Endoscopic Management of Iatrogenic Duodenal Perforation With Linear-Probe Echoendoscope

Endoscopic ultrasound (EUS) has become an important diagnostic and interventional tool for endoscopists. Initially used for diagnostic purposes, EUS has now expanded to include interventional applications (such as fine-needle aspiration, biopsy, and injection), therapeutic drainage procedures, and neurolysis.¹ In the hands of an experienced operator, complication rates are similar to those associated with standard upper endoscopy.¹ Possible complications of EUS include infection, sedation-related cardiopulmonary events, and perforation. Traditionally, iatrogenic gastrointestinal perforations secondary to endoscopy have been managed surgically. Recently, data have emerged revealing that immediate endoscopic treatment with an endoclipping device may be an acceptable alternative to surgery in select cases.^2–5 Multiple reports have described perforation management, but there are limited data reporting cases of duodenal perforation specifically attributable to linearprobe echoendoscopes.^2–5 We describe the endoscopic management of iatrogenic duodenal perforation during linear EUS examination with endoscopic clipping device placement and subsequent medical management.

Case #1

A 68-year-old woman with a history of gastric carcinoid presented for evaluation. The linear-probe echoendoscope (Pentax Medical Co.) was advanced to the second portion of the duodenum. A 7-mm transmucosal tear was immediately visualized in the postbulbar duodenum (Figure 1). A standard upper endoscope was used to deploy 7 endoscopic clipping devices (QuickClip HX-200U-135; Olympus America, Inc.) to fully approximate the lesion (Figure 2). After admission, broad-spectrum antibiotics were administered and surgical consultation was obtained. Computed tomography (CT) scan at 24 hours did not reveal contrast leakage. The patient was discharged 3 days postprocedure, and follow-up at 120 days was uneventful.

Figure 1.

Transmucosal tear immediately visualized in the postbulbar duodenum.

Figure 2.

Endoclip deployment to the approximate defect.

Case #2

An 87-year-old woman underwent diagnostic EUS for evaluation of a suspected pancreatic mass. CT scan revealed common biliary duct dilatation. Examination with a linear-probe echoendoscope (Pentax Medical Co.) revealed a 1.5-cm irregular mass in the head of the pancreas. Prior to intervention, a 5-mm tear was visualized at the junction of the duodenal bulb and descending duodenum. The defect was approximated entirely with multiple endoscopic clipping devices (QuickClip HX-200U-135; Olympus America, Inc.). Of note, the lumen was very difficult to insufflate prior to clipping; after clipping was performed, normal insufflation was possible (Figure 3). After admission, broad-spectrum antibiotics were administered. CT scan at 24 hours revealed no contrast leakage. The patient was discharged on postprocedure day 6, and follow-up at 120 days was uneventful.

Figure 3.

A 5-mm tear at the junction of the duodenal bulb and descending duodenum after deployment of multiple endoclips.

Discussion

Our cases demonstrate the feasibility of endoscopic management in linear EUS-related perforations. In select patients, surgery can be avoided by using endoscopic clipping devices to repair the defect. The success of endoclipping depends on immediate recognition and closure of the defect. Theoretically, immediate closure of intestinal perforation minimizes leakage of digestive enzymes, bile, and bacteria to surrounding organs. With less inflammatory damage and bacterial contamination, antibiotics and symptomatic management may be sufficient to prevent further morbidity. In addition, the use of endoscopic clipping devices is safe, with a much lower rate of complications compared to emergent surgical intervention. Endoscopic clipping devices do not impair healing or re-epithelialization of the closed defect margins, and there have been no reports of clip impaction, perforation, or other significant complications.² The use of the endoscopic clipping device as a therapeutic tool was initially described in 1975 by Hayashi and colleagues, though clinical application of the device was initially limited. Advances in instrument design by the late 1980s enabled routine use for hemostasis in gastrointestinal bleeding.⁶ Uses have now expanded to include closure of tissue defects, perforations, and anastomotic leaks, as well as prevention of postpolypectomy bleeding, placement of enteral feeding tubes, and marking sites for surgical excision and endoscopic re-examination.⁷ Binmoeller and col-3). leagues reported the use of an endoscopic clipping device into close an iatrogenic gastric perforation after snare resection of a leiomyoma.⁸ Consequently, Kaneko and coworkers described endoscopic closure of a duodenal perforation secondary to endoscopy in 1999.⁹ Regarding EUS-related perforation, Seibert described the use of endoscopic closure of iatrogenic duodenal perforation during EUS examination.³ Our results concur with previous data that endoscopic intervention is a useful alternative for small erforations that are recognized immediately. In addition, these results are reported only in association with linear-probe echoendoscopes, for which data are limited.

The most commonly reported complication associated with EUS is esophageal perforation.¹⁰ Duodenal perforations have been reported with much less frequency, and data stratifying perforation by probe type on, (radial versus linear) are more limited. A recent retro-spective study of 11,539 EUS procedures reported an use overall complication rate of 0.12%, which included a 0.046% complication rate among 10,731 diagnostic EUS procedures.¹¹ With regard to probe type (radial versus linear), the study showed a 0.033% complication rate for radial-probe EUS, a 0.1% complication rate for diagnostic EUS using a linear probe, and a 1.11% use complication rate after interventional EUS using a linear robe. Other data have reported a complication rate of less than 0.1% for diagnostic EUS and approximately 2% for therapeutic EUS.^1,12 In a prospective study of 20,000 EUS procedures, 7 deaths (0.00035%) were attributable to duodenal perforation, all of which occurred with linear probes.¹ The study did not mention whether endoscopic intervention was attempted in any of the fatal cases. Our endoscopy unit (2 endosonographers) has performed 3,791 EUS procedures over the last 4 years (average, 947/year). Eight perforations have occurred, resulting in a perforation rate of 0.13%, similar to the above study. In total, both endosonographers have performed nearly 11,000 EUS procedures.

Previous data have suggested that risk factors for duodenal perforation during EUS include operator inexperience (<100 cases) and the presence of duodenal diverticula (Table 1).^11,13 The presence of these risk factors should be cause for additional caution during EUS examination. In addition to the above, the sudden inability to insufflate the lumen (case #2) should also raise suspicion for perforation.

Table 1.

Factors Associated With Increased Risk of Duodenal Perforation During Endoscopic Ultrasound

Operator inexperience (<100 cases total) Anatomical variation (ie, duodenal diverticula) Use of linear probe

If iatrogenic duodenal perforation is suspected, we suggest the use of an endoclipping device combined with medical management as a potential alternative to emergency surgery. This management, however, should only be pursued in the following cases: immediate recognition and visualization of a tear during the procedure; the presence of an endoscopist adequately trained in the use of the endoclipping device; perforation size less than 10 mm; and availability of surgical backup in-house. In cases of high surgical risk, these criteria may serve as guidance for nonsurgical management. Endoscopic intervention should generally be followed by inpatient medical management, including a nil-per-os regimen, nasogastric suctioning, intravenous fluids and antibiotics, and observation for signs of decompensation, including peritonitis, subcutaneous emphysema, hemodynamic instability, and/or sepsis. Contrast should be obtained 24 hours postprocedure, or earlier as dictated by signs and symptoms, to evaluate for contrast leakage.¹⁴ If imaging does not reveal extravasation of contrast, a liquid diet may be initiated shortly thereafter.

Our cases demonstrate the feasibility of endoscopic management in linear-probe EUS-related duodenal perforations. As EUS with linear probe becomes more common, clinicians should be aware of management strategies that may potentially preclude surgical intervention and subsequent morbidity.