Linking tumor immune infiltration to enhanced longevity in recurrence-free breast cancer

Abstract

Background: The infiltration of tumor-infiltrating B cells and plasma cells in early-stage breast cancer has been associated with a reduced risk of distant metastasis. However, the influence of B-cell tumor infiltration on overall patient survival remains unclear.

Materials and methods: This study explored the relationship between an antitumor immune response, measured by a 14-gene B-cell/immunoglobulin (IGG) signature, and mortality risk in 9638 breast cancer patients across three datasets. Associations with tumor subtype, stage, and age were examined. IGG was characterized using spatial GeoMx profiling and single-cell RNA sequencing, and its relationship with tertiary lymphoid structures (TLSs) was evaluated. The predictive value of each of the 14 IGG genes for B-cell receptor (BCR) and T-cell receptor (TCR) clonality and longevity was also assessed, along with its association with longevity in other cancer types.

Results: High IGG signature expression was significantly associated with a 41%-47% reduction in death risk in breast cancer survivors (P < 0.001), regardless of age, tumor stage, or subtype. Similar associations were observed in other cancers, including melanoma. In breast cancer, the IGG signature was significantly linked to overall survival without relapse in patients aged 41-70 years at diagnosis. Additionally, IGG expression correlated with the presence of TLSs and higher B- and T-cell polyclonality. A specific subset of seven IGG genes strongly correlated with BCR and TCR clonality, with predictive power for identifying clonality and improved longevity, especially when combining two of these genes.

Conclusions: This study uncovers a significant link between immune gene expression in tumors and extended longevity in breast cancer survivors, even in the absence of recurrence. The IGG signature, particularly its key gene subset, emerges as a powerful marker of sustained antitumor immunity and overall patient fitness. These findings pave the way for personalized treatment strategies that enhance both survival and long-term health outcomes.

Keywords: breast cancer; clonality; immune gene expression; longevity.

Figure 1

Assessment of the 14-gene IGG signature for predicting longevity in breast cancer survivors without a documented recurrence. (A) Schematic representation of the study approach. (B) Cumulative risk plots with censored recurrences illustrating the association of IGG expression with OS in all patients from METABRIC (n = 1940), SCAN-B (n = 6652), and TCGA datasets (n = 1082). (C) Correlation analysis depicting the relationship between the IGG signature and age in all patients without documented breast cancer recurrence during follow-up in METABRIC (n = 1133), SCAN-B (n = 4470), and TCGA datasets (n = 936). (D) Distribution of the IGG signature (low, med, high) across age groups in all patients without documented breast cancer recurrence. Chi-square P value <0.001. (E) Forest plot presenting the association of the IGG signature with OS in all patients without documented breast cancer recurrence across different age groups, and in the pooled analysis of the three datasets (in bold). (F) IGG signature expression across breast cancer clinical subtypes in all patients without documented breast cancer recurrence. (G) Distribution of the IGG signature (low, med, high) and age groups, stratified by breast cancer subtype in all patients without documented breast cancer recurrence. CI, confidence interval; HER2, human epidermal growth factor receptor 2; HR, hazard ratio; HR+, hormone receptor-positive; IGG, 14-gene B-cell/immunoglobulin; med, medium; OS, overall survival.

Figure 2

Characterization of the 14-gene IGG signature in breast cancer. (A) Exemplification of protein-based spatial profiling in an IGG-high tumor versus an IGG-low tumor using the GeoMx digital spatial profiling platform. Green denotes tumor cells, red represents immune cells, and blue signifies cell nuclei (DNA). Selected ROIs are highlighted in yellow. The heatmap represents the expression on immune proteins in CD45+ ROIs in each sample. (B) Hematoxylin–eosin staining image depicting a TLS breast tumor (left); IGG signature expression in tumors without TLSs (n = 69) versus tumors with TLSs (n = 66) (middle); and distribution of IGG signature high, medium, and low expression groups in TLS-negative and TLS-positive tumors (right). (C) UMAP plots showing cell types identified using single-cell RNA sequencing analysis in one IGG-high (BC177) and one IGG-low (BC192) breast cancer sample (left); bar plot distribution of all cell types in each sample (right). (D) IHC staining of CD20 (B-cell marker), CD38 (plasma cell marker), and CD3 (T-cell marker) in one IGG-high (BC177) and one IGG-low (BC192) breast cancer sample. AUC, area under the curve; IGG, 14-gene B-cell/immunoglobulin; IHC, immunohistochemistry; ROC, receiver operating characteristic; ROIs, regions of interest; TLS, tertiary lymphoid structure; UMAP, uniform manifold approximation and projection.

Figure 3

Association between IGG and BCR/TCR clonality and identification of CIGGs related to longevity in patients with breast cancer without documented recurrence. (A) BCR Shannon entropy across IGG-high (n = 10), IGG-medium (n = 7), and IGG-low (n = 6) tumors (left) and TCR Shannon entropy across IGG-high (n = 18), IGG-medium (n = 7), and IGG-low (n = 8) tumors assessed by immunoSEQ assay. (B) Pearson correlation between BCR and TCR clonality measures and IGG levels in the HER2-positive breast cancer CALGB-40601 cohort. (C) Venn diagram of the significant individual genes associated with longevity in univariate Cox regression models in METABRIC (n = 1133), SCAN-B (n = 4470), and TCGA datasets (n = 936). (D) Expression of seven CIGGs and IGG signature in the METABRIC dataset. (E) Amount of variation explained in the prognosis of the expression of one single CIGG or the expression of the combination of two CIGGs (i.e. addition of the log base 2 expression values) as defined by chi-square statistics obtained from likelihood ratio testing. (F) ROC AUCs of IGG signature and the combination of the two CIGGs IL2RG + IGJ to predict BCR Shannon entropy across (left) and TCR Shannon entropy (right). (G) Schematic representation of the characteristics of IGG-low and IGG-high tumors. Spearman correlation: ∗∗∗P < 0.001, ∗∗P < 0.01, ∗P < 0.05. AUC, area under the curve; BCR, B-cell receptor; CIGG, core IGG gene; DC, dendritic cell; IGG, 14-gene B-cell/immunoglobulin; NK, natural killer; ROC, receiver operating characteristic; TCR, T-cell receptor.