Surgeon vs Pathologist for Prediction of Pancreatic Fistula: Results from the Randomized Multicenter RECOPANC Study.
Aged
Aged, 80 and over
Female
Fibrosis
Humans
Male
Margins of Excision
Middle Aged
Pancreas
/ pathology
Pancreatic Fistula
/ epidemiology
Pancreatic Neoplasms
/ pathology
Pancreaticoduodenectomy
/ adverse effects
Pathologists
/ statistics & numerical data
Postoperative Complications
/ epidemiology
Prospective Studies
Reproducibility of Results
Risk Factors
Surgeons
/ statistics & numerical data
Journal
Journal of the American College of Surgeons
ISSN: 1879-1190
Titre abrégé: J Am Coll Surg
Pays: United States
ID NLM: 9431305
Informations de publication
Date de publication:
06 2021
06 2021
Historique:
received:
05
02
2021
revised:
07
03
2021
accepted:
10
03
2021
pubmed:
23
4
2021
medline:
9
10
2021
entrez:
22
4
2021
Statut:
ppublish
Résumé
Surgically assessed pancreatic texture has been identified as the strongest predictor of postoperative pancreatic fistula. However, texture is a subjective parameter with no proven reliability or validity. Therefore, a more objective parameter is needed. In this study, we evaluated the fibrosis level at the pancreatic neck resection margin and correlated fibrosis and all clinico-pathologic parameters collected over the course of the Pancreatogastrostomy vs Pancreatojejunostomy for RECOnstruction (RECOPANC) study. The RECOPANC trial was a multicenter randomized prospective trial of patients undergoing pancreatoduodenectomy. There were 261 hematoxylin and eosin-stained slides allocated for histopathologic analyses. Pancreatic fibrosis was scored from 0 to III (no fibrosis up to severe fibrosis) by 2 blinded independent pathologists. All variables possibly associated with POPF were entered into a generalized linear model for multivariable analysis. The fibrosis grade and pancreatic texture were scored in all 261 patients. In POPF B/C (postoperative pancreatic fistula grade B or C) patients, 71% had a soft pancreas, and fibrosis grades were distributed as follows: 48% with score 0, 28% with score I, 20% with score II, and 7% with score III, respectively. Fibrosis grading showed substantial inter-rater reliability (kappa = 0.74) and correlated positively with hard pancreatic texture (p < 0.05). In univariable analysis, area under the curve (AUC) for POPF B/C prediction was higher for fibrosis grade than for pancreatic texture (0.71 vs 0.59). In multivariate analysis, the following predictors were selected: sex, surgeon volume, pancreatic texture, and fibrosis grade. However, the addition of pancreatic texture only led to an incremental improvement (AUC 0.794 vs 0.819). Histologically evaluated pancreatic fibrosis is an easily applicable and highly reproducible POPF predictor and superior to surgically evaluated pancreatic texture. Future studies might use fibrosis grade for risk stratification in pancreatoduodenectomy.
Sections du résumé
BACKGROUND
Surgically assessed pancreatic texture has been identified as the strongest predictor of postoperative pancreatic fistula. However, texture is a subjective parameter with no proven reliability or validity. Therefore, a more objective parameter is needed. In this study, we evaluated the fibrosis level at the pancreatic neck resection margin and correlated fibrosis and all clinico-pathologic parameters collected over the course of the Pancreatogastrostomy vs Pancreatojejunostomy for RECOnstruction (RECOPANC) study.
STUDY DESIGN
The RECOPANC trial was a multicenter randomized prospective trial of patients undergoing pancreatoduodenectomy. There were 261 hematoxylin and eosin-stained slides allocated for histopathologic analyses. Pancreatic fibrosis was scored from 0 to III (no fibrosis up to severe fibrosis) by 2 blinded independent pathologists. All variables possibly associated with POPF were entered into a generalized linear model for multivariable analysis.
RESULTS
The fibrosis grade and pancreatic texture were scored in all 261 patients. In POPF B/C (postoperative pancreatic fistula grade B or C) patients, 71% had a soft pancreas, and fibrosis grades were distributed as follows: 48% with score 0, 28% with score I, 20% with score II, and 7% with score III, respectively. Fibrosis grading showed substantial inter-rater reliability (kappa = 0.74) and correlated positively with hard pancreatic texture (p < 0.05). In univariable analysis, area under the curve (AUC) for POPF B/C prediction was higher for fibrosis grade than for pancreatic texture (0.71 vs 0.59). In multivariate analysis, the following predictors were selected: sex, surgeon volume, pancreatic texture, and fibrosis grade. However, the addition of pancreatic texture only led to an incremental improvement (AUC 0.794 vs 0.819).
CONCLUSIONS
Histologically evaluated pancreatic fibrosis is an easily applicable and highly reproducible POPF predictor and superior to surgically evaluated pancreatic texture. Future studies might use fibrosis grade for risk stratification in pancreatoduodenectomy.
Identifiants
pubmed: 33887486
pii: S1072-7515(21)00240-4
doi: 10.1016/j.jamcollsurg.2021.03.024
pii:
doi:
Types de publication
Comparative Study
Journal Article
Multicenter Study
Randomized Controlled Trial
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
935-945.e2Investigateurs
Louisa Bolm
(L)
Ruediger Braun
(R)
Hryhoriy Lapshyn
(H)
Commentaires et corrections
Type : CommentIn
Informations de copyright
Copyright © 2021 American College of Surgeons. Published by Elsevier Inc. All rights reserved.