A 17-year-old gelding was presented with mild to moderate signs of colic. Exploratory laparotomy revealed an intussusception of ileum and jejunum into the caecum. The starting point of the intussusception was an intramural mass in the jejunum. Half a metre of jejunum was resected and the horse recovered without complications. Histopathology and immunohistochemistry revealed a myxoma with no myogenic or neurogenic origin.
Intestinal neoplasia in horses is rare (Taylor et al. 2006). The most common equine intestinal neoplasm is the alimentary form of lymphoma; other intestinal neoplasms reported in equids include adenocarcinoma, leiomyoma, leiomyosarcoma, myxosarcoma, ganglioneuroma, nerve sheath tumour and carcinoid (Taylor et al. 2006).
To the authors' knowledge, a jejunal myxoma causing a jejunoileo-caecal intussusception has not been reported previously in horses.
A 17-year-old, 570 kg German Warmblood gelding was presented to the Equine Clinic of the Faculty of Veterinary Medicine, Free University Berlin with colic of 2 hours' duration. On admission, the horse showed no signs of colic. Heart rate (HR) was at 44 beats/min, respiratory rate (RR) 12 breaths/min. Mucus membranes were pink and capillary refill time was 2 s. Body temperature was 37.9°C. Borborygmi were reduced in all abdominal quadrants. On rectal examination the pelvic flexure was not palpable and the presumption was made that a ventral colon impaction with a retroflexion could be the cause for it. Haematology found packed cell volume of 41%, total protein (TP) of 68 g/l and white blood cell count (WBC) of 8.8 × 109 cells/l. Nasogastric intubation revealed no reflux.
The horse was treated with sodium metamizole (Novaminsulfon1, 35 mg/kg bwt i.v.). In addition, an infusion with 9 l of Ringer's solution (10 ml/kg bwt/h) was administered. A 1.5 l dose of mineral oil was given by nasogastric intubation. During the first 10 h after the initial treatment, the gelding showed mild signs of colic and was treated with sodium metamizole 2 more times. During the next few hours more severe signs of colic were observed. Sixteen hours following admission the horse's HR was 64 beats/min and RR was 32 breaths/min. Abdominal paracentesis revealed an orange cloudy peritoneal fluid with an increase of WBC to 5.2 × 109 cells/l and TP to 30 g/l (reference values: WBC <5 × 109 cells/l, TP <25 g/l). No reflux was obtained by nasogastric intubation and examination of the abdomen per rectum had no new findings. Transabdominal ultrasound examination revealed no abnormalities. Due to the continuing signs of colic and the findings in the peritoneal fluid, exploratory laparotomy was recommended.
The horse was premedicated with amoxicillin2 (10 mg/kg bwt b.i.d. i.v.), gentamicin sulphate3 (6.6 mg/kg bwt s.i.d. i.v.), and flunixin meglumine4 (1.1 mg/kg bwt s.i.d. i.v.). Sedation was administered using medetomidine5 (Domitor 0.007 mg/kg bwt i.v.), and anaesthesia was induced using diazepam6 (0.04 mg/kg bwt bwt i.v.) and ketamine7 (2.2 mg/kg bwt i.v.) After intubation, balanced anaesthesia was maintained with isoflurane in oxygen/air and a constant rate infusion of medetomidine (Domitor 0.0035 mg/kg bwt/h).
Abdominal exploration immediately showed a moderately dilated small intestine with transmural haemorrhage. Lidocaine8 was first given as a bolus (1.5 mg/kg bwt i.v.) and then as an infusion (0.05 mg/kg bwt/min i.v.). Gas was removed from the small intestine by mechanical aspiration with a surgical vacuum. Further exploration of the abdominal cavity revealed an invagination of the jejunum and ileum into the caecum. The intussusception was released manually by pulling the ileum and jejunum slowly out of the caecum. Once this was done a second partial invagination of the jejunum became visible (Fig. 1). Palpation revealed the presence of a firm-elastic mass at the site of the partial jejunal invagination. The intestinal wall of the invaginated jejunum and ileum was slightly reddish discoloured and without visible signs of a wall oedema and further pathologies. Moreover, a tear in the ileal and jejunal mesentery along a mesenteric vein became apparent, measuring about 40 cm. The mesentery tear was closed using No. 2-0 polyglactin 910 (Vicryl)9 and continuous suture pattern. In order to remove the mass in the jejunum, a resection of approximately 50 cm of the distal jejunum was performed using a 2 layered hand sutured jejuno-jejunal end-to-end anastomosis with No. 2-0 polyglacctin 910 (Vicryl) and continuous suture pattern.
Post operative treatment and outcome
The horse recovered well from anaesthesia. Post operative medication included amoxicillin (10 mg/kg bwt b.i.d. i.v.), gentamicin sulphate (6.6 mg/kg bwt s.i.d. i.v.) and flunixin meglumine (1.1 mg/kg bwt s.i.d. i.v.) for 5 days. Lidocaine (0.05 mg/kg bwt/min i.v.) was given as an infusion for 2 days. Ringer's solution was administered for 5 days in a constant rate infusion (2 ml/kg bwt/h). Two days post surgery the horse received small amounts of hay every other hour; starting with very small portions that were slowly increased. Recovery was uneventful and the horse was discharged from the clinic 10 days post surgery. Ten months post surgery the owner was contacted by telephone and stated that the gelding's performance was at the same level as it was prior to surgery.
The mass in the removed jejunum measured 5.9 × 6.0 × 3.5 cm and showed a haemorrhagic infiltration. Its surface was brownish and uneven (Fig. 2).
The jejunal mass was fixed by immersion in 4% neutral-buffered formalin for 24 h and paraffin-embedded. Adjacent sections of 4 µm thickness were stained with haematoxylin and eosin (H&E) or Alcian blue stain. Histologically, the neoplastic mass (Fig. 3, asterisks) was detected in the submucosa, compressing and distending the muscle layer (Fig. 3, M) but not penetrating the serosa. It consisted of monomorphic spindle-shaped cells arranged in whirls and bundles in a loose fibrovascular myxomatous stroma. Many of the neoplastic cells had bipolar cytoplasmic extensions. The nuclei of the tumour cells were mainly round and centrally located (Fig. 4). Mitotic activity was low (<1 mitosis per high power field). The neoplasm had no capsule and neoplastic cells partially invaded adjacent tissues. However, no vascular invasion was observed. In addition, multifocal areas of necrosis, haemorrhage and mucosal ulceration as well as moderate suppurative inflammation characterised by the presence of numerous viable or degenerate neutrophils were present.
There was strong positive staining for Alcian Blue, indicating a mucin-rich matrix between the tumour cells (Fig. 4, inset).
Sections of 4 µm thickness were mounted on adhesive glass slides and subjected to immunohistochemistry. Vimentin labelling was visualised using the streptavidin-biotin complex method with a monoclonal anti-vimentin antibody diluted 1:50 in PBS10 with 15 min microwave heating in citrate buffer pH 6.0 at 560 W as the retrieval method. A biotinylated goat anti-mouse antibody diluted 1:200 was used as secondary antibody. Colour development was performed using diaminobenzidine tetrahydrochloride with haemalaun as counterstain. There was strong homogenous to slightly granular cytoplasmic expression of vimentin by virtually all tumour cells (not shown), verifying the mesenchymal differentiation of the tumour cells. Immunohistochemical sections for smooth muscle actin and S-100 were negative (not shown).
In the reported case, a jejunoileo-caecal intussusception was released and a myxoma removed by midline laparotomy. Intussusception is an invagination of one segment of intestine and its mesentery into the lumen of an adjacent segment (Lewis and Ellison 1987). The invaginated segment is the intussusceptum while the enveloping segment is the intussuscipiens (Edwards 1986). Predisposing factors that may lead to segmental motility differences are enteritis, mesenteric arteritis, severe ascarid impaction, tapeworm infection (Anoplocephala perfoliata) and sudden dietary changes (Ekert et al. 2003; Freeman 2006).
A toehold for jejunojejunal intussusception can be a transverse enterotomy, a functional side-to-side anastomosis or an end-to-end anastomosis (Freeman 2006). Furthermore, changes in the mucosal wall such as intramural leiomyoma, pedunculated papilloma and cryptococcal granuloma may form the leading edge for an intussusception (Collier and Trent 1983; Boulton and Williamson 1984).
In the reported case, an intramural myxoma changed the mucosal wall. Myxomas are benign tumours composed of stellate cells in a mucinous matrix with some loose collagen fibres (Head et al. 1994). According to Kirchhof myxomas belong to the group of gastrointestinal stromal tumours defined as nonepithelial gastrointestinal neoplasms of neural, smooth muscle, myofibroblast or undifferentiated mesenchymal origin (Banerjee et al. 1991; Kirchhof et al. 1996).
Based on the histopathological findings, the mass was a gastrointestinal stromal tumour (GIST). Del Piero (et al. 2001) describes GISTs as a loosely arranged network of spindle cells separated by myxoid matrix. Immunohistochemistry provided more detailed information. Sections stained for smooth muscle actin and S-100 were negative, indicating no myogenic or neurogenic components in the tumour.
Intussusceptions can cause acute or chronic colic symptoms depending on the length of the intussusception (Freeman 2006) and the secondary complications in the mesentery and the intestinal wall, e.g. inflammation, oedema or haemorrhagic infarction. In the reported case, the acute deteriorating clinical signs of colic can be explained with the long segments that were invaginated causing a tension on the mesentery and intestinal obstruction (Freeman 2006).
That the dilated small intestine was not identified on either per rectum palpation or transabdominal ultrasound may be explained by the relatively large size of the horse. On palpation per rectum, normally only 25–30% of the abdominal cavity can be palpated (Singer and Smith 2002). However, transabdominal ultrasound is considered very liable to detect small intestinal changes (Singer and Smith 2002). In the reported case, approximately 3 h lay between the last examination and the beginning of surgery because the owner needed some time to make a decision. During that time the small intestine could have dilated even more while, at the time of ultrasound, the dilated small intestine was not ventral enough to be detected by ultrasound. However, the caecum was not examined ultrasonographically and it is possible that the intussusceptum could have been visible earlier.
In conclusion, it seems very likely that in the reported case the jejunal myxoma acted as a starting point for the intussusception (Fig. 1). The ileum was the intussiscipiens to the jejunum that subsequently passed through into the caecum, and the mesentery probably tore as a result of the intussuseption.
Authors' declaration of interests
No conflicts of interest have been declared.
1 Vetoquinol GmbH, Ravensburg, Germany.
2 Selectavet Dr. Otto Fischer GmbH, Weyarn-Holzolling, Germany.
3 Bela-Pharm GmbH & Co. KG, Vechta, Germany.
4 Intervet Deutschland GmbH, Unterschleißheim, Germany.
5 Orion Corporation, Espoo, Finland.
6 Ratiopharm GmbH, Ulm, Germany.
7 CP-Pharma Handelsges. mbH, Burgdorf, Germany.
8 Bela-Pharm GmbH & Co. KG, Vechta, Germany.
9 Johnson & Johnson Intl., St. Stevens-Woluwe, Belgium.
10 DAKO North America, Inc., Carpinteria, California, USA.