Bleeding risk and prevention after colorectal polypectomy: What differs between cold and hot resection?

Abstract

In this editorial, we discuss the retrospective cohort study published in World Journal of Gastroenterology by Hwang et al, which identified immediate bleeding and trainee involvement as predictors of delayed post-polypectomy bleeding. These findings are clinically important because post-polypectomy bleeding remains the most common adverse event after colorectal polypectomy, yet its mechanisms and risk profiles differ according to resection technique. Cold techniques, including cold forceps polypectomy and cold snare polypectomy, are generally associated with a low risk of delayed bleeding because they avoid thermal injury, although the risk may still increase with larger lesions, pedunculated morphology, antithrombotic therapy, and unfavorable location. Current guidance therefore favors cold snare resection for most diminutive and small colorectal polyps. In contrast, hot techniques such as endoscopic mucosal resection and endoscopic submucosal dissection exhibit a distinct bleeding pattern due to vessel damage from electrocautery and tissue necrosis after the procedure. In endoscopic mucosal resection, bleeding risk is associated with larger lesion size, proximal or right-sided colonic location, antithrombotic use, and extensive mucosal defects; in selected high-risk right-sided lesions, prophylactic clip closure may reduce delayed bleeding. In colorectal endoscopic submucosal dissection, delayed bleeding is more often seen in patients with large lesions, rectal location, antithrombotic exposure, advanced age, and greater comorbidity burden. Overall, the findings by Hwang et al reinforce that post-polypectomy bleeding should be assessed and prevented in a procedure-specific and individualized manner, integrating resection method, lesion characteristics, antithrombotic use, intraprocedural bleeding events, and operator-related factors.

Key Words: Colorectal polypectomy; Delayed post-polypectomy bleeding; Cold snare polypectomy; Endoscopic mucosal resection; Endoscopic submucosal dissection; Prophylactic clipping

Core Tip: Bleeding risk after colorectal polypectomy differs between cold and hot resection because the mechanisms of tissue injury are distinct. Cold resection generally lowers delayed bleeding by avoiding thermal injury, whereas endoscopic mucosal resection and endoscopic submucosal dissection carry bleeding risks related to electrocautery and post-procedural necrosis. Prevention should be individualized according to resection method, lesion characteristics, antithrombotic exposure, and the potential benefit of prophylactic closure.

This editorial refers to "Immediate post-polypectomy bleeding is associated with an increased risk of delayed post-polypectomy bleeding: A large retrospective cohort study" by Hwang et al, 2026; https://doi.org/10.3748/wjg.118775.

---

INTRODUCTION

A recent large retrospective study published in World Journal of Gastroenterology by Hwang et al[1] demonstrated that immediate post-polypectomy bleeding is associated with an increased risk of delayed bleeding, highlighting the importance of identifying high-risk patients during endoscopic resection. Colorectal cancer continues to represent a major global health burden, and early detection and endoscopic removal of colorectal polyps are key strategies for colorectal cancer prevention[2,3]. A wide range of endoscopic resection techniques is now available, including cold snare polypectomy (CSP), cold endoscopic mucosal resection (C-EMR), hot endoscopic mucosal resection (H-EMR), and endoscopic submucosal dissection (ESD). Nevertheless, post-polypectomy bleeding remains a clinically important adverse event, with delayed bleeding being the most common major complication after colorectal endoscopic resection. Because cold resection and electrocautery-based hot resection differ in their mechanisms of bleeding[4], risk factors, and preventive strategies, these approaches should be evaluated separately when considering post-polypectomy bleeding[5-9].

BLEEDING RISK AND PREVENTION AFTER COLD RESECTION

Cold and hot resection differ in their mechanisms of tissue injury and associated bleeding. Cold resection does not require electrocautery, thereby limiting damage to the submucosa and deeper mural vessels. Bleeding is usually immediate, capillary in nature, and self-limited, rather than delayed or clinically significant[5]. In contrast, hot resection transects tissue using cautery, causing thermal injury to submucosal vessels and deeper layers; as the eschar separates and necrosis develops, delayed bleeding may occur[4,5]. CSP has increasingly become the preferred approach for diminutive and small colorectal polyps because it avoids cautery-related injury and carries a very low risk of delayed bleeding[6-9]. Whether this favorable safety profile justifies broader application to lesions measuring 10-19 mm remains unclear, and guideline recommendations differ within this size range.

CSP has a favorable overall safety profile. In a Japanese multicenter retrospective study of 18007 patients undergoing CSP for colorectal polyps ≤ 10 mm, the overall delayed bleeding rate was 0.14% (26/18007). Multivariable analysis identified antiplatelet use [odds ratio (OR) = 4.521, P = 0.001] and anticoagulant use (OR = 7.866, P = 0.003) as independent patient-related risk factors. Lesion-related risk factors included polyp size ≥ 5 mm (OR = 3.251, P = 0.005), rectal location (OR = 3.674, P = 0.007), and polypoid morphology (OR = 7.087, P = 0.002)[10]. Similarly, in a multicenter retrospective study from China including 10650 patients who underwent CSP, the delayed bleeding rate was 0.24%. Multivariable Cox analysis showed that antithrombotic use, particularly antiplatelet therapy, as well as intraprocedural hematoma, larger wound size, and pedunculated morphology (Ip), were strongly associated with delayed bleeding[11]. A prospective Korean study further suggested that cold resection may remain feasible during uninterrupted single-antiplatelet therapy in carefully selected patients, provided that intraprocedural hemostasis is meticulous[12]. Nevertheless, antithrombotic therapy should remain central to risk stratification rather than be considered uniformly safe across regimens.

Recent studies have increasingly examined the feasibility of C-EMR. Nevertheless, these data do not support the routine application of cold resection to all larger adenomatous lesions, as any reduction in bleeding risk must be weighed against the potential for residual or recurrent neoplasia and compromised oncologic adequacy. The choice of resection technique should therefore depend not only on lesion size, but also on histologic subtype, morphology, and malignant potential. For nondysplastic sessile serrated lesions, including selected larger lesions suitable for piecemeal resection, cold piecemeal endoscopic mucosal resection (EMR) may be appropriate because the therapeutic goal is complete eradication rather than en bloc histologic assessment[5]. By contrast, for adenomatous lesions, especially when optical assessment raises concern for superficial submucosal invasion, cold or piecemeal resection is generally unsuitable, because histologic interpretability and oncologic completeness should take priority.

A recent meta-analysis incorporating three randomized controlled trials and four nonrandomized comparative studies found that, in lesions ≥ 20 mm, C-EMR was linked to markedly lower risks of delayed bleeding, immediate bleeding, and perforation compared with H-EMR[13]. However, recent high-quality evidence suggests that C-EMR may carry a higher risk of residual or recurrent neoplasia, particularly in larger adenomatous lesions. Recent randomized evidence in flat adenomatous large nonpedunculated colorectal polyps measuring 15-50 mm further highlights this trade-off. In a trial of 177 lesions randomized to C-EMR or H-EMR, technical success was comparable (98.9% vs 100%; P = 0.31), but recurrence was substantially higher with C-EMR (18.4% vs 1.1%; relative risk = 16.6, 95%CI: 2.24-122; P < 0.001)[14]. Similarly, in a multicenter randomized trial of 660 patients with large (≥ 20 mm) nonpedunculated colon polyps, C-EMR was associated with fewer severe adverse events and no perforations, without a significant difference in post-procedural bleeding, but at the cost of a markedly higher recurrence rate than H-EMR[15]. Similar results were reported in a further multicenter randomized trial of 396 nonpedunculated colorectal polyps ≥ 20 mm, in which C-EMR reduced thermal injury-related complications, including perforation, but resulted in a higher residual adenoma burden than H-EMR[16]. Retrospective data are in line with these findings. Although cold piecemeal EMR appears safe across lesion subgroups, its efficacy is lower for large adenomas (≥ 20 mm), which had a 6-month recurrence rate of 16.1%, compared with 4.1% for large sessile serrated lesions and 3.0% for medium-sized adenomas[17].

BLEEDING RISK AND PREVENTION AFTER EMR

Delayed bleeding remains one of the most clinically important adverse events after EMR, particularly following H-EMR. Antithrombotic therapy is one of the most consistently reported predictors of post-EMR bleeding, while older age, hypertension, renal disease, and impaired baseline hemostatic reserve also contribute meaningfully to risk[18]. In a Korean cohort including 21562 patients and 41930 polyps, a platelet count below 90000/μL was associated with a 2.67-fold increase in immediate bleeding and a 9.66-fold increase in delayed bleeding. For lesions larger than 20 mm treated with H-EMR or ESD, the authors proposed a platelet threshold of 100000/μL[19]. Cirrhosis also appears to influence bleeding risk in a stage-dependent manner, with the excess risk largely confined to patients with decompensated disease[20].

Among lesion-related factors, size and location are the most consistent predictors of bleeding. In a study of 1962 polyps, the risk increased by approximately 4.5-fold for lesions measuring 1-2 cm and by 22.4-fold for lesions ≥ 3 cm when compared with diminutive polyps[21]. This association is further supported by studies using the size, morphology, site, and access classification, which showed that in laterally spreading tumors ≥ 20 mm, greater procedural complexity was associated with higher rates of intraprocedural bleeding and clinically significant delayed bleeding (CSDB)[22].

Procedure-related factors may further increase this risk. Growing evidence suggests that characteristics of the post-resection defect also carry prognostic significance. Visible muscle fibers, cherry-red spots, and multiple exposed submucosal vessels have each been identified as independent predictors of delayed bleeding after piecemeal H-EMR of large nonpedunculated lesions[21,23,24].

Prophylactic clipping after EMR is best considered a selective, risk-adapted strategy rather than a routine measure for all cases. The strongest evidence of benefit is seen in large right-sided lesions treated with H-EMR. In a retrospective study of 524 large colorectal lesions, complete defect closure reduced delayed bleeding from 9.7% to 1.8%[25]. Likewise, a 2022 randomized trial of right-sided nonpedunculated colorectal polyps ≥ 20 mm found that clipping reduced clinically significant bleeding from 10.6% to 3.4%[26]. Individual patient data meta-analyses have shown similar directional findings, particularly for proximal lesions and in patients receiving anticoagulants or dual antiplatelet therapy[27]. Cost-effectiveness analyses also favor selective rather than routine clipping[28,29]. The CLIPPER trial found no reduction in delayed bleeding with routine clipping after EMR of proximal lesions ≥ 20 mm[30], and a meta-analysis found no overall significant benefit, although a favorable signal persisted in high-volume tertiary centers[31]. Other adjunctive measures, including prophylactic coagulation, injection therapy, powders, and self-assembling peptide gels, remain investigational rather than standard preventive strategies[32-36].

BLEEDING RISK AND PREVENTION AFTER ESD

According to the 2024 European Society of Gastrointestinal Endoscopy guideline, ESD may be considered an alternative resection strategy for large nonpedunculated colorectal polyps measuring ≥ 20 mm, preferably in high-volume centers[5]. Delayed bleeding after colorectal ESD is a result of cumulative risk rather than a single determinant. The most consistently supported predictors are antithrombotic exposure, particularly anticoagulants, rectal or rectosigmoid location, and larger lesions or post-ESD defects[37-40]. In a multicenter study of 3142 patients, age ≥ 75 years, American Society of Anesthesiologists class III-IV, anticoagulant use, antiplatelet use, and lesion size > 50 mm were independently associated with CSDB[40]. A randomized trial likewise identified rectal location and specimen area > 870 mm² as independent risk factors[38].

Against this background, prophylactic defect closure appears beneficial in selected settings, although the benefit is unlikely to be universal. In a multicenter randomized controlled trial, prophylactic clipping significantly reduced both overall and severe delayed bleeding and remained independently protective on multivariable analysis[38]. A retrospective study reported similar findings, with lower delayed bleeding after closure[41]. In a prospective Japanese study of anticoagulated patients, complete closure was achieved in 97.6%, with an overall delayed bleeding rate of only 4.9% under minimally interrupted anticoagulation[42]. Similarly, an ABCD-J database study found lower bleeding rates after complete closure in patients receiving DOACs or warfarin, with the clearest benefit observed in right-sided lesions[37]. Meta-analytic evidence is generally supportive. A meta-analysis of four randomized controlled trials demonstrated a marked reduction in CSDB[43], and a systematic review of nine studies involving 2404 patients also reported a lower risk of delayed bleeding, particularly in studies including larger lesions or a greater proportion of rectal lesions[44]. However, these results are not uniform. A large multicenter study of 3142 patients found no significant reduction in CSDB with prophylactic complete closure, either overall or after propensity matching, and no clear benefit in anticoagulated or other high-risk subgroups[40]. Taken together, current evidence strongly supports selective rather than routine closure, particularly in patients with anticoagulant exposure, large lesions, or high-risk lesion location.

A PRACTICAL FRAMEWORK FOR CHOOSING COLD VERSUS HOT RESECTION

In routine clinical practice, the choice between cold and hot resection should begin with distinguishing pedunculated from nonpedunculated lesions[3,5]. For pedunculated polyps, the key technical issue is reliable control of stalk vessels. Accordingly, cold snare resection is generally appropriate only for very small lesions with a thin stalk. In contrast, for larger pedunculated polyps, particularly those with a head diameter ≥ 20 mm or a thick stalk, electrocautery-based snare resection remains the preferred approach. For nonpedunculated lesions, cold snare resection is the preferred technique for most lesions ≤ 10 mm[3,5]. Cold resection is also well suited to nondysplastic sessile serrated lesions, including selected larger lesions amenable to piecemeal cold EMR, because it preserves a favorable safety profile without compromising the therapeutic objective in lesions without suspected dysplasia. However, for nonpedunculated adenomatous lesions measuring 10-19 mm, the role of cold resection remains uncertain, and guideline recommendations diverge in this size range[13-17]. When complete excision and reliable histologic assessment are the primary priorities, electrocautery-based resection remains the more dependable default strategy.

For adenomatous lesions measuring ≥ 20 mm, although cold resection may reduce thermal injury and improve procedural safety, this advantage must be weighed against the greater likelihood of residual or recurrent neoplasia[13-17]. By contrast, lesions with endoscopic or optical features suggestive of superficial carcinoma, superficial submucosal invasion, or substantial fibrosis, as well as lesions requiring en bloc resection, should not be managed with a routine cold resection strategy. In these situations, oncologic adequacy becomes the overriding priority, and ESD should be considered when lesion characteristics and local expertise permit. Important uncertainties nevertheless remain. Key controversies and evidence gaps in bleeding prevention after colorectal endoscopic resection are summarized in Table 1.

Table 1 Key controversies and evidence gaps in bleeding prevention after colorectal endoscopic resection.

Clinical issue	What current evidence supports	What remains uncertain
Cold vs hot resection for 10-19 mm nonpedunculated lesions	Cold resection reduces thermal injury and delayed bleeding, but current guidelines are not fully aligned: ESGE 2024 favors hot snare polypectomy for 10-19 mm nonpedunculated adenomatous polyps, whereas USMSTF 2020 permits either cold or hot snare resection in this size range	Whether the bleeding advantage of cold resection is sufficient to outweigh concerns regarding complete excision, margin negativity, and histologic adequacy in adenomatous lesions remains unresolved
Whether cold EMR should expand to lesions ≥ 20 mm	Cold EMR appears to have a favorable safety profile in selected lesions, particularly sessile serrated lesions without suspected dysplasia. ESGE 2024 supports piecemeal CSP or cold EMR for SSLs of any size without suspected dysplasia, while reserving cold piecemeal resection of large flat adenomatous lesions for selected situations	Whether this safety advantage justifies broader use in large adenomatous lesions despite higher residual or recurrent neoplasia
Prophylactic clipping after EMR	ESGE 2024 recommends clip closure after conventional EMR of large nonpedunculated lesions in the right colon but suggests against routine clipping for lesions < 20 mm and for ≥ 20 mm lesions in the left colon	Which lesions outside this clearly high-risk subgroup derive meaningful benefit from closure remains uncertain, particularly with respect to intermediate-sized lesions, left-sided lesions, and varying antithrombotic risk
Routine vs selective closure after colorectal ESD	Recent studies and meta-analyses suggest that prophylactic closure after colorectal ESD may reduce CSDB, although the evidence remains heterogeneous	Whether routine closure is justified remains unclear, because the balance of benefit, feasibility, procedure time, cost, lesion size, location, and antithrombotic exposure varies substantially across studies and practice settings

Bleeding risk should be assessed in a procedure-specific rather than uniform manner, because the relative importance of individual risk factors differs across cold resection, endoscopic mucosal resection, and endoscopic submucosal dissection. EMR: Endoscopic mucosal resection; ESD: Endoscopic submucosal dissection; ESGE: European Society of Gastrointestinal Endoscopy; USMSTF: United States Multi-Society Task Force; CSP: Cold snare polypectomy; CSDB: Clinically significant delayed bleeding; SSLs: Sessile serrated lesions.

Prophylactic clip use should also be individualized according to the resection modality. After CSP, routine clip closure is generally unnecessary because delayed bleeding is uncommon and most bleeding is immediate and self-limited[10,11]. When additional hemostatic treatment is needed, it is usually driven by ongoing intraprocedural bleeding or specific high-risk findings rather than by routine closure of the resection defect. Routine clip closure is likewise not generally warranted after cold EMR, given the low risk of delayed bleeding, although selective use may be reasonable for large defects in patients at high bleeding risk. In contrast, the best evidence for prophylactic clipping comes from conventional hot EMR of large nonpedunculated lesions[25-27], particularly proximal or right-sided colonic lesions ≥ 20 mm when complete closure is feasible, especially in patients receiving anticoagulants or dual antiplatelet therapy[28-31]. For colorectal ESD, prophylactic defect closure may be considered in selected high-risk situations[37,38], particularly in the setting of anticoagulant exposure and/or a large post-ESD defect when complete closure is technically feasible[40-43]. However, the current evidence remains heterogeneous and does not support universal closure of all colorectal ESD defects across routine practice[40,43,44].

CONCLUSION

The risk of delayed bleeding after colorectal endoscopic resection is determined by the interplay of patient-, lesion-, and procedure-related factors (Tables 2 and 3). Current guidelines are broadly aligned for diminutive and small lesions, supporting CSP as the preferred technique for most polyps ≤ 9 mm. Recommendations diverge, however, for nonpedunculated adenomatous lesions measuring 10-19 mm. The 2020 United States Multi-Society Task Force allows either cold or hot snare resection, whereas the 2024 European Society of Gastrointestinal Endoscopy guideline favors hot snare polypectomy for this subgroup and reserves cold piecemeal approaches for selected situations[3,5]. Preventive strategies should therefore be selective rather than routine. In line with current guideline recommendations, this includes careful optical assessment to exclude submucosal invasion, thoughtful selection of cold vs hot resection, meticulous inspection of the post-resection defect, individualized management of antithrombotic therapy, and targeted prophylactic closure of large right-sided EMR defects. However, routine prophylactic clipping of all resection defects is not supported by the available evidence.

Table 2 Main bleeding risk factors and practical considerations after cold resection.

Risk domain	Main risk factors	Suggested approach
Patient-related	Antithrombotic therapy	Assess bleeding risk and anticipated long-term oncologic benefit before selecting the resection technique
Lesion-related	Pedunculated morphology, larger size, rectal location, larger wound size	Restrict cold resection to appropriate indications. Use caution in pedunculated and larger lesions; when adenomatous lesions are treated with cold resection, closer post-resection surveillance may be appropriate
Procedure-related	Intraprocedural hematoma, persistent immediate bleeding, large cold defects	Routine prophylactic hemostasis is unnecessary in most cases, but selected high-risk defects may warrant endoscopic treatment

Table 3 Main bleeding risk factors and practical considerations after hot resection.

Risk domain	Main risk factors	Suggested approach
Patient-related	Antithrombotic therapy, thrombocytopenia, renal dysfunction, cirrhosis, advanced age	Consider these factors when planning electrocautery-based resection. In higher-risk patients, limit unnecessary thermal injury and consider selective closure of large right-sided H-EMR defects or selected ESD defects
Lesion-related	Large lesion size, right-sided colonic location, rectal/rectosigmoid location, pedunculated polyps with a large head or thick stalk	Use diluted epinephrine injection and/or mechanical prophylaxis for pedunculated polyps with a large head or thick stalk. Consider selective closure after H-EMR of large right-sided lesions and after selected large rectal or rectosigmoid ESD defects
Procedure-related	Thermal injury, piecemeal EMR, large mucosal defects, extensive submucosal dissection	Carefully inspect the resection defect, treat exposed vessels when present, and reserve prophylactic closure for selected high-risk defects

H-EMR: Hot endoscopic mucosal resection; ESD: Endoscopic submucosal dissection; EMR: Endoscopic mucosal resection.

Future studies should move beyond broad size-based classifications and focus on clinically relevant subgroups in whom the choice between cold and hot resection remains genuinely uncertain. In our view, the highest priority is multicenter randomized trials in patients with nonpedunculated adenomatous lesions, stratified by lesion location, including right-sided colonic and rectal lesions, antithrombotic exposure, and long-term oncologic risk, with complete excision, recurrence at surveillance, histologic interpretability, and the need for rescue therapy incorporated as co-primary or hierarchical endpoints. Prophylactic clip closure also merits further investigation in prespecified high-risk bleeding subgroups, particularly patients with large lesions, antithrombotic exposure, and right-sided colonic defects, while clarifying whether additional populations beyond those currently supported by guideline-level evidence may benefit. At the same time, further research is needed to develop novel hemostatic materials that are effective, easy to apply, cost-conscious, and scalable for routine clinical use, especially for large post-resection defects and other high-risk bleeding settings.