1. INTRODUCTION

Endoscopic retrograde cholangiopancreatography (ERCP) has been a reliable and widely accepted procedure for pancreaticobiliary disease since it was first reported in 1968.1 However, ERCP continues to be associated with several adverse events, which are attributed to the procedural invasiveness.2,3 Due to the blind insertion of duodenoscope and coagulation abnormality as well as thrombocytopenia in patients with liver cirrhosis, variceal bleeding precipitated by ERCP is a major concern, but there is no robust evidence to highlight the risk.

Esophageal variceal bleeding was one of the major complications of liver cirrhosis and had a 6-week mortality rate between 15% and 25% despite the advancement of treatment.4–7 A previous study showed no event of esophageal variceal bleeding in 23 cirrhotic patients undergoing ERCP. The overall procedure-associated bleeding risk was 13.5% in cirrhotic patients, regardless of the existence of esophageal varices (EV).8 The other four studies enrolled cirrhotic patients who received ERCP and reported rates of procedure-associated bleeding of 1.1%, 2.3%, 2.5%, and 5.7%, respectively.9–12 However, these studies were focused on the overall adverse events in cirrhotic patients without special emphasis on the etiology of bleeding, especially variceal bleeding, or making comparisons to patients without varices. Therefore, clarifying the risk of variceal bleeding precipitated by ERCP is important.

Here, we compare the risk of post-ERCP variceal bleeding as well as other adverse events between cirrhotic patients with and without EV.

2. METHODS 2.1. Patients

This study is a retrospective study. A total of 4445 cases who received ERCP were screened from October 2010 to November 2017, and 75 adult patients with liver cirrhosis were enrolled in a single tertiary medical center. Data were collected within 30 days of the procedure.13 The diagnosis of liver cirrhosis was based on liver biopsy or clinical, biochemical, and radiological findings. The decompensation of liver cirrhosis was defined by the presence of ascites, variceal bleeding, encephalopathy, and/or jaundice.14 The study was performed in accordance with the Declaration of Helsinki and current ethical guidelines. It was also approved by the Institutional Review Board (2021-06-005AC).

2.2. Endoscopic evaluation

The size of EV was recorded following the criteria proposed by Beppu et al15 small and straight varices were recorded as F1, moderately sized and tortuous varices were recorded as F2, and large and tumorous varices were recorded as F3. The EV size of F2, F3, and F1 with red-color signs or Child-Turcotte-Pugh C hepatic reserve were defined as high-risk EV.5,16

2.3. Standard of medical care and procedure

The indications for ERCP, liver function tests, and image studies, including abdominal sonography, computer tomography, or magnetic resonance imaging of all the patients before the procedure, were evaluated by experienced endoscopists, gastroenterologists, and radiologists. Premedication with local pharyngeal 10% lidocaine spray and intravenous injection of midazolam and hyoscine-N-butylbromide were routinely prescribed if there was no contraindication. Antibiotic prophylaxis was not given. Therapeutic interventions were performed when indicated, such as endoscopic sphincterotomy, endoscopic papillary balloon dilation, endoscopic retrograde biliary drainage, and nasobiliary drainage. All patients under antiplatelet therapy stopped 7 days before the procedure. The patient under warfarin therapy was discontinued 5 days before the procedure with low-molecular-weight heparin bridging therapy and gave the last dose 24 hours before the procedure.

2.4. Definition of bleeding

Overall gastrointestinal (GI) bleeding is defined as any signs of bleeding from the GI tract, including endoscopic and clinical findings. Clinically significant GI bleeding was defined as clinical evidence of melena or hematemesis, with hemoglobin level dropping by more than 2 g/dL or the need for blood transfusion.13,17 The definition of variceal bleeding was patients with clinical manifestations of GI bleeding, such as hematemesis, coffee-ground vomitus, hematochezia, or melena with endoscopic evidence of active bleeding, including adherent blood clots or erosions on varices and/or large varices with a red-color sign in the absence of other sources of bleeding.

2.5. Statistical analysis

The baseline characteristics to be evaluated with outcomes were selected and expressed as mean ± standard deviation for parametric variables or median ± interquartile range for nonparametric variables. The Mann-Whitney U test and independent t test were used for nonparametric variables and parametric variables, respectively. Pearson’s chi-square analysis or Fisher’s exact test were used to compare categorical variables, as appropriate. Multivariate analysis was done with those variables returning a p < 0.1 in the univariate analyses. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 21.0 (IBM Corp., Armonk, NY, USA).

3. RESULTS 3.1. Baseline clinical characteristics

Among the 75 patients study population, 60.0% patients (n = 45) had EV and 40.0% patients (n = 30) had no EV. The demographic characteristics of the study population are shown in Table 1. There is no difference in age, sex, etiology, model for end-stage liver disease (MELD) score, Child-Turcotte-Pugh, and serum biochemistry between patients with EV (EV group) and those without EV (non-EV group). Thirty-three of 45 (73.3%) patients had high-risk EV, and nine patients had concomitant gastric varices (GV). The characteristics of varices are shown in Table 2.

Table 1 - Baseline characteristics of patients Total EV No EV Baseline characteristics n = 75 n = 45 n = 30 p Age, median (range), y 62.1 ± 1.3 60.3 ± 1.7 64.8 ± 2.0 0.101 Sex (male), n (%) 49 (65.3) 29 (64.4) 20 (66.7) 0.843 Etiology, n (%)  HBV 27 (36.0) 15 (33.3) 12 (40.0) 0.556  HCV 17 (22.7) 12 (26.7) 5 (16.7) 0.311  Alcohol 8 (10.7) 5 (11.1) 3 (10.0) 0.879  Othersa 23 (30.7) 13 (28.9) 10 (33.3) 0.683 Child-Turcotte-Pugh classification, n (%)  A 24 (32.0) 12 (26.7) 12 (40.0) 0.225  B 37 (49.3) 22 (48.9) 15 (50.0) 0.925  C 14 (18.7) 11 (24.4) 3 (10.0) 0.116 MELD score 15 (11–19) 15 (12–19) 15 (10–16) 0.219 HCC, n (%) 27 (36.0) 17 (37.8) 10 (33.3) 0.694 Other malignancy, n (%) 7 (9.3) 5 (11.1) 2 (6.67) 0.517 Chronic kidney disease, n (%) 7 (9.3) 3 (6.7) 4 (13.3) 0.331  End-stage renal disease, n 0 0 0 n/a Using of antithrombotic agent, n (%)b 8 (10.7) 4 (8.9) 4 (13.3) 0.541 Serum lab data  Prothrombin time, INR 1.21 (1.06–1.29) 1.22 (1.17–1.28) 1.20 (1.05–1.21) 0.075  Platelet count, K/cumm 89 (70–153) 83 (69–127) 120 (78–165) 0.055  Creatinine, mg/dL 0.90 (0.76–1.20) 0.85 (0.73–1.13) 0.99 (0.76–1.26) 0.747  ALT, U/L 49 (30–145) 51 (29–143) 47 (30–165) 0.893  AST, U/L 70 (42–128) 71 (43–131) 65 (39–116) 0.395  Albumin, g/dL 3.4 (2.8–3.7) 3.3 (2.9–3.7) 3.4 (2.8–3.7) 0.894  Total bilirubin, mg/dL 2.8 (1.4–7.7) 3.4 (1.5–11.0) 2.2 (1.2–5.0) 0.057

ALT = alanine transaminase; AST = aspartate transaminase; EV = esophageal varices; HBV = hepatitis B virus; HCV = hepatitis C virus; HCC = hepatocellular carcinoma; MELD = model for end-stage liver disease.

aOthers etiology including three patients had autoimmune hepatitis, three patients had concomitant HBV infection and alcoholism, two patients had concomitant HBV and HCV infection, three patients had nonalcoholic steatosis hepatitis, three patients had primary biliary cholangitis, nine patients had cryptogenic cirrhosis.

bSeven patients using antiplatelet (four clopidogrel and three aspirin) and one patient using anticoagulant of warfarin.

Table 2 - Characteristic and prophylaxis of varices n = 45 EV size, n (%)a  Trace 3 (6.7)  F1 17 (37.8)  F2 17 (37.8)  F3 8 (17.8) Red-color signs, n (%) 18 (40.0) High-risk EV, n (%) 33 (73.3) Concomitant GV, n (%) 9 (20.0) Primary prevention, n (%)  EVL 16 (35.6)  NSBB 2 (4.4) Secondary prevention, n (%)  EVL + NSBB 9 (20.0)  EVL alone 3 (6.7)

EV = esophageal varices; EVL = endoscopic variceal ligation; GV = gastric varices; NSBB = nonselective beta-blocker.

aEV size defined as largest size of EV.

3.2. Indications and findings during the procedure

Indications of ERCP were choledocholithiasis (30/75, 40.0%), benign bile duct stricture (17/75, 22.7%), malignant tumor compression (9/75, 12.0%), cholecystitis (7/75, 9.3%), pancreatitis (6/75, 8.0%), and undetermined reasons (6/75, 8.0%). Successful cannulation was achieved in 72 (96.0%) patients, including 43 patients with EV and 29 patients without EV. Dilated common bile duct (CBD) was found in 17 patients (37.8%) with EV and 11 patients (36.7%) without EV. A negative ERCP result was detected in two patients (4.4%) with EV and one patient (3.3%) without EV. The number of patients undergoing therapeutic ERCP was also no different. About half of the patients received endoscopic balloon dilatation or endoscopic sphincterotomy. The details of therapeutic ERCP are shown in Table 3.

Table 3 - Therapeutic interventions during endoscopic retrograde cholangiopancreatography EV No EV n = 45 n = 30 p Therapeutic interventions, n (%)a  EST 7 (15.6) 8 (26.7) 0.239  EPBD 13 (28.9) 7 (23.3) 0.594  ERBD 11 (24.4) 8 (26.7) 0.828   Plastic stent 6 (13.3) 5 (16.7) 0.689   Metallic stent 5 (11.1) 3 (10.7) 0.879  NBD 4 (8.9) 0 (0) 0.093

EPBD = endoscopic papillary balloon dilation; ERBD = endoscopic retrograde biliary drainage; ERCP = endoscopic retrograde cholangiopancreatography; EST = endoscopic sphincterotomy; NBD = nasobiliary drainage.

aOne patient received EPBD and ERBD with metallic stent placement at the same time.

3.3. Bleeding related to the procedure

The risk of overall GI bleeding (5/45, 11.1% vs 3/30, 10.0%, p = 0.879) and significant GI bleeding (3/45, 6.7% vs 1/30, 3.3%, p = 0.529) was not different between patients with and without EV. However, a patient with EV experienced melena 2 days after endoscopic sphincterotomy and partial cover metallic stenting for distal CBD stricture (Fig. 1). Endoscopy disclosed gastric variceal bleeding at the posterior wall of the high body near the cardia (Fig. 2). He underwent endoscopic cyanoacrylate injection with N-butyl-2-cyanoacrylate for acute gastric variceal bleeding and endoscopic variceal ligation for EV. The patient was discharged 5 days later without further bleeding. Another patient experienced significant duodenal ulcer bleeding 7 days after ERCP. Endoscopic hemostasis was performed, which revealed a 0.8-cm Forrest class Ib ulcer located at the second portion of the duodenum and hemostasis was successfully achieved using epinephrine injection plus thermal therapy. The duodenal ulcer was not seen in the previous ERCP examination, and the patient did not use prophylaxis NSAID before and after ERCP. The details of the bleeding are shown in Table 4.

Table 4 - Characteristic of gastrointestinal bleeding EV No EV n = 45 n = 30 p Overall GI bleeding 5 (11.1) 3 (10.0) 0.879  Variceal bleeding 1 (2.2) 0 (0) 0.411  Duodenal ulcer bleeding 1 (2.2) 0 (0) 0.411  Gastric Dieulafoy’s lesion bleeding 1 (2.2) 0 (0) 0.411  Undetermineda 2 (4.4) 3 (10.0) 0.345 Clinically significant GI bleeding 3 (6.7) 1 (3.3) 0.529  Duodenal ulcer bleeding 1 (2.2) 0 (0) 0.411  Gastric Dieulafoy’s lesion bleeding 1 (2.2) 0 (0) 0.411  Undetermined 1 (2.2) 1 (3.3) 0.770 Hb dropped, median (range), g/dL 1.8 (0.6–2.6) 2.3 (2.2–2.4) 0.836

GI bleeding = gastrointestinal bleeding; Hb = hemoglobin.

aOne patient received upper and lower GI endoscopy and no bleeding source was noted in the EV group. Two patients in the no EV group and one patient in the EV group did not receive endoscopy for bleeding survey and improved after pharmacological therapy. Another patient in the no EV group did not perform endoscopy for bleeding survey and improved after blood transfusion and pharmacological therapy.

Fig. 1:

Esophagogastroduodenoscopy findings. A, Tumor-like gastric varices at fundus and posterior wall of high body. B, A suspicious active oozing site at gastric varices (black arrow).

Fig. 2:

Endoscopic retrograde cholangiopancreatography findings. A, A 4-cm narrowing at distal common bile duct in endoscopic retrograde cholangiography (ERC). B, A 6-cm partially covered metallic stent was deployed in ERC.

3.4. Other adverse events related to the procedure

Adverse events of ERCP were comparable between patients with EV and those without EV, regardless of counting by events (14 vs 11, p = 0.659) or patients (26.7% vs 36.7%, p = 0.358). The details of adverse events are shown in Table 5. There was no procedure-related perforation or mortality during the study period.

Table 5 - Adverse events related to the procedure EV No EV n = 45 n = 30 p Patients with adverse events, n (%) 12 (26.7) 11 (36.7) 0.358 Numbers of adverse events, total n 14 11 0.659 Characteristics of adverse events, n (%)  Overall GI bleeding 5 (11.1) 3 (10.0) 0.879  Sepsis 6 (13.3) 5 (10.0) 0.689  Post-ERCP pancreatitis 3 (6.7) 3 (10.0) 0.602

ERCP = endoscopic retrograde cholangiopancreatography; EV = esophageal varices; GI bleeding = gastrointestinal bleeding.

3.5. Predictors of complications

On univariate and multivariate analyses, no indicator (including cirrhosis status, MELD score, age, sex, and hepatocellular carcinoma) could be identified to determine the adverse events (Supplementary Table 1, https://links.lww.com/JCMA/A158).

4. DISCUSSION

With regard to thrombocytopenia, coagulopathy, and potential comorbidity in patients with cirrhosis, these are major concerns for the occurrence of complications in the patients undergoing invasive procedures, particularly when ERCP is performed for patients with EV. In the literature, there are limited data regarding the adverse events in patients with concomitant cirrhosis and EV undergoing ERCP. Two previous retrospective cohort studies without special emphasis on EV demonstrated similar overall adverse events between cirrhotic and noncirrhotic patients.8,9 Subgroup analysis showed an increased risk of adverse events only in patients with decompensated cirrhosis.8,9 Nevertheless, another two studies showed increased overall adverse events in cirrhotic patients compared with noncirrhotic patients.10,12 Therefore, the risk of an adverse event in patients with cirrhosis is controversial. Moreover, no study has focused on the influence of the existence of EVs. The serious concern of bleeding risk in patients with EV has never been well analyzed.

The lack of consistency in the definition of post-ERCP bleeding in cirrhotic patients might explain the wide range of bleeding rates in previous studies: from 1.1% to 13.5%.8–12 In this study, the overall bleeding risk was 10.7% in the whole population, and the clinically significant GI bleeding risk was 5.3%, comparable with the previous study.12 Again, details of EV, including the case number of patients with EV and high-risk EV, were not revealed in previous studies.8–12 We found no esophageal variceal bleeding during and after diagnostic or therapeutic ERCP, even in those with high-risk EV. Interestingly, one patient with concomitant gastro-EV experienced gastric variceal bleeding 2 days after the procedure. It is possible that the side-view duodenoscope caused the unintended traumatic injury of gastric varix. Luckily, current treatment standards successfully controlled the acute bleeding, including antibiotic, vasoactive agent, and endoscopic cyanoacrylate injection with N-butyl-2-cyanoacrylate. Taken together, diagnostic or therapeutic ERCP is generally safe in patients with cirrhosis and EV.

According to five previous studies, the reported rates of post-ERCP pancreatitis in cirrhotic patients were from 4.4% to 8.3%. Only one previous study documented no prophylactic pharmacological therapy was used in that study.12 Other prophylactic therapies, such as pancreatic duct stenting, were not well demonstrated in previous studies.8–12 In this study, three patients received a prophylactic rectal nonsteroidal anti-inflammatory drug (two in the EV group and one in the no EV group), and one in the EV group received prophylactic pancreatic duct stenting for post-ERCP pancreatitis prophylaxis. Despite a small number of patients receiving prophylactic therapy for post-ERCP pancreatitis, the incidence rate of post-ERCP pancreatitis was 8.0% (6.7% in the EV group and 10.0% in the no EV group) in this study which is still comparable to previous studies.

No predictor of adverse events was consistently identified according to previous studies. Two studies reported decompensated cirrhosis as a risk factor,9,10 but this was not confirmed in the third study.12 In our study, decompensated cirrhosis was not a predictor of adverse events. This might be explained by the heterogenicity of populations, operator-dependable technique, and different definitions of adverse events.

This study has some limitations. First, selection bias cannot be prevented because of the study’s retrospective nature. However, we believe the bleeding risk is very low, even if more than 70% of high-risk varices were included in this study. Second, the case number of this study is the largest among studies focused on varices. Nevertheless, this was not large enough to uncover the details of risk factors.

In conclusion, there was no difference in overall adverse events and GI bleeding between cirrhotic patients with and without EV receiving ERCP. Although ERCP is generally safe and esophageal variceal bleeding is rare even in patients with high-risk EV, particular attention should be paid to unintended injury of GV caused by side-view duodenoscope.

ACKNOWLEDGMENTS

The authors declared this study supported by the Taipei Veterans General Hospital Research Program (V110C-101) and the Ministry of Science and Technology Project (MOST109-2314-B-075-096). The authors declared that there is no conflict of interest.

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data related to this article can be found at https://links.lww.com/JCMA/A158.

REFERENCES 1. McCune WS, Shorb PE, Moscovitz H. Endoscopic cannulation of the ampulla of vater: a preliminary report. Ann Surg. 1968;167:752–6. 2. Chandrasekhara V, Khashab MA, Muthusamy VR, Acosta RD, Agrawal D, Bruining DH, et al. Adverse events associated with ERCP. Gastrointest Endosc. 2017;85:32–47. 3. Vandervoort J, Soetikno RM, Tham TC, Wong RC, Ferrari AP Jr, Montes H, et al. Risk factors for complications after performance of ERCP. Gastrointest Endosc. 2002;56:652–6. 4. European Association for the Study of the Liver. Electronic address: [email protected]; European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of patients with decompensated cirrhosis. J Hepatol. 2018;69:406–60. 5. North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices. N Engl J Med. 1988;319:983–9. 6. D’Amico G, Pasta L, Morabito A, D’Amico M, Caltagirone M, Malizia G, et al. Competing risks and prognostic stages of cirrhosis: a 25-year inception cohort study of 494 patients. Aliment Pharmacol Ther. 2014;39:1180–93. 7. Augustin S, Muntaner L, Altamirano JT, González A, Saperas E, Dot J, et al. Predicting early mortality after acute variceal hemorrhage based on classification and regression tree analysis. Clin Gastroenterol Hepatol. 2009;7:1347–54. 8. Macías-Rodríguez RU, Ruiz-Margáin A, Rodriguez-Garcia JL, Zepeda-Gómez S, Torre A. Risk factors associated with complications in cirrhotic patients undergoing endoscopic retrograde cholangio-pancreatography. Eur J Gastroenterol Hepatol. 2017;29:238–43. 9. Adler DG, Haseeb A, Francis G, Kistler CA, Kaplan J, Ghumman SS, et al. Efficacy and safety of therapeutic ERCP in patients with cirrhosis: a large multicenter study. Gastrointest Endosc. 2016;83:353–9. 10. Inamdar S, Berzin TM, Berkowitz J, Sejpal DV, Sawhney MS, Chutanni R, et al. Decompensated cirrhosis may be a risk factor for adverse events in endoscopic retrograde cholangiopancreatography. Liver Int. 2016;36:1457–63. 11. Solanki S, Kichloo A, Dahiya DS, Solanki D, Singh J, Wani F, et al. Endoscopic retrograde cholangiopancreatography (ERCP) in patients with liver cirrhosis: analysis of trends and outcomes from the national inpatient sample database. J Clin Gastroenterol. 2022;56:618–26. 12. Leal C, Prado V, Colan J, Chavez-Rivera K, Sendino O, Blasi A, et al. Adverse events and acute chronic liver failure in patients with cirrhosis undergoing endoscopic retrograde cholangiopancreatography: a multicenter matched-cohort study. Am J Gastroenterol. 2019;114:89–97. 13. Freeman ML, Nelson DB, Sherman S, Haber GB, Herman ME, Dorsher PJ, et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med. 1996;335:909–18. 14. D’Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol. 2006;44:217–31. 15. Beppu K, Inokuchi K, Koyanagi N, Nakayama S, Sakata H, Kitano S, et al. Prediction of variceal hemorrhage by esophageal endoscopy. Gastrointest Endosc. 1981;27:213–8. 16. Garcia-Tsao G, Abraldes JG, Berzigotti A, Bosch J. Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the Study of Liver Diseases. Hepatology. 2017;65:310–35. 17. Schulman S, Kearon C; Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005;3:692–4.