SEARCH

SEARCH BY CITATION

Keywords:

  • biomarker;
  • nonsmall cell lung cancer;
  • YKL-40;
  • bone lesions

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

BACKGROUND:

The glycoprotein YKL-40 is synthesized both by cancer cells and by tumor-associated macrophages and plays a functional role in tumor progression. Consequently, high serum YKL-40 levels have been associated with a poor prognosis in patients with several cancer types. However, the role of YKL-40 has not been established in nonsmall cell lung cancer (NSCLC).

METHODS:

Pretreatment serum levels of YKL-40 were determined in 189 patients with NSCLC (143 men and 46 women; median age, 62 years;, age range, 41-76 years). Twelve percent of patients had stage IIIB disease, and 88% had stage IV disease. Ninety-eight patients received combined gemcitabine and vinorelbine, and 91 received combined gemcitabine, vinorelbine, and cisplatin as first-line chemotherapy. The median overall survival was 37 weeks.

RESULTS:

Patients had a median serum YKL-40 level of 209 ng/mL (range, 19-2153 ng/mL). No correlation was observed between overall survival and the type of chemotherapy regimen used, tumor stage, sex, or histologic types. Patients with high serum YKL-40 levels (greater than the median level for all patients [209 ng/mL]) had a significantly shorter survival than patients with serum YKL-40 levels below the median (median survival, 32 weeks vs 41 weeks; P = .007). In multivariate analysis, the serum YKL-40 level, the presence of bone lesions, and the serum lactate dehydrogenase level were independent, statistically significant prognostic factors.

CONCLUSIONS:

The pretreatment serum YKL-40 level was identified as a new, independent prognostic biomarker in patients with metastatic NSCLC and may help to determine the individual prognosis of these patients. Cancer 2010. © 2010 American Cancer Society.

Among all malignant diseases, lung cancer is the most frequent cause of death in Europe and the United State, and smoking is the predominant risk factor.1, 2 Histologically, lung cancer contains a heterogeneous group of tumors: Approximately 75% of these tumors are nonsmall cell lung cancers (NSCLC), and 25% are small cell lung cancer (SCLC). On the basis of histomorphology, the NSCLC group can be subdivided into adenocarcinomas, squamous cell carcinomas, and large cell carcinomas. Recent epidemiologic studies have demonstrated an increasing incidence of adenocarcinoma in women and the younger population, whereas the incidence of SCLC is slightly decreasing.3

At the time of diagnosis, >33% of patients have metastatic disease with malignant effusions or distant organ metastases.2, 4 According to the tumor-lymph node-metastasis (TNM) classification, these patients have stage IIIB or IV disease, for which only palliative treatment is available. Patients with stage IV disease have a poor prognosis, including a median overall survival of 9 months and a 5-year survival rate of only 3%.4 Only a few prognostic factors, such as sex, performance status, and pretreatment weight loss, have been identified.5, 6 The objective of the current study was to identify new prognostic factors with which to assess individual risk profiles and to identify targets for new anticancer treatment strategies.

YKL-40, a phylogenetically highly conserved protein, belongs to the family of 18 glycosyl hydrolases.7 YKL-40 is secreted by activated neutrophils, macrophages in later stages of differentiation, arthritic chondrocytes, differentiated vascular smooth muscle cells, and fibroblast-like synovial cells.8-11 Elevated serum levels of YKL-40 have been reported in several nonmalignant diseases characterized by inflammation and tissue remodeling, for instance, severe bacterial inflammation, rheumatoid arthritis, liver fibrosis, and asthma.12-14

The YKL-40 gene is overexpressed in cancer cells arising from bone, brain, breast, cartilage, cervix, colon, germ cell, head and neck, kidney, liver, lung, lymph node, ovary, pancreas, prostate, skin, and thyroid compared with the respective normal tissues or cells (dbEST database; National Center for Biotechnology Information, US National Library of Medicine, Bethesda, Md). Immunohistochemical studies have demonstrated YKL-40 protein expression in many types of cancer cells, including adenocarcinoma cells from breast, colon, liver, and pancreas; and squamous carcinoma cells from the uterine cervix, the head/neck area, melanoma cells, and glioblastoma cells.15 Furthermore, YKL-40 is produced by tumor-associated macrophages, mast cells, and leukocytes.15 For instance, in patients with SCLC, YKL-40 messenger RNA expression has been observed in tumor-associated macrophages.16

In cancer, the exact biologic functions of YKL-40 are not known. It has been suggested that YKL-40 plays a role in proliferation and differentiation of malignant cells and is regulated by tumor necrosis factor alpha and nuclear factor κB.17 YKL-40 protects against apoptosis18; it is a cellular survival factor, as supported by observations in human glioblastoma,19 and promotes tumor angiogenesis.20, 21 Several studies have demonstrated elevated serum levels of YKL-40 in cancer patients.15 At the time of diagnosis from 13% to 65% of patients with primary localized disease and from 39% to 91% of patients with metastatic disease had elevated serum levels of YKL-40. Furthermore, serum YKL-40 was not highly correlated with other prognostic serum biomarkers (eg, carcinoembryonic antigen, cancer antigen [CA 125], CA 15-3, human epidermal growth factor receptor 2, or prostate-specific antigen), and elevated serum YKL-40 was identified as an independent prognostic biomarker of shortened overall survival in patients with cancer.15, 22 Therefore, serum YKL-40 may reflect other aspects of tumor growth and spread than the classic markers. It was demonstrated recently that high plasma YKL-40 levels were associated with an increased risk of developing gastrointestinal cancer in the general population.23 Furthermore, serum YKL-40 levels are stable both in the short term and in the long term in healthy individuals, and it has been suggested that YKL-40 may be a useful marker for monitoring patients with cancer during and after treatment.22 The objective of the current study was to investigate the prognostic value of pretreatment serum YKL-40 levels for survival in patients with metastatic NSCLC.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Patients

Serum samples from 189 patients with cytologically or histologically confirmed, advanced NSCLC were investigated. The patients were a nonselected subset of 287 patients who were treated in a randomized phase 3 trial that comparing combination chemotherapy with gemcitabine plus vinorelbine (GV) versus combination chemotherapy with gemcitabine, vinorelbine, plus cisplatin (GVP).24 The 189 patients who participated in the current study did not differ in terms of relevant patient characteristics, such as Karnofsky performance status, sex, disease stage, tumor histology, and tumor grade, from the other 98 patients who were not included. This also was true for outcome. The enrollment of patients into the study took place between September 1999 and June 2001. Informed consent was obtained from all patients before study enrollment.

Twenty-three patients (12%) with stage IIIB disease who had malignant pleural effusions and 166 patients (88%) with stage IV disease were enrolled in the current investigation. No previous chemotherapeutic treatment was allowed. The study was approved by the local ethic committee and by the Protocol Committee of the German Cancer Society.

Processing Serum Samples and YKL-40 Analysis

Twenty milliliters of venous blood were taken before chemotherapy from 189 nonfasting patients with NSCLC. The blood subsequently was spun at 3400 revolutions per minute for 7 minutes at 4°C. The supernatant was transferred into microtubes and stored at −70°C until analysis.

The serum concentration of YKL-40 was determined by using a commercial, 2-site, sandwich-type enzyme-linked immunosorbent assay (ELISA) (Quidel, Santa Clara, Calif) using streptavidin-coated microplate wells, a biotinylated-Fab monoclonal capture antibody, and an alkaline phosphatase-labeled polyclonal detection antibody. The sensitivity of the ELISA was 10 ng/mL. The intra-assay coefficient of variation (CV) was ≤5%, and the interassay CV was ≤6%. Each serum sample was analyzed in duplicate. The samples were analyzed by an investigator who was blinded to the clinical parameters and study endpoint.

Statistical Analysis

Patients' overall survival was measured from the time of randomization to the date of either death (from any cause) or last contact for patients who remained alive (censored observation). The data were analyzed on an intent-to-treat basis. Graphic presentations using Kaplan-Meier survival estimates were produced that grouped the patients according to serum YKL-40 levels, which were dichotomized above and below the median serum YKL-40 level for all patients (209 ng/mL). Median survival was obtained from Kaplan-Meier survival curves. The log-rank test was used to compare survival curves.

In this study, we investigated 2 outcomes, overall survival and progression-free survival, without adjusting statistical significance for this bivariate endpoint. Spearman correlation coefficients (r) and their P values of significance versus zero correlation were calculated to analyze the relations between quantitative serum concentrations of YKL-40, tumor stage, Karnofsky performance status, and metastatic sites. A statistically significant correlation (P < .05) was judged as low, moderate, or high if the Spearman correlation coefficient exceeded 0.25, 0.5, or 0.75, respectively. An inverse correlation was determined when the Spearman correlation coefficient was negative.

To determine whether factors were associated with survival, a univariate Cox regression analysis was performed for age, sex, tumor grade, disease stage, therapy, serum YKL-40 concentration, Karnofsky performance status, and several metastatic sites (bone, liver, lung, lymph nodes, adrenals, and malignant pleural effusions). In univariate survival analyses, serum YKL-40 levels, thrombocytes, and serum lactate dehydrogenase (LDH) levels were dichotomized as above versus below the median.

In addition, multivariate Cox regression analysis was performed to adjust for potential confounders. Covariates were selected when they were significant in the univariate analysis at the level of 5%. To identify the best subset of predictive factors in the multivariate Cox model, a backward variable selection procedure was used. Regression coefficients are reported with their standard errors, P values were calculated using the Wald test, and the respective hazard ratios (HRs) are provided with 95% confidence intervals (CIs). Statistical analyses were performed using the statistical software packages SAS for Windows (version 9; SAS Institute, Cary, NC) and R (version 2.1.1; R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Patient characteristics and the relations between pretreatment serum YKL-40 and standard prognostic variables are listed in Table 1. The median patient age was 62 years (range, 41-76 years), and the median Karnofsky performance status was 90%. The majority of patients were men (76%). In the past, 31 patients (16%) had undergone surgery, and 13 patients (7%) had received radiotherapy. After inclusion in the phase 3 trial for combination chemotherapy, 98 patients received GV, and 91 patients received GVP. Both treatment arms were well balanced with regard to age, sex, performance status, tumor histology, tumor grade, and disease stage.

Table 1. Patient Characteristics and Baseline Serum Concentrations of YKL-40 in 189 Patients With Metastatic Nonsmall Cell Lung Cancer: Basic Prognostic Variables
CharacteristicNo. of Patients (%)Median Serum YKL-40 [range], ng/mLPa
  • NSCLC indicates nonsmall cell lung cancer.

  • a

    The Kruskal-Wallis test was used to compare the actual levels in the strata.

  • b

    Because of multiple counts, no medians, ranges, or P values were calculated.

Age, y  .94
 40-4919 (21)242 [19-756] 
 50-5948 (34)193 [37-1055] 
 60-6988 (27)205 [32-1358] 
 70-7934 (17)235 [33-2153] 
Sex  .80
 Men143 (76)209 [32-2153] 
 Women46 (24)208 [19-1055] 
Karnofsky performance status  .001
 100%25 (13)156 [19-601] 
 90%77 (41)193 [54-1358] 
 80%62 (33)253 [33-1212] 
 70%25 (13)341 [37-2153] 
Stage  .88
 IIIB23 (12)217 [47-763] 
 IV166 (88)209 [19-2153] 
Histology  .01
 Adenocarcinoma94 (50)178 [19-1358] 
 Squamous cell carcinoma63 (33)242 [66-1212] 
 Large cell carcinoma14 (7)307 [62-504] 
 Unclassified NSCLC18 (10)228 [86-2153] 
Grading  .11
 Well differentiated: Grade 17 (4)342 [168-831] 
 Moderate differentiated: Grade 262 (33)190 [19-1212] 
 Pure differentiated: Grade 369 (36)205 [33-1055] 
 Undifferentiated: Grade 419 (10)314 [37-1358] 
 Unknown32 (17)217 [32-2153] 
Site of metastasesb   
 Lymph nodes77 (41)  
 Bone61 (32)  
 Lung53 (28)  
 Liver50 (26)  
 Pleura30 (16)  
 Adrenal glands24 (13)  
Chemotherapy regimen  .18
 Gemcitabine and vinorelbine98 (52)189 [32-1212] 
 Gemcitabine, vinorelbine, and cisplatin91 (48)239 [19-2153] 

Pretreatment Serum YKL-40

The 189 patients with metastatic NSCLC had a median serum YKL-40 level of 209 ng/mL (range, 19-2153 ng/mL). The pretreatment serum YKL-40 level was not associated with age, sex, treatment, tumor grade, or disease stage. There was a significant correlation between pretreatment serum YKL-40 level and Karnofsky performance status (P = .001), and patients who had a Karnofsky performance status of 100% had the lowest serum YKL-40 levels. There also was a trend toward an association with histology (P = .01): The lowest serum YKL-40 levels were observed in patients with adenocarcinoma (Table 1).

Low correlations were observed between serum YKL-40 levels and LDH (r = 0.18; P = .02), and inverse correlations were observed between serum YKL-40 levels and Karnofsky performance status (r = −0.28; P < .0001) and hemoglobin (r = −0.22; P = .002). No correlations were observed between serum YKL-40 levels and tumor stage (r = −0.01; P = .9) or the presence of bone metastases (r = 0.01; P = .9).

Pretreatment Serum YKL-40 and Response to Chemotherapy

To evaluate the impact of the pretreatment serum YKL-40 level on response to chemotherapy, data from 146 patients were available. In these patients, the pretreatment serum YKL-40 level had no impact on response to chemotherapy. Patients who achieved a complete or partial response (n = 35) had a median serum YKL-40 level of 189 ng/mL, whereas patients who had stable or progressive disease (n = 111) had a median serum YKL-40 level of 196 ng/mL (P = .62).

Serum YKL-40 and Progression-Free Survival

The median progression-free survival for all 189 patients was 24 weeks (95% CI, 19-32 weeks). No difference in progression-free survival was observed with regard to the chemotherapy regimen used (P = .4), tumor stage (P = .9), sex (P = .2), histologic types (P = .4), or grade of tumor differentiation (P = .6). Patients who had a pretreatment serum YKL-40 level >209 ng/mL (the median level for all patients) had significantly shorter progression-free survival than patients who had a serum YKL-40 level ≤209 ng/mL (median progression-free survival, 18 weeks vs. 32 weeks; P = .009) (Fig. 1). Patients who had a Karnofsky performance status of 70% to 80% had a shorter progression-free survival than patients who had a performance status of 90% to 100% (17 weeks vs 32 weeks; P = .004), as expected.

thumbnail image

Figure 1. These Kaplan-Meier survival curves illustrate the association between progression-free survival and serum levels of YKL-40 in samples that collected before chemotherapy in 189 patients with metastatic nonsmall cell lung cancer. The serum YKL-40 level was dichotomized according to the median level for all patients (209 ng/mL). P values were calculated by using the log-rank test for equality of strata.

Download figure to PowerPoint

Serum YKL-40 and Overall Survival

The median overall survival for all 189 patients was 37 weeks (95% CI, 33-42 weeks). No significant difference in overall survival was observed with regard to chemotherapy regimen (P = .8), tumor stage (P = 1.0), sex (P = .4), histologic type (P = .4), or tumor grade of differentiation (P = 1.0). Patients who had a pretreatment serum YKL-40 level >209 ng/mL had significantly shorter overall survival than patients who had a pretreatment serum YKL-40 level ≤209 ng/mL (median overall survival, 32 weeks vs 41 weeks; P = .007) (Fig. 2). Patients who had a Karnofsky performance status of 70% to 80% had a shorter overall survival than patients who had a performance status of 90% to 100% (32 weeks vs 41 weeks; P = .01).

thumbnail image

Figure 2. These Kaplan-Meier survival curves illustrate the association between overall survival and serum levels YKL-40 in samples that collected before chemotherapy in 189 patients with metastatic nonsmall cell lung cancer. The serum YKL-40 level was dichotomized according to the median level for all patients (209 ng/mL). P values were calculated by using the log-rank test for equality of strata.

Download figure to PowerPoint

In univariate Cox analysis, the pretreatment serum YKL-40 level (dichotomized as >209 ng/mL vs ≤209 ng/mL) was associated significantly with overall survival (HR, 1.58; P = .008). In univariate Cox regression analysis, Karnofsky performance status, low/high serum LDH, and the presence of bone metastases also had prognostic significance for survival in all patients with NSCLC (Table 2).

Table 2. Univariate Cox Regression Analysis of Potential Prognostic Factors for Overall Survival in 189 Patients With Nonsmall Cell Lung Cancer
FactorParameter EstimateStandard ErrorPHR [95% CI]
  • HR indicates hazard ratio; CI, confidence interval; GV, gemcitabine and vinorelbine; LDH, lactate dehydrogenase.

  • a

    Reference group.

  • b

    This variable was dichotomized according to the median serum level in 189 patients.

Age−0.0010.011.911.0 [0.98-1.02]
Sex (mena)−0.1610.202.420.85 [0.57-1.26]
Karnofsky performance status (70%-80%a)−0.440.171.010.65 [0.46-0.91]
Grade (poorly differentiated and undifferentiateda)−0.010.191.960.99 [0.68-1.44]
Tumor stage (IIIBa)−0.010.27.970.99 [.58-1.70]
Therapy (baseline GVa)0.050.17.751.06 [0.76-1.47]
Serum LDH (median, 39 U/L)b0.480.18.0071.62 [1.14-2.29]
Serum YKL-40 (median, 209 ng/mL)b0.460.17.0081.58 [1.13-2.21]
Metastases    
 Lung−0.240.19.210.78 [0.54-1.15]
 Lymph nodes0.070.17.691.07 [0.76-1.51]
 Pleura−0.260.25.290.77 [0.47-1.25]
 Bone0.560.18.0021.76 [1.24-2.49]
 Liver0.361.89.051.44 [1.00-2.08]
 Adrenal glands0.300.24.221.35 [0.84-2.17]
 Others0.440.25.081.56 [0.95-2.57]

In multivariate Cox analysis, only the pretreatment serum YKL-40 level (dichotomized as >209 ng/mL vs ≤209 ng/mL), the serum LDH level (dichotomized as >1100 pg/mL vs ≤1100 pg/mL), and the presence of bone metastases were identified as independent prognostic factors (Table 3). When LDH and YKL-40 levels were analyzed as quantitative parameters (details of the Cox regression not shown) LDH (quantitatively scaled) no longer was a statistically significant factor (P = .05) in univariate Cox regression and was not included in the multivariate analysis, whereas YKL-40 (quantitative scale; P = .0002) and bone metastases (P = .001) were identified as significant, independent prognostic factors.

Table 3. Multivariate Cox Regression Analysis Using Factors With Sufficient Statistical Power (P < .05) in Univariate Analysis and Backward Variable Selection Procedure Results for Those Factors
FactorParameter EstimateStandard ErrorPHR (95% CI)
  • HR indicates hazard ratio; CI, confidence interval; GV, gemcitabine and vinorelbine; LDH, lactate dehydrogenase.

  • a

    This variable was dichotomized according to the median serum level in 189 patients.

LDHa0.380.18.0361.46 (1.03-2.08)
YKL-40a0.400.18.0291.48 (1.04-2.10)
Metastases: Bone0.530.18.0041.69 (1.18-2.42)

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

The objective of the current investigation was to study the association between the pretreatment serum YKL-40 level and the clinicopathologic parameters and clinical outcomes of patients with metastatic NSCLC (stage IIIB with malignant pleural effusions or stage IV). In general, serum YKL-40 levels were higher in patients with metastatic NSCLC compared with the levels reported in most other types of cancer.15, 22, 25-36 The median serum YKL-40 levels were lower in patients with metastatic melanoma, glioblastoma, SCLC, prostate cancer, colorectal cancer, breast cancer, and ovarian cancer compared with the levels observed in patients with NSCLC, although they also were elevated compared with healthy individuals, and some of these patients also had very elevated levels. Indeed, our cutoff value for YKL-40 was data-driven and cannot be generalized. Further investigation will be needed to determine how to incorporate the serum YKL-40 level into future clinical studies.

Mechanistically, the very high serum YKL-40 level in patients with metastatic NSCLC may be related to their very poor prognosis, as illustrated by the short progression-free survival and overall survival in the group of NSCLC patients that had the highest serum YKL-40 levels. It is noteworthy that a high serum YKL-40 level was an independent prognostic biomarker of poor survival in this patient population, in accordance with previous findings from others studies on the prognostic value of serum YKL-40, for example, in patients with breast cancer, colorectal cancer, cervical cancer, head and neck cancer, renal cell cancer, prostate cancer, ovarian cancer, endometrial cancer, SCLC, glioblastoma, melanoma, and acute myeloid leukemia, in which high serum YKL-40 levels have been associated with a poor prognosis.15, 22, 25-36

Serum YKL-40 was a strong prognostic biomarker of poor prognosis and was independent of Karnofsky performance status and bone metastases. Another study that investigated patients with high-grade glioma who underwent surgery also observed significantly better survival for patients who had lower serum YKL-40 levels and suggested the use of YKL-40 as a biomarker for prognosis in that patient population.31

The biologic functions of YKL-40 in cancer cells are unknown, and very few studies have evaluated the functional role of YKL-40 expression in cancer cells. It has been suggested that YKL-40 plays a role in the proliferation and differentiation of malignant cells, protects the cells from undergoing apoptosis,18 stimulates angiogenesis,21 has an effect on extracellular tissue remodeling, and stimulates fibroblast activity/proliferation surrounding the cancer cells, although in vivo proof of these hypotheses has yet to be obtained.15 Membrane receptors that mediate the biologic effects of YKL-40 are not known. However, the activation of cytoplasmic signal-transduction pathways suggests that YKL-40 interacts with 1 or several signaling components on the cell membrane.17 In the future, YKL-40 may become an important target for anticancer treatment strategies in NSCLC. By blocking YKL-40 or its receptor using a monoclonal antibody, angiogenesis and cancer progression may be inhibited. Currently, more knowledge about the biology, mechanisms of action, and function of YKL-40 is mandatory for using YKL-40 as a target for cancer therapy. These investigations are currently under study in our laboratory.

In conclusion, the current study has identified the pretreatment serum concentration of YKL-40 as an independent prognostic biomarker for survival in patients with metastatic NSCLC. Together with classic prognostic markers like the Karnofsky performance status, the pretreatment serum YKL-40 level can help to assess individual risk profiles for this patient population.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

We thank Tonni Løve Hansen and Debbie Nadelmann at Herlev Hospital for excellent technical assistance.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

This study was supported by grants from “Direktør Jens Aage Sørensen og Hustru Edith Ingeborg Sørensens Mindefond” and “Læge Sophus Carl Emil Friis og hustru Olga Doris Friis'Legat.”

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES