NLCECA score: a serum inflammatory-tumor biomarker score to predict survival of advanced perihilar cholangiocarcinoma after hepatic arterial infusion chemotherapy

Prognostic features in advanced perihilar cholangiocarcinoma (pCCA) patients who received first-line hepatic arterial infusion chemotherapy (HAIC) are unknown. The purpose of our study was to develop an applicable score based on serum inflammatory-tumor biomarkers to predict the survival of advanced pCCA patients who received first-line HAIC. In total, 106 advanced pCCA patients were enrolled as the training cohort. The optimal cutoff values of baseline variables were defined by the receiver operating characteristic method or according to previous publications. According to the results of Cox regression analysis, baseline neutrophil-to-lymphocyte ratio (NLR) > 3.19, carcinoembryonic antigen (CEA) > 10 ng/mL, and carbohydrate antigen 19-9 (CA19-9) > 200 U/mL were identified as independent survival predictors, which were used to develop the NLCECA score (NLR, CEA, and CA19-9). When including the NLCECA score in the multivariate analysis, the NLCECA score was the only independent predictor of survival. The risk of survival decreased by 111.9% for each 1-point increase in the NLCECA score. Additionally, the NLCECA score could also predict survival in another 33 patients in the validation cohort (P < 0.001). In summary, the NLCECA score is a potential biomarker system for predicting the survival of advanced pCCA patients who received first-line HAIC.

www.nature.com/scientificreports/peripheral blood tests has been used as a convenient inflammatory biomarker because of its association with system inflammation and tumor microenvironment.Numerous reports and meta-analyses have found that an elevated NLR before treatment is related to worse OS in several malignancies [9][10][11] .However, whether NLR is closely related to the survival of pCCA patients who received HAIC remains unclear.
Therefore, the purpose of the study was first to evaluate the neutrophil-to-lymphocyte ratio at baseline (bNLR) as a survival biomarker and then develop an applicable score based on serum inflammatory-tumor biomarkers to predict treatment outcomes for pCCA patients receiving first-line HAIC.

Study design
This retrospective study was approved by the institutional review board of Peking University Cancer Hospital (approval protocol number: 2021KT144), and the requirement for informed consent was waived by the institutional review board of Peking University Cancer Hospital.All methods were performed in accordance with the Declaration of Helsinki.The data on pCCA patients from a single center for 10 consecutive years (January 2011 to January 2020) were reviewed.
The inclusion criteria were: (1) unresectable pCCA patients, including those with Blumgart T3 Stage lesions 12 , N2 lymph node metastasis, intrahepatic or distant metastasis, liver cirrhosis, or decreased liver function; (2)  patients with a total bilirubin level less than 100 μmol/L and albumin level greater than 30 g/L prior to HAIC treatment; (3) patients who received first-line HAIC for at least two cycles using the 3cir-OFF regimen; (4) patients who underwent peripheral blood tests within 7 days before the first cycle of HAIC procedure; and (5)  patients who had abdominal contrast-enhanced computed tomography (CT) or magnetic resonance (MR) imaging within 1 month before the first cycle of HAIC and after every two cycles.
The exclusion criteria were: (1) coexistent malignancies; (2) received systemic chemotherapy, radiation therapy, resection, and other previous local treatments; (3) received HAIC other than the 3cir-OFF regimen, concomitant to other local treatment or systemic treatment; (4) absence of tumor response evaluation; and (5) a follow-up period of less than 6 months.
A total of 228 advanced pCCA patients were treated from January 2011 to January 2020, of which 31 patients only received percutaneous transhepatic cholangial drainage (PTCD), 49 patients had previously received systemic chemotherapy, radiation therapy, resection, or other local treatments, and 12 patients received HAIC with different regimens.Among the other 136 patients who received HAIC (3cir-OFF), 129 patients had abdominal imaging before the initiation of HAIC.Of the 129 patients, three had coexisting malignancies, seven did not receive a response evaluation of HAIC, and 13 patients only received one cycle of HAIC.Finally, 106 pCCA patients were enrolled (Fig. 1).According to the same inclusion and exclusion criteria, an additional 33 advanced pCCA patients who received first-line HAIC treatment at our center from January 2020 to January 2022 served as the validation cohort.www.nature.com/scientificreports/

HAIC procedure
HAIC was performed by percutaneous implantation of indwelling port catheter systems as previously described 13 .First, all extrahepatic arteries branching out from the hepatic artery, such as the right gastric artery and accessory left gastric artery, were embolized with 0.018-inch micro-coils.Second, hepatic arterial blood flow was redistributed using 0.018-inch micro-coils to convert multiple arteries into a single arterial blood supply in cases of variation involving multiple hepatic arteries.Finally, the catheter tip was fixed in the gastroduodenal artery, and a side hole was opened in the distal part of the common hepatic artery.Subsequently, the indwelling catheter was connected to the port.The HAIC regimen was 3cir-OFF, consisting of oxaliplatin (40 mg/m 2 for 2 h), 5-fluorouracil (800 mg/m 2 for 22 h), and folinic acid (200 mg/m 2 , 2 h from the beginning of 5-fluorouracil infusion) on days 1-3 every 4 weeks [14][15][16][17][18][19] .A maximum of six consecutive HAIC cycles were performed on patients not showing disease progression.Maintenance therapy with oral capecitabine was followed until tumor progression.

Follow-up and assessments
Overall survival was calculated from the initiation of HAIC to death or last follow-up.PFS was defined as the period from the date of HAIC to disease progression, death, or last follow-up, whichever happened first.All patients underwent contrast-enhanced CT or MR imaging and routine laboratory studies at baseline.During the treatment phase, patients underwent regular laboratory studies, serum carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) level evaluation, and CT or MR imaging after every two HAIC cycles and every 3 months during capecitabine treatment.Survival follow-up was conducted every 2 months by telephone call until September 1, 2021 (January 1, 2024, for the validation cohort), or until the occurrence of death/loss to follow-up.

Statistical analysis
CEA and CA19-9 cutoff values were used similar to previous publications (CEA: 10 ng/mL; CA199: 200 U/mL) 5,6 .The cutoff values of other baseline variables were defined by the receiver operating characteristic (ROC) method.Univariate and multivariate analyses were performed by the Cox proportional hazards regression method.Variables with a P-value < 0.05 on univariable analysis were considered for multivariable analysis.Variables identified as independent predictors for OS were used to construct the survival scoring system.OS/PFS was assessed by Kaplan-Meier analysis.Differences with a P-value < 0.05 were considered statistically significant.All analyses were performed by SPSS v.23.0 software (IBM Corp, Armonk, NY, USA).

Ethics declarations
This retrospective study was approved by the institutional review board of Peking University Cancer Hospital (approval protocol number: 2021KT144).

Consent to participate
This retrospective study was approved by the institutional review board, and the requirement for informed consent was waived.

Patients
All 106 pCCA patients received HAIC (3cir-OFF) as first-line treatment.1).The characteristics of patients in the validation cohort are presented in Table S1.

Discussion
This retrospective study demonstrated NLR as a biomarker of survival for advanced pCCA patients receiving HAIC.Based on NLR, CEA, and CA19-9, which were identified as independent predictors of survival of pCCA after HAIC treatment, a simple and easily applicable clinical serum inflammatory-tumor biomarker system, the NLCECA score, consisting of NLR, CEA, and CA19-9, was established.When it was incorporated, the NLCECA score was the sole independent predictor of survival, and the risk of survival decreased by 111.9% for each 1-point increase in NLCECA score.It is usually difficult to perform a core needle biopsy to obtain pCCA tissues due to its particular perihilar anatomical site and its periductal infiltration growth pattern along the bile duct wall.Most pathological diagnoses of advanced pCCA are based on cytopathology.Hence, for patients with advanced unresectable pCCA, there is a need for biomarkers of survival features.CEA and CA19-9 are widely used traditional prognostic biomarkers for pCCA.However, for the clinical application of CEA or CA19-9 alone, only 30% of CCA patients exhibit elevated CEA levels 20 and CA19-9 is often falsely elevated in benign biliary disease or cholangitis 21 .Thus, a single indicator of CEA or CA19-9 may not well reflect the prognosis of pCCA patients.
Inflammation, which is one of the seven essential characteristics of tumors, can contribute to cancer progression 22 .CCA is also associated with biliary tract inflammation.Most cases of CCA arise during chronic infection of the biliary tree such as cholelithiasis, infection with liver fluke Clonorchis sinensis, or primary sclerosing cholangitis 23 .Various inflammatory biomarkers such as NLR, lymphocyte-to-monocyte ratio, and plateletto-lymphocyte ratio could effectively reflect the degree of inflammation and immune response and are recommended as predictive biomarkers for cancer patients 24,25 .Of the various inflammatory biomarkers, NLR has drawn a large amount of attention because of the function and characteristics of neutrophils and lymphocytes in tumor development.Previous studies have revealed that high neutrophil counts are associated with immunosuppression conditions 26 .Lymphocytes recognize and eliminate tumor cells, as well as inhibit cancer cell proliferation by producing cytokines 27 .Therefore, low lymphocyte counts may be related to impaired adaptive www.nature.com/scientificreports/immunity activation 28 .Thus, high NLR levels serve as a potential biomarker of worse prognosis, which is similar to the results of our study.The prognosis of advanced pCCA is associated with the proliferation and metabolism of the tumor, as well as the local and systemic inflammatory immune status 29 .CEA and CA19-9 are the main protein products of tumor metabolism secreted into the peripheral blood during tumor proliferation and progression 30,31 and have been used as prognosis-related biomarkers in biliary tract cancer 32 .NLR in peripheral blood is associated with tumor immune microenvironment and system inflammation, which has been used as a convenient inflammatory biomarker [9][10][11] .In our study, a new serum inflammatory-tumor biomarker system, the NLCECA score based on NLR, CEA, and CA19-9, was established, and it accurately reflected the tumor growth and proliferation level, as well as the body and tumor inflammatory immune status.When it was incorporated, the NLCECA score was the sole independent predictor of survival, with the risk of survival decreasing by 111.9% for each 1-point increase in the NLCECA score.As a liquid marker, it could aid in the selection of advanced pCCA patients for first-line HAIC in clinical practice.Moreover, due to the easily accessible clinical laboratory test indexes adopted in this score, it is convenient for clinical application.
Our research has several limitations.First, because this was single-center retrospective research, results require further validation in more extensive, multi-center studies.Second, the association of neutrophils to each subgroup lymphocyte ratio with the survival of pCCA patients was not further explored in this study.
In summary, as an immune and inflammatory indicator, low bNLR is an independent predictor of better survival of pCCA after HAIC treatment.A new established serum inflammatory-tumor biomarker system, the NLCECA score, consisting of NLR, CEA, and CA199, may be a reliable survival prediction system for advanced pCCA after HAIC treatment.Prospective validation of the NLCECA score is also warranted.Table 4. Multivariate analyses of factors related to overall survival.*Extent of disease, macroscopic growth patterns, carcinoembryonic antigen (CEA) level, carbohydrate antigen 19-9 (CA19-9) level, neutrophil-tolymphocyte ratio (NLR) level, and NLCECA score were included in the initial multivariate analysis; extent of disease (P = 0.055), macroscopic growth patterns (P = 0.126), CEA level (P = 0.783), CA19-9 level (P = 0.894), and NLR level (P = 0.820) lost their significance and were therefore removed.

Figure 2 .Figure 3 .
Figure 2. (A,B) Cumulative survival curves of hepatic arterial infusion chemotherapy for perihilar cholangiocarcinoma.(C,D) Cumulative survival curves of hepatic arterial infusion chemotherapy for perihilar cholangiocarcinoma low neutrophil-to-lymphocyte ratio at baseline and high neutrophil-to-lymphocyte ratio at baseline groups.bNLR neutrophil-to-lymphocyte ratio at baseline, CI confidence interval, PFS progression-free survival.

3 Figure 4 .
Figure 4. (A,B) Cumulative survival curves of hepatic arterial infusion chemotherapy for 106 perihilar cholangiocarcinoma patients at different time points.(C,D) Cumulative survival curves of hepatic arterial infusion chemotherapy for perihilar cholangiocarcinoma in the validation cohort at different time points.

Table 1 .
Summary of patient baseline characteristics.

Table 2 .
Univariate and multivariate analyses of factors related to overall survival.