Prognostic role of inflammatory markers in patients with primary biliary cholangitis

， ， ，

Xijing Hospital of Digestive Diseases，Air Force Medical University，Xi’an 710032，China

［Abstract］Objective To investigate whether inflammatory markers can predict the prognosis of patients with primary biliary cholangitis (PBC).Methods A retrospective study was conducted to explore the correlation between inflammatory markers and liver function through Spearman correlation analysis.The optimal cut-off value of inflammatory markers was determined by X-tile software.The effects of each inflammatory marker on prognosis were compared by Log-rank test and their respective Kaplan-Meier curves were drawn,and finally the inflammatory markers affecting prognosis were further analyzed by Cox survival analysis.Results The gamma-glutamyl transpeptidase-to-platelet ratio (GPR) was an independent risk factor for the prognosis of PBC patients,and higher GPR (≥1.9) prompted a poor prognosis.Conclusion GPR level at baseline has a correlation with the histological stage and poor prognosis of PBC patients.GPR (≥1.9) can be utilized as an independent predictor to suggest negative prognosis of PBC patients.

［Key words］inflammatory markers; primary biliary cholangitis; prognosis

1 Introduction

Primary biliary cholangitis (PBC) is a chronic liver disease characterized by the inflammation of the small bile ducts and the development of fibrosis.This inflammation can eventually lead to cirrhosis and even liver failure，especially in middle-aged women[1-2].Ursodeoxycholic acid (UDCA) is currently used to treat subjective symptoms in patients with PBC，including fatigue and pruritus，both of which have been shown in previous studies to significantly reduce health-related quality of life[3].It has been demonstrated through researches that UDCA has the potential to significantly boost biochemical markers and transplant-free survival[4-5].Nevertheless，some UDCA-treated patients do not show adequate reactions，and a research indicates that the patients who fail to show a biochemical reaction to UDCA have a greater risk of mortality[6].Therefore，there is an urgent need to find biomarkers for prognosis and prediction of PBC for better clinical treatment and prognosis.

Recent years，more and more studies have shown systemic inflammatory markers are closely correlated with the prognosis in various cancers and liver diseases[7-8].However，the prognostic value of systemic inflammatory markers in PBC is not clearly clarified.A previous study by LIN et al[9]showed that neutrophil-to-lymphocyte ratio (NLR) was closely related to short-term mortality in patients with PBC.Unfortunately，the evidence is based on small numbers of participants(88 patients with PBC)and short duration of follow-up (12 months)，and another weakness of the study was that it was unclear which systemic inflammatory biomarker was most associated with prognosis.Aiming to further clarify the prognostic value of systemic inflammatory markers in PBC，we conducted this retrospective study.

2 Methods

2.1 Study population

We conducted this study in Department of Gastroenterology，Xijing Hospital，Air Force Medical University.The data of patients were collected from January 2004 to December 2020.We enrolled patients who were diagnosed with PBC and had treatment-naive for UDCA at presentation.The diagnosis of PBC was established by meeting at least two out of the following three criteria：①Elevated levels of alkaline phosphatase (ALP) or gamma-glutamyl transpeptidase (GGT) for a period of 6 months，indicating cholestasis；②Presence of positive anti-mitochondrial antibodies as confirmed by a blood test；③Liver tissue examination showing characteristic features of PBC，such as non-suppurative cholangitis and injury to the interlobular bile ducts[10].Only PBC patients who received continuous UDCA treatment for a minimum of 1 year following diagnosis were included in our analysis.Excluded from the study were patients who had a primary clinical endpoint within 6 months，alcoholic liver disease，primary sclerosing cholangitis，steatohepatitis，overlapping autoimmune hepatitis，and viral hepatitis (hepatitis B or C).A total of 401 patients were eventually included in our investigation，as shown in Figure 1.

Figure 1 Flowcharts of the study

2.2 Data collection

At the beginning of the study，we collected demographic and clinical information，such as age at diagnosis，gender，histological stage，and various laboratory examinations.These examinations included neutrophil count，lymphocyte count，platelet count，monocyte count，white blood cell count，hemoglobin，alanine aminotransferase (ALT)，aspartate aminotransferase (AST)，ALP，GGT，total bilirubin (TBIL)，albumin (ALB)，international normalized ratio，prothrombin time，fibrinogen，and immunoglobulin.Upon PBC diagnosis，all patients enrolled in our study were treated with UDCA at stable doses of 13-15 mg/(kg·d) and regularly followed up for more than 12 months.Whether to response to UDCA treatment was assessed by Paris I criteria.The primary clinical endpoint for our cohort study was defined as hepatic decompensation (ascites，variceal bleeding，and encephalopathy for the first time，respectively)，liver transplantation，or liver-related death.Data were censored at the time of occurring of the above events.

2.3 Definitions

The concrete measurement of the systemic inflammatory markers[11-13]were calculated as follows：SII (×109/L)，platelet count×neutrophil count/lymphocyte count；NLR (×109/L)；MLR (×109/L)，monocyte-to-lymphocyte ratio；GPR (×109/L)，gamma-glutamyl transpeptidase-to-platelet ratio；FAR (×109/L)，fibrinogen-to-albumin ratio；ALR (×109/L)，aspartate aminotransferase-to-lymphocyte ratio；PLR (×109/L)，platelet-to-lymphocyte ratio.

2.4 Statistical analysis

3 Results

3.1 Baseline characteristics of patients

In strict adherence to our inclusion criteria and exclusion criteria for the study，401 patients were eventually enrolled in our retrospective study.The demographic and clinical features of the PBC patients treated with UDCA at baseline were summarized in Table 1，which contained 301 patients (UDCA responders，75.1%) according to Paris I criteria.At baseline，the median age was 50 (45，58)，83.5% of the cohort as a whole was female，and the average length of follow-up was 55 months.Baseline histopathological stage，systemic inflammatory markers as well as more details were showed in Table 1.

Table 1 The demographic and clinical features of the PBC patients at baseline (n=401)

3.2 Association between baseline systemic inflammatory markers and liver function-related indicators

Associations between liver function impairment and systemic inflammatory markers were presented in Figure 2A，which displayed GPR and ALR were most closely correlated with liver function.To further explore the relationship between GPR，ALR and liver function，we applied Spearman correlation analysis，and the results of the correlation analysis were presented in Figure 2B-G.

A：The relationship between inflammatory markers and liver function indicators；B-D：The correlations between GPR and AST，ALP，TBIL，respectively；E-G：The correlations between ALR and AST，TBIL，ALP，respectively.ALT：alanine aminotransferase；AST：aspartate aminotransferase；ALP：alkaline phosphatase；GGT：gamma-glutamyl transpeptidase；TBIL：total bilirubin；ALB：albumin；PLT：platelets；NLR：neutrophil-to-lymphocyte ratio；PLR：platelet-to-lymphocyte ratio；MLR：monocyte-to-lymphocyte ratio；SII：platelet count×neutrophil count/lymphocyte count；GPR：gamma-glutamyl transpeptidase-to-platelet ratio；ALR：aspartate aminotransferase-to-lymphocyte ratio；FAR：fibrinogen-to-albumin ratio.

3.3 Factors associated with the occurrence of adverse events

X-tile software was used to determine the optimal cut-off values for systemic inflammatory markers of the adverse outcome-free survivals，which were NLR 1.8，PLR 138.5，MLR 0.2，SII 95.8，GPR 1.9，ALR 55.1，FAR 0.1.On the basis of these cut-off values，the total number of patients was then divided into two categories (low-valuevshigh-value) in each inflammatory marker，and Kaplan-Meier survival analysis was constructed.Differences between the two subgroups had been assessed with Log-rank tests，which were shown in Figure 3.Kaplan-Meier survival analysis revealed that NLR<1.8，GPR<1.9 and ALR<55.1 were substantially linked to prolongation of time free of adverse events (P=0.02，P=0.001，P<0.001，respectively).To recognize the key elements that were associated with the clinical outcomes of patients，univariable and multivariable Cox regression analysis were carried out as presented in Table 2.Univariable Cox regression analysis showed that the factors associated with the occurrence of adverse events were higher ALT level (HR：1.231，95%CI：1.052-1.440，P=0.009)，higher AST level (HR：1.326，95%CI：1.161-1.514，P<0.001)，higher ALP level (HR：1.291，95%CI：1.132-1.472，P<0.001)，higher GGT level (HR：1.079，95%CI：1.032-1.127，P=0.001)，higher TBIL level (HR：1.533，95%CI：1.331-1.766，P<0.001)，lower ALB level (HR：0.003，95%CI：0.000-0.005，P<0.001)，higher NLR (≥1.8) level (HR：2.007，95%CI：1.117-3.607，P=0.020)，higher ALR (≥55.1) level (HR：3.852，95%CI：2.072-7.163，P<0.001)，higher GPR (≥1.9) level (HR：3.035，95%CI：1.537-5.992，P=0.001) at baseline.To further investigate the independence of NLR，ALR and GPR as a potential marker for outcome，multivariable Cox regressive analysis was carried out，which indicated higher GPR (≥1.9) level (HR：2.125，95%CI：1.055-4.282，P=0.035) was an independent prognostic factor in patients with PBC.In addition，multivariable regression analysis showed that TBIL and ALB were independent factors associated with occurrence of adverse events (HR：1.361，95%CI：1.134-1.633，P=0.001；HR：0.032，95%CI：0.001-0.701，P=0.029，respectively).

A：All cohort；B：NLR；C：PLR；D：MLR；E：SII；F：GPR；G：ALR；H：FAR.NLR：neutrophil-to-lymphocyte ratio；PLR：platelet-to-lymphocyte ratio；MLR：monocyte-to-lymphocyte ratio；SII：platelet count×neutrophil count/lymphocyte count；GPR：gamma-glutamyl transpeptidase-to-platelet ratio；ALR：aspartate aminotransferase-to-lymphocyte ratio；FAR：fibrinogen-to-albumin ratio.

Table 2 Univariate and multivariate Cox regression analyses of factors associated with the occurrence of adverse events

4 Discussion

PBC has a global distribution and can occur in all racial and ethnic groups[14-16].A recent meta-analysis showed that the incidence and prevalence of PBC are increasing globally，with annual incidence ranging from 0.23/100 000 to 5.31/100 000 and prevalence ranging from 1.91/100 000 to 40.2/100 000，with the highest in North American and Northern European countries[17].A recent meta-analysis estimated the prevalence of PBC in China was 20.5/100 000，ranking second in the Asia-Pacific region after Japan[18].UDCA is the first-line treatment for PBC.For patients with an inadequate biochemical reaction to UDCA，the long-term prognosis and survival rate are low，and second-line treatment should be considered[4-5].There are multiple international criteria for evaluating the biochemical response after UDCA therapy[19].However，these criteria are dependent on biochemical markers 6 months or 1 year after UDCA treatment.Previous studies have suggested that inflammatory markers may be associated with the progression and survival of digestive disorders，such as hepatocellular carcinoma and colorectal cancer[20-22].However，the role of inflammatory factors in PBC is not so clear since the number of patients included in the existing studies is small，and the inclusion of inflammatory markers is not comprehensive.Therefore，it is also unclear which inflammatory marker is most closely associated with the survival of patients with PBC.

We conducted this retrospective cohort study to describe the adverse outcome-free survival and determine the prognostic value of systemic inflammatory markers in UDCA-treated PBC patients at a large tertiary care hospital in Midwest China.Eventually，consistent with previous studies，TBIL was associated with survival of the patients in our cohort[23-25].More importantly，our results showed that elevated pre-treatment ALR and GPR were associated with shorter adverse event-free survival in PBC recipients treated with UDCA，independently of other prognostic factors.To our utmost knowledge，this is currently the largest cohort of inflammatory markers validated in the patients with PBC，and at the same time，this is the first time that SII，NLR，MLR，GPR，FAR，ALR，and PLR have been monitored to identify their association with pathological stages.Moreover，the indicators involved in the inflammatory markers are readily available.

Different from the findings of LIN et al[9]，in our study，multivariate Cox regression analysis did not suggest a statistically significant difference between NLR and the occurrence of adverse event-free survival，which may require further validation in large cohorts.

Our study still has a small number of restrictions.As our study was a single-center post hoc trial，sampling bias was unavoidable.Meanwhile，our study was observational，and the data at baseline did not reflect some dynamic results.Therefore，our results will then be validated in larger，prospectively recruited cohorts of patients.

In conclusion，our study indicated that baseline ALR and GPR were correlated with the occurrence of adverse events.Baseline higher GPR (≥1.9) could be a useful inflammatory marker to identify the patients who tend to have a poor response to UDCA and predict the occurrence of cirrhosis-related complications and reduced survival.