BACKGROUND: Despite numerous studies of peripancreatic inflammatory fluid collection (PIFC) that report on the relevance of the drain amylase concentration (D-AMY), early prediction using this assay is problematic. This study aimed to investigate the clinical significance of measuring the D-AMY at 3 h after gastrectomy (POD0) for gastric cancer. METHODS: This retrospective analysis included consecutive patients who underwent gastrectomy combined with peripancreatic lymph node dissection. The predictive value of D-AMY on POD0 and postoperative day 1 (POD1) for clinically relevant PIFC was evaluated together or individually. RESULTS: Analyses were performed in 204 patients. Twenty (9.8%) patients experienced PIFC. D-AMY cutoffs of 721 IU/L on POD0 and 1695 IU/L on POD1 were determined using the receiver operating characteristic curve analysis for predicting PIFC. The D-AMY on POD0 had higher sensitivity (80%) but lower specificity (66.3%) for prediction of PIFC, compared with those of D-AMY on POD1 (65%, 89.1%, respectively). When combination marker analysis was performed, the highest risk group (D-AMY ≥ the cutoff values of POD0 and POD1) were associated with an elevated rate of occurrence (44%) and a high positive likelihood ratio (7.36) compared with those of the single cutoff group. The lowest risk group (D-AMY < the cutoff values on POD0 and POD1) was associated with a low rate of occurrence (2.5%) and low negative likelihood ratio (0.24) compared with those of the single cutoff group. CONCLUSIONS: Combined measurements of D-AMYs on POD0 and POD1 enhanced early prediction of PIFC after gastrectomy.
BACKGROUND: Despite numerous studies of peripancreatic inflammatory fluid collection (PIFC) that report on the relevance of the drain amylase concentration (D-AMY), early prediction using this assay is problematic. This study aimed to investigate the clinical significance of measuring the D-AMY at 3 h after gastrectomy (POD0) for gastric cancer. METHODS: This retrospective analysis included consecutive patients who underwent gastrectomy combined with peripancreatic lymph node dissection. The predictive value of D-AMY on POD0 and postoperative day 1 (POD1) for clinically relevant PIFC was evaluated together or individually. RESULTS: Analyses were performed in 204 patients. Twenty (9.8%) patients experienced PIFC. D-AMY cutoffs of 721 IU/L on POD0 and 1695 IU/L on POD1 were determined using the receiver operating characteristic curve analysis for predicting PIFC. The D-AMY on POD0 had higher sensitivity (80%) but lower specificity (66.3%) for prediction of PIFC, compared with those of D-AMY on POD1 (65%, 89.1%, respectively). When combination marker analysis was performed, the highest risk group (D-AMY ≥ the cutoff values of POD0 and POD1) were associated with an elevated rate of occurrence (44%) and a high positive likelihood ratio (7.36) compared with those of the single cutoff group. The lowest risk group (D-AMY < the cutoff values on POD0 and POD1) was associated with a low rate of occurrence (2.5%) and low negative likelihood ratio (0.24) compared with those of the single cutoff group. CONCLUSIONS: Combined measurements of D-AMYs on POD0 and POD1 enhanced early prediction of PIFC after gastrectomy.