Oncologists’ use of tumor sequencing findings and patient derived benefit.

Abstract

e13016 Background: Despite increasing use of tumor genomic sequencing in clinical oncology, there is currently limited understanding of how oncologists interpret and apply this information to a patient’s clinical management. Methods: We surveyed oncologists who referred patients with refractory cancers to the Michigan Oncology Sequencing (MI-ONCOSEQ) program which uses integrative clinical sequencing to detect multiple classes of somatic and germline molecular aberrations including point mutations, amplifications, insertions/deletions, gene fusions, and outlier gene expression from June 2014-February 2015. Patients were also surveyed approximately 2 weeks after notification of completion of sequencing. In addition, we conducted chart reviews to ascertain how many patients’ clinical management was changed or derived clinical benefit as a result of sequencing. Results: Forty-three oncologists (37% female, 98% academic medical center practice), referring 112 patients ( M = 2, range: 1-9) completed the survey (response rate = 93%). Oncologists reported an intention to change 22% of patients’ treatment plans on the basis of sequencing results, with referral to a clinical trial (N = 12; 50%) as the most frequently endorsed intended change. However, few patients (N = 9, 38%) had an actual change in clinical management. Barriers identified included patient factors, ineligibility for identified treatment, and loss to follow-up/deceased. Of the 57 patients who completed post-sequencing surveys, 34 (60%) reported that results of sequencing had not been disclosed. However, patient records had documentation that results disclosure had indeed taken place in 41% of cases. Conclusions: Although expectation of enrollment in clinical trials is frequently proposed as the indication for genomic sequencing, practice barriers remain. These include lack of access to studies, severity of patient illness and challenges to effective communication of results.