Dynamic monitoring of circulating tumor DNA can predict chemotherapy response and prognosis in metastatic triple-negative breast cancer patients

Abstract Background: Chemotherapy is the main treatment for metastatic triple-negative breast cancer (mTNBC). However, mTNBC patients (pts) is often associated with chemotherapy resistance and poor prognosis. Biomarkers that can effectively predict the efficacy of chemotherapy for mTNBC are currently lacking. Detection of circulating tumor DNA (ctDNA) in plasma by liquid biopsy is a minimally invasive and highly sensitive method, which has been wildly used in the clinic and plays an important role in tumor diagnosis, efficacy evaluation, residual and recurrent tumor detection, etc. But yet, the clinical application of ctDNA in mTNBC remains relatively scarce. Here, for the first time, we set to explore the possibility of ctDNA as a biomarker for predicting chemotherapy response and prognosis in mTNBC.Methods: From May 2018 to October 2020, 38 mTNBC pts who received less than 3rd line standard chemotherapy were prospectively included. Tumor tissues were obtained prior chemotherapy, and plasma for ctDNA were collected at baseline, a day before the 3rd cycle chemotherapy and at disease progression. Next-generation sequencing (NGS,457 genes panel) was performed on all samples for mutation detection. And the ctDNA fraction, maximum variant allele frequency (max-VAF), tumor mutation burden (TMB) and other variate were calculated, further combined with clinical data, chemotherapy response and survival of pts for statistical analysis.Results: Finally, 109 blood samples and 13 tissue samples were detected by NGS. A total of 214 mutation genes and 397 mutation sites were detected. The mutation types included missense mutation, nonsense mutation, non-/frameshift insertion, non-/frameshift deletion, non-/frameshift substitution, splicing mutation and so on. The genes with the top 5 mutation frequencies were TP53 (32/38, 84.21%), PIK3CA (14/38, 36.84%), KMT2C (8/38, 21.05%), PTEN (6/38, 15.79%), NOTCH4 (6/38, 15.79%), respectively. The same variants were identified in 12 of 13 pts in paired plasma and tissue, with a concordance rate of 92.3%. The mutation rate in plasma ctDNA was significantly lower than that in tissues (15.26% ± 12.52% vs. 28.88 ± 16.54%, P < 0.001) but was still positively correlated with that in tissues (r = 0.306, P = 0.049). The median progression-free survival (mPFS) of pts with GNAS mutation was shorter than GNAS wild-type pts (3.000 vs. 6.100 months, P = 0.014). The area under curve of ctDNA fraction (0.812, P = 0.043), max-VAF (0.817, P = 0.043) and TMB (plasma) (0.759, P = 0.053) in predicting chemotherapy resistance or disease progression were larger than CEA (0.491, P = 0.888), CA125 (0.574, P = 0.243) and CA153 (0.482, P = 0.778). CtDNA fraction (r = 0.482, P < 0.001), max-VAF (r = 0.489, P < 0.001), TMB (plasma). (r = 0.419, P < 0.001) were correlated with chemotherapy response measured by RECIST v1.1 in CT imaging, while CEA (r = -0.024, P = 0.808), CA125 (r = 0.111, P = 0.266) and CA153 (r = -0.017, P = 0.865) had no correlation with chemotherapy response. The mPFS of pts with ctDNA fraction ≤ 50% or max-VAF ≤ 0.4 at baseline was significantly longer than that of pts with ctDNA fraction > 50% (6.100 vs. 3.430 months, P = 0.006) or max-VAF > 0.4 (6.100 vs. 3.430 months, P = 0.047). The elimination of mutations or the decrease of mutation rate in plasma ctDNA after 2 cycles of chemotherapy showed better chemotherapy response, while recurrence of mutations, increase of mutation rate and emergence of new mutations showed chemotherapy resistance.Conclusions: Mutations in tumor tissues and plasma ctDNA of mTNBC pts detected by NGS have high consistency. And compared with CT imaging and traditional tumor markers, dynamic monitoring ctDNA can more aptly reflect the change of whole-body tumor burden, better predict the chemotherapy response and prognosis in mTNBC. Citation Format: Huihui Li, Yajing Chi, Sha Yin, Bo Yu, Mu Su, Baoxuan Zhang, Ling Qiang, Guohua Ren, Lihua Song, Bing Bu, Shu Fang, Mao Shang, Qiaorui Tan, Xiaochu Man. Dynamic monitoring of circulating tumor DNA can predict chemotherapy response and prognosis in metastatic triple-negative breast cancer patients [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-08-27.