Background: Post-pancreatectomy liver injury (PPLI) is a poorly defined but increasingly recognized complication in the modern pancreatic surgery. This study aimed to determine the incidence and clinical impact of PPLI and to derive validated transaminase thresholds for its early prediction. Methods: This international retrospective study included consecutive patients who underwent pancreatoduodenectomy, left pancreatectomy with celiac axis resection, or total pancreatectomy at six high-volume centers. A prespecified geographic training-validation framework was used. Clinically PPLI was defined as grade B/C liver failure according to the International Study Group on Liver Surgery, or as radiologically confirmed liver infarction, necrosis, or abscess requiring therapeutic intervention. Prediction models were based on postoperative AST/ALT levels measured between postoperative days (POD) 1 to 3. Results: Overall, 2341 patients were included. In the Western training cohort (n=1457), PPLI occurred in 47 (3.2%) patients. PPLI patients had higher rates of reoperation (46.8% vs 8.7%), major complications (76.6% vs 21.1%), and 90-day mortality (31.9% vs 2.8%) (all p<0.001). PPLI incidence increased with the vascular resection extent (none 1.9%; portal vein 4.1%; hepatic artery 22.2%; p<0.001). Single-marker discrimination increased from POD1 to POD3, with optimal POD1 thresholds being 220 U/L (AST) and 160 U/L (ALT). Combining POD1 and POD3 resulted in the best discrimination for each enzyme with AUCs of 0.90 (AST) and 0.87 (ALT) (biochemical PPLI). In the Eastern validation cohort (n=884) PPLI occurred in 10 (1.1%) patients, and biochemical PPLI was validated achieving AUCs of 0.95 (AST) and 0.96 (ALT). PPLI was an independent predictor of mortality (aOR 5.05, 95% CI 1.89–13.45), and mortality increased progressively from biochemical PPLI (4.4%), to clinically relevant PPLI with liver infarction/abscess (19.4%), and severe PPLI with liver failure (42.3%) (p<0.001). Conclusions: PPLI is an uncommon but deadly complication and occurs more frequently, despite not exclusively, after vascular resection. Early postoperative transaminases accurately identify at-risk patients and support a proposed 3-tier classification into biochemical (grade A), clinically relevant (grade B), and severe (grade C).
Background: Post-pancreatectomy liver injury (PPLI) is a poorly defined but increasingly recognized complication in the modern pancreatic surgery. This study aimed to determine the incidence and clinical impact of PPLI and to derive validated transaminase thresholds for its early prediction. Methods: This international retrospective study included consecutive patients who underwent pancreatoduodenectomy, left pancreatectomy with celiac axis resection, or total pancreatectomy at six high-volume centers. A prespecified geographic training-validation framework was used. Clinically PPLI was defined as grade B/C liver failure according to the International Study Group on Liver Surgery, or as radiologically confirmed liver infarction, necrosis, or abscess requiring therapeutic intervention. Prediction models were based on postoperative AST/ALT levels measured between postoperative days (POD) 1 to 3. Results: Overall, 2341 patients were included. In the Western training cohort (n=1457), PPLI occurred in 47 (3.2%) patients. PPLI patients had higher rates of reoperation (46.8% vs 8.7%), major complications (76.6% vs 21.1%), and 90-day mortality (31.9% vs 2.8%) (all p<0.001). PPLI incidence increased with the vascular resection extent (none 1.9%; portal vein 4.1%; hepatic artery 22.2%; p<0.001). Single-marker discrimination increased from POD1 to POD3, with optimal POD1 thresholds being 220 U/L (AST) and 160 U/L (ALT). Combining POD1 and POD3 resulted in the best discrimination for each enzyme with AUCs of 0.90 (AST) and 0.87 (ALT) (biochemical PPLI). In the Eastern validation cohort (n=884) PPLI occurred in 10 (1.1%) patients, and biochemical PPLI was validated achieving AUCs of 0.95 (AST) and 0.96 (ALT). PPLI was an independent predictor of mortality (aOR 5.05, 95% CI 1.89–13.45), and mortality increased progressively from biochemical PPLI (4.4%), to clinically relevant PPLI with liver infarction/abscess (19.4%), and severe PPLI with liver failure (42.3%) (p<0.001). Conclusions: PPLI is an uncommon but deadly complication and occurs more frequently, despite not exclusively, after vascular resection. Early postoperative transaminases accurately identify at-risk patients and support a proposed 3-tier classification into biochemical (grade A), clinically relevant (grade B), and severe (grade C).
POST- PANCREATECTOMY LIVER INJURY: Incidence, Outcomes, and Biochemical Signature.
MORBIN, CECILIA
2025/2026
Abstract
Background: Post-pancreatectomy liver injury (PPLI) is a poorly defined but increasingly recognized complication in the modern pancreatic surgery. This study aimed to determine the incidence and clinical impact of PPLI and to derive validated transaminase thresholds for its early prediction. Methods: This international retrospective study included consecutive patients who underwent pancreatoduodenectomy, left pancreatectomy with celiac axis resection, or total pancreatectomy at six high-volume centers. A prespecified geographic training-validation framework was used. Clinically PPLI was defined as grade B/C liver failure according to the International Study Group on Liver Surgery, or as radiologically confirmed liver infarction, necrosis, or abscess requiring therapeutic intervention. Prediction models were based on postoperative AST/ALT levels measured between postoperative days (POD) 1 to 3. Results: Overall, 2341 patients were included. In the Western training cohort (n=1457), PPLI occurred in 47 (3.2%) patients. PPLI patients had higher rates of reoperation (46.8% vs 8.7%), major complications (76.6% vs 21.1%), and 90-day mortality (31.9% vs 2.8%) (all p<0.001). PPLI incidence increased with the vascular resection extent (none 1.9%; portal vein 4.1%; hepatic artery 22.2%; p<0.001). Single-marker discrimination increased from POD1 to POD3, with optimal POD1 thresholds being 220 U/L (AST) and 160 U/L (ALT). Combining POD1 and POD3 resulted in the best discrimination for each enzyme with AUCs of 0.90 (AST) and 0.87 (ALT) (biochemical PPLI). In the Eastern validation cohort (n=884) PPLI occurred in 10 (1.1%) patients, and biochemical PPLI was validated achieving AUCs of 0.95 (AST) and 0.96 (ALT). PPLI was an independent predictor of mortality (aOR 5.05, 95% CI 1.89–13.45), and mortality increased progressively from biochemical PPLI (4.4%), to clinically relevant PPLI with liver infarction/abscess (19.4%), and severe PPLI with liver failure (42.3%) (p<0.001). Conclusions: PPLI is an uncommon but deadly complication and occurs more frequently, despite not exclusively, after vascular resection. Early postoperative transaminases accurately identify at-risk patients and support a proposed 3-tier classification into biochemical (grade A), clinically relevant (grade B), and severe (grade C).| File | Dimensione | Formato | |
|---|---|---|---|
|
Morbin_Cecilia.pdf
Accesso riservato
Dimensione
1.85 MB
Formato
Adobe PDF
|
1.85 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
https://hdl.handle.net/20.500.12608/116689