Optimal Timing of Single-Stage Retrograde Endoscopic Common Bile Duct Stone Removal in Mild and Moderate Acute Cholangitis: A Prospective Trial


 Background: In this study, we aimed to compare the efficacy and safety of removing a single-stage, retrograde, endoscopic common bile duct stone in patients with mild and moderate acute cholangitis associated with choledocholithiasis.Methods: We enrolled 196 endoscopic retrograde cholangiopancreatography (ERCP)-naïve patients diagnosed with acute cholangitis and choledocholithiasis between September 2018 and February 2020 at a single hospital. For eligible patients, single-stage treatment involved stone removal at initial ERCP. Early ERCP was defined as ERCP performed ≤ 72 hours following diagnosis in the emergency room.Results: The final analysis included 138 patients. The success rate of complete stone extraction was similar in patients with mild and moderate cholangitis (88.5% vs. 91.7%; p = 0. 536). Complication rates were also comparable between the two groups. In the moderate cholangitis group, the length of hospitalization declined significantly among patients who underwent early single-stage ERCP (10.6 ± 6.1 vs. 18.7 ± 12.5 days; p = 0.001) compared with patients treated with delayed ERCP. In the multivariate analysis, early ERCP indicated shorter hospitalization times (≤ 10 days) (odds ratio (OR), 7.689; p = 0.030), while endoscopic retrograde biliary drainage, for acute cholangitis only, indicated longer hospitalization times (OR, 0.358; p = 0.030). A stone size larger than 1.5 cm was an independent risk factor for stone extraction failure (OR, 24.507; p = 0.009).Conclusions: Single-stage, retrograde, endoscopic common bile duct stone removal may be safe and effective for patients with mild and moderate cholangitis. The benefit of early single-stage ERCP (≤ 72 hours) was reflected mainly by reduced hospitalization time and costs.Trial registration: ClinicalTrials.gov: NCT03754491.

Further, this treatment averted the need for a second session of ERCP, thus reducing medical expenses (7)(8)(9)(10). Therefore, the 2018 Tokyo guidelines recommend simultaneous CBD stone treatment and biliary drainage in patients with mild to moderate choledocholithiasis (11). However, evidence supporting the feasibility of single-stage stone removal in patients with moderate acute cholangitis remains insu cient.
Whether or not early stone removal can reduce the bacterial load, thereby enabling better infection control and further reducing the length of hospitalization, remains unclear. Moreover, the optimal common bile stone size suitable for removal during single-stage ERCP in mild and moderate acute cholangitis is unknown. Therefore, we conducted this prospective trial to evaluate the e cacy and safety of singlestage, retrograde, endoscopic CBD stone removal in patients with mild and moderate acute cholangitis associated with choledocholithiasis. papilla not found (n = 4), and tumor-related obstruction (n = 3). We gathered written informed consent from all included patients before the trial. Before ERCP, we recorded the following demographic and clinical variables: age, sex, history of coexisting comorbidities, alcohol consumption, smoking habits, American Society of Anesthesiologists (ASA) score, and serum levels of albumin, C-reactive protein (CRP), total bilirubin, prothrombin time (PT), activated partial thromboplastin time (APTT), and liver function enzymes, as well as a complete and differential blood counts obtained in the emergency room (ER) before and after ERCP for evaluation of complications. Additionally, endoscopic ndings were recorded, including papilla type (12), juxtapapillary diverticulum, CBD stone size and number, and procedure methods used.

Patients and assessments
As part of the single-stage treatment, we performed stone removal during the initial ERCP session. Early ERCP was de ned as ERCP occurring 72 hours or less following diagnosis in the ER (5). Primary outcomes were as follows: 1) operation time during ERCP, de ned as the period ranging from the beginning of cannulation to complete stone removal; 2) success rate of complete bile duct stone removal, 3) signi cant complications, including post-ERCP pancreatitis (amylase levels higher than three times the upper reference limit accompanied by abdominal pain), perforation, and bleeding; 4) length of hospitalization, and 5) costs. Secondary outcomes included the development of pneumonia within 30 days and mortality within 30 days after ERCP. Bleeding complications could present as melena or hematemesis, with a decrease in hemoglobin concentration of at least 2 g/dL. The de nition of bleeding degree for patients who did not require transfusion was 'mild bleeding degree.' Cases requiring up to four units of blood were de ned as 'moderate bleeding degree,' and those requiring ve or more units of blood for transfusion, surgery, or angiography were de ned as 'severe bleeding degree' (13).

Stages of acute cholangitis
The stage classi cation of acute cholangitis was based on the 2013 and 2018 Tokyo guidelines (6,11), with a sensitivity of 91.8% and a speci city of 77.7% (14). Grade II (moderate) acute cholangitis was that associated with any two of the following conditions: 1) abnormal white blood count (WBC) count (> 12,000/mm 3 or < 4,000/mm 3 ), 2) high fever (≥ 39 °C), 3) age (≥ 75 years), 4) hyperbilirubinemia (total bilirubin ≥ 5 mg/dL), and 5) hypoalbuminemia (< SD × 0.7, where standard deviation (SD) is the lower limit of the standard value). Patients with Grade III (severe) acute cholangitis exhibited dysfunction of at least one of the following organs/systems: 1) cardiovascular system, suggested by hypotension, requiring a dopamine infusion of 5 µg/kg/min or higher or any dose of norepinephrine; 2) nervous system, indicated by a disturbance in consciousness; 3) respiratory system, with a PaO 2 /FiO 2 ratio > 300; 4) renal system, with oliguria and an s-creatinine level > 2 mg/dL; 5) hepatic system, with a platelet international normalized ratio > 1.5; and 6) hematological system, with a platelet count < 100,000/mm 3 . Finally, Grade I acute cholangitis was de ned as acute cholangitis, not ful lling the criteria of either Grade III (severe) or Grade II (moderate) acute cholangitis at the time of the initial diagnosis. The balloon was in ated with saline solution to reach 8 to 20 mm diameter and to dilate the papilla with a progressive increase in pressure from 3 to 8 atmospheres for three minutes depending on CBD stone size (15). Endoscopic sphincterotomy (EST) was performed using standard pull-sphincterotomes (ENDO-FLEX GmbH, Voerde, Germany). In cases where biliary cannulation was di cult, we performed limited precut EST or stulotomy (Needle Knife, pointed type, ENDO-FLEX GmbH, Voerde, Germany) or a transpancreatic EST (16). A pancreas duct stent was placed for preventing post-ERCP pancreatitis (PEP) if the pancreas duct cannulation occurred twice or more. At the same time, 100 mg of indomethacin was administered anally to all patients who did not have a history of allergy (17). Aggressive intravenous hydration (including 3 mL/kg/h during ERCP, a 20-mL/kg bolus, and 3 mL/kg/h for eight hours after ERCP) with lactated Ringer's solution was administered to all patients without contraindications (18). CBD stones were extracted using a balloon and/or basket catheter. A retrograde biliary drain with a plastic stent was inserted if CBD stone extraction could not be performed within one hour of the procedure, if the contrast medium bile ow was poor with papilla swelling after stone extraction, or if pus bile was noted. All patients underwent empiric antimicrobial treatment for acute cholangitis. All patients were asked to fast for at least 12 hours after ERCP and received intravenous proton-pump inhibitors (PPIs), corresponding to two doses after ERCP, and oral PPIs once daily for seven days.

Statistical analysis
Descriptive statistics, including distributions, absolute frequencies, relative frequencies, medians with ranges, and means ± standard deviations (SDs) were calculated depending on the variable type. Betweengroup differences for quantitative variables with normal distribution were compared using Student's ttest. The differences between categorical data proportions were evaluated with Fisher's exact test when there were fewer than ve expected cases; otherwise, we used the chi-square test. We included factors with probability (p) values < 0.3 in the univariate analysis in the logistic regression analysis. A multivariate logistic regression model was adopted to identify independent factors of procedural success and major adverse events. A p value < 0.05 was considered to indicate statistical signi cance in all analyses.

Population characteristics
The study included 138 patients who ful lled the inclusion criteria, including 78 and 60 patients in the mild and moderate cholangitis groups, respectively ( Fig. 1). The were no differences in sex, personal habits (i.e., alcohol use and smoking), white and platelet blood counts, APTT, levels of alanine transaminase, albumin, bilirubin, and alkaline phosphatase; estimated glomerular ltration rate (eGFR), papilla type, ratio of the juxtapapillary diverticulum, endoscopic papillary balloon dilation (EPBD), and EST between the two groups. Meanwhile, age, comorbidity rate (i.e., diabetes and hypertension), ASA score, initial body temperature (°C) in the ER, PT, and CRP were higher in the moderate cholangitis group than in the mild cholangitis group (Table 1). Stone size (0.96 ± 0.45 vs. 1.17 ± 0.55 cm; p = 0.128), mean CBD diameter (1.31 ± 0.40 vs. 1.50 ± 0.39 cm; p = 0.897), frequency of endoscopic retrograde biliary drainage (ERBD) use (20.5% vs. 33.3%; p = 0.089), and procedure time (24.3 ± 11.1 vs. 24.8 ± 11.3 min; p = 0. 929) did not differ between the two groups. The number of stones was higher in the mild cholangitis group than in the moderate group (1.81 ± 1.22 vs. 1.37 ± 1.00 stones; p = 0.012) ( Table 2).   Table 3. There were no fatalities in any of the two groups. One patient experienced coffee-ground vomiting and a mild degree of bleeding after EPBD. Five patients with a mild degree of PEP were treated with conservative care, with symptoms spontaneously resolving.   Stone size during single-stage stone extraction by ERCP In 14 cases, we were unable to extract stones. Eight cases were due to di cult cannulation, even with precut EST or stulotomy use, while six cases were a simple failure of stone extraction. We conducted univariate and multivariate analyses of the factors associated with CBD stone extraction failure. After excluding cannulation failure cases, stone sizes greater than 1.5 cm was an independent factor suggesting CBD stone extraction failure (OR, 24.507; 95% CI, 2.186-274.708; p = 0.009) ( Table 6). In the subgroup analysis, the rate of complete CBD stone removal varied according to CBD stone size (i.e., ≤ 1.5 cm or > 1.5 cm). Thus, after excluding cannulation failure cases, there were higher success rates (98.6% and 100%, respectively) among patients with CBD stones ≤ 1.5 cm than there were among patients with stones > 1.5 cm (81.8% and 70.0%, respectively), regardless of whether ERCP was performed early or not (p = 0.005 and p = 0.001, respectively) (Fig. 3). Factors with p values less than 0.3 were included in the logistic regression analysis. Abbreviations: CBD, common bile duct; OR, odd ratio; CRP, C-reactive protein; ERCP, endoscopic retrograde cholangiopancreatography; EST, endoscopic sphincterotomy; EPBD, endoscopic papillary balloon dilation.

Discussion
To our knowledge, this is the rst prospective trial that aimed to identify suitable ERCP timing and the distribution of CBD stone sizes, as observed during single-stage, retrograde, endoscopic stone removal in patients with acute cholangitis and to compare differences between the mild and moderate stages of acute cholangitis that may affect treatment strategy. The most worrying issue under review in this investigation was whether direct stone extraction in mild to moderate acute cholangitis would increase complications, such as bleeding and pancreatitis. The institutional review board and ethics committee recommended that we adopt standard prevention strategies to reduce injuries, such as guidewire-based selective cannulation, long-term dilatation during EPBD, pancreas duct stenting, indomethacin use, aggressive hydration, and PPI administration.
Endoscopists skilled in ERCP can clear CBD stones successfully after EST in the early stages of mild cholangitis. However, adding EST may introduce a higher risk of unforeseen complications, such as bleeding (4-14.5%) (19,20). Acute cholangitis seems to be an independent risk factor for post-EST bleeding (14,21). Therefore, biliary drainage without sphincterotomy is recommended in patients with severe acute cholangitis (19). In a national, population-based study by Hung et al. (22), EPBD was the preferred method to decrease the risk of post-ERCP hemorrhage, especially in patients with liver cirrhosis or impaired renal function.  (Fig. 2). In the multivariate analysis, early ERCP was an independent factor predicting shorter hospitalization (OR, 7.689; p = 0.030).
Meanwhile, initial PTBD for acute cholangitis did not reduce the length of hospitalization. We suspect that patients with moderate cholangitis who receive early or delayed ERCP have different in ammation severity levels. In the sub-analysis of baseline characteristics of patients with moderate acute cholangitis who underwent early and delayed ERCP, respectively, there was no difference between the two groups in terms of age, renal function, albumin, WBC count, liver function, or bilirubin. However, patients with moderate cholangitis, who underwent early ERCP relative to delayed ERCP (125.5 ± 70.3 vs. 117.6 ± 93.9; p = 0.029), experienced a higher level of CRP. Therefore, it appears reasonable to suggest that the optimal timing of single-stage stone removal in both mild and moderate cholangitis is within 72 h.
On the other hand, there was a negative association between biliary drainage only (ERBD) for acute cholangitis and successful infection control as well as reduced hospital stay (OR: 0.358, p = 0.030) ( Table 5). As shown in other studies addressing acute cholecystitis management, an early approach to emergency cholecystectomy within 72 h of symptom onset reduces operative time, decreases hospitalization length, is associated with fewer adverse postoperative outcomes, and reduces mortality (24)(25)(26). The circumstance might be related to the management of foreign-body infection (27): Removing debris from the site of injury reduces the bacterial load and thereby facilitates control infection. Bactibilia (the presence of bacteria in the biliary tract) increases in the presence of biliary obstruction, mainly partial obstruction, and in the presence of foreign bodies like stones (28). The most common bacteria linked to ascending cholangitis are Escherichia coli (29), Klebsiella (30), Enterobacter (31), and Enterococcus (32), which form a bio lm covering the surfaces of stones. This bio lm protects the bacteria from antibacterial agents and phagocytic leucocytes (33). Therefore, prompt removal of infected stones in cases of acute cholangitis is preferable.
The evaluation of stone size during single-stage removal in acute cholangitis is important. We determined that stones that it was not challenging to remove stones up to 1.5 cm (34), and successful stone removal (98.6-100%) was higher than when the stones were larger than 1.5 cm (success rate, 70.0-81.8%), as long as biliary cannulation was successful (Fig. 3). Thus, we recommend direct removal of stones in patients with mild and moderate cholangitis only if the stone size is 1.5 cm or smaller.
The limitations of the current study need to be acknowledged. First, this study was initially designed as a randomized controlled trial to compare single-stage and two-stage ERCP-based stone removal (biliary drainage rst and bile stone removal one week later) in patients with moderate acute cholangitis.
However, most patients refused to undergo two-stage ERCP because of the need for more than one session. Therefore, we altered the trial design to a prospective trial of single-stage ERCP in patients with mild and moderate cholangitis. The results indicated that treatment by biliary drainage only (ERBD) in acute cholangitis (OR: 0.358, p = 0.030) was negatively associated with a shorter hospital stay. Second, although limited EST plus endoscopic papillary large-balloon dilation to remove large bile duct stones (> 1.5 cm) was associated with a high success rate (98.3%) in our previous study (35), a single-stage treatment for larger stones (> 1.5 cm) in patients with moderate cholangitis might be more complicated.
More research is required to assess the bene ts and risks. De nitive treatment with removal of large stones is still recommended, but only after the patient's general condition becomes stable per the established guidelines.

Conclusions
We conclude that patients with moderate acute cholangitis have more comorbidity and a more severe degree of in ammation, resulting in longer hospitalization and higher hospitalization costs. Single-stage, retrograde, endoscopic CBD stone removal in mild and moderate cholangitis with choledocholithiasis may be safe and effective, especially if the stone is 1.5 cm or smaller, which can obviate the requirement for a second ERCP session, thus reducing medical expenses. In this study, we demonstrated the bene t of single-stage emergency ERCP by shorter hospital stays and reduced costs for patients with mild and moderate cholangitis. However, large-scale, randomized clinical trials are necessary to clarify the safety and e cacy of the single-stage approach, especially when dealing with large CBD stones larger than 1.5 cm. The ow diagram of the prospective enrollment of consecutive patients with mild/moderate acute cholangitis.

Figure 3
Relationship between bile duct stone size and stone removal success rate after excluding cannulation failure cases.