Validation of original, expanded Baveno VI, and stepwise & platelet-MELD criteria to rule out varices needing treatment in compensated cirrhosis from various etiologies

Validation of a Machine Learning Algorithm, EVendo, for Predicting Esophageal Varices in Hepatocellular Carcinoma

BACKGROUND

Treatment with atezolizumab and bevacizumab has become standard of care for advanced unresectable hepatocellular carcinoma (HCC) but carries an increased gastrointestinal bleeding risk. Therefore, patients are often required to undergo esophagogastroduodenoscopy (EGD) to rule out esophageal varices (EV) prior to initiating therapy, which can delay care and lead to unnecessary procedural risks and health care costs. In 2019, the EVendo score was created and validated as a noninvasive tool to accurately screen out patients who were at low risk for having EV that required treatment. We sought to validate whether the EVendo score could be used to accurately predict the presence of EV and varices needing treatment (VNT) in patients with HCC.

METHODS

This was a retrospective multicenter cohort study of patients with HCC from 9/2004 to 12/2021. We included patients who underwent EGDs within 1 year after their HCC diagnosis. We collected clinical parameters needed to calculate an EVendo score at the time of EGD and compared the EVendo model prediction to the gold standard endoscopic report in predicting presence of VNT.

RESULTS

112 with HCC were recruited to this study, with 117 qualifying EGDs. VNT occurred in 39 (33.3%) patients. The EVendo score had a sensitivity of 97.4% and a negative predictive value of 96.9%, supporting the validity in applying EVendo in predicting VNT in HCC.

CONCLUSION

In this study, we validated the use of the EVendo score in ruling out VNT in patients with HCC. The application of the EVendo score could safely defer about 30% of EGDs for EV screening in HCC patients. Although additional validation cohorts are needed, this suggests that EVendo score can potentially be applied in patients with HCC to avoid unnecessary EGDs, which can ultimately mitigate healthcare costs and delays in initiating HCC treatment with atezolizumab and bevacizumab.

Performance of Baveno VII criteria for the screening of varices needing treatment in patients with biliary atresia

OBJECTIVE

Biliary atresia (BA) is the leading cause of liver cirrhosis and chronic liver insufficiency in children in the world. Gastroesophageal varices bleeding is an ominous complication of cirrhosis in BA patients and is associated with high morbidity and mortality. In this study, we aimed to investigate the utility of noninvasive Baveno VI and Baveno VII criteria for the screening of varices need treatment (VNT) and the need for liver transplantation in BA patients.

METHODS

This study enrolled 48 BA patients (23 females and 25 males) who underwent an esophagogastroduodenoscopy (EGD) and transient elastography at a mean age of 11.18 ± 1.48 years; the clinical data were surveyed in a retrospective design.

RESULTS

The sensitivity and negative predictive value of Baveno VI and Baveno VII criteria for the prediction of VNT in BA patients are both 100% and 100%, respectively. The VNT missing rate of Baveno VI and Baveno VII criteria are both 0% in our cohort. The Baveno VI, expanded Baveno VI, and Baveno VII criteria are also predictive of the need for liver transplantation in our cohort (OR = 10.33, 4.24, and 21.33; p = 0.009, 0.03, and 0.007, respectively).

CONCLUSION

The Baveno VI and Baveno VII criteria are useful for the screening of VNT and minimize non-necessary invasive EGD in BA patients with low VNT missing rates. The Baveno VI, expanded Baveno VI, and Baveno VII criteria are associated with the need for liver transplantation.

Prediction of Severe Esophageal Varices in Patients With Cirrhosis Based on Levitt's CO Breath Test: A Proof of Concept Study

GOALS

This study investigated the feasibility of using erythrocyte (RBC) lifespan determined by Levitt's CO breath test (LCOBT) to predict esophageal varices needing treatment (VNT) in patients with cirrhosis.

BACKGROUND

Esophageal varix bleeding is a common fatal complication of cirrhosis and portal hypertension. The gold standard for identifying VNT is esophagogastroduodenoscopy (EGD), an invasive procedure with low patient compliance. VNT screening based on Baveno VI criteria has mediocre specificity.

STUDY

RBC lifespan was determined by LCOBT in 53 cirrhotic patients (13 without varices, 11 mild/moderate varices, and 29 severe varices). Correlation of varix severity with RBC lifespan and other variables was analyzed. Rates of shortened RBC lifespan and thrombocytopenia (Baveno VI criteria) were compared.

RESULTS

RBC lifespan correlated inversely with severity of varices ( r =-0.793, P <0.001). Mean RBC lifespans were 129±31, 96±21, and 59±21 days for Nonvarix, Mild/Moderate, and Severe groups. Shortened RBC lifespan (<75 d) was observed in 79.3% (23/29) of patients with severe varices, a frequency similar or identical to thrombocytopenia rates [original Baveno VI criteria, 86.2% (25/29), P =0.487; expanded criteria, 79.3% (23/29), P >0.999]. Among 24 patients without severe varices, shortened RBC lifespan was observed in 1 patient whereas thrombocytopenia was detected in 13 and 8 patients based on the original ( P <0.001) and expanded criteria ( P =0.010), respectively.

CONCLUSIONS

RBC lifespan correlates inversely with varix severity in patients with cirrhosis. LCOBT may enable specific screening for VNT.

Validation of the Baveno VI criteria to rule out high-risk varices using hepatic shear-wave elastography

INTRODUCTION AND OBJECTIVES

Liver stiffness measurement (LSM) by transient elastography has been validated to predict high-risk varices (HRV). We aimed to evaluate the accuracy of shear-wave elastography (SWE) and platelet count (Baveno VI criteria) to rule out HRV in patients with compensated advanced chronic liver disease (c-ACLD).

METHODS

This retrospective study analyzed data of patients with c-ACLD (transient elastography ≥ 10 kPa) submitted to two-dimensional SWE (2D-SWE) (GE-LOGIQ-S8) and/or point SWE (p-SWE) (ElastPQ) who had a gastrointestinal endoscopy within 24 months. HRV definition was a large size and presence of red wale marks or sequelae from previous treatment. Optimal thresholds of SWE systems for HRV were identified. The proportion of spared gastrointestinal endoscopies and missing HRV considering a favorable SWE Baveno VI criteria were assessed.

RESULTS

Eighty patients [36% male, median age = 63 (interquartile range, 57-69) years] were included. The prevalence of HRV was 34% ( n  = 27/80). The optimal thresholds to predict HRV were 10 kPa and 12 kPa for 2D-SWE and p-SWE, respectively. A favorable 2D-SWE Baveno VI criteria (LSM < 10 kPa and platelets count > 150 × 10 9 /mm 3 ) avoided 19% of gastrointestinal endoscopies without missing HRVs. A favorable p-SWE Baveno VI criteria (LSM < 12 kPa and platelets count > 150 × 10 9 /mm 3 ) spared 20% of gastrointestinal endoscopy without missing HRVs. Using a lower threshold of platelet count (<110 × 10 9 /mm 3 , expanded Baveno VI), 2D-SWE (<10 kPa) avoided 33% of gastrointestinal endoscopy with 8% of missing HRVs, while p-SWE (<12 kPa) avoided 36% of gastrointestinal endoscopy with 5% of missing HRVs.

CONCLUSION

LSM by p-SWE or 2D-SWE combined with platelet count (Baveno VI criteria) can spare a considerable number of gastrointestinal endoscopies missing a negligible proportion of HRV.

Automated Multimodal Machine Learning for Esophageal Variceal Bleeding Prediction Based on Endoscopy and Structured Data

Esophageal variceal (EV) bleeding is a severe medical emergency related to cirrhosis. Early identification of cirrhotic patients with at a high risk of EV bleeding is key to improving outcomes and optimizing medical resources. This study aimed to evaluate the feasibility of automated multimodal machine learning (MMML) for predicting EV bleeding by integrating endoscopic images and clinical structured data. This study mainly includes three steps: step 1, developing deep learning (DL) models using EV images by 12-month bleeding on TensorFlow (backbones include ResNet, Xception, EfficientNet, ViT and ConvMixer); step 2, training and internally validating MMML models integrating clinical structured data and DL model outputs to predict 12-month EV bleeding on an H2O-automated machine learning platform (algorithms include DL, XGBoost, GLM, GBM, RF, and stacking); and step 3, externally testing MMML models. Furthermore, existing clinical indices, e.g., the MELD score, Child‒Pugh score, APRI, and FIB-4, were also examined. Five DL models were transfer learning to the binary classification of EV endoscopic images at admission based on the occurrence or absence of bleeding events during the 12-month follow-up. An EfficientNet model achieved the highest accuracy of 0.868 in the validation set. Then, a series of MMML models, integrating clinical structured data and the output of the EfficientNet model, were automatedly trained to predict 12-month EV bleeding. A stacking model showed the highest accuracy (0.932), sensitivity (0.952), and F1-score (0.879) in the test dataset, which was also better than the existing indices. This study is the first to evaluate the feasibility of automated MMML in predicting 12-month EV bleeding based on endoscopic images and clinical variables.

Nomogram based on spleen volume expansion rate predicts esophagogastric varices bleeding risk in patients with hepatitis B liver cirrhosis

BACKGROUND

We aimed to explore the risk factors for hemorrhage of esophagogastric varices (EGVs) in patients with hepatitis B cirrhosis and to construct a novel nomogram model based on the spleen volume expansion rate to predict the risk of esophagogastric varices bleeding.

METHODS

Univariate and multivariate logistic regression analysis was used to analyze the risk factors for EGVs bleeding. Nomograms were established based on the multivariate analysis results. The predictive accuracy of the nomograms was assessed using the area under the curve (AUC or C-index) of the receiver operating characteristic (ROC) and calibration curves. Decision curve analysis was used to determine the clinical benefit of the nomogram. We created a nomogram of the best predictive models.

RESULTS

A total of 142 patients' hepatitis B cirrhosis with esophagogastric varices were included in this study, of whom 85 (59.9%) had a history of EGVs bleeding and 57 (40.1%) had no EGVs bleeding. The spleen volume expansion rate, serum sodium levels (mmol/L), hemoglobin levels (g/L), and prothrombin time (s) were independent predictors for EGVs bleeding in patients with hepatitis B liver cirrhosis (P < 0.05). The above predictors were included in the nomogram prediction model. The area under the ROC curve (AUROC) of the nomogram was 0.781, the C-index obtained by internal validation was 0.757, and the calibration prediction curve fit well with the ideal curve. The AUROCs of the PLT-MELD and APRI were 0.648 and 0.548, respectively.

CONCLUSION

In this study, a novel nomogram for predicting the risk of EGVs bleeding in patients with hepatitis B cirrhosis was successfully constructed by combining the spleen volume expansion rate, serum sodium levels, hemoglobin levels, and prothrombin time. The predictive model can provide clinicians with a reference to help them make clinical decisions.

Endoscopic diagnosis and management of esophagogastric variceal hemorrhage: European Society of Gastrointestinal Endoscopy (ESGE) Guideline

1: ESGE recommends that patients with compensated advanced chronic liver disease (ACLD; due to viruses, alcohol, and/or nonobese [BMI < 30 kg/m²] nonalcoholic steatohepatitis) and clinically significant portal hypertension (hepatic venous pressure gradient [HVPG] > 10 mmHg and/or liver stiffness by transient elastography > 25 kPa) should receive, if no contraindications, nonselective beta blocker (NSBB) therapy (preferably carvedilol) to prevent the development of variceal bleeding.Strong recommendation, moderate quality evidence. 2: ESGE recommends that in those patients unable to receive NSBB therapy with a screening upper gastrointestinal (GI) endoscopy that demonstrates high risk esophageal varices, endoscopic band ligation (EBL) is the endoscopic prophylactic treatment of choice. EBL should be repeated every 2-4 weeks until variceal eradication is achieved. Thereafter, surveillance EGD should be performed every 3-6 months in the first year following eradication.Strong recommendation, moderate quality evidence. 3: ESGE recommends, in hemodynamically stable patients with acute upper GI hemorrhage (UGIH) and no history of cardiovascular disease, a restrictive red blood cell (RBC) transfusion strategy, with a hemoglobin threshold of ≤ 70 g/L prompting RBC transfusion. A post-transfusion target hemoglobin of 70-90 g/L is desired.Strong recommendation, moderate quality evidence. 4 : ESGE recommends that patients with ACLD presenting with suspected acute variceal bleeding be risk stratified according to the Child-Pugh score and MELD score, and by documentation of active/inactive bleeding at the time of upper GI endoscopy.Strong recommendation, high quality of evidence. 5 : ESGE recommends the vasoactive agents terlipressin, octreotide, or somatostatin be initiated at the time of presentation in patients with suspected acute variceal bleeding and be continued for a duration of up to 5 days.Strong recommendation, high quality evidence. 6 : ESGE recommends antibiotic prophylaxis using ceftriaxone 1 g/day for up to 7 days for all patients with ACLD presenting with acute variceal hemorrhage, or in accordance with local antibiotic resistance and patient allergies.Strong recommendation, high quality evidence. 7 : ESGE recommends, in the absence of contraindications, intravenous erythromycin 250 mg be given 30-120 minutes prior to upper GI endoscopy in patients with suspected acute variceal hemorrhage.Strong recommendation, high quality evidence. 8 : ESGE recommends that, in patients with suspected variceal hemorrhage, endoscopic evaluation should take place within 12 hours from the time of patient presentation provided the patient has been hemodynamically resuscitated.Strong recommendation, moderate quality evidence. 9 : ESGE recommends EBL for the treatment of acute esophageal variceal hemorrhage (EVH).Strong recommendation, high quality evidence. 10 : ESGE recommends that, in patients at high risk for recurrent esophageal variceal bleeding following successful endoscopic hemostasis (Child-Pugh C  ≤ 13 or Child-Pugh B > 7 with active EVH at the time of endoscopy despite vasoactive agents, or HVPG > 20 mmHg), pre-emptive transjugular intrahepatic portosystemic shunt (TIPS) within 72 hours (preferably within 24 hours) must be considered.Strong recommendation, high quality evidence. 11 : ESGE recommends that, for persistent esophageal variceal bleeding despite vasoactive pharmacological and endoscopic hemostasis therapy, urgent rescue TIPS should be considered (where available).Strong recommendation, moderate quality evidence. 12 : ESGE recommends endoscopic cyanoacrylate injection for acute gastric (cardiofundal) variceal (GOV2, IGV1) hemorrhage.Strong recommendation, high quality evidence. 13: ESGE recommends endoscopic cyanoacrylate injection or EBL in patients with GOV1-specific bleeding.Strong recommendations, moderate quality evidence. 14: ESGE suggests urgent rescue TIPS or balloon-occluded retrograde transvenous obliteration (BRTO) for gastric variceal bleeding when there is a failure of endoscopic hemostasis or early recurrent bleeding.Weak recommendation, low quality evidence. 15: ESGE recommends that patients who have undergone EBL for acute EVH should be scheduled for follow-up EBLs at 1- to 4-weekly intervals to eradicate esophageal varices (secondary prophylaxis).Strong recommendation, moderate quality evidence. 16: ESGE recommends the use of NSBBs (propranolol or carvedilol) in combination with endoscopic therapy for secondary prophylaxis in EVH in patients with ACLD.Strong recommendation, high quality evidence.

Noninvasive assessment oesophageal varices: impact of the Baveno VI criteria

PURPOSE OF REVIEW

In 2015, as a consequence of the high development in noninvasive tests, Baveno VI consensus recommended for the first time the use of a prediction rule (liver stiffness <20kPa and platelet count > 150000) to identify patients at low risk of having varices and that could circumvent endoscopy. These became known as the Baveno VI criteria. We review here the data validating Baveno VI criteria and we discuss the attempts of expanding these criteria.

RECENT FINDINGS

We report 28 studies assessing the performance of Baveno VI criteria showing a pooled 99% negative predictive value for ruling out high-risk varices. Performance is not affected by the cause of cirrhosis. Different attempts at expanding these criteria show suboptimal performance. Nonelastography-based criteria require further validation.

SUMMARY

Baveno VI criteria can be safely used to avoid endoscopy in a substantial proportion of patients with compensated cirrhosis. The progressive change in approach to the management of compensated cirrhosis, progressively focusing on treating portal hypertension with beta-blockers independently of the presence of varices, might render these criteria less relevant.

Unlocking the benefits of the Baveno VI guidance when screening for varices: an audit of clinical practice across London

Background

The Baveno VI consensus identifies patients with compensated advanced chronic liver disease (cACLD) who can safely avoid screening endoscopy. However, concordance in clinical practice with this guidance is unknown. We audited clinical practice and the provision of transient elastography (TE) aiming to identify potential cost savings and benefits.

Methods

Retrospective data collection from 12 sites across London over 6 months by reviewing oesophagogastroduodenoscopy (OGD) reports, platelet count and TE results as well as information on site-specific provision of TE.

Results

Three-hundred and fifty-one screening procedures were identified; 177 (50.43%) had a TE test performed within the preceding 12 months; 142 (80.23%) patients with a recent TE test did not meet criteria for screening OGD. TE provision varied widely between sites.

Conclusion

Improving concordance with the Baveno criteria through improved provision of TE would have benefits for patients, healthcare systems and the environment and would help to address the challenges of moving on from the COVID-19 pandemic.

Doppler hemodynamic liver index and hepatic elastic modulus: Advanced ultrasonographic techniques for non-invasive prediction of esophageal varices in cirrhosis

PURPOSE

Currently esophagoscopy is the gold standard for assessment of esophageal varices in cirrhosis. Predicting the presence of esophageal varices, varices needing treatment (VNT) and variceal grade by advanced ultrasonographic techniques using a combination of Doppler hemodynamic liver index (HDLI) (quantifying portal hypertension) and hepatic elastic modulus (quantifying hepatic fibrosis) would be a cost-effective and non-invasive alternative to routine endoscopy.

METHODS

Our cross-sectional study consisted of cirrhotic patients diagnosed using clinical features and laboratory parameters. Portal venous Doppler and liver sonoelastography were performed in selected subjects for obtaining measurements of HDLI (portal vein diameter/mean velocity) and hepatic elastic modulus respectively. Within 3 days of ultrasound, the subjects underwent upper GI endoscopy for assessment of presence, VNT and grade (F1, F2, F3) of varices. Subjects were divided into two groups (without and with varices) and data analyzed using XLSTAT.

RESULTS

A total of 60 subjects (26 without and 34 with varices) were evaluated. Mean Doppler HDLI of subjects with varices was significantly higher (0.72 vs. 0.59, p < 0.0001) with progressive increase in values with variceal grade, being highest in grade-F3 (mean 0.77). Likewise, mean hepatic elastic modulus was also higher in subjects with varices (28.9 vs. 12.6 kPa, p < 0.0001) and showed progressive increase with grade (51.1 kPa for F3). For predicting presence of varices, maximum accuracy of elastic modulus was at cut-off of 14.5 kPa (sensitivity, specificity 83% and 84.6%) and Doppler HDLI was at 0.66 (66% and 92.3%). Good inter-rater agreement was present (κ 0.66).

CONCLUSION

Combination of Doppler HDLI and hepatic elastic modulus is an excellent non-invasive method for predicting the presence, VNT, and variceal grade and may obviate need for routine endoscopic screening in cirrhosis.

A New Scoring System for Predicting In-hospital Death in Patients Having Liver Cirrhosis With Esophageal Varices

Introduction: Liver cirrhosis is caused by the development of various acute and chronic liver diseases. Esophageal varices is a common and serious complication of liver cirrhosis during decompensation. Despite the development of various treatments, the prognosis for liver cirrhosis with esophageal varices (LCEV) remains poor. We aimed to establish and validate a nomogram for predicting in-hospital death in LCEV patients.

Methods: Data on LCEV patients were extracted from the Medical Information Mart for Intensive Care III and IV (MIMIC-III and MIMIC-IV) database. The patients from MIMIC-III were randomly divided into training and validation cohorts. Training cohort was used for establishing the model, validation and MIMIC-IV cohorts were used for validation. The independent prognostic factors for LCEV patients were determined using the least absolute shrinkage and selection operator (LASSO) method and forward stepwise logistic regression. We then constructed a nomogram to predict the in-hospital death of LCEV patients. Multiple indicators were used to validate the nomogram, including the area under the receiver operating characteristic curve (AUC), calibration curve, Hosmer-Lemeshow test, integrated discrimination improvement (IDI), net reclassification index (NRI), and decision curve analysis (DCA).

Results: Nine independent prognostic factors were identified by using LASSO and stepwise regressions: age, Elixhauser score, anion gap, sodium, albumin, bilirubin, international normalized ratio, vasopressor use, and bleeding. The nomogram was then constructed and validated. The AUC value of the nomogram was 0.867 (95% CI = 0.832–0.904) in the training cohort, 0.846 (95% CI = 0.790–0.896) in the validation cohort and 0.840 (95% CI = 0.807–0.872) in the MIMIC-IV cohort. High AUC values indicated the good discriminative ability of the nomogram, while the calibration curves and the Hosmer-Lemeshow test results demonstrated that the nomogram was well-calibrated. Improvements in NRI and IDI values suggested that our nomogram was superior to MELD-Na, CAGIB, and OASIS scoring system. DCA curves indicated that the nomogram had good value in clinical applications.

Conclusion: We have established the first prognostic nomogram for predicting the in-hospital death of LCEV patients. The nomogram is easy to use, performs well, and can be used to guide clinical practice, but further external prospective validation is still required.

Validation and Refinement of the Baveno VI Criteria for Ruling Out High-Risk Varices

In the past decade, numerous studies have evaluated the roles of noninvasive methods for diagnosing or excluding varices and high-risk varices in patients with liver cirrhosis. The Baveno VI criteria recommend the use of a simple algorithm based on a liver stiffness measurement < 20 kPa through transient elastography and a platelet count > 150 × 10⁹/L for ruling out high-risk varices in patients with compensated advanced chronic liver disease. A large number of studies have validated the clinical usefulness of Baveno VI criteria for excluding high-risk varices. Several strategies have been proposed to refine the Baveno VI criteria; however, currently there is no review to summarize the diagnostic accuracy and limitations of the Baveno VI criteria after extensive validation. In this review, we summarize the diagnostic accuracy and limitations of the Baveno VI criteria after extensive validation. We also discuss methods to refine these criteria.

Deciding Among Noninvasive Tools for Predicting Varices Needing Treatment in Chronic Liver Disease: An Analysis of Asian Cohort

INTRODUCTION

Both transient elastography (TE)-based and non-TE-based criteria exist for detection of varices needing treatment (VNT) in patients with asymptomatic advanced chronic liver disease (CLD). However, their performance in clinical settings at different risk thresholds of detection of VNT and in regions where elastography is not widely available is unknown. We aimed to validate existing noninvasive criteria in our patients with CLD and identify best TE- and non-TE-based criteria for VNT screening at usual risk thresholds.

METHODS

Patients with compensated advanced CLD (cACLD) who underwent esophagogastroduodenoscopy and TE within 3 months were included. Diagnostic performance of Baveno VI, expanded Baveno VI, platelet-model for end-stage liver disease, and platelet-albumin (Rete Sicilia Selezione Terapia-hepatitis C virus) criteria were estimated. Decision curve analysis was conducted for different predictors across range of threshold probabilities. A repeat analysis including all patients with compensated CLD (cACLD and non-cACLD) was performed to simulate absence of TE.

RESULTS

A total of 1,657 patients (cACLD, 895; non-cACLD, 762) related to hepatitis B virus (38.2%), hepatitis C virus (33.4%), nonalcoholic steatohepatitis (14.7%), and alcohol (11.8%) were included. Baveno VI identified maximum VNT (97.3%) and had best negative predictive value (96.9%), followed by platelet-albumin criteria. Expanded Baveno VI and platelet-model for end-stage liver disease had intermediate performance. At threshold probability of 5%, Baveno VI criteria showed maximum net benefit, and platelet-albumin criteria was next best, with need for 95 additional elastographies to detect 1 additional VNT. Similar results were obtained on including all patients with compensated CLD irrespective of TE.

DISCUSSION

Baveno VI criteria maximizes VNT yield at 5% threshold probability. An acceptable alternative is the platelet-albumin criteria in resource-limited settings.