Stuck scope: Tackling the adherent scope situation after hemostatic powder application: A case report and review of literature

Abstract

BACKGROUND

TC-325 (Hemospray) is a mineral-based hemostatic powder that is increasingly used for nonvariceal upper gastrointestinal bleeding, particularly diffuse or malignancy-related bleeding. Adhesion of the powder to the endoscope with inability to withdraw the instrument is a rare but potentially serious adverse event that prompted a manufacturer device-correction notice in 2023, and it has most often been reported when the powder is applied in retroflexion. Descriptions of how to free an entrapped endoscope remain limited.

CASE SUMMARY

An 84-year-old man with metastatic jejunal mucinous adenocarcinoma presented with hematemesis and a hemoglobin of 7.5 g/dL. Upper endoscopy with push enteroscopy identified a 3-cm cratered gastric body ulcer with stigmata of recent bleeding (Forrest class IIc) and the known obstructing jejunal tumor. TC-325 was applied to the ulcer with the endoscope in partial retroflexion. On attempted withdrawal the endoscope was fixed: Hydrated powder had coated the insertion tube and bonded it to the mucosa at the gastroesophageal junction. Water irrigation, viscous lidocaine, gentle torque and forceps clearance through the entrapped endoscope did not release it. A second ultrathin (XP) gastroscope was advanced alongside the entrapped endoscope, and biopsy forceps passed through the ultrathin scope were used to strip the adherent powder from the shaft under direct vision, after which the endoscope was withdrawn without resistance and without mucosal injury. The patient had no further clinically evident bleeding during the admission and was discharged to home hospice in accordance with his goals of care.

CONCLUSION

TC-325 applied in retroflexion can bond the endoscope to the mucosa. Early recognition, avoidance of forceful traction, irrigation, controlled de-retroflexion and, when these fail, rescue with a parallel ultrathin endoscope can free the instrument safely.

Key Words: Endoscopy; Hemostatic powder; Hemospray; Endoclot; Gastrointestinal bleeding; Stuck endoscope; Gastroenterology; Case report

Core Tip: Gastrointestinal bleeding can be challenging to treat even in the most experienced hands. The modalities available to gastroenterologists for achieving hemostasis have expanded over the years, and this area continues to evolve. Hemostatic powders are one such valuable innovation; they have become widely popular and are often used. However, awareness of unique risks, such as failure to withdraw the endoscope because of mucosal adherence, and knowledge of how to prevent and overcome this challenge are crucial to continued use.

INTRODUCTION

Acute upper gastrointestinal (GI) bleeding remains a leading cause of hospital admission. Although mortality has declined over the past two decades, it remains approximately 1%-5% in contemporary series and is higher in older patients with comorbidity[1,2]. Endoscopic hemostasis is first-line therapy and has traditionally relied on injection, thermal and mechanical modalities, each of which requires precise targeting of a discrete bleeding point and carries a 5%-10% rebleeding risk[3,4]. TC-325 (Hemospray; Cook Medical, Winston-Salem, NC, United States) is an inorganic, mineral-based hemostatic powder approved by the United States Food and Drug Administration in 2018 for nonvariceal upper GI bleeding. On contact with luminal moisture and blood it absorbs water and forms a cohesive, adhesive mechanical barrier that tamponades the bleeding site and concentrates clotting factors[3,5]. Because it requires neither en-face orientation nor identification of a single vessel, it is well suited to diffuse, malignant or endoscopically inaccessible bleeding. Its guideline position should be stated precisely: The 2021 American College of Gastroenterology guideline suggests TC-325 for actively bleeding ulcers as a conditional recommendation based on very-low-quality evidence[4]; the 2021 European Society of Gastrointestinal Endoscopy guideline suggests a topical hemostatic spray/powder for bleeding refractory to standard modalities (weak recommendation, low-quality evidence)[6]; and the 2020 American Gastroenterological Association clinical practice update, an expert review rather than a formal guideline, considers it useful as rescue or temporizing therapy, particularly for malignant bleeding[7]. Randomized and registry data support its efficacy in malignancy-related bleeding[8-12].

Reported adverse events are uncommon and include catheter or working-channel blockage, transient biliary obstruction, perforation and, rarely, difficulty withdrawing the endoscope[8,13,14]. In May 2023 the manufacturer issued an urgent medical device correction warning that Hemospray powder may adhere to the distal endoscope and result in difficulty or inability to maneuver or remove the instrument, with potential for mucosal tear, perforation, hemorrhage and death; in some, but not all, reported instances the powder had been sprayed with the endoscope in a retroflexed position[15]. Retroflexion, routinely used to inspect and treat lesions tangential to the forward view such as the gastric cardia and fundus or the distal rectum[16], is itself low-risk but can cause perforation even in experienced hands[17,18].

Endoscope adhesion after TC-325 application has been described previously, including during retroflexion, and prior reports were resolved by traction or torque, prolonged irrigation, delayed removal after the powder had sloughed or, in one report, a fiberoptic bronchoscope advanced alongside the entrapped gastroscope (Table 1)[19-27]. The contribution of the present report is therefore not the complication itself but its management: After irrigation, lubrication, torque and forceps clearance through the entrapped endoscope had failed, the instrument was freed by advancing a second ultrathin gastroscope alongside it and stripping the adherent powder under direct vision. To our knowledge, the use of a second endoscope with a working channel, through which the adherent powder is stripped from the entrapped instrument under direct vision, has not previously been reported for powder-related entrapment. In the only prior report in which a second instrument was used, Crowley[23] freed an entrapped gastroscope with a fiberoptic bronchoscope advanced alongside it, dislodging the adhesive with gentle back-and-forth movements of the bronchoscope; a second endoscope was mentioned as a theoretical option in one pediatric report but was not attempted[25], and the double-endoscopy technique described by Huang et al[28] addressed mechanical impaction of a retroflexed gastroscope rather than an adhesive bond. This report is written in accordance with the CARE guidelines for case reports[29,30].

Table 1 Published reports of endoscope entrapment or retention attributed to TC-325 (Hemospray) adhesion, distinguishing full case reports from conference abstracts.

Ref.	Report type	Hemostatic agent	Indication/bleeding source	How the scope became stuck	How it was freed/managed	Complications/outcome
Smith et al[19] and Smith et al[20]	Conference abstract (SEAL prospective registry interim data, n = 82), with the same two events reported in the subsequent full registry publication (n = 63)	TC-325	Nonvariceal upper GI bleeding (UGIB), various lesions	Endoscope became adherent to esophageal mucosa after application in retroflexion (2 patients)	Not detailed	Not detailed; listed as technical complications
Yii et al[21]	Full case report	TC-325 (Hemospray)	Bleeding gastric cardia malignancy (lymphoma)	Powder adhesion after application; endoscope retained at the gastroesophageal junction	Endoscope left in situ under intensive care and withdrawn after approximately 48 hours once the powder had sloughed	Prolonged retention (approximately 48 hours); mechanical ventilation/critical care
Shaikh and Parikh[22]	Conference abstract	TC-325	Bleeding of gastric cardia and fundus	Adhesion after retroflexed application; endoscope retained approximately 6 hours	Traction after transfer to a tertiary center	Scope freed; further outcome not detailed
Crowley[23]	Full case report (anesthesia journal)	TC-325	Oozing esophageal mass (nonvariceal UGIB; hemoglobin 6.1 g/dL)	Gastroscope could be neither advanced nor withdrawn after application and was found adherent to the esophagus; position of the gastroscope during spraying (retroflexion) not stated	Sedation stopped to protect the airway; fiberoptic video bronchoscope passed orally alongside the gastroscope, secretions suctioned and the adhesive dislodged from the gastroscope with gentle back-and-forth movements; freed in under 15 minutes	No complication; procedure completed; report written from the anesthesia (airway-management) perspective
Syed et al[24]	Full case report	TC-325	Bleeding gastric GIST near the gastroesophageal junction (fundus in the abstract; lesser curve in the case description), treated in retroflexion	Powder coated the scope at the gastroesophageal junction and adhered to mucosa	Torque maneuvering plus withdrawal force (described as moderate in the abstract and as significant in the case description)	Scope freed; no recognized injury, although the authors could not exclude tissue trauma from the force applied; started on imatinib
Edwards et al[25]	Full case report (pediatric)	TC-325	Pediatric upper GI bleeding (22-month-old)	Transient adhesion of the distal endoscope to the esophagus after diffuse application in retroflexion	Constant traction for several seconds with gentle side-to-side rotation; a second endoscope was considered but not feasible because of patient size	No esophageal trauma on re-intubation and no immediate or long-term adverse effects; normal repeat endoscopy at 3 months
Guzik et al[26]	Conference abstract	TC-325	Retained endoscope, case series (upper GI bleeding); number of patients not stated in abstract	Powder adhesion	Not detailed (conference abstract)	Not detailed
Echeverria et al[27]	Full case report	TC-325 (with PuraStat gel)	Gastric cardia/fundus AVMs (arteriovenous malformations), upper GI bleeding, INR 7 (retroflexion)	Powder adhesion after retroflexed application	Aggressive water irrigation; scope removed after 90 minutes	Prolonged retention (90 minutes); no further GI bleeding documented; lost to outpatient follow-up
Present case	Full case report	TC-325	Bleeding gastric body ulcer (partial retroflexion)	Insertion tube coated with hydrated powder and bonded to cardia/gastroesophageal junction mucosa after application in partial retroflexion	Irrigation, viscous lidocaine, torque and forceps through the entrapped scope failed; second ultrathin (XP) gastroscope advanced alongside and forceps used to clear the adherent powder under direct vision	Scope freed after 30 minutes; no mucosal injury; no clinically evident rebleeding during admission; discharged to hospice

INR: International normalized ratio; GI: Gastrointestinal.

CASE PRESENTATION

Chief complaints

An 84-year-old man presented to the emergency department with several episodes of hematemesis as well as melena of a few days’ duration.

History of present illness

The patient reported several episodes of sudden-onset hematemesis as well as melena of a few days’ duration. He was not on nonsteroidal anti-inflammatory drugs and he was not taking antiplatelet or anticoagulant therapy.

History of past illness

He had a history of colon cancer (in the remote past, with details unknown) and of jejunal mucinous adenocarcinoma diagnosed a few months earlier. At that time he had undergone exploratory laparotomy, which revealed a large jejunal mass and extensive peritoneal carcinomatosis with metastatic deposits up to 10 cm; an internal bypass from the jejunum to the mid-transverse colon had been created to relieve obstruction. He had subsequently started palliative chemotherapy with FOLFOX and had received one cycle.

Personal and family history

He had no history of tobacco or alcohol use. No known family history of GI malignancy or bleeding diathesis.

Physical examination

On admission the patient was tachycardic (110 beats/minute) and hypotensive (83/64 mmHg). He appeared pale on evaluation and abdominal examination revealed mid-abdominal and epigastric tenderness.

Laboratory examinations

Hemoglobin was 7.5 g/dL compared with a recent baseline of 9 g/dL. The platelet count of 405 × 10³/μL and international normalized ratio of 1.1 were unremarkable. He was also noted to have acute-on-chronic kidney injury (creatinine 1.4 mg/dL at baseline to 1.9 mg/dL). He received 3 units of packed red blood cells before endoscopy and was started on intravenous proton pump inhibitor therapy in the form of a loading dose 80 mg intravenously followed by continuous infusion at a rate of 8 mg/hour. The high-dose acid suppression therapy was initiated on presentation and continued for 48 hours post procedure before transitioning to 40 mg twice daily.

Imaging examinations

Computed tomography of the abdomen and pelvis showed no active contrast extravasation but demonstrated a large intra-abdominal mass involving the lumen of the duodenojejunal junction, consistent with the known malignancy (Figure 1).

Figure 1  Computed tomography of the abdomen and pelvis showing a large intra-abdominal mass involving the lumen of the duodenojejunal junction.

Endoscopic findings and procedural timeline

The procedure was performed on the day of presentation to the hospital within about 7 hours of admission under monitored anesthesia care with an anesthesia team in attendance. Upper endoscopy with push enteroscopy was performed with a pediatric colonoscope (Olympus PCF-H190 L) using carbon dioxide insufflation. A large, approximately 3 cm, cratered ulcer with heaped-up margins, several pigmented spots and areas of necrosis was identified in the gastric body at the junction of the greater curvature and posterior wall and was examined under white light and narrow-band imaging (Figures 2 and 3). The ulcer showed stigmata of recent bleeding without active spurting or oozing at the time of inspection (Forrest class IIc). Cold-forceps biopsies were taken from the ulcer edges. In the proximal jejunum an obstructing mass with overlying mucus was seen, consistent with the known jejunal adenocarcinoma (Figure 4); no bleeding was identified from the tumor. The endoscope was withdrawn from the small bowel into the stomach without resistance.

Figure 2  White-light view of the large cratered gastric ulcer with stigmata of recent bleeding in the gastric body at the junction of the greater curvature and posterior wall.

Figure 3  Narrow-band imaging view of the cratered gastric ulcer.

Figure 4  Obstructing mass with overlying mucus in the proximal jejunum.

The decision to treat the gastric ulcer with TC-325 rather than, or in addition to, conventional injection, thermal or mechanical therapy was based on the following considerations: The lesion was large with a broad necrotic base and friable heaped-up margins suspicious for malignant infiltration, there was no single targetable vessel, and the patient's advanced malignancy and goals of care favored a rapid, non-contact approach. TC-325 was delivered through the Hemospray catheter in short bursts with the endoscope in partial retroflexion to bring the entire ulcer into view (Figure 5).

Figure 5  Application of TC-325 to the cratered gastric ulcer with the endoscope in partial retroflexion.

Immediately after application, attempted withdrawal met firm resistance and the endoscope could be neither advanced nor withdrawn. In the retroflexed view the insertion tube was coated with hydrated powder that bonded it to the mucosa of the cardia and gastroesophageal junction (Figure 6; the adhesion site is indicated by an arrow). Entrapment was recognized within a few minutes, forceful traction was deliberately avoided, insufflation was minimized, and the anesthesia team was informed so that the airway remained secured. The patient did not require intubation but was closely monitored. The following maneuvers were then attempted in sequence over approximately 20 minutes: (1) Vigorous water irrigation through the working channel and water-jet directed at the interface in the retroflexed view; (2) Instillation of viscous lidocaine over the adherent region in an attempt to lubricate the interface, an empirical choice by the endoscopist without published supporting data; (3) Gentle torque and attempted controlled de-retroflexion in both retroflexed and forward views, which produced no movement; and (4) Passage of biopsy forceps through the entrapped endoscope to peel powder off the shaft in the retroflexed view, which was only partially effective because of the limited working angle (Figure 7). Fluoroscopy and surgical backup were available and considered but not required.

Figure 6 Retroflexed view showing the insertion tube coated with hydrated TC-325 and bonded to the mucosa at the cardia/gastroesophageal junction. Arrow indicates the adhesion site.

Figure 7  Attempt to remove adherent TC-325 from the insertion tube with biopsy forceps passed through the entrapped endoscope in the retroflexed view.

A second, ultrathin gastroscope (Olympus GIF-XP190N) was then passed orally alongside the entrapped pediatric colonoscope. The ultrathin scope provided an independent view of the adhesion site from the esophageal side, and biopsy forceps passed through its working channel were used to strip the adherent powder from the insertion tube and the adjacent mucosa under direct vision (Figure 8). Care was taken to grasp powder rather than mucosa and to avoid contact with the lens and bending section of the entrapped instrument. Once the bond was cleared, the pediatric colonoscope moved freely and was withdrawn without resistance. The total duration of entrapment was approximately 30 minutes and the total procedure time was approximately 60 minutes.

Figure 8  Ultrathin (XP) gastroscope advanced alongside the entrapped pediatric colonoscope; biopsy forceps passed through the ultrathin scope are used to clear the adherent TC-325 from the insertion tube under direct vision.

The ultrathin gastroscope was then used to re-examine the entire esophagus, the gastroesophageal junction in forward and retroflexed views, and the stomach; there was no mucosal tear, laceration, hematoma or perforation, and the powder remained in place over the ulcer.

FINAL DIAGNOSIS

Bleeding gastric body ulcer, moderately differentiated adenocarcinoma on biopsy, in a patient with metastatic jejunal mucinous adenocarcinoma, complicated by TC-325-related endoscope entrapment during partial retroflexion. The pathology results supported the likelihood that this was a metastatic ulcer given the underlying history.

TREATMENT

Treatment consisted of endoscopic hemostasis with TC-325 monotherapy, intravenous proton pump inhibitor therapy with continuous infusion for 72 hours, transfusion of a total of three units of packed red blood cells, and rescue of the entrapped endoscope as described above. No further endoscopic therapy was required.

OUTCOME AND FOLLOW-UP

The patient had no further hematemesis or melena, and hemoglobin remained stable at 10 g/dL over the remaining 5 days of hospitalization; no repeat transfusion was required. In view of his advanced malignancy he was discharged to home hospice on the last hospital day. Follow-up consisted of close monitoring with home hospice during which no clinically evident rebleeding was reported. Because hospice enrollment precluded surveillance endoscopy or routine laboratory testing, we characterize the outcome as absence of clinically evident rebleeding during the follow-up period rather than confirmed durable hemostasis.

DISCUSSION

TC-325 achieves hemostasis through the physical property that also makes it hazardous: On contact with moisture it transforms from a free-flowing powder into a cohesive, adherent mass[3,5]. The same coalescence that seals a bleeding surface can bind to the insertion tube, bending section or distal cap of the endoscope. Edwards et al[25] proposed that inactive powder layered onto the endoscope surface becomes activated when the straightened instrument is withdrawn against moist esophageal mucosa. In retroflexion this risk is amplified: The shaft is acutely angulated within a confined lumen, the tip and shaft lie in close apposition to the mucosa and to each other, and powder dispersed in a retroflexed position settles onto the scope rather than away from it. A layer of hydrating powder bridging the angulated scope and the surrounding wall can mechanically lock the instrument. Overapplication, spraying close to the shaft and aspiration of powder into the working channel (which can also block suction) probably compound the effect. In our case the endoscope was in partial rather than full retroflexion, indicating that partial retroflexion does not eliminate the risk. These observations concern TC-325 only; we found no published reports of endoscope entrapment with EndoClot or other topical agents, and whether they carry a similar risk is unknown.

Endoscope entrapment after TC-325 has now been described in at least 10 patients across five full case reports and three conference abstracts (Table 1)[19-27]. Of these, eight patients are individually documented (two in the SEAL registry[19,20] and one in each of six single-patient reports[21-25,27]); the remaining conference abstract describes a case series without stating the number of patients[26], so the total of at least 10 assumes a minimum of two patients in that series. The complication also appears among events reported to the MAUDE database[13,14] and in the manufacturer’s 2023 device correction[15]. The reported rescues have relied on traction or torque[22,24,25], prolonged irrigation for up to 90 minutes[27], or deferred removal under intensive care after approximately 48 hours once the powder had sloughed[21], with attendant airway and critical-care implications when an upper endoscope cannot be withdrawn[23]. In the anesthesia report by Crowley[23], the gastroscope was freed in under 15 minutes by a fiberoptic video bronchoscope passed orally alongside it: Secretions were suctioned and the adhesive was dislodged from the gastroscope with gentle back-and-forth movements of the bronchoscope. That report is written from the airway-management perspective and does not describe endoscopic details such as the position of the gastroscope during spraying. Our case shares the parallel-instrument principle but differs in the instrument and in the method of release: Rather than applying additional force, waiting for the bond to degrade or rubbing the bond loose with the second instrument itself, a second ultrathin gastroscope was advanced alongside the entrapped instrument, providing an independent endoscopic view of the adhesion site from the esophageal side and a working channel through which biopsy forceps stripped the powder from the insertion tube under direct vision. This differs from the double-endoscopy technique of Huang et al[28], which used a second endoscope to release a retroflexed gastroscope mechanically impacted in the distal esophagus and did not involve an adhesive bond, and from the report of Yii et al[21], in which the endoscope was retained until the powder detached. A second endoscope was raised as a theoretical option by Edwards et al[25] but was not feasible in their 22-month-old patient. Taken together, the published reports indicate a reproducible, mechanism-based event rather than an isolated anomaly.

Prevention

It is important to separate the manufacturer's instructions, published evidence and our own inference. The instructions for use state that when spraying in the retroflexed position the powder may adhere to the outside of the endoscope, and the 2023 device correction adds that adhesion may also occur in the esophagus or stomach and that providers should be prepared to take immediate steps to manage it[15]. Product labeling and technical reviews advise delivering the powder in short bursts from a short distance (approximately 1-2 cm) without contact between the catheter and the target[3,8]. The published case reports support limiting the amount of powder applied in retroflexion and near the gastroesophageal junction, moistening any pale, inactivated powder visible on the endoscope before withdrawal, and treating in a straightened position whenever the lesion can be reached that way[24,25,27]. Beyond these points, the following are our own suggestions based on the mechanism and on this case and have not been validated: Extend the catheter beyond the endoscope tip and aim it at the target rather than across the shaft; use the smallest effective dose, because efficacy does not require coating the field; and take particular care in confined spaces such as the cardia, duodenal bulb and rectal vault, where powder can bridge the endoscope and the opposing wall.

Management

We propose the following sequence for an endoscope that will not withdraw after TC-325 application, drawn from the prior reports, the manufacturer’s notice and our experience; except where referenced, it represents expert opinion rather than tested evidence: (1) Recognize entrapment immediately and stop; forceful traction risks mucosal tear, perforation and, when the endoscope is in the esophagus, airway compromise[15,23]; (2) Alert the anesthesia team at once so that the airway can be secured or maintained and the patient kept still and comfortable for what may be a prolonged rescue[23]; (3) Reduce insufflation to decompress the lumen (our suggestion); (4) Apply copious water irrigation at the interface; the powder does not truly dissolve, but hydration and mechanical loosening allowed release after prolonged irrigation in one report[27] and were suggested as a countermeasure in another[25]; (5) Apply gentle torque and attempt controlled de-retroflexion under direct vision, as reported by Syed et al[24], rather than abrupt pulling; (6) Attempt mechanical clearance of the bond with forceps through the entrapped endoscope; (7) If these fail, advance a second ultrathin or pediatric endoscope alongside the entrapped instrument to clear the bond under direct vision, as in the present case, or advance a fiberoptic bronchoscope alongside the gastroscope to dislodge the adhesive, as reported by Crowley[23]; conceptually this resembles the double-endoscopy approach to mechanical impaction[28], although that report did not involve powder; (8) The ultrathin endoscope can be safely introduced adjacent to the adult gastroscope given its thin caliber and flexibility. Though it is flexible, there can be some difficulty with maneuverability within the esophagus. However, instruments can be safely and cautiously introduced (such as forceps) through the ultrathin scope as well if needed to aid in disimpaction. One of the limitations is not being able to traverse the gastroesophageal junction itself given the entrapped endoscope. If the endoscope remains fixed, fluoroscopy may help define the configuration of the instrument (our suggestion, not tested), the surgical and interventional radiology teams should be made aware, and, as reported by Yii et al[21], the endoscope may be left in situ under intensive care until the powder detaches, which occurs within 24-72 hours[12,25]; and (9) After release, inspect the esophagus, gastroesophageal junction and stomach for injury and maintain a low threshold for cross-sectional imaging if perforation is suspected. Viscous lidocaine was an empirical choice in our case and we cannot attribute any benefit to it.

This complication should be weighed against the independent, low but real perforation risk of retroflexion[17,18] and against the established efficacy and generally favorable safety profile of TC-325 across upper, lower and malignant GI bleeding[8-12]. The lesson is not to avoid the powder but to apply it with deliberate technique and to recognize entrapment early. Limitations of this report include its single-patient design, the retrospective recording of some procedural details, and the short and non-endoscopic follow-up dictated by hospice care, which limits conclusions about durability of hemostasis.

CONCLUSION

TC-325 applied with the endoscope in retroflexion, even partial retroflexion, can bond the instrument to the mucosa at the gastroesophageal junction. The practical learning points are early recognition, avoidance of forceful traction, prompt involvement of anesthesia, copious irrigation and controlled de-retroflexion under direct vision, with a second ultrathin endoscope advanced alongside the entrapped instrument as a rescue option when simpler measures fail.