Background: Clinically relevant postoperative pancreatic fistula (CR-POPF) is the main determinant of morbidity after pancreatic resection. Intraoperative risk scores, such as the Fistula Risk Score (FRS), provide a static estimate of risk that does not account for the postoperative course. This study evaluated whether local, systemic, and immune-hematological biomarkers, considered in the chronological order in which they become available, can progressively refine the risk of CR-POPF. Methods: Retrospective single-center cohort of patients undergoing pancreatic resection at the Surgical Oncology of the Digestive Tract Unit, Veneto Institute of Oncology (IOV – IRCCS), between April 2018 and July 2026. Total pancreatectomies were excluded. Drain fluid amylase (DFA), C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and procalcitonin were collected from postoperative day (POD) 1 to POD5. The primary endpoint was CR-POPF (ISGPS 2016 grade B/C); secondary endpoints were POPF of any grade and CR-POPF after pancreatoduodenectomy (PD). Discrimination was assessed with ROC analysis; incremental value was tested with sequential logistic models (end-of-surgery variables, then DFA on POD1, CRP on POD3, and NLR on POD5), using multiple imputation and bootstrap internal validation. Results: Of 171 patients (116 PD, 55 distal pancreatectomies), 42 (24.6%) developed CR-POPF, which was associated with a higher rate of major complications (50.0% vs 10.9%), reoperation, and readmission. DFA on POD1 was the strongest early predictor (AUC 0.78; NPV 92.9% at 830 U/L) and discriminated better than CRP on POD3, NLR on POD5, and the FRS. Adding DFA on POD1 to end-of-surgery variables increased the optimism-corrected AUC from 0.73 to 0.77 (OR 2.26 per tenfold increase); after PD, the FRS lost significance once DFA was included (corrected AUC 0.69 to 0.79). CRP and NLR did not significantly improve the model, although NLR declined more slowly in patients with CR-POPF and CRP showed modest discrimination after PD (AUC 0.67–0.69). A three-marker count stratified CR-POPF risk from 2.9% to 81.3%, without outperforming DFA alone. Conclusions: Postoperative information substantially improves the estimate of CR-POPF risk made at the end of surgery, and the decisive update occurs on POD1 through a local marker. DFA on POD1 is particularly useful to identify low-risk patients; whether systemic markers can refine the assessment of patients with high DFA remains to be tested in larger cohorts. These exploratory findings require external validation.

Background: Clinically relevant postoperative pancreatic fistula (CR-POPF) is the main determinant of morbidity after pancreatic resection. Intraoperative risk scores, such as the Fistula Risk Score (FRS), provide a static estimate of risk that does not account for the postoperative course. This study evaluated whether local, systemic, and immune-hematological biomarkers, considered in the chronological order in which they become available, can progressively refine the risk of CR-POPF. Methods: Retrospective single-center cohort of patients undergoing pancreatic resection at the Surgical Oncology of the Digestive Tract Unit, Veneto Institute of Oncology (IOV – IRCCS), between April 2018 and July 2026. Total pancreatectomies were excluded. Drain fluid amylase (DFA), C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and procalcitonin were collected from postoperative day (POD) 1 to POD5. The primary endpoint was CR-POPF (ISGPS 2016 grade B/C); secondary endpoints were POPF of any grade and CR-POPF after pancreatoduodenectomy (PD). Discrimination was assessed with ROC analysis; incremental value was tested with sequential logistic models (end-of-surgery variables, then DFA on POD1, CRP on POD3, and NLR on POD5), using multiple imputation and bootstrap internal validation. Results: Of 171 patients (116 PD, 55 distal pancreatectomies), 42 (24.6%) developed CR-POPF, which was associated with a higher rate of major complications (50.0% vs 10.9%), reoperation, and readmission. DFA on POD1 was the strongest early predictor (AUC 0.78; NPV 92.9% at 830 U/L) and discriminated better than CRP on POD3, NLR on POD5, and the FRS. Adding DFA on POD1 to end-of-surgery variables increased the optimism-corrected AUC from 0.73 to 0.77 (OR 2.26 per tenfold increase); after PD, the FRS lost significance once DFA was included (corrected AUC 0.69 to 0.79). CRP and NLR did not significantly improve the model, although NLR declined more slowly in patients with CR-POPF and CRP showed modest discrimination after PD (AUC 0.67–0.69). A three-marker count stratified CR-POPF risk from 2.9% to 81.3%, without outperforming DFA alone. Conclusions: Postoperative information substantially improves the estimate of CR-POPF risk made at the end of surgery, and the decisive update occurs on POD1 through a local marker. DFA on POD1 is particularly useful to identify low-risk patients; whether systemic markers can refine the assessment of patients with high DFA remains to be tested in larger cohorts. These exploratory findings require external validation.

From Surgical Risk to Postoperative Biology: Sequential Risk Stratification of Clinically Relevant Postoperative Pancreatic Fistula

PANIZZA, FRANCESCO
2024/2025

Abstract

Background: Clinically relevant postoperative pancreatic fistula (CR-POPF) is the main determinant of morbidity after pancreatic resection. Intraoperative risk scores, such as the Fistula Risk Score (FRS), provide a static estimate of risk that does not account for the postoperative course. This study evaluated whether local, systemic, and immune-hematological biomarkers, considered in the chronological order in which they become available, can progressively refine the risk of CR-POPF. Methods: Retrospective single-center cohort of patients undergoing pancreatic resection at the Surgical Oncology of the Digestive Tract Unit, Veneto Institute of Oncology (IOV – IRCCS), between April 2018 and July 2026. Total pancreatectomies were excluded. Drain fluid amylase (DFA), C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and procalcitonin were collected from postoperative day (POD) 1 to POD5. The primary endpoint was CR-POPF (ISGPS 2016 grade B/C); secondary endpoints were POPF of any grade and CR-POPF after pancreatoduodenectomy (PD). Discrimination was assessed with ROC analysis; incremental value was tested with sequential logistic models (end-of-surgery variables, then DFA on POD1, CRP on POD3, and NLR on POD5), using multiple imputation and bootstrap internal validation. Results: Of 171 patients (116 PD, 55 distal pancreatectomies), 42 (24.6%) developed CR-POPF, which was associated with a higher rate of major complications (50.0% vs 10.9%), reoperation, and readmission. DFA on POD1 was the strongest early predictor (AUC 0.78; NPV 92.9% at 830 U/L) and discriminated better than CRP on POD3, NLR on POD5, and the FRS. Adding DFA on POD1 to end-of-surgery variables increased the optimism-corrected AUC from 0.73 to 0.77 (OR 2.26 per tenfold increase); after PD, the FRS lost significance once DFA was included (corrected AUC 0.69 to 0.79). CRP and NLR did not significantly improve the model, although NLR declined more slowly in patients with CR-POPF and CRP showed modest discrimination after PD (AUC 0.67–0.69). A three-marker count stratified CR-POPF risk from 2.9% to 81.3%, without outperforming DFA alone. Conclusions: Postoperative information substantially improves the estimate of CR-POPF risk made at the end of surgery, and the decisive update occurs on POD1 through a local marker. DFA on POD1 is particularly useful to identify low-risk patients; whether systemic markers can refine the assessment of patients with high DFA remains to be tested in larger cohorts. These exploratory findings require external validation.
2024
From Surgical Risk to Postoperative Biology: Sequential Risk Stratification of Clinically Relevant Postoperative Pancreatic Fistula
Background: Clinically relevant postoperative pancreatic fistula (CR-POPF) is the main determinant of morbidity after pancreatic resection. Intraoperative risk scores, such as the Fistula Risk Score (FRS), provide a static estimate of risk that does not account for the postoperative course. This study evaluated whether local, systemic, and immune-hematological biomarkers, considered in the chronological order in which they become available, can progressively refine the risk of CR-POPF. Methods: Retrospective single-center cohort of patients undergoing pancreatic resection at the Surgical Oncology of the Digestive Tract Unit, Veneto Institute of Oncology (IOV – IRCCS), between April 2018 and July 2026. Total pancreatectomies were excluded. Drain fluid amylase (DFA), C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and procalcitonin were collected from postoperative day (POD) 1 to POD5. The primary endpoint was CR-POPF (ISGPS 2016 grade B/C); secondary endpoints were POPF of any grade and CR-POPF after pancreatoduodenectomy (PD). Discrimination was assessed with ROC analysis; incremental value was tested with sequential logistic models (end-of-surgery variables, then DFA on POD1, CRP on POD3, and NLR on POD5), using multiple imputation and bootstrap internal validation. Results: Of 171 patients (116 PD, 55 distal pancreatectomies), 42 (24.6%) developed CR-POPF, which was associated with a higher rate of major complications (50.0% vs 10.9%), reoperation, and readmission. DFA on POD1 was the strongest early predictor (AUC 0.78; NPV 92.9% at 830 U/L) and discriminated better than CRP on POD3, NLR on POD5, and the FRS. Adding DFA on POD1 to end-of-surgery variables increased the optimism-corrected AUC from 0.73 to 0.77 (OR 2.26 per tenfold increase); after PD, the FRS lost significance once DFA was included (corrected AUC 0.69 to 0.79). CRP and NLR did not significantly improve the model, although NLR declined more slowly in patients with CR-POPF and CRP showed modest discrimination after PD (AUC 0.67–0.69). A three-marker count stratified CR-POPF risk from 2.9% to 81.3%, without outperforming DFA alone. Conclusions: Postoperative information substantially improves the estimate of CR-POPF risk made at the end of surgery, and the decisive update occurs on POD1 through a local marker. DFA on POD1 is particularly useful to identify low-risk patients; whether systemic markers can refine the assessment of patients with high DFA remains to be tested in larger cohorts. These exploratory findings require external validation.
POPF
Drain fluid amylase
C-reactive protein
NRL
Dynamic risk
File in questo prodotto:
File Dimensione Formato  
Panizza_Francesco.pdf

Accesso riservato

Dimensione 1.09 MB
Formato Adobe PDF
1.09 MB Adobe PDF

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/119695