Genetic determinants and clinical significance of circulating and tumor-specific levels of insulin-like growth factor binding protein 7 (IGFBP7) in a Swedish breast cancer cohort

Abstract

Previous research indicates that insulin-like growth factor binding protein 7 (IGFBP7) protein levels in breast cancer tissue and blood are prognostic. However, genetic determinants of IGFBP7 in breast cancer remain largely unexplored. We examined IGFBP7 in a cohort of 1701 patients with first breast cancer from Sweden, enrolled prior to surgery 2002-16 and followed for up to 15 years. Genotyping was performed on blood samples using OncoArray. Tumor-specific protein levels of IGFBP7, insulin receptor (InsR), and IGF-I receptor (IGFIR) were assessed on tumor tissue microarrays in 964 patients. Furthermore, 275 patients had plasma IGFBP7 levels measured. A genetic proxy marker for circulating IGFBP7 levels was constructed from five candidate single-nucleotide polymorphisms (SNPs) (rs6852762, rs1714014, rs9992658, rs10004910, and rs4865180) based on number of recessive genotypes. Age-adjusted linear regression was used to evaluate SNPs and tumor-specific IGFBP7 levels in relation to circulating IGFBP7 levels. Cox regression adjusted for age, tumor characteristics, and adjuvant treatments was used to assess associations with clinical outcomes. Circulating and tumor-specific IGFBP7 levels were significantly positively correlated. High circulating and genetically predicted IGFBP7 levels were associated with increased risk for distant metastasis and all-cause mortality. A significant interaction between high tumor-specific IGFBP7 levels and membrane-bound InsR resulted in a four-fold increased risk of breast cancer events and distant metastases. Both measured and genetically predicted IGFBP7 levels were independent prognostic biomarkers in breast cancer.

Keywords: IGFBP7; breast cancer; cohort study; genotype; prognostic biomarker.

Graphical Abstract

Figure 1.

Genomic region of IGFBP7 and IGFBP7-AS1 (A) with Chromosome:base pair coordinates (GRCh37 assembly) of the five pQTL for IGFBP7 levels (rs6852762, rs1714014, rs4865180, rs9992658, and rs10004910) used to construct a genetic proxy of circulating IGFBP7 levels. (B–G) Violin plots with overlaying box plots illustrate the distribution of circulating IGFBP7 levels by each individual SNP (rs6852762 (B), rs1714014 (C), rs4865180 (D), rs9992658 (E), and rs10004910 (F)) and the genetic proxy (G) with corresponding forest plots of the age-adjusted linear regression models. In the box plots, the boundary of the box closest to zero indicates the 25th percentile, a black line within the box marks the median, and the boundary of the box farthest from zero indicates the 75th percentile. Points above the whiskers (Q3 + 1.5 * interquartile range (IQR)) and below the (Q1 − 1.5 * IQR) indicate outliers

Figure 2.

Violin plot (A) with overlaying box plot illustrates the distribution of circulating IGFBP7 levels by tumor-specific IGFBP7 levels and (B) corresponding forest plots of the age-adjusted linear regression model. In the box plot, the boundary of the box closest to zero indicates the 25th percentile, a black line within the box marks the median, and the boundary of the box farthest from zero indicates the 75th percentile. Points above the whiskers (Q3 + 1.5 * IQR) and below the (Q1 − 1.5 * IQR) indicate outliers. IGFIR status (C) in breast cancers by tumor-specific IGFBP7 levels. The y-axis indicates IGFIR status in tumors, and tumor-specific IGFBP7 levels are indicated on the x-axis. IGFIR membrane status (D) in breast cancers by tumor-specific IGFBP7 levels. The y-axis indicates IGFIR membrane status in tumors, and tumor-specific IGFBP7 levels are indicated on the x-axis. InsR status (E) in breast cancers by tumor-specific IGFBP7 levels. The y-axis indicates InsR status in tumors, and tumor-specific IGFBP7 levels are indicated on the x-axis. InsR membrane status (F) in breast cancers by tumor-specific IGFBP7 levels. The y-axis indicates InsR membrane status in tumors, and tumor-specific IGFBP7 levels are indicated on the x-axis. The rs10004910 genotype (G) by tumor-specific IGFBP7 levels. The y-axis indicates rs10004910 genotype, and tumor-specific IGFBP7 levels are indicated on the x-axis

Figure 3.

Kaplan–Meier estimates of (A) breast cancer-free interval with corresponding (B) forest plots of adjusted hazard ratios (HR) (95% confidence intervals (CI)), distant metastasis-free interval (C) with corresponding (D) forest plots of adjusted HR (95% CI), and overall survival (E) with corresponding (F) forest plots of adjusted HR (95% CI) in relation to tertiles of circulating IGFBP7 protein levels in patients with available plasma analytes (n = 275). The number of patients is indicated at each time-point. The study is ongoing; thus, the number of patients decreases with time

Figure 4.

Kaplan–Meier estimates of (A) breast cancer-free interval with corresponding (B) forest plots of adjusted HR (95% CI), distant metastasis-free interval (C) with corresponding (D) forest plots of adjusted HR (95% CI), and overall survival (E) with corresponding (F) forest plots of adjusted HR (95% CI) in relation to the genetic proxy of circulating IGFBP7 levels in patients with available genotype information (n = 1701). The genotype variant ‘1+’ is a proxy for higher IGFBP7 levels, the genotype variant ‘0’ is a proxy for intermediate IGFBP7 levels, and the genotype variant ‘−1’ is a proxy for lower IGFBP7 levels. The number of patients is indicated at each time-point. The study is ongoing; thus, the number of patients decreases with time

Figure 5.

Kaplan–Meier estimates of distant metastasis-free interval with corresponding forest plots of adjusted HR (95% CI) in relation to rs10004910 genotype in patients with ER + tumors only treated with aromatase inhibitors (A, B), only treated with tamoxifen (C, D), treated with sequential therapy (E, F), and not treated with endocrine therapy (G, H). The number of patients is indicated at each time-point. The study is ongoing; thus, the number of patients decreases with time