Korean Journal of
Veterinary Service

A gastro-subcutaneous fistula caused by the migration of a foreign body is rarely reported in small animals (Choi and Han, 2017). Cases of fistula formation due to foreign body (FB) migration are challenging yet crucial for diagnosis and evaluation because the extent of organ damage depends on the type of foreign body and its origin within the body (Appleby et al., 2015). The clinical signs in such cases vary depending on the organ affected by the migrating foreign body. It may lead to peritonitis in the abdominal cavity, pyothorax or pneumothorax may occur in the thoracic cavity, and lameness may develop when the spine or pelvis is involved (Choi and Han, 2017). Clinical presentations can range from asymptomatic to life threatening. Common imaging modalities used in these cases include plain radiography, contrast radiography, and ultrasonography, and CT scans have become increasingly valuable in recent years (Appleby et al., 2015). When fistula formation occurs because of foreign body migration, there is a high likelihood of foreign body translocation into the abdominal or thoracic cavity, necessitating precise preoperative imaging and abdominal exploration (Hunt et al., 2004; Rubin et al., 2015). In recent veterinary practice, laparoscopy-assisted techniques have been used effectively for the diagnosis, evaluation, and removal of foreign bodies in complex cases (Bleedorn et al., 2013; Otomo et al., 2019).

In conventional laparoscopic gastrotomy for foreign body removal, the recommended placement of the four ports is as follows: 2 cm caudal to the xiphoid process, 2 cm cranial to the umbilicus, at the midpoint between these two locations, and at the midpoint between the midline and the axillary line. Typically, in laparoscopy-assisted gastrostomy designed for minimally invasive surgery with a minimized number of ports, two ports are utilized: an optical port and a surgical port. In this report, the port for laparoscopic exploration was placed on the midline, 1∼2 cm caudal to the umbilicus.

This case report introduces a novel technique for selecting incision sites to enable simultaneous laparoscopic abdominal exploration and foreign body removal from the stomach. This new approach utilizes CT 3D reconstruction images to successfully perform laparoscopy-assisted minilaparotomy. This report presents a case of gastrosubcutaneous fistula caused by foreign body migration treated with this novel technique.

A 4-year-old neutered male Shar-Pei weighing 20 kg presented to the Veterinary Teaching Hospital, College of Veterinary Medicine, Kangwon National University, with the chief complaint of a subcutaneous chest wall mass on the right thorax. The mass was first noticed by the owner three months prior and had progressively increased in size since then (Fig. 1A). Upon medical history review, no other clinical symptoms were reported. Physical examination revealed that the dog was in general good health, with a body condition score (BCS) of 4/9. The rectal temperature was 39.0℃, heart rate was 100 beats per minute, respiratory rate was 28 breaths per minute, and blood pressure was 110 mm Hg, all within normal limits. No significant findings were noted upon physical examination.

Figure 1. (A) Gross appearance of the right thoracic mass (white dotted circle). (B) Ventro-dorsal abdominal radiograph of a dog with a gastric foreign body (arrow). (C, D) Thoracic images from computed tomography. A subcutaneous mass (white dotted circle) was observed near the 7th and 8th ribs on the right thorax. Reactive changes were identified in the 7th and 8th ribs, which were in contact with the foreign body and soft tissue formation within the thoracic cavity. (E) Ventro-dorsal radiograph of a dog with a gastric foreign body (arrow). (F) Linear foreign body observed in the abdominal image from a ventro-dorsal radiograph (arrow). (G, H) Computed tomography images.

Clinical examination revealed a smooth, rounded mass measuring 12×7 cm in the right thoracic wall (Fig. 1A). Blood tests revealed an elevated white blood cell count (WBC) of 29.09 K/uL (reference range: 12∼15 K/uL), with neutrophils at 23.27 K/uL (reference range: 12∼15 K/uL) and monocytes at 1.44 K/uL (reference range: 12∼15 K/uL), both above the normal range. No significant abnormalities were observed in the serum biochemistry results.

Thoracic radiographs revealed a round, soft tissue-dense mass located in the region of the 7th to 8th right ribs, with periosteal reaction and lytic lesions observed in the adjacent 7th to 8th right ribs (Fig. 1B). Abdominal radiographs showed gastric distension due to gas and a round soft tissue density mass in the pyloric region, along with a foreign body measuring approximately 15 cm x 0.7 cm in the shape of a stick within the stomach (Fig. 1E, 1F).

A 16-slice computed tomography (CT) scan was performed in both the abdominal and thoracic cavities. Thoracic CT revealed a small pleural effusion, along with atrophy and atelectasis of the right middle, caudal, and accessory lung lobes (Fig. 1C, 1D). Additionally, a periosteal reaction was observed in the right 7th and 8th ribs, with a 9 mm lytic lesion in the 7th rib. Abdominal computed tomography (CT) revealed a foreign body measuring 6 mm in diameter and 146 mm in length, extending from the gastric fundus to the costodiaphragmatic recess (Fig. 1G, 1H). The foreign body penetrated the pylorus and was positioned near the diaphragm, causing displacement of the pylorus and the adjacent duodenum toward the diaphragm. Degeneration of adjacent soft tissues was also observed. Fine-needle aspiration (FNA) of the subcutaneous mass was performed for cytological examination and bacterial culture. Cytological analysis revealed that the mass was a granulomatous abscess comprising of hypersegmented neutrophils, vacuolated macrophages, and a small number of bacteria. Bacterial culture identified Escherichia coli, which was sensitive to all antibiotics except erythromycin, azithromycin, and clindamycin. The subcutaneous mass in the right thorax was diagnosed as a chronic granulomatous abscess caused by a gastro-subcutaneous fistula resulting from foreign body migration. Laparoscopic exploration was planned to accurately assess the extent of organ damage and identify adhesion sites.

A surgical plan was established using CT 3D reconstructed images to determine the laparoscopic port and incision sites for foreign body removal (Fig. 2). In the ventrodorsal (VD) view, the distance from the final sternal end to the xiphoid process was measured to precisely mark the location (Fig. 2A, 2B). In the lateral view, a virtual circle was drawn using the length of the foreign body as the radius centered around the fixed portion of the linear foreign object (Fig. 2D). A virtual line was drawn along the costal cartilage, and the intersection of the circle and line was selected as the incision site. The planned incision site (Fig. 2E, I point) was located at the midline 14 cm caudal to the xiphoid process.

Figure 2. The image used for surgical planning was extracted from the 3D reconstruction image obtained via computed tomography. (A) In the ventro-dorsal view, the distance from the caudal edge of the 7th sternum (a) to the end of the xiphoid process (b) was measured to accurately mark the end of the xiphoid process, which serves as the surgical landmark. (B) The lateral view image was selected from the 3D reconstruction where the ventral part of the 7th sternum was exposed. The distance between (a) and (b) was measured (c), and the exact location of (b) was marked on the lateral view. (C) Laparoscopic abdominal exploration using a 5 mm telescope. (D) The end of the xiphoid process (b) and the foreign body (small dotted line) in the stomach are indicated. A circle was drawn using the foreign body as the diameter. A virtual line (yellow line) representing the abdominal skin was drawn following the margin of the costal cartilage line. (E) The distance (d) between the point where the foreign body makes contact with the abdominal skin (arrow) and the end of the xiphoid process (b) was measured. (F) The intra-abdominal view of (C), showing adhesions (arrows) between the right abdominal wall and the omentum (*) covering the fistula of the pylorus. The right lateral liver lobe was intact.

The measurements obtained from the imaging data used for surgical planning demonstrated that the linear measurements from three-dimensional (3D) images generated by multidetector computed tomography (MDCT) showed no significant differences when applied to real-life conditions. To minimize measurement errors, a single observer conducted the measurements ten times, and the mean value with standard deviation was calculated. The measurements were 145±3 mm. During surgery, an incision was made at the midline, 14 cm caudal to the xiphoid process, considering factors such as skin mobility, skin and abdominal wall thickness, and irregularity of the costal cartilage line. Minor discrepancies were resolved during the surgical procedure.

Surgical procedure

The patient was premedicated with midazolam (Midazolam^®, 0.2 mg/kg IV; Bukwang Pharm, Korea) for sedation and anxiolysis. Anesthesia was induced with alfaxalone (Alfaxan^®, 3 µg/kg IV; Jurox Pharmaceuticals, Australia) and maintained with isoflurane inhalation. During the surgery, cephazolin (Cephazolin^®, 22 mg/kg IV; Jongeundang PharmCO, Korea) and tramadol (Tramadol HCl, 4 mg/kg IV; Yuhan Corporation, Korea) were administered. The patient was placed in a dorsal recumbent position. An incision was made at the planned site, and a cannula (TERNAMIAN ENDOTIP Cannula, 5 mm; Karl Storz Veterinary Endoscopy) was inserted, followed by the introduction of a telescope (Hopkins^® Straight Forward Telescope 0º; Karl Storz Veterinary Endoscopy) (Fig. 2C). The abdominal cavity was insufflated with carbon dioxide to maintain a maximum pressure of 10 mmHg.

The abdominal cavity was examined using a single telescope. Upon exploration, the pylorus was found to be displaced, contacting the right diaphragm and right abdominal wall due to the foreign body, with the omentum adhering to the area (Fig. 2F). No signs of inflammation or necrosis were observed in the rest of the abdominal cavity apart from the adhesion site. The hepatic parenchyma and other solid organs were unaffected by foreign bodies.

Laparoscopic exploration revealed that the gastric foreign body could be removed via the existing laparoscopic port without the need for additional ports. The laparoscopic telescope and cannula were removed and the incision site was extended to approximately 2 cm. A new cannula (TERNAMIAN ENDOTIP Cannula, 10 mm; Karl Storz Veterinary Endoscopy) was then inserted, followed by the installation of a LAPspay single port (Hopkins^® Operating Telescope 0º; Karl Storz Veterinary Endoscopy). A section of the stomach surrounding the end of the foreign body was grasped and exteriorized using laparoscopic Babcock grippers (CLICKLINE BABCOCK Grasping Forceps), at which point the cannula and telescope were removed.

The exteriorized portion of the stomach was secured with 2-0 nylon (Blue Nylon; Ailee, Busan, Korea), which was held by an assistant surgeon to expose the stomach. A 5-mm incision was made in the exposed region of the stomach where the foreign body was protruding. The end of the foreign body was grasped using mosquito forceps and gently pulled out of the stomach and abdominal cavity. The stomach was sutured in two layers using 3-0 PDS II (polydioxanone; Ethicon Endo-Surgery, Johnson & Johnson, Cincinnati, Ohio, USA) and the securing sutures were removed. The abdominal cavity was closed using the standard technique.

The foreign body removed was a discolored wooden stick, 145 mm in length and 5 mm in diameter, with one end pointed at and the other end blunt. A negative pressure drain was placed in the right thorax for 3 days postoperatively.

The patient recovered from anesthesia without any complications and received postoperative treatment during the 4-day hospitalization period. The treatment regimen included tramadol (5 mg/kg IV, three times daily) (Tramadol HCL; YUHAN, Korea), cefazolin (22 mg/kg IV, twice daily) (Cephazolin^®; Jongeundang PharmCO, Korea,), and metronidazole (15 mg/kg IV, twice daily) (Metronidazole; JW Pharmaceutical, Korea).

This report presents the first case of a linear foreign body originating from the stomach that formed a gastrosubcutaneous fistula over an extended period. The condition was diagnosed using radiography and CT tomography, and a preoperative plan for the incision site was established. Exploratory laparoscopy was performed at this planned site, followed by laparoscopy-assisted mini-laparotomy through the same incision to minimize the size of the incision.

Similar cases involving fistula formation have been reported, including one case in which an ice cream stick caused a gastrocutaneous fistula, leading to recurrent skin lesions (Brennan et al., 2004). In the present case, the foreign body was a wooden stick approximately 15 cm long with a pointed end. The fistula formed in the direction of the pylorus, rather than in the gastric fundus. This was likely because the pointed end of the foreign body was in contact with the pyloric region, whereas the blunt end was in contact with the fundus. The pointed ends of linear foreign bodies, as in this case, can cause various injuries depending on their movement, potentially affecting the abdominal cavity, thoracic cavity, or skeletal system (Hunt et al., 2004).

The right lung lobes sustained minor damage due to the foreign body; however, there were no penetrating injuries to the lung parenchyma. A granulomatous abscess was formed around the area where the foreign body was fixed. In a similar case, a linear wooden foreign body invaded the thoracic cavity of a dog, causing chronic coughing and pyothorax, and necessitating lobectomy (Choi and Han, 2017). However, in this case, there was no penetrating injury to the lung parenchyma, and the inflammation had not progressed extensively in the thoracic cavity or lung lobes; therefore, surgical intervention in the thoracic cavity was not required.

Typically, laparoscopic surgery for the removal of gastric foreign bodies requires three or four ports (Lew et al., 2005; Matyjasik et al., 2011). In cases where foreign bodies are removed laparoscopically with assistance, two ports or a single-incision laparoscopic surgery (SILS) port may be used (Lew et al., 2005). However, in this case, the optimal incision site for foreign body removal was planned preoperatively using 3D CT imaging. Exploratory laparoscopy was performed through a single port, followed by the removal of the gastric foreign body using laparoscopic-assisted mini-laparotomy through the same incision.

Planning the location of the initial port incision is crucial, as in this case, no additional incisions were necessary following exploratory laparoscopy for foreign body removal. Laparoscopic-assisted surgical procedures offer clear advantages over traditional open laparotomy, including minimal incisions, reduced soft tissue damage, decreased postoperative pain, and shorter hospitalization periods (Matyjasik et al., 2011).

In this case, during the preoperative planning, the port location was selected based on the point at which the stomach surrounding the mobile end of the foreign body was expected to contact the abdominal wall in the VD position. Although this may vary depending on the obesity level of the patient, the abdominal wall generally extends to the costal margin in the VD position. Using this method, placing the exploratory laparoscopic port at the location where the foreign body can be removed with minimal incision may eliminate the need for additional incisions.

If the port is placed more than 14 cm from the xiphoid process, it becomes difficult to exteriorize the foreign body through an incision. Conversely, positioning the port closer than 14 cm increases the risk of excessive manipulation, which could damage the foreign body or the opposite fistula site. Therefore, placing the port slightly closer to the optimal distance is crucial to maintaining a balance between accessibility and minimizing risks. Based on this assumption, the surgical plan was developed to ensure effective removal of the foreign body while avoiding complications associated with standard laparoscopic incision sites.

In the process of planning the incision site, the point where the virtual skin line and the imaginary circular path of the foreign body intersect was designated as the incision site. However, the fact that the actual skin is not a perfect straight line and that the rotation path of the foreign body is, in reality, spherical were factors that could potentially introduce error. Furthermore, the degree of gastric distension due to intragastric gas could also influence the procedure. Nevertheless, these factors were compensated for during the actual surgical process and did not have a significant impact on the planned surgical outcome.

This case report describes a novel planning technique using CT 3D imaging for the minimally invasive treatment of a patient with a gastrosubcutaneous fistula. Despite the absence of highly invasive procedures, the patient showed favorable short- and long-term outcomes with no lesion recurrence. At the 8-month follow-up, the owner reported that the patient remained healthy without any notable abnormalities.

No potential conflict of interest relevant to this article was reported.