Hiroshima University Introduces Framework to Identify Cancer-Causing Mutations in Genomic Profiling Data

The Challenge of Variants of Uncertain Significance in Cancer Genomics

Comprehensive genomic profiling (CGP), a targeted sequencing method that examines hundreds of cancer-related genes simultaneously, has revolutionized oncology since its introduction in Japan in 2019. By identifying genomic alterations, CGP guides physicians toward genome-matched therapies like PARP inhibitors for BRCA-mutated cancers. However, a major hurdle persists: variants of uncertain significance (VUS). These genetic changes cannot be classified as benign or pathogenic with current evidence, comprising up to 75% of detected BRCA1/2 variants in real-world data.

In Japan, where over 100,000 patients have undergone CGP by 2025, VUS outnumber confirmed pathogenic variants, complicating clinical decisions. For BRCA1 and BRCA2 genes—linked to hereditary breast and ovarian cancer (HBOC)—this uncertainty delays precision treatments and hereditary risk assessments.

Hiroshima University's Groundbreaking Prioritization Framework

Researchers at Hiroshima University Hospital's Department of Clinical and Molecular Genetics have developed a practical framework to triage BRCA1/2 VUS from CGP data. Published in the European Journal of Human Genetics on March 2, 2026, the study analyzed 2,172 CGP tests from 13 institutions in Japan's Chugoku-Shikoku region between September 2019 and July 2025.

Led by Hiroaki Niitsu, M.D., Ph.D., specially appointed associate professor, and co-authors including Hikaru Nakahara and Takao Hinoi, the team identified 526 BRCA1/2 variants: 130 pathogenic and 396 VUS (153 unique). Their approach integrates computational predictions with functional validation to pinpoint likely pathogenic candidates efficiently. Read the full paper here.

Step-by-Step Methodology: How the Framework Works

The framework employs a sequential filtering process, making it accessible for clinical labs without advanced resources. Here's the breakdown:

• Step 1: Variant Identification - From CGP panels (e.g., FoundationOne CDx, OncoGuide NCC Oncopanel), extract BRCA1/2 variants and classify per ClinVar/ACMG guidelines, excluding common polymorphisms (allele frequency >0.1% in gnomAD/jMorp/ToMMo).
• Step 2: In Silico Prediction - Score 153 VUS using 10 predictors: CADD (≥28.1 damaging), SIFT (≤0.00), PolyPhen-2 (1.0), SpliceAI (≥0.8), REVEL (≥0.5), BayesDel (≥0.6), MetaSVM (≥0.585), PrimateAI (≥0.796), phyloP100 (≥3.0), MaxEntScan (≥3 delta). Thresholds based on Franklin scores for high confidence.
• Step 3: Prioritization - Select top-scoring VUS (e.g., 10 candidates: 8 missense, 2 splice-site) concordant across predictors.
• Step 4: Functional Validation - Perform RNA-based assays like RT-PCR and Sanger sequencing on patient samples to confirm impacts (e.g., splicing defects).
• Step 5: Clinical Correlation - Link to patient outcomes, such as hypersensitivity to platinum chemotherapy or PARP inhibitors, indicating homologous recombination deficiency (HRD).

This systematic method narrowed candidates efficiently, outperforming single-tool approaches like ENIGMA BRCA VCEP.

Key Findings: Reclassifying VUS as Pathogenic

Among prioritized variants, BRCA2 c.67G>C (p.D23H) stood out. Found in multiple patients, functional analysis revealed it causes exon 2 skipping, disrupting the reading frame and BRCA2 function. Two patients—one with duodenal cancer, another prostate—showed exceptional responses to platinum-based therapy and PARP inhibitors, consistent with HRD phenotypes.

The 10 candidates aligned with prior lab studies (e.g., HDR/SGE assays showing loss-of-function), validating the framework's accuracy. VUS were prevalent in non-HBOC cancers, highlighting CGP's tumor-agnostic value.

Hiroshima University's Legacy in Cancer and Genomics Research

Hiroshima University, with its Research Institute for Radiation Biology and Medicine, has long excelled in oncology due to its historical context post-1945 atomic bombings. The Graduate School of Biomedical and Health Sciences drives precision medicine initiatives, including genomic oncology labs focusing on tumor biology and immunotherapy.

The Department of Clinical and Molecular Genetics supports Japan's national CGP reimbursement since 2019, contributing to over 100,000 tests. Collaborations with local hospitals amplify real-world data collection, positioning Hiroshima as a hub for translational genomics. University press release.

Implications for Precision Oncology in Japan

Japan faces ~1 million new cancer cases annually, with breast cancer (~95,000 women/year) and prostate prominent. BRCA1/2 pathogenic variants occur in 5-15% globally, but East Asian databases underrepresent them, inflating VUS rates.

This framework accelerates VUS reclassification, enabling germline testing, family screening, and targeted therapies. It supports Japan's Project HOPE (genomic medicine promotion), potentially expanding to Lynch syndrome genes like MLH1.

Global Relevance and Comparisons

While tailored to Japanese cohorts, the framework's reliance on standard predictors (e.g., SpliceAI) makes it adaptable worldwide. It complements ACMG/AMP guidelines and large-scale assays like HDR, addressing VUS bottlenecks seen in U.S./EU studies (40-50% VUS rates).

East Asian-specific tools like jMorp enhance accuracy, countering gnomAD biases.

Expert Perspectives and Quotes

"This study was motivated by a clinical observation," Niitsu notes. "We encountered a patient who showed an exceptionally strong response to platinum-based chemotherapy." He emphasizes VUS challenges: "In many cases, VUS outnumber clearly pathogenic variants." Future goals include multi-gene expansion and routine in silico integration.

Future Outlook: Expanding the Framework

Niitsu's team plans validation across syndromes and integration into CGP workflows. With Japan's aging population (cancer peak 2020s), such tools are vital. Hiroshima eyes AI enhancements for predictor fusion.

For academics, this underscores Japan's rising genomics prowess, with opportunities in computational biology and functional genomics.

Photo by Andy Wang on Unsplash

Careers in Cancer Genomics at Japanese Universities

Hiroshima University and peers like UTokyo seek experts in bioinformatics, molecular genetics. Roles in precision medicine labs blend research and clinical translation, fueled by MEXT grants. Japan's 2026 genome projects offer postdocs, faculty positions amid CGP expansion.