Sequential drain amylase to guide drain removal following pancreatectomy

Optimal cutoff values of drain amylase for predicting pancreatic fistula are different between open and laparoscopic distal pancreatectomy

BACKGROUND

Drainage fluid amylase (DFA) is useful for predicting clinically relevant postoperative pancreatic fistula (CR-POPF) after distal pancreatectomy (DP). However, difference in optimal cutoff value of DFA for predicting CR-POPF between open DP (ODP) and laparoscopic DP (LDP) has not been investigated. This study aimed to identify the optimal cutoff values of DFA for predicting CR-POPF after ODP and LDP.

METHODS

Data for 294 patients (ODP, n = 127; LDP, n = 167) undergoing DP at Kobe University Hospital between 2010 and 2021 were reviewed. Propensity score matching was performed to minimize treatment selection bias. Receiver operating characteristic (ROC) analysis was performed to determine the optimal cutoff values of DFA for predicting CR-POPF for ODP and LDP. Logistic regression analysis for CR-POPF was performed to investigate the diagnostic value of DFA on postoperative day (POD) three with identified cutoff value.

RESULTS

In the matched cohort, CR-POPF rates were 24.7% and 7.9% after ODP and LDP, respectively. DFA on POD one was significantly lower after ODP than after LDP (2263 U/L vs 4243 U/L, p < 0.001), while the difference was not significant on POD three (543 U/L vs 1221 U/L, p = 0.171). ROC analysis revealed that the optimal cutoff value of DFA on POD one and three for predicting CR-POPF were different between ODP and LDP (ODP, 3697 U/L on POD one, 1114 U/L on POD three; LDP, 10564 U/L on POD one, 6020 U/L on POD three). Multivariate analysis showed that DFA on POD three with identified cutoff value was the independent predictor for CR-POPF both for ODP and LDP.

CONCLUSIONS

DFA on POD three is an independent predictor for CR-POPF after both ODP and LDP. However, the optimal cutoff value for it is significantly higher after LDP than after ODP. Optimal threshold of DFA for drain removal may be different between ODP and LDP.

Evaluation and Recalibration of Risk-Stratified Pancreatoduodenectomy Drain Fluid Amylase Removal Criteria

BACKGROUND

Risk-stratified drain fluid amylase cutoff values for postoperative day 1 (POD1) (DFA1) and POD3 (DFA3) can guide early drain removal after pancreatoduodenectomy (PD). The aim of this study was to evaluate and recalibrate cutoff values instituted in Feb 2019 using a prospective sequential cohort.

METHODS

We performed a single-institution prospective cohort study of consecutive patients who underwent pancreatoduodenectomy following implementation of institution-specific DFA cutoffs in February 2019 through April 2022. DFA values, drain removal, and clinically relevant postoperative pancreatic fistulas (CR-POPF) were analyzed. Receiver operating characteristic (ROC) curve analysis determined optimal cutoff values.

RESULTS

In total, 267 patients, 173 (65%) low-risk and 94 (35%) high-risk, underwent 228 (85%) open and 39 (15%) robotic pancreatoduodenectomies. Seven (4%) low-risk patients and 21 (22%) high-risk patients developed CR-POPF. Drains were removed in 147 (55%) patients before/on POD3, with 1 (0.7%) CR-POPF. In low-risk patients, CR-POPF was excluded with 100% sensitivity if DFA1 < 286 (area under curve, AUC = 0.893, p = 0.001) or DFA3 < 97 (AUC = 0.856, p = 0.002). DFA1 < 137 (AUC = 0.786, p < 0.001) or DFA3 < 56 (AUC = 0.819, p < 0.001) were 100% sensitive in high-risk patients. Previously established DFA1 cutoffs of 100 (low-risk) and < 26 (high-risk) were 100% sensitive, while DFA3 cutoffs of 300 (low-risk) and 200 (high-risk) had 57% and 91% sensitivity.

CONCLUSIONS

Within a learning health system, we recalibrated post-PD drain removal thresholds to DFA1 ≤ 300 and DFA3 ≤ 100 for low-risk and DFA1 ≤ 100 and DFA3 ≤ 50 for high-risk patients. This methodology is generalizable to other centers for developing institution-specific criteria to optimize safe early drain removal.

Drain fluid volume combined with amylase level predicts clinically relevant postoperative pancreatic fistula after pancreaticoduodenectomy: A retrospective clinical study

BACKGROUND AND AIM

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.

METHODS

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.

RESULTS

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).

CONCLUSION

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

The risk of clinically-relevant pancreatic fistula after pancreaticoduodenectomy is better predicted by a postoperative trend in drain fluid amylase compared to day 1 values in isolation

BACKGROUND

Recent studies support early drain removal after pancreaticoduodenectomy in patients with a drain fluid amylase on postoperative day 1 (DFA1) level of ≤5,000. The use of DFA1 to guide drain management is increasingly common among pancreatic surgeons; however, the benefit of checking additional drain fluid amylases beyond DFA1 is less known. We sought to determine whether a change in drain fluid amylase (ΔDFA) is a more reliable predictor of clinically relevant postoperative fistula than DFA1 alone.

METHODS

Using the American College of Surgeons National Surgical Quality Improvement Plan, pancreaticoduodenectomy patients with intraoperative drain placement, known DFA1, highest recorded drain fluid amylase value on postoperative day 2 to 5 (DFA2nd), day of drain removal, and clinically relevant postoperative fistula status were reviewed. Logistic models compared the predictive performance of DFA1 alone versus DFA1 + ΔDFA.

RESULTS

A total of 2,417 patients with an overall clinically relevant postoperative fistula rate of 12.6% were analyzed. On multivariable regression, clinical predictors for clinically relevant postoperative fistula included body mass index, steroid use, operative time, and gland texture. These variables were used to develop model 1 (DFA1 alone) and model 2 (DFA1 + ΔDFA). Model 2 outperformed model 1 in predicting the risk of clinically relevant postoperative fistula. According to model 2 predictions, the risk of clinically relevant postoperative fistula increased with any rise in drain fluid amylase, regardless of whether the DFA1 was above or below 5,000 U/L. The risk of clinically relevant postoperative fistula significantly decreased with any drop in drain fluid amylase, with an odds reduction of approximately 50% corresponding with a 70% decrease in drain fluid amylase (P < .001). A risk calculator was developed using DFA1 and a secondary DFA value in conjunction with other clinical predictors for clinically relevant postoperative fistula.

CONCLUSION

Clinically relevant postoperative fistula after pancreaticoduodenectomy is more accurately predicted by DFA1 and ΔDFA versus DFA1 in isolation. We developed a novel risk calculator to provide an individualized approach to drain management after pancreaticoduodenectomy.

C-reactive protein postoperative values to predict clinically relevant postoperative pancreatic fistula after distal pancreatectomy

INTRODUCTION

despite significant medical and technological advances, the incidence of postoperative pancreatic fistula (POPF) after distal pancreatectomy (DP) is reported to be between 3-45 %. The main objective of this study was to analyze the early post-surgical risk factors for developing POPF after DP.

MATERIAL AND METHODS

a retrospective observational study was performed on a prospective basis of patients undergoing DP in a tertiary hospital from January 2011 to December 2021. Sociodemographic, preoperative analytical, tumor-related and postoperative complications variables were analyzed.

RESULTS

of the 52 patients analyzed, 71.8 % of the sample had postoperative drains amylase elevation. However, 25.7 % of the total had grade-B and/or grade-C POPF. Univariate logistic regression with the variables studied showed the following as risk factors for B-C or clinically relevant POPF: amylase values in drainage at the 5th postoperative day (POD) (p = 0.097; 1.01 [1-1.01]), preoperative BMI (p = 0.015; 1.27 [1.04-1.55]) and C-reactive protein (CRP) value at the 3rd POD (p = 0.034; 1.01 [1.01-1.02]). The ROC curve of CRP value at the 3rd POD showed an area under the curve of 0.764 (95 % CI: 0.6-0.93) and the best cut-off point was 190 mg/l (sensitivity 89 % and specificity 67 %).

CONCLUSIONS

CRP value at the 3rd POD is a predictive factor for POPF after DP. Early detection of patients at risk of POPF based on these characteristics could have an impact on their postoperative management.

Adjusting Drain Fluid Amylase for Drain Volume Does Not Improve Pancreatic Fistula Prediction

INTRODUCTION

Drain fluid amylase (DFA) levels have been used to predict clinically relevant postoperative pancreatic fistula (CR-POPF) and guide postoperative drain management. Optimal DFA cutoff thresholds vary between studies, thereby prompting investigation of an alternative assessment technique. As DFA measurements could, in theory, be distorted by variations in ascites fluid production, we hypothesized that adjusting DFA for volume corrected drain fluid amylase (vDFA) would improve CR-POPF predictive models.

METHODS

A single-institution retrospective cohort study of patients, who underwent pancreatoduodenectomies (PD) and distal pancreatectomies (DP) between 2013 and 2019, was performed. DFAs and vDFAs were measured on postoperative day (POD) 3. Clinicopathologic variables were compared between cohorts by univariable and multivariable analyses and Receiver operating characteristic (ROC) curves.

RESULTS

Patients developing a CR-POPF were more likely to be male and have elevated DFA, vDFA, and body mass index (BMI). vDFA use did not contribute to a superior CR-POPF predictive model compared to DFA-a finding consistent on subanalysis of surgery type PD versus DP. In CR-POPF predictive models, DFA, vDFA, and male sex significantly improved CR-POPF predictive models when considering both surgery subtypes, while only DFA and vDFA significantly improved models when cohorts were segregated by surgery type.

CONCLUSIONS

Postoperative DFA remains a preferred method of predicting CR-POPF as the proposed vDFA assessment technique only adds complexity without increased discriminability.

Novel Calculator to Estimate the Risk of Clinically Relevant Postoperative Pancreatic Fistula Following Distal Pancreatectomy

BACKGROUND

Drain management algorithms are based on studies that predict clinically relevant postoperative pancreatic fistula (CR-POPF) using drain fluid amylase level on POD1 (DFA1). These studies are focused on pancreaticoduodenectomy which is inherently different than distal pancreatectomy. Moreover, the change of DFA between POD1 and POD3 (ΔDFA) is underutilized despite its importance in predicting CR-POPF. We sought to generate a calculator that estimates the risk of CR-POPF following distal pancreatectomy.

METHODS

The 2014-2018 pancreas-targeted ACS-NSQIP database was used to identify patients who underwent elective distal pancreatectomy. Models to predict CR-POPF were constructed using DFA1 with/without ΔDFA. The fittest model was used to construct a calculator.

RESULTS

Out of 12,042 distal pancreatectomies, 692 patients met the study's inclusion criteria. The risk of CR-POPF was 15.9% in the included cohort versus 14.8% in the excluded one (P = 0.421). The predictors of the CR-POPF were age, operative time, DFA1, and ΔDFA. Adding ΔDFA decreased the Akaike's information criterion of the model (507.7 vs 544.7)-indicating a significantly better model fit-and improved the cross-validated area under the curve from 0.731 to 0.791. An easy-to-use calculator was created for surgeons to estimate the risk of CR-POPF based on the abovementioned variables. A sensitivity/specificity table was created at various cutoffs to direct clinical decision-making with respect to early drain removal.

CONCLUSIONS

This study highlights the importance of ΔDFA, in addition to DFA1, in predicting CR-POPF. The provided calculator will facilitate predicting CR-POPF and postoperative drain management following distal pancreatectomy.

Machine learning algorithms as early diagnostic tools for pancreatic fistula following pancreaticoduodenectomy and guide drain removal: A retrospective cohort study

OBJECTIVE

Clinically relevant postoperative pancreatic fistula (CR-POPF) remains the major cause of morbidity following pancreaticoduodenectomy (PD). Several model score systems such as the Fistula Risk Score (FRS) have been developed to predict CR-POPF using preoperative and intraoperative data. Machine learning (ML) algorithms are increasingly applied in the medical field and they could be used to assess the risk of CR-POPF, identify clinically meaningful data and guide drain removal.

METHODS

Data from consecutive patients who underwent PD between January 1, 2010 and March 31, 2021 at a single high-volume center was collected retrospectively in this study. Demographics, clinical features, intraoperative parameters, and laboratory values were used to conduct the ML model. Four different ML algorithms (CatBoost, lightGBM, XGBoost and Random Forest) were used to train this model with cross-validation.

RESULTS

A total of 2421 patients with 62 clinical parameters were enrolled in this ML model. The majority of patients (76.3%) underwent open PD while others underwent robot-assisted PD. CR-POPF occurred in 424 (17.5%) patients. The CatBoost algorithm outperformed other algorithms with a mean area under the receiver operating characteristic curve (AUC) of 0.81 (95% confidence interval: 0.80-0.82) from the 5-fold cross-validation procedure. In the test dataset, the CatBoost algorithm also achieved the best mean-AUC of 0.83. The most important value was mean drain fluid amylase (DFA) in the first seven postoperative days (POD). The performance of models that used only preoperative data and intraoperative data was marginally lower than that of models that used combined data.

CONCLUSION

Our ML algorithms could be applied as early diagnostic tools for CR-POPF in patients who underwent PD. Such real-time clinical decision support tools can identify patients with a high risk of CR-POPF, help in developing the perioperative management plan and guide the optimal timing of drain removal.

Innovations in pancreatic anastomosis technique during pancreatoduodenectomies

Purpose

Pancreatic fistula following pancreatic resections is still a relevant complication. The present work shows the efforts of a single institute to decrease this problem.

Methods

A total of 130 patients (63 men, 67 women) with a mean age of 60 (range: 23–81) years were operated on between January 2013 and March 2020. The most frequent type of pancreatic resection was a Whipple procedure with partial antrectomy. During all operations, an innovative method was used, namely a modification of the purse-string suture pancreatojejunostomy. Moreover, an early drain removal policy was applied, based on the drain amylase level on the first and subsequent postoperative days.

Results

Mean postoperative hospital stay was 13 days (range: 7–75). The overall morbidity rate was 43.8%; the clinically relevant (grade B/C) pancreatic fistula (CR-POPF) rate was 6.9%. Delayed gastric emptying (DGE) was observed in 4% of the patients. The ratio of operative mortality was 0.7%; the reoperation rate was 5.3%. Based on the drain amylase level on the first postoperative day, two groups could be established. In the first one, the drain was removed early, on the fourth day in average (range: 2–6). In the other group, the drain was left in situ protractedly or reinserted later on.

Conclusion

A single center’s experience proves that the refinement of the technique can improve the results of pancreatic surgery.

Prospective study on predictability of complications by pancreatic surgeons

PURPOSE

We evaluated the intuition of expert pancreatic surgeons, in predicting the associated risk of pancreatic resection and compared this "intuition" to actual operative follow-up. The objective was to avoid major complications following pancreatic resection, which remains a challenge.

METHODS

From January 2015 to February 2018, all patients who were 18 years old or more undergoing a pancreatic resection (pancreaticoduodenectomy [PD], distal pancreatectomy [DP], or central pancreatectomy [CP]) for pancreatic lesions were included. Preoperatively and postoperatively, all surgeons completed a form assessing the expected potential occurrence of clinically relevant postoperative pancreatic fistula (CR-POPF: grade B or C), postoperative hemorrhage, and length of stay.

RESULTS

Preoperative intuition was assessed for 101 patients for 52 PD, 44 DP, and 5 CP cases. Overall mortality and morbidity rates were 6.9% (n = 7) and 67.3% (n = 68), respectively, and 38 patients (37.6%) developed a POPF, including 27 (26.7%) CR-POPF. Concordance between preoperative intuition of CR-POPF occurrence and reality was minimal, with a Cohen's kappa coefficient (κ) of 0.175 (P value = 0.009), and the same result was obtained between postoperative intuition and reality (κ = 0.351; P < 0.001). When the pancreatic parenchyma was hard, surgeons predicted the absence of CR-POPF with a negative predictive value of 91.3%. However, they were not able to predict the occurrence of CR-POPF when the pancreas was soft (positive predictive value 48%).

CONCLUSIONS

This study assessed for the first time the surgeon's intuition in pancreatic surgery, and demonstrated that pancreatic surgeons cannot accurately assess outcomes except when the pancreatic parenchyma is hard.

Early postoperative drain fluid amylase in risk-stratified patients promotes tailored post-pancreatectomy drain management and potential for accelerated discharge

BACKGROUND

First postoperative day drain fluid amylase (DFA1) <5000 U/L is commonly used for early drain removal. We manage patients with risk-stratified pancreatectomy care pathways determined preoperatively by risk for postoperative pancreatic fistula. We hypothesized that preoperative risk stratification would yield unique DFA1/DFA3 cutoffs for safe early drain removal.

METHODS

Patients with DFA1/DFA3 values after pancreaticoduodenectomy or distal pancreatectomy were identified. Patients were risk stratified as "low-risk pancreaticoduodenectomy," "high-risk pancreaticoduodenectomy," or "distal pancreatectomy." Receiver operator characteristic analyses yielded clinically relevant sensitivity thresholds for International Study Group on Pancreatic Surgery grade B/C postoperative pancreatic fistulas.

RESULTS

From October 2016 to April 2018, 174 patients were preoperatively stratified as low-risk pancreaticoduodenectomy (n = 78, 45%), high-risk pancreaticoduodenectomy (n = 51, 29%), and distal pancreatectomy (n = 45, 26%). B/C postoperative pancreatic fistulas developed in 3% (n = 2) of low-risk pancreaticoduodenectomies, 37% (n = 19) of high-risk pancreaticoduodenectomies, and 24% (n = 11) of distal pancreatectomies (low- vs high-risk pancreaticoduodenectomy P < .001, low-risk pancreaticoduodenectomy versus distal pancreatectomy P = .004, high-risk pancreaticoduodenectomy versus distal pancreatectomy P = .25). B/C postoperative pancreatic fistulas occurred in 16% (n = 21) pancreaticoduodenectomy patients (high- + low-risk pancreaticoduodenectomy), and B/C postoperative pancreatic fistulas were excluded in pancreaticoduodenectomy with 100% sensitivity if DFA1 ≤ 136 or DFA3 ≤ 93. DFA1 < 5000 excluded B/C postoperative pancreatic fistulas with only 57% sensitivity after pancreaticoduodenectomy. Exclusion of B/C postoperative pancreatic fistulas occurred with 100% sensitivity if DFA1 ≤ 661 or DFA3 ≤ 141 in low-risk pancreaticoduodenectomy patients, DFA1 ≤ 136 or DFA3 ≤ 93 in high-risk pancreaticoduodenectomy patients, and DFA1 < 49 or DFA3 < 26 in distal pancreatectomy patients.

CONCLUSION

Preoperative risk stratification results in unique DFA1/DFA3 thresholds to exclude B/C postoperative pancreatic fistulas, thus allowing for safe drain removal and potential for accelerated discharge. Rather than applying generic DFA cutoffs based on national databases, we propose institution-specific DFA1 and DFA3 values tailored to 3 replicable postoperative pancreatic fistula-risk pathways.

Evolving the Paradigm of Early Drain Removal Following Pancreatoduodenectomy

BACKGROUND

Recent data illustrates improved outcomes when adhering to early drain removal following pancreatoduodenectomy (PD). This study aims to explore the potential benefits of expanding the timeframe for early drain removal.

METHODS

Six hundred forty PDs were originally managed by selective drain placement and early removal. Outcomes were reappraised in the framework of a novel proposal; intraoperative drains were omitted based on a low-risk profile (Fistula Risk Score 0-2), followed by drain removal at PODs 1, 3, and 5 if drain fluid amylase (DFA) fell below specific cutoffs based on optimized negative predictive values (NPV) for clinically relevant postoperative pancreatic fistula (CR-POPF). Characteristics of the remaining cohort with drains in situ on POD5 were examined using multivariable analysis (MVA).

RESULTS

Intraoperative FRS would preclude drains from 230 (35.9%) negligible/low-risk cases with a cohort CR-POPF rate of 1.7%. Of the remaining patients, 30.5% would have drains removed on POD1 based on a DFA threshold of 300 IU/L (NPV = 98.4%), demonstrating a 1.6% CR-POPF rate. On POD3, drains could be removed in the residual cohort from 21.1% of patients with DFA ≤ 150 IU/L (NPV = 96.6%), reflecting a 3.4% CR-POPF rate. On POD5, a DFA threshold of 50 IU/L (NPV = 84%) identified 16.3% more patients whose drains could be removed. The remaining cohort (POD5 DFA > 50 IU/L), "enriched" for fistula development and reflecting just 18.4% of the original patients, displays a 61% CR-POPF rate. Among these patients on POD5, a DFA threshold > 2000 IU/L best predicted subsequent CR-POPF (PPV = 89.5%), and MVA revealed a positive association between pancreatic cancer/pancreatitis (OR = 4.37, p = 0.022) and longer operations (OR = 3.74, p = 0.014) with CR-POPF development.

CONCLUSION

Early drain removal is a dynamic concept and can be employed throughout the postoperative time course using conditional thresholds to better identify patients at risk for CR-POPF.