Impact of periampullary diverticulum on the incidence of post-endoscopic retrograde cholangiography pancreatitis

Abstract

BACKGROUND

Periampullary diverticulum (PAD) is a common anatomical variant, but its association with post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) remains controversial. While PAD may alter ampullary anatomy, increasing technical difficulty during ERCP, existing studies report inconsistent findings on its role in PEP pathogenesis. We hypothesize that PAD presence, particularly type B, shows a significant association with PEP development and may interact with procedural factors like pancreatic duct guidewire insertion.

AIM

To examine the association between PAD (including subtypes A/B) and PEP incidence after ERCP for choledocholithiasis.

METHODS

We conducted a retrospective cohort study of 615 patients undergoing ERCP at two tertiary hospitals from 2023 to 2025. Participants were stratified into PAD (n = 183; subtype A = 125, subtype B = 58) and non-PAD (n = 432) groups. The primary outcome was PEP incidence. Multivariable logistic regression adjusted for age, sex, hypertension, diabetes, gallbladder surgery, and guidewire insertion. Statistical significance was set at P < 0.05 (two-tailed).

RESULTS

PAD prevalence was 29.8% (183/615). PEP occurrence was more frequent in PAD patients [15.3% (28/183)] than in non-PAD patients [4.2% (18/432)], odds ratio (OR) = 3.86, 95% confidence interval: 2.03-7.35, P < 0.001. Type B PAD showed a stronger association with PEP than type A (OR = 14.16, 95% confidence interval: 5.84-34.34, P < 0.001). Guidewire pancreatic duct entry was linked to higher PEP odds in PAD patients (adjusted OR = 5.02, P < 0.05). Hypertension also demonstrated an association with PEP in the PAD subgroup (P = 0.012).

CONCLUSION

PAD, particularly type B, is independently associated with PEP after ERCP. Patients with these features, especially those with hypertension or pancreatic duct instrumentation, may benefit from enhanced monitoring and prophylaxis.

Key Words: Periampullary diverticulum; Endoscopic retrograde cholangiopancreatography; Post-endoscopic retrograde cholangiopancreatography pancreatitis; Choledocholithiasis; Risk stratification; Anatomical variation

Core Tip: This study demonstrates that periampullary diverticulum (PAD), particularly type B, shows a significant association with post-endoscopic retrograde cholangiopancreatography pancreatitis (PEP). The findings suggest that PAD subtype evaluation may help identify patients with higher likelihood of PEP development. The results indicate potential clinical value of enhanced postoperative monitoring for patients with specific diverticular characteristics and optimization of endoscopic retrograde cholangiopancreatography techniques. These observations contribute to refining PEP prevention strategies and support consideration of individualized procedural approaches.

INTRODUCTION

Periampullary diverticulum (PAD) is a mucosal outpouching lesion located within 2.5 cm of the major duodenal papilla in the ampulla of Vater region[1]. PAD has gained increasing recognition among endoscopists with the widespread adoption of endoscopic retrograde cholangiopancreatography (ERCP). Current clinical research primarily focuses on PAD’s influence on technical aspects of ERCP, particularly its association with increased cannulation difficulty[2,3]. However, evidence regarding PAD’s potential role in post-ERCP pancreatitis (PEP) development, especially across different PAD subtypes, remains limited. Previous studies have suggested that the spatial relationship between PAD and the duodenal papilla may correlate with PEP risk, though systematic validation of this hypothesis is lacking[4-6]. The study of Boix et al[5], which categorizes PAD based on anatomical relationship with the papilla, has limited clinical utility due to its complexity. Panteris et al[6] subsequently proposed a simplified classification system (types A and B). Utilizing Panteris et al[6] criteria, this study aims to: (1) Evaluate PAD’s association with PEP incidence; (2) Compare PEP rates between PAD subtypes (A vs B); and (3) Identify potential strategies for optimizing ERCP safety in high-risk populations.

MATERIALS AND METHODS

Subjects

This retrospective cohort study included 774 consecutive patients diagnosed with choledocholithiasis who underwent ERCP at The Central Hospital of Wuhan and The Third People’s Hospital of Hubei Province between January 2023 and March 2025. The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of The Central Hospital of Wuhan (Approval No. WHZXKYL2025-135). Informed consent was waived due to the retrospective nature of this research. The exclusion criteria were as follows: (1) Pre-existing acute or chronic pancreatitis or hyperamylasemia; (2) Previous ERCP intervention; (3) History of gastrointestinal reconstruction surgery (including Billroth II, Roux-en-Y anastomosis, or pancreaticoduodenectomy); (4) Concurrent biliary strictures (benign or malignant), pancreatic lesions, or duodenal protrusions; (5) Age under 18 years; (6) Technically unsuccessful procedures or incomplete clinical data; and (7) Prophylactic administration of PEP prevention measures (e.g., intravenous hydration or rectal indomethacin).

Main measures

Baseline data included the following: Age, gender, medical history of hypertension, diabetes, coronary heart disease, and viral hepatitis (including hepatitis B and hepatitis C), and history of gallstone surgery. Intraoperative indicators included the use of cannulation method (≥ 1 pancreatic duct guidewire placement) and presence/absence of PAD. Patients with PAD were allocated to the PAD group, while those without PAD were assigned to the non-PAD group. According to the typing criteria proposed by Panteris et al[6], the PAD group was further stratified into two subgroups: The type A PAD subgroup (papilla located at the edge of the diverticulum or within a radius of 2 cm from the diverticular edge) and the type B PAD subgroup (papilla located inside the diverticulum or lying between two adjacent diverticula; Figure 1). Postoperative indicators were the presence of PEP and/or hyperamylasemia.

Figure 1 Schematic diagram of periampullary diverticulum: Based on the positional relationship between periampullary diverticulum and the duodenal papilla, periampullary diverticulum is classified into two subtypes. A: The papilla is located at the edge of the diverticulum or within 2 cm of the diverticular margin; B: The papilla is situated inside the diverticulum or between two adjacent diverticula. PAD: Periampullary diverticulum.

Definitions of outcomes

The primary outcome measure was PEP, and the secondary outcome measure was hyperamylasemia. PEP was defined as follows: (1) New or worsened abdominal pain within 24 hours after ERCP; (2) A serum amylase level ≥ 3 times normal; and/or (3) Imaging evidence of pancreatitis. The diagnosis of PEP required two of the three criteria, as per the Atlanta Consensus[7]. Hyperamylasemia refers to serum amylase levels above the upper normal limit at 4 hours or 12 hours after ERCP, without abdominal pain or related imaging findings[8,9].

Statistical analysis

The main clinical outcome of this study was the occurrence of postoperative pancreatitis (PEP), and all patients were divided into the PEP and non-PEP groups accordingly. We first compared the baseline characteristics between the two groups. Categorical variables were expressed as frequency (percentage) and compared using the χ² test or Fisher’s exact test. Continuous variables were assessed for normality using the Shapiro-Wilk test. Normally distributed variables were expressed as mean ± SD and compared using the independent samples t-test. Non-normally distributed variables were expressed as median (interquartile range) and compared using the Mann-Whitney U test.

To handle missing data, we used the multiple imputation by chained equations method for covariates with < 20% missing data. After addressing missing data, we constructed univariate and multivariate logistic regression models to explore independent risk factors for PEP. Results were presented as odds ratios (ORs) and their 95% confidence intervals (CIs). All variables that showed potential associations (P < 0.10) or had significant clinical relevance individually in univariate analysis were included in the final multivariate model. To clearly demonstrate the impact of the main study variables, such as diverticulum type, after adjusting for confounding factors, we constructed stratified models. Model 1 was a crude analysis without any adjustments, model 2 adjusted for age and gender, and model 3 further adjusted for all other significant covariates identified in univariate analysis based on model 2. Finally, we assessed the robustness of the main results through sensitivity analysis. All statistical analyses were performed using R software (version 4.2.2), and a two-sided P value < 0.05 was considered statistically significant.

RESULTS

Baseline characteristics of the subjects

Following application of the predefined inclusion and exclusion criteria, 615 patients were included in the final analysis. The study population comprised 183 patients with PAD (type A: n = 125; type B: n = 58) and 432 patients without PAD (Figure 2). Comparative analysis of demographic and clinical characteristics revealed similar distributions between groups for sex, hypertension prevalence, diabetes mellitus, coronary artery disease, hepatitis status, and cholecystectomy history (all P > 0.05). However, patients in the PAD group showed higher mean age compared to the non-PAD group (69.0 ± 12.8 years vs 61.6 ± 15.9 years, P < 0.001; Table 1).

Figure 2 Patient screening flowchart. PAD: Periampullary diverticulum; A-PAD: Type A periampullary diverticulum; B-PAD: Type B periampullary diverticulum.

Table 1 Baseline characteristics of the study participants, mean ± SD/n (%).

Variable	Total (n = 615)	Absence of diverticulum (n = 432)	Presence of diverticulum (n = 183)	Statistical value (t/χ2)	P value
Sex				1.298	0.255
Male	311 (50.6)	212 (49.1)	99 (54.1)
Female	304 (49.4)	220 (50.9)	84 (45.9)
Age	63.8 ± 15.4	61.6 ± 15.9	69.0 ± 12.8	-6.082	< 0.001
Hypertension				0.942	0.332
No	341 (55.4)	245 (56.7)	96 (52.5)
Yes	274 (44.6)	187 (43.3)	87 (47.5)
Diabetes				0.576	0.448
No	518 (84.2)	367 (85.0)	151 (82.5)
Yes	97 (15.8)	65 (15.0)	32 (17.5)
Coronary heart disease				0.777	0.378
No	526 (85.5)	373 (86.3)	153 (83.6)
Yes	89 (14.5)	59 (13.7)	30 (16.4)
Hepatitis				0.766	0.382
No	573 (93.2)	405 (93.8)	168 (91.8)
Yes	42 (6.8)	27 (6.2)	15 (8.2)
History of cholecystectomy				0.999	0.317
No	486 (79.0)	346 (80.1)	140 (76.5)
Yes	129 (21.0)	86 (19.9)	43 (23.5)

Univariate analysis of postoperative pancreatitis

The overall PEP incidence in our cohort of 615 patients was 6.83%. A significant association was observed between PAD presence and PEP occurrence (OR = 3.86, 95%CI: 2.03-7.35, P < 0.05). Among patients with PAD, guidewire entry into the pancreatic duct showed a positive association with PEP development (OR = 4.17, 95%CI: 2.17-7.99, P < 0.01). Notably, an inverse relationship was identified between postoperative hyperamylasemia and PEP incidence, with hyperamylasemia cases exhibiting significantly lower PEP rates (OR = 0.09, 95%CI: 0.01-0.65, P = 0.017). No statistically significant associations were found between PEP incidence and other examined variables, including sex, age, hypertension, diabetes mellitus, coronary artery disease, or hepatitis status (all P > 0.05; Table 2).

Table 2 Univariate analysis of the risk factors associated with post-endoscopic retrograde cholangiopancreatography pancreatitis.

Variable	OR (95%CI)	P value
Sex: Male vs female	1.39 (0.74-2.61)	0.313
Age (years)	1.01 (0.99-1.03)	0.451
PAD: Yes vs no	3.86 (2.03-7.35)	< 0.001
Guidewire insertion into the pancreatic duct: Yes vs no	4.17 (2.17-7.99)	< 0.001
Hypertension: Yes vs no	0.75 (0.39-1.42)	0.372
Diabetes mellitus: Yes vs no	0.71 (0.27-1.85)	0.483
Coronary heart disease: Yes vs no	0.99 (0.4-2.42)	0.979
Hepatitis: Yes vs no	0.32 (0.04-2.37)	0.263
History of cholecystectomy: Yes vs no	0.38 (0.13-1.08)	0.07
Postoperative hyperamylasemia: Yes vs no	0.09 (0.01-0.65)	0.017

PAD: Periampullary diverticulum; OR: Odds ratio; CI: Confidence interval.

Multivariate logistic regression analysis of the association between PAD and PEP

After adjusting for age, sex, and other baseline characteristics (model 1), the presence of PAD showed a strong positive association with PEP occurrence (adjusted OR = 4.13, 95%CI: 2.18-11.00, P < 0.001). In model 2, which incorporated additional adjustments for guidewire entry into the pancreatic duct, comorbidities (hypertension, diabetes mellitus, coronary artery disease, and hepatitis), and procedural factors (cholecystectomy history and PEP prevention measures), this association remained robust (adjusted OR = 4.23, 95%CI: 2.06-8.69, P < 0.001). These results suggest that the relationship between PAD and PEP persists even after accounting for potential confounders. Notably, type B PAD demonstrated particularly strong associations with PEP (OR = 14, all P < 0.001), while type A PAD showed no significant correlation (P > 0.5). The significant trend across PAD types (P = 0.039 for trend) indicates a potential graded relationship between PAD classification and PEP occurrence (Table 3).

Table 3 Multivariate Logistic regression analysis of periampullary diverticulum and post-endoscopic retrograde cholangiopancreatography pancreatitis incidence.

Variable	Crude		Model 12		Model 23
	OR (95%CI)	P value	OR (95%CI)	P value	OR (95%CI)	P value
PAD
No	1.01		1.01		1.01
Yes	3.86 (2.03-7.35)	< 0.001	4.13 (2.1-8.11)	< 0.001	4.23 (2.06-8.69)	< 0.001
PAD type
Non-PAD	1.01		1.01		1.01
A-PAD	1.22 (0.47-3.17)	0.68	1.32 (0.5-3.49)	0.572	1.52 (0.56-4.17)	0.411
B-PAD	13.38 (6.39-28.03)	< 0.001	14.65 (6.63-32.33)	< 0.001	14.16 (5.84-34.34)	< 0.001
P value for trend		0.052		0.056		0.039

¹1.0 was used as the reference group, indicates the reference category.

²Model 1: Adjusted for age and sex.

³Model 2: Adjusted for model 1 + periampullary diverticulum, guidewire insertion into the pancreatic duct during endoscopic retrograde cholangiopancreatography, hypertension, diabetes mellitus, coronary heart disease, hepatitis, history of cholecystectomy, post- endoscopic retrograde cholangiopancreatography pancreatitis, Postoperative hyperamylasemia.

PAD: Periampullary diverticulum; A-PAD: Type A periampullary diverticulum; B-PAD: Type B periampullary diverticulum; OR: Odds ratio; CI: Confidence interval.

Association between PAD and PEP in subgroup analyses

The analysis demonstrated that PAD was independently associated with PEP across multiple subgroups. Regarding guidewire positioning, PAD showed a significant association with PEP when the guidewire did not enter the pancreatic duct (adjusted OR = 5.02, 95%CI: 2.06-12.26, P < 0.001). However, no significant association was observed between PAD and PEP when the guidewire entered the pancreatic duct (P = 0.233), suggesting that intubation difficulty may attenuate the observed association with diverticula. In the hypertension-stratified analysis, the association between PAD and PEP appeared stronger in hypertensive patients (adjusted OR = 7.64) compared to non-hypertensive patients (adjusted OR = 2.52), with an interaction P value of 0.162. Although this interaction did not reach statistical significance, the findings may indicate potential effect modification. In certain subgroups (e.g., hepatitis and hyperamylasemia), extremely large OR values (e.g., > 1 × 10⁸) due to sparse event numbers limited further analysis and interpretation (Table 4).

Table 4 Subgroup analyses of the association between periampullary diverticulum and incidence of post-endoscopic retrograde cholangiopancreatography pancreatitis, n (%).

Subgroup	Total	Event	Crude OR (95%CI)	Crude P value	Adjusted OR (95%CI)	Adjusted P value	P value for interaction
Males							0.797
No	212	7 (3.3)	1.01		1.01
Yes	99	11 (11.1)	3.66 (1.37-9.75)	0.009	3.74 (1.38-10.1)	0.009
Females
No	220	10 (4.5)	1.01		1.01
Yes	84	14 (16.7)	4.2 (1.79-9.88)	0.001	4.49 (1.78-11.33)	0.002
Without hypertension							0.162
No	245	13 (5.3)	1.01		1.01
Yes	96	13 (13.5)	2.8 (1.25-6.27)	0.013	2.52 (1.07-5.92)	0.034
With hypertension
No	187	4 (2.1)	1.01		1.01
Yes	87	12 (13.8)	7.32 (2.29-23.42)	0.001	7.64 (2.36-24.71)	0.001
Without diabetes mellitus							0.398
No	367	16 (4.4)	1.01		1.01
Yes	151	21 (13.9)	3.54 (1.79-7)	< 0.001	3.46 (1.71-6.99)	0.001
With diabetes mellitus
No	65	1 (1.5)	1.01		1.01
Yes	32	4 (12.5)	9.14 (0.98-85.53)	0.052	11.54 (1.15-116.01)	0.038
Without coronary heart disease							0.91
No	373	15 (4)	1.01		1.01
Yes	153	21 (13.7)	3.8 (1.9-7.59)	< 0.001	3.71 (1.79-7.67)	< 0.001
With coronary heart disease
No	59	2 (3.4)	1.01		1.01
Yes	30	4 (13.3)	4.38 (0.75-25.48)	0.1	3.59 (0.6-21.49)	0.162
Without hepatitis							0.374
No	405	17 (4.2)	1.01		1.01
Yes	168	24 (14.3)	3.8 (1.99-7.29)	< 0.001	3.95 (2.01-7.75)	< 0.001
With hepatitis
No	27	0 (0)	1.01		1.01
Yes	15	1 (6.7)	02	*	7.82 (0.65-94.5)	0.105
With history of cholecystectomy							0.485
No	346	15 (4.3)	1.01		1.01
Yes	140	23 (16.4)	4.34 (2.19-8.59)	< 0.001	4.43 (2.17-9.04)	< 0.001
Without history of cholecystectomy
No	86	2 (2.3)	1.01		1.01
Yes	43	2 (4.7)	2.05 (0.28-15.07)	0.481	2.01 (0.27-14.96)	0.495
Guidewires enter the pancreatic duct							0.315
No	67	8 (11.9)	1.01		1.01
Yes	47	11 (23.4)	2.25 (0.83-6.13)	0.112	1.93 (0.66-5.67)	0.233
Guidewires did not enter the pancreatic duct
No	365	9 (2.5)	1.01		1.01
Yes	136	14 (10.3)	4.54 (1.92-10.75)	0.001	5.02 (2.06-12.26)	< 0.001
Without hyperamylasemia							0.419
No	355	17 (4.8)	1.01		1.01
Yes	136	24 (17.6)	4.26 (2.21-8.22)	< 0.001	4.27 (2.18-8.38)	< 0.001
With hyperamylasemia
No	77	0 (0)	1.01		1.01
Yes	47	1 (2.1)	02	02	5.91 (0.52-67.8)	0.158

¹1.0 was used as the reference group, indicates the reference category.

²Statistical value (the crude odds ratio, its 95% confidence interval, and the corresponding crude P value) could not be calculated due to the characteristics of the raw data.

OR: Odds ratio; CI: Confidence interval.

DISCUSSION

This retrospective cohort study found that PAD was associated with an increased incidence of PEP, though not all PAD subtypes showed this trend. type B PAD, but not type A PAD, was significantly linked to PEP. Potential explanations include anatomical distortion from diverticula complicating endoscopic maneuvers and mechanically irritating the papilla; reduced ductal tension due to absent smooth muscle support, impairing drainage; and stasis of food residues promoting bacterial overgrowth and local inflammation, further obstructing secretion flow[10-13]. However, compared with the findings of Mohammad Alizadeh et al[14], the PEP incidence was higher in our study, possibly due to the stricter exclusion criteria and larger sample size. Clinically, PAD assessment before ERCP appears warranted.

By comparing the incidence of PEP in different PAD subtypes, we found that the distance from the nipple to PAD also affects its occurrence. Preliminary studies have shown that PAD close to the nipple exacerbates anatomical compression of the ampulla of Vater, causing distortion of the bile and pancreatic ducts and obstructing their drainage, which significantly increases the risk of postoperative complications[10,15,16]. Type B diverticula are closer to the nipple, which partially explains why the incidence of PEP in type B PAD patients is higher than that in type A PAD and non-PAD patients. In addition, the more times the guidewire enters the pancreatic duct, the higher the probability of PEP occurrence; this study has confirmed this result. The OR of PEP after the guidewire enters the pancreatic duct during the procedure was 4.17 in this study, which is consistent with previous research findings[17]. This may be because PAD alters the normal anatomy around the duodenal papilla and the course of the common bile duct and pancreatic duct, thereby affecting the operator’s judgment[4]. As the procedure becomes more difficult, repeated guidewire insertion may injure the pancreatic duct opening and sphincter, resulting in pancreatitis. We also found a significant negative correlation between postoperative hyperamylasemia and PEP. Patients with hyperamylasemia had a 91% reduced risk of PEP (OR = 0.09). Hyperamylasemia is a biochemical marker of acinar cell damage after ERCP, reflecting early injury, while PEP is the result of damage exceeding the body’s tolerance, triggering an uncontrollable inflammatory response[18]. Many patients with elevated serum amylase did not develop PEP, mainly because their pancreas had a strong reserve function and effective protective mechanisms; thus, the injury did not reach a critical stage, and they lacked persistent clinical abdominal pain symptoms. In addition, the strict diagnostic criteria for PEP restrict the identification of cases, potentially underestimating its true incidence.

By analyzing the baseline characteristics of the patients, we found that the age of the PAD group was higher than that of the non-PAD group. The main mechanism is that the local anatomical structure of the duodenal wall in patients shows defects. With increasing age and continuous increases in intestinal cavity pressure, the smooth muscle tension of the intestinal wall at the defect site further weakens. The combined effect of these two factors causes the defective intestinal wall to protrude continuously, which leads to the formation of diverticula. This mechanism explains why patients with diverticula tend to be older, a conclusion confirmed by Suda et al[19]. A theory suggests that changes in bile duct pressure after cholecystectomy may lower the risk of postoperative pancreatitis complications[20,21]. Future studies with larger sample sizes are needed to validate this view. Table 4 shows that the OR for pancreatitis risk in patients with diverticulum and hypertension is 7.64, about three times higher than in patients with diverticulum but without hypertension (OR = 2.52). Although the interaction did not reach significance (P = 0.162), this trend suggests that hypertension and diverticulum may have a potential synergistic effect, increasing the risk of PEP. Due to the small sample size in this study, further research with more participants is needed for in-depth validation.

The results of this study are of great significance for clinical practice. When assessing the risk of progressive PEP, it is important to consider both the presence and the type of PAD. If preoperative imaging assessment or endoscopy reveals a type B diverticulum, it should be regarded as a high-level warning signal, and the operator should adopt a more cautious strategy and take adequate preventive measures, such as placing a pancreatic duct stent, using non-steroidal anti-inflammatory drugs preoperatively, or performing the procedure with a more experienced endoscopist[22]. These measures may help reduce the incidence of PEP. Consequently, they can improve patient prognosis, shorten hospital stays, alleviate economic burdens, and enhance quality of life. In contrast, patients with type A diverticulum generally do not warrant heightened concern regarding their risk.

This study has yielded some results on the relationship between PAD and PEP, but limitations remain. First, as this is a retrospective analysis, selection bias may exist. Second, the strict exclusion criteria led to a small sample size, limiting detailed study of the relationship between PAD subtypes and PEP. This is especially clear in the analysis of type B PAD, where the estimated OR = 14.16 is high but the 95%CI is wide (5.84-34.34). This suggests that although the risk appears high, the exact magnitude is uncertain, likely due to the small size of the type B PAD group. Therefore, future large, multicenter studies are needed to more accurately estimate its effects. The relatively short follow-up for some patients prevented comprehensive long-term assessment, possibly missing cases of postoperative hyperamylasemia. Since the incidence of postoperative hemorrhage, perforation, and postoperative infection is extremely low, we did not include these complications in the analysis.

CONCLUSION

This study shows that peripapillary diverticula independently increase the risk of PEP. Type B PAD significantly increases the risk of PEP. Clinicians should thoroughly assess the procedure-related risks for diverticulum patients during ERCP and take appropriate preventive measures to reduce PEP occurrence. Although the interaction between diverticulum and hypertension in patients is not statistically significant (P = 0.162), a mild synergistic effect on PEP incidence cannot be ruled out. Therefore, strengthened prospective monitoring and management of this population are recommended. Future prospective, multicenter, large-sample studies are needed to investigate how hypertension, diabetes, cholecystectomy history, and peripapillary diverticula affect the risk of postoperative pancreatitis, providing more reliable evidence for clinical practice.