Pancreatic fistula is the most common complication of pancreatic surgeries that causes more serious conditions, including bleeding due to visceral vessel erosion and peritonitis.

To develop a machine learning (ML) model for postoperative pancreatic fistula and identify significant risk factors of the complication.

A single-center retrospective clinical study was conducted which included 150 patients, who underwent pancreatoduodenectomy. Logistic regression, random forest, and CatBoost were employed for modeling the biochemical leak (symptomless fistula) and fistula grade B/C (clinically significant complication). The performance was estimated by receiver operating characteristic (ROC) area under the curve (AUC) after 5-fold cross-validation (20% testing and 80% training data). The risk factors were evaluated with the most accurate algorithm, based on the parameter "Importance" (Im), and Kendall correlation, P < 0.05.

The CatBoost algorithm was the most accurate with an AUC of 74%-86%. The study provided results of ML-based modeling and algorithm selection for pancreatic fistula prediction and risk factor evaluation. From 14 parameters we selected the main pre- and intraoperative prognostic factors of all the fistulas: Tumor vascular invasion (Im = 24.8%), age (Im = 18.6%), and body mass index (Im = 16.4%), AUC = 74%. The ML model showed that biochemical leak, blood and drain amylase level (Im = 21.6% and 16.4%), and blood leukocytes (Im = 11.2%) were crucial predictors for subsequent fistula B/C, AUC = 86%. Surgical techniques, morphology, and pancreatic duct diameter less than 3 mm were insignificant (Im < 5% and no correlations detected). The results were confirmed by correlation analysis.

This study highlights the key predictors of postoperative pancreatic fistula and establishes a robust ML-based model for individualized risk prediction. These findings contribute to the advancement of personalized perioperative care and may guide targeted preventive strategies.

Clinically relevant postoperative pancreatic fistula (CR-POPF) represents a significant complication after pancreaticoduodenectomy (PD). Therefore, the early prediction of CR-POPF is of paramount importance. Based on above, this study sought to develop a CR-POPF prediction model that amalgamates radiomics and clinical features to predict CR-POPF, utilizing Shapley Additive explanations (SHAP) for visualization.

Extensive radiomics features were extracted from preoperative enhanced Computed Tomography (CT) images of patients scheduled for PD. Subsequently, feature selection was performed using Least Absolute Shrinkage and Selection Operator (Lasso) regression and random forest (RF) algorithm to select pertinent radiomics and clinical features. Last, 15 CR-POPF prediction models were developed using five distinct machine learning (ML) predictors, based on selected radiomics features, selected clinical features, and a combination of both. Model performance was compared using DeLong's test for the area under the receiver operating characteristic curve (AUC) differences.

The CR-POPF prediction model based on the XGBoost predictor with the combination of the radiomics and clinical features selected by Lasso regression and RF exhibited superior performance among these 15 CR-POPF prediction models, achieving an accuracy of 0.85, an AUC of 0.93. DeLong's test showed statistically significant differences (P < 0.05) when compared to the radiomics-only and clinical-only models, with recall of 0.63, precision of 0.65, and F1 score of 0.64.

The proposed CR-POPF prediction model based on the XGBoost predictor with the combination of the radiomics and clinical features selected by Lasso regression and RF can effectively predicting the CR-POPF and may provide strong support for early clinical management of CR-POPF.

Pancreatoduodenectomy (PD) is a challenging procedure which is associated with high morbidity rates. This study was performed to make an overview of risk factors included in risk stratification methods both logistic regression models and models based on artificial intelligence algorithms to predict postoperative pancreatic fistula (POPF) and other complications following PD and to provide insight in the extent to which these tools were validated.

Five databases were searched to identify relevant studies. Calculators, equations, nomograms, and artificial intelligence models that addressed POPF and other complications were included. Only PD resections were considered eligible. There was no exclusion of the minimally invasive techniques reporting PD resections. All other pancreatic resections were excluded.

90 studies were included. Thirty-five studies were related to POPF, thirty-five studies were related to other complications following PD and twenty studies were related to artificial intelligence predication models after PD. Among the identified risk factors, the most used factors for POPF risk stratification were the main pancreatic duct diameter (MPD) (80%) followed by pancreatic texture (51%), whereas for other complications the most used factors were age (34%) and ASA score (29.4%). Only 26% of the evaluated risk stratification tools for POPF and other complications were externally validated. This percentage was even lower for the risk models using artificial intelligence which was 20%.

The MPD was the most used factor when stratifying the risk of POPF followed by pancreatic texture. Age and ASA score were the most used factors for the stratification of other complications. Insight in clinically relevant risk factors could help surgeons in adapting their surgical strategy and shared decision-making. This study revealed that the focus of research still lies on developing new risk models rather than model validation, hampering clinical implementation of these tools for decision support.

Prophylactic abdominal drainage (PAD) is considered a routine procedure after pancreatoduodenectomy (PD) to prevent and detect severe complications at an early stage. However, the drainage itself may cause adverse consequences. Thus, the optimal strategy of PAD after PD remains controversial.

The present paper summarizes the latest research on the strategies of PAD following PD, mainly focusing on 1) the selective placement of PAD, 2) the optimal drainage types, 3) the early removal of drainage (EDR), and 4) novel strategies for PAD management.

Accurate stratifications based on the potential risk factors of clinically relevant-postoperative pancreatic fistula (CR-POPF) facilitates the selective placement of PAD and the implementation of EDR, with postoperative outcomes superior or similar to routine PAD placement. Both active and passive drainage methods are feasible in most patients after PD, with similar prognostic outcomes. Novel predictive models with accurate, dynamic, and individualized performance further guide the management of PAD and afford a better prognosis.

Evidence-based risk stratification of CR-POPF aids in the management of PAD in patients undergoing PD. Novel dynamic and individualized PAD strategies might be the next hotspot in drain-relevant explorations.

Postoperative computed tomography imaging has been shown to play an important role in avoiding failure-to-rescue. We sought to examine the impact of the timing of such imaging studies on outcomes after pancreatectomy.

Patients who underwent pancreatic resection at our institution from 2017 to 2022 were reviewed retrospectively to identify those undergoing computed tomography for any indication before discharge. Patients were subdivided by the postoperative day that the first computed tomography scan was obtained: immediate (postoperative day <3), early (postoperative day 3-7), and delayed (postoperative day >7).

Of 370 patients, 110 (30%) had a computed tomography during the initial surgical stay. The 3 timing groups were similar in age, comorbidities, pathology, operative time, and number of scans. When comparing the early with the delayed group, we found that antibiotic usage, percutaneous drainage, and overall invasive interventions during surgical stay were all similar. However, those patients who were scanned in the early period had significantly shorter length of stay (17.05 vs 22.82, P = .0008) and fewer composite days hospitalized (20.1 vs 24.9, P = .01) relative to the delayed group. Importantly, early computed tomography imaging was found to be the only independent predictor of a postoperative length of stay ≤15 days on multivariate analysis. Surgical stay mortality rates were significantly lower in the early compared with delayed group (0% vs 11%, P = .02). A change in treatment was observed in 59% after computed tomography, with 15% undergoing invasive interventions, 27% treated medically, and 16% with expectant management.

In our cohort, patients imaged early after pancreatectomy experienced shorter hospital stays and lower inpatient mortality relative to those scanned after the first postoperative week.

The advancement of laparoscopic technology has broadened the application of laparoscopic pancreaticoduodenectomy (LPD) for treating pancreatic head and ampullary tumors. Despite its benefits, postoperative pancreatic fistula (POPF) and postpancreatectomy hemorrhage (PPH) remain significant complications. Ligamentum teres hepatis wrapping around the gastroduodenal artery (GDA) stump show limitations in reducing POPF and PPH.

This study retrospectively analyzed patients undergoing LPD from January 2016 to October 2023, We compared the effectiveness of the two-parts wrapping (the ligamentum teres hepatis wrapping of the gastroduodenal artery stump and the omentum flap wrapping of the pancreatojejunal anastomosis) and ligamentum teres hepatis wrapping around the gastroduodenal artery (GDA) in reducing postoperative pancreatic fistula (POPF) and postpancreatectomy hemorrhage (PPH), using propensity score matching for the analysis.

A total of 172 patients were analyzed, showing that the two-parts wrapping group significantly reduced the rates of overall and severe complications, POPF, and PPH compared to ligamentum teres hepatis wrapping around the GDA group. Specifically, the study found lower rates of grade B/C POPF and no instances of PPH in the two-parts wrapping group, alongside shorter postoperative hospital stays and drainage removal times. These benefits were particularly notable in patients with soft pancreatic textures and pancreatic duct diameters of < 3 mm.

The two-parts wrapping technique significantly reduce the risks of POPF and PPH in LPD, offering a promising approach for patients with soft pancreas and pancreatic duct diameter of < 3 mm.

Postoperative pancreatic fistula (POPF) is a frequent complication after pancreatectomy, leading to increased morbidity and mortality. Optimizing prediction models for POPF has emerged as a critical focus in surgical research. Although over sixty models following pancreaticoduodenectomy, predominantly reliant on a variety of clinical, surgical, and radiological parameters, have been documented, their predictive accuracy remains suboptimal in external validation and across diverse populations. As models after distal pancreatectomy continue to be progressively reported, their external validation is eagerly anticipated. Conversely, POPF prediction after central pancreatectomy is in its nascent stage, warranting urgent need for further development and validation. The potential of machine learning and big data analytics offers promising prospects for enhancing the accuracy of prediction models by incorporating an extensive array of variables and optimizing algorithm performance. Moreover, there is potential for the development of personalized prediction models based on patient- or pancreas-specific factors and postoperative serum or drain fluid biomarkers to improve accuracy in identifying individuals at risk of POPF. In the future, prospective multicenter studies and the integration of novel imaging technologies, such as artificial intelligence-based radiomics, may further refine predictive models. Addressing these issues is anticipated to revolutionize risk stratification, clinical decision-making, and postoperative management in patients undergoing pancreatectomy.

Postoperative pancreatic fistula is a life-threatening complication with an unmet need for accurate prediction. This study was aimed to develop preoperative artificial intelligence-based prediction models. Patients who underwent pancreaticoduodenectomy were enrolled and stratified into model development and validation sets by surgery between 2016 and 2017 or in 2018, respectively. Machine learning models based on clinical and body composition data, and deep learning models based on computed tomographic data, were developed, combined by ensemble voting, and final models were selected comparison with earlier model. Among the 1333 participants (training, n = 881; test, n = 452), postoperative pancreatic fistula occurred in 421 (47.8%) and 134 (31.8%) and clinically relevant postoperative pancreatic fistula occurred in 59 (6.7%) and 27 (6.0%) participants in the training and test datasets, respectively. In the test dataset, the area under the receiver operating curve [AUC (95% confidence interval)] of the selected preoperative model for predicting all and clinically relevant postoperative pancreatic fistula was 0.75 (0.71-0.80) and 0.68 (0.58-0.78). The ensemble model showed better predictive performance than the individual ML and DL models.

When they encounter various highly related postoperative complications, existing risk evaluation tools that focus on single or any complications are inadequate in clinical practice. This seriously hinders complication management because of the lack of a quantitative basis. An interpretable multilabel model framework that predicts multiple complications simultaneously is urgently needed.

The authors included 50 325 inpatients from a large multicenter cohort (2014-2017). The authors separated patients from one hospital for external validation and randomly split the remaining patients into training and internal validation sets. A MARKov-EmbeDded (MARKED) multilabel model was proposed, and three models were trained for comparison: binary relevance, a fully connected network (FULLNET), and a deep neural network. Performance was mainly evaluated using the area under the receiver operating characteristic curve (AUC). The authors interpreted the model using Shapley Additive Explanations. Complication-specific risk and risk source inference were provided at the individual level.

There were 26 292, 6574, and 17 459 inpatients in the training, internal validation, and external validation sets, respectively. For the external validation set, MARKED achieved the highest average AUC (0.818, 95% CI: 0.771-0.864) across eight outcomes [compared with binary relevance, 0.799 (0.748-0.849), FULLNET, 0.806 (0.756-0.856), and deep neural network, 0.815 (0.765-0.866)]. Specifically, the AUCs of MARKED were above 0.9 for cardiac complications [0.927 (0.894-0.960)], neurological complications [0.905 (0.870-0.941)], and mortality [0.902 (0.867-0.937)]. Serum albumin, surgical specialties, emergency case, American Society of Anesthesiologists score, age, and sex were the six most important preoperative variables. The interaction between complications contributed more than the preoperative variables, and formed a hierarchical chain of risk factors, mild complications, and severe complications.

The authors demonstrated the advantage of MARKED in terms of performance and interpretability. The authors expect that the identification of high-risk patients and the inference of the risk source for specific complications will be valuable for clinical decision-making.

This study aimed to analyze the predictive value of umbilical cord blood Interleukin-6 (UCB IL-6) for the severity-graded BPD and to establish machine learning (ML) predictive models in a Chinese population based on the 2019 NRN evidence-based guidelines.

In this retrospective analysis, we included infants born with gestational age <32 weeks, who underwent UCB IL-6 testing within 24 h of admission to our NICU between 2020 and 2022. We collected their medical information encompassing the maternal, perinatal, and early neonatal phases. Furthermore, we classified the grade of BPD according to the 2019 NRN evidence-based guidelines. The correlation between UCB IL-6 and the grades of BPD was analyzed. Univariate analysis and ordinal logistic regression were employed to identify risk factors, followed by the development of ML predictive models based on XGBoost, CatBoost, LightGBM, and Random Forest. The AUROC was used to evaluate the diagnostic value of each model. Besides, we generated feature importance distribution plots based on SHAP values to emphasize the significance of UCB IL-6 in the models.

The study ultimately enrolled 414 preterm infants, with No BPD group (n = 309), Grade 1 BPD group (n = 73), and Grade 2-3 BPD group (n = 32). The levels of UCB IL-6 increased with the grades of BPD. UCB IL-6 demonstrated clinical significance in predicting various grades of BPD, particularly in distinguishing Grade 2-3 BPD patients, with an AUROC of 0.815 (95% CI: 0.753-0.877). All four ML models, XGBoost, CatBoost, LightGBM, and Random Forest, exhibited Micro-average AUROC values of 0.841, 0.870, 0.851, and 0.878, respectively. Notably, UCB IL-6 consistently appeared as the most prominent feature across the feature importance distribution plots in all four models.

UCB IL-6 significantly contributes to predicting severity-graded BPD, especially in grade 2-3 BPD. Through the development of four ML predictive models, we highlighted UCB IL-6's importance.

To evaluate the performance of published fistula risk models by external validation, and to identify independent risk factors for postoperative pancreatic fistula (POPF).

Multiple risk models have been developed to predict POPF after pancreatoduodenectomy. External validation in high-quality prospective cohorts is, however, lacking or only performed for individual models.

A post hoc analysis of data from the stepped-wedge cluster cluster-randomized Care After Pancreatic Resection According to an Algorithm for Early Detection and Minimally Invasive Management of Pancreatic Fistula versus Current Practice (PORSCH) trial was performed. Included were all patients undergoing pancreatoduodenectomy in the Netherlands (January 2018-November 2019). Risk models on POPF were identified by a systematic literature search. Model performance was evaluated by calculating the area under the receiver operating curves (AUC) and calibration plots. Multivariable logistic regression was performed to identify independent risk factors associated with clinically relevant POPF.

Overall, 1358 patients undergoing pancreatoduodenectomy were included, of whom 341 patients (25%) developed clinically relevant POPF. Fourteen risk models for POPF were evaluated, with AUCs ranging from 0.62 to 0.70. The updated alternative fistula risk score had an AUC of 0.70 (95% confidence intervals [CI]: 0.69-0.72). The alternative fistula risk score demonstrated an AUC of 0.70 (95% CI: 0.689-0.71), whilst an AUC of 0.70 (95% CI: 0.699-0.71) was also found for the model by Petrova and colleagues. Soft pancreatic texture, pathology other than pancreatic ductal adenocarcinoma or chronic pancreatitis, small pancreatic duct diameter, higher body mass index, minimally invasive resection and male sex were identified as independent predictors of POPF.

Published risk models predicting clinically relevant POPF after pancreatoduodenectomy have a moderate predictive accuracy. Their clinical applicability to identify high-risk patients and guide treatment strategies is therefore questionable.

Several indicators are recognized in the development of clinically relevant postoperative pancreatic fistula (CR-POPF) after pancreaticoduodenectomy (PD). However, drain fluid volume (DFV) remains poorly studied. We aimed to discover the predictive effects of DFV and guide clinical management.

We retrospectively reviewed the clinical data of patients that received PD between January 2015 and December 2019 in a high-volume center. DFV was analyzed as a potential risk factor and postoperative short-term outcomes as well as drain removal time were compared stratified by different DFV levels. Receiver operating characteristic curves and area under curves (AUC) were compared for DFV alone and DFV combined with drain fluid amylase (DFA). Subgroup analysis of DFV stratified by DFA evaluated the predictability of CR-POPF.

CR-POPF occurred in 19.7% of 841 patients. Hypertension, postoperative day 3 (POD3) DFA ≥ 300 U/L, and POD3 DFV ≥ 30 mL were independent risk factors, while pancreatic main duct diameter ≥ 3 mm was a protective factor. POD3 DFV ≥ 30 mL increased the overall occurrences of CR-POPF and major complications (P = 0.017; P = 0.029). POD3 DFV alone presented a low predictive value (AUC 0.602), while POD3 DFV combined with DFA had a high predictive value (AUC 0.759) for CR-POPF. Subgroup analysis showed that the combination of POD3 DFV ≥ 30 mL and DFA ≥ 300 U/L led to higher incidences of CR-POPF (P = 0.003).

CR-POPF is common after PD, and high DFV combined with DFA may predict its occurrence and facilitate appropriate management.

Pancreaticoduodenectomy (PD) nowadays serves as a standard treatment for patients with disorders of the pancreas, intestine, and bile duct. Although the mortality rate of patients undergoing PD has decreased significantly, postoperative complication rates remain high. Dexamethasone, a synthetic glucocorticoid with potent anti-inflammatory and metabolic effects, has been proven to have a favorable effect on certain complications. However, the role it plays in post-pancreatectomy patients has not been systematically evaluated. The aim of this study is to assess the effect of dexamethasone on postoperative complications after PD.

The PANDEX trial is an investigator-initiated, multicentric, prospective, randomized, double-blinded, placebo-control, pragmatic study. The trial is designed to enroll 300 patients who are going to receive elective PD. Patients will be randomized to receive 0.2 mg/kg dexamethasone or saline placebo, administered as an intravenous bolus within 5 min after induction of anesthesia. The primary outcome is the Comprehensive Complication Index (CCI) score within 30 days after the operation. The secondary outcomes include postoperative major complications (Clavien-Dindo≥3), postoperative pancreatic fistula (POPF), post-pancreatectomy acute pancreatitis (PPAP), infection, and unexpected relaparotomy, as well as postoperative length of stay, 30-day mortality, and 90-day mortality.

The PANDEX trial is the first randomized controlled trial concerning the effect of dexamethasone on postoperative complications of patients undergoing PD, with the hypothesis that the intraoperative use of dexamethasone can reduce the incidence of postoperative complications and improve short-term outcomes after PD. The results of the present study will guide the perioperative use of dexamethasone and help improve the clinical management of post-pancreatectomy patients.

ClinicalTrials.gov NCT05567094. Registered on 30 September 30 2022.

Postoperative pancreatic fistula (POPF) is the leading cause of morbidity and early mortality in patients undergoing pancreatic resection. In addition, recent studies have identified postoperative acute pancreatitis (POAP) as an independent contributor to morbidity. Most perioperative mitigation strategies experimented for POPF have been shown to be in vain with no consensus on the best perioperative management. Clinical prediction models have been developed with the hope of identifying high POPF risk patients with the leading idea of finding subpopulations possibly benefiting from pre-existing or novel mitigation strategies. The aim of this review was to map out the existing prediction modeling studies to better understand the current stage of POPF prediction modeling, and the methodology behind them.

A narrative review of the existing POPF prediction model studies was performed. Studies published before September 2022 were included.

While the number of POPF prediction models for pancreatoduodenectomy has increased, none of the currently existing models stand out from the crowd. For distal pancreatectomy, two unique POPF prediction models exist, but due to their freshness, no further external validation or adoption in clinics or research has been reported. There seems to be a lack of adherence to correct methodology or reporting guidelines in most of the studies, which has rendered external validity-if assessed-low. Few of the most recent studies have demonstrated preoperative assessment of pancreatic aspects from computed tomography (CT) scans to provide relatively strong predictors of POPF.

Main goal for the future would be to reach a consensus on the most important POPF predictors and prediction model. At their current state, few models have demonstrated adequate transportability and generalizability to be up to the task. Better understanding of POPF pathophysiology and the possible driving force of acute inflammation and POAP might be required before such a prediction model can be accessed.