The study design was a prospective cohort. Serum samples from a total of 200 female patients with histologically confirmed invasive breast cancer at Songklanagarind Hospital and adequate pathological data were collected from 2017 to 2019 after informed consent was obtained. The exclusion criteria included those with abnormal renal function, those who had previously received any form of treatment for breast cancer and those with other cancers. Pathologic stage and cancer subtype were identified according to the AJCC 8th Edition (25) and St. Gallen International Expert Consensus 2013(26), respectively. Metastatic work-up included abdominal ultrasound and chest computerized tomography. The metastatic status of all patients was confirmed by radiology and histopathological evidence. Treatment of breast cancer in our institute followed the Adult Cancer Treatment Guideline of Thai National Health Security Office (2018) with individual adjustment according to the functional status. Patients were appointed every 3 months during the period of active treatment, then every 6 months, thereafter. The Human Research Ethics Committee of the Faculty of Medicine, Prince of Songkla University, approved the study protocol (Reference no. 60-040-10-1).

Serum β2-M level was measured once at the time of breast cancer diagnosis. Blood was collected under aseptic precautions. Serum was separated and immediately analyzed by immunological agglutination with latex reaction enhancement assay using Immunoturbidimetric Assay kit (Roche/Hitachi). The ratio of reaction is 2:180:80 for sample:buffer:latex suspension. The concentration of β2-microglobulin was evaluated by measuring the agglutination reaction at 570 nm compared to the absorbance of standard β2-microglobulin.

The expression of ER, PR, HER-2 and Ki67 in tumor tissues was evaluated by IHC staining with the following primary antibodies: Anti-ER (Thermo Scientific, Inc., clone SP1, 1/250 dilution), anti-PR (Leica Biosystems, 1/2500 dilution), anti-Ki-67 antigen (Dako Glostrup, 1/250 dilution), and anti-HER2 (Ventana). The staining results were reported by a certified pathologist as the number of positive cells per 100 cancer cells. In cases with equivocal HER2 results, the specimens were further studied by HER-2 dual in situ hybridization (HER2-DISH).

HER2 and chromosome 17 probes were used for two-colour chromogenic in situ hybridization of formalin-fixed paraffin-embedded human breast cancer specimens following the VENTANA BenchMark XT automated slide staining protocol. The slides were evaluated under light microscopy for HER2 and chromosome 17 signals in at least 20 nuclei. Calculation of the HER2/chromosome 17 ratio was performed by dividing the total number of HER2 signals in the target area by the total number of chromosome 17 signals in the same area.

The cell migration assay was conducted in a 24-well plate with Transwell chambers (8-µm pore PET membrane; Falcon, Fisher Scientific). The low migratory cell line, MCF-7, obtained from American Type Culture Collection (ATCC; Catalog HTB-2), and routinely cultured on monolayers at ≤80% confluence in RPMI-1640 medium (Gibco) containing 10% FBS at 37˚C, 5% CO₂ was used in this study. After starvation for 24 h, 2x10⁵ cells in 100 µl 1% FBS RPMI-1640 medium with or without recombinant β2-microglobulin (Abcam) were added to the upper compartment of the chamber, whereas the lower chamber contained RPMI-1640 with 10% FBS. After 72 h of incubation at 37˚C, the cells on the upper surface were removed using a cotton swab. The membranes were fixed with 70% ethanol for 10 min at room temperature and stained with 0.1% crystal violet for 10 min. The number of migrated cells was quantified by counting cells in five different fields of view under a light microscope at a magnification of 200x. The data are presented as the mean ± SD.

The association between β2-M levels and clinicopathological factors was analyzed by using unpaired Student's t-test or one-way ANOVA with Tukey's post hoc test. As age at diagnosis was a major confounder, patients were subcategorized into 2 groups on the basis of age, those aged 25-55 years and those aged >55 years, for analysis of the association between metastatic status and β2-M level. For the Transwell migration study, one-way ANOVA followed by Turkey's post hoc test was used to determine statistical significance between groups. A receiver operating characteristic (ROC) curve was plotted using the sensitivity and specificity of each β2-M cut-off that predicted metastatic status. Survival analysis used log rank test and Kaplan-Meier survival plot with cancer related death used as a censor in overall survival (OS) analysis. Beginning date used operative date and survival data were as of December 2020. All data were analyzed with Statistical Package Stata 14.0 (Stata Corporation). A P-value of less than 0.05 was considered statistically significant.

A total of 200 patients were included in this study. The mean age at the time of diagnosis was 54 years (range 25-88 years). The most common histological type was invasive ductal carcinoma (188 cases; 94%), where the most frequent tumor grade was grade III (83 cases; 41.5%). The percentage of patients with positive lymphovascular invasion was 36.5% (73 cases). Lymph node involvement was most frequently seen in the N0 group (121 cases; 60.5%). T2 was the highest group among the T stages (108 cases; 54%). The percentage of patients with distant metastasis was 4% (8 cases). Regarding tumor molecular subtypes, luminal B was the most common tumor subtype, followed by luminal A (Table I). The example of HER-2 staining and HER-2-DISH was showed as Fig. S1.

The serum β2-M levels in female breast cancer patients according to each pathological parameter are shown in Table I. Statistically significant differences in β2-M levels were found to be associated with age, histologic type, and metastatic status. As our data showed that serum β2-M levels were correlated with age, the level was re-analyzed with age stratification. The association between β2-M levels and metastatic status held true only in those aged >55 years (Table II).

To evaluate the correlation between serum β2-M levels and the prediction of metastatic status in breast cancers, a receiver operating curve (ROC) was constructed, as shown in Fig. 1. The area under the curve (AUC) was 0.78. If a serum β2-M level of 1.9 µg/ml was used as a cut-off, the sensitivity to diagnose the metastatic status was 87.5%, the specificity was 65.0%, and the diagnostic odds ratio was 2.47. Details of the diagnostic value of β2-M using different cut-off values are shown in Table SI. When the ROCs of β2-M performance for diagnosing metastasis were replotted by age subgroup, the AUC of the group of age 25-to 55-year-old group and the >55-year-old group were 0.75 (logistic regression P-value 0.22) and 0.82 (P-value 0.03), respectively. Median follow-up duration was 27.8 months and 24-month OS in all cases was 95.7%. Two-year overall survival (2-year OS) analysis was performed by cancer types, stages, and β2-M levels. The results showed that no significance of difference was found in 2-years OS when patients were sub-grouped by cancer types (log rank P-value=0.07) (Fig. 2A). However, when sub-grouping by stage of cancer, the significance of 2-year OS was found with the percentage of 96.4, 99.0, 88.2 and 75.0% for stages I-III and IV, respectively (log rank P-value <0.01) (Fig. 2B). Two-year OS in cases with serum β2-M levels lower than 1.9 µg/ml (98.3%) was also significantly higher than that of high serum β2-M (89.3%, log rank P-value <0.01) (Fig. 2C).

To study the influence of β2-M on the metastatic capability of breast cancer, the low migratory MCF-7 cell line was subjected to the Transwell cell migration assay. Cells were treated with human recombinant β2-M at different concentrations and incubated for 72 h. The results revealed that ectopic treatment with β2-M stimulated MCF-7 breast cancer cell migration, especially at 30 µg/ml, and that the number of migrated cells was decreased at lower concentrations (Fig. 3). These findings indicated that β2-M had a direct functional effect on MCF-7 cell movement.