Abstract #20

Presenter

Dr. Conway is a Clinical Assistant Professor of Pathology at the University of Michigan. He received his M.D. from the University of California – San Diego School of Medicine in 2016, after which he completed his training in anatomic pathology and neuropathology at the University of Michigan in 2020. He was faculty in the Department of Pathology at the University of Iowa from 2020-2022 and has since been faculty at the Department of Pathology at the University of Michigan. Prior to his career in medicine, Dr. Conway received his J.D. from the University of Wisconsin Law School in 2008 and practiced law for several years, including judicial clerkships with the Minnesota Court of Appeals and Wisconsin Supreme Court. He has a broad spectrum of clinical and scholarly interests in neuropathology, with a particular interest in the ethical and medicolegal aspects of pathology.

Authors

Kyle S. Conway¹, Pouya Jamshidi², Lauren B. Smith¹

1 Department of Pathology, University of Michigan, Ann Arbor, MI

2 Department of Pathology, Northwestern University, Chicago, IL

CanMEDS:
Medical Expert (the integrating role)

Ethical considerations in the use of DNA methylation profiling for tumor diagnostics

Abstract

DNA methylation profiling has become essential for CNS tumor classification. Unlike traditional molecular testing, methylation profiling does not interrogate specific genetic alterations. Rather, it evaluates a genome-wide methylation profile, typically in conjunction with a machine-learning classifier that assigns a diagnostic category. Patients benefit from an unprecedented ability to accurately diagnose and classify tumors, but there is little formal evaluation of the unique ethical considerations raised by this modality. Our objective is to apply principles of medical ethics to the use of methylation profiling, with a focus on three principles: (1) Informed consent: The diagnostic paradigm of methylation is complex for patients and clinicians to understand and risks confusion regarding the diagnostic significance of a methylation profile. Patients may be unaware of what data is stored and how it will be stored, transferred, and re-analyzed for clinical and research purposes. (2) Harm to patients:  Methylation profiling generates epigenetic data that exists indefinitely and can be analyzed in a dynamic fashion. As new classifiers are implemented, a previously unclassifiable tumor may become classifiable without re-sequencing or a clinically significant change in methylation class may occur. (3) Justice and resource utilization: The extensive granularity of diagnostic categories defined by methylation profiling creates a risk of over-testing that may not generate clinically actionable results. The use of machine learning carries specific considerations of justice, including inherent biases in the populations used for classifier generation. Neuropathologists should keep these ethical considerations in mind when implementing methylation as a clinical test or in clinical practice.