A study of circulating interleukin 10 in prognostication of unresectable hepatocellular carcinoma

Authors

  • Stephen L. Chan MRCP,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Frankie K. F. Mo PhD,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Cesar S. C. Wong PhD,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Charles M. L. Chan PhD,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Linda K. S. Leung MRCP,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Edwin P. Hui MRCP,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Brigette B. Ma MRCP,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Anthony T. C. Chan MD,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Tony S. K. Mok MD,

    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Winnie Yeo MD

    Corresponding author
    1. State Key Laboratory in Oncology in South China, Sir Y. K. Pao Center for Cancer, Department of Clinical Oncology, Hong Kong Cancer Institute and Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
    • Department of Clinical Oncology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong
    Search for more papers by this author
    • Fax: (011) 852-2648-7097


Abstract

BACKGROUND:

The level of circulating interleukin 10 (IL-10) is elevated in a proportion of patients with hepatocellular carcinoma (HCC). The objective of the current study was to evaluate the prognostic significance of serum the IL-10 level in patients with unresectable HCC.

METHODS:

Patients with unresectable HCC who provided serum at the time of diagnosis were enrolled prospectively in the study. The level of circulating IL-10 in serum samples was determined by enzyme-linked immunosorbent assay. The association of the IL-10 level with overall survival was evaluated in relation to sociodemographics, liver function, hepatitis B viral load, and tumor staging.

RESULTS:

In total, 222 patients were recruited; of these, 82.4% were positive for hepatitis B virus surface antigen, and 65.8% had Barcelona Clinic Liver Cancer stage C disease. The mean log IL-10 level was 1.1 pg/mL, and 146 patients had an IL-10 level >1 pg/mL (high IL-10 group). The high IL-10 group had worse overall survival than the low IL-10 group (5.0 months vs 14.9 months; hazard ratio, 2.192; P < .0001). The IL-10 level was associated with worse hepatic function and with a high alanine transaminase (ALT) level. The IL-10 level remained an independent prognostic factor (hazard ratio, 1.824; P = .0005) after adjustment for sociodemographics, tumor staging, treatment, Child-Pugh stage, and ALT level. The IL-10 level also subdivided patients into 2 populations with distinct survival (10.2 months vs 3.5 months; P = .0027).

CONCLUSIONS:

The serum IL-10 level was identified as an independent prognostic factor for unresectable HCC. The current findings suggested that an elevated IL-10 level may be related to hepatic injury caused by cirrhotic processes rather than tumor load. The authors concluded that the IL-10 level offers additional prognostic value to the existing tumor staging systems. Cancer 2012. © 2011 American Cancer Society.

INTRODUCTION

Hepatocellular carcinoma (HCC) is the sixth most common cancer and the third leading cause of cancer-related death globally.1, 2 Surgical treatment remains the standard treatment for patients with early HCC.3 However, >60% of patients with HCC present with disease at an advanced stage that is beyond surgical treatment.4, 5 It is known that both clinical behavior and tumor biology of inoperable HCC are different from those of early HCC because of comorbid cirrhosis and tumor load.6-9 Several tumor staging systems have been established for predicting the prognosis of patients with different severities of HCC.10-13

Interleukin 10 (IL-10) is a pleiotropic cytokine that plays a central role in modulation of the immune system in humans.14 It has been suggested that IL-10 is involved in viral-related inflammation in the liver,15-17 and circulating IL-10 is detected in a proportion of patients with HCC.18-20 In patients with resectable HCC, a surgical series reported that an IL-10 level >12 pg/mL was associated with a worse postoperative outcome.18 However, for patients with unresectable HCC, there has been a paucity of data on the prognostic value of circulating IL-10. Furthermore, the clinical significance of circulating IL-10 on the established tumor staging systems for HCC is unclear.

From January 2007 to June 2009, we prospectively enrolled treatment-naive patients with inoperable HCC, and all patients consented to the archiving of serum samples at the time of their HCC diagnosis. In the current study, by quantifying the serum IL-10 level at baseline, our objectives were to determine the prognostic significance of serum the IL-10 level on overall survival in patients with unresectable HCC and to evaluate the clinical significance of serum IL-10 level when incorporated into the currently established tumor staging systems.

MATERIALS AND METHODS

Study Population

The study population was composed of consecutive patients with unresectable HCC who were diagnosed at the Joint Hepatoma Clinic in the Prince of Wales Hospital, Hong Kong, from January 2007 to June 2009. The diagnosis was confirmed by histology or by the presence of characteristic radiologic findings in patients with alpha-fetoprotein (AFP) levels >500 ng/mL and evidence of cirrhosis. Patients who had tumors with mixed histology or those who had received prior therapy for HCC at the time of their first consultation at the clinic were excluded from the study. Patients who were not amenable to resection or liver transplantation would be offered locoablative therapy (radiofrequency ablation, percutaneous ethanol injection), palliative therapy (transarterial therapy, systemic agents), or supportive care. All eligible patients were consented before study entry and were followed prospectively. The study was approved by the Institutional Review Board of the Joint Chinese University of Hong Kong–New Territories East Cluster Clinical Research Ethics Committee and was conducted in accordance with the Declaration of Helsinki of 1975.

Quantification of Serum Interleukin 10 Levels

Serum samples were collected from all study patients within 4 weeks of their HCC diagnosis and before they received treatment for HCC. The samples were stored at −80°C and thawed immediately before determination of the IL-10 level. The IL-10 level was measured by using an enzyme-linked immunosorbent assay (no. 900-036; Enzo Life Sciences, Plymouth Meeting, Pa) according to the manufacturer's instructions. For each reaction, 100 μL of serum were added to each well of a microtiter plate that was precoated with a monoclonal antibody against IL-10 in duplicate. The assay sensitivity was 3.75 pg/mL. The laboratory staff was blinded to all clinical data from the serum samples.

Study Variables and Tumor Staging

In univariate analysis, the following factors were evaluated: patient factors (age, sex, Eastern Cooperative Oncology Group [ECOG] performance status), tumor factors (maximum tumor size, tumor number, vascular thrombosis/invasion, tumor staging), treatment modality (locoablative therapy, palliative treatment and supportive care), liver function parameters (bilirubin, albumin, international normalized ratio [INR], alkaline phosphatase [ALP], Child-Pugh stage, and Model for End-Stage Liver Disease [MELD] score), viral activity (serum hepatitis B virus [HBV] DNA level and alanine transaminase [ALT]), and other laboratory parameters (hemoglobin, white blood cells, and AFP). The MELD score was calculated based on the equation: 9.57 × (creatinine mg/dL) + 3.78 × ln(bilirubin mg/dL) + 11.2 × ln(INR) + 6.43. The maximal serum creatinine level within the MELD score equation was 4.0 mg/dL. For tumor staging, each patient was staged at the time of consent with the following staging systems: American Joint Committee on Cancer tumor-lymph node-metastasis (TNM) staging,10 Barcelona Clinic Liver Cancer Classification (BCLC) system,11 Cancer of the Liver Italian Program (CLIP) score,12 and Chinese University Prognostic Index (CUPI).13 To examine the robustness of the prognostic significance of serum IL-10 and to minimize the problem of hypercollinearity, 2 multivariate analyses with different sets of data were conducted. For the first multivariate analysis, the following factors were included: sociodemographics (age, sex, ECOG performance status), log IL-10, tumor staging (BCLC, CLIP, TNM, Okuda, CUPI), treatment modality (locoablative therapy vs palliative therapy), Child-Pugh stage, ALT level, and MELD score. The second multivariate analysis included sociodemographics (age, sex, ECOG performance status), log IL-10, tumor factors (size, number, and vascular involvement), treatment modality (locoablative therapy vs palliative therapy), Child-Pugh stage, ALT level, and MELD score.

Statistical Analysis

Statistical analysis was performed by using the SAS statistical software package (version 8.2; SAS Institute Inc., Cary, NC). Continuous variables were expressed as means with standard deviations. Baseline continuous variables were compared using the Student t test, and categorical variables were compared using chi-square tests. The database was frozen on June 30, 2010. Overall survival was calculated from the date of HCC diagnosis to the date of death from any cause or was censored at the date of last follow-up if the patient remained alive at the time of analysis. Serum levels of IL-10 were presented in a dichotomized manner (ie, high IL-10 group vs low IL-10 group). In this study, distribution of the serum IL-10 level followed a log-normal distribution. For the convenience of clinical application, we decided to adopt the nearest integral mean logarithmic value of IL-10 as a cutoff level for defining the high and low IL-10 groups. Survival curves were constructed by using the Kaplan-Meier method, and the differences between log IL-10 values were compared using log-rank tests. Clinical factors that were related to overall survival were examined with a Cox proportional hazards model. The hazard ratio (HR) and corresponding 95% confidence intervals (CIs) are provided. Multivariate analysis was performed using a stepwise model-building procedure based on a significance value of P < .05 for both inclusion and exclusion of the prognostic factors and the log IL-10 level. Two sets of multivariate analyses were performed. Factors that were included in the first model were sociodemographic factors (age, sex, ECOG performance status), tumor staging system, treatment modality, Child-Pugh stage, MELD score, and ALT level. Factors that were included in the second model were similar to those included in the first model, except the summarized score of the tumor staging system was replaced by individual tumor factors (tumor number, tumor size, and vascular invasion).

RESULTS

Baseline Characteristics

In total, 222 patients were enrolled. The median follow-up was 31.0 months (95% CI, 26.9-34.1 months). The baseline characteristics of the patients are summarized in Table 1. Most patients (89.1%) were men. The mean age was 59.9 years. One hundred sixty-four patients (73.9%) had an ECOG performance status of worse than 1, and 144 patients (64.9%) had radiologic evidence of cirrhosis. One hundred forty-four patients (64.9%) had Child-Pugh liver function stage A, whereas 65 patients (29.3%) and 13 patients (5.8%) had Child-Pugh liver function stage B and C disease, respectively. The mean MELD score was 9.14. Regarding treatment modality, 32 patients (14.4%) received locoablative therapy (including radiofrequency ablation or percutaneous ethanol treatment) as first-line therapy, 99 patients (44.6%) received palliative therapy (including transarterial therapy or systemic agents), and 91 patients (41.0%) were not suitable for aggressive therapy and received best supportive therapy. Tumor staging details are summarized in Table 2. Fifty-six patients (25.2%) had BCLC intermediate (stage B) disease, and 146 patients (65.8%) had BCLC had advanced (stage C) disease. The overall TNM tumor stage for most patients (n = 136; 61.3%) was stage IV.

Table 1. Patient Characteristics at Diagnosis of Hepatocellular Carcinoma
 No. of Patients (%) 
VariableLog IL-10 >1.00 pg/mL, n = 146Log IL-10 ≤1.00 pg/mL, n = 76Total, N = 222P
  1. Abbreviations: AFP, alpha-fetoprotein; ALP, alkaline phosphatase; ALT, alanine transaminase; ECOG, Eastern Cooperative Oncology Group; HBsAg, hepatitis B virus surface antigen; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; IL-10, interleukin 10; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; SD, standard deviation; TACE, transarterial chemoembolization.

Sex: Men:Women132:1466:10198:24.4165
Age: Mean ± SD, y59.7 ± 12.160.4 ± 12.859.9 ± 12.3.6345
ECOG performance status    
 026 (17.8)32 (42.1)58 (26.1).0009
 1110 (75.3)42 (55.3)152 (68.5) 
 28 (5.5)1 (1.3)9 (4.1) 
 32 (1.4)1 (1.3)3 (1.3) 
HBsAg positive117 (80.1)66 (86.8)183 (82.4).2129
Anti-HCV positive9 (6.2)2 (2.6)11 (4.9).2498
Log HBV DNA: Mean ± SD, log copies/mL6.00 ± 1.34 [n = 117]5.01 ± 1.59 [n = 66]5.64 ± 1.51 [n = 183].0161
Hemoglobin: Mean ± SD, g/dL12.7 ± 2.113.2 ± 1.712.9 ± 2.0.0640
White cell count: Mean ± SD, (109/L7.7 ± 3.26.2 ± 2.47.2 ± 3.1.0003
Platelets: Mean ± SD, (109/L218 ± 126201 ± 126212 ± 126.3615
Bilirubin: Mean ± SD, μmol/L46 ± 9321 ± 1738 ± 77.0015
ALT: Mean ± SD, IU/L95 ± 8461 ± 3683 ± 73<.0001
ALP: Mean ± SD, IU/L217 ± 152146 ± 73193 ± 135<.0001
Albumin: Mean ± SD, g/L34.5 ± 6.338.5 ± 4.936.0 ± 6.2<.0001
INR [n = 220]1.23 ± 0.681.10 ± 0.121.18 ± 0.56.0269
AFP: Mean ± SD, ng/mL107107 ± 39482958127 ± 41731990339 ± 402400.3907
Ascites52 (35.6)6 (7.9)58 (26.1)<.0001
Cirrhosis: Radiologic evidence98 (67.1)46 (60.5)144 (64.9).3286
Child-Pugh stage    
 A78 (53.4)66 (86.8)144 (64.9)<.0001
 B55 (37.7)10 (13.2)65 (29.3) 
 C13 (8.9)0 (0)13 (5.8) 
Vascular involvement51 (34.9)21 (27.6)72 (32.4).2703
MELD score: Mean ± SD9.77 ± 4.677.93 ± 2.979.14 ± 4.25.0021
First-line treatment    
 Locoablative treatment: RFA, PEI17 (11.6)15 (19.8)32 (14.4).0112
 Palliative treatment: TACE, systemic therapy59 (40.4)40 (52.6)99 (44.6) 
 Supportive care70 (48)21 (27.6)91 (41) 
Table 2. Tumor Staging
 No. of Patients (%) 
VariableLog IL-10 >1.00 pg/mL, n = 146Log IL-10 ≤1.00 pg/mL, n = 76Total, N = 222P
  1. BCLC indicates Barcelona Clinic Liver Cancer staging; CLIP, Cancer of Liver Italian Program; CUPI, Chinese University Prognostic Index; IL-10, interleukin 10; SD, standard deviation; TNM, tumor-lymph node-metastasis.

Tumor size: Mean ± SD, cm8.8 ± 5.17.6 ± 5.08.4 ± 5.0.0826
No. of tumors   .1113
  Single44 (30.1)31 (40.8)75 (33.8) 
  Multiple102 (69.9)45 (59.2)147 (66.2) 
TNM staging    
 Tumor classification   .2671
  T12 (1.4)3 (4)5 (2.2) 
  T232 (21.9)21 (27.6)53 (23.9) 
  T323 (15.8)15 (19.7)38 (17.1) 
  T489 (61)37 (48.7)126 (56.8) 
 Lymph node status   .0507
  N0123 (84.3)71 (93.4)194 (87.4) 
  N123 (15.7)5 (6.6)28 (12.6) 
 Metastasis classification   .3115
  M0119 (81.5)66 (86.8)185 (83.3) 
  M127 (18.5)10 (13.2)37 (16.7) 
 TNM overall stage   .2238
  I+II29 (19.9)22 (29)51 (23) 
  III21 (14.4)14 (18.4)35 (15.8) 
  IVA65 (44.5)30 (39.5)95 (42.8) 
  IVB31 (21.2)10 (13.1)41 (18.4) 
Okuda   .0008
  A52 (35.6)41 (54)93 (41.9) 
  B70 (48)34 (44.7)104 (46.8) 
  C24 (16.4)1 (1.3)25 (11.3) 
BCLC   .0007
  A10 (6.9)6 (7.9)16 (7.2) 
  B25 (17.1)31 (40.8)56 (25.2) 
  C107 (73.3)39 (51.3)146 (65.8) 
  D4 (2.7)0 (0)4 (1.8) 
CLIP   .0018
  010 (6.8)16 (21.1)26 (11.7) 
  117 (11.6)16 (21.1)33 (14.9) 
  234 (23.3)15 (19.7)49 (22.1) 
  343 (29.5)19 (25)62 (27.9) 
  427 (18.5)9 (11.8)36 (16.2) 
  515 (10.3)1 (1.3)16 (7.2) 
CUPI    
  Low52 (35.6)46 (60.5)98 (44.1).0002
  Intermediate67 (45.9)28 (36.9)95 (42.8) 
  High27 (18.5)2 (2.6)29 (13.1) 

Serum Interleukin 10 Level as a Prognostic Factor for Advanced Hepatocellular Carcinoma

The mean log IL-10 level was 1.10 pg/mL (ie, 12.59 pg/mL). In total, 146 patients had a serum log IL-10 level >1.00 pg/mL. The median overall survival for the cohort was 8.3 months (95% CI, 6.2-9.8 months). The median overall survival for patients who had a log IL-10 level >1.00 pg/mL (the high IL-10 group) was 5.0 months (95% CI, 3.7-7.8 months). For patients who had a log IL-10 level ≤1.00 pg/mL the (low IL-10 group), the median overall survival was 14.9 months (95% CI, 10.5-22.3 months). The overall survival of the high IL-10 group was significantly worse than that of the low IL-10 group (P < .0001; HR, 2.192; 95% CI, 1.6-3.0) (Figure 1).

Figure 1.

Cumulative survival is plotted using the Kaplan-Meier method and is compared between groups using log-rank tests. Patients with serum log interleukin 10 (IL-10) levels >1.0 pg/mL had worse overall survival than patients with log IL-10 levels ≤1.0 pg/mL (P < .0001). The solid lines indicates log IL-10 ≤ 1.0 pg/mL; dashed line, log IL-10 >1.0 pg/mL.

Univariate and Multivariate Analyses

In univariate analysis (Table 3), the following were identified as prognostic factors for overall survival: ECOG performance status, tumor features (tumor size, tumor number, and vascular invasion), tumor staging (BCLC, CUPI, CLIP, TNM, Okuda, TNM staging), treatment modality (local ablative/palliative therapies vs supportive care), serum HBV DNA level, ALT level, liver function parameters (ALP, bilirubin, albumin, ascites, and overall Child-Pugh stage), MELD score, log AFP, platelets, hemoglobin, and white blood cells. In both multivariate analyses (Table 4), serum IL-10 remained an independent prognostic factor after adjustment for sociodemographic factors (age, sex), tumor features (tumor staging or individual factors), treatment modality (locoablative and palliative therapies), Child-Pugh stage, MELD score, and ALT level.

Table 3. Univariate Analysis
VariablePHR (95% CI)
  1. Abbreviations: AFP, alpha-fetoprotein; ALP, alkaline phosphatase; ALT, alanine transaminase; BCLC, Barcelona Clinic Liver Cancer; CI, confidence interval; CLIP, Cancer of Liver Italian Program; CUPI, Chinese University Prognostic Index; ECOG, Eastern Cooperative Oncology Group; HBV, hepatitis B virus; HR, hazard ratio; IL-10, interleukin 10; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; TACE, transarterial chemoembolization; TNM, tumor-lymph node-metastasis.

Sociodemographics  
 Age.36630.995 (0.983-1.006)
 Sex.54201.164 (0.715-1.894)
 ECOG performance status<.00011.761 (1.406-2.204)
Marker  
 Log IL-10<.00012.192 (1.582-3.037)
Tumor factors  
 No. of tumors.01701.473 (1.072-2.026)
 Tumor size<.00011.079 (1.049-1.110)
 Vascular invasion<.00013.784 (2.742-5.222)
Tumor staging  
 BCLC<.00012.469 (1.889-3.228)
 Okuda<.00012.311 (1.830-2.918)
 TNM overall stage<.00011.623 (1.394-1.889)
 CLIP<0.00011.896 (1.678-2.142)
 CUPI<.00012.521 (2.054-3.095)
Hepatitis  
 Log HBV DNA.00651.160 (1.043-1.292)
 ALT.00581.002 (1.001-1.004)
Liver function  
 Bilirubin<.00011.005 (1.003-1.007)
 ALP<.00011.003 (1.002-1.004)
 Albumin<.00010.936 (0.915-0.957)
 INR.48841.066 (0.889-1.279)
 Ascites<.00012.246 (1.622 – 3.112)
 Child-Pugh stage<.00012.006 (1.587-2.534)
Treatment  
 Grouped treatment<.00010.357 (0.281-0.454)
 Locoablative therapy: RFA, PEI<.00010.125 (0.069-0.226)
 Palliative treatment: TACE,  systemic therapy<.00010.362 (0.265-0.493)
 Supportive care1
Others  
 Log AFP<.00011.360 (1.241-1.490)
 Hemoglobin.00030.859 (0.792-0.933)
 White cell count<.00011.118 (1.068-1.171)
 Platelet.00041.002 (1.001-1.003)
 MELD score.00021.068 (1.032-1.106)
Table 4. Multivariate Analysis
 Method 1, N = 220Method 2, N = 222
VariablePHR (95% CI)PHR (95% CI)
  1. Abbreviations: ALT, alanine transaminase; BCLC, Barcelona Clinic Liver Cancer staging; CI, confidence interval; CLIP, Cancer of Liver Italian Program; CUPI, Chinese University Prognostic Index; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; IL-10, interleukin 10; MELD, Model for End-Stage Liver Disease; NA, not applicable; PEI, percutaneous ethanol injection; RFA, radiofrequency ablation; TACE, transarterial chemoembolization; TNM, tumor-lymph node-metastasis.

Sociodemographics    
 Age    
 Sex    
 ECOG performance status    
Marker    
 Log IL-10.00051.824 (1.303-2.554).00041.881 (1.324-2.673)
Tumor factors    
 No. of tumorsNA .02881.466 (1.040-2.067)
 Tumor sizeNA <.00011.074 (1.040-1.109)
 Vascular invasionNA <.00012.331 (1.671-3.251)
Tumor staging    
 BCLC  NA 
 TNM  NA 
 Okuda  NA 
 TNM.00941.253 (1.057-1.485)NA 
 CLIP<.00011.598 (1.397-1.829)NA 
 CUPI    
Treatment    
 Locoablative therapy: RFA, PEI<.00010.211 (0.114-0.390)<.00010.195 (0.103-0.367)
 Palliative treatment: TACE, systemic therapy<.00010.398 (0.288-0.550)<.00010.400 (0.282-0.567)
 Supportive care    
Others    
 Child-Pugh stage  .01501.422 (1.071-1.889)
 ALT    
 MELD score    

Prognostic Role of Serum Interleukin 10 Level in Tumor Staging for Hepatocellular Carcinoma

The prognostic significance of the serum IL-10 level in each tumor staging system is listed in Table 5. For the TNM system, the patients who had stage I and stage II disease were combined for this analysis because of the small number of patients in each category. Patients who had stage III and stage IV disease also were grouped for evaluation because of the similar outcomes in this group, as demonstrated in a previous study.13 Within both groups of patients (stages I/II and stages III/IV), the IL-10 level further stratified patients into 2 categories with statistically significant, distinct overall survival (P = .0066 and P = .0002, respectively).

Table 5. Subgroup Analysis of Prognostic Role of Interleukin 10 in Each Tumor Staging System
 High IL-10Low IL-10 
Staging SystemNo. of PatientsMedian OS (95% CI), moNo. of PatientsMedian OS (95% CI), moLog-Rank P
  1. Abbreviations: BCLC, Barcelona Clinic Liver Cancer; CI, confidence interval; CLIP, Cancer of Liver Italian Program; CUPI, Chinese University Prognostic Index; NR, not reached; OS, overall survival; TNM tumor-lymph node-metastasis.

TNM     
 I+II2911.45 (8.9-25.0)22— (22.3-NR).0066
 III+IV1173.84 (3.02-5.01)5411.28 (8.13-14.93).0002
BCLC     
 A+B3514.70 (9.79-22.34)3726.44 (16.46-NR).0335
 C1073.45 (2.31-4.65)3910.21 (6.96-13.33).0027
 D43.19 (0.85-7.32)0
CLIP     
 0-26114.03 (9.24-19.90)6624.55 (18.24-43.42).0137
 3433.90 (3.25-4.87)198.94 (4.65-12.85).0174
 4-5421.54 (0.94-1.85)103.12 (2.24-9.62).0231
CUPI     
 Low5214.34 (10.05-20.42)4622.31 (15.41-NR).0620
 Intermediate673.74 (2.02-4.46)288.13 (5.27-11.28).0005
 High271.53 (0.68-2.31)26.60 (1.11-12.10).6703
 Intermediate and high942.76 (1.82-3.80)308.13 (4.00-11.28).0003
Okuda     
 A527.77 (4.39-14.16)4126.44 (20.13-NR).0003
 B704.65 (3.32-8.62)349.15 (6.96-12.10).1105
 C241.02 (0.62-3.28)12.24 (—).8861
 B+C943.80 (2.67-6.21)358.94 (5.43-12.10).0272

For the BCLC staging system, patients with stage A and stage B disease were grouped together for analysis because of the small sample sizes in each of these stages; and, with this combination, the high IL-10 group had worse overall survival than the low IL-10 group (14.7 months vs 26.4 months; P = .0335). For patients with stage C disease, the high IL-10 group had an overall survival of 3.5 months, which was significantly shorter than that for the low IL-10 group at 10.2 months (P = .0027). Because there were only 4 patients with stage D disease, all of whom were in the high IL-10 group, an evaluation of the prognostic significance of IL-10 was not feasible.

For the CLIP score, analyses were conducted in subgroups with scores of 0 to 2, 3, and 4 to 6 based on previous publications that defined clinically meaningful subgroups.21, 22 In all of the subgroups, the IL-10 level was a statistically significant prognostic factor. Among the patients who had a CLIP score of 3, the high IL-10 group had an overall survival of 3.9 months, which was shorter than the overall survival of 8.9 months for the low IL-10 group (P = .0174).

Regarding the CUPI staging system, in the low-risk category, patients in the high IL-10 group tended to have worse overall survival than those in the low IL-10 group, but the association was not statistically significant. The IL-10 level was prognostic in the intermediate risk category (P = .0005). Because of the small number of CUPI high-risk patients in the low IL-10 group, an analysis was conducted on the combined group of high-risk plus intermediate risk categories. The prognostic significance of IL-10 remained statistically significant (2.8 months vs 8.1 months; P = .0003) when the high-risk patients were combined with the intermediate risk group.

For the Okuda system, in patients with stage A disease, the median overall survival of the high IL-10 group was significantly worse than that of the low IL-10 group (7.8 months vs 26.4 months; P = .0003). Patients who had stage C disease were grouped together with those who had stage B disease for this analysis because of the small numbers of patients in the low IL-10 group. The median overall survival of the high IL-10 group was 3.8 months, which was shorter than the 8.9 months for the low IL-10 group (P = .0272).

Serum Interleukin 10 Levels Are Associated With More Active Hepatitis and Poorer Liver Reserves

The differences in various parameters between the high IL-10 group and the low IL-10 group are summarized in Tables 1 and 2. In summary, there were no differences in age or sex between the 2 groups. High IL-10 was associated with worse liver reserves, as evidenced by higher INR, lower albumin, a higher rate of ascites, worse Child-Pugh function, and a higher MELD score. Patients in the high IL-10 group also had more active hepatitis, as evidenced by higher HBV DNA and serum ALT levels. Patients in the high IL-10 group had worse ECOG performance status and higher white cell counts. There were no statistically significant differences in TNM tumor staging, in the proportion of vascular invasion, or in tumor size or numbers. For tumor staging systems that incorporated liver function (BCLC, Okuda, CUPI, CLIP), high-IL-10 was associated with more advanced tumor stage.

DISCUSSION

In the current study, have we demonstrated that the serum IL-10 level is a significant prognostic factor for unresectable HCC. The median overall survival of patients in the low IL-10 group was more than double that of patients in the high IL-10 group. Our results echo a previous study indicating that the serum IL-10 level was an adverse prognostic factor for patients with HCC after surgical resection.18 With respect to inoperable HCC, to our knowledge, there has only been 1 retrospective series based on 74 patients; in that series, it was demonstrated that a serum IL-10 level >10 pg/mL (ie, log IL-10 level >1 pg/mL) was a negative prognostic factor.23 However, in that study, the independent prognostic significance of IL-10 was not determined by multivariate analysis. In the current study, we demonstrated that the prognostic significance of IL-10 was independent of liver function, tumor staging, and treatment for HCC. To our knowledge, this is the first study to systematically evaluate the independent prognostic impact of the serum IL-10 level in patients with unresectable HCC.

We also demonstrated that the prognostic significance of IL-10 is clinically meaningful in the context of currently established tumor staging systems. For both earlier stage HCC (eg, TNM stages I and II, BCLC stages A and B, CLIP scores of 0-2) and advanced stage HCC (eg, TNM stages III and IV, BCLC stage C, CLIP scores of 4-5, intermediate CUPI stage, and Okuda stages B and C), IL-10 could be used to stratify patients into 2 subgroups with distinctly different survival. Although tumor staging attempts to pool patients with similar prognoses into the same category, it is known that patients with more advanced stage HCC (eg, BCLC stage C, intermediate CUPI stage) have highly heterogeneous clinical outcomes with a wide range of overall survival that may overlap with other categories.13, 24 In the current study, we demonstrated that the median survival of the low IL-10 group with TNM stages III and IV, BCLC stage C, and intermediate CUPI stage was longer than 8 months, whereas the survival of the high IL-10 group was only approximately 3 months, which was very close to the survival of patients with terminal HCC. This ability of IL-10 to further differentiate patients with more advanced stage HCC is of particular clinical relevance, because it may aid in decision-making when considering the appropriate provision of care for individual patients. For instance, although aggressive treatment may be offered to suitable patients, this treatment can be withheld in those with terminal disease to avoid unnecessary toxicities. Such characteristics could assist clinicians in refining the prognostication of patients with unresectable HCC.

The results from the current study may provide some indirect clues about the possible association between the IL-10 level and the prognosis of patients with HCC. Some preclinical studies have demonstrated that IL-10 is secreted by HCC cell lines; hence, the IL-10 level theoretically was reflective of tumor load.18, 25 Some other studies have demonstrated that circulating IL-10 came from immune cells in response to viral hepatitis and that the level could be associated with hepatic injury from hepatitis or cirrhosis.15, 26, 27 In the current study, although we did not observe any association between serum IL-10 and tumor characteristics, the IL-10 level was associated with worse Child-Pugh liver function and worse MELD scores. In addition, among patients with chronic HBV infection, IL-10 was associated with higher HBV DNA and ALT levels. It appears that the elevated level of circulating IL-10 may be related to hepatic injury caused by cirrhotic processes rather than by the tumor load per se.

In the current study, despite better hepatic reserves and a higher proportion of locoablative therapy in the low IL-10 group, the prognostic significance of IL-10 was independent of Child-Pugh stage, MELD score, and treatment modality in multivariate analyses. This finding implies that IL-10 is involved in additional pathologic processes that contribute to the progression of HCC. Because it is known that IL-10 exerts an anti-inflammatory response and suppresses the immune system, 1 postulation is that IL-10 may render an immunosuppressive microenvironment that is favorable to tumor progression.20, 28 Another possibility is that IL-10 may interact directly with cancer cells and promote cellular proliferation, and this has been demonstrated in other non-HCC models.29 Therefore, suppression of IL-10 is potentially therapeutic in HCC through the mechanism of restoring immunity in the host as well as minimizing its contribution to cancer progression. However, these postulations remain hypothetical, and further studies will be required to address the detailed pathologic mechanism of IL-10 in HCC.

There were several limitations in the current study. First, this study has not addressed the impact of IL-10 polymorphism. However, it has to be noted that the intent of this study was to evaluate a marker that could be widely used in clinical practice; the technical ease of serum quantification and the readily reproducible results mean that IL-10 measurement is readily applicable in the routine clinical setting. Second, the cohort was composed of an Asian population, and >80% of patients had chronic HBV infection. It is unclear whether the results could be generalized to western HCC populations and to patients with other etiologic factors. Third, serial levels of serum IL-10 were not available; therefore, we could not monitor the interval change during therapy for HCC and its relation to any change in clinical parameters. Further studies are being planned at our center to measure serial levels of individual patients along with monitoring the clinical course of their disease.

In conclusion, the circulating IL-10 level is an independent prognostic factor for patients with unresectable HCC and provides additional prognostic information to the currently existing tumor staging systems. The findings of current study suggest that an elevated IL-10 level may be related to hepatic injury caused by cirrhotic processes rather than because of tumor load.

FUNDING SOURCES

This work was supported by a direct grant from the Chinese University of Hong Kong (Project no. 2041409) and by the Research Fund for the Control of Infectious Diseases Health and Health Services Research Fund (Project 6902728).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary