Purpose: This study aimed to evaluate whether the three protein markers that we had developed previously for early breast cancer diagnosis, are useful for postoperative follow-up.
Methods: Of 111 patients with breast cancer that were prospectively enrolled and underwent blood tests before, 8-weeks, 6-months, or 1-year after surgery, 53 underwent serial tests preoperatively to 1-year postoperatively. Protein biomarker changes were analyzed before and after surgery. Moreover, 145 patients who underwent surgery >1 year previously, were prospectively enrolled into the recurrence and non-recurrence (no evidence of disease; NED) groups. Using traditional and artificial intelligence analysis methods, we determined whether the biomarkers could identify disease presence or absence, and evaluated whether significant correlation occurred in biomarker changes.
Results: Of the 111 included patients, 105 and 53 were followed-up for 8 weeks and 6 months to 1 year after surgery, respectively. Preoperatively, biomarker diagnostic sensitivity was 73.0%; 8-weeks and 6-months to 1-year after surgery, normalized blood biomarker level occurred in 67.6% and 86.8% of patients, respectively. Among 53 patients with serial tests, preoperative test sensitivity was 73.6%, which normalized after 6 month in 86.8%. Greater than 1-year after surgery, 38/63 patients with recurrence developed cancer, and 66/81 NED patients were cancer-free (62.5% sensitivity; 81.5% specificity). The upgraded version of biomarker analysis confirmed a further improved performance (with recurrence, 82.5% sensitivity; with NED, 87.7% specificity).
Conclusions: Blood protein signatures (Mastocheck) developed for early breast cancer diagnosis normalized over time postoperatively, and could be used as a biological monitoring test after breast cancer surgery.