CANALOGRAPHIC EVALUATION OF THE EXTERNAL EAR CANAL IN DOGS

Canalography was performed to measure the diameter of the horizontal ear canal and to assess the clinical relevance of the measurements in identifying abnormalities of the ear canal. The diameter of the proximal and distal end of the annular cartilage in the horizontal ear canal was measured using canalography in 222 ears. The tympanic membrane could not be visualized with otoscopic examination in 70 ears even after ear cleansing. These canals were classified as being stenotic. The diameter of the proximal annular cartilage in the stenotic canals ranged from 0.8 to 3.1 mm (mean = 2.6 +/- 0.8 mm). The proximal annular cartilage was consistently smaller in diameter than the distal annular cartilage. The ratio between the diameter of the proximal and distal annular cartilage varied between dogs. In stenotic ear canals (70 ears) the ratio was less than 0.65. Total ear canal ablations and histopathological analysis were performed in 70 ear canals classified as stenotic after canalography. Hyperplasia of the epidermal layer was found in 56 stenotic canals. Stenosis of the canal due to otitis externa was found in the other 14 canals. Three canals with severe otitis externa were also diagnosed with canalography as having a ruptured tympanic membrane. The results suggest that canalography can be used to measure the diameter of the horizontal ear canal and to assess whether or not the tympanic membrane is ruptured prior to conservative or surgical therapy.     

PREVALENCE AND CHARACTERIZATION OF SMALL TYMPANIC BONE SPICULES AND DRUMSTICK‐LIKE HYPEROSTOTIC TYMPANIC BONE SPICULES IN THE MIDDLE EAR CAVITY OF DOGS

Rounded, sessile, hyperattenuating structures detected in computed tomography (CT) studies of canine tympanic bullae have been termed “otoliths.” These have been proposed to represent dystrophic mineralizations or heterotopic bone formations in the middle ear that are potentially related to chronic otitis media. Aims of the current study were to describe the prevalence, macroscopic, and histological features of structures consistent with “otoliths” in the canine tympanic cavity. Tympanic bullae from 50 routinely necropsied dogs and 139 retrospectively retrieved CT scans of canine clinical cases were examined. Small tympanic bone spicules with pointed or clubbed tips essentially arising from the free margin of the septum bullae were bilaterally present in the tympanic cavities of all 50 of the necropsied dogs. In 48% of the dogs, “otolith”‐like CT‐detectable bone spicules carrying drumstick‐like hyperostoses that were 1–6 mm in diameter were also present. In the retrospective survey of bulla CT scans of 139 cases, the prevalence of hyperostotic tympanic bone spicules (HTBS) was 20%. Findings from the current study indicated that the presence of small tympanic bone spicules in adult dogs is most likely due to physiological bone growth in the septum bullae and that HTBS represent osseous proliferations of small tympanic bone spicules. However, the factors inducing formation of hyperostotic spicules from small tympanic bone spicules remain unknown. The high prevalence of HTBS displaying a similar appearance in bulla CT scans in dogs suggests that these spicules should be included in a differential diagnosis list for “otoliths.”     

ULTRASONOGRAPHIC EVALUATION OF THE PROSTATE IN HEALTHY INTACT DOGS

Sagittal and transverse ultrasonographic images of the prostate gland were obtained in 100 healthy adult intact male dogs. Prostatic length, width, and height on transverse and sagittal images as well as the presence of prostatic cysts were determined. Linear regression and correlation analysis were performed between prostatic parameters (length, width, height on sagittal and transverse images, and estimated volume) and parameters related to body size (body weight, body height, left kidney length and aortic diameter) and age of the dogs. Significant positive correlations were found between all prostatic parameters and parameters related to body size and age. Maximum predicted values for prostatic parameters for a given body weight and age were determined based on the upper limit of the 95% confidence interval of the mean predicted values. Such values should represent a useful tool for ultrasonographic evaluation of the prostate in the dog. Prostatic cysts were found in 14% of the dogs.     

ULTRASONOGRAPHIC EVALUATION OF THE PARATHYROID GLANDS IN HYPERCALCEMIC DOGS

Two hyperplastic parathyroid glands and three solitary parathyroid adenomas were identified using high-resolution ultrasonography in five adult dogs with persistent hypercalcemia. Ultrasonographic features of parathyroid adenomas included visualization of a round or oval, 5 mm or larger, hypoechoic mass in the cranial pole of one thyroid lobe. Each mass had well-defined margins between the thyroid gland and parathyroid adenoma, reduced echogenicity of the adenoma compared to surrounding thyroid parenchyma, and distal enhancement. Ultrasonographic features of the hyperplastic parathyroid glands included hypoechogenicity, compared to surrounding thyroid parenchyma, and a size of approximately 2 mm. Hyperplastic parathyroids were well marginated in one dog and poorly marginated in another dog.     

ULTRASONOGRAPHIC CHARACTERISTICS OF THE ABDOMINAL ESOPHAGUS AND CARDIA IN DOGS

Differential diagnoses for regurgitation and vomiting in dogs include diseases of the gastroesophageal junction. The purpose of this cross‐sectional study was to describe ultrasonographic characteristics of the abdominal esophagus and gastric cardia in normal dogs and dogs with clinical disease involving this region. A total of 126 dogs with no clinical signs of gastrointestinal disease and six dogs with clinical diseases involving the gastroesophageal junction were included. For seven euthanized dogs, ultrasonographic features were also compared with gross pathology and histopathology. Cardial and abdominal esophageal wall thicknesses were measured ultrasonographically for all normal dogs and effects of weight, sex, age, and stomach filling were tested. Five layers could be identified in normal esophageal and cardial walls. The inner esophageal layer was echogenic, corresponding to the cornified mucosa and glandular portion of the submucosa. The cardia was characterized by a thick muscularis, and a transitional zone between echogenic esophageal and hypoechoic gastric mucosal layers. Mean (±SD) cardial wall thicknesses for normal dogs were 7.6 mm (±1.6), 9.7 mm (±1.8), 10.8 mm (±1.6), 13.3 mm (±2.5) for dogs in the <10 kg, 10–19.9 kg, 20–29.9 kg and ≥30 kg weight groups, respectively. Mean (±SD) esophageal wall thicknesses were: 4.1 mm (±0.6), 5.1 mm (±1.3), 5.6 mm (±1), and 6.4 mm (±1.1) for the same weight groups, respectively. Measurements of wall thickness were significantly correlated with dog weight group. Ultrasonography assisted diagnosis in all six clinically affected dogs. Findings supported the use of transabdominal ultrasonography as a diagnostic test for dogs with suspected gastroesophageal disease.     

VIRTUAL CT OTOSCOPY OF THE MIDDLE EAR AND OSSICLES IN DOGS

Virtual otoscopy enables noninvasive 3D endoluminal imaging of the middle ear through postprocessing of computed tompgraphy (CT) data. A standardized imaging approach for the middle ear was established in six normal dogs in an attempt to optimize the clinical application and student education. High‐resolution CT data were obtained. Virtual otoscopic images of the middle ear cavity and ossicles were generated using commercially available software. The views of the four different directions (the ear canal, tympanic bulla, eustachian tube, and ossicular chain) were made for virtual otoscopy. The promontory, cochlea window, tympanic bulla, septum bulla, and auditory tube were distinguished easily and clearly. One of the ossicles, the malleus, was visualized accurately. However, small structures such as the incus and stapes always could not be seen. The main advantage of virtual otoscopy is not only to provide diagnostic information but also to enhance the quality and efficiency of student education, because it contributes to an understanding of the anatomy of the middle ear. We describe the normal topographical 3Dimages of the middle ear of the dog using virtual otoscopy.     

ULTRASONOGRAPHIC FEATURES OF PYONEPHROSIS IN DOGS

Pyonephrosis refers an infected hydronephrotic kidney which arise from pyelonephritis followed by exudate accumulation in a dilated renal pelvis or hydronephrosis followed by ascending infection. Pyonephrosis may cause serious systemic complications, making prompt and reliable diagnosis critical. Clinical and ultrasonographic findings are used for the diagnosis of pyonephrosis in humans, but these findings have not been investigated in dogs. We reviewed ultrasonographic features in pyonephrosis in 18 dogs. Ten dogs with hydronephrosis were also evaluated to compare with the pyonephrosis patients. In most dogs with pyonephrosis, hyperechoic contents completely filled the dilated renal pelvis (n=8) or a fluid‐debris level was observed (n=8). Hyperechoic contents were dispersed in renal pelvis in only two of the 18 dogs. Hyperechoic, edematous mesentery, and peritoneal and retroperitoneal effusion, which represented peritoneal and retroperitoneal inflammation, were observed in the perinephric region in 11 dogs. Compared with pyonephrosis, and as expected, hydronephrosis was characterized by anechoic contents within the urine‐filled collecting system and there were no definitive findings to suspect peritonitis. Thus, there is a distinct difference in the sonographic appearance of pyonephrosis vs. hydronephrosis in dogs.     

ULTRASONOGRAPHIC FEATURES OF INTESTINAL ENTRAPMENT IN DOGS

The clinical and ultrasonographic features of postoperative intestinal entrapment were assessed in five dogs. Four had vomiting and lethargy, and one had peracute collapse and hematochezia. Ultrasonographic findings in four of five dogs were similar, being characterized by focally hyperechoic mesentery and abdominal effusion, surrounding a single loop of amotile and dilated intestine. In some dogs, the affected intestinal loop had a thickened or corrugated wall, or alteration of wall layering. In one dog, the site of entrapment could be directly visualized. In the most severely affected dog, a large volume of echogenic peritoneal effusion was present, as well as fluid dilation of multiple intestinal loops. The ultrasonographic appearance of intestinal entrapment is similar to that of intestinal perforation or infarction by other causes.     

ULTRASONOGRAPHIC APPEARANCE OF OVARIAN TUMORS IN 10 DOGS

Signalment, clinical history, physical examination, clinicopathologic, readiographic and ultrasonographic findings of 10 female dogs with histologically confirmed ovarian neoplasms were reviewed. Ultrasonographic images and reports were reviewed for (1) location, size, outer margins, and echogenicity of the mass(es), (2) presence of free abdominal fluid, (3) evidence of uterine abnormalities, and (4) signs of metastatic disease. The masses were classified according to their ultrasonographic pattern in solid, solid with cystic component, and cystic. The masses were ultrasonographically reported as being of ovarian origin in eight dogs, and this origin was included in the list of differentials in the remaining two dogs. When present, abdominal effusions and uterine abnormalities were diagnosed by means of ultrasound.     

ULTRASONOGRAPHIC DIAGNOSIS OF PORTOSYSTEMIC SHUNTING IN DOGS AND CATS

The value of ultrasonography was evaluated in 85 dogs and 17 cats presented with a clinically suspected portosystemic shunt (PSS). A PSS was confirmed in 50 dogs and nine cats (single congenital extrahepatic in 42, single congenital intrahepatic in 11, and multiple acquired in six). Six dogs and one cat had hepatic microvascular dysplasia, and 29 dogs and seven cats had a normal portal system. Ultrasonography was 92% sensitive, 98% specific, and had positive and negative predictive values of 98% and 89%, respectively, in identifying PSS, with an overall accuracy of 95%. When a PSS was identified with ultrasonography, extrahepatic, intrahepatic, and multiple acquired PSS could be correctly differentiated in 53/54 patients (98%). The combination of a small liver, large kidneys, and uroliths had positive and negative predictive values of 100% and 51% for the presence of a congenital PSS in dogs. The portal vein/aorta (PV/Ao) and portal vein/caudal vena cava (PV/ CVC) ratios were smaller in animals with extrahepatic PSSs compared with animals with microvascular dysplasia, intrahepatic PSSs and those without portal venous anomalies (P<0.001). All dogs and cats with a PV/Ao ratio of < or = 0.65 had an extrahepatic PSS or idiopathic noncirrhotic portal hypertension. Dogs and cats with PV/Ao and PV/CVC ratios of > or = 0.8 and > or = 0.75, respectively, did not have an extrahepatic PSS. Reduced or reversed portal flow was seen in four of four patients with multiple acquired PSSs secondary to portal hypertension. The presence of turbulence in the caudal vena cava of dogs had positive and negative predictive values of 91% and 84%, respectively, for the presence of any PSS terminating into that vein.     

APPROACHES FOR ULTRASONOGRAPHIC EVALUATION OF LONG BONES IN THE DOG

The objective of this study was to identify access portals for ultrasonographic evaluation of canine long bones (humerus, radius, ulna, femur, and tibia). A 7.5 MHz linear transducer was used in five medium-sized dogs (range of 25-35 kg body weight). Good approaches could be identified for all long bones. For the proximal humerus, a cranial portal, and the distal humerus, a lateral portal is suggested. For the radius and ulna, a craniolateral approach seemed to be the best. In the hindlimb, a medial approach for the femur and a craniolateral approach for the tibia were the most effective approaches.     

RETROSPECTIVE ULTRASONOGRAPHIC EVALUATION OF ADRENAL LESIONS IN 26 DOGS

A review was performed of ultrasonographic findings in 26 dogs with confirmed adrenal lesions. Adrenal shape, size, echogenicity, laterality, and the presence of vascular invasion were evaluated. Histopathologic diagnoses were obtained in all dogs. Adrenal lesions were confirmed as pheochromocytomas (9), adenocarcinomas (6), a poorly differentiated blastoma (l), bilateral adrenal metastases of a carcinoma (l), adenomas-one of which was bilateral-(4) and hyperplasia (6). Size and shape were extremely variable and not specific to lesion type. There was a tendency for pheochromocytomas (7), adenocarcinomas (5) and poorly differentiated blastoma (1) to be rounded masses. Adenomas (4), hyperplasia (7) and adrenal metastases (2) presented predominantly as nodules. No specificity in echogenicity was noted. Mineralization and bilaterality were present in both benign and malignant lesions. Vascular extension or the presence of a thrombus were suggestive but not specific signs of malignancy. Based on our prelimiary study, ultrasonography is an effective method for localizing adrenal lesions and is helpful in assessing their extension. However, no definitive differentiation between benign and malignant lesions was possible using ultrasonographic criteria alone.     

COMPARISON BETWEEN COMPUTED TOMOGRAPHIC CHARACTERISTICS OF THE MIDDLE EAR IN NONBRACHYCEPHALIC AND BRACHYCEPHALIC DOGS WITH OBSTRUCTIVE AIRWAY SYNDROME

Prevalence of subclinical middle ear lesions in dogs that undergo computed tomography (CT) and magnetic resonance imaging of the head has been reported up to 41%. A predisposition in brachycephalics has been suggested, however evidence‐based studies are lacking. Aims of this retrospective cross‐sectional study were to compare CT characteristics of the middle ear in groups of nonbrachycephalic and brachycephalic dogs that underwent CT of the head for conditions unrelated to ear disease, and test associations between thickness of the soft palate and presence of subclinical middle ear lesions. One observer recorded CT findings for each dog without knowledge of group status. A total of 65 dogs met inclusion criteria (25 brachycephalic, 40 nonbrachycephalic). Brachycephalic dogs had a significantly thicker bulla wall (P = 2.38 × 10^?26) and smaller luminal volume (P = 5.74 × 10^?20), when compared to nonbrachycephalic dogs. Soft palate thickness was significantly greater in the brachycephalic group (P = 2.76 × 10^?9). Nine of 25 brachycephalic dogs had material in the lumen of the tympanic cavity, compared to zero of 45 of nonbrachycephalics. Within the brachycephalic group, a significant difference in mean soft palate thickness was identified for dogs with material in the middle ear (12.2 mm) vs. air‐filled bullae (9 mm; P = 0.016). Findings from the current study supported previous theories that brachycephalic dogs have a greater prevalence of subclinical middle ear effusion and smaller bulla luminal size than nonbrachycephalic dogs. Authors recommend that the bulla lumen volume formula previously developed for mesaticephalic dogs, (?0.612 + 0.757 [lnBW]) be adjusted to 1/3(?0.612 + 0.757 [lnBW]) for brachycephalic breeds.     

PREVALENCE OF EAR DISEASE IN DOGS UNDERGOING MULTIDETECTOR THIN‐SLICE COMPUTED TOMOGRAPHY OF THE HEAD

Previous reports describing the prevalence of ear diseases in dogs have primarily been based on dogs presenting with clinical signs of disease. The prevalence of subclinical ear disease remains unknown. The purpose of this cross‐sectional retrospective study was to describe the prevalence of lesions consistent with middle and external ear disease in dogs presented for multidetector computed tomography (CT) of the head and/or cranial cervical spine at our hospital during the period of July 2011 and August 2013. For each included dog, data recorded were signalment, CT findings, diagnosis, and treatment. A total of 199 dogs met inclusion criteria. Nineteen dogs (9.5%) were referred for evaluation of suspected ear disease and 27 dogs (13.5%) had histories or physical examination findings consistent with otitis externa. A total of 163 dogs (81.9%) had CT lesions consistent with external ear disease (i.e. ear canal mineralization, external canal thickening, and/or narrowing of the external canal). Thirty‐nine dogs (19.5%) had CT lesions consistent with middle ear disease (i.e. soft tissue attenuating/fluid material in the tympanic bullae, bulla wall thickening or lysis, and/or periosteal proliferation of the temporal bone). Findings from this study indicated that the prevalence of external and middle ear disease in dogs could be higher than that previously reported.     

ULTRASONOGRAPHY VERSUS RADIOGRAPHY FOR DETECTION OF FLUID IN THE CANINE TYMPANIC BULLA

The tympanic bullae of 20 previously healthy cadaver animals were randomly filled with sterile saline or air. The heads were positioned using tape ties and a rostrocaudal open mouth radiograph taken. The resulting radiographs were assessed by a blinded ECVDI diplomate, with each tympanic bulla being recorded as being either air or fluid filled. Ultrasound examination of the bullae was performed, by a blinded assessor, using a B-mode 5 mHz sector probe (Sonotron-VingMed Sound A/S, PO Box 141, N3191, Horten, Norway) operating at 3 mHz. Each bulla was recorded as being either air or fluid filled. Results of radiography and ultrasonography were each compared to the actual status of the bulla in a chi square analysis. Sensitivity of radiography was 80%, with specificity of 65%. Ultrasonographic sensitivity was 100% and specificity 100%. Ultrasonography may in future provide a cheap, noninvasive, rapid and widely available method for diagnosing otitis media in clinical cases.     

ULTRASONOGRAPHIC FINDINGS IN CHOLECYSTOKININ-INDUCED PANCREATITIS IN DOGS

Pancreatitis can be induced in dogs by intravenous infusion of supraphysiologic doses of synthetic cholecystokinin octapeptide (CCK-8). Eight anesthetized female beagles were given either CCK-8 or saline in a randomized, blinded fashion and had abdominal ultrasonography to study the development of pancreatic lesions. Pancreatic lesions were apparent 2 hours after the start of CCK-8 infusion and included swelling, interlobular and subcapsular fluid accumulation and patchy hypoechogenicity of the gland parenchyma. CCK-8 had a variable effect on gallbladder emptying. No signs of biliary obstruction or duodenal lesions were identified. Dogs given CCK-8 had severe edematous pancreatitis at necropsy six hours after the start of the infusion. Compared to the oleic acid model of pancreatitis, CCK-8 infusion induces a rapid onset of pancreatitis and an ultrasonographic appearance that reflects diffuse edema of the gland. The short time-scale of the experiment may account for the relative absence of secondary lesions affecting the biliary tract or duodenum compared to the naturally-occurring disease.     

ASSESSMENT OF THREE RADIOGRAPHIC PROJECTIONS FOR DETECTION OF FLUID IN THE RABBIT TYMPANIC BULLA

Otitis media in the rabbit commonly results in fluid accumulation in the tympanic bulla. Radiographic detection of fluid in the tympanic bulla is only moderately sensitive in the dog and cat. The purpose of this study was to investigate the accuracy of radiographic detection of fluid in the rabbit tympanic bulla for three different radiographic projections. Forty rabbit cadavers were used. Half of the tympanic bullae were randomly allocated to be filled with soft tissue material. Following this procedure, each specimen was radiographed in three projections: dorsoventral (DV), rostro 40° ventral-caudodorsal (R40°V-CdDO), and left and right latero 40° ventral-laterodorsal (Lat40°V-LatDO). After imaging, each specimen was frozen and sectioned to determine the content of the tympanic bullae. Images were interpreted and scored independently by two board-certified radiologists. There was no significant difference in sensitivity or specificity between the three projections when compared with the gross findings following sectioning, but observer confidence was highest for the DV projection. Accuracy of radiographic fluid detection was similar to that reported in the dog and cat. The DV and R40°V-CdDO allow comparison of both tympanic bullae on a single radiograph, but the DV was subjectively easier to position, while the Lat40°V-LatDO requires two radiographs for comparison.     

COMPARATIVE PERFORMANCE OF RADIOGRAPHY AND COMPUTED TOMOGRAPHY IN THE DIAGNOSIS OF MIDDLE EAR DISEASE IN 31 DOGS

The purpose of this study was to compare computed tomography (CT) and radiography for diagnosing the presence and severity of middle ear disease in dogs with a history of chronic otitis externa. Thirty-one dogs undergoing a total ear canal ablation and bulla osteotomy were studied. Three normal dogs served as controls. All dogs were examined using radiography and CT. Three radiologists independently evaluated imaging studies in random order. A visual analog scale method was used for scoring certainty and severity of middle ear disease. Surgical findings were recorded intra-operatively. Bulla lining samples were submitted for histopathologic evaluation and scored by a single pathologist who also used a visual analog scale system. Findings from both imaging modalities agreed more closely with surgical findings than with histopathologic findings. With either surgical or histopathologic findings as the gold standard, CT was more sensitive than and as specific as radiographs for predicting presence and severity of middle ear disease. Observer performance with CT was more consistent than the performance with radiographs in the detection of changes that occur with middle ear disease. Both radiography and CT were more accurate for predicting the severity of the disease than its presence. Findings indicate that CT is more accurate and reliable than radiography in diagnosing middle ear disease for dogs having concurrent otitis externa, but only when severity of disease is moderate or high. With low severity of disease, diagnostic certainty for both modalities becomes more variable.     

ULTRASONOGRAPHIC DIAGNOSIS OF PORTAL VEIN THROMBOSIS IN FOUR DOGS

Ante mortem diagnosis of portal vein thrombosis was determined ultrasonographically in four dogs. In each dog the thrombus was visible in two-dimensional, grey-scale images of the portal vein obtained through a right intercostal window. Duplex-Doppler measurements and color-Doppler images provided information about the effects of thrombosis on portal blood flow. Reduced portal blood flow compatible with portal hypertension was detected in three dogs. A hypercoagulable state was probably involved in the pathogenesis of portal vein thrombosis in two dogs, one with pancreatitis and gastrointestinal blood loss and another with protein-losing nephropathy and probable immune-mediated anemia. The third dog had chronic ehrlichiosis; thrombosis was probably secondary to vasculitis. The remaining dog had thrombosis secondary to invasion of the portal vein by a recurrent duodenal neoplasm. This dog was euthanized because the tumor was considered inoperable. The dog with pancreatitis developed acute portal hypertension due to obstruction of the portal vein by the thrombus and was euthanized. The dogs with protein-losing nephropathy and ehrlichiosis were treated medically and recovered. Although portal vein thrombosis is uncommon, this complication should be considered in dogs with a variety of abdominal or systemic disorders. Ultrasonography is a practical method for diagnosis of portal vein thrombosis and detection of the underlying cause.     

ULTRASONOGRAPHIC FINDINGS IN DOGS AND CATS WITH GASTROINTESTINAL PERFORATION

A retrospective study was performed to evaluate the sonographic features of gastrointestinal (GI) perforation in dogs and cats. Sonographic findings in 19 animals (14 dogs and 5 cats) included regional bright mesenteric fat (19), peritoneal effusion (16), fluid-filled stomach or intestines (12), GI wall thickening (11), presence of free air (9), loss of GI wall layering (9), regional lymphadenopathy (8), reduced GI motility (7), pancreatic changes (4), corrugated intestines (4), presence of a mass (3), presence of a foreign body (3), and mineralization of the gastric wall (1). In 14 patients, "perforation" was listed as a differential diagnosis by the sonographer. Abdominal radiographs and radiographic reports were available for 14 patients. Radiographic findings were decreased serosal detail (12), free air (8), peritoneal contrast medium (1), and suspected foreign body (1). GI perforation was listed as radiographic diagnosis in eight patients, seven of which had evidence of pneumoperitoneum, and one had leakage of contrast material on an upper GI study. In 9/14 patients with radiography, "GI perforation" was listed as a sonographic diagnosis. In three patients in which free air was diagnosed sonographically, radiographs were either not available (2) or the presence of free air was not detected at presentation (1). Peritoneal fluid analysis was performed in nine patients, five of which were identified as septic inflammation, and the remaining four were classified as neutrophilic inflammation with no etiologic agent identified. The histologic or surgical diagnoses were as follows: three intestinal surgical dehiscence; one percutaneous endoscopic gastrostomy tube site leakage; one duodenal adenocarcinoma; one ileocolic lymphoma; one trichobezoar; one ascarid impaction; and one bobby pin foreign body. In the remaining 10 patients, a focal area of gastric/intestinal ulceration or transmural necrosis with perforation was identified without evidence of an underlying cause.