Radiation-induced circulating microRNAs linked to echocardiography parameters after radiotherapy

Figures

• 

  Figure 1. Results of miRNA expression analysis in the primary group. (A) Impact of radiotherapy for breast cancer on serum microRNA levels at three timepoints (1 – before first fraction of RT, 2 – 24 hours after the last RT fraction, 3 – 1 month after the last RT fraction) with at least one significant difference (after adjustment for multiple hypothesis testing) between either of the three timepoints. Next we identified miRNAs with fold changes > 1.5 or < 0.67 and unadjusted p values < 0.05 (represented by red dots) in between-timepoint comparisons: (B)) Volcano plot of miRNAs differentially expressed between the pre- and 24-hour post-RT timepoints. (C) Volcano plot of miRNAs differentially expressed between the 24-hour post-RT and 1-month post-RT timepoints. (D)) Volcano plot of miRNAs differentially expressed between the pre- and 1-month post-RT timepoints. (E) Among miRNAs with significantly different expression between the 24h post-RT timepoint and the two other timepoints we identified ones with consistent patterns of change and represented them on a Venn diagram.

• 

  Figure 2. (A) Correlation matrix between post-RT miRNA expression levels and dose-volume histogram characteristics of the heart and lungs. Spearman correlation coefficients marked in bold are significant at p < 0.05. (B) Positive correlation between serum levels of miR-30d and mean dose delivered to the heart (r=0.73, p=0.01). (C) Negative correlation between serum levels of miR-15b and mean dose delivered to the left ventricle (r=-0.57; p=0.01). (D) Negative correlation between serum levels of miR-22 and mean lung dose (R=-0.50; p=0.02).

• 

  Figure 3. (A) Correlation between levels of miR-15b and thickness of the left ventricle wall in diastole at follow-up examination (r=0.42; p=0.03). (B) Correlation between levels of miR-424 and thickness of the left ventricle wall in diastole at follow-up examination (r=0.42; p=0.03). (C) Correlation between levels of miR-24 and thickness of the left ventricle wall in diastole at follow-up examination (r=-0.42; p=0.03). (D) Correlation between levels of miR-15b and the E/E’ ratio (an index that can be used to evaluate the LV filling pressure as defined in Supplementary Table 1 – presented without units as a ratio of velocities) at follow-up examination (r=-0.41; p=0.03). (E) Correlation between levels of miR-30d and the E/E’ ratio at follow-up examination (r=0.35; p=0.06). (F) Correlation between levels of miR-451 and the E/A ratio (a marker of the function of the left ventricle of the heart as defined in Supplementary Table 1 – presented without units as a ratio of velocities) at follow-up examination (r=0.32; p=0.09).

• 

  Figure 4. (A) Correlation of post-/pre-irradiation expression level fold changes in the primary and validation cohort. Both sets show similar pattern and magnitude of miRNA expression changes elicited by radiation exposure and subsequent normalization. (B) Serum expression levels of miR-451, miR-424a, miR-22-3p and miR-15b-5p at baseline, 24 hours and 1 month after completion of radiotherapy. In all cases the drop 24h post-RT was significant vs baseline values and reverted to values non-significantly different from the baseline at 1 month, with the exception of miR-15b-5p which remained decreased.