Li-Peng Zhang, Hu

Ren, Yong-Xing Du, Cheng-Feng Wang

Abstract

Key Words: Pancreatic ductal adenocarcinoma; R0 resection; Fibrinogen-to-albumin ratio; Biomarker; Prognosis; Survival

INTRODUCTION

Of all the malignancies, pancreatic cancer has the worst prognosis, with mortality roughly comparable to the incidence[1]. Pancreatic cancer is the sixth most common cause of cancer-related deaths (2.82%) in China[2]. Globally, pancreatic cancer is the fourth leading cause of cancer-related death[3]. With the annual increase in pancreatic cancer-related deaths, it is estimated that by 2030, just a decade from now, malignant tumors will be the second leading cause of cancer-related deaths in the United States[4]. Based on the latest data from the Surveillance, Epidemiology, and End Results database of the National Cancer Institute, the 5-year overall survival (OS) rate for pancreatic cancer patients is only 10% in the United States[5,6]. At present, radical surgical resection is still an important method, or even the only way, to prolong significantly the survival time of pancreatic cancer patients[7]. Nevertheless the 5-year OS rate of patients undergoing radical resection is still less than 20%[8,9], despite the concept of surgery-based comprehensive treatment being constantly practiced and updated in clinical work[6]. Moreover, only 20% of patients diagnosed with pancreatic cancer have access to surgical treatment[10]. The poor prognosis of pancreatic cancer patients is related to the early asymptomatic nature of the disease, leading to the late stage of disease diagnosis and the high possibility of distant metastasis in the early stage of pancreatic cancer[11]. Since pancreatic cancer has a poor overall prognosis and many patients may choose to refuse treatment, it is particularly important to explore simple and cost-effective biomarkers to help better predict patient OS before treatment, especially for patients who have the opportunity to undergo surgery.

The systemic inflammatory response is a key factor in promoting the occurrence and metastasis of malignant tumors, including pancreatic cancer[12-14]. Pancreatic cancer can also lead to a hypercoagulable state and cause obvious thrombosis in the patient[15], in which fibrinogen plays a vital role. At the same time, the plasma albumin level is also affected by the systemic inflammatory response[16], and it is also a key indicator of nutritional status in malignant tumor patients[17,18].

Recently, growing evidence has confirmed that fibrinogen-to-albumin ratio (FAR), the ratio of preoperative fibrinogen concentration to plasma albumin level, is an important predictor of the OS of patients with esophageal cancer[19], breast cancer[20,21], and liver cancer[22]after radical surgery. However, to date, there have been no reports discussing the association between FAR and the prognosis of patients with pancreatic ductal adenocarcinoma (PDAC) undergoing radical R0 resection. Here, we conducted a retrospective study to explore whether FAR has prognostic value in these patients.

MATERIALS AND METHODS

Patients

The inclusion criteria were as follows: (1) Positron emission tomography/computed tomography (CT) examination or enhanced magnetic resonance imaging and/or CT scan confirmed no distant metastasis; (2) Radical pancreaticoduodenectomy or distal pancreatectomy with splenectomy was performed; (3) The pathologic diagnosis was PDAC and radical R0 resection (no tumor cells within 1 mm from the resection margins) was performed[23]; and (4) Complete clinical information and follow-up data were available.

The exclusion criteria were as follows: (1) A history of cancer or other malignant tumors at the same time; (2) Received neoadjuvant therapy before surgery; (3) Presence of acute or chronic inflammatory diseases; (4) Presence of anemia or other blood system diseases; (5) Received anticoagulant therapy or albumin infusion before surgery; (6) Abnormal liver function or complicated with other liver diseases; and (7) Death occurred during the perioperative period, 1 mo after the operation.

In total, we retrospectively analyzed the data of 282 patients with PDAC who underwent radical R0 resection at The National Cancer Center/Cancer Hospital of the Chinese Academy of Medical Sciences from January 2010 to December 2019.

Data collection

All clinicopathological characteristics included age, sex, blood type, clinical symptoms (including jaundice, pain, digestive symptoms, weight loss, fatigue,etc.), diabetes, smoking status, alcohol consumption, family history of cancer, approaches of open surgery, tumor information (including tumor location, degree of differentiation, lymphovascular invasion, perineural invasion, capsular invasion, maximum tumor diameter, number of metastatic lymph nodes), T stage, N stage, Tumor-Node-Metastasis (TNM) stage, adjuvant therapy, preoperative cancer antigen 19-9 ( CA19-9 ) level (U/mL), preoperative plasma fibrinogen concentration (g/L), preoperative plasma albumin level (g/L), and preoperative FAR.

The age of the patients refers to the age at which the primary ductal adenocarcinoma of the pancreas was clearly diagnosed. T staging, N staging, and TNM staging were defined according to the eighth edition of the TNM classification standard of the American Joint Committee on Cancer[24].

Ethical statement

The study was conducted according to the ethical standards of the World Medical Association Declaration of Helsinki (9thedition of July 2018) and was approved by the Medical Ethics Committee of The National Cancer Center/Cancer Hospital of the Chinese Academy of Medical Sciences (approval No. 17-168/1424). All patients in the group also signed a written informed consent form.

Fibrinogen and albumin measurements

The patient’s blood samples were collected before breakfast within 7 d before operation and processed quickly (within 2 h). Fibrinogen was processed and measured by a CA7000 automatic coagulation analyzer (Sysmex Corporation, Kobe, Japan). Serum albumin was measured by the bromocresol green dye method. The normal reference value of plasma fibrinogen concentration is 2.0-4.0 g/L, and the normal reference value of plasma albumin is 40.0-55.0 g/L.

FAR

FAR was defined as the ratio of the preoperative plasma fibrinogen concentration to the preoperative plasma albumin level.

Treatments and follow-up

All patients underwent radical surgery for PDAC, including pancreaticoduodenectomy and distal pancreatectomy with splenectomy, at The National Cancer Center/Cancer Hospital of Chinese Academy of Medical Sciences.

All patients were followed up by reviewing outpatient medical records and by telephone follow-up. The patients were followed up every 3 mo for the first 2 years and every 6 mo thereafter, and we recorded the results of physical, laboratory, and imaging examinations in detail. The date on which the patient underwent surgery was the beginning of the follow-up period, and the OS time was defined from the date of surgery to the date of death or the deadline for follow-up. The last follow-up date was August 16, 2020.

Statistical analysis

Continuous data with a normal distribution are shown as the mean ± standard deviation (Kolmogorov-Smirnov test,P> 0.05), and those with a nonnormal distribution are expressed as the median (minimum-maximum). Frequencies and percentages were used for the categorical variables. Statistical analyses were performed using RStudio software (version 1.3.1073, http://www.rstudio.org), SPSS software (version 25.0; IBM Corp., Armonk, NY, United States), and GraphPad Prism software (version 8.02; GraphPad Software, Inc., La Jolla, CA, United States). The surv_cutpoint function was used to determine the optimal cut-off value for preoperative plasma fibrinogen, albumin, and FAR. Clinicopathological characteristics at baseline were compared by the chi-square test and independent samplest-test. The Kaplan-Meier method was used to calculate the cumulative survival rate, and the logrank test was used to compare survival curves. A multivariate Cox proportional hazards model was used to test further prognostic factors proven statistically significant in the univariate analysis. The results were calculated with 95% confidence intervals (CIs), and a two-tailedP< 0.05 was considered statistically significant.

RESULTS

Patient characteristics

A total of 282 patients with PDAC who underwent radical R0 resection at The National Cancer Center/Cancer Hospital of the Chinese Academy of Medical Sciences from January 2010 to December 2019 were included in this retrospective study, with a median follow-up of 14.98 mo. A total of 217 patients died during the follow-up period, with an estimated median OS of 17.43 mo (range: 1.30-100.07 mo). The 1-, 2-, 3-, and 5-year postoperative survival rates were 67.5%, 35.9%, 20.4%, and 10.2%, respectively. We analyzed the clinical indicators of all patients who met the inclusion criteria and found that the median age of these patients at diagnosis was 61 years (range: 31-81 years), with 136 (48.2%) of the patients being over 60 years of age. Of these patients, 151 (53.5%) were male, and 225 (79.8%) patients were seen for the presence of clinical symptoms that included jaundice, pain, digestive symptoms, weight loss, and fatigue. A total of 130 (46.1%) patients had tumors located in the head or neck of the pancreas and underwent pancreaticoduodenectomy, and 158 (56.0%) patients received further adjuvant therapy after surgery. A total of 217 (77.0%) patients had a histologic diagnosis of moderate differentiation. Based on the TNM staging criteria, only 29 (10.3%) patients were classified as stage III. Details of the patients' baseline characteristics are shown in Table 1.

Determination of optimal cut-off values for survival analysis

Based on the R programming language, R packages offer a collection of functions and data sets. The R package survminer provides convenient survival analysis and visualization capabilities, in which the surv_cutpoint function determines the optimal cut-off value for one or more continuous variables and plots survival curves for these variables, calculating the hazard ratio and the upper and lower limits of the 95%CIs. We used the surv_cutpoint function to determine the optimal cut-off values for preoperative fibrinogen, albumin, and FAR, which were 3.31 g/L, 45.2 g/L, and 0.08, respectively, and the optimal cut-off values for these indicators were also validated by their respective receiver operating characteristic curves, as shown in Figure 1.

Table 1 Baseline characteristics of 282 pancreatic ductal adenocarcinoma patients who underwent R0 resection, n (%)

Present 212 (75.2)Capsular invasion Absent 49 (17.4)Present 233 (82.6)Maximum tumor diameter in cm> 4 88 (31.2)≤ 4 194 (68.8)T stage T1 34 (12.1)T2 159 (56.4)T3 89 (31.6)Number of metastatic lymph nodes Absent 161 (57.1)Present 121 (42.9)N stage N0 161 (57.1)N1 92 (32.6)N2 29 (10.3)TNM stage IA 20 (7.1)IB 92 (32.6)IIA 49 (17.4)IIB 92 (32.6)III 29 (10.3)Adjuvant therapy Absent 124 (44.0)Present 158 (56.0)Preoperative CA19-9 level in U/mL 172.4 (0.6-55412.0)> 336.4 77 (27.3)≤ 336.4 205 (72.7)Preoperative plasma fibrinogen concentration in g/L 3.02 (1.20-6.70)> 3.31 141 (50.0)≤ 3.31 141 (50.0)Preoperative plasma albumin level in g/L 42.6 (23.2-54.0)> 45.2 85 (30.1)≤ 45.2 197 (69.9)Preoperative FAR 0.07 (0.03-0.21)> 0.08 126 (44.7)≤ 0.08 156 (55.3)FAR: Fibrinogen-to-albumin ratio; TNM: Tumor-Node-Metastasis.

Figure 1 Receiver operating characteristics curve analysis based on preoperative fibrinogen concentration, albumin level, and fibrinogen-to-albumin ratio for overall survival. A: The area under the receiver operating characteristics curve (AUC) indicates the diagnostic power of preoperative plasma fibrinogen concentration. In this model, the optimum cut-off point for fibrinogen concentration was 3.31 g/L, AUC was 0.714 (95% confidence interval (CI): 0.649-0.779), with a sensitivity of 61.14% and a specificity of 79.37% by the Youden’s index; B: The AUC indicates the diagnostic power of preoperative plasma albumin level. In this model, the optimum cut-off point for albumin level was 45.2 g/L, AUC was 0.659 (95%CI: 0.580-0.738), with a sensitivity of 75.58% and a specificity of 53.85% by the Youden’s index; C: The AUC indicates the diagnostic power of preoperative fibrinogen-to-albumin ratio (FAR). In this model, the optimum cut-off point for FAR was 0.08, AUC was 0.743 (95%CI: 0.677-0.810), with a sensitivity of 59.91% and a specificity of 87.69% by the Youden’s index; D: By comparing AUC, the diagnostic power of FAR was the most powerful.

The median preoperative fibrinogen concentration for all patients included in this study was 3.02 g/L (range: 1.20-6.70 g/L) (Table 1), with an optimal cut-off value of 3.31 g/L. As shown in Figure 1A, the area under the curve (AUC) of preoperative fibrinogen was 0.714 (95%CI: 0.649-0.779), and the sensitivity and specificity at the maximum Youden’s Index were 61.14% and 79.37%, respectively. Based on this cut-off value, 141 patients had preoperative fibrinogen concentrations > 3.31/L, as detailed in Table 2.

The median preoperative albumin level for all patients enrolled in this study was 42.6 g/L (range: 23.2-54.0 g/L) (Table 1), and the optimal cut-off value was 45.2 g/L. As shown in Figure 1B, the AUC of preoperative albumin was 0.659 (95%CI: 0.580-0.738), and the sensitivity and specificity at the maximum Youden’s Index were 75.58% and 53.85%, respectively. Based on this cut-off value, 197 patients had a preoperative albumin level ≤ 45.2 g/L, and 85 patients had a preoperative albumin level > 45.2 g/L, the details of which are shown in Table 3.

The median preoperative FAR for all patients included in this retrospective study was 0.07 (range: 0.03-0.21) (Table 1), and the optimal cut-off value was 0.08. As shown in Figure 1C, the AUC of preoperative FAR was 0.743 (95%CI: 0.677-0.810), with a sensitivity and specificity of 59.91% and 87.69%, respectively, at the maximum Youden’s Index. Based on this optimal cut-off value, 156 patients had a preoperativeFAR ≤ 0.08, and 126 patients had a preoperative FAR > 0.08, as detailed in Table 4.

Table 2 Correlation between preoperative plasma fibrinogen concentration and clinicopathological characteristics in pancreatic ductal adenocarcinoma patients who underwent R0 resection, n (%)

Perineural invasion 0.054 Absent 28 (9.9)42 (14.9)Present 113 (40.1)99 (35.1)Capsular invasion 0.271 Absent 21 (7.4)28 (9.9)Present 120 (42.6)113 (40.1)Maximum tumor diameter in cm 1.000> 4 44 (15.6)44 (15.6)≤ 4 97 (34.4)97 (34.4)T stage 0.991 T1 17 (6.0)17 (6.0)T2 80 (28.4)79 (28.0)T3 44 (15.6)45 (16.0)Number of metastatic lymph nodes 0.718 Absent 82 (29.1)79 (28.0)Present 59 (20.9)62 (22.0)N stage 0.110 N0 82 (29.1)79 (28.0)N1 40 (14.2)52 (18.4)N2 19 (6.7)10 (3.5)TNM stage 0.099 IA 8 (2.8)12 (4.3)IB 44 (15.6)48 (17.0)IIA 30 (10.6)19 (6.7)IIB 40 (14.2)52 (18.4)III 19 (6.7)10 (3.5)Adjuvant therapy 0.631 Absent 64 (22.7)60 (21.3)Present 77 (27.3)81 (28.7)Preoperative CA19-9 level in U/mL 0.023> 336.4 47 (16.7)30 (10.6)≤ 336.4 94 (33.3)111 (39.4)Preoperative plasma albumin level in g/L 0.006> 45.2 32 (11.3)53 (18.8)≤ 45.2 109 (38.2)88 (31.2)Preoperative FAR < 0.001> 0.08 114 (40.4)12 (4.3)≤ 0.08 27 (9.6)129 (45.7)FAR: Fibrinogen-to-albumin ratio; TNM: Tumor-Node-Metastasis.

Correlations of indicators with clinicopathological factors

As shown in Table 2, all patients could be categorized into low or high value groupsbased on the optimal cut-off value of preoperative fibrinogen concentration. The increase in preoperative plasma fibrinogen concentration was significantly correlated with clinical symptoms, tumor location, surgical approaches, preoperative CA19-9 level, preoperative plasma albumin level, and preoperative FAR. However, preoperative fibrinogen concentration did not significantly correlate with age, sex,blood type, diabetes, smoking status, alcohol consumption, family history of cancer, degree of tumor differentiation, lymphovascular invasion, perineural invasion, capsular invasion, maximum tumor diameter, number of metastatic lymph nodes, T stage, N stage, TNM stage, or adjuvant therapy. The survival curves for preoperative fibrinogen concentrations showed shorter OS times in patients with preoperative fibrinogen concentrations > 3.31 g/L than in patients with preoperative fibrinogen concentrations ≤ 3.31 g/L (Figure 2A).

Table 3 Correlation between preoperative plasma albumin level and clinicopathological characteristics in pancreatic ductal adenocarcinoma patients who underwent R0 resection, n (%)

Perineural invasion 0.741 Absent 20 (7.1)50 (17.7)Present 65 (23.0)147 (52.1)Capsular invasion 0.673 Absent 16 (5.7)33 (11.7)Present 69 (24.5)164 (58.2)Maximum tumor diameter in cm 0.330> 4 30 (10.6)58 (20.6)≤ 4 55 (19.5)139 (49.3)T stage 0.530 T1 8 (2.8)26 (9.2)T2 47 (16.7)112 (39.7)T3 30 (10.6)59 (20.9)Number of metastatic lymph nodes 0.363 Absent 52 (18.4)109 (38.7)Present 33 (11.7)88 (31.2)N stage 0.127 N0 52 (18.4)109 (38.7)N1 29 (10.3)63 (22.3)N2 4 (1.4)25 (8.9)TNM stage 0.145 IA 4 (1.4)16 (5.7)IB 34 (12.1)58 (20.6)IIA 14 (5.0)35 (12.4)IIB 29 (10.3)63 (22.3)III 4 (1.4)25 (8.9)Adjuvant therapy 0.003 Absent 26 (9.2)98 (34.8)Present 59 (20.9)99 (35.1)Preoperative CA19-9 level in U/mL 0.416> 336.4 26 (9.2)51 (18.1)≤ 336.4 59 (20.9)146 (51.8)Preoperative plasma fibrinogen concentration in g/L 0.006> 3.31 32 (11.3)109 (38.7)≤ 3.31 53 (18.8)88 (31.2)Preoperative FAR < 0.001> 0.08 15 (5.3)111 (39.4)≤ 0.08 70 (24.8)86 (30.5)FAR: Fibrinogen-to-albumin ratio; TNM: Tumor-Node-Metastasis.

Table 4 Correlation between preoperative fibrinogen-to-albumin ratio and clinicopathological characteristics in pancreatic ductal adenocarcinoma patients who underwent R0 resection, n (%)

Perineural invasion 0.082 Absent 25 (8.9)45 (16.0)Present 101 (35.8)111 (39.4)Capsular invasion 0.360 Absent 19 (6.7)30 (10.6)Present 107 (37.9)126 (44.7)Maximum tumor diameter in cm 0.391> 4 36 (12.8)52 (18.4)≤ 4 90 (31.9)104 (36.9)T stage 0.462 T1 13 (4.6)21 (7.4)T2 76 (27.0)83 (29.4)T3 37 (13.1)52 (18.4)Number of metastatic lymph nodes 0.988 Absent 72 (25.5)89 (31.6)Present 54 (19.1)67 (23.8)N stage 0.902 N0 72 (25.5)89 (31.6)N1 40 (14.2)52 (18.4)N2 14 (5.0)15 (5.3)TNM stage 0.494 IA 6 (2.1)14 (5.0)IB 40 (14.2)52 (18.4)IIA 26 (9.2)23 (8.2)IIB 40 (14.2)52 (18.4)III 14 (5.0)15 (5.3)Adjuvant therapy 0.531 Absent 58 (20.6)66 (23.4)Present 68 (24.1)90 (31.9)Preoperative CA19-9 level in U/mL 0.132> 336.4 40 (14.2)37 (13.1)≤ 336.4 86 (30.5)119 (42.2)Preoperative plasma fibrinogen concentration in g/L < 0.001> 3.31 114 (40.4)27 (9.6)≤ 3.31 12 (4.3)129 (45.7)Preoperative plasma albumin level in g/L < 0.001> 45.2 15 (5.3)70 (24.8)≤ 45.2 111 (39.4)86 (30.5)FAR: Fibrinogen-to-albumin ratio; TNM: Tumor-Node-Metastasis.

As shown in Table 3, all patients could be categorized into a low value group (≤ 45.2 g/L) or a high value group (> 45.2 g/L) based on the optimal cut-off value of preoperative plasma albumin level. The increase in preoperative albumin level was significantly correlated with clinical symptoms (P= 0.012), tumor location (P< 0.001), surgical approaches (P< 0.001), adjuvant therapy (P= 0.003), preoperative fibrinogen concentration (P= 0.006), and preoperative FAR (P< 0.001). The survival curves for preoperative albumin level showed longer OS times in patients with preoperative plasma albumin level > 45.2 g/L than in those with preoperative albumin level ≤ 45.2 g/L (Figure 2B).

As shown in Table 4, according to the optimal cut-off value of preoperative FAR, all patients can be divided into a low-value group (≤ 0.08) or a high-value group (> 0.08). Higher preoperative FAR was significantly associated with clinical symptoms (P= 0.001), tumor location (P< 0.001), surgical approaches (P< 0.001), preoperative plasma fibrinogen concentration (P< 0.001), and preoperative plasma albumin level (P< 0.001). The survival curves for preoperative FAR showed shorter OS times in patients with FAR > 0.08 than in patients with FAR ≤ 0.08 (Figure 2C).

Univariate analysis and multivariate analysis results

In the univariate Cox analysis (Table 5), age [hazard ratio (HR): 1.358, 95%CI: 1.036-1.780,P= 0.027], clinical symptoms (HR: 0.600, 95%CI: 0.424-0.848,P= 0.004), degree of tumor differentiation (P< 0.001), capsular invasion (HR: 0.609, 95%CI: 0.420- 0.885,P= 0.009), maximum tumor diameter (HR: 1.403, 95%CI: 1.058-1.862,P= 0.019), T stage (P= 0.035), number of metastatic lymph nodes (HR: 0.590, 95%CI: 0.449-0.775,P< 0.001), N stage (P= 0.001), TNM stage (P= 0.003), adjuvant therapy (HR: 1.625, 95%CI: 1.244-2.123,P< 0.001), preoperative CA19-9 level (HR: 1.971, 95%CI: 1.469-2.644,P< 0.001), preoperative plasma fibrinogen concentration (HR: 1.888, 95%CI: 1.438-2.479,P< 0.001), preoperative plasma albumin level (HR: 0.650, 95%CI: 0.475-0.890,P= 0.007), and preoperative FAR (HR: 2.257, 95%CI: 1.725-2.952,P< 0.001) were important prognostic risk factors for OS in patients with PDAC who underwent radical R0 resection.

However, sex, blood type, diabetes, smoking status, alcohol consumption, family history of cancer, surgical approaches, tumor location, lymphovascular invasion, and perineural invasion were not prognostic risk factors for OS in patients with PDAC who underwent radical R0 resection (P> 0.05).

In the multivariate Cox regression analysis (Table 6), degree of tumor differentiation (P< 0.001), number of metastatic lymph nodes (HR: 0.678, 95%CI: 0.509-0.904,P= 0.008), adjuvant therapy (HR: 1.604, 95%CI: 1.214-2.118,P= 0.001), preoperative CA19-9 level (HR: 1.740, 95%CI: 1.288-2.352,P< 0.001), and preoperative FAR (HR: 2.258, 95%CI: 1.720-2.963,P< 0.001) were independent risk factors for poor prognosis in patients with PDAC who underwent radical R0 resection.

DISCUSSION

Our results showed that preoperative FAR was closely related to OS in patients with PDAC after radical R0 resection. To our knowledge, this is the first report of FAR in patients with PDAC who underwent radical R0 resection. A higher preoperative FAR predicted poor prognosis in these patients. Further multivariate analysis showed that preoperative FAR (P< 0.001), degree of differentiation (P< 0.001), number of metastatic lymph nodes (P= 0.008), adjuvant therapy (P= 0.001), and preoperative CA19-9 level (P< 0.001) were independent prognostic factors for the OS of patients with PDAC after radical R0 resection.

As a newly discovered tumor prognostic marker, the prognostic value of preoperative FAR in patients with malignant tumors undergoing radical surgery is increasing. The prognostic role of FAR has been confirmed in patients who underwent radical resection of tumors, such as esophageal cancer[19], breast cancer[20,21]and liver cancer[22]. It has also been reported that FAR (HR: 1.634; 95%CI: 1.359-1.964;P< 0.001) is an important prognostic factor for locally advanced or metastatic pancreatic cancer[25]. With regard to the optimal cut-off value of FAR, the value in patients who underwent radical operation for esophageal cancer was the same as ours, whereas that in patients who underwent radical operation for liver cancer was 0.062, which was lower than our value. The optimal cut-off value of FAR in patients with locally advanced or metastatic pancreatic cancer was 0.079, which was similar to our value.However, the reported optimal cut-off value of preoperative FAR in patients who underwent radical mastectomy was between 0.071 and 0.084, indicating that the optimal cut-off value may be different even in different patient groups with the sametumor type. The reasons for the above differences may be related to the different biological behaviors of different tumors or the selection bias of different patient groups with the same tumor type, so more basic and clinical studies are needed to verify this. Although preoperative FAR has a clear role in evaluating the prognosis of patients undergoing radical resection of malignant tumors[19-22], the molecular mechanisms related to the occurrence and development of tumors are still not completely clear.

Table 5 Univariate analysis for overall survival in pancreatic ductal adenocarcinoma patients who underwent R0 resection

Present Capsular invasion 0.609 (0.420-0.885)0.009 Absent Present Maximum tumor diameter in cm 1.403 (1.058-1.862)0.019> 4≤ 4 T stage —0.035 T1 T2 T3 Number of metastatic lymph nodes 0.590 (0.449-0.775)< 0.001 Absent Present N stage 0.001 N0 N1 N2 TNM stage 0.003 IA IB IIA IIB III Adjuvant therapy 1.625 (1.244-2.123)< 0.001 Absent Present Preoperative CA19-9 level in U/mL 1.971 (1.469-2.644)< 0.001> 336.4≤ 336.4 Preoperative plasma fibrinogen concentration in g/L 1.888 (1.438-2.479)< 0.001> 3.31≤ 3.31 Preoperative plasma albumin level in g/L 0.650 (0.475-0.890)0.007> 45.2≤ 45.2 Preoperative FAR 2.257 (1.725-2.952)< 0.001> 0.08≤ 0.08 CI: Confidence interval; FAR: Fibrinogen-to-albumin ratio; HR: Hazard ratio.

Table 6 Multivariate analysis for overall survival in pancreatic ductal adenocarcinoma patients who underwent R0 resection

Fibrinogen is a soluble glycoprotein with a relative molecular mass of 340 kDa that is composed of three different polypeptide chains, namely α, β, and γ. Normally, fibrinogen is synthesized by the liver and released into the bloodstream[26], where it plays an important role in both the inflammatory response and blood coagulation.

Figure 2 Kaplan-Meier curves were generated by grouping preoperative fibrinogen concentration, albumin level, and fibrinogen-toalbumin ratio using the optimal cut-off values. A: In the comparison of preoperative fibrinogen concentration > 3.31 g/L with preoperative fibrinogen concentration ≤ 3.31 g/L (P < 0.05), the blue line represents the > 3.31 g/L group and the red line represents the ≤ 3.31 g/L group; B: In the comparison of preoperative albumin level > 45.2 g/L and preoperative albumin level ≤ 45.2 g/L (P < 0.05), the bule line represents the > 45.2 g/L group and the red line represents the ≤ 45.2 g/L group; C: In the comparison of preoperative fibrinogen-to-albumin ratio (FAR) > 0.08 with preoperative FAR ≤ 0.08 (P < 0.05), the bule line represents the FAR > 0.08 group and the red line represents the FAR ≤ 0.08 group.

Inflammation is an important marker of cancer[14], and numerous clinical and experimental studies have convincingly supported the notion that inflammation is an important component of tumor progression[12,13]. The link between colitis and colon cancer has been fully confirmed, and researchers have directly tested the hypothesis of a link between fibrinogen and colitis-associated colon cancer in mice through animal experiments[27]. Preoperative hyperfibrinogenemia has been reported to be significantly associated with shorter OS (HR: 3.138, 95%CI: 1.077-9.139,P= 0.036) in patients with nonmetastatic colon cancer[28]. Hyperfibrinogenemia is associated with systemic inflammation and can predict poor prognosis in advanced pancreatic cancer (HR: 2.184, 95%CI: 1.574-3.032,P< 0.001)[29]. It has also been reported that fibrinogen concentration is positively correlated with the stage of pancreatic cancer, and the increase in plasma fibrinogen concentration before surgery may be a useful indicator for predicting the distant metastasis of pancreatic cancer (distant metastasis groupvsno distant metastasis group: 4.41 ± 0.84 g/Lvs3.76 ± 1.04 g/L,P< 0.01)[30]. Survival analyses reported in the literature show that plasma fibrinogen concentration is an independent prognostic factor for operable patients with non-small cell lung cancer. Patients with high fibrinogen concentrations (fibrinogenemia) had a higher risk of disease progression (HR: 1.49, 95%CI: 1.07-2.05,P< 0.05) and death (HR: 1.64, 95%CI: 1.06-2.53,P< 0.05) than those with normal fibrinogen concentrations[31]. In a large-scale observation study of 2563 patients with nonmetastatic primary nasopharyngeal carcinoma, a combined increase in Epstein-Barr virus deoxyribonucleic acid copy number and fibrinogen concentration was found to be significantly associated with reduced disease-free survival (HR: 8.51, 95%CI: 5.34-13.56,P< 0.001) and OS (HR 12.77, 95%CI: 5.08-32.12,P< 0.001)[32]. Thus, elevated plasma fibrinogen concentrations before treatment were associated with a significant reduction in the survival of patients with solid tumors, which also suggests that clinical trials are needed to determine whether plasma fibrinogen can be included in the cancer staging system and whether reducing the concentration of plasma fibrinogen has a beneficial effect on disease recurrence and mortality[33].

Pancreatic cancer has also been found to cause hypercoagulability, which may lead to significant thromboembolism in patients with pancreatic cancer, with a reported prevalence of venous thromboembolism of 12%-36%[34,35]. Fibrinogen not only plays a key role in hemostasis but also plays an important role in tumor biology. Fibrin deposition provides a temporary medium-like matrix rich in growth factors that promote the growth of cancer cells. The importance of fibrinogen in tumor biology has been demonstrated in studies of fibrinogen-deficient mice[36]. Studies have shown that fibrinogen can promote platelet adhesion to tumor cells; at the same time, platelets promote the accumulation of more fibrinogen around tumor cells by forming thrombin, and these components cooperate with each other to protect tumor cells from natural killer cells[37]. Fibrinogen can be cleaved by fibrinolytic enzymes, and its protein hydrolysate can promote angiogenesis, which is the premise of tumor growth and metastasis[38]. Fibrinogen has the ability to bind directly to vascular endothelial growth factor (VEGF), platelet-derived growth factor, fibroblast growth factor family members, and transforming growth factor-β, which also reflects the important role fibrinogen plays in angiogenesis, tumor cell proliferation, the hematogenous metastasis of malignant tumors, and epithelial-mesenchymal transformation[39-41]. By inducing epithelial-mesenchymal transformation, fibrinogen promotes the migration of tumor cells and enhances the invasion and metastasis of tumor cells[42].

In cancer patients, the diagnosis of malnutrition is very common and has a very negative impact on prognosis[17]. Progressive malnutrition is closely related to the occurrence and development of cachexia, especially in patients with pancreatic cancer and gastric cancer[43]. Epidemiological studies on the relationship between the level of plasma albumin and the survival of cancer patients have found that the decrease in plasma albumin level is associated with reduced survival in patients[18]. The relationship between pretreatment plasma albumin level and prognosis has been reported in head and neck cancer[44], non-small cell lung cancer[45], kidney cancer[46], gastric cardia adenocarcinoma[47], and ovarian cancer[48]. Inflammation is known to play a critical role in the development, progression, and prognosis of colorectal tumors, and plasma albumin concentration has been shown to be a valuable prognostic factor for colorectal tumors[49]. Tumor cells produce and release proinflammatory cytokines (interleukin (IL)-1, IL-6, tumor necrosis factor-α, interferon-γ) and hormones that are associated with the pathogenesis of both malnutrition and cachexia[43,50], and these inflammatory cytokines can interfere with the normal synthesis of albumin in the liver[16]. Among them, IL-6 has been shown to regulate the secretion of VEGF in glioblastoma cells, while VEGF can lead to an increase in vascular permeability, which indirectly leads to a decrease in plasma albumin level[51].

Therefore, preoperative FAR, as an indicator that reflects the ratio between the fibrinogen concentration and albumin level, can be considered a risk factor for the prognosis of patients undergoing radical R0 resection of PDAC.

This study has some limitations. First, this study is a single-center, retrospective study with a small sample, which may have selection bias. In the future, it is necessary to design more multicenter, prospective studies with a large sample to verify the current conclusions. Second, due to the relatively small number of patients, we did not divide the patients into a training group and test group for statistical verification, and there was a lack of external verification, which needs further study. In our study, we focused on the prognostic significance of preoperative FAR, while the relationship between the change in FAR after surgery and the prognosis of these patients was not been studied. In addition, many patients received multiple treatments due to tumor recurrence during the follow-up period, which also affected OS.

CONCLUSION

In conclusion, in this first report of preoperative FAR in PDAC, we have found that preoperative FAR, defined as the ratio of the fibrinogen concentration to the albumin level, is a useful indicator for assessing the prognosis of patients undergoing radical R0 resection of PDAC. Its practicability, convenience, and low cost make FAR a promising serum biomarker for predicting the prognosis of patients with PDAC. It may provide important help in treatment decision-making for patients with resectable PDAC in the future. However, more relevant studies are needed to validate further the relevant conclusions.

ARTICLE HIGHLIGHTS