Higher IGFBP-1 to IGF-1 serum ratio predicts unfavourable survival in patients with nasopharyngeal carcinoma.

ABSTRACT

BACKGROUND: The insulin-like growth factor (IGF) system plays an important role in the development and progression of cancer. However, little is known about the expression of the IGF system components and their clinicopathological significance and prognostic value in nasopharyngeal carcinoma (NPC). METHODS: IGF system components (IGF-1, IGF-2, IGF-1SR, IGFBP-1, IGFBP-2, IGFBP-3, IGFBP-4 and IGFBP-6) were quantified from the plasma of NPC patients and healthy individuals using the RayBio Human Cytokine Antibody Array. IGFBP-1 and IGF-1 mRNA levels were quantified by real-time qPCR, and protein expression was detected by western blot in nine NPC cell lines and four immortalized nasopharyngeal epithelial (NPE) cell lines. Tissue-specific expression of IGFBP-1 and IGF-1 was detected by immunohistochemistry in paraffin-embedded NPC tissues. ELISA analysis was used to measure the serum levels of IGFBP-1 and IGF-1 in 142 NPC patients and 128 healthy controls and determine potential correlation with clinicopathological parameters. RESULTS: Significantly higher levels of circulating IGFBP-1 and lower levels of IGF-1 and IGF-2 were detected in NPC patients compared to healthy controls by Cytokine Antibody Array analyses (P = 0.034, 0.012, 0.046, respectively). IGFBP-1 expression was detected in the majority of NPC cell lines, but not in NPE cell lines, and was shown to localize to the nucleus of tumour cells, in contrast to the cytoplasmic staining observed in normal cells. Importantly, IGFBP-1 expression was stronger in NPC tumour tissues compared to peritumoural tissues. In contrast, IGF-1 expression was weak or absent in NPC and NPE cell lines, with the exception of the EBV-infected C666 cell line, and was found to be expressed at lower levels in tumour tissues compared to tumour-adjacent normal tissue. Levels of serum IGFBP-1 were shown to be significantly higher in patients with NPCs compared to healthy control individuals (55.23 ± 41.25 µg/L vs. 32.08 ± 29.73 µg/L, P < 0.001), whereas serum levels of IGF-1 were significantly lower in NPC patients compared to healthy controls (98.14 ± 71.48 µg/L vs. 164.01 ± 92.08 µg/L, P = 0.001). Consistently, the IGFBP-1/IGF-1 serum ratio was shown to be significantly higher in NPC patients compared to healthy control individuals (P = 0.002). Serum levels of IGFBP-1 and the IGFBP-1/IGF-1 ratio significantly correlated with age (P = 0.020; P = 0.016), WHO histological classification (P = 0.044; P = 0.048), titre of EA (EB Virus Capsid Antigen-IgA) and NPC (P = 0.015; P = 0.016). In contrast, higher IGFBP-1 serum levels and IGFBP-1/IGF-1 ratio significantly correlated with poor RFS (P = 0.046; P = 0.037) and OS (P = 0.038; P = 0.009). Multivariate analysis revealed that the IGFBP-1/IGF-1 ratio, but not serum IGFBP-1 level, represents an independent risk factor for poor RFS (P = 0.044) and OS (P = 0.035). CONCLUSIONS: A higher IGFBP-1/IGF-1 serum ratio is significantly associated with poor prognosis in NPC patients.