Tech advances are enabling faster, more accurate pathological assessments of cancer


During the Educational Session Pathology for Cancer Researchers on Saturday, April 9, a panel of pathologists reviewed the important and evolving role of pathology in cancer research and discussed how advances in technology, including artificial intelligence (AI), are improving the accuracy and specificity of tumor assessment. The session can be viewed on the virtual platform by registered meeting participants through July 13, 2022.

Jiaoti Huang, MD, PhD
Jiaoti Huang, MD, PhD

Jiaoti Huang, MD, PhD, Duke University School of Medicine, opened the session with an overview of how tumors are pathologically assessed and characterized. Accuracy in classifying and grading tumors, he said, is key to determining a prognosis and staging of the cancer.

“A given organ can give rise to multiple different tumor types, and classification is generally based on morphology. But, these days, molecular changes and genetic alterations are often considered together with morphology to help us form an accurate classification,” Huang said. “Classification has important clinical significance because different tumors have different biological behaviors, and the prognosis and treatment differ accordingly. Similarly, when it comes to grading, two different patients can have the same tumor type but their prognosis can be very different due to different grades of tumors. Additionally, different tumors in different organs have different grading systems.”

Beatrice S. Knudsen, MD, PHD
Beatrice S. Knudsen, MD, PhD

Beatrice S. Knudsen, MD, PhD, University of Utah, followed with a presentation about the growing field of computational pathology and the utilization of pathomics, which combines AI and digitalized pathology to analyze histopathology images.

“Recent advances in machine learning and AI have provided us with new tools to turn pathology slides into large, pathomics datasets that can be used for improving the accuracy of cancer detection, grading, and prognosis,” Knudsen said. “In 2021, for example, the U.S. Food and Drug Administration approved an algorithm for prostate cancer diagnosis. The software improved detection of prostate cancer by 7.3 percent.”

This is an example, she said, of how a pathologist-inspired algorithm can improve the diagnostic accuracy of community pathologists and reduce the use of immunohistochemistry to confirm the diagnosis.

Angelo Michael De Marzo, MD, PhD
Angelo Michael De Marzo, MD, PhD

Angelo Michael De Marzo, MD, PhD, Johns Hopkins University School of Medicine, concluded the session with a look at how molecular pathology tools are leading to novel insights and informing tissue-based genomics in cancer. He focused on prostate cancer precursor and advanced lesions to showcase molecular, pathology-based tools and approaches used to inform, guide, and complement genomic approaches.

“Tissue-based genomic studies can be powerful tools for studying and understanding molecular pathogenesis, therapeutic targeting, and drug-resistance mechanisms in cancer,” De Marzo said. “Mitochondrial DNA (mtDNA) copy number, for example, has been associated with a variety of aging-related diseases, including prostate cancer. Several studies using inhibitors of mtDNA replication and transcription to treat cancer have shown promise.”

[sub-post-content]