Prospective Decision Analysis Study of Clinical Genomic Testing in Metastatic Breast Cancer

Impact on Outcomes and Patient Perceptions

Daniel G. Stover, MD; Raquel E. Reinbolt, MD; Elizabeth J. Adams; Sarah Asad, MPH; Katlyn Tolliver, RN, BSN; Mahmoud Abdel-Rasoul, MS, MPH; Cynthia D. Timmers, PhD; Susan Gillespie, RN, BSN; James L. Chen, MD; Siraj Mahamed Ali, MD, PhD; Katharine A. Collier, MD; Mathew A. Cherian, MBBS; Anne M. Noonan, MBBChBAO, MSc; Sagar Sardesai, MD; Jeffrey VanDeusen, MD, PhD; Robert Wesolowski, MD; Nicole Williams, MD; Clara N. Lee, MD, MPP; Charles L. Shapiro, MD; Erin R. Macrae; Bhuvaneswari Ramaswamy, MD, MRCP; Maryam B. Lustberg, MD


JCO Precis Oncol. 2019;3 

In This Article

Abstract and Introduction


Purpose: To evaluate the impact of targeted DNA sequencing on selection of cancer therapy for patients with metastatic breast cancer (MBC).

Patients and Methods: In this prospective, single-center, single-arm trial, patients with MBC were enrolled within 10 weeks of starting a new therapy. At enrollment, tumor samples underwent next-generation sequencing for any of 315 cancer-related genes to high depth (> 500×) using FoundationOne CDx. Sequencing results were released to providers at the time of disease progression, and physician treatment recommendations were assessed via questionnaire. We evaluated three prespecified questions to assess patients' perceptions of genomic testing.

Results: In all, 100 patients underwent genomic testing, with a median of five mutations (range, 0 to 13 mutations) detected per patient. Genomic testing revealed one or more potential therapies in 98% of patients (98 of 100), and 60% of patients (60 of 100) had one or more recommended treatments with level I/II evidence for actionability. Among the 94 genomic text reports that were released, there was physician questionnaire data for 87 patients (response rate, 92.6%) and 31.0% of patients (27 of 87) had treatment change recommended by their physician. Of these, 37.0% (10 of 27) received the treatment supported by genomic testing. We did not detect a statistically significant difference in time-to-treatment failure (log-rank P = .87) or overall survival (P = .71) among patients who had treatment change supported by genomic testing versus those who had no treatment change. For patients who completed surveys before and after genomic testing, there was a significant decrease in confidence of treatment success, specifically among patients who did not have treatment change supported by genomic testing (McNemar's test of agreement P = .001).

Conclusion: In this prospective study, genomic profiling of tumors in patients with MBC frequently identified potential treatments and resulted in treatment change in a minority of patients. Patients whose therapy was not changed on the basis of genomic testing seemed to have a decrease in confidence of treatment success.


Among patients diagnosed with breast cancer, approximately 30% eventually develop metastatic breast cancer (MBC),[1] and the estimated 150,000 patients living in the United States with stage IV breast cancer are relatively understudied.[2–7] Evaluating multiple somatic mutations in MBC may identify mechanisms of resistance, prognostic genomic biomarkers, or molecular targets to exploit.[8–10] Personalized cancer medicine—tailoring medical decisions by integrating clinical features and demographic factors with genetic information—is promising but its value remains unclear for malignancies such as breast cancer, which harbor few truly predictive genomic alterations.[8,10,11]

Most of our understanding of somatic genomic alterations in breast cancer comes from studies of primary breast cancer,[12–14] but our understanding of genomic data on MBC are not nearly as well developed.[15] In the largest clinico-genomic analysis of MBC to date, 1,918 tumors prospectively underwent targeted DNA panel sequencing.[15] Mutations in frequently altered genes were similar to those in primary cancers, such as TP53 in triple-negative breast cancer (TNBC) and PIK3CA and CDH1 in hormone receptor (HR)–positive, human epidermal growth factor receptor 2 (HER2)–negative patients.[15] In HR-positive/HER2-negative patients, the total number of alterations was only moderately higher in MBC tumors compared with primary tumors.[15] Recent evaluation of genomic targets in breast cancer suggests there are nine alterations with level I/II evidence for actionability:[16]ERBB2 amplifications (level IA), germline pathogenic BRCA1 or BRCA2 mutations (IA), PIK3CA hotspot-activating missense mutations (IA), microsatellite instability (IC), NTRK fusions (IC), ESR1 hotspot-activating missense mutations (IIA), PTEN loss (IIA), AKT1 mutations (IIB), and ERBB2 hotspot-activating missense mutations (IIB).[16]

Although genomic testing is promising for improving treatment efficacy, it is unclear how frequently it changes choice of treatment.[17–19] Complex interactions between genomic testing and patient perception of care make it imperative to understand the impact of genetic sequencing in clinical oncology.[17,20,21] To date, we have little understanding of how patients perceive the value or accuracy of genomic testing, and whether changes in treatment recommendation based on genomic testing affect patients' perceptions of care. There are discrepancies between cancer patient and physician expectations of therapy efficacy; patients demonstrate greater optimism about therapy efficacy.[22–25] Expectations of patients with MBC and their motivations for undergoing somatic genetic testing have not yet been explored.

Clinical tumor genomic analyses typically rely on targeted next-generation sequencing (NGS) of a panel of specific, typically actionable cancer-related genes to analyze somatic gene alterations from biopsy specimens. Most commercial targeted panel sequencing approaches, including FoundationOne CDx (Foundation Medicine, Cambridge, MA) in this study, provide treatment suggestions that are based on genomic alterations and available clinical literature.[26] We hypothesized that prospective implementation of FoundationOne CDx testing into clinical care for patients with MBC will identify patient-specific approaches that offer improved patient outcomes and perceptions of care. In this study, patients with MBC received FoundationOne CDx testing when a new treatment was initiated and genomic testing results released to the provider at the next progression event. The primary objectives of this study were to assess the proportion of patients whose subsequent cancer-related therapy was based on the FoundationOne CDx test results (FoundationOne CDx-supported treatment change), to evaluate time-to-treatment failure (TTF) and overall survival (OS) in patients whose therapy was genomically directed versus not, and to assess patient perceptions of genomic testing.