Little has been reported in the veterinary literature about spontaneous perforations of the upper digestive tract in small animals. Although their potential toxicity has been well described, the role that recently developed non-steroidal anti-inflammatory drugs (NSAIDs) may play in this disease is unknown. Our aim was to describe the clinical, clinicopathological, imaging and surgical findings, as well as the outcome following management of the cases of spontaneous gastro-duodenal perforations in dogs and cats presenting at our institution. We hypothesised that NSAIDs, and more particularly meloxicam, would be identified as the causative factor in the majority of cases and outcome following surgical management would be good.

For this retrospective clinical study all the consecutive cases of spontaneous gastro-duodenal perforations who presented at the Queen Mother Hospital for Animals between 2002 and 2007 were included.

Seventeen cases (15 dogs and 2 cats), all pure-bred, were identified. Mainly large- to giant-breed dogs were affected. They essentially presented with non-specific signs with a duration ranging from 12 hours to 4 weeks. Hematemesis was rarely observed. Clinicopathological findings were varied and inconsistent, although panhypoproteinaemia and neutrophilia were common. Ascites was identified in all patients examined ultrasonographically, whereas signs of gastropathy were uncommonly (4/11 cases) demonstrated. Cytological evaluation of the fluid obtained by abdominocentesis had a high diagnostic yield. The perforations were located to the stomach (13), including its fundus (1) and body (5), the duodenum (5), the pylorus (2), and were multiple in 3 cases. Surgery, consisting of debridement of the ulcer and primary reconstruction was carried on in 12 cases. Ten patients survived. All had isolated gastric perforations. All of the patients with duodenal or pyloric perforations were euthanized or died postoperatively. NSAIDs were identified as the causative factor for perforation in 11 cases (including 7 due to meloxicam), corticosteroids in 2, tumour in 1 and a combination of previous surgery and drugs in 3 cases. The survivors had a favourable long-term outcome. Survival was not related to the recognised causative factor but was associated with the location and the number of perforations.

Gastro-duodenal perforations seemed to be mainly related to the use of NSAIDs and not to be that uncommon with regards to the incidence of cases of septic peritonitis treated at our institution. However, surgical treatment is likely to be successful, especially for gastric lesions.