Impact of chronic proliferative cholangitis on stone recurrence and biliary strictures after intrahepatic stone treatment: A meta-analysis

Table 1 General characteristics of included observational studies (six cohorts, hepatolithiasis management and outcomes

Ref.	Country	Journal	Study design	Sample size	Follow-up duration	CPC/proxy reported?
Jin et al[12], 2016	China	Surgery	Retrospective cohort (LLH vs OLH)	201	NR (comparative outcomes reported)	No - not reported
Tsutsumi et al[13], 2017	Japan	Ther Adv Gastroenterol	Retrospective cohort (PTCS vs sDBE after HJ)	40	Median 7.2 years (PTCS)/3.1 years (sDBE)	Yes - post-HJ stricture/CPC-proxy
Ishihara et al[14], 2021	Japan	Surg Endosc	Retrospective cohort (balloon-assisted endoscopy after HJ)	90	Median 30.5 months (IQR 14.7-55.5)	Yes - recurrent cholangitis and anastomotic narrowing
Torres et al[15], 2021	Brazil (6 centers)	Asian J Surg	Multicenter retrospective cohort (hepatectomy)	127	NR (recurrence assessed)	No - no CPC label; surgical clearance focus
Yan et al[16], 2024	China	J Evid Based Med	Retrospective cohort (PTCS in decompensated cirrhosis)	21	Median 18 months (up to 40)	Yes - recurrent cholangitis (post-procedure)
Hakuta et al[17], 2024	Japan	Surg Endosc	Retrospective cohort (BE-ERCP after HJS)	131	1-year, 3-year, and 5-year recurrence estimates reported	Yes - HJS anatomy with CPC-related stricture

LLH: Laparoscopic left hepatectomy; OLH: Open left hepatectomy; HJ: Hepaticojejunostomy; PTCS: Percutaneous transhepatic cholangioscopy; HJS: Hepaticojejunostomy stricture; BE: Balloon-enteroscopy; ERCP: Endoscopic retrograde cholangiopancreatography; NR: Not reported; sDBE: Short double-balloon enteroscopy; IQR: Interquartile range; CPC: Chronic proliferative cholangitis.

Table 2 Baseline characteristics of study populations

Ref.	Age (mean)	Sex distribution	Key comorbidities	Intervention/exposure	Comparator/control
Jin et al[12], 2016	NR	NR	Primary hepatolithiasis	LLH	OLH
Tsutsumi et al[13], 2017	NR	NR	Prior HJ; bilioenteric anastomotic stricture common	PTCS	sDBE
Ishihara et al[14], 2021	NR	NR	Prior HJ; stone burden and strictures modeled	Balloon-assisted endoscopy	None (single-arm outcomes)
Torres et al[15], 2021	46.1	Male 40.9%	Cholangitis 51.9%; atrophy 24.4%; biliary stenosis 14.1%	Hepatectomy (various)	None (single-arm outcomes)
Yan et al[16], 2024	NR (MELD reported)	NR	Decompensated cirrhosis; transplant-level MELD at baseline	PTCS (one- or two-step)	None (single-arm outcomes)
Hakuta et al[17], 2024	NR	NR	Prior HJS; malignant or congenital indications	BE-ERCP	None (single-arm outcomes)

Sex = male proportion reported if available. NR: Not reported; MELD: Model for End-Stage Liver Disease; HJ: Hepaticojejunostomy; HJS: Hepaticojejunostomy stricture; LLH: Laparoscopic left hepatectomy; PTCS: Percutaneous transhepatic cholangioscopy; BE: Balloon-enteroscopy; ERCP: Endoscopic retrograde cholangiopancreatography; OLH: Open left hepatectomy; sDBE: Short double-balloon enteroscopy.

Table 3 Endpoints and outcomes of individual studies (author-reported statistics)

Ref.	Endpoints assessed	Outcome summary
Jin et al[12], 2016	Stone clearance, recurrence, complications, transfusion, blood loss, LOS	LLH (n = 96) vs OLH (n = 105) showed lower blood loss (383 ± 281 mL vs 554 ± 517 mL; P = 0.005) and lower transfusion (8.3% vs 30.5%; P < 0.001). No significant differences were seen in operation time, hospital stay, postoperative complication rate, residual stone rate, or recurrent stone rate. Findings suggest laparoscopic resection can match open surgery in clearance and recurrence, with perioperative benefits. Specific recurrence percentages not stated in abstract. CPC not reported; likely CPC-negative surgical cohort
Tsutsumi et al[13], 2017	Technical/clinical success, adverse events, LOS, stone-free survival	Initial access success 100% (PTCS) vs 91% (sDBE). All 40 patients achieved clinical success after crossover when needed. Adverse events were lower with sDBE (10%) than PTCS (45%; P = 0.025). Median LOS for complete clearance was shorter with sDBE (10 days vs 35 days; P < 0.001). Stone-free probabilities were similar at 1-3 years (P = 0.919). Results favor sDBE on safety and LOS while maintaining comparable long-term stone control. CPC-enriched post-HJ cohort with anastomotic stricture
Ishihara et al[14], 2021	Complete removal, recurrence, adverse events, re-intervention, multivariable risk factors	In 90 HJ patients, complete removal and recurrence were tracked over a median of 30.5 months (IQR 14.7-55.5). Endoscopic adverse events were reported. Multivariable modeling identified factors associated with recurrence (e.g., stricture/stone burden as per abstract). The study provides time-to-event recurrence and standardized endoscopic adverse events, enabling comparison with other BE-ERCP series. Detailed percentages for clearance and AEs are summarized in the text and tables of the article. CPC-enriched cohort with recurrent cholangitis and anastomotic narrowing
Torres et al[15], 2021	Stone-free status, recurrence, complications, mortality, procedures	Among 127 resections, postoperative complications occurred in 33.0% (biliary fistula 13.3%; SSI 7.0%). Mortality was 0.7%. Recurrence was 7.8% overall, with higher risk in bilateral stones and/or with hepaticojejunostomy. Stone location and atrophy were detailed. The series demonstrates durable stone control after resection with low mortality but notable morbidity. Mostly CPC-negative resection cohort with definitive duct clearance
Yan et al[16], 2024	Technical success, stone clearance, MELD change, recurrence, survival	In 21 decompensated cirrhosis patients, technical success was 100%. Most stones were cleared (90.48%). MELD decreased at 3 months (10.81 ± 3.31 vs 17.24 ± 3.40; P < 0.05) and was lowest at 6 months (9.94 ± 4.31). Median follow-up was 18 months; recurrence was 28.57% (6/21). Thirteen patients survived without transplant; five died due to liver failure or cancer. CPC-proxy positive (cohort with recurrent cholangitis in cirrhosis)
Hakuta et al[17], 2024	Scope/stone removal success, adverse events, recurrence (1/3/5-year), risk factors	In 131 patients after HJS, scope insertion succeeded in 89% and complete stone removal in 73%. Early adverse events occurred in 9.9%. Recurrence after complete removal was 17% at 1 year, 20% at 3 years, and 31% at 5 years. Past HJS (> 10 years) predicted failed stone removal (OR = 10.4, 95%CI: 2.99-36.5). CPC-enriched post-HJS anatomy with stricture predisposition
Hakuta et al[17], 2024	Scope/stone removal success, adverse events, recurrence (1/3/5-year), risk factors	In 131 patients after HJS, scope insertion succeeded in 89% and complete stone removal in 73%. Early adverse events occurred in 9.9%. Recurrence after complete removal was 17% at 1 year, 20% at 3 years, and 31% at 5 years. Past HJS (> 10 years) predicted failed stone removal (OR = 10.4, 95%CI: 2.99-36.5). CPC-enriched post-HJS anatomy with stricture predisposition

Endpoints assessed and principal findings in each included observational study. LOS: Length of hospital stay; MELD: Model for End-Stage Liver Disease; LLH: Laparoscopic left hepatectomy; OLH: Open left hepatectomy; CPC: Chronic proliferative cholangitis; PTCS: Percutaneous transhepatic cholangioscopy; sDBE: Short double-balloon enteroscopy; HJ: Hepaticojejunostomy; IQR: Interquartile range; BE: Balloon-enteroscopy; ERCP: Endoscopic retrograde cholangiopancreatography; AE: Adverse event; SSI: Surgical site infection; HJS: Hepaticojejunostomy stricture; OR: Odds ratio; CI: Confidence interval.

Table 4 Comparative outcomes and risk of bias across studies

Ref.	Effect size	95%CI	P value	Secondary outcomes	Subgroup analyses/notable findings	Risk of bias	Notes
Jin et al[12], 2016	NR (no OR/RR for recurrence in abstract)			Lower blood loss (LLH vs OLH)2; lower transfusion1	No significant difference in recurrence/complications between LLH and OLH	Some concern	Benefit in perioperative metrics; recurrence not different
Tsutsumi et al[13], 2017	AE difference (sDBE 10% vs PTCS 45%)1		0.025	LOS shorter with sDBE2 (10 days vs 35 days; P < 0.001)	Stone-free survival similar (P = 0.919)	Some concern	Direct method comparison after HJ; crossover allowed
Ishihara et al[14], 2021	NR (author multivariable HRs for recurrence reported in text)			Clearance, AEs, re-intervention reported	Multivariable model for recurrence risk	Some concern	Time-to-event recurrence over approximately 30 months
Torres et al[15], 2021	Recurrence proportion 7.8%			Complications 33.0%; mortality 0.7%	Higher recurrence in bilateral stones/HJ	Some concern	Multicenter non-Asian surgery cohort
Yan et al[16], 2024	Clearance 90.48%; recurrence 28.57%		-	MELD improvement2 (P < 0.05); survival described	Decompensated cirrhosis; transplant eligibility	High risk (small n, severity)	High-risk population; detailed follow-up
Hakuta et al[17], 2024	OR 10.4 (failed removal in past HJS)	2.99-36.5		AEs 9.9%; recurrence 17%, 20%, 31% at 1/3/5-year	Past HJS (> 10 years) drives technical failure	Some concern	Large HJS cohort; Kaplan-Meier recurrence

¹Categorical variable compared using χ² or Fisher’s exact test.

²Continuous variable compared using t-test or Mann-Whitney U test as reported by authors.

NR: Not reported; OR: Odds ratio; RR: Risk ratio; AE: Adverse event; sDBE: Short double-balloon enteroscopy; PTCS: Percutaneous transhepatic cholangioscopy; HR: Hazard ratio; HJS: Hepaticojejunostomy stricture; CI: Confidence interval; LLH: Laparoscopic left hepatectomy; OLH: Open left hepatectomy; LOS: Length of hospital stay; MELD: Model for End-Stage Liver Disease; HJ: Hepaticojejunostomy.

Table 5 Newcastle-Ottawa Scale quality assessment of included observational studies

Ref.	Selection (Max 4)	Comparability (Max 2)	Outcome (Max 3)	Total NOS Score (Max 9)	Quality interpretation	Comments
Jin et al[12], 2016	4	1	2	7	Moderate quality	Retrospective surgical cohort; good case selection; limited adjustment for confounders
Tsutsumi et al[13], 2017	3	1	3	7	Moderate quality	Clear outcome assessment and follow-up; limited comparability between treatment approaches
Ishihara et al[14], 2021	3	2	3	8	High quality	Adequate follow-up; multivariable analysis used to assess recurrence risk
Torres et al[15], 2021	4	1	2	7	Moderate quality	Multicenter cohort; good selection; limited control for confounding
Yan et al[16], 2024	3	1	2	6	Moderate quality	Small sample size; high-risk population; outcome reporting adequate
Hakuta et al[17], 2024	4	2	3	9	High quality	Large cohort; clear outcome definitions; long-term follow-up and risk factor analysis

Max: Maximum score; NOS: Newcastle-Ottawa Scale.