Zygomatic sialocele secondary to infarction treated with sialoadenectomy in a dog

Abstract

Zygomatic salivary gland disease is not commonly reported in dogs and there is a paucity of literature reporting salivary gland disease secondary to infarction in dogs. A 9-year-old German wirehaired pointer presented with left eye exophthalmos, 3rd eyelid elevation, negative retropulsion, and pain upon opening of the mouth. Computed tomography revealed a mass extending from the left zygomatic salivary gland, consistent with a sialocele. A left-sided zygomatic sialoadenectomy was performed successfully. Histopathologic diagnosis concluded zygomatic salivary gland infarction. The dog had no signs of recurrence 20 mo after surgery.

Key clinical message:

To the authors’ knowledge, this is the first case report with long-term outcome of a zygomatic sialocele secondary to salivary gland infarction in a dog treated by zygomatic sialoadenectomy via zygomatic osteotomy.

Salivary gland and duct disease is not commonly reported in cats or dogs. Salivary gland and duct disease represented less than 0.3% (245 of 87,392) of diagnostic case submissions presented to 1 laboratory (1), spanning 41 mo. In dogs, the main salivary gland disorders include primary or secondary neoplasia, followed by nonspecific sialoadenitis/sialoadenosis, sialocele, necrotizing sialometaplasia, and sialolithiasis (1,2). In a study of 245 canine and feline histopathologic salivary gland samples, malignant neoplasms were detected in 30%, sialadenitis in 26%, normal salivary glands in 16%, sialoceles in 9%, and salivary gland infarct in 8%. The remaining 11% had various other causes, including degenerative or fibrotic lesions, ductal ectasia, sialolithiasis, edema, benign neoplasia, and secondary salivary involvement with systemic or cervical lymphosarcoma or with sarcoma in the head and neck (1). A sialocele is an accumulation of mucoid saliva in subcutaneous or submucosal tissue without an epithelial lining (3), due to a tear in the salivary glandduct complex (1). Sialoceles may be caused by trauma, sialolith, penetrating foreign body, neoplasia (4), immune-mediated disease (5), tooth extraction (6), and infarction (7). Although a sialocele can occur in any of the major or minor salivary glands, the most often reported are the sublingual and mandibular salivary gland complex, with rare reports of zygomatic sialoceles (1).

Clinical signs of zygomatic sialoceles include divergent strabismus, trismus, retrobulbar swelling, a painless orbital swelling, 3rd eyelid protrusion, and exophthalmos (3). Causes of zygomatic sialoceles are the same as for any sialocele; however, a zygomatic sialocele following a caudal hemimaxillectomy has also been described (8). Salivary infarct is a rarely reported cause of a zygomatic sialocele (9). Causes of salivary infarcts are not well-understood and there is a paucity of literature on zygomatic salivary infarcts.

There are both surgical and non-surgical treatments described for the treatment of sialoceles. Non-surgical treatment options include phenobarbital administration (10), corticosteroid therapy (11), and intracanalicular injections of N-acetylcysteine (12). Surgical treatment involves removal of the affected salivary gland. Zygomatic sialoadenectomy may be performed via a modified lateral orbitotomy (13), zygomatic arch osteotomy, or ostectomy. To the authors’ knowledge, this is the first report of long-term follow-up of zygomatic sialoadenectomy via zygomatic osteotomy for treatment of zygomatic sialocele secondary to infarction in a dog.

Case description

A 9-year-old spayed female (28.5 kg) German wirehaired pointer was presented to the University of Minnesota Veterinary Teaching Hospital with acute exophthalmos and chemosis of the left eye 4 d following a routine ovariohysterectomy. The patient had a slight serous discharge, protruded nictitating membrane, and blepharospasm of the left eye that had begun on the day of presentation. On physical examination, the dog had a temperature of 39.6°C, pain upon opening of the oral cavity, lateral strabismus of the left eye, asymmetric retropulsion, a Grade II/ VI left systolic murmur with no arrhythmia, enlarged mammary glands with no masses or pain on palpation, and an unremarkable ventral abdominal incision from the previous surgery.

Complete blood (cell) count (CBC) revealed a mild leukocytosis [21.01 × 10³ μL, reference range (RR): 3.88 to 14.57 × 10³ μL) characterized by a mild mature neutrophilia (18.7 × 10³ μL, RR: 2.10 to 11.20 × 10³ μL), and mild monocytosis (1.26 × 10³ μL, RR: 0 to 1.20 × 10³ μL). Rare large platelets and platelet clumping were noted. All other values were within normal ranges. A chemistry profile revealed a mildly elevated alkaline phosphatase (ALP) 159 U/L (RR: 8 to 139 U/L), and glucose 120 mg/dL (RR: 75 to 117 mg/dL). All other values were within normal range. Thoracic radiographs revealed a moderate bronchial pattern throughout the lungs (suspected age-related changes); gas in the peritoneal space (suspected from recent laparotomy); and a 6.0 × 4.0 cm soft tissue opacity extending from the ventral thoracic wall (suspected to arise from mammary tissue). There was no radiographic evidence of cardiovascular disease.

The dog was placed under general anesthesia for pre- and post-contrast computed tomography (CT) (CT angiography) under the supervision of a Board-certified veterinary anesthesiologist. The dog was co-induced with fentanyl (Sublimaze; Hospira, Lake Forest, Illinois, USA), 115 μg (3.9 μg/kg BW, IV) and propofol (PropoFlo; Diprivan; Sagent Pharmaceuticals, Schaumburg, Illinois, USA), 50 mg (1.7 mg/kg BW, IV). Anesthesia was maintained with isoflurane 0.75 to 1.5% in 100% oxygen, and a constant rate infusion (CRI) of fentanyl (Sublimaze; Hospira), 5 μg/kg, BW, per hour, IV; Lactated Ringer’s Solution (LRS) (Baxter Healthcare, Deerfield, Illinois, USA) was administered for the duration of the procedure at a rate of 4 mL/ kg, BW, per hour, IV. Anesthetic complications included bradycardia, which was treated with atropine (VetOne, Boise, Idaho, USA), 0.46 mg (0.016 mg/kg BW, IV) and a delirious recovery, which was treated with propofol (PropoFlo; Diprivan; Sagent Pharmaceuticals), 20 mg (0.7 mg/kg BW, IV) and dexmedetomidine (Dexdomitor; Zoetis, Kalamazoo, Michigan, USA), 15 μg (0.52 μg/kg BW, IV).

A CT of the skull revealed a well-marginated, homogeneously soft tissue/fluid attenuating structure, 26.8 (L) × 21.3 (H) × 19.8 mm (W), causing retro-dorsal displacement of the left globe and dorsal displacement of extraocular muscles and optic nerve (Figures 1, 2). Left zygomatic salivary gland volume seemed decreased compared to the right side. The findings concluded a non-contrast enhancing structure within the left retrobulbar space most consistent with a thin-walled cyst, appearing to be extending from the left zygomatic salivary gland. This finding was consistent with a sialocele with no identified cause; neoplasia was considered unlikely. Incidentally, the crown of the left maxillary canine (#204) was fractured, without widening of the pulp cavity. The tips of crowns of both mandibular canine teeth (#304 and 404) were also missing and flattened. These were presumed to be unrelated chronic changes. The left exophthalmos, therefore, was secondary to a retrobulbar mass effect with edema of the surrounding adipose and soft tissues.

Figure 1.

Sagittal (a) and coronal (b) CT images of the skull. Arrows indicate non-contrast-enhancing structure within the left retrobulbar space, most consistent with a thin-walled cyst.

Figure 2.

Sagittal (a) and axial (b) CT images of the skull. Arrows indicate non-contrast-enhancing structure within the left retrobulbar space, most consistent with a thin-walled cyst.

A left-sided zygomatic sialoadenectomy was recommended. The dog was placed under general anesthesia. Co-induction was initiated with fentanyl (Sublimaze; Hospira), 290 μg (10 μg/k g BW, IV) and propofol (PropoFlo; Diprivan; Sagent Pharmaceuticals), 90 mg (3.1 mg/kg BW, IV). Regional anesthesia (infraorbital block) was performed with a combination of bupivacaine (Sensorcaine; Marcaine, Hospira, Lake Forest, Illinois, USA), 10 mg (0.35 mg/kg BW) and dexmedetomidine (Dexdomitor; Zoetis), 5 μg (0.17 μg/kg BW). The dog was maintained on isoflurane 1 to 2%, in 100% oxygen and CRI of fentanyl (Sublimaze; Hospira), 10 μg/kg, BW, per hour, IV, and ketamine (Ketaset; Ketaflo; Vetalar; Zetamine; VetOne, Boise, Idaho, USA), 1 mg/ kg, BW, per hour, IV; Lactated Ringer’s Solution (Baxter Healthcare Corporation) was administered at 5 mL/ kg, BW, per hour, IV, for the first hour, and 3 mL/kg, BW, per hour, IV, thereafter. Cefazolin (Ancef; Kefzol; Zolicef; Apotex, Weston, Florida, USA), 22 mg/kg BW, IV, was administered 20 min pre-operatively and readministered every 90 min thereafter, until completion of the procedure.

A 15-cm horizontal incision dorsal to the zygomatic arch was made and Metzenbaum scissors were used to dissect further, taking care to preserve the optic ligament, as reported (3,14). The masseter muscle attachment to the ventral zygomatic arch was incised and reflected off the ventral surface of the arch. Prior to osteotomy, 4 drill holes were made (2 cranial and caudal at the proximal and distal positions) approximately 1 cm away from the cranially proposed osteotomy site, as well as the caudally proposed osteotomy site (8 holes in total) using a 2.5-mm drill bit. The cranial and caudal osteotomy were made and the osteotomized zygomatic arch was elevated and retroflexed dorsally. The masseter muscle was reflected ventrally and during this process, dark red gelatinous to viscous fluid egressed from deep to the masseter (Figure 3), consistent with inspissated saliva, which was submitted for aerobic culture and sensitivity. Any abnormal glandular tissue was removed using a combination of blunt and sharp dissection until remaining tissue was grossly normal; the surgical site was then copiously lavaged with warm sterile isotonic saline. Cerclage wire (22-gauge) was pre-placed in all drill holes (4 wires total) and secured by twisting and pulling simultaneously. The wires were trimmed and flattened on top of the arch and away from the palpebral nerve. The periosteum of the zygomatic arch was apposed to the masseter muscle using 3-0 poliglecaprone-25 cruciate suture pattern. Subcutaneous tissue was apposed using 3-0 poliglecaprone-25 simple continuous suture pattern and the intradermal layer was apposed with 4-0 poliglecaprone-25 simple continuous suture pattern.

Figure 3.

Intraoperative photograph demonstrating egressed glandular tissue removed deep to the masseter muscle.

Intraoperative complications included bradycardia which was managed with atropine (VetOne), 0.54 μg (0.019 μg/ kg BW, IV) and glycopyrrolate (Robinul; American Regent, Shirley, New Jersey, USA), 0.3 mg (0.01 mg/kg BW, IV) and hypotension which was managed with a 150 mL (5.3 mL/kg BW, IV) bolus of Ringer’s lactate solution (Baxter Healthcare). Post-operative medications included acepromazine (PromAce; VetOne), 0.15 mg (0.005 mg/kg BW, IV) and carprofen (Rimadyl; Vetprofen; Carorieve; Norocarp; Novox; quellin; Rovera; Zoetis, Lincoln, Nebraska, USA), 60 mg (2 mg/kg BW, SQ). The dog experienced peri-anesthetic dysphoria and a second dose of acepromazine (PromAce; VetOne), 0.15 mg (0.005 mg/kg BW, IV) was administered. Post-operative care consisted of carprofen (Rimadyl; Vetprofen; Carorieve; Norocarp; Novox; quellin; Rovera; Zoetis), 50 mg (1.75 mg/kg BW, PO, q12h), trazodone (Desyrel; Teva Pharmaceuticals USA, Parsippany, New Jersey, USA), 100 mg (3.5 mg/kg BW, PO, q8h), and acepromazine (PromAce; VetOne), 0.4 mg (0.014 mg/kg BW, SQ, prn) for sedation and anti-anxiety control, cephalexin (Cefalexin Rilexine; Keflex, Ascend Laboratories, LLC, Parsippany, New Jersey, USA), 750 mg (26.3 mg/kg BW, PO, q12h) and cefazolin (Ancef; Kefzol; Apotex), 620 mg (21.8 mg/kg BW, IV, q8h). Supportive postoperative care included ice packing the left eye q6h as tolerated by the dog. The Colorado State University pain scale was used q6h to assess pain during recovery. The dog was discharged 3 d after surgery and returned to the hospital 7 d postoperatively for medical boarding 5 d in duration, at which time the dog continued to recover uneventfully. Aerobic culture was negative for bacterial growth.

Histopathology of the left zygomatic salivary gland (Figure 4) revealed necrotic salivary lobules with retained lobular architecture (coagulation necrosis), large numbers of intact and degenerated neutrophils intermixed with variable numbers of lymphocytes, plasma cells, and macrophages, some viable hyperplastic epithelial cells lining acini and salivary ducts, and mildly dilated salivary gland interlobular ducts containing mucin and some neutrophils. Some necrotic lobules were surrounded by extensive hemorrhage and fibrin deposition with infiltrates of neutrophils. Necrosis involved surrounding adipose tissue and incorporated several blood vessels. These histopathologic findings were consistent with a diagnosis of zygomatic salivary gland infarction with severe fibrinous, hemorrhagic and neutrophilic inflammation. Suspected etiologies for salivary gland infarction included traumatic damage or immune-mediated damage to blood vessels.

Figure 4.

Hematoxylin and eosin (H&E)-stained sections from an excised left zygomatic salivary gland. Note areas of salivary gland coagulative necrosis (black arrows) adjacent to areas of extensive fibrin deposition, hemorrhage, and neutrophilic inflammation (yellow arrows).

A 20-month postoperative re-evaluation revealed a left zygomatic scar, no evidence of facial asymmetry, and no oral pain with maximal mouth opening (Figure 5). There remained a slab fracture of tooth #204. Skull radiographs revealed no long-term implant complications and healed left zygomatic osteotomy sites (Figure 6). No recurrence or de novo sialoceles were noted.

Figure 5.

Front (a) and lateral (b) views of 20 mo postoperative healed incision site.

Figure 6.

Ventrodorsal (a) and orthogonal (b) skull radiographs obtained 20 mo postoperatively, demonstrating healed zygomatic osteotomy site.

Discussion

This case highlights the successful surgical management of a histopathologically confirmed left-sided salivary gland infraction by sialoadenectomy via a zygomatic arch osteotomy approach. The dog had minimal perioperative complications with no evidence of complications at the 20-month re-evaluation. There was evidence of bone healing and static implants at follow-up with no signs of recurrence, based on a physical examination and orthogonal skull radiographs. To the authors’ knowledge, there are no other reports of long-term follow-up past 4 mo, at which time there was clinical recurrence of disease, in a case of a sialocele secondary to infarction in any salivary gland (15), and this is the first case report with long-term post-operative skull radiographs.

Salivary gland infarction is an uncommon cause of sialoceles. One case report discusses salivary gland infarction in 2 adult dogs characterized by ischemic necrosis, capsular fibrosis, and regenerative hyperplasia of surviving ductal epithelium, with necrosis of arterial tunica media and thrombosis of the infarcted parts of the salivary glands on histopathology. The first dog was an 8-year-old male neutered Jack Russell terrier that had excessive amounts of unusually viscous saliva and dysphagia for 7 wk and upon examination had weight loss and a hard swelling in the left submandibular region. The left submandibular salivary gland was excised. The second dog was a 5-year-old female West Highland white terrier that owners noted to have excessive salivation and hardening of the submandibular region several weeks after upper respiratory tract signs were noted. Two months later, the dog presented for examination and the left submandibular and sublingual salivary glands were excised. One week after excision, the right submandibular salivary gland was also excised. A third case of salivary gland necrosis was detected in a 9-year-old Jack Russell terrier after this case report was prepared. The cause of infarction was not determined in either case (9).

Two case reports describe necrotizing sialometaplasia (NSM) in a submandibular salivary gland in a 15-month-old, male beagle dog (7) and a parotid salivary gland in a 7-year-old, neutered male cocker spaniel dog (15). These describe NSM as a benign self-limiting inflammatory lesion of the salivary glands with 5 proposed stages in the etiopathogenesis of NSM: infraction, sequestration, ulceration, reparation, and healing. In our case, the left zygomatic salivary gland had evidence of infarction with severe fibrinous, hemorrhagic, and neutrophilic inflammation. This case could have represented the first stage of NSM.

The exact cause of salivary gland infarctions is unknown, with suspected etiologies including traumatic damage or immunemediated damage to blood vessels. It is important to discuss the location of the fractured tooth (#204) noted on CT and the location where the zygomatic salivary gland ducts exit as a potential cause of trauma. Given that the zygomatic salivary gland ducts exit at the caudolateral aspect of the last maxillary molar (#210) (16), no widening of the pulp chamber noted on CT, and the appearance of the fracture being chronic upon physical examination, it is less likely that the tooth fracture served as a cause of the infarct causing the zygomatic sialocele. The remainder of the physical examination was unremarkable with no evidence of other facial trauma as a potential cause of infarction.

Besides trauma, another differential for infarction is metabolic or immune-mediated disease. The dog described here did not have any clinical or hematological signs of these predisposing causes. In addition, no other infarcts were detected at the time or since surgery.

Computed tomographic angiography was performed in this case for guidance and surgical planning. The major salivary glands are well-defined and organized, and can be readily identified on advanced imaging studies, with recent literature describing the appearance and location of normal major canine salivary glands on CT sialography and magnetic resonance imaging (MRI). In addition, MRI is more sensitive than CT angiography in identifying infarcts in the brain, which is the site studied most often in the canine population (17). Although plain radiographs are of limited usefulness in salivary gland assessment, advanced imaging techniques have led to improvements in diagnosis and management of salivary gland disease (18). Ultrasonography was used to support the diagnosis of submandibular sialoceles (19); however, ultrasonographic characteristics may vary depending on the chronological stage of disease and given the anatomic location of the zygomatic salivary gland deep to the zygomatic arch, ultrasonography was not offered as a diagnostic in this case.

Treatment consisted of a left-sided zygomatic sialoadenectomy via zygomatic arch osteotomy. This was discussed as the treatment of choice, due to limited publications reporting success with non-surgical treatment options and given the success rates of sialoadenectomy reported for other locations of sialoceles (20,21). A study reporting the outcome of a non-surgical treatment option involving intracanalicular injection of N-acetylcysteine as an adjunctive treatment for sialoceles in dogs (12) was published after this case was presented.

In our patient, intra-operative surgical complications encountered were the presence of red-black gelatinous fluid, hypotension, bradycardia, pain, and dysphoria during anesthetic recovery, which were all managed accordingly. There are limited reports on intra-operative complications associated with sialoadenectomy. Transient pain and anxiety were encountered post-operatively. Short-term post-operative sialoadenectomy complications reported include incisional swelling (13) and drainage (8), periorbital swelling, and anorexia (6).

In any case report, there are limitations in treatment choices, applicability to other cases, and the retrospective nature of the report. Other treatment options including sialocentesis, corticosteroid (11), phenobarbital administration (10), or N-acetylcystine infiltration (12) were not chosen, and therefore cannot be discussed. An infarct was not identified using CT angiography, and upon retrospective review, MRI may represent a more sensitive imaging option in similar case presentations. In addition, predisposing causes for infarction were not investigated as thoroughly as possible. Lastly, follow-up would ideally be lifelong.

Zygomatic sialoceles are not commonly reported in dogs, and infarction is a rarely reported cause of sialocele formation in any salivary gland, although this may be the initial stage to necrotizing sialometaplasia. This case highlighted the importance of physical examination findings in conjunction with advanced imaging techniques in diagnosis and surgical correction, with excellent long-term results of a zygomatic salivary gland infraction. CVJ