Clinicopathological characteristics

The distribution of fibrinogen, pre-albumin, neutrophils, and lymphocytes for the entire cohort are provided in the Supplementary Table 1. The optimal cutoff point of FPR was established at 14.77, leading to the stratification of the study cohort into two distinct groups: a high FPR group (FPR ≥ 14.77) and a low FPR group (FPR < 14.77), as depicted in Fig. 1. The demographic and clinical profiles of the 563 participants incorporated in this analysis are comprehensively tabulated in Table 1. A comparative evaluation revealed that the high FPR cohort predominantly comprised patients with tumors located in the pancreatic head (p < 0.001) and exhibited adverse clinical features, as evidenced by higher pN stages (p = 0.010), elevated NLR (p < 0.001), and increased PLR (p < 0.001).

Fig. 1

The optimal cutoff point of fibrinogen-to-prealbumin ratio

Table 1 Characteristics of the 575 patients with PDACSurvival analysis of RFS and OS across two groups

The Kaplan–Meier method was employed to assess and contrast the survival outcomes between patients stratified into high and low FPR groups. Analysis encompassing the entire cohort revealed a significantly poorer prognosis associated with the high FPR group, as evidenced by both diminished RFS (median RFS: 9 months for the high FPR group vs. 16 months for the low FPR group, p < 0.001, Fig. 2A) and reduced OS (median OS: 18 months for the high FPR group vs. 29 months for the low FPR group, p < 0.001, Fig. 2B). Next, we will employ PSM to balance the baseline data of two groups (Supplementary Table 2). In the PSM cohort, survival analysis reaffirmed that patients exhibiting high FPR were associated with inferior outcomes in terms of RFS (median RFS: 9 months for the high FPR group vs. 16 months for the low FPR group, p < 0.001, Fig. 2C) and OS (median OS: 20 months for the high FPR group vs. 28 months for the low FPR group, p < 0.001, Fig. 2D). Further exploratory subgroup analyses elucidated that patients within the high FPR cohort consistently exhibited inferior RFS and OS compared to their low FPR counterparts across tumor stages pT1 to pT3 (Fig. 3A-3F, all p < 0.05). These findings underscore the pronounced prognostic significance of FPR within the PDAC patient population, highlighting its potential as a pivotal prognostic marker.

Fig. 2

The Kaplan–Meier curves of recurrence‐free survival (A) and overall survival (B) according to preoperative level of fibrinogen-to-prealbumin ratio in the whole cohort before propensity score matching analysis; The Kaplan–Meier curves of recurrence‐free survival (C) and overall survival (D) according to preoperative level of fibrinogen-to-prealbumin ratio in the whole cohort after propensity score matching analysis

Fig. 3

The Kaplan–Meier curves of recurrence‐free survival and overall survival according to preoperative level of fibrinogen-to-prealbumin ratio in pT1 stage (A-B), pT2 stage (C-D), and pT3 stage (E–F)

Univariate and multivariate cox regression analyses of FPR and other prognostic factors for RFS and OS

Optimal cutoff points for the NLR and PLR were established at 1.82 and 170.8, respectively, as illustrated in Supplemental Fig. 2. The subsequent univariate and multivariate Cox regression analyses, aimed at elucidating the prognostic implications of various clinical parameters for RFS and OS within the entire patient cohort, are delineated in Tables 2 and 3. The univariate analysis pertaining to RFS identified significant associations with several variables, including preoperative levels of CA19-9, CEA, NLR, and PLR, alongside tumor differentiation, pN stage, lymphovascular and extrapancreatic invasion, perineural invasion, resectability status, and the administration of adjuvant chemotherapy. Subsequent multivariate analysis refined these findings, pinpointing preoperative CA19-9 and CEA levels, NLR, FPR, tumor differentiation, lymphovascular invasion, resectability status, and adjuvant chemotherapy as independent prognostic indicators for RFS.

Table 2 Univariate and multivariate Cox regression of prognostic variables for recurrence-free survival in the cohortTable 3 Univariate and multivariate Cox regression of prognostic variables for overall survival in the cohort

Similarly, the univariate analysis for OS established significant correlations with preoperative CA19-9 and CEA levels, NLR, PLR, FPR, tumor differentiation, pT stage, pN stage, lymphovascular and extrapancreatic invasion, resectability status, and the use of adjuvant chemotherapy. Further refinement through multivariate analysis highlighted preoperative CA19-9 levels, NLR, FPR, tumor differentiation, resectability status, and adjuvant chemotherapy as independent determinants of OS. Thus, the analyses conclusively demonstrate the FPR as an independent prognostic factor affecting both RFS and OS in patients undergoing resection for PDAC, underscoring its potential utility in the prognostic stratification and management of PDAC patients.

Evaluating the prognostic predictive efficacy of FPR relative to NLR

Within this study, the NLR, a conventional marker of inflammation, emerged as an independent prognostic variable influencing both RFS and OS. To ascertain the relative prognostic precision of the FPR vs NLR, we employed time-dependent ROC curve analyses. These analyses illuminated that FPR consistently exhibited a superior area under the curve (AUC), indicative of enhanced predictive capability for 2-year RFS and 3-year OS, in comparison to NLR (Fig. 4A-B).

Fig. 4

A-B The time-dependent receiver operating characteristic curves of 2-year recurrence-free survival and 3-year overall survival for neutrophil-to-lymphocyte ratio and fibrinogen-to-prealbumin ratio. C-D The decision curve analysis of 2-year recurrence-free survival and 3-year overall survival for neutrophil-to-lymphocyte ratio and fibrinogen-to-prealbumin ratio

Subsequently, DCAs were conducted to meticulously assess and juxtapose the clinical applicability and advantages proffered by FPR and NLR. The DCAs revealed that, across a spectrum of risk thresholds, FPR conferred greater net benefits in terms of 2-year RFS and 3-year OS predictions relative to NLR (Fig. 4C-D), thereby underscoring the superior prognostic utility of FPR in the context of PDAC. Finally, we integrated NLR and FPR to stratify patients into four distinct groups. As depicted in the Supplementary Fig. 3A-B, patients characterized by elevated FPR and NLR demonstrated significantly shorter RFS and OS. This trend was followed by those with high FPR and low NLR, then individuals exhibiting high NLR and low PLR, with the group having low FPR and NLR showing the most favorable outcomes.