Abstract

Circulating tumor DNA (ctDNA) during and after neoadjuvant chemotherapy and prior to surgery is a powerful prognostic factor in triple-negative breast cancer (TNBC).

Author
person Luca Cavallone Segal Cancer Centre, Lady Davis Institute, Jewish General Hospital, Montreal, QC, Canada info_outline Luca Cavallone, Adriana Aguilar, Mohammed Aldamry, Josiane Lafleur, Susie Brousse, Cathy Lan, Najmeh Alirezaie, Eric Bareke, Jacek Majewski, Manuela Pelmus, Cristiano Ferrario, Elizabeth A. Marcus, Andre Robidoux, Federico Discepola, Mark Basik
Full text
Authors person Luca Cavallone Segal Cancer Centre, Lady Davis Institute, Jewish General Hospital, Montreal, QC, Canada info_outline Luca Cavallone, Adriana Aguilar, Mohammed Aldamry, Josiane Lafleur, Susie Brousse, Cathy Lan, Najmeh Alirezaie, Eric Bareke, Jacek Majewski, Manuela Pelmus, Cristiano Ferrario, Elizabeth A. Marcus, Andre Robidoux, Federico Discepola, Mark Basik Organizations Segal Cancer Centre, Lady Davis Institute, Jewish General Hospital, Montreal, QC, Canada, Segal Cancer Centre/Jewish General Hospital and McGill University, Montreal, QC, Canada, Jewish General Hospital, Lady Davis Institute/Segal Cancer Center, Montreal, QC, Canada, Segal Cancer Center/Jewish General Hospital, Montreal, QC, Canada, Segal Cancer Center, Montreal, QC, Canada, Department of Human Genetics, McGill University, Montreal, QC, Canada, Jewish General Hospital, Montreal, QC, Canada, Segal Cancer Centre, Jewish General Hospital, Montreal, QC, Canada, Cook Cty Hosp, Chicago, IL, Centre Hospitalier de l'Université de Montréal, Montréal, QC, Canada, Department of Oncology, McGill University, Montreal, QC, Canada Abstract Disclosures Research Funding Other Quebec Breast Cancer Foundation Background: TNBC, the most aggressive form of breast cancer, is treated primarily with chemotherapy, even before surgery (neoadjuvant chemotherapy or NAC). The prognosis and need for adjuvant therapy depends greatly on the tumor response assessed by pathology (pCR). Highly sensitive and specific ctDNA assays have been shown to be of prognostic value in the metastatic settingbut not yet in earlier settings. Methods: Tissue was collected from 26 Q-CROC-03 clinical trial TNBC patients before, during and after NAC, prior to surgery. Whole exome sequencing on tumor tissues was used to select single nucleotide variants with high allele frequency (VAF), prioritizing TP53, to generateindividual digital droplet PCR (ddPCR) assays. An average of 5 variants (range 1-12) per patient were tested, for a total of 121 variants. A detection threshold was defined for each variant from a pool of normal controls. Median follow-up was 55 months. Results: ctDNA was detectable in 96% of patients at baseline, but 20% of the 121 variants were not detectable at any time point. At baseline, the mean VAF of all analyzed variants, but not of TP53 variants alone, was significantly correlated (p < 0.05) with tumor factors (tumor size, stage, grade, nodal status before and at surgery, RCB score) but not with patient age or BRCA1/2 mutation status. 87 variants (74%) were detected at baseline and their VAF fell by 86% after 1 cycle of chemotherapy (T1). The detection of ctDNA at T1 was associated with DFS (p = 0.027) while the detection of ctDNA at the last post-chemotherapy pre-surgery time point (T4) was strongly associated with pathological complete response (pCR) and both DFS (p = 0.013) and OS(p = 0.006). At this time point, 5 of 41 variants (12%) were detected in pCR patients vs 42 of 80 (53%) in non-pCR, while only 6 of the 15 (40%) non-pCR patients had detectable TP53 variants. Interestingly, for variants detected at baseline, the positive predictive value of T4 ctDNA for disease recurrence was 69%, similar to that of non-pCR, while the negative predictive value of no ctDNA at T4 was 89% for disease recurrence vs 80% for pCR. Conclusions: ctDNA detection after NAC prior to surgery is strongly predictive of disease-free survival and overall survival and is comparable to pCR as a prognostic factor in our cohort (NCT01276899).