Colonoscopy is a commonly performed procedure. The rate of adverse events is 2.8 per 1000 screening colonoscopies. These adverse events include cardiovascular and pulmonary events, abdominal pain, hemorrhage, perforation, postpolypectomy syndrome, infection, and death. Serious adverse events, such as hemorrhage and perforation, occur most frequently when colonoscopy is performed with polypectomy. This article highlights the prevention and management of adverse events associated with polypectomy and endoscopic mucosal resection of colonic lesions.
Key points
- •
Adverse events from endoscopic mucosal resection (EMR) and polypectomy include immediate and delayed bleeding, perforation, and postpolypectomy syndrome.
- •
Intraprocedural bleeding can be managed effectively with a variety of endoscopic modalities, including clips, detachable snares, and contact thermal probes or graspers, with or without epinephrine injection; the selection of one or a combination of techniques depends on the type of lesion, completeness of resection, device availability, and operator preference.
- •
Delayed postpolypectomy bleeding is self-limited in most cases and can be managed conservatively; endoscopic hemostasis is reserved for recurrent or ongoing bleeding.
- •
Immediate recognition and closure of EMR- or polypectomy-induced perforations are key determinants for a successful outcome. Endoscopic versus surgical management of perforations depends on defect size and access, presence of extraluminal contamination, and clinical status of the patient.
- •
Mucosal clip placement constitutes the mainstay of endoscopic therapy for perforation, although newer devices, such as the over-the-scope clip and endoscopic suturing, have expanded the options for closure.
- •
Close monitoring post perforation closure is essential in the context of a multidisciplinary approach, with prompt surgical intervention in the presence of clinical deterioration.
- •
Postpolypectomy syndrome is a transmural thermal injury that can mimic perforation, but whose prognosis is excellent with conservative management.
Introduction
Colonoscopy is a commonly performed procedure. The rate of adverse events (AEs) is 2.8 per 1000 screening colonoscopies. These AEs include cardiovascular and pulmonary events, abdominal pain, hemorrhage, perforation, postpolypectomy syndrome (PPS), infection, and death. Serious AEs, such as hemorrhage and perforation, occur most frequently when colonoscopy is performed with polypectomy. This article highlights the prevention and management of AEs associated with polypectomy and endoscopic mucosal resection (EMR) of colonic lesions.
Introduction
Colonoscopy is a commonly performed procedure. The rate of adverse events (AEs) is 2.8 per 1000 screening colonoscopies. These AEs include cardiovascular and pulmonary events, abdominal pain, hemorrhage, perforation, postpolypectomy syndrome (PPS), infection, and death. Serious AEs, such as hemorrhage and perforation, occur most frequently when colonoscopy is performed with polypectomy. This article highlights the prevention and management of AEs associated with polypectomy and endoscopic mucosal resection (EMR) of colonic lesions.
Polypectomy and endoscopic mucosal resection of colonic lesions
Standard polypectomy techniques involve hot or cold snaring without submucosal fluid injection. Lesions that are less than or equal to 1 cm in size can be resected safely via cold snare, whereas hot snare is usually used for larger lesions. A blended current is commonly used during hot snare polypectomy; however, there are proponents for the use of pure coagulation current.
The use of hot biopsy forceps is not recommended because of increased risk of complications, such as PPS and delayed perforation, and the availability of safer polypectomy techniques.
EMR is a modified version of saline-assisted polypectomy that is used in the colon to facilitate lesion resection and mitigate the risk of perforation associated with the removal of large sessile polyps. Although technical variations exist, most EMR techniques are centered on the concept of injecting a solution to provide a cushion between the mucosal and deeper layers of the colon wall. Specialized band ligation and cap-assisted EMR devices are commonly used in the esophagus and stomach, but have limited applicability in the colon. EMR in the colon usually consists of freehand snare resection following submucosal fluid injection. En bloc snare resection is preferred so that the depth and lateral margins of the resected specimen can be accurately assessed at histopathology, and this is generally feasible if the lesion is less than 2 cm in size. For lesions that are 2 cm or larger, piecemeal resection is recommended to decrease the risk of perforation.
The submucosal lift can be performed using a variety of solutions ( Table 1 ). A commonly used injectate consists of saline stained with a few drops of a dye (indigo carmine or methylene blue), with or without dilute epinephrine. A post-EMR defect that uniformly stains blue confirms that the resection plane is limited to the submucosal layer ( Fig. 1 ). A wide range of dilute epinephrine (1:10,000–1:100,000) in the mixture has been reported. Although epinephrine minimizes the risk of immediate bleeding and facilitates endoscopic visualization by maintaining a dry resection field, it does not prevent delayed bleeding. A longer-lasting fluid cushion can be obtained with the use of viscous solutions, such as hetastarch, succinylated gelatin, and hydroxypropyl methylcellulose. These solutions may reduce procedural time and the number of resections needed for completing piecemeal EMR.
Solution | Cushion Duration | Cost | Tissue Damage |
---|---|---|---|
Saline | + | Cheap | No |
Hypertonic saline | ++ | Cheap | Yes |
50% Dextrose | ++ | Cheap | Yes |
Glycerol | ++ | — | No |
Hyaluronic acid | +++ | Expensive | No |
Hydroxypropyl methylcellulose | +++ | Cheap | No |
Adverse events of endoscopic mucosal resection and polypectomy
The major AEs related to EMR and polypectomy include hemorrhage, perforation, and PPS. Clinically relevant stricture formation can result from wide or circumferential EMR, but this is more of an issue in the esophagus than in the colon.
Hemorrhage
Hemorrhage is the most common AE of colonoscopy with polypectomy, with reported incidences ranging from 0.1% to 0.5% for clinically significant bleeding. Hemorrhage can occur at the time of the procedure (immediate bleeding) or hours to weeks (delayed bleeding) after the procedure. However, most delayed bleeding events occur within 2 weeks. Intraprocedural and delayed bleeding caused by EMR of large colorectal lesions occur in 1% to 11% of cases, although a wide range of EMR-related bleeding estimates have been reported (0%–45%).
Predictors of hemorrhage
Several factors related to the patient, lesion, and resection technique can predict the risk of hemorrhage following polypectomy and EMR. In a large prospective study involving 9336 polypectomies in 5152 patients, immediate postpolypectomy bleeding occurred in 4% of cases. Significant risk factors for immediate bleeding included age 65 and older, use of anticoagulants, comorbid cardiovascular or chronic renal disease, polyp size greater than 1 cm, lesions featuring pedunculated polyps or laterally spreading tumors, suboptimal bowel preparation, and use of cutting current. In another study involving 6617 polypectomies in 3138 patients, delayed postpolypectomy bleeding occurred in 38 (0.57%) lesions and 37 (1.2%) patients. Hypertension and polyp size (10.0 ± 6.9 vs 5.6 ± 3.8 mm; P <.0001) were associated with increased risk of delayed bleeding, whereas lesion location and resection method did not seem to affect the bleeding risk. Other studies, however, have shown an increased risk for bleeding following removal of right-sided colonic lesions. In one small case-control study, polyp location and size were found to be independent risk factors for delayed bleeding. Polyps located in the right colon had an odds ratio (OR) of 4.67 (95% confidence interval [CI], 1.88–11.61; P = .001) for delayed hemorrhage. Similarly, polyp location proximal to the splenic flexure (OR, 2.9; 95% CI, 1.05–8.1) and size (OR, 1.3; 95% CI, 1.1–1.7 for each 10 mm increase in size) were associated with delayed bleeding in another report of colonic EMR for large (≥2 cm) lesions.
In a study involving 1657 patients and a polypectomy-related bleeding rate of 2.2%, warfarin was found to be an independent risk factor (OR, 13.37; 95% CI, 4.20–43.65) for polypectomy-induced bleeding, but not the use of nonsteroidal anti-inflammatory drugs, aspirin, and other antiplatelet agents. Data on the use of uninterrupted clopidogrel and postpolypectomy bleeding are conflicting. In a recent meta-analysis, continued clopidogrel use was shown to increase the risk of delayed but not immediate postpolypectomy bleeding, with a pooled relative risk ratio of 4.66 (95% CI, 2.37–9.17; P <.01). In contrast, a case-control study demonstrated no significant differences in the rates of delayed bleeding among clopidogrel users and nonusers, although the study findings were mostly applicable to small polyps (<1 cm).
There is no standardization regarding the use of electrosurgical current during snare polypectomy, and evidence-based data on which to base the optimal setting for EMR and polypectomy are lacking. In one study, all immediate postpolypectomy bleeding occurred when a blended current was used, whereas all delayed postpolypectomy bleeding occurred with the use of coagulation current. In a recent study, clinically significant delayed postpolypectomy hemorrhage was associated with the use of an electrosurgical current not controlled by a microprocessor (OR, 2.03; P = .038).
In summary, several variables have been associated with the risk for postpolypectomy bleeding. Risk factors that seem to be consistent among studies include polyp size greater than 1 to 2 cm, flat or laterally spreading lesions and pedunculated polyps with thick stalks, right-sided colonic lesions, resection technique (including type of electrosurgical current utilized), and coagulation status.
Management of postpolypectomy hemorrhage
Immediate (intraprocedural) hemorrhage
Constricting the residual stump with the snare and holding pressure can usually control immediate bleeding following transection of a pedunculated polyp with a thick stalk. If significant bleeding recurs following snare loosening, reconstricting the stalk for an additional period of time may result in hemostasis or, alternatively, dilute epinephrine (1:10,000 solution) can be injected in the base of the stump to reduce or stop bleeding and enable clearing of the field of view for more definitive therapy, such as placement of clips, a detachable snare (endoloop), or direct suture ligation.
In the situation where the residual stalk is too short to grasp with the electrosurgical snare or an endoloop, several hemostatic measures can be used, alone or in combination. Dilute epinephrine can be injected in and around the bleeding point, but care should be undertaken to avoid overinjection on the aboral side to avoid lifting the lesion away from view and placing it in a difficult position for subsequent therapy. The use of epinephrine is only a temporary measure and this should be followed by more definitive therapy. When technically feasible, the application of mechanical hemostatic devices is preferable because they do not extend the depth of tissue injury, which occurs with the use of thermal devices. Endoscopic clips can be placed directly on the bleeding point or on the residual stalk. Alternatively, a detachable snare can be used if access to and length of the residual stalk are favorable. If thermal probes are used, coaptive coagulation for 3 to 5 seconds is recommended at settings of 15 J for the heater probe and 12 to 15 W for the bipolar coagulation probe. The tip of the snare can also be used to control bleeding, but caution using this technique is warranted to avoid deep monopolar thermal injury. The Coagrasper device (Olympus Corp., Tokyo, Japan) is a monopolar coagulation forceps designed for grasping, tenting, and sealing of nonbleeding and bleeding vessels. Although it is used primarily during endoscopic submucosal dissection (ESD) for coagulation of submucosal vessels, it can be effective at grasping and sealing the bleeding vessel atop the resected stalk or polypectomy base, provided the grasped tissue can be tented.
Treatment of bleeding that occurs during en bloc or piecemeal EMR may not be necessary (eg, mild transient oozing at the resection edge). Endoscopic therapy should be reserved for active bleeding that interferes with completion of the procedure, or persistent oozing that has not ceased by the end of the procedure. An endoscope with forward water jet irrigation is helpful to precisely identify the location of the bleeding point. An actively bleeding vessel within an EMR defect is best treated with the grasping coagulation forceps (Coagrasper) using a soft coagulation mode at a power setting that produces prompt tissue coagulation (this can vary depending on the power setting [30–60 W] and type of electrosurgical unit; [CR] ) or with clip placement ( Fig. 2 ), with or without prior epinephrine injection. Clip placement, however, should be avoided if it has the potential to interfere with completion of resection. Argon plasma coagulation and contact thermal probes can extend the depth of tissue injury and should be used with caution.
Delayed hemorrhage
In the setting of delayed postpolypectomy hemorrhage, the basic tenets regarding management of gastrointestinal (GI) bleeding should be followed, including triage to the appropriate service (ward or intensive care unit) based on the severity of bleeding and patient comorbidities, fluid and blood transfusions as appropriate, withholding anticoagulation and antiplatelet agents, and correction of coagulopathy if present.
Patients in whom bloody bowel movements have ceased at the time of admission can usually be managed conservatively because rebleeding is uncommon. If there are no signs of ongoing bleeding during administration of the colon preparation, colonoscopy may be deferred with the exception of patients requiring prompt resumption of antithrombotic therapy. In patients in whom bleeding recurs during observation or in those with ongoing bloody stools, urgent colonoscopy should be performed.
The significance and rebleeding rates of stigmata of recent hemorrhage found in postpolypectomy ulcer beds are not as well studied as those related to bleeding peptic ulcers. In general, the ulcer experience is translatable to the postpolypectomy ulcer. Postpolypectomy ulcers with clean bases or flat, pigmented spots do not require intervention. Pigmented protuberances (visible vessels), adherent clots, and active bleeding require endoscopic therapy. Clip placement or thermal therapy, alone or in combination with epinephrine injection, can be used to treat the underlying bleeding stigmata. Mucosal clips can be applied directly on the vessel or the entire postpolypectomy defect ( [CR] ). The latter may not be possible because of induration of the polypectomy base margins ( [CR] ). Similarly, it may be difficult for the Coagrasper device to grasp and tent a visible vessel within the indurated ulcer bed. This can result in avulsion of the vessel with subsequent bleeding. Some of these lesions may be amenable to placement of an over-the-scope clip (OTSC; Ovesco Endoscopy AG, Tübingen, Germany) ( [CR] ), but practical issues may limit the use of this device because removal of the endoscope is required for loading of the OTSC. Similarly, direct suture ligation may be performed after withdrawing the colonoscope (OverStitch; Apollo Endosurgery, Austin, TX), provided the bleeding site can be reached with a double-channel upper endoscope. An indurated polypectomy base provides a safety cushion for the use of contact thermal probes. These modalities may be more effective in this setting (see [CR] ). A hemostatic spray (Hemospray; Cook Medical Inc, Bloomington, IN) is effective in active bleeding but is not available in the United States.
Endoscopic therapy is usually successful in postpolypectomy bleeding. If bleeding cannot be controlled at the time of endoscopy, salvage angiographic embolization should be considered as the next step. Endoscopically placed clips are effective markers of the bleeding site and facilitate subsequent supraselective angiographic embolization using microcoils to minimize the risk of colonic ischemia. Surgical intervention for torrential postpolypectomy hemorrhage is a rare occurrence.
Prevention of hemorrhage
Placement of a detachable snare (PolyLoop; Olympus Corp., Tokyo, Japan) or injection of epinephrine into the stalk of particularly large pedunculated polyps (≥2 cm) can significantly reduce bleeding AEs after snare polypectomy ( [CR] ). Moreover, the combined use of epinephrine injection followed by detachable snare placement has been reported superior to epinephrine injection alone in decreasing the proportion of early postpolypectomy bleeding episodes in patients with large pedunculated polyps. Placement of the detachable snare before polyp resection can be problematic for a pedunculated lesion with a large polyp head situated in a narrow lumen or at an angulated location. In addition, the detachable snare cannot be reopened once constrained and can become entangled inadvertently in the polyp head rather than the stalk, making subsequent attempts at snare resection difficult or impossible. For these reasons, prophylactic placement of the detachable snare following polyp transection is more desirable as long as a residual stalk of sufficient length is maintained. Detachable snares for large pedunculated polyps (≥2 cm) with thick stalks should be considered in patients in need of antithrombotic therapy. Endoscopic clip placement across the base of the stalk is an alternative treatment option in the patient on antithrombotic therapy, provided that near complete cross-sectional compression of the stalk can be achieved. Contact between the clips and the snare results in transmission of the electrical coagulating current into the clip during polypectomy. In one randomized trial, prophylactic clip placement was as effective as the detachable snare in the prevention of postpolypectomy bleeding in large pedunculated polyps ( [CR] ).
In two studies, prophylactic clip placement did not prevent delayed bleeding from postpolypectomy ulcers, although the mean size of the ulcers was small in one of the studies. A retrospective study assessing the role of prophylactic clipping following resection of large flat or sessile polyps (≥2 cm) suggested that not clipping the EMR defect was associated with an increased risk for delayed bleeding (OR, 6.0; 95% CI, 2.0–18.5). Until more robust data are available to further define the role of prophylactic clip placement for large EMR defects, universal use of clips cannot be endorsed currently. However, consideration should be given for clipping EMR defects greater than 2 cm in size that either may have demonstrated oozing or contain an indeterminate focal (vascular) spot, particularly in the right colon and in patients who require immediate antithrombotic therapy to decrease the risk of postpolypectomy bleeding ( Fig. 3 ).