lack of sensitivity in detecting the intraspinal portion of the sinus tract and extramedullary tumors raises questions regarding the optimal way to image dermal sinuses.^3,9

Preoperative CT scans have been used to define the extent of dermoid sinuses, detect soft tissue swelling and associated bony abnormalities such as block vertebrae, hemivertebrae, and spina bifida.⁵ In human medicine, a CT scan and radiographs are used for nasal dermoid sinuses.⁸ When a CT scan has been used to assess nasal dermoid sinuses, the extent of the sinus and the presence of bifidity are easily discerned.¹¹ Reports also suggest that sometimes CT is preferred to MRI to characterize the degree of bony involvement and intracranial extension.¹¹ In the cases presented here, a single-phase contrast CT scan was performed to evaluate the dermoid sinus. On the CT scan, the dermoid sinus tract was a contrast enhancing lesion, and the complete extent of each tract was established without myelography. Surgical findings confirmed the accuracy of these findings. Cross-sectional CT imaging allowed the detection of concurrent spina bifida in both dogs, providing information about the pertinent anatomy prior to surgery. It should be noted that the image detail acquired using a 64 slice CT scanner is extremely high, and this level of detail contributed to the diagnostic quality of the images. Thus, although CT imaging is more widely available and cost effective as compared with MRI, it is not clear that CT scanners with less resolution will provide the same level of anatomic detail in imaging dermoid sinuses.

The dermoid sinus in case 1 was in the thoracolumbar region with extension to the 12th thoracic vertebrae, a location not previously reported. Previously, dermoid sinuses in the dog were reported in the cervical, cranial thoracic, lumbosacral, and sacrococcygeal, with the sacral and nasal regions being less frequent.^{2,4–10,13–17,19,20} The cervical region is the most common area, as was seen in case 2.^4,6,13,18 Typically in dogs, dermoid sinuses have been cranial and caudal to the ridge.^6,13,20 All but one dermoid sinus reported in breeds other than Rhodesian ridgeback occurred cranial or in the cranial thoracic region and none at the thoracolumbar junction.^7,13,16,18 It is unknown whether the atypical breed (bloodhound mix) contributed to the atypical location.

The dermoid sinus is a hereditary congenital neural tube abnormality in the Rhodesian ridgeback.^{4–7,9,12,19} Although it is most common in Rhodesian ridgebacks, it has been reported in Rhodesian ridgeback mixes and other breeds, although not a Rhodesian ridgeback–bloodhound mix.^{4,6,12,13,15–18} Other breeds include Yorkshire terrier, shih tzu, boxer, golden retriever, chow chow, English bulldog, Boerboel, Kerry blue terrier, English springer spaniel, English cocker spaniel, Siberian husky, Great Pyrenees, German shepherd dog, rottweiler, Swedish vallhund, Chinese crusted, Dalmatian, Victorian bulldog, American cocker spaniel, dachshund, and domestic cats.^{2,4–6,8,10,14,15,18,19} The genetic transmission of dermoid sinuses in the Rhodesian ridgeback is complex, irregular in expression, and probably dependent on several genes.^9,12,13,20 Although there is debate about genetic transmission, if dermoid sinuses are recessive, then the bloodhound would have to have been a carrier. As shown by case 1, a dermoid sinus should not be ruled out based on signalment.

Conclusion

In conclusion, single-phase contrast CT should be considered as a minimally invasive diagnostic modality when determining the extent of a dermoid sinus in dogs. Application in larger numbers of dogs and with other imaging modalities would allow further evaluation of the sensitivity and specificity of this technique.