Dynamic analysis of circulating tumor DNA to predict the prognosis and monitor the treatment response of patients with metastatic triple-negative breast cancer: A prospective study

Background: Limited data are available on applying circulating tumor DNA (ctDNA) in metastatic triple-negative breast cancer (mTNBC) patients. Here, we investigated the value of ctDNA for predicting the prognosis and monitoring the treatment response in mTNBC patients. Methods: We prospectively enrolled 70 Chinese patients with mTNBC who had progressed after <= 2 lines of chemotherapy and collected blood samples to extract ctDNA for 457-gene targeted panel sequencing. Results: Patients with ctDNA+, defined by 12 prognosis-relevant mutated genes, had a shorter progression-free survival ( PFS) than ctDNA- patients (5.16 months vs. 9.05 months, p=0.001), and ctDNA +was independently associated with a shorter PFS (HR, 95% CI: 2.67, 1.2- 5.96; p=0.016) by multivariable analyses. Patients with a higher mutant-allele tumor heterogeneity (MATH) score (>= 6.316) or a higher ctDNA fraction (ctDNA%>= 0.05) had a significantly shorter PFS than patients with a lower MATH score (5.67 months vs.11.27 months, p=0.007) and patients with a lower ctDNA% (5.45 months vs. 12.17 months, p<0.001), respectively. Positive correlations with treatment response were observed for MATH score (R=0.24, p=0.014) and ctDNA% (R=0.3, p=0.002), but not the CEA, CA125, or CA153. Moreover, patients who remained ctDNA +during dynamic monitoring tended to have a shorter PFS than those who did not (3.90 months vs. 6.10 months, p=0.135). Conclusions: ctDNA profiling provides insight into the mutational landscape of mTNBC and may reliably predict the prognosis and treatment response of mTNBC patients.