Detection of somatic mutations in tumor biopsies and circulating tumor DNA during neoadjuvant endocrine therapy in breast cancer patients

Abstract

Background and aim: Breast cancer is one of the most prevalent forms of cancer in the world. About 70% of breast cancer patients have hormone-dependent tumors, determined by estrogen receptor positivity. Endocrine therapy is used to target estrogen-dependent tumors, however resistance to these therapies has become an increasingly large problem. Although tumor biopsies are helpful when monitoring the effect of cancer treatments, collecting multiple tissue biopsies is invasive for the patient, and the samples do not account for tumor heterogeneity. Liquid biopsies give rise to the opportunity of analyzing tumor derived components that are released into the bloodstream. As liquid biopsies can be obtained from patients in a non-invasive manner, they make it possible to frequently monitor the effect of an ongoing endocrine therapy on the tumor. The aim of this pilot study was to evaluate the use of circulating tumor-derived DNA from liquid biopsies to monitor the acquired tumor mutations during neoadjuvant endocrine therapy in breast cancer patients.

Materials and method: Plasma samples and tissue samples collected from breast cancer patients who had received two different aromatase inhibitor treatments were used for this project. The samples were collected before treatment was administered, and after six months of treatment. The nucleic acids in the samples were extracted and then sequenced by two different sequencing platforms using sequencing panels that target key cancer related genes. A total of 18 patients were included in this study. Plasma samples from both timepoints were available from all patients, while only 15 tissue samples were available, as there lacked samples due to either lack of tissue in the FFPE sample, or complete response to treatment leaving no residual tumor to sample. The mutations identified in the solid tumors and liquid biopsies before and after aromatase inhibitor therapy were assessed in relation to aromatase inhibitor treatment and resistance. The mutations in the liquid biopsies were also evaluated to assess if they were consistent with the mutations in the tissue biopsies.

Results: Assessment of the mutations observed in the tissue biopsies showed that three patients experienced either amplification of FGF3 and MDM2, loss-of-function in SETD2 due to an insertion, or a hotspot mutation in KRAS after six months of AI-treatment. The most frequent mutation was observed in the PIK3CA gene, which was present both prior to and after therapy in three patients. In one patient mutations in CDK2, CDKN1B and PMS2 disappeared after treatment, while another experienced loss of a PTEN mutation.

Assessment of the results from the plasma samples showed variants were only observed in the two samples that contained the highest concentration of cell free DNA. No variants were observed in the remaining 16 samples. When comparing the mutations observed in the tissue samples and the plasma samples there was little consistency. For one patient a mutation in AKT1 was detected only in the plasma sample, while a mutation in KRAS was only observed in the tissue sample. For the other patient the consistency was better. A mutation in ERBB3 was only observed in the plasma sample, while mutations in BRAF, PIK3CA and RET were observed in both sample types.

Conclusion: Sequencing of plasma samples and tumor tissue revealed the presence of mutations prior to treatment that could potentially impact the response to AIs. The emergence of new mutations post-treatment, possibly acquired as an adaptive mechanism to avoid the effects of treatment, were also observed. Comparison of the observed mutations in the tissue and plasma samples showed varying consistency, however this project proved that it is in principle possible to detect acquired tumor mutations during neoadjuvant therapy of breast cancer patients using ctDNA. The low amount of ctDNA seemed to have limited the results.