As cancer becomes more complex, obtaining an accurate, timely diagnosis through imaging and pathology becomes critically important to treatment, said speakers at a workshop meeting of the National Cancer Policy Forum (NCPF) of the National Academies of Sciences, Engineering, and Medicine in Washington, D.C. The February meeting explored strategies to ensure that all cancer patients have access to the best in high-quality diagnostic and treatment technologies; a written summary report on the meeting's findings will be published in several months.
The NCPF meeting was the first in a series of two. The second, to be held in the fall of 2018, will highlight the clinical applications of computational methods and big data in precision oncology. Deliberations from the first meeting will undergird the second.
Industry Trends
"The most important thing you can do for a patient is to get the diagnosis right," said Richard L. Schilsky, MD, FACP, FSCT, FASCO, Senior Vice President and Chief Medical Officer of ASCO and a meeting planning committee member for the event.
"We all believe that good care starts with a good diagnosis. There will not be precision medicine without precision diagnosis," agreed meeting chair Hedvig Hricak, MD, PhD, Drhc, a member of the NCPF and Chair of the Department of Radiology at Memorial Sloan Kettering Cancer Center, Professor in the Gerstner Sloan Kettering Graduate School of Biomedical Sciences, and Professor of Radiology at Weill Medical College of Cornell University, New York, N.Y. "Imaging is on everyone's mind," she added, noting that imaging provides a road map that is essential for every stage of cancer care: screening, diagnosis, treatment planning, and follow-up.
Today, about half of all cases in any imaging practice are cancer-related, noted Hricak. But, the need for excellent training and expertise in oncologic imaging is "not widely recognized or acknowledged." She said there are few role models in oncologic imaging and a lack of appreciation for the value of additional knowledge gained on the part of those who specialize in oncologic imaging.
Agreeing on the need for better access to high-quality diagnostic imaging and pathology was Otis W. Brawley, MD, MACP, FASCO, FACE, Chief Medical and Scientific Officer of the American Cancer Society, Professor of Hematology, Oncology, Medicine and Epidemiology at Emory University, Atlanta, and a member of the NCPF. Today, he stressed, diagnosing cancer has become far more complex because a 21st century definition of cancer includes genomics as well as histology.
"Our health system is incredibly inefficient," said Brawley, noting that some people receive too little care and some not enough. For example, inadequate staging has led some to conclude that colon cancer is more aggressive among blacks. Understaging among poor people with cancer is all too common, he said, and "is totally related to workload."
The rapidly evolving area of molecular diagnostics has made the field of pathology much more complex, confirmed meeting planning member Michael Cohen, MD, Anatomic Pathologist and Interim Chair at Wake Forest University School of Medicine, Winston-Salem, N.C. He noted that today there are numerous "boutique" vending companies that offer proprietary diagnostic algorithms. But he said the interpretation of the results of these tests "can be challenging," especially if the relevant expertise is not readily at hand. He stressed the need for robust quality control of such tests.
The field of pathology is becoming much more complex and technical today, agreed Richard Friedberg, MD, PhD, FCAP, MHCM, CPE, Chair of the Department of Pathology, Diagnostic Medicine Services, and Professor and Chair of the Department of Pathology at the University of Massachusetts Medical School-Baystate. But, he noted, pathology is being held back by late adoption of information technology. "Pathology is 3 decades or so behind radiology in terms of digital technology," he said. "Part of the problem we have is that pathology has not gone digital." But it will, he emphasized.
Multidisciplinary Care, Education
Speakers agreed that diagnostic professionals need to be more fully integrated into cancer care, rather than working in their own silos. Fully integrating diagnostics into cancer care "is no question our future," said Hricak. She said that moving toward such integration to improve diagnosis will require collaboration and a commitment to change. The Association of Community Cancer Centers plans to launch a new integrated program, starting with pathologists, that will bring medical oncologists, pathologists, and radiologists together at the community level where the majority of cancer patients receive care.
Multidisciplinary tumor conferences are an important tool for integrated care, noted Ritu Nayar, MD, Professor and Vice Chair of Pathology and Director of Cytopathology at Northwestern University Feinberg School of Medicine, Chicago, and a trustee of the American Board of Pathology. Collaboration and communication on cancer cases, however they occur, are necessary to decrease diagnostic errors, agreed Dana Siegal, RN, CPHRM, CPPS, Director of Patient Safety and Interim AVP for CRICO Strategies, a division of the Risk Management Foundation of the Harvard Medical Institutions.
A better understanding of the systemic nature and genomic basis of cancer has led to the need for cancer imaging subspecialty training, stressed Fiona Fennessy, MD, PhD, Program Director of the Cancer Imaging Program at Dana-Farber Cancer Institute, Brigham and Women's Hospital, and Harvard Medical School, Boston. She noted that subspecialty training is necessary today given the need for imaging specialists to be familiar with genomic classification of tumor types, types of tumor response criteria, toxicity and adverse events to molecularly targeted therapies, the role of radiology in early-phase clinical trials, and imaging's role in patient management. Fennessy said that digitized pathology correlated with radiology is the future of cancer diagnostics.
Hricak agreed on the need for oncologic subspecialization in radiology. "Subspecialization in radiology is absolutely essential," she said. She advocated for more fellowship training and an oncologic imaging certificate of added qualification (CAQ) based upon increased education and training.
Improving Cancer Diagnostics
Several speakers spoke on the potential of effective clinical decision support (CDS) tools in making an accurate cancer diagnosis. CDS tools need to be efficient, evidence-based, brief, and actionable, said Ramin Khorasani, MD, MPH, Professor and Vice Chair of Radiology at Harvard Medical School, Distinguished Chair of Medical Informatics, and co-founder and Director of the Center for Evidence-Based Imaging at Brigham and Women's Hospital, Boston.
But there are barriers to the use of CDS, said planning committee member Kojo S.J. Elenitoba-Johnson, MD, Founding Director of Penn Medicine's Center for Personalized Diagnostics, Chief of the Division of Precision & Computational Diagnostics, inaugural Peter C. Nowell, MD, Professor at the University of Pennsylvania Perelman School of Medicine, Philadelphia, and NCPF member. These barriers include lack of feedback on CDS use; the difficulty of incorporating guidelines into CDS, since guidelines are not written in a way computers can understand; lack of leadership for CDS; and risk of physician burnout. On the latter point, CDS "alert fatigue" from computerized flags and pop-ups is a real problem for physicians, said Brian H. Shirts, MD, PhD, Assistant Professor of Laboratory at the University of Washington, Seattle. Also, he said, CDS interoperability across institutions remains a major challenge.
Clinicians are overwhelmed by integrating all the diagnostic information from different sources for one patient into a cohesive whole, said Mary Zutter, MD, Vice President for Integrative Diagnostics, the Louise B. McGavock Chair and Professor of Pathology, Microbiology, and Immunology, as well as Cancer Biology at Vanderbilt University School of Medicine, and co-leader of the Host Tumor Interaction Program at Vanderbilt Ingram Cancer Center, Nashville, Tenn. One solution for that work overload: Diagnostic Management Teams (DMT), a collaborative effort among pathologists, clinicians, and biomedical informatics. DMTs generate a comprehensive report, which Zutter said physicians love because it saves them time, and which has saved money for insurers by reducing unnecessary testing. Vanderbilt has pioneered such teams in hematology, and "We're now moving into solid tumors," she noted.
Another strategy to improve cancer diagnostics, especially in rural areas, is telehealth, exemplified by Project ECHO, said Ellen Baker, MD, MPH, Director of Community-Based Cancer Prevention and Control Initiatives, Director of Project Extension for Community Healthcare Outcomes, and Operations Lead for Cervical Cancer Prevention in Resource Limited Settings at the University of Texas MD Anderson Cancer Center, Houston. Project ECHO is a telehealth program originally established in New Mexico to respond to the hepatitis C crisis. This is provider-to-provider telehealth/telementoring/teletraining on cancer, said Baker, with the goal of "demonopolizing" knowledge in big academic cancer centers and bringing it out to the community.
Physicians in the community have responded very favorably to the program, she noted. Clinicians participating in Project ECHO, which has been effective in lowering high cervical cancer rates in rural southern Texas, can earn continuing medical education credits.
Yet another strategy to improve diagnostic accuracy in oncology is seeking second opinions. Schilsky told Oncology Times that some large employers routinely seek second opinions from major medical centers for employees diagnosed with cancer before therapy starts to make sure the diagnosis is accurate and their employees receive the right treatment.
Grand Rounds is a data-driven approach to achieving an accurate diagnosis by matching patients and expert physicians through phone, Web, video chat, and a mobile app. One of its aims is to arrive at an accurate, timely diagnosis without causing the patient to have to travel, said Lawrence Hofmann, MD, co-founder of Grand Rounds, Chief of Interventional Radiology at Stanford Hospital, and Professor of Radiology at Stanford University School of Medicine, Stanford, Calif.
"Low-quality cancer care drives high costs for employers," noted Hofmann. Grand Rounds, which is used by some major employers, has developed physician quality-rating models based on publicly available and proprietary data to measure expertise. In breast oncology, for example, the model finds that physicians with higher expertise are more likely to perform genomic tests.
Speakers at the NCPF meeting noted that insurance reimbursement policies need to be aligned with strategies designed to help integrate high-quality diagnostic radiology and pathology services more fully into the continuum of cancer care. "I think it's going to come if the insurance incentives are aligned," noted Cohen.
Peggy Eastman is a contributing writer.