One in Five Blood Cancer Diagnoses Gets Changed by Expert Review—Can AI Close the Gap?

When Matthew Matasar, MD, reviews blood cancer diagnoses from community pathologists, he changes the diagnosis in a way that alters guideline-directed treatment in approximately one out of every five cases. That sobering 20% discordance rate underscores a diagnostic precision challenge in hematologic malignancies—and highlights why AI-assisted digital pathology may be less a future promise than a present-day necessity.

Matasar, Chief of the Division of Blood Disorders at Rutgers Cancer Institute and Professor of Medicine at Rutgers Robert Wood Johnson Medical School, characterizes AI-enhanced digital pathology for hematologic cancers as a "here-and-now, slash near future" technology rather than a distant innovation. Given the "extremely heterogeneous" nature of blood cancer diagnoses and their critical impact on patient outcomes, the stakes for diagnostic accuracy are substantial.

The Diagnostic Accuracy Challenge

The 20% management-altering discordance between community pathology and expert hematopathology review represents more than a statistical observation—it translates directly into treatment decisions that can affect patient outcomes. Hematologic malignancies present diagnostic challenges due to their complexity and heterogeneity, requiring specialized expertise that may not be uniformly available across community healthcare settings.

This accuracy gap creates a two-tier system where patients diagnosed at academic centers with specialized hematopathologists may receive different diagnoses than those evaluated in community practices. AI-assisted digital pathology offers a potential solution by democratizing access to expert-level diagnostic precision, effectively bringing specialized knowledge to every microscope through algorithmic assistance.

The Governance Challenge

Despite the clinical readiness of digital pathology AI, a fragmented governance landscape may slow its implementation. Major institutions including Mayo Clinic, Memorial Sloan Kettering, and Rutgers are currently developing isolated, institution-specific AI governance frameworks—an approach that creates operational inefficiency and interoperability challenges.

Matasar and colleague Arturo Loaiza-Bonilla, MD, MSEd, FACP—Systemwide Chief of Hematology and Oncology at St. Luke's University Health Network and Co-founder of Massive Bio—advocate for professional societies to establish standardized governance priorities. Organizations including the American Society for Clinical Oncology (ASCO), American Society of Hematology (ASH), and American Medical Association (AMA) are positioned to create unified frameworks that could accelerate adoption while ensuring appropriate oversight.

The discussion also touched on emerging applications, including research into retinal imaging that might predict hematologic malignancy risk based on blood vessel characteristics—illustrating the expanding frontier of AI-assisted diagnostics beyond traditional pathology slides.

Implications for Healthcare Delivery

For healthcare providers, the documented diagnostic variability represents both a quality improvement opportunity and an operational consideration. AI-assisted digital pathology could potentially reduce the need for routine expert second opinions while improving initial diagnostic accuracy, streamlining care pathways and reducing time to appropriate treatment.

For health technology companies developing hematopathology AI solutions, the market opportunity is clear but complicated by governance uncertainty. Vendors must navigate institution-specific requirements while anticipating potential society-led standards that could reshape compliance landscapes.

For patients diagnosed with hematologic malignancies in community settings, standardized AI implementation could mean access to expert-level diagnostic precision without geographic or institutional barriers—potentially transforming the 20% discordance rate from an accepted reality into an addressable problem.

Sources

1. Is Digital Pathology Ready to Transform Hematologic Diagnosis? — Cancer Network