This report describes the case management, histopathological and post mortem findings in a 23-year-old gelding with a peri-rectal mass. The mass was debulked surgically and submitted samples revealed it to be a poorly differentiated carcinoma. In the post operative period the horse developed signs of abdominal pain and dysuria and was subjected to euthanasia. Post mortem examination revealed a large infiltrative mass located between the rectum and urethra, consistent with a carcinoma of an accessory genital gland, most likely the seminal vesicle.
Peri-rectal masses are uncommon in the horse. Clinical signs are often consistent with extra-luminal obstruction of the rectum. These include abdominal pain, rectal impaction, dyschezia, dysuria and reduced faecal output (Torkelson 2002; Elce 2006; Freeman 2012). Other clinical signs that may be observed include tenesmus, anorexia and pyrexia (Elce 2006). Differential diagnoses include lymphadenopathy, haematoma, neoplasia or abscess formation, subsequent to foreign body penetration or following intramuscular (i.m.) injection in the gluteal region (Sanders-Shamis 1985; Magee et al. 1997; Torkelson 2002; Elce 2006). Melanomas and squamous cell carcinomas are the most common forms of perineal and peri-rectal neoplasia in the horse (Wilson 1994; Freeman 2012). However, rectal polyps (Freeman 2012), adenocarcinoma (Turner and Fessler 1980) and leiomyosarcoma (Clem et al. 1987) have also been reported. To our knowledge there are no previous reports of primary neoplasia of an accessory genital gland in the horse.
A 23-year-old Camargue gelding presented to the Philip Leverhulme Equine Hospital with a 36 h history of intermittent, mild abdominal pain and reduced faecal output. Prior to referral the horse was treated with enteral fluids and phenylbutazone (Equipalazone)1 3 mg/kg bwt i.v. Rectal examination by the referring veterinary surgeon revealed a rectal impaction, which prevented further palpation. The horse had no history of colic, weight loss, trauma, i.m. injection or rectal palpation prior to the onset of clinical signs. The horse was at pasture 24 h per day and had not undergone any recent changes in management.
On clinical examination the gelding was bright and alert with a body condition score of 3/5 (Anon 2005) and weighed 625 kg. Rectal temperature and heart rate were within normal limits. The respiratory rate was 20 breaths/min and thoracic auscultation was unremarkable. Abdominal auscultation revealed intestinal sounds to be reduced in all quadrants. Haematology and serum biochemistry were unremarkable.
Rectal examination revealed a 10 x 15 cm mass protruding from the right hand side into the rectal lumen. The mass was covered by rectal mucosa and significantly narrowed the rectal lumen. The mass was firm and nonpainful on palpation. Cranial to the mass, a large quantity of dry faecal material was palpable, preventing palpation beyond this point. Abdominocentesis yielded a sample of peritoneal fluid that was normal in appearance and had a white blood cell count of 1.1 × 109 cells/l (ref: <5 × 109cells/l) and total protein of 5 g/l (ref: <20 g/l). Transrectal ultrasonography was performed using a 5 MHz linear transducer. This revealed the margins of the mass to be poorly circumscribed and of mixed echogenicity (Fig1), with a linear hyperechogenic region evident, possibly consistent with gas or fibrosis. Endoscopic examination of the rectum revealed no gross abnormalities of the rectal mucosa overlying the mass.
A peri-rectal abscess was considered the most likely differential diagnosis based on findings following clinical and ultrasonographic examination. The horse was treated with enteral fluids and was started on a course of doxycycline (Pulmodox)1 10 mg/kg bwt per os q. 12 h, metronidazole (Metronidazle)2 15 mg/kg bwt per os q. 8 h and phenylbutazone (Equipalazone)3, 2.2 mg/kg bwt per os q. 12 h). He was fed only grass and short chopped fibre, passed normal faeces and demonstrated no signs of abdominal pain. Repeat rectal examination 24 and 72 h after admission to the hospital revealed no alteration in the size or consistency of the mass or in its ultrasonographic appearance.
Surgical management of the case was discussed but was initially declined by the owner. Since there was no alteration in the character or size of the mass over the first 4 days of therapy, antimicrobial therapy was switched to rifampicin (Rifampicin)4 10 mg/kg bwt per os q. 12 h, trimethoprim sulphonamide (Trimediazine)5, 30 mg/kg bwt per os q. 12 h and potassium iodide (15 g per os q. 24 h). On repeat examination, 7 days later, palpation per rectum revealed the mass to be unchanged in size or consistency. The decision was made to resect or debulk the mass in order to pursue a diagnosis and prevent recurrence of a rectal impaction.
Food was withheld for 12 h preoperatively and the gelding received preoperative flunixin meglumine (Meflosyl)6, 1.1 mg/kg bwt i.v. and acepromazine maleate (ACP Injection)7, 0.03 mg/kg bwt i.m. The gelding was sedated with romifidine hydrochloride (Sedivet)8, 0.06 mg/kg bwt i.v. An epidural injection of morphine sulphate (Morphine Sulphate)9 0.1 mg/kg bwt, methadone (Methodone)9 0.1 mg/kg bwt and mepivicaine hydrochloride (Intra-Epicaine)3 0.2 mg/kg bwt, was administered via the first intercoccygeal space (Schumacher 2006).
The rectum was evacuated and the perineal region was prepared aseptically and draped. A 10 cm elliptical incision was created to the right of the anal sphincter. A combination of blunt and sharp dissection was performed with the surgeon's right hand, in a cranial direction, lateral to the rectum, until a firm capsule was reached, approximately 10 cm cranial to the anus. During this dissection the surgeon's left hand was placed within the rectum to reduce the risk of penetration of the rectal mucosa. The rectal mucosa was not penetrated at any stage of the procedure. A fine needle aspirate (FNA) was taken from the mass, using a 16 gauge needle and syringe. This yielded a white, semi-solid, odourless material. Sharp dissection was then used to incise into the capsule. This allowed more semi-solid, white, odourless material to be digitally evacuated from the mass, which was submitted for histopathology. The cranial extent of the cavity could not be palpated. Therefore, complete excision of the mass was not possible. The mass was debrided using a spoon curette and digital dissection and lavaged with sterile saline. The cavity and defect caudal to the cavity were packed with sterile gauze bandages soaked in metronidazole.
Post operative treatment and progress
Flunixin (Meflosyl6, 1.1 mg/kg bwt i.v) was administered every 12 h following surgery and the gelding continued to receive rifampicin, trimethoprim sulphonamide and potassium iodide.
During the first 24 h post operatively the gelding demonstrated mild signs of abdominal pain, dysuria and oliguria. Catheterisation of the bladder was performed without any difficulty, and a large quantity of urine was obtained. Endoscopic examination of the urethra and bladder did not reveal any abnormalities. Thirty-six hours post operatively the horse became tachycardic with a heart rate of 60 beats/min and developed signs of severe abdominal pain. No improvement was seen following catheterisation of the bladder or following administration of morphine sulphate (Morphine Sulphate)9 0.2 mg/kg bwt i.v. Palpation per rectum revealed multiple distended loops of small intestine. The owner declined further surgical intervention and, due to unrelenting signs of severe abdominal pain, the horse was subjected to euthanasia.
Post mortem examination
On post mortem examination, a large (approximately 20 x 10 x 15 cm), lobulated, poorly demarcated, soft mass was present, compressing the right side of the rectum and occupying the space between the bladder and rectum (Fig2). The mass was adhered to the dorsolateral aspect of the wall of the urinary bladder and urethra and to the ventral aspect of the rectum. It was not possible to separate the mass from these structures. No involvement of the mucosa of the bladder, urethra or rectum was observed. On sectioning, the cut surface of the mass was white to pale tan with multifocal areas of haemorrhage. The left seminal vesicle, prostate and bulbourethral glands were not identified because the neoplasm occupied the periurethral space completely effacing the normal structures.
Tissue samples were fixed in buffered formalin (10%), processed routinely and embedded in paraffin wax. Sections were cut at 4 μm and stained with haematoxylin-eosin. Immunohistochemistry was performed with antibodies to Pan-cytokeratin (Abcam ab17155-1; 1:10), cytokeratin 20 (Dako M7019; 1:50), vimentin (Dako M0725; 1:100), desmin (Dako M0760; 1:2000) and myoglobin (Dako A0324; 1:500).
Histological examination of the mass revealed a cellular neoplasm effacing and infiltrating the seminal vesicle. The mass was heterogeneous in nature, comprising predominantly of polygonal or spindle-shaped epithelial cells, approximately 15–20 μm in size, with variably distinct cytoplasmic borders and scant to abundant eosinophilic cytoplasm. The cell nuclei were large (approximately 10–15 μm), centrally located, vesicular and round or irregular in shape, with a single prominent magenta nucleolus. The cells showed mild to moderate pleomorphism with moderate anisocytosis and marked anisokaryosis. There was an average of one mitotic figure per high power field and apoptotic bodies were frequently observed (Fig3). In some sections the cells were arranged in dense clusters, whilst in others they were arranged in cords, tubules, or acinar-like structures, embedded in a scirrhous reaction. Giant cells, with a single giant nucleus (up to 60 μm) and clusters of perivascular lymphocytes were also observed. In addition, focal areas of extensive coagulative necrosis and acute haemorrhage were present. At the margins of the mass islands of atrophied tubular and glandular, simple to pseudostratified columnar epithelium were observed. The neoplastic cells stained uniformly, mildly, positive for Pan-cytokeratin (Fig4), and negative for cytokeratin 20 (Dako M7019; 1:50), vimentin (Dako M0725; 1:100), desmin (Dako M0760; 1:2000) and myoglobin (Dako A0324; 1:500).
The clinical and gross findings in this case were consistent with a neoplasm within the pelvic cavity. Histological examination indicated that this most likely arose in a seminal vesicle.
Differential diagnoses of a peri-rectal mass include abscess, lymphadenopathy, haematoma or neoplasia (Sanders-Shamis 1985; Magee et al. 1997; Torkelson 2002; Elce 2006). Rectal examination should be performed to determine the location and consistency of the mass and any palpable communication with the rectum. Careful, digital palpation of the rectal mucosa, with a bare hand, may reveal the presence of a defect in the rectal mucosa. Due to the abundant mucosal folds, digital palpation is often more sensitive than visual inspection, either endoscopically, or with a speculum (Freeman 2012). Abdominocentesis should be carried out to determine the presence and magnitude of any inflammatory or infectious process involving the abdominal cavity, secondary to the presence of the mass (Sanders-Shamis 1985; Elce 2006; Freeman 2012). Ultrasonographic examination per rectum, as utilised in the case described, assists with the determination of the extent and nature of the mass and in monitoring response to treatment (Elce 2006). A preoperative fine needle aspirate (FNA) or biopsy may have assisted in this case, enabling material to be submitted for cytology and culture and sensitivity. If the mass had been infectious in origin, positive culture and sensitivity would have assisted with the selection of appropriate antimicrobial therapy (Elce 2006). In the case of a neoplastic mass, preoperative biopsy may have enabled an earlier definitive diagnosis. A percutaneous, ultrasonographically-guided approach, reduces the risk of iatrogenic infection, obtains a more reliable sample for culture and ensures penetration of the catheter stylette into the mass (Magee et al. 1997; Elce 2006).
The initial medical management of this case included facilitating defaecation, and the administration of antimicrobial and nonsteroidal anti-inflammatory medication. Surgical intervention, in addition to debulking the mass, allowed a biopsy and a definitive diagnosis. However, the subsequent post operative regional inflammation resulted in dysuria, secondary to neuritis or local physical obstruction, and uncontrollable signs of abdominal pain. Colostomy, which has been previously described as an option for large peri-rectal neoplasms (Wilson 1994), may have avoided these post operative complications, but was unsuitable in this case, due to the owners' wishes and the infiltrative nature of the mass.
On histological examination, the mass comprised clusters of epithelial cells, arranged in cords, tubules or acinar-like structures with atrophic, glandular tissue at its margins. The neoplasm was positive with handset keratin, and negative with vimentin, desmin and myoglobin, confirming its epithelial origin. Carcinoma arising in the urinary or alimentary tracts was excluded, since gross and microscopic involvement of the mucosa of the urinary tract and colon were not identified. It was considered that the origin was most likely to be the seminal vesicle. The atrophic glandular tissue did not appear to produce mucin (a common finding in the rare bulbourethral gland adenocarcinoma [http://pathologyoutlines.com/prostate.html]. In human medicine, several markers are available to distinguish between neoplasia of urothelial, prostate and seminal vesicle origin (Leroy et al. 2003; Stenzel et al. 2011). To the best of the authors' knowledge, only CK20, which is frequently expressed in prostate and urothelial tumours in man (Alsheikh et al. 2001; Thiel and Effert 2002; Navallas et al. 2011), has been shown to react with equine tissue (http://www.dako.com/cross-reactivity_study_3287.pdf). CK20 was negative in the present case, excluding these organs as a possible origin of the mass.
To the authors' knowledge there are no previous reports of primary neoplasia of an accessory sex gland in the horse. The findings described in this case indicate that neoplasia of an accessory sex gland should be considered as a differential diagnosis for peri-rectal masses in the male horse.