Abstract
Background/Aim: The prognosis of patients with pancreatic cancer remains poor, despite recent advances in surgical techniques, perioperative care, neoadjuvant and adjuvant chemotherapy. This study aimed to investigate the preoperative neutrophil-to-lymphocyte ratio (NLR) as a prognostic factor and determine the optimal cutoff value in surgical patients with pancreatic cancer. Patients and Methods: We retrospectively enrolled 461 patients with pancreatic cancer who underwent resection between January 2013 and December 2022 in the Department of Gastrointestinal Surgery at Kanagawa Cancer Center. The association between continuous or categorical variables and NLR was analyzed using the Mann–Whitney U-test and Fisher’s exact test. Overall survival (OS) and relapse-free survival (RFS) rates were calculated using the Kaplan–Meier method. Univariate and multivariate analyses were performed using Cox proportional-hazard regression models. Results: The optimal cutoff value for the preoperative NLR was 3.2. The NLR≥3.2 was associated with a large tumor size (p=0.005), poor histological differentiation (p=0.002), and less adjuvant chemotherapy (p=0.048). The NLR≥3.2 had an important influence on the decreased OS (21.6 vs. 25.8 months), and RFS (10.3 vs. 14.3 months). In univariate and multivariate analyses, the preoperative NLR was an independent prognostic factor for OS (p=0.022) and RFS (p=0.002). Conclusion: Preoperative NLR (cutoff value: 3.2) within two weeks before surgery is a prognostic factor for OS and RFS in surgical patients with pancreatic cancer. This study could help establish evidence on the immune system’s impact and a unified treatment strategy pre-surgery, potentially improving the prognosis for patients with pancreatic cancer.
Pancreatic cancer is the seventh leading cause of cancer-related deaths (1), with more than 330,000 deaths annually worldwide (2). The incidence of pancreatic cancer is increasing annually and is estimated to become the second leading cause of cancer-related deaths by 2030 (3). In addition, the 5-year survival rate is less than 5%, which is the lowest among all cancers (4), and the median survival time is 6-8 months (5).
Only 15-20% of patients with pancreatic cancer are diagnosed with resectable tumors at the initial diagnosis, and surgical residual tumor absent (R0) resection remains the sole curative treatment for this condition (6). The gold standard treatment for pancreatic cancer worldwide, including in Japan, is neoadjuvant chemotherapy (NAC), radical pancreatectomy, and adjuvant chemotherapy (AC). Although recent advances in surgical techniques, perioperative care, NAC, and AC have improved prognosis, the prognosis of patients with pancreatic cancer remains poor. Therefore, searching for prognostic factors is important to improve prognosis.
Currently, the most reliable prognostic factors for pancreatic cancer include lymph node status, tumor size, residual tumors, and histological type (7). However, most of these clinicopathological parameters are determined after surgery and are not valid for predicting the prognosis of patients with pancreatic cancer before treatment. Therefore, simple and practical preoperative biomarkers are required. Recently, inflammation-based prognostic scores, such as the C-reactive protein-to-albumin ratio, neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and modified Glasgow Prognostic Score have attracted attention as prognostic factors for various malignancies, including pancreatic cancer (8-11). Among these, the NLR is considered the most valuable prognostic factor for pancreatic cancer (12-14). However, few studies included surgical patients, with the largest sample size being <400 (8). Furthermore, the timing of NLR evaluation and the optimal cutoff value for NLR have not been established.
The purpose of this study was to investigate whether preoperative NLR is a prognostic factor and to determine the optimal cutoff value in patients with pancreatic cancer who underwent surgical resection at our high-volume center using blood test data within two weeks before surgery.
Patients and Methods
Patients. Between January 2013 and December 2022, patients with pancreatic cancer who underwent pancreatic resection at the Department of Gastrointestinal Surgery, Kanagawa Cancer Center, Kanagawa, Japan were included in this study. All clinical records were retrospectively reviewed. The inclusion criteria were a pathological diagnosis of pancreatic ductal adenocarcinoma (PDAC), pancreatic cancer resection with lymphadenectomy, and pathological residual tumor as R0 or present (R1) resection. Patients with initial diagnosis “unresectable” pancreatic cancer were excluded from the study according to the Japanese General Rules for the Clinical and Pathological Study of Pancreatic Cancer, 7th Edition (JGR). Moreover, patients diagnosed with pathological Stage IV (pStage IV) cancer, or those who underwent R2 resection were also excluded.
Surgical treatment. All patients underwent standard pancreatic resection with lymphadenectomy according to the 7th edition of the JGR, and all procedures were performed by gastrointestinal surgeons. The surgical procedures for pancreatic resection were pancreaticoduodenectomy (PD), distal pancreatectomy (DP), and total pancreatectomy (TP), and were chosen based on the location of the tumor. Our standard procedure for PD consisted of open subtotal stomach-preserving pancreaticoduodenectomy (SSPPD) with digestive tract reconstruction using a modified child procedure.
Patient characteristics and definitions of clinical variables. Patient and clinicopathological characteristics including age, sex, preoperative serum carbohydrate antigen 19-9 (CA19-9) level, administration of NAC, surgical procedure, duration of surgery, intraoperative blood loss, postoperative complications, tumor size, histological type, pathological N factor, pathological tumor-node-metastasis (pTNM) stage, residual tumor, administration of AC, and NLR were collected to evaluate the risk of overall mortality. The pTNM stage was determined according to the 7th edition of the JGR staging system.
The NLR was calculated by dividing the blood neutrophil count (/μl) by the lymphocyte count (/μl) obtained from blood test data within 2 weeks before surgery. For continuous variables, such as NLR, the optimal cutoff value of each factor was defined using receiver operating characteristic (ROC) curve analysis. Continuous variables were classified into two groups as follows: age <69 or ≥69 years, preoperative CA19-9 level <100 or ≥100, duration of operation <432 or ≥432 min, intraoperative blood loss <346 or ≥346 g, tumor size <20 or ≥20 mm, NLR <3.2 or ≥3.2. Categorical variables were classified into two groups as follows: sex as male and female, NAC as no or yes, postoperative complications according to Clavien–Dindo (C-D) classification as minor (C-D, I-II) and major (C-D, III-V), histological type as well/moderate and poor, pathological N factor as absent (N0) or present (N1), residual tumor as R0 or R1.
Ethical approval and informed consent to participate. All study protocols were approved by the Ethics Committee of Kanagawa Cancer Center (approval number: 2023 epidemiologic study-78) and all procedures were conducted according to the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all patients included in the study.
Statistical analyses. Continuous variables are described as medians and interquartile ranges (IQR) and analyzed using the Mann–Whitney U-test. Categorical variables are described as totals and percentages and compared using Fisher’s exact test, as appropriate. Overall survival (OS) and relapse-free survival (RFS) rates after pancreatic cancer resection were calculated according to the Kaplan–Meier method, meanwhile, the log-rank test was used to evaluate the statistical significance of differences between NLR <3.2 and ≥3.2 groups. Furthermore, variables considered significant (p<0.05) in the univariate analysis were candidates for multivariate analysis with Cox proportional-hazard regression models, and the results were displayed as hazard ratios (HRs) and 95% confidence intervals (CIs). Moreover, p<0.05 was considered statistically significant. All data were analyzed using the EZR version 4.20 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) (15).
Results
Study population. A total of 527 patients who were pathologically diagnosed with PDAC underwent pancreatic cancer resection and lymphadenectomy at our institution between January 2013 and December 2022. Additionally, 45 patients were diagnosed with “unresectable” cancer according to the 7th edition of the JGR, 20 patients were diagnosed with pStage IV, and one patient underwent R2 resection. Therefore, only 461 patients were included in this study.
Patient characteristics. The patient characteristics are displayed in Table I as median, IQR, total, and percentages. The median age was 71 years (IQR=39-88), and 243 patients (52.7%) were male. The proportion of surgical procedures for pancreatic resection was PD in 301 (65.3%), DP in 138 (29.9%), and TP, including total remnant pancreatectomy, in 22 (4.8%) cases. Postoperative complications were observed in 278 cases (60.3%), consisting of bile leakage in four (0.9%), paralytic ileus in five (1.1%), pulmonary complications in eight (1.7%), cholangitis in 19 (4.1%), delayed gastric emptying in 37 (8.0%), and postoperative pancreatic fistula with grade B or C in 114 (24.7%) cases. Major complications (C-D, III-V) occurred in 109 patients (23.6%). The median tumor size was 30 mm (range=0-120 mm). A total of 304 (65.9 %) patients had pathologically metastatic lymph nodes. Moreover, NAC and AC were performed in 183 (47.3%) and 412 (89.4%) patients, respectively. Most NAC regimens consisted of gemcitabine/TS-1 therapy (130 patients, 71.0%), while the majority of AC regimens were TS-1 monotherapy (384 patients, 93.2%). The optimal cutoff value of NLR was 3.2 using ROC curve analysis. Furthermore, 108 (23.4%), patients had a value of 3.2 or higher of preoperative NLR (NLR ≥3.2).
Patient characteristics.
Clinicopathological characteristics in relation to NLR. Clinicopathological characteristics in relation to NLR with a cutoff of 3.2 are displayed in Table II. The median tumor size was bigger in patients with NLR ≥3.2 (p=0.005) compared to patients with NLR <3.2. Compared to patients with NLR ≥3.2, those with NLR <3.2 had better histological differentiation (p=0.002) and received more AC (p=0.048).
Clinicopathological characteristics in relation to neutrophil-lymphocyte ratio with a cutoff of 3.2.
Univariate and multivariate analyses of overall survival. The univariate and multivariate analyses of clinicopathological variables in relation to OS after pancreatic cancer resection are demonstrated in Table III. In univariate analysis of OS, significant (p<0.05) variables were as follows: preoperative CA19-9 level ≥100 (p<0.001), duration of operation ≥432 min (p=0.003), intraoperative blood loss ≥346 g (p=0.001), tumor size ≥20 mm (p=0.002), well or moderate differentiation (p=0.025), N1 (p<0.001), R1 (p<0.001), NLR≥3.2 (p=0.005). In the multivariate analysis, considering significant variables from the univariate analysis, independent prognostic factors of OS included preoperative CA19-9 level ≥100 (p=0.021), well or moderate differentiation (p=0.043), N1 (p<0.001), R1 (p<0.001), and NLR ≥3.2 (p=0.022). Median OS in patients with NLR <3.2 and NLR ≥3.2 was 25.8 months (IQR=1.6-121.8), and 21.6 months (IQR=2.4-124.6), respectively. The Kaplan–Meier curves of OS after pancreatic cancer resection are displayed in Figure 1. In patients with NLR ≥3.2, OS was low (HR=1.413, 95%CI=1.052-1.897, p=0.022).
Univariate and multivariate analyses of clinicopathological variables in relation to overall survival after pancreatic cancer resection.
Kaplan–Meier curve for overall survival between neutrophil-to-lymphocyte ratio <3.2 and ≥3.2 groups.
Univariate and multivariate analyses of relapse-free survival. The univariate and multivariate analyses of clinicopathological variables in relation to RFS after pancreatic cancer resection are demonstrated in Table IV. In univariate analysis of OS, significant (p<0.05) variables were as follows: preoperative CA19-9 level ≥100 (p<0.001), duration of operation ≥432 min (p=0.004), intraoperative blood loss ≥346 g (p<0.001), tumor size ≥20 mm (p<0.001), well or moderate differentiation (p=0.006), N1 (p<0.001), R1 (p=0.001), NLR≥3.2 (p<0.001). In multivariate analysis, considering significant variables from the univariate analysis, preoperative CA19-9 level ≥100 (p=0.019), tumor size ≥20 mm (p=0.012), well or moderate differentiation (p=0.042), N1 (p<0.001), R1 (p=0.007) and NLR ≥3.2 (p=0.002) were independent prognostic factors of RFS. Median RFS in patients with NLR <3.2 and NLR ≥3.2 was 14.3 months (IQR=1.3-121.8), and 10.3 months (IQR=1.3-124.6), respectively. The Kaplan–Meier curves of RFS after pancreatic cancer resection are displayed in Figure 2. In patients with NLR ≥3.2, RFS was low (HR=1.55, 95%CI=1.179-2.038, p=0.002).
Univariate and multivariate analyses of clinicopathological variables in relation to relapse-free survival after pancreatic cancer resection.
Kaplan–Meier curve for relapse-free survival between neutrophil-to-lymphocyte ratio <3.2 and ≥3.2 groups.
Discussion
This study aimed to investigate whether NLR, known as a significant inflammation-based prognostic score, serves as a prognostic factor in our high-volume center, and to determine the optimal cutoff value of NLR. The major finding was that a high preoperative NLR was an independent poor prognostic factor for OS and RFS, along with the preoperative CA19-9 level, tumor size, histological type, pathological N factor, and residual tumor. Furthermore, the NLR cutoff value in this study was 3.2. Therefore, our results suggest that preoperative NLR is a useful prognostic factor. In addition, by defining the timing of NLR evaluation within 2 weeks before surgery, as in this study, a unified treatment strategy can be established in the period leading up to surgery.
In the present study, we demonstrated that NLR was a promising prognostic factor, and HR for OS was 1.413 (95%CI=1.052-1.897, p=0.022). Similar results have been reported in previous studies. Fang et al. evaluated the prognostic impact of preoperative NLR in 389 patients with pancreatic cancer diagnosed as pStage I-III (16). They identified that a high NLR was significantly associated with short OS and RFS. The median OS was 13.7 months in patients with high NLR and 22 months in those with low NLR (p=0.001). In addition, the median RFS was 8.2 months in patients with high NLR, and 14.9 months in those with low NLR (p=0.039). Furthermore, multivariate analysis demonstrated that high NLR was associated with a significantly worse prognosis than low NLR for both OS (p=0.001) and RFS (p=0.022).
Possible explanations are present as to why a high NLR status affects long-term oncological outcomes. The first explanation is that NLR status is related to aggressive tumor status. Fang et al. reported that NLR was related to the TNM stage and distant metastasis and that a high TNM stage tended to be associated with high NLR (16). This is because neutrophils produce or release various chemokines, including vascular endothelial growth factor (VEGF), proteases, reactive oxygen species, and nuclear factor kappa B (17), which induce angiogenesis, suppress antitumor activity, and promote tumor invasion and growth (17-19). Additionally, lymphocytes are associated with the direct death of cancer cells (20). Therefore, a high NLR is associated with cancer growth, invasion, metastasis, and progression, leading to a poor prognosis. In this case, clinicopathological features, such as tumor size and histological type correlated with a high NLR; that is, the increase in neutrophil count may have led to tumor growth and malignancy.
Second, the NLR status is associated with the clinical course of AC. Merlo et al. reported that the AC completion rate was related to OS (21). However, no reports demonstrating the association between preoperative NLR with AC administration or completion rate in patients with pancreatic cancer are available. In this case, in patients with pancreatic cancer, high NLR was significantly associated with a low administration rate of AC (HR=0.53, 95%CI=0.35-0.8, p=0.002). This is because a decrease in lymphocyte count is believed to suppress antitumor immunity and cause resistance to chemotherapy (22-24). The proportion of patients who could not receive AC owing to recurrence was 9/31 (29%) in the low-NLR group and 9/17 (53%) in the high-NLR group. In addition, the proportion of patients who discontinued treatment due to adverse events was 5/321 (1.6%) in the low-NLR group and 4/91 (4.4%) in the high-NLR group. This suggests that a high NLR may cause cancer invasion, metastasis, and low continuation rates of AC, resulting in poor prognosis for surgical patients with pancreatic cancer.
The strength of our study was the large sample size. Approximately, 44 reports on the clinical impact of NLR on pancreatic cancer were available from 2010 to 2023 (8), but most reports were on chemotherapy and chemoradiotherapy for stage IV cancer, and only a few reports on surgery. Among reports on surgical cases, our study has the largest sample size (461 patients).
However, the present study has certain limitations. First, the study was a single-center retrospective analysis, which could have caused selection bias when collecting data. Second, NAC and AC were mixed in the data. Other limitations include the various chemotherapy regimens, various types of surgical procedures, and the effects of preoperative pancreatitis or cholangitis.
In conclusion, a high NLR was associated with poor prognosis in surgical patients with pancreatic cancer. To optimize the treatment for pancreatic cancer, together with clinical and imaging diagnoses, simple and practical biomarkers are needed. In the present study, we identified that preoperative NRL may be a good biomarker. Additionally, establishing the timing of NLR evaluation, as shown in this study, could enable the implementation of a cohesive treatment approach before surgery. Furthermore, in patients with a poor NLR, modification of the NAC regimen, more detailed preoperative staging, and stratification in clinical trials may lead to an improved prognosis for patients with pancreatic cancer. Further validation is required in prospective studies.
Acknowledgements
The Authors express their gratitude to the staff of our department for their assistance in the collection and registration of data and samples. This study was supported in part by the nonprofit Yokoyama Surgical Research Group. The Authors would also like to thank Editage (www.editage.jp) for English language editing.
Footnotes
Authors’ Contributions
Shinnosuke Kawahara, TA, and MM made substantial contributions to the conception and design. Shinnosuke Kawahara, MM, RK, DT, YK, MK, and Naoto Yamamoto made substantial contributions to data collection and entry. Shinnosuke Kawahara, TA, MM, IH, YM, Satoshi Kobayashi, MU, and SM made substantial contributions to the data analysis and interpretation. Shinnosuke Kawahara, TA, MM, IH, AK, YM, KK, SS, HT, TO, Norio Yukawa, YR, SM, and AS made substantial contributions to drafting the article and figures. All Authors contributed to writing the manuscript and approved this submission.
Conflicts of Interest
All Authors declare that they have no conflicts of interest in relation to this study.
- Received October 26, 2023.
- Revision received November 21, 2023.
- Accepted November 22, 2023.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- Clinical Relevance of the CALLY Index in Prognostic Stratification of Intrahepatic Cholangiocarcinoma
- Novel UF-5000-based Detection, Integrating Inflammatory Parameters, for Urothelial Carcinoma of the Urinary Bladder
- Prognostic Significance of Inflammation-based Scores in Pancreatic Cancer Patients Treated With Palliative Chemotherapy: A Single Institution Experience
- Combined Neutrophil-to-Lymphocyte Ratio Score Is Associated With Chemotherapeutic Response and Predicts Prognosis in Patients With Advanced Pancreatic Cancer