This prospective multicenter study investigated the clinical, biochemical, and imaging predictors of common bile duct (CBD) stones in patients with a history of cholecystectomy. Our results demonstrate that male gender, age > 63 years, liver function test (LFT) alterations, and CBD dilation on ultrasound are independent risk factors for the presence of CBD stones, as confirmed by endoscopic ultrasound (EUS) [9,10,11,12]. Based on these variables, we developed a simplified scoring model that effectively stratifies patients by risk of CBD stones and may assist clinicians in guiding further diagnostic and therapeutic strategies [2,3,4].

These findings carry several important implications for post-cholecystectomy patients, a group frequently encountered in clinical practice but often under-represented in diagnostic algorithms. While existing guidelines predominantly focus on populations with gallbladders in situ, our study highlights the necessity of adapting risk stratification tools to reflect the altered anatomy and physiology following cholecystectomy, including the common occurrence of mild baseline CBD dilation and potentially different symptom profiles. In detail, our findings align with existing literature by confirming age, LFT alterations, and CBD dilation as key predictors for CBD stones while also highlighting male gender as a novel independent risk factor in post-cholecystectomy patients, adding depth to the understanding of stone recurrence in this population [1,12]. Surprisingly, the identification of CBD stones by US was not correlated with the presence of stones according to multivariate analysis due to the low diagnostic accuracy; indeed, the optimal (up to 100%) specificity of US in diagnosing CBD stones was counterbalanced by a very low (<50%) sensitivity.

The identification of four independent risk factors, namely male gender, age > 63 years, LFT alterations, and CBD dilation on US, could have several clinical implications. Indeed, our model showed a clear correlation between the number of risk factors and the likelihood of CBD stones: from 9.5% in patients with no risk factors up to 100% in those with all four. This linear gradient underscores the potential utility of a stepwise pre-test probability approach to optimize clinical decision-making. In high-probability cases (≥3 risk factors), a one-step diagnostic and therapeutic procedure using same-session EUS and ERCP could be justified, reducing unnecessary imaging such as MRCP, shortening time to treatment, and improving patient flow through endoscopy units. For example, we hypothesized that patients presenting with three or more risk factors can be directly scheduled for same-session EUS and ERCP based on the high pre-test probability of CBD stones (66.7% and 100%, respectively).

Although hyperbilirubinemia was significantly associated with the presence of CBD stones in the univariate analysis (OR 1.86, p = 0.04), it did not retain independent predictive value in the multivariate model. This finding likely reflects statistical collinearity between hyperbilirubinemia and other more robust indicators of biliary obstruction, such as LFT alterations and CBD dilation. Furthermore, our study excluded patients with acute cholangitis, who typically presents with jaundice. The exclusion of these cases may have reduced the overall variability and strength of association between bilirubin levels and CBD stones in our cohort. These considerations may explain why hyperbilirubinemia, despite its clinical relevance, did not emerge as an independent predictor in the adjusted model.

The finding that male gender was an independent predictor of CBD stones in this population adds a novel dimension to the current literature, which often underestimates the influence of sex-based differences in biliary pathology. Although, traditionally, female gender is more associated with gallstone formation, our results suggest that, post-cholecystectomy, the risk landscape may shift, potentially due to changes in bile composition, hormone metabolism, or delayed presentation among male patients [13,14,15,16,17].

Interestingly, while abdominal pain and acute pancreatitis were frequent presentations, they were not significantly associated with the presence of CBD stones in either univariate or multivariate analysis. This suggests that symptomatology alone may not be sufficient to discriminate between patients with and without CBD stones, particularly in a post-cholecystectomy setting. Similarly, CBD stones identified by transabdominal ultrasound had low sensitivity, despite high specificity, reinforcing the known limitations of ultrasound in this diagnostic scenario and the value of EUS for definitive diagnosis [18,19,20].

Periampullary diverticulum (PD) and angulation of the distal common bile duct (ADCBD) are anatomical factors known to contribute to biliary stasis and have been implicated in the pathogenesis of common bile duct stones. In our cohort, PD was observed in 7.1% of patients, while ADCBD was present in 6.3% of patients who underwent ERCP after confirmation of CBD stones. However, these findings were not incorporated into the predictive model for several reasons. Firstly, PD was an endoscopic finding (during EUS in this study) and, thus, cannot be used as a criterion for selecting patients for EUS examination. Secondly, ADCBD is typically identified during ERCP after biliary cannulation or on magnetic resonance cholangiopancreatography (MRCP), meaning these data were only available for patients with already confirmed CBD stones in this study. Consequently, ADCBD and PD do not serve as practical predictive factors in the initial diagnostic pathway and were, therefore, excluded from the multivariate analysis. These considerations highlight the challenges of integrating into early predictive models anatomical variations that are identifiable only after the confirmation of stones [12,13,14].

This combined diagnostic and therapeutic approach of performing EUS and ERCP in the same session is crucial for optimizing patient outcomes and spare time in the endoscopic schedule and in reducing the carbon footprint of the procedures [21,22]. Furthermore, it lowers anesthesiologic risks by minimizing exposure to repeated sedations, especially in elderly or high-risk patients [23]. The ability to address both diagnosis and treatment concurrently ensures more efficient care and reduces the likelihood of recurrent biliary events, which are common in patients with untreated CBD stones [24].

Moreover, it is also noteworthy that patients in the validation set had a significantly higher prevalence of hyperbilirubinemia and LFT alterations compared to those in the training set. This reinforces the importance of these clinical signs as strong indicators of CBD stones in patients with previous cholecystectomy. Although abdominal pain and acute pancreatitis did not emerge as statistically significant predictors, they were frequent clinical presentations in our cohort, suggesting their continued relevance in the clinical evaluation of these patients, even if they are not independently associated with the presence of stones.

A major strength of this study is the prospective design, which enhances the generalizability of the results; it has been emphasized in the most recent guidelines that most studies in this field have poor methodological quality. Additionally, we used an independent validation set to test our model. On the other hand, our study presents several limitations. First of all, the relatively small size of both populations may affect the statistical power of the logistic regression analyses; even if this study lacks an a priori sample size calculation, the sample size for this prospective study was based on the number of outcome events needed for multivariate logistic regression analysis. According to the EPV rule, a minimum of 10 outcome events are required per independent variable. With four predictors in our final model and 96 patients with confirmed CBD stones (events), our study exceeded the minimum requirement of 40 events, ensuring adequate statistical power and model stability. Although a sample size calculation was not performed a priori, the post hoc event-based analysis supports the validity of our multivariable approach. Moreover, the exclusion of patients presenting with acute cholangitis requiring urgent ERCP could have impacted on some of the observed results; surprisingly, we failed to identify the presence of hyperbilirubinemia as an independent risk factor. Finally, we excluded all patients with previous CBD interventions (ERCP, percutaneous); with previous surgery; and with hepatic, pancreatic, or biliary neoplasia in order to enroll a homogeneous population. We acknowledge that this tight selection does not reflect the real-life setting of patients with suspected CBD stones, who usually have several comorbidities and medical history. Finally, the design of this study does not allow us to assess possible risk factors related to previous surgery, such as between the cholecystectomy and the enrollment and the possible difficulties encountered during surgery.

Several methodological issues must be kept in mind during the interpretation of the results of this proof-of-concept study. In detail, the lack of a predefined sample size calculation, the 2-year study period, and the single-center nature of the validation cohort limit the statistical power of the results and model reliability. Since all the authors agreed that confirmation by a large multicenter prospective study is required, a prospective study has already been planned, as demonstrated by the synopsis shown in Figure 7.

Future research should aim to validate our model in larger and different populations and explore whether additional clinical or laboratory markers could further refine the performance of the model. In particular, we suggest an external validation of this model in larger, multicenter, and more diverse populations, including those with prior CBD interventions or comorbidities, as well as assessment of the role of additional biomarkers (e.g., inflammatory markers or bile biochemistry) to enhance predictive accuracy, investigation of the cost-effectiveness and patient outcomes of using this risk-based model to guide same-session EUS and ERCP, the exploration of gender-specific differences in CBD stone recurrence and pathophysiology after cholecystectomy [25,26,27,28,29,30,31], and the possible application of novel approaches to treat benign and malignant biliary conditions [32,33].