Serum microRNA-210 as a predictor of neoadjuvant chemotherapy resistance in breast cancer

Subjects

Forty pathologically proven BC patients, who were presented to the oncology department, Alexandria Main University Hospital during the year 2020, were prospectively enrolled in this study. Moreover, 10 age-matched healthy women were recruited from the outpatient clinic as a control group. Patients with BC were allocated to three different groups. The first group included 15 patients with chemosensitive BC to NAC. The second group enrolled 15 patients with chemoresistant BC to NAC. The third group comprised 10 newly diagnosed BC patients before start of treatment. The first two groups were allowed to receive four cycles of Docetaxel and Cyclophosphamide (TC) regimen as NAC. Patients with CR and partial response (PR) were considered as chemosensitive and those with stable disease (SD) and progression disease (PD) were defined as chemoresistant based on response evaluation criteria in solid tumors (RECIST), version 1.1.^[20] All controls underwent mammography and/or ultrasonography which had to be free. The demographic and clinicopathological data were obtained. The study was conducted in accordance with the local institutional ethical guidelines and a signed informed consent was obtained from each woman included in the study.

Sample Collection

From every patient and control subject, 5 mL of venous blood was withdrawn. Each blood sample was drawn on a plain test tube. Blood was allowed to clot, and then centrifuged at 1200 g for 10 minutes to separate serum sample, which was kept frozen at −20°C until use.

Extraction of total RNA

Total RNA isolation from serum samples was implemented with the Qiagen® miRNeasy Mini Kit (Qiagen, CA) according to the manufacturer’s instructions. QIAzol Lysis Reagent is added to serum samples. After addition of chloroform, the lysate splits into aqueous and organic phases by centrifugation (RNA partitions to the upper aqueous phase, DNA partitions to the interphase, and proteins to the lower organic phase).

The upper, aqueous phase is extracted, and ethanol is added to facilitate binding for all RNA molecules from approximately 18 nucleotides upward. The sample is then added to the RNeasy MinElute spin column, where the total RNA binds to the membrane and phenol and other contaminants are efficiently washed away. High-quality RNA is extracted as RNA elution in a small volume of RNase-free water.

To normalize the serum RNA extraction efficiency, a synthesized Caenorhabditis elegans miRNA mimic (cel-miR-39) was added to samples during the extraction step to control for variations during the preparation of total RNA and subsequent steps and for normalization of real-time RT-PCR results of endogenous miRNAs in the sample.

RNA quantitation and quality assessment

The concentration and purity of RNA were measured at 230, 260, and 280 nm by the use of NanoDrop 2000/2000c spectrophotometer (Thermo Scientific, USA). Ratio of A260/A280 = 1.8–2.1 and A260/A230 = 1.8–2.1 indicates highly pure RNA.

Real-time quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Quantification of miRNA-210 expression using the TaqMan MiRNA Assays was carried out using RT-PCR. cDNA was reverse transcribed from purified RNA samples by the use of specific miRNA stem loop primers from the TaqMan MiRNA Assays and reagents from the TaqMan® MiRNA Reverse Transcription Kit. Then, PCR products were amplified from cDNA samples using the TaqMan miRNA Assay together with the TaqMan® Universal PCR Master Mix. RT-PCR was carried out using Applied Biosystems StepOne™ Real-time PCR System. In the real-time PCR system software, a 20 µL sample using standard run mode under thermal cycling conditions was created.

Relative quantification of miRNA-210 expression

For each sample, the difference in CT values for miRNA-210 (target) and miRNA-39 (spike in control) is calculated (the ΔCT). Next, subtraction of the average Δ CT of control group from the calculated ΔCT of each sample yields the ΔΔCT. The negative value of this subtraction (the –ΔΔCT) is used as the exponent of two in the equation (2^–ΔΔCT) and represents the difference in “corrected” number of cycles to threshold.^[21] If relative quantification of miRNA-210 >1.0, this was considered as high expression in BC patients relative to the control. While relative quantification of miRNA-210 <1.0 was considered as low expression relative to the control.

Statistical analysis

The analysis was done using SPSS software version 22 (http://www.ibm.com). Age and miRNA-210 were considered as continuous variables. Patients were categorized into different groups according to tumor characteristics as follows; estrogen receptors (negative versus positive), progesterone receptors (negative versus positive), Her2 (negative versus positive), lymphovascular invasion (negative versus positive), tumor grade (II versus III), and clinical stage (II versus III). Chi-square analyses were used for comparisons of categorical variables. For continuous variables, the nonparametric Kruskal–Wallis test was performed for comparison between different groups. Mann–Whitney test was used for pairwise comparison, with Tukey correction as post-hoc analysis. A univariate logistic regression analysis was also performed to estimate odds ratios (ORs) and 95% confidence intervals (CIs). For prediction of cutoff values of miRNA-210, the receiver operating characteristic (ROC) curve was generated and the area under the curve (AUC) was calculated. A p-value < 0.05 was considered statistically significant.

Characteristics of patients

There was no statistically significant difference between different BC groups regarding the different clinical parameters, namely, age, estrogen receptors, progesterone receptors, Her2, lymphovascular invasion, grade and clinical stage, as shown in Table 1.

Serum miRNA-210 expression level in the studied groups

As demonstrated in Figure 1a, the median value of miRNA-210 level was 1.53, 0.27, 0.52, and 1.35 in chemosensitive, chemoresistant, newly diagnosed and control groups, respectively.

Close

Figure 1:

(a) Box plot shows the difference between different groups regarding the miRNA-210 expression levels; (b) ROC curve shows the sensitivity and specificity of miRNA-210 in prediction of chemoresistance.

Export to PPT

Using Kruskal–Wallis Test, a statistically significant difference (Chi square = 9.801, p = 0.020, df = 3) was found in the mean serum miRNA-210 among the four different groups (chemosensitive, chemoresistant, newly diagnosed, and control). There was a statistically significant decrease in the expression level of miRNA-210 in the chemoresistant group compared to the control (p = 0.009) and chemosensitive groups (p = 0.015) using Mann–Whitney U test for pairwise comparison, with Tukey correction [Table 2].

Sensitivity and specificity of serum miRNA-210 as a predictive biomarker

Using logistic regression model, low serum miRNA-210 expression level was found to be a predictive factor of chemoresistance with p = 0.038 (OR of 0.103, 95% CI = 0.012 –0.886). By plotting ROC curve [Figure 1b], it was found that the AUC was 0.747 (95% CI: 0.550 – 0.943) and the optimum cutoff value was 0.525 for serum miRNA-210 to predict chemoresistance with a sensitivity of 100% and a specificity of 66.7%.