|Home | About | Journals | Submit | Contact Us | Français|
A Labrador retriever dog was presented for intestinal obstruction resulting in devitalization of portions of the duodenum. A severe perforation, accounting for 70% duodenal circumference, was present at the level of the duodenal papilla. A vascularized jejunal graft was used to close the perforation, representing novel utilization of this grafting technique.
Greffe jéjunale d’un pédicule vascularisé pour la fermeture d’un grand défaut duodénal chez un canidé. Un Labrador retriever a été présenté pour un blocage intestinal qui produisait la dévitalisation de portions du duodénum. Une perforation grave, représentant 70 % de la circonférence duodénale, était présente au niveau de la papille duodénale. Une greffe jéjunale vascularisée a été utilisée pour fermer la perforation, ce qui représente une nouvelle utilisation de cette technique de greffe.
(Traduit par Isabelle Vallières)
Jejunal grafting was used to repair a large defect in the duodenum in which direct closure could not be achieved. This technique provides a novel alternative to more invasive options including gastroenterotomy or gastrojejunotomy with cholecystoenterotomy. This grafting technique is useful in cases of compromised duodenal tissue involving regions in proximity to vital structures including the duodenal papilla and pancreas and may also be applied to cases in which standard closure or resection and anastomosis failed.
A 4-year-old, 40 kg castrated male Labrador retriever dog was presented with a 1-day history of vomiting. Vomitus periodically contained small pieces of cloth material. The dog had a history of intestinal resection and anastomosis, gastrotomy, and gastropexy approximately 1 y earlier. Further history included steroid administration (trimeprazine with prednisolone) for 1 y due to allergic skin disease, along with a protein-restricted fish and potato diet for food sensitivity.
On physical examination, the dog was quiet, alert, and responsive with appropriate body condition and an estimated 5% to 7% dehydration. Mild ptyalism was noted along with mild discomfort on caudal abdominal palpation. The remainder of the physical examination was within normal limits.
A complete blood (cell) count (CBC) was consistent with hemoconcentration (hematocrit 61.9%), and serum chemistry profile revealed hyperglycemia [8.3 mmol/L; reference interval (RI): 3.9 to 7.9 mmol/L], hyperalbuminemia (44 g/L; RI: 22 to 39 g/L), hypokalemia (3.1 mmol/L; RI: 3.5 to 5.8 mmol/L), and elevated alanine aminotransferase (ALT; 122 U/L; RI: 10 to 100 U/L) and alkaline phosphatase (ALP; 1206 U/L; RI: 23 to 212 U/L). Ventrodorsal and right lateral abdominal radiographs were consistent with soft tissue opacity within the gastric lumen and potential plication of small intestine in the caudal abdomen. Abdominal ultrasound revealed fluid distension of the stomach and proximal duodenum. No foreign material was identified, and an intestinal obstruction could not be confirmed.
A nasogastric tube was placed with gastric suctioning every 4 h, along with intravenous fluid therapy with potassium supplementation [20 mEq KCl/L, 120 mL/kg body weight (BW)/d], metoclopramide (Hospira, Lake Forrest, Illinois, USA), 2 mg/kg BW/d, constant rate infusion (CRI), pantoprazole (Wyeth, Philadelphia, Pennsylvania, USA), 1 mg/kg BW, IV, q24h, and maropitant citrate (Zoetis, Kalamazoo, Michigan, USA), 1 mg/kg BW, SQ, q24h. There was a decrease in the volume of suctioned gastric contents obtained and ptyalism improved through the overnight period. The nasogastric tube was removed the following afternoon, and a repeat abdominal ultrasound revealed resolution of gastrointestinal fluid distension. The dog was discharged after 24 h hospitalization with metoclopramide (0.5 mg/kg BW, PO, q8h) after demonstrating willingness to eat in the hospital. Recheck ultrasound examination prior to discharge revealed a stomach mildly distended with gas and a normal appearance to the intestinal tract. There was no evidence of mechanical obstruction or foreign material.
The dog was re-presented 7 d after discharge for protracted anorexia and vomiting. The owners had noted hyporexia since discharge, and all prehension reportedly induced bilious vomitus. Two bowel movements had been observed since discharge: the first of normal consistency and the second soft but formed. On physical examination, the dog was bright, alert, and responsive. Temperature, heart rate, respiratory rate, and pulse quality were within normal limits. The abdomen was soft and not painful on palpation, with no masses or organomegaly noted. A limited blood panel was performed, revealing electrolytes, ionized calcium, blood glucose, blood urea nitrogen, and creatinine within normal limits. The dog was mildly hemoconcentrated (hematocrit: 51%). An IV catheter was placed and Multiple Electrolytes Injection, Type 2, USP (Plasma-Lyte A; Abbott Animal Health, Abbott Park, Illinois, USA), 12.5 mL/kg BW, IV, was administered as a bolus, followed by a continuous infusion of 120 mL/kg BW per day, IV. Maropitant citrate (Zoetis), 1 mg/kg BW, IV, and famotidine (West-Ward, Eatontown, New Jersey, USA), 1 mg/kg BW, IV, were administered once; nothing was given per os overnight.
The following morning a CBC revealed a mild normochromic normocytic anemia (RBC 4.5 × 106/μL; RI: 4.8 to 9.3 × 106/μL), hemoglobin (HGB; 111 g/L; RI: 12.1 to 20.3 g/L), hematocrit (HCT; 32%; RI: 36% to 60%). Serum chemistry profile revealed hypoproteinemia (39 g/L; RI: 50 to 74 g/L) characterized by hypoalbuminemia (19 g/L; RI: 27 to 44 g/L), elevated ALP (213 U/L; RI: 5 to 131 U/L], hyperglycemia (7.9 mmol/L; RI: 3.9 to 7.7 mmol/L), hypocalcemia (2.1 mmol/L; RI: 2.2 to 2.9 mmol/L), hypomagnesemia (0.7; RI: 0.75 to 1.25 mmol/L), and elevated creatine kinase (CK; 944 U/L; RI: 59 to 895 U/L).
Repeat abdominal ultrasound examination revealed a hyperechoic distal shadowing object measuring 2.5 cm in diameter within the proximal duodenum. Thickened duodenum with hyperechoic surrounding mesentery was present orad to this region. The remainder of the small bowel and colon were within normal limits, with no peritoneal effusion noted and the remaining organs were unremarkable. Due to the high suspicion of a duodenal foreign body causing complete intestinal obstruction, along with the chronicity of clinical signs, it was elected to pursue an exploratory laparotomy.
A ventral midline abdominal approach was made from xiphoid to pubis. Findings included a moderate amount of peritoneal effusion and foreign material palpable in the stomach, duodenum, and proximal jejunum. The duodenum was plicated, and areas were discolored, suggestive of vascular congestion or impaired blood flow, while some areas were grossly necrotic. A gastrotomy was performed through the ventral stomach wall midway between the lesser and greater curvatures, in order to remove foreign material from the stomach and relieve the anchor site for the remaining linear foreign material. This was closed in 2 layers, with mucosa and submucosa apposed with a simple continuous pattern of 3-0 PDS (Ethicon, Guaynabk, Puerto Rico) and muscularis and serosa closed in an inverting Cushing’s pattern of 3-0 PDS.
Two large perforations were noted in the proximal duodenum. The first was 3 to 5 cm distal to the pylorus, and a second perforation was noted at the caudal duodenal flexure. Once necrotic tissue was debrided from the orad perforation, a 5 cm × 3 cm defect was present involving approximately 70% of the duodenal circumference, with the region of viable tissue involving the mesenteric border (Figure 1). The major duodenal papilla was clearly visible in the area of the defect. The aborad perforation in the caudal duodenal flexure was smaller and measured 2 cm × 3 cm, once tissue was debrided.
Initially, primary closure of the orad perforation was attempted using 4-0 PDS in a simple interrupted pattern. This resulted in a narrowed lumen and there was tension on the closure that resulted in sutures tearing through the tissues. A leak-free closure could not be obtained. In order to maintain luminal diameter and decrease tension, a jejunal grafting technique was utilized. A 5-cm segment of healthy jejunum, including jejunal artery and vein, was isolated and mobilized from the middle of the jejunum. The jejunal segment was incised at the anti-mesenteric border to create a rectangular shaped graft which was trimmed as needed (Figure 2), transposed, and sutured in place with 4-0 PDS in a simple interrupted pattern to cover the orad duodenal perforation (Figure 3). Orad-to-aborad orientation was maintained. The graft harvest site was anastomosed using 4-0 PDS in 2 simple continuous appositional patterns. A second routine resection and anastomosis was performed at the aborad duodenal perforation. The abdomen was lavaged with 5 L of sterile saline and a Jackson Pratt drain was placed in the peritoneal cavity prior to closure. The linea alba was closed in a continuous pattern using 0 PDS and the skin was apposed in an intradermal pattern using 2-0 Monocryl (Ethicon) and skin staples.
The dog recovered in the ICU on lactated ringer’s solution (Hospira), 112 mL/kg BW per day, hydromorphone (West-Ward), 0.1 mg/kg BW, IV, q6h, ampicillin/sulbactam (Sagent Pharmaceuticals, Schaumberg, Illinois, USA), 30 mg/kg BW, IV, q8h, enrofloxacin (Bayer, Shawnee Mission, Kansas, USA), 10 mg/kg BW, IV, q24h, pantoprazole (Wyeth), 1 mg/kg BW, IV, q24h, and famotidine (West-Ward), 1 mg/kg BW, IV, q24h. Blood pressure was recorded every 6 h, and packed cell volume and total protein were recorded every 12 h after surgery. Both parameters remained within normal limits.
Cytology of abdominal effusion obtained from the Jackson-Pratt drain was monitored daily to assess for evidence of septic peritonitis. On the third day after surgery, there was a significant increase in degenerate neutrophils in the abdominal effusion with no intracellular bacteria noted. Peripheral venous blood glucose was 10.5 mmol/L, with a peritoneal fluid glucose of 4.8 mmol/L. Due to the nature of the repair and the changes in abdominal effusion a second exploratory laparotomy was performed.
The incision was opened and the previous surgery site was examined. There was < 100 mL of peritoneal effusion present. A 2 to 3 cm, suspect pancreatic abscess was present near the proximal duodenal repair. The anastomosis sites were intact. The jejunal graft was pink and healthy with no evidence of necrosis. The duodenal repair was isolated with Doyen forceps and saline was injected into the lumen to assess for leakage. Approximately 1 to 2 mm of the graft site repair leaked saline when pressure tested. The defect was reinforced with a single interrupted suture using 4-0 PDS, and the suspect pancreatic abscess was debrided. The peritoneal cavity was lavaged using 6 L of sterile saline and omentum was manually placed in the region of the pancreatic abscess. The Jackson-Pratt drain was replaced and an abdominal culture, revealing abundant Escherichia coli and a few Enterococcus species, was obtained. Closure was performed in an identical fashion to the previous procedure. The patient recovered in the ICU, receiving medications as previously described.
A CBC and serum chemistry were rechecked 1 day after revision. The CBC revealed a continued normocytic normochromic anemia (RBC 4.3 × 106/μL; HGB 102 g/L; HCT 31%) and leukocytosis (WBC 23 600/μL; RI: 4000 to 15 500/μL). Serum chemistry profile revealed progressive hypoproteinemia (27 g/L) characterized by hypoalbuminemia (13 g/L) and hypoglobulinemia (14 g/L; RI: 16 to .6 g/L), decreased ALT (9 U/L; RI: 12 to 118 U/L), hypocalcemia (1.9 mmol/L), hypomagnesemia (0.7 mmol/L, and elevated amylase (1171 U/L; RI: 290 to 1125 U/L).
The dog was hospitalized with a Jackson-Pratt drain for the following 4 d. The drain was removed based on decreased production and cytologic evaluation of the abdominal fluid. The dog was discharged 5 d following the second surgery. Medications prescribed included amoxicillin/clavulanate potassium (14 mg/kg BW, PO, q12h) and enrofloxacin (6.8 mg/kg BW, PO, q24h).
Upon recheck examination 8 d later, the dog was bright, alert, and responsive with vital signs within normal limits. The incision was intact and appropriate, and skin staples were removed. The owners reported improved appetite with no vomiting, normal water intake, and normal feces. The dog was permitted to return to normal activity and recheck as needed. At the last telephone follow-up, 12 wk following revision surgery, the dog was reported to be doing well with no clinical signs.
This case represents a successful method of duodenal perforation repair, using a pedicled jejunal graft. This was necessitated by the proximity of the duodenal papilla and inability to perform a standard duodenal closure. To the authors’ knowledge, there are no previous reports of this procedure in the veterinary literature. Previously described surgical proceduress for severe proximal duodenal defects include gastroenterotomy, cholecystoenterotomy, serosal patching, and pancreaticoduodenectomy. Because of the potential of long-term complications and potential higher morbidity associated with these procedures, the reported procedure was performed as an alternative.
The jejunum was selected as a suitable candidate for duodenal repair due to a number of intrinsic factors. The jejunum is abundant, allowing a standard resection and anastomosis of the tissue with minimal to no functional effect on the dog. The intestine allows length modification as needed for varying defects (1). The tissue has a low rate of disease and retains resorptive and peristaltic function after resection and grafting (1–3). Additionally, the mesenteric vasculature can be included and provides adequate length to be used in a pedicle fashion (1–3). In the present case, no tension of the vascular pedicle was noted after implantation of the graft. The inherent mobility of the mesentery allowed vascular transposition without visible vascular obstruction or irregular angulation resulting in obstruction of the vasculature.
Perforating ulceration of the duodenum has been reported in humans most commonly with peptic ulcer disease or trauma (4). The small perforations can be repaired with single layer closure and/or omental patching. However, a small subset (1% to 2%) represents giant gastroduodenal ulcerations, defined as greater than 2 to 3 cm in human patients, and presents a challenge for surgical repair (4–6). Repair methods reported include partial gastrectomy, jejunal serosal patch, jejunal pedicled graft, gastrojejunostomy, and gastric disconnection (4–6). A gastrojejunostomy reconstruction with gastric patching utilizing the remnant pyloric tissue was reported successfully in 1 case (7), with a second novel method consisting of the use of the ligamentum teres hepatis to combine laparoscopic and endoscopic modalities in a minimally invasive approach (8). To the authors’ knowledge, reports of similar procedures allowing preservation of surrounding structures in dogs are limited.
In humans, jejunal grafting has been used as pedicled, augmented, free segment, or a combination of these (2). The most common procedures reported include pharyngolaryngectomy defect, tracheal, esophageal, and urethral reconstruction, which are currently considered the optimal methods of circumferential pharyngolaryngectomy reconstruction (2,9–12). Reports of jejunal grafting in the veterinary literature are less frequent, though much information in the human literature has been derived from the use of canine models. Two reports examined canine models of tracheal replacement using autologous jejunal free flaps with various intra- and extra-luminal stents, and presented conflicting reports of success (13,14). Aside from thrombosis of the micro vessel anastomosis of the graft and mild secretory effects of the tissue, the remaining morbidity and mortality in these studies are assumed largely associated with the stenting techniques and the specific qualities of the organ being replaced.
The complications reported are believed inapplicable to our case, as microvascular anastomosis was not required, and secretory function of the graft’s mucosal layer does not present pathologic complications as with grafting into respiratory tissues. Other reported complications, including stricture formation and vascular thrombosis, were not encountered in this patient, and the second surgery allowed a unique opportunity for visualization of a post-operative graft in situ to definitively exclude these as short-term complications.
Jejunal autografts used for cervical and thoracic esophageal replacement in the dog demonstrated that accepted grafts retained gross and microscopic normalcy, along with peristalsis coordinating with esophageal contractions (15). A second study demonstrated frequency of fistulization and functional stenosis and dismotility in canines with cervical esophageal reconstruction using jejunal grafts (16). These grafts were shown in 1 canine study to maintain baseline motility (3). However, dysphagia is often noted due to incoordination of contractions and sustained local contractions, with the use of metoclopramide demonstrating an improvement in coordinated functional motility (3). Due to the potential demonstrated in these studies, the authors elected to maintain an orad to aborad orientation of the jejunal graft. Treatment with metotoclopramide was not pursued in this case due to low suspicion of ileus and no clinical signs deemed secondary to graft dismotility; however, the implication of this decision is unknown.
Further studies are required to determine clinical implications along with the morbidity and mortality associated with the use of jejunal pedicle grafting to repair duodenal defects. Controlled prospective in vivo videofluoroscopic and endoscopic monitoring of graft sites could be considered in order to further demonstrate the functionality associated with the described graft. Though promotility medications were not utilized in this case and no adverse results were noted, further studies may demonstrate the implications of such drugs for coordinated peristalsis and their effects at this location.
This case represents the novel use of a jejunal graft for successful repair of a large circumferential duodenal perforation at the level of the duodenal papilla in a dog. It is unclear if the septic peritonitis that developed after the initial procedure was due to graft leakage or the development of a pancreatic abscess. The leakage seen in the revision surgery was small and may have been due to supra-physiologic pressure applied during leak testing. The results support the use of this technique as a potential method to maintain physiologically relevant structures in this region. CVJ
Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (gro.vmca-amvc@nothguorbh) for additional copies or permission to use this material elsewhere.