Tranexamic acid in acute gastrointestinal bleeding – A cautionary tale

Abstract

Gastrointestinal bleeding remains a major global cause of morbidity and mortality. The underlying etiology is heterogenous and generally grouped by site, to upper or lower gastrointestinal tract. Acute, upper gastrointestinal bleeding has a reported annual incidence of 84 to 172 per 100 000. In the United Kingdom alone, this leads to in excess of 50 000 hospital admissions, and an estimated 5000 deaths each year.1.National Clinical Guideline CentreNICE Acute Upper Gastrointestinal Bleeding: Management. National Clinical Guideline Centre at Royal College of Physicians, London, UK2012https://www.nice.org.uk/guidance/cg141/evidence/full‐guideline‐pdf‐186534541Google Scholar, 2.Hearnshaw S.A. Logan R.F. Lowe D. Travis S.P. Murphy M.F. Palmer K.R. Acute upper gastrointestinal bleeding in the UK: patient characteristics, diagnoses and outcomes in the 2007 UK audit.Gut. 2011; 60: 1327-1335Crossref PubMed Scopus (417) Google Scholar Peptic ulcer disease and gastroesophageal varices secondary to liver disease are the most frequent cause. Rebleeding is common (25% post variceal and 10% post nonvariceal bleeds) and is strongly associated with mortality. There has been little change in rebleeding rates and mortality over time, with a need for further strategies to address this.3.van Leerdam M.E. Epidemiology of acute upper gastrointestinal bleeding.Best Pract Res Clin Gastroenterol. 2008; 22: 209-224Crossref PubMed Scopus (273) Google Scholar Early management includes resuscitation with blood product transfusion, alongside medical, radiological, endoscopic, and surgical intervention. Access to diagnostic and/or therapeutic investigation is variable and may influence outcome.4.Tripathi D. Stanley A.J. Hayes P.C. et al.UK guidelines on the management of variceal haemorrhage in cirrhotic patients.Gut. 2015; 64: 1680-1704Crossref PubMed Scopus (348) Google Scholar Lower gastrointestinal bleeding is less frequent with an estimated annual incidence of 33 to 87 per 100 000, accounting for 3% of all emergency surgical referrals.5.Hreinsson J.P. Gumundsson S. Kalaitzakis E. Bjornsson E.S. Lower gastrointestinal bleeding: incidence, etiology, and outcomes in a population‐based setting.Eur J Gastroenterol Hepatol. 2013; 25: 37-43Crossref PubMed Scopus (109) Google Scholar, 6.Oakland K. Chadwick G. East J.E. et al.Diagnosis and management of acute lower gastrointestinal bleeding: guidelines from the British Society of Gastroenterology.Gut. 2019; 68: 776-789Crossref PubMed Scopus (107) Google Scholar The most common underlying causes are diverticular bleeding and other benign anorectal conditions. Mortality is substantially increased in those with bleeding onset while hospitalized but is most often related to underlying comorbidity, with death directly attributable to bleeding uncommon.7.Oakland K. Guy R. Uberoi R. et al.Acute lower GI bleeding in the UK: patient characteristics, interventions and outcomes in the first nationwide audit.Gut. 2018; 67: 654-662PubMed Google Scholar Tranexamic acid is an antifibrinolytic agent acting via inhibition of plasminogen activation, and is widely used for surgical bleeding.8.Hunt B.J. The current place of tranexamic acid in the management of bleeding.Anaesthesia. 2015; 70: 50-53, e18Crossref PubMed Scopus (109) Google Scholar The findings of CRASH‐2 a decade ago were practice‐changing. Early administration of tranexamic acid, a cheap and widely available drug, significantly reduced mortality due to bleeding in major trauma.9.CRASH‐2 Trial CollaboratorsEffects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH‐2): a randomised, placebo‐controlled trial.Lancet. 2010; 376: 23-32Abstract Full Text Full Text PDF PubMed Scopus (2108) Google Scholar The WOMAN study followed with a survival advantage in women with postpartum hemorrhage receiving early tranexamic acid.10.WOMAN Trial CollaboratorsEffect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post‐partum haemorrhage (WOMAN): an international, randomised, double‐blind, placebo‐controlled trial.Lancet. 2017; 389: 2105-2116Abstract Full Text Full Text PDF PubMed Scopus (719) Google Scholar There were no excess adverse events, particularly thromboembolic events, irrespective of timing of administration. These findings led some to hail tranexamic acid as a “wonder drug” for major bleeding,11.Franchini M. Mannucci P.M. The never ending success story of tranexamic acid in acquired bleeding.Haematologica. 2020; 105: 1201-1205Crossref Google Scholar embedding it in major hemorrhage protocols.12.Chay J. Koh M. Tan H.H. et al.A national common massive transfusion protocol (MTP) is a feasible and advantageous option for centralized blood services and hospitals.Vox Sang. 2016; 110: 36-50Crossref Scopus (12) Google Scholar, 13.NHS Blood and Transplant. National Comparative Audit of Blood Component Use; 2018 Audit of the Management of Major Bleeding. Available at https://nhsbtdbe.blob.core.windows.net/umbraco-assets-corp/19130/2018-major-haemorrhage-audit-full-report.pdf. Accessed July 18, 2020.Google Scholar Others urged caution with a need for randomized studies in other areas of major bleeding to confirm applicability and safety.14.Stanworth S.J. Manno D. Shakur H. et al.Extending evidence for the use of tranexamic acid from traumatic haemorrhage to other patients with major bleeding: do we need more than one haemorrhage protocol? The case of gastrointestinal bleeding.Transfus Med. 2015; 25: 198-200Crossref Scopus (1) Google Scholar In both the CRASH‐2 and WOMAN studies, the greatest survival advantage was derived with early administration and lost when delayed more than 3 hours after bleeding onset.15.Gayet‐Ageron A. Prieto‐Merino D. Ker K. Shakur H. Ageron F.X. Roberts I. Effect of treatment delay on the effectiveness and safety of antifibrinolytics in acute severe haemorrhage: a meta‐analysis of individual patient‐level data from 40 138 bleeding patients.Lancet. 2018; 391: 125-132Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar Of note, fibrinolytic activation is a recognized feature of both traumatic hemorrhage16.Gall L.S. Davenport R.A. Fibrinolysis and antifibrinolytic treatment in the trauma patient.Curr Opin Anaesthesiol. 2018; 31: 227-233Crossref Scopus (26) Google Scholar and postpartum hemorrhage.17.Roberts I. Shakur H. Fawole B. et al.Haematological and fibrinolytic status of Nigerian women with post‐partum haemorrhage.BMC Pregnancy Childbirth. 2018; 18: 143Crossref PubMed Scopus (25) Google Scholar Despite minimal evidence to support fibrinolysis as a feature of gastrointestinal bleeding, there are multiple systematic reviews and meta‐analyses of small, randomized controlled trials investigating mortality following tranexamic acid versus placebo in upper gastrointestinal bleeding. The most recent identified 10 studies including 2013 patients and reported a significant reduction in mortality in those receiving tranexamic acid (relative risk [RR], 0.59; 95% confidence interval [CI], 0.43‐0.82; P = 0.001).18.Twum‐Barimah E. Abdelgadir I. Gordon M. Akobeng A.K. Systematic review with meta‐analysis: the efficacy of tranexamic acid in upper gastrointestinal bleeding.Aliment Pharmacol Ther. 2020; 51: 1004-1013Crossref PubMed Scopus (14) Google Scholar There were numerous methodological weaknesses of the included studies with small sample sizes. Reporting of adverse outcomes was even less consistent with no increase in thromboembolic events apparent but assessed as low‐quality evidence. The authors acknowledge their findings cannot be considered conclusive.18.Twum‐Barimah E. Abdelgadir I. Gordon M. Akobeng A.K. Systematic review with meta‐analysis: the efficacy of tranexamic acid in upper gastrointestinal bleeding.Aliment Pharmacol Ther. 2020; 51: 1004-1013Crossref PubMed Scopus (14) Google Scholar There are no prior studies of tranexamic acid in lower gastrointestinal bleeding. HALT‐IT is a large multinational randomized double‐blind, placebo‐controlled trial investigating the effect of high‐dose tranexamic acid on mortality from gastrointestinal bleeding.19.HALT‐IT Trial CollaboratorsEffects of a high‐dose 24‐h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT‐IT): an international randomised, double‐blind, placebo‐controlled trial.Lancet. 2020; 395: 1927-1936Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar A total of 12 009 patients were randomized to tranexamic acid administered as a bolus of 1 g, followed by 3 g over 24 hours versus blinded placebo. The original planned primary outcome of all‐cause mortality was amended to death from bleeding within 5 days, following the observation toward the end of recruitment that the majority of deaths were unrelated to bleeding. The target recruitment was correspondingly increased to enable adequate power to detect a 25% reduction in relative risk of death from early bleeding. The primary endpoint was met in 4% (222/5956) of participants in the tranexamic acid group and 4% (226/5981) in the placebo group (RR, 0.99; 95% CI, 0.82‐1.18). Further analysis of secondary outcomes (deaths from bleeding within 24 hours and 28 days, all‐cause mortality, and rebleeding rates at the same timepoints) revealed similar findings. Arterial thromboembolism was similar between groups but venous thromboembolism was increased in the tranexamic acid group (0.8%, 48/5952) compared with placebo (0.4%, 26/5977; RR, 1.85; 95% CI, 1.15‐2.98). This finding of not only no benefit but also increased harm, vis à vis venous thromboembolic events will have surprised many, and lays the foundation for caution in using tranexamic acid outside areas with good quality evidence to support both efficacy and safety. Potential factors influencing the study findings are presented in Table 1. The investigators acknowledge a number of differences in the gastrointestinal bleeding population, and considered these in the study design. It is recognized that patients with gastrointestinal bleeding present later.14.Stanworth S.J. Manno D. Shakur H. et al.Extending evidence for the use of tranexamic acid from traumatic haemorrhage to other patients with major bleeding: do we need more than one haemorrhage protocol? The case of gastrointestinal bleeding.Transfus Med. 2015; 25: 198-200Crossref Scopus (1) Google Scholar The median time to randomization from bleeding onset in HALT‐IT was 21 hours, with <20% of participants randomized within 3 hours of symptom onset.19.HALT‐IT Trial CollaboratorsEffects of a high‐dose 24‐h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT‐IT): an international randomised, double‐blind, placebo‐controlled trial.Lancet. 2020; 395: 1927-1936Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar Secondary analysis of pooled data from CRASH‐2 and WOMAN studies highlighted the importance of early administration with a 10% fall in benefit for each 15‐minute delay in administration.15.Gayet‐Ageron A. Prieto‐Merino D. Ker K. Shakur H. Ageron F.X. Roberts I. Effect of treatment delay on the effectiveness and safety of antifibrinolytics in acute severe haemorrhage: a meta‐analysis of individual patient‐level data from 40 138 bleeding patients.Lancet. 2018; 391: 125-132Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar Administration of tranexamic acid 3 hours or more after bleeding onset had no survival benefit.15.Gayet‐Ageron A. Prieto‐Merino D. Ker K. Shakur H. Ageron F.X. Roberts I. Effect of treatment delay on the effectiveness and safety of antifibrinolytics in acute severe haemorrhage: a meta‐analysis of individual patient‐level data from 40 138 bleeding patients.Lancet. 2018; 391: 125-132Abstract Full Text Full Text PDF PubMed Scopus (189) Google Scholar Of note, in these studies, the majority of deaths occurred within 12 hours of bleeding onset, with a peak at 2 to 3 hours after childbirth in WOMAN. In contrast, in HALT‐IT, the median time to death was 2.4 days after randomization.19.HALT‐IT Trial CollaboratorsEffects of a high‐dose 24‐h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT‐IT): an international randomised, double‐blind, placebo‐controlled trial.Lancet. 2020; 395: 1927-1936Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar The later presentation of those with gastrointestinal bleeding is difficult to overcome, particularly because determining bleeding onset from the upper gastrointestinal tract is challenging. The use of a higher dose and prolonged infusion was aimed at prolonging therapeutic exposure during the highest risk period of rebleeding.20.Roberts I. Coats T. Edwards P. et al.HALT‐IT–tranexamic acid for the treatment of gastrointestinal bleeding: study protocol for a randomised controlled trial.Trials. 2014; 15: 450Crossref PubMed Scopus (69) Google Scholar Additionally, there are differences in patient characteristics; those with traumatic hemorrhage and postpartum hemorrhage are young (>75% younger than 44 years in CRASH‐2 and 80% younger than 34 years in WOMAN),9.CRASH‐2 Trial CollaboratorsEffects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH‐2): a randomised, placebo‐controlled trial.Lancet. 2010; 376: 23-32Abstract Full Text Full Text PDF PubMed Scopus (2108) Google Scholar, 10.WOMAN Trial CollaboratorsEffect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post‐partum haemorrhage (WOMAN): an international, randomised, double‐blind, placebo‐controlled trial.Lancet. 2017; 389: 2105-2116Abstract Full Text Full Text PDF PubMed Scopus (719) Google Scholar with likely little comorbid disease. In contrast, those with gastrointestinal bleeding are older (mean age 57 years in HALT‐IT, with only 13% <40 years) and more than 70% of participants had significant comorbidities (including 41% with liver disease and 7% malignancy).19.HALT‐IT Trial CollaboratorsEffects of a high‐dose 24‐h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT‐IT): an international randomised, double‐blind, placebo‐controlled trial.Lancet. 2020; 395: 1927-1936Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar Increasing age, malignancy, and liver disease are known to be associated with an increased risk of venous thrombosis.21.Rosendaal F.R. Risk factors for venous thrombotic disease.Thromb & Haemost. 1999; 82: 610-619Crossref PubMed Scopus (357) Google Scholar, 22.Lisman T. Porte R.J. Pathogenesis, prevention, and management of bleeding and thrombosis in patients with liver diseases.Res Pract Thromb Haemost. 2017; 1: 150-161Abstract Full Text Full Text PDF PubMed Scopus (82) Google ScholarTable 1Potential factors influencing efficacy of tranexamic acid and venous thrombosis risk in gastrointestinal bleedingFactors Potentially Reducing EfficacyFactors Potentially Increasing VTE•Delayed presentation from bleeding onset, and therefore delayed administration of tranexamic acid•Potential delays in definitive management of underlying bleeding lesion (eg, variable access to endoscopy/interventional radiology)•Possible lack of fibrinolysis activation in gastrointestinal bleeding•Heterogeneity in underlying cause of bleeding (not confirmed on subgroup analyses)•Pre‐existing hypofibrinolysis in patients with variceal bleeding/liver disease•Higher dose and prolonged infusion of tranexamic acid•Patient characteristicsa.Older ageb.Increased comorbidityc.Inclusion of patients with chronic liver disease•Potential use of fibrinogen replacement (concentrate/cryoprecipitate)Abbreviation: VTE, venous thromboembolism. Open table in a new tab Abbreviation: VTE, venous thromboembolism. Although coagulopathy is common in patients presenting with acute gastrointestinal bleeding, this is frequently explained by either anticoagulant use or underlying liver disease.23.Jairath V. Kahan B.C. Stanworth S.J. et al.Prevalence, management, and outcomes of patients with coagulopathy after acute nonvariceal upper gastrointestinal bleeding in the United Kingdom.Transfusion. 2013; 53: 1069-1076Crossref Scopus (39) Google Scholar Contemporary evaluation of fibrinolysis in gastrointestinal bleeding is nonexistent. There are comprehensive laboratory studies of coagulation and fibrinolysis in the nonbleeding, chronic liver disease population. It is now widely accepted that despite an apparent “coagulopathy,” with prolonged prothrombin time/INR on routine testing, overall coagulation is rebalanced when measured with global coagulation assays.24.Lisman T. Porte R.J. Rebalanced hemostasis in patients with liver disease: evidence and clinical consequences.Blood. 2010; 116: 878-885Crossref PubMed Scopus (425) Google Scholar Although there are reports of enhanced fibrinolysis in patients with liver disease, a recent study of acutely ill patients demonstrates a mixed fibrinolytic pattern, with some patients exhibiting marked hypofibrinolysis.22.Lisman T. Porte R.J. Pathogenesis, prevention, and management of bleeding and thrombosis in patients with liver diseases.Res Pract Thromb Haemost. 2017; 1: 150-161Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar, 25.Blasi A. Patel V.C. Adelmeijer J. et al.Mixed fibrinolytic phenotypes in decompensated cirrhosis and acute‐on‐chronic liver failure with hypofibrinolysis in those with complications and poor survival.Hepatol. 2020; 71: 1381-1390Crossref PubMed Scopus (47) Google Scholar Hypofibrinolysis was more prevalent with increasing illness severity and of note, there was no evidence of increased fibrinolysis in the subgroup of patients with bleeding complications. Additionally, a small study in those with active variceal bleeding (n = 14) demonstrated preserved thrombin generation compared with both nonbleeding patients with cirrhosis and healthy controls.26.Jairath V. Harrison P. Stanworth S. Collier J. Murphy M. Barnes E. Thrombin generation is normal in cirrhotics with acute variceal haemorrhage: results from a prospective study.Gut. 2012; 61: A418Crossref Google Scholar At the time of HALT‐IT launch, there was little published with regard to epidemiology and outcomes from lower gastrointestinal bleeding. Subsequent studies have confirmed bleeding is rarely life‐threatening with the majority of deaths attributed to comorbid disease rather than bleeding.7.Oakland K. Guy R. Uberoi R. et al.Acute lower GI bleeding in the UK: patient characteristics, interventions and outcomes in the first nationwide audit.Gut. 2018; 67: 654-662PubMed Google Scholar Of note, in HALT‐IT, only 4% of all early bleeding deaths (within 5 days) were due to lower gastrointestinal bleeding (despite comprising 10% of all participants).19.HALT‐IT Trial CollaboratorsEffects of a high‐dose 24‐h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT‐IT): an international randomised, double‐blind, placebo‐controlled trial.Lancet. 2020; 395: 1927-1936Abstract Full Text Full Text PDF PubMed Scopus (135) Google Scholar In contrast, 72% of all early bleeding deaths occurred in those with variceal bleeding or liver disease. Although there was no heterogeneity in the study findings on subgroup analysis by bleeding site, and this is the largest randomized controlled trial to date, the authors acknowledge a small increase/decrease in mortality from bleeding cannot be excluded. It seems likely that if a small benefit does exist, this is most plausible in nonvariceal upper gastrointestinal bleeding. In HALT‐IT, a greater proportion (69%) of participants received blood products compared with CRASH‐2 and WOMAN, where approaching one‐half did not require transfusion. In HALT‐IT of those transfused, 98% received red blood cells/whole blood, with 23% and 6% receiving frozen plasma and platelets respectively. Data regarding the use of cryoprecipitate/fibrinogen concentrate were not collected. A previous national audit of management of major bleeding in England (2018) reported cryoprecipitate use in 15% of major hemorrhage from gastrointestinal bleeding (with 67% and 32% receiving frozen plasma and platelets. respectively; n = 165).13.NHS Blood and Transplant. National Comparative Audit of Blood Component Use; 2018 Audit of the Management of Major Bleeding. Available at https://nhsbtdbe.blob.core.windows.net/umbraco-assets-corp/19130/2018-major-haemorrhage-audit-full-report.pdf. Accessed July 18, 2020.Google Scholar The audit defined major hemorrhage on the basis of number of red blood cell units transfused and is not directly comparable to the HALT‐IT population (inclusion based on clinical criteria for severe bleeding). Nevertheless, cryoprecipitate use is common and while fibrinogen replacement is likely to have been similarly balanced across groups in HALT‐IT, its effect on outcomes in gastrointestinal bleeding is of interest. The HALT‐IT findings serve as a reminder of the care needed in using meta‐analyses as the gold standard for establishing evidence‐based practice. Overuse and lack of clinical utility of systematic reviews and meta‐analysis is recognized.27.Ioannidis J.P.A. The mass production of redundant, misleading and conflicted systematic reviews and meta‐analyses.Milbank Q. 2016; 94: 485-514Crossref PubMed Scopus (738) Google Scholar To further illustrate, a study in perioperative medicine of 16 interventions identified 18 large randomized controlled trials with 44 preceding meta‐analyses.28.Sivakumar H. Peyton P.J. Poor agreement in significant findings between meta‐analyses and subsequent large randomized trials in perioperative medicine.Br J Anaesth. 2016; 117: 431-441Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar The preceding meta‐analysis accurately predicted the outcome of the subsequent, large randomized controlled trial in only 64%, highlighting a significant proportion with discordant outcome.28.Sivakumar H. Peyton P.J. Poor agreement in significant findings between meta‐analyses and subsequent large randomized trials in perioperative medicine.Br J Anaesth. 2016; 117: 431-441Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar In this case, preceding meta‐analyses reported a large survival benefit associated with tranexamic acid use in upper gastrointestinal bleeding. The apparent benefit is likely due to methodological issues in the primary studies, the small sample size, and potential publication bias of positive studies. The HALT‐IT collaborative acknowledges their study cannot exclude a small survival benefit but the large effect reported from meta‐analyses can be confidently discounted. Although HALT‐IT subgroup analysis did not identify heterogeneity in mortality outcomes based on etiology, death from bleeding was uncommon following lower gastrointestinal bleeding (16/1328, 1.2%). Subgroup analysis highlighted the increased risk of venous thrombosis in those with suspected variceal bleeding or liver disease. Future interventions to reduce bleeding deaths should focus on upper gastrointestinal bleeding and investigate variceal bleeding as a distinct entity. Finally, care is needed in extrapolating practice from other bleeding settings, particularly trauma. Just as major obstetric hemorrhage has distinct features,29.Allard S. Green L. Hunt B.J. How we manage the haematological aspects of major obstetric haemorrhage.Br J Haematol. 2013; 164: 177-188Crossref Scopus (39) Google Scholar so may major gastrointestinal bleeding. Audits in the United Kingdom demonstrate increased use of tranexamic acid use in gastrointestinal bleeding from <1% in 2007 to 67% in 2018.13.NHS Blood and Transplant. National Comparative Audit of Blood Component Use; 2018 Audit of the Management of Major Bleeding. Available at https://nhsbtdbe.blob.core.windows.net/umbraco-assets-corp/19130/2018-major-haemorrhage-audit-full-report.pdf. Accessed July 18, 2020.Google Scholar, 20.Roberts I. Coats T. Edwards P. et al.HALT‐IT–tranexamic acid for the treatment of gastrointestinal bleeding: study protocol for a randomised controlled trial.Trials. 2014; 15: 450Crossref PubMed Scopus (69) Google Scholar Major hemorrhage protocols must be updated to recommend tranexamic acid only where evidence supports its use, namely in the context of surgical, traumatic, and obstetric bleeding. This raises further questions regarding the optimal early management of major gastrointestinal hemorrhage, some of which is similarly based on studies in traumatic hemorrhage. For example, is a fixed ratio of red blood cells and fresh frozen plasma appropriate and what is the optimal ratio? Can viscoelastic testing better guide early blood product replacement? An exploratory randomized controlled study of thromboelastography to guide blood product replacement versus usual care in variceal bleeding (n = 60) reduced transfusion requirements with no apparent adverse impact on mortality or early rebleeding.30.Rout G. Shalimar G.D. Mahapatra S.J. Kedia S. Garg P.K. Nayak B. Thromboelastography‐guided blood product transfusion in cirrhosis patients with variceal bleeding: a randomized controlled trial.J Clin Gastroenterol. 2020; 54: 255-262Crossref PubMed Scopus (50) Google Scholar Does fibrinogen replacement (cryoprecipitate or concentrate) have a role in the management of variceal bleeding or might it be similarly associated with an excess risk of thrombosis? There is an urgent need to address these uncertainties in carefully designed multicenter randomized controlled trials. HALT‐IT has demonstrated such studies are possible in this patient group, thereby setting the standard and future direction for establishing best practice in the early management of gastrointestinal bleeding. Dr. Roberts has received educational grants from Sanofi and Bayer for conference attendance. Lara N. Roberts researched and wrote the manuscript. The author is grateful to colleagues at King's Thrombosis Centre for their review and helpful comments.