Endoscopic management of upper non-variceal and lower gastrointestinal bleeding: Where do we stand?

Abstract

Non-variceal upper gastrointestinal bleeding (GIB) remains a significant clinical challenge with a 30-day mortality of up to 11%. Peptic ulcers are the most common cause, followed by other conditions like Mallory-Weiss syndrome, Dieulafoy’s lesions, and gastric neoplasms. Treatment strategies include acid-suppressive therapy, endoscopic interventions, and surgical or radiological procedures. Endoscopic techniques such as over-the-scope clips, coagulation graspers, and endoscopic ultrasound-guided treatments have significantly improved outcomes, reducing rebleeding rates and the need for surgery. Injectable therapies, mechanical hemostasis via clips, and thermal modalities (e.g., electrocoagulation, argon plasma coagulation) remain standard approaches for active bleeding. Newer hemostatic powders, such as TC-325, offer promising non-contact treatments, particularly in cases of refractory bleeding or malignancy. Doppler endoscopic probes aid in risk stratification by detecting residual arterial blood flow, improving the efficacy of endoscopic therapy and reducing rebleeding risks. For small bowel bleeding, endoscopic management with enteroscopy and thermal therapies remains key, though medical therapies are evolving. Lower GIB, which often involves conditions like diverticular disease and angioectasia, requires a comprehensive approach combining endoscopic, radiologic, and surgical interventions. Pharmacologic management focuses on balancing antithrombotic therapy with bleeding risks, with reversal agents playing a crucial role in life-threatening bleeding episodes. This review highlights advances in diagnostic tools and endoscopic therapies that have enhanced management outcomes for GIB across various etiologies.

Key Words: Upper non-variceal gastrointestinal bleeding; Lower gastrointestinal bleeding; Complications; Endoscopic hemostasis; Chemical hemostasis; Mechanical hemostasis

Core Tip: Gastrointestinal bleeding (GIB) is a serious condition requiring prompt diagnosis and intervention. Advances in endoscopic therapy, such as clips, thermocoagulation, and hemostatic powders, have improved outcomes by reducing rebleeding, surgeries, and mortality. Non-variceal upper GIB, often caused by peptic ulcers, has a 30-day mortality of 11%, while lower GIB, with diverse causes like diverticular bleeding and ischemic colitis, usually resolves spontaneously but can recur. Accurate diagnosis, early treatment, and careful management are essential to improve outcomes and reduce complications.

INTRODUCTION

We recently read with interest the article by Alhumayyd et al[1] published in the World Journal of Gastroenterology. In this article, the authors reported the incidence of acute overt gastrointestinal bleeding (GIB) in patients with hematological malignancies. The article offers a detailed and in-depth approach to this condition. More specifically, this retrospective single-center study assessed the role of endoscopic intervention in inpatients with hematologic malignancies and severe thrombocytopenia (< 50 × 10³/mL) who developed overt GIB. The study suggests that mortality and clinical outcomes in this population are primarily driven by the severity of the underlying hematologic malignancy rather than the use of endoscopic intervention. The risks of endoscopy and anesthesia, particularly in thrombocytopenic patients, should be carefully weighed against potential benefits. Although urgent endoscopy (< 24 hours) was associated with lower rates of recurrent bleeding, it did not impact mortality or transfusion requirements. Additionally, patients in the non-endoscopy group had higher hemoglobin levels and lower platelet counts, suggesting that perceived procedural risks influenced the decision to withhold endoscopy. The findings challenge current guidelines recommending urgent endoscopy in GIB patients with malignancy and highlight the need for a more individualized approach. Future research should explore alternative platelet thresholds, point-of-care platelet function testing (e.g., thromboelastography), and a prospective multi-center study to validate these findings and refine management strategies for GIB in patients with hematologic malignancies and severe thrombocytopenia. By reviewing the available literature, we provide a brief overview of the pathophysiology, prevalence as well as available methods of GIB, in an effort to provide clinicians with the necessary knowledge to improve their daily practice. Table 1 summarizes the endoscopic management of GIB.

Table 1 Endoscopic management of gastrointestinal bleeding.

Category	UGIB	LGIB
Risk stratification	Use GBS for pre-endoscopy assessment. GBS ≤ 1: Safe outpatient management. GBS ≥ 7: High risk, requires intervention	Use risk scores like Oakland Score to determine the need for hospitalization and intervention
Timing of endoscopy	Early endoscopy (≤ 24 hours) recommended after resuscitation. Urgent endoscopy (≤ 12 hours) only in hemodynamically unstable patients	Colonoscopy within 24 hours for hemodynamically stable patients. Urgent colonoscopy in massive bleeding with ongoing instability
Initial hemostasis	Combination therapy for active bleeding ulcers (e.g., epinephrine + thermal/mechanical therapy). Nonbleeding visible vessels (FIIa) treated with thermal, mechanical, or sclerosing agents	Endoscopic therapy for visible bleeding sources (e.g., clips, coagulation, injection therapy)
Persistent or recurrent bleeding	Second endoscopic hemostasis attempt if initial therapy fails. Consider transcatheter arterial embolization or surgery if endoscopic treatment fails	Repeat endoscopic therapy for ongoing bleeding. If unsuccessful, angiography with embolization or surgery may be needed
Pharmacologic management	High-dose PPIs post-endoscopy (e.g., continuous infusion or alternative regimens)	No routine use of PPIs; focus on treating underlying cause (e.g., anticoagulant reversal if applicable)
Anticoagulation management	Resume anticoagulation within 7 days of controlled bleeding, balancing thrombotic risk	Resume anticoagulation based on bleeding severity and thrombotic risk assessment
Special considerations	Consider hemostatic sprays or cap-mounted clips for challenging cases	For diverticular bleeding or angiodysplasia, thermal therapy or endoscopic clips may be used

UGIB: Upper gastrointestinal bleeding; LGIB: Lower gastrointestinal bleeding; GBS: Glasgow-Blatchford Score; PPI: Proton pump inhibitor.

MANAGEMENT OF NON-VARICEAL UPPER GIB

Non-variceal upper GIB-epidemiology, pathophysiology and causes

Non-variceal upper GIB (UGIB) refers to bleeding from non-variceal diseases of the gastrointestinal tract located above the ligament of Treitz, including the esophagus, stomach, duodenum, pancreatic and bile ducts, and post-gastrojejunostomy anastomotic sites. While its annual incidence decreased from 78 to 61 per 100000 persons between 2001 and 2009, 30-day mortality remains as high as 11%. Peptic ulcers are the most common cause, accounting for 20%-67% of cases, with other causes including erosive esophagitis, Dieulafoy’s lesion, Mallory-Weiss syndrome, gastric antral vascular ectasia (GAVE), and neoplasms. Established treatments for non-variceal UGIB (NVUGIB) include acid-suppressive drugs, endoscopic therapy (ET), and radiological or surgical interventions. The widespread adoption of ET has significantly reduced rebleeding rates, the need for surgery, and mortality. Additionally, novel endoscopic techniques, such as over-the-scope clips (OTSC), endoscopic suturing, coagulation graspers, and endoscopic ultrasound (EUS)-guided treatments, have further improved hemostasis outcomes[2].

GUIDELINES AND STRATIFICATION TOOLS

The European Society of Gastrointestinal Endoscopy (ESGE) recommends the Glasgow-Blatchford Score (GBS) for pre-endoscopy risk stratification in acute upper gastrointestinal hemorrhage (UGIH), with very low-risk patients (GBS ≤ 1) safely managed as outpatients. In patients on low-dose aspirin for cardiovascular prophylaxis, therapy should not be interrupted or restarted within 3-5 days if stopped. After hemodynamic resuscitation, early endoscopy (≤ 24 hours) is advised, while urgent endoscopy (≤ 12 hours) is not beneficial. Actively bleeding ulcers (FIa, FIb) require combination therapy, while nonbleeding visible vessels (FIIa) can be treated with thermal, mechanical, or sclerosing agents, alone or with epinephrine. Persistent bleeding may necessitate topical hemostatic spray or cap-mounted clips, and recurrent bleeding may require a second endoscopic attempt or, if unsuccessful, transcatheter embolization or surgery. High-dose proton pump inhibitors (PPIs) are recommended post-endoscopy through continuous infusion or alternative regimens. Anticoagulation should resume within 7 days of controlled bleeding, considering the rapid onset of direct oral anticoagulants (DOACs)[3]. Risk stratification scores, such as GBS, AIMS65, Rockall, and PNED, are essential for managing UGIB in emergency settings. GBS, AIMS65, and admission Rockall are pre-endoscopic scores, whereas PNED and full Rockall include endoscopic findings. GBS predicts intervention needs, AIMS65 predicts mortality and hospital stay, and the age, blood tests, and comorbidities (ABC) score, incorporating age, urea, albumin, creatinine, comorbidities, and ASA classification, categorizes risk as low (≤ 3), moderate (4-7), or high (≥ 8). Studies indicate that GBS ≥ 7 is the most sensitive and specific for predicting ET needs, while GBS ≤ 1 effectively identifies low-risk patients for outpatient management, misclassifying < 1% of high-risk cases. AIMS65 is superior for mortality and intensive care unit (ICU) admission prediction, whereas GBS outperforms Rockall for identifying patients at high risk for intervention, allowing for optimized clinical decisions and resource utilization[4,5].

ENDOSCOPIC APPROACH

Injection therapy

Injectable therapies for hemostasis include dilute epinephrine, sclerosing agents (e.g., ethanol, ethanolamine), and tissue adhesives (e.g., thrombin, cyanoacrylate). Dilute epinephrine (1:10000 or 1:20000) is the most used, primarily providing tissue tamponade and vasoconstriction. It is injected in 0.5- to 2-mL aliquots around the ulcer base. However, epinephrine monotherapy is less effective than other modalities, such as bipolar electrocoagulation or clips, and is associated with a higher risk of rebleeding (RR: 2.2, 95%CI: 1.04-4.64). Combining epinephrine injection with another ET significantly reduces rebleeding risk (RR: 0.34, 95%CI: 0.23-0.50). Its ease of use and ability to temporarily slow active bleeding make it particularly valuable for providing a clear field for definitive endoscopic interventions like clipping or electrocoagulation[6].

Mechanical endoscopic hemostasis

The ESGE recommends endoscopic clipping for non-bleeding ulcers and epinephrine injection with mechanical or thermal methods for active bleeding. Two-pronged clips achieve an 85%-100% success rate, though chronic ulcers and challenging factors like large size, difficult locations, or comorbidities may reduce efficacy. For refractory bleeding ulcers, OTSCs show high success (85%-100%) with minimal recurrence, but they are typically reserved for cases where conventional methods fail. For Mallory-Weiss syndrome, endoscopic clipping, band ligation, or thermocoagulation is effective for active bleeding, with superior rebleeding prevention compared to injection therapy. Dieulafoy lesions, often in the stomach or duodenum, are best managed with mechanical hemostasis (clipping or band ligation), which reduces rebleeding and aids future interventions. Preventive clipping is used to reduce delayed bleeding after endoscopic mucosal resection, with guidelines supporting its role despite cost-effectiveness debates.

Through-the-scope (TTS) clips are essential tools in managing GIB, functioning by compressing vessels through tissue grasping. These clips vary in material (e.g., stainless steel, titanium, nitinol), arm length, closure angle, opening width, and prong design. Modern TTS clips are preloaded, rotatable, and allow multiple openings and closings before deployment, with working lengths of 225-250 cm and magnetic resonance imaging compatibility up to 3.0 Tesla. Multi-firing clip devices enable sequential application of 3-4 clips without device exchange, further enhancing efficiency. Comparative studies on popular models, including Instinct, Resolution, DuraClip, SureClip, and QuickClip Pro, revealed distinct advantages.

The OTSC, developed by Ovesco (Tübingen, Germany), is a nitinol-based device mounted on the distal end of an endoscope, similar to rubber-band ligation systems. OTSCs address limitations of TTS clips, such as small arm size, by offering larger and stronger mechanical compression capable of managing large vessels and lesions effectively. These clips are highly effective in achieving hemostasis, particularly for iatrogenic perforations, massive bleeding, and post-resection repairs. However, a notable drawback is the need to remove the endoscope for reloading. The nitinol structure provides a permanent closing force for sufficient tissue compression, as demonstrated in animal and phantom models. OTSCs are available in two configurations: Nontraumatic clips with blunt teeth and traumatic clips with sharp teeth. Another variant, the Padlock, features a nitinol hexagonal ring with six inner needles, providing 360° radial tissue compression. This system ensures precise tissue control to maintain blood flow, promoting lesion healing while minimizing tissue damage[7].

Hemostatic powders

Topical hemostatic agents represent a significant advancement in managing UGIB, offering non-contact, localized treatment with broad application to diffusely bleeding lesions, such as malignancies. The most extensively studied agent, TC-325, a silica-based inert powder, achieves > 90% immediate hemostasis as primary or rescue therapy. However, its temporary effect (lasting 12-24 hours) raises concerns about recurrent bleeding. Randomized studies, including a large trial by Lau[8] with 224 patients, demonstrated that TC-325 is non-inferior to conventional ET for controlling bleeding (90.1% vs 81.4%), with comparable 30-day rebleeding (8.1% vs 8.8%) and mortality rates (12.6% vs 12.4%). Challenges include catheter blockage during application, mitigated by adhering to manufacturer guidelines.

Other agents like EndoClot Polysaccharide Hemostatic System and PuraStat also show promise. EndoClot, a polymer powder, achieves 64%-100% immediate hemostasis, with rebleeding rates of 3.3%-20%. PuraStat, a synthetic peptide forming a transparent hydrogel, allows multi-modality therapy and promotes ulcer healing, achieving hemostasis in > 80% of cases with a 17% rebleeding rate. Emerging agents such as Nexpowder and Ankaferd Blood Stopper have limited data but similar mechanisms.

Current ESGE and ACG guidelines recommend topical agents for refractory bleeding and conditionally for actively bleeding ulcers. Future guidelines may position them as first-line treatments, particularly for malignant bleeding, supported by growing evidence of their efficacy compared to traditional methods[6].

THERMAL THERAPIES

Thermal therapies achieve hemostasis by coagulating tissue and contracting blood vessels through the application of heat. Common contact thermal devices include bipolar electrocoagulation, heater probes, and monopolar soft coagulation. Bipolar electrocoagulation probes are effective for coaptive coagulation, requiring firm pressure and heat application for 8-10 seconds with a 3.2-mm probe. Meta-analysis of 15 randomized controlled trials (RCTs) demonstrated that thermal contact devices significantly reduce rebleeding (RR: 0.44, 95%CI: 0.36-0.54) and mortality (RR: 0.58, 95%CI: 0.34-0.98) compared to no ET.

Monopolar hemostatic forceps soft coagulation

Monopolar hemostatic forceps soft coagulation, initially developed for managing bleeding during endoscopic resection, is also effective for other bleeding lesions by applying heat directly to the bleeding site or grasping the vessel.

Argon plasma coagulation

Argon plasma coagulation (APC), a non-contact modality producing superficial coagulation (1-2 mm), is commonly used for vascular lesions like angiodysplasia and GAVE. Evidence for APC in peptic ulcer bleeding is less robust. While some RCTs showed it reduces rebleeding compared to water injection and is comparable to heater probes or TTS clips, the ACG provides only a conditional recommendation for its use in peptic ulcer bleeding[6].

Cryotherapy

Cryotherapy, which induces tissue necrosis through localized freezing, has been proposed for GAVE refractory to APC. A pilot study reported complete resolution in 50% of cases and partial response in the remainder after three endoscopic sessions. However, its limited availability and the need for further studies restrict its widespread use[4].

Radiofrequency ablation

Radiofrequency ablation, widely used for Barrett's esophagus, is a promising alternative to APC for GAVE. A systematic review of 72 patients showed a 74% clinical response rate with minimal adverse events (4.2%), suggesting it as a safe and effective option[4].

Coagrasper

The Coagrasper, a thermal-mechanical hemostatic forceps, offers effective hemorrhage control with reduced perforation risk due to its low-voltage operation. It has demonstrated superior efficacy compared to endoclips and heater probes in randomized trials for peptic ulcer bleeding, making it a valuable tool for endoscopic hemostasis[4].

DOPPLER ENDOSCOPIC PROBE

The Doppler endoscopic probe (DEP), introduced in 1982 and United States Food and Drug Administration-approved, assists in risk stratification and guides ET by detecting arterial blood flow beneath lesions. DEP is particularly useful for stigmata of recent hemorrhage, such as non-bleeding visible vessels, oozing bleeding, and flat spots, which are often challenging to assess visually. It also confirms the eradication of arterial blood flow post-treatment, reducing the risk of rebleeding. DEP is easy to use, safe, and cost-effective (150-200 USD per probe), though its adoption has been limited by the need for training and variability in signal interpretation.

Clinical evidence supports DEP's role in reducing rebleeding and improving outcomes in non-variceal UGIH (NVUGIH). A pivotal randomized trial reported significantly lower 30-day rebleeding rates (11.1% vs 26.3%, P = 0.021) and confirmed that residual arterial blood flow post-treatment strongly predicts rebleeding (89% vs 0%, P < 0.001). Meta-analyses have shown DEP reduces rebleeding (OR: 0.13, 95%CI: 0.04-0.44; P = 0.001), need for surgery (OR: 0.05, 95%CI: 0.01-0.24; P < 0.001), and bleeding-related mortality (OR: 0.19, 95%CI: 0.05-0.81; P = 0.025)[9].

Α total of 62 patients were included in the study conducted by Nielsen et al[9] which the results included the following: At second-look endoscopy, 91% (29/32) of patients had a positive DEP signal at the ulcer base and were treated with contact thermal therapy (n = 29), injection of diluted adrenaline (n = 23), and haemoclips (n = 7). Among the 32 patients evaluated with DEP, only one re-bled (3%) compared to four patients (13%) in the control group (P = 0.20), with no differences in secondary outcomes or complications related to DEP evaluation[9]. Moreover, a systematic review of original articles published until December 2019 assessed the use of DOP-US in peptic ulcer bleeding. Eight studies showed that DOP-US signal-guided therapy reduced rebleeding, mortality, and surgical intervention compared to standard visual evaluation, with a significantly higher risk of rebleeding [48.5% (95%CI: 29.5-67.9%)] if the signal persisted despite ET[11]. Another study, by Shiratori et al[12], evaluated Doppler monitoring systems in gastric endoscopic submucosal dissection. A comparative study of DOP patients (50) vs no-DOP patients (151) found lower rates of delayed bleeding (2.0% vs 8.6%, P = 0.11), readmission due to bleeding (0% vs 2.7%), and blood transfusion (2.0% vs 3.3%) in the DOP group. In the total study population (n = 245), the log-rank test indicated a lower incidence of delayed bleeding (P = 0.036), with the Cox regression model showing a marginally significant effect (HR: = 0.17, 95%CI: 0.022-1.26, P = 0.082). No procedure-related adverse events were observed[12]. Despite its strong evidence base and excellent safety profile, DEP's integration into routine practice has been slow. Broader adoption could improve the management of high-risk lesions in NVUGIH, offering precise risk stratification and superior clinical outcomes when used alongside standard endoscopic therapies[3].

Small bowel bleeding: Causes, diagnosis and treatment

Small bowel bleeding is an emerging clinical entity influenced by various pathophysiological mechanisms, with the microbiota playing a critical role. Experimental studies demonstrate that antibiotics can prevent non-steroidal anti-inflammatory drug (NSAID)-induced small bowel damage, and rifaximin, a microbiota modulator, reduces diclofenac-induced erosions and ulcers in healthy volunteers. Probiotics, such as Bifidobacterium breve Bif195, have also shown protective effects against ASA-induced injury in randomized trials. Conversely, PPIs, by altering gastric pH, may exacerbate NSAID-associated injury, as seen in a study where PPIs increased erosions in patients on COX-2 inhibitors. Other factors contributing to small bowel bleeding include antiplatelet and anticoagulant use, angiodysplasias, and conditions like Heyde syndrome.

Common causes of small bowel bleeding include angiodysplasias, inflammatory bowel disease, drug-induced ulcers, and tumors. Capsule endoscopy is a pivotal diagnostic tool, with randomized trials showing higher detection rates (64.3% vs 31.1%, P < 0.01) compared to standard endoscopy. Optimal preparation for capsule endoscopy includes a low-fiber diet or polyethylene glycol ingestion for better mucosal visualization. Fecal immunochemical tests may correlate with small bowel lesions but lack strong evidence to guide capsule endoscopy use, as per ESGE guidelines[3].

Endoscopic management of small bowel bleeding often requires enteroscopy, with balloon-assisted or spiral systems aiding lesion access. Thermal therapies, like laser-argon treatment, are effective but carry high rebleeding rates (34%-45%). Medical therapies, including misoprostol for low-dose aspirin injury and octreotide for angiodysplasias, show promise, though robust evidence is limited[4].

MANAGEMENT OF LOWER GIB

LGIB-epidemiology, pathophysiology and causes

Lower GIB (LGIB) accounts for 20%-30% of major GIB cases, with an annual incidence of 0.03% that increases sharply with age. The mean presentation age is 63-77 years, and 35.7 per 100000 adults in the United States are hospitalized annually, with most cases involving self-limited bleeding and uncomplicated hospitalizations. Compared to upper GIB, LGIB patients have higher hemoglobin levels, less shock, and reduced transfusion needs. Mortality ranges from 2%-4%, mainly due to comorbidities, and recent data show declining incidence and case fatality rates. Historically, LGIB has been defined as bleeding from a source distal to the ligament of Treitz; however, small-bowel sources are now categorized as midgut bleeding, with LGIB redefined as bleeding distal to the ileocecal valve. Acute LGIB refers to bleeding of recent onset (< 3 days), potentially causing hemodynamic instability, anemia, or the need for transfusion. Chronic LGIB involves prolonged or intermittent bleeding, presenting as occult fecal blood, intermittent melena, maroon stools, or scant bright red blood per rectum.

LGIB has a wide range of etiologies, both common and rare. Common causes include diverticular bleeding, ischemic colitis, angioectasia, hemorrhoids, colorectal neoplasia, and postpolypectomy bleeding, while other significant contributors are inflammatory bowel disease and infectious colitis. Less common causes include NSAID colopathy, radiation proctopathy, stercoral ulcers, rectal varices, and Dieulafoy lesions. Rectal ulcers are observed in 8% of patients with severe hematochezia and 32% in those who develop LGIB after ICU admissions. These patients often have comorbidities like end-stage renal disease, respiratory failure, cirrhosis, or malignancy. Endoscopic findings vary from clean-based ulcers to active bleeding, and early rebleeding occurs in 44%-48% of cases, with mortality ranging from 33% to 48% in patients with high-risk features. LGIB is seen in 4%-13% of patients with radiation proctopathy and in 1.2%-6% of Crohn’s disease patients, more often in those with colon involvement. Spontaneous resolution occurs in up to 50% of cases, though recurrence is common. In HIV patients, LGIB is linked to AIDS-related thrombocytopenia, with an inpatient mortality rate of 28%, often caused by opportunistic infections. An upper gastrointestinal source may be identified in 11%-15% of suspected LGIB cases, while small-bowel sources account for 2%-15%. Accurate identification of the underlying cause is essential for effective management and improving patient outcomes[11].

Guidelines

The ESGE recommends a comprehensive initial assessment for acute LGIB, including history, physical examination, laboratory markers, and risk scores to assist but not replace clinical judgment. Patients with self-limited bleeding and an Oakland score ≤ 8 may be discharged for outpatient follow-up. In hemodynamically stable patients, a restrictive red blood cell transfusion strategy is advised for those without cardiovascular disease (threshold ≤ 7 g/dL, target 7-9 g/dL), while a more liberal strategy is recommended for those with cardiovascular disease (threshold ≤ 8 g/dL, target ≥ 10 g/dL). Colonoscopy should be performed during hospitalization, though early colonoscopy does not improve outcomes. For hemodynamic instability with suspected ongoing bleeding, computed tomography (CT) angiography is preferred to localize the bleeding site. Anticoagulants should be managed cautiously, with vitamin K antagonists (VKA) withheld and coagulopathy corrected using prothrombin complex concentrators (PCC) or fresh frozen plasma; DOACs should also be temporarily paused. Aspirin for secondary cardiovascular prevention should generally not be withheld and resumed within 5 days if interrupted, while dual antiplatelet therapy requires cardiology consultation, continuing aspirin and temporarily withholding the P2Y12 receptor antagonist if necessary[12].

Initial assessment

Initial management of moderate to severe LGIB focuses on hemodynamic stability and blood resuscitation. Blood transfusion is recommended when hemoglobin drops below 7 g/dL, unless the patient has heart or cerebrovascular conditions, in which case the target is above 8 g/dL. Colonoscopy remains the preferred diagnostic tool with a diagnostic yield ranging from 42% to 90%, but its timing remains debated. Some studies suggest ET within 12 hours has a 29% success rate, while no benefit is seen after 48 hours. In cases of rectal bleeding and hemodynamic instability, a CT angiography (CTA) is preferred for its high sensitivity and specificity in identifying bleeding sources[13].

Prognostic scoring for LGIB is still evolving, with early predictors such as tachycardia, low blood pressure, and rectal bleeding helping to identify high-risk patients. The Oakland score, developed in 2017, helps assess the risk for adverse outcomes and the safety of discharge for low-risk patients. This score has been validated for predicting death, rebleeding, transfusion needs, and therapeutic interventions. Endoscopic treatment for LGIB includes injection, mechanical, thermal, and topical therapies. For diverticular bleeding, mechanical and injection therapy are preferred, with thermal therapy used cautiously to avoid perforation, particularly in the right colon[13].

When endoscopic methods fail or in cases of hemodynamic instability, angiographic embolization or surgery may be necessary. For vascular angiectasia, commonly found in the right colon, laser argon coagulation is the typical treatment. Additionally, Hemospray can be used for diffuse bleeding sources like neoplasms or inflammatory bowel disease, though its use in the lower GI tract remains unapproved in some countries[4].

Stratification tools

Acute LGIB is a frequent cause of hospital admission, often requiring blood transfusions and interventions like colonoscopy, radiological procedures, and surgery. Effective management depends on factors such as comorbidities, medication use, symptoms, vital signs, and lab data for risk stratification. Early colonoscopy can improve the identification of bleeding sources and increase the rate of endoscopic interventions compared to elective procedures. Contrast-enhanced CT before colonoscopy may help detect recent hemorrhage, especially in patients with recent hematochezia. Careful consideration is needed regarding the use of NSAIDs and antithrombotic agents after hemostasis to prevent rebleeding and thromboembolic events. Discontinuing aspirin and NSAIDs after LGIB is generally recommended. Further studies are needed to identify patients who need early colonoscopy and hemostatic intervention[14]. An international cohort study developed a risk score for predicting mortality in UGIB and LGIB patients. The ABC score was more strongly associated with mortality [areas under the receiver-operating-characteristics curves (AUROCs): 0.81-0.84] than existing scores (AUROCs: 0.65-0.75). In UGIB, the 30-day mortality rates for low (≤ 3), medium (4-7), and high (≥ 8) ABC scores were 1.0%, 7.0%, and 25%, respectively. In LGIB, the in-hospital mortality rates for low, medium, and high ABC scores were 0.6%, 6.3%, and 18%, respectively[15].

Management

Colonoscopy identifies the source of lower gastrointestinal bleeding in over 75% of patients and allows therapeutic intervention. Early colonoscopy, typically within 24 hours of admission, is recommended after adequate bowel preparation. Treatments for diverticular bleeding include adrenaline injections (1:10000 in 1-2 mL aliquots) at the site of bleeding. If a patient cannot tolerate colonoscopy, CTA is a fast and widely available alternative, detecting bleeding rates of 0.3-0.5 mL/min with 85% sensitivity. Catheter angiography is reserved for hemodynamically unstable patients or those with recurrent bleeding, while radionuclide imaging can detect intermittent bleeding using Technetium-99m with repeated scans over 24 hours. Surgery, such as emergency segmental resection or subtotal colectomy, is necessary when endoscopic and radiologic interventions fail.

Endoscopic treatments for diverticular bleeding include thermal methods (e.g., heater probes, bipolar coagulation), epinephrine injections, and mechanical options like clip placement. Clips effectively stop bleeding and close diverticular orifices, while band ligation is less reliable due to high rebleeding rates. Thermal coagulation is used cautiously in the thin-walled right colon to avoid perforation (reported in up to 2.5% of cases). Endoscopic angioectasia treatment often involves APC, which has a high success rate with minimal complications, improving hemoglobin levels and reducing transfusion needs.

CTA is recommended before angiography or surgery for hemodynamically unstable patients or those unable to undergo bowel preparation. It detects vascular lesions and bleeding sources in 35%-76% of cases when used before colonoscopy, significantly aiding diagnosis. Angiography achieves hemostasis in 40%-100% of cases, with rebleeding rates of 15%. However, risks include bowel ischemia (1%-4%) and contrast-induced nephropathy. Surgery is reserved for patients with ongoing bleeding requiring more than six blood units within 24 hours or when other interventions fail. Localizing the bleeding site is critical to reduce rebleeding and surgical complications, as total colectomies carry higher mortality (20%-40%) than limited resections (7%-22%).

Emerging therapies include high-dose barium impaction, which has shown promise in preventing rebleeding from diverticular sources. Novel methods using barium and balloon-assisted devices for right-sided diverticular bleeding are under investigation, demonstrating potential advantages for controlling severe or recurrent bleeding[16,17].

Pharmacologic management

Management of medications in LGIB requires balancing bleeding risks with thromboembolic events. For patients on antithrombotic agents experiencing life-threatening bleeding, cessation is recommended, with resumption once hemostasis is achieved. A multidisciplinary approach is critical, especially for managing dual antiplatelet or anticoagulant therapy. Non-aspirin NSAIDs significantly increase LGIB risk and recurrence (HR: 2.0, 95%CI: 1.2-3.3) and should be discontinued, particularly in diverticular bleeding. Switching to COX-2 inhibitors does not effectively reduce recurrence risk. Antiplatelet agents triple LGIB risk compared to UGIB. Aspirin should continue for secondary prevention but not for primary. Dual antiplatelet therapy increases myocardial infarction and death risk within 30 days of stenting or 90 days of acute coronary syndrome, though temporary discontinuation of the second agent for up to 7 days may be acceptable. An international normalized ratio (INR) < 2.5 is safe for endoscopy. Reversal agents like 4F-PCC or idarucizumab may help in severe bleeding, but their efficacy in LGIB is unclear. Heparin bridging is ineffective and increases bleeding risk. Resuming anticoagulation lowers thrombotic events (HR: 0.68) and mortality (HR: 0.76) without significantly increasing rebleeding risk, though resumption within 7 days raises rebleeding (HR: 3.27) while reducing thromboembolism (HR: 0.76) and mortality (HR: 0.56). Adjustments like switching to apixaban or reducing dabigatran doses may help, but better bleeding risk models are needed[15].

For life-threatening gastrointestinal bleeding, guidelines recommend reversal agents. In VKA patients with INR > 2.5, 4F-PCC is preferred over fresh frozen plasma. For DOAC-related bleeding, targeted reversal agents are first-line, though 4F-PCC is an alternative. Anticoagulation should resume after bleeding stops to prevent thromboembolism[16]. A GRADE review advises PCC over fresh frozen plasma for warfarin users with GI bleeding and against vitamin K. DOAC users should avoid PCC, idarucizumab (dabigatran), and andexanet alfa (rivaroxaban/apixaban). Antiplatelet users should not receive platelet transfusions, and cardiac ASA for secondary prevention should continue or resume after hemostasis. In elective endoscopy, warfarin should usually continue but may be paused for high-risk procedures, with bridging only for mechanical heart valves. DOACs should be temporarily stopped. For dual antiplatelet therapy, the P2Y12 inhibitor may be paused, but ASA monotherapy should continue. Insufficient evidence prevents recommendations on PCC use for warfarin-related gastrointestinal bleeding, P2Y12 inhibitor interruption in monotherapy, or optimal anticoagulant resumption timing[17].

DOAC interruption and kidney function

DOAC interruption in patients with impaired kidney function varies based on the type of drug and creatinine clearance (mL/minute). For dabigatran, discontinuation before a procedure is recommended at 2-3 days for creatinine clearance > 80 mL/minute, 2-3 days for 50-80 mL/minute, 3-4 days for 30-49 mL/minute, and 4-6 days for < 29 mL/minute. For apixaban, patients with creatinine clearance > 60 mL/minute should stop the drug 1-2 days before the procedure, while those with 30-59 mL/minute should discontinue 3 days prior, and those with 15-29 mL/minute should stop 4 days before. Rivaroxaban should be discontinued > 1 day before a procedure for creatinine clearance > 90 mL/minute, 2 days for 60-90 mL/minute, 3 days for 30-59 mL/minute, and 4 days for 15-29 mL/minute. Proper timing of DOAC interruption helps reduce bleeding risks while balancing thromboembolic protection[18].

Thromboembolism risk in anticoagulation therapy

Thromboembolism risk in anticoagulated patients varies by condition and classification systems (BSG-ESGE, ASGE, APAGE-APSDE). Low-risk patients include those with atrial fibrillation (AF) without valvular disease, venous thromboembolism (VTE) > 3 months prior, or a bileaflet aortic valve prosthesis without AF. Medium-risk includes VTE within 3-12 months, recurrent VTE, non-severe thrombophilia, active cancer, or a bileaflet aortic valve prosthesis with risk factors [AF, cerebrovascular accident (CVA), transient ischemic attack (TIA), hypertension, diabetes, chronic heart failure, age > 75]. High-risk patients have AF with mitral stenosis, VTE < 3 months, severe thrombophilia, a prosthetic mitral valve, or recent CVA/TIA (< 6 months). Very high-risk includes ACS or PCI within 6 weeks and prosthetic valves with AF or metallic mitral valves. Proper risk assessment is essential for anticoagulation management, especially periprocedurally[18].

CONCLUSION

Endoscopic management remains the cornerstone in the diagnosis and treatment of both upper non-variceal GIB and LGIB. Advances in endoscopic technology, including novel hemostatic devices, Doppler-assisted risk stratification, and emerging topical agents, have significantly improved bleeding control, reduced rebleeding rates, and minimized the need for surgical intervention. While traditional techniques such as injection therapy, thermal coagulation, and mechanical clipping continue to be effective, newer modalities like OTSC, hemostatic powders, and EUS-guided therapies offer additional options, particularly in refractory or high-risk cases. For LGIB, especially from diverticular sources or angioectasias, a tailored approach combining endoscopic, radiologic, and pharmacologic strategies is often required. Management of antithrombotic therapy and optimization of comorbid conditions remain critical to improving outcomes. Ongoing innovations—including artificial intelligence for enhanced lesion detection and risk prediction, as well as bioengineered hemostatic materials—promise to further refine endoscopic therapy. Future research should focus on standardizing treatment algorithms, developing predictive tools for rebleeding risk, and integrating personalized medicine approaches to optimize patient care in GIB.