The association between immune cells and breast cancer: insights from Mendelian randomization and meta‐analysis

Background: Breast cancer (BC) is the most common cancer among women worldwide, with 2.3 million new cases and 685 000 deaths annually. It has the highest incidence in North America, Europe, and Australia and lower rates in parts of Asia and Africa. Risk factors include age, family history, hormone replacement therapy, obesity, alcohol consumption, and lack of physical activity. BRCA1 and BRCA2 gene mutations significantly increase the risk. The 5-year survival rate is over 90% in developed countries but lower in developing ones. Early screening and diagnosis, using mammography and MRI, are crucial for reducing mortality. In recent years, significant progress has been made in studying BC immunophenotyping, particularly in multicolor flow cytometry, molecular imaging techniques, and tumor microenvironment analysis. These technologies improve diagnosis, classification, and detection of minimal residual disease. Novel immunotherapies targeting the tumor microenvironment, like CAR-T cell therapy, show high efficiency and fewer side effects. High levels of tumor-infiltrating lymphocytes correlate with better prognosis, while immune checkpoint molecules (PD-1, PD-L1) help cancer cells evade the immune system. Tumor-associated macrophages promote invasion and metastasis. Blocking molecules like CTLA-4, LAG-3, and TIM-3 enhance antitumor responses, and cytokines like IL-10 and TGF-β aid tumor growth and immune evasion. Mendelian randomization (MR) studies use genetic variants to reduce confounding bias and avoid reverse causation, providing robust causal inferences about immune cell phenotypes and BC. This approach supports the development of precision medicine and personalized treatment strategies for BC. Methods: This study aims to conduct MR analysis on 731 immune cell phenotypes with BC in the BCAC and Finngen R10 datasets, followed by a meta-analysis of the primary results using the inverse-variance weighted (IVW) method and multiple corrections for the significance P -values from the meta-analysis. Specifically, the study is divided into three parts: First, data on 731 immune cell phenotypes and BC are obtained and preprocessed from the GWAS Catalog and Open GWAS (BCAC) and the Finngen R10 databases. Second, MR analysis is performed on the 731 immune cell phenotypes with BC data from the BCAC and Finngen R10 databases, followed by a meta-analysis of the primary results using the IVW method, with multiple corrections for the significance P -values from the meta-analysis. Finally, the positively identified immune cell phenotypes are used as outcome variables, and BC as the exposure variable for reverse MR validation. Results: The study found that two immune phenotypes exhibited strong significant associations in MR analysis combined with meta-analysis and multiple corrections. For the immune phenotype CD3 on CD28+ CD4-CD8- T cells, the results were as follows: in the BCAC dataset, the IVW result was odds ratio (OR) = 0.942 (95% CI: 0.915–0.970, P =6.76×10 -5 ), β =−0.059; MR Egger result was β =−0.095; and the weighted median result was β =−0.060. In the Finngen R10 dataset, the IVW result was OR=0.956 (95% CI: 0.907–1.01, P =0.092), β =−0.045; MR Egger result was β =−0.070; and weighted median result was β =−0.035. The β values were consistent in direction across all three MR methods in both datasets. The meta-analysis of the IVW results from both datasets showed OR=0.945 (95% CI: 0.922–0.970, P =1.70×10 -5 ). After Bonferroni correction, the significant P- value was P =0.01, confirming the immune phenotype as a protective factor against BC. For the immune phenotype HLA DR on CD33- HLA DR+, the results were as follows: in the BCAC dataset, the IVW result was OR=0.977 (95% CI: 0.964–0.990, P =7.64×10 -4 ), β =−0.023; MR Egger result was β =−0.016; and the weighted median result was β =−0.019. In the Finngen R10 dataset, the IVW result was OR=0.960 (95% CI: 0.938–0.983, P =6.51×10 -4 ), β =−0.041; MR Egger result was β =−0.064; and weighted median result was β =−0.058. The β values were consistent in direction across all three MR methods in both datasets. The meta-analysis of the IVW results from both datasets showed OR=0.973 (95% CI: 0.961–0.984, P =3.80×10 -6 ). After Bonferroni correction, the significant P -value was P =0.003, confirming this immune phenotype as a protective factor against BC. When the immune cell phenotypes CD3 on CD28+ CD4-CD8- T cells and HLA DR on CD33- HLA DR+ were used as outcomes and BC was used as exposure, the data processing and analysis procedures were the same. The MR analysis results are as follows: data from the FinnGen database regarding the effect of positive immune phenotypes on malignant neoplasm of the breast indicated a β coefficient of −0.011, OR = 0.99 (95% CI: −0.117–0.096, P =0.846); data from the BCAC database regarding favorable immune phenotypes for BC demonstrated a β coefficient of −0.052, OR=0.095 (95% CI: −0.144–0.040, P =0.266). The results suggest insufficient evidence in both databases to indicate that BC inversely affects these two immune cell phenotypes. Conclusions: Evidence suggests that the immune cell phenotypes CD3 on CD28+ CD4-CD8- T cells and HLA DR on CD33- HLA DR+ protect against BC. This protective effect may be achieved through various mechanisms, including enhancing immune surveillance to recognize and eliminate tumor cells; secreting cytokines to inhibit tumor cell proliferation and growth directly; triggering apoptotic pathways in tumor cells to reduce their number; modulating the tumor microenvironment to make it unfavorable for tumor growth and spread; activating other immune cells to boost the overall immune response; and inhibiting angiogenesis to reduce the tumor’s nutrient supply. These mechanisms work together to help protect BC patients and slow disease progression. Both immune cell phenotypes are protective factors for BC patients and can be targeted to enhance their function and related pathways for BC treatment.