Serial sequencing of plasma DNA noninvasively identifies mutations associated with drug resistance.
Major finding: Serial sequencing of plasma DNA noninvasively identifies mutations associated with drug resistance.
Approach: Exome sequencing of plasma DNA was performed before and after multiple systemic therapies.
Impact: Serial liquid biopsies to track tumor evolution in response to treatment are feasible.
Serial tumor biopsies can identify underlying causes of resistance to cancer therapies, but are invasive and frequently unattainable. The amount of cell-free circulating tumor DNA isolated from plasma is a potential biomarker for overall tumor burden or for tumor recurrence and metastasis, but whether information on mechanisms of resistance to targeted or cytotoxic agents can be gained from so-called “liquid” biopsies remains untested. Murtaza and colleagues performed whole-exome sequencing on plasma samples collected from 6 patients with advanced cancer before and after multiple consecutive therapies over the course of up to 2 years. For 2 cases, tumor samples collected at initial diagnosis and metastatic tumor biopsies collected concurrently with plasma samples were also available, and many mutations were found in both tumor and plasma biopsies. The authors compared the mutant allele fraction in plasma before and after each treatment and identified a median of 49 acquired mutations in each comparison. One patient with breast cancer showed an increased abundance of a MED1 mutation and an ATM mutation following combined treatment with tamoxifen and trastuzumab and an increased abundance of a GAS6 mutation and a TP63 mutation after subsequent lapatinib and capecitabine treatment. Other examples included an RB1 mutation that arose in the plasma DNA of a patient with ovarian cancer treated with cisplatin, and a secondary EGFR mutation that appeared in the plasma DNA of a patient with EGFR-mutant lung cancer that progressed on gefitinib. These findings provide proof of principle that serial whole-exome sequencing of plasma DNA can identify mutations that are representative of primary and metastatic tumors and be used repeatedly and noninvasively to characterize tumor evolution in response to cancer therapy.