Specialist multidisciplinary input maximises rare disease diagnoses from whole genome sequencing

The diagnostic bottleneck in rare diseases has shifted from accessing genetic testing to interpreting the data. Previously, diagnosis relied on clinical assessment and targeted gene testing. Now, whole genome sequencing (WGS) or exome sequencing is commonly used, often employing virtual gene panels (VGPs) to filter variants. The 100,000 Genomes Project (100kGP) in England has integrated clinical WGS with research data, forming the basis for the National Health Service Genomic Medicine Service (NHS GMS). This centralized model combines large datasets and bioinformatics with local clinical analysis. Other healthcare systems use more localized or private models. A major challenge is balancing WGS implementation with sufficient data analysis. Genomic medicine is an emerging discipline using genomic data to inform clinical management, similar to radiology's use of medical imaging. In rare diseases, genomic medicine roles may be fulfilled by clinical geneticists or mainstream medical specialists. However, WGS reports can be complex: WGS may miss variants without a broad approach; reports may include variants of uncertain significance (VUS); and post-WGS investigations (like reverse phenotyping and advanced bioinformatics) may improve diagnostic yield. These complexities necessitate a specialist multidisciplinary team (MDT) to interpret WGS results. Researcher-identified potential diagnoses (RIPs) improve diagnostic rates but lack systematic application and equitable access. This study proposes an integrated clinical solution for patients with PMDs following WGS analysis within English national healthcare genetics services, improving diagnostic confirmation rates.

Existing literature highlights the challenges of interpreting WGS data in rare diseases, particularly in complex disorders like primary mitochondrial diseases (PMDs). Studies have shown that standard, automated analysis often misses potential diagnoses, underscoring the need for more comprehensive approaches. The use of multidisciplinary teams and specialized bioinformatics techniques has been shown to significantly improve diagnostic yields in similar studies. However, inconsistencies in research practices and the variable involvement of clinicians have hindered the development of systematic, equitable solutions for the analysis of WGS data. This paper addresses the existing gap by proposing a standardized clinical approach.

This study involved 102 adult patients (55.9% female, 44.1% male; age range 17-81 years, mean age 47.3 years) with suspected PMDs. Patients underwent WGS (Illumina TruSeq, HsSeq 250x) through the 100kGP. Initial analysis involved applying VGPs from PanelApp, prioritizing variants based on ACMG classifications. This standard approach yielded a 16.7% diagnostic rate. An expanded approach involved a genomic medicine clinician reassessing clinical phenotypes, reviewing 'red' status genes from PanelApp, and determining if additional panels were needed. VUSs were reassessed. Data were reanalyzed with expanded gene panels including 'amber' and 'red' genes, considering CNVs and mtDNA. Variants were annotated with CADD and Splice AI scores. Reverse phenotyping and pedigree review were incorporated. Functional validation (qPCR, Western blot) was performed for specific variants. Sanger sequencing validated findings. The study compared the diagnostic yield of the standard and expanded approaches.

The standard approach identified molecular diagnoses in 17/102 (16.7%) patients. The expanded approach, which included a comprehensive review of phenotypes and pedigrees, application of supplementary filters, reassessment of VUSs, and functional validation, significantly improved the diagnostic rate to 31.4% (32/102). An additional 3.9% (4/102) showed strong candidate genes (VUSs). The enhanced approach identified several key factors leading to improved diagnoses: missed intronic variants in recessive genes (MOLCN1, POLR3, MYH2); the need for functional validation for specific variants; the utility of applying additional gene panels based on refined phenotypes; the importance of pedigree review and variant segregation; the contribution of reverse phenotyping; the identification of heteroplasmic mtDNA variants and a mosaic variant in DM2; and the value of updated literature information. Of note, specific genes emerged as crucial findings, such as MYH2, for which functional validation demonstrated reduced transcript levels; COX7B, with increased transcript levels; and other genes identified through additional gene panel applications. Several cases showcased the combined effect of various methodologies, such as integration of reverse phenotyping with literature review or the use of updated HPO terms to improve variant prioritization. The improved diagnostic rate led to significant management implications: eligibility for clinical trials, modification of systemic complication screening, and changes in childhood screening strategies.

This study demonstrates that a specialist, clinically directed approach to WGS significantly improves diagnostic yields in patients with suspected PMDs. The increased diagnostic rate, from 16.7% to 31.4%, highlights the limitations of relying solely on standard, semi-automated analysis. The integration of genomic medicine clinicians within diagnostic genetic services is crucial for maximizing the diagnostic utility of WGS data. The study's findings support advocating for a robust MDT approach, supplementing high-throughput analysis with clinical and bioinformatic oversight. Comparison with a researcher-led study showed that a clinically-driven MDT approach is more effective. The substantial management implications for the newly diagnosed patients highlight the importance of this integrated approach for improving patient care.

This study demonstrates the significant benefit of integrating specialist multidisciplinary input into the analysis of whole genome sequencing data for rare disease diagnosis. The substantially improved diagnostic rate and consequent management changes highlight the value of this model. Future research should focus on implementing and evaluating this model in other disease areas and healthcare systems, exploring optimal MDT composition and workflow. Further investigation into the cost-effectiveness and scalability of this approach would also be valuable.

The study focused on patients with suspected PMDs, limiting the generalizability of the findings to other rare disease categories. The study's reliance on a specific healthcare system (NHS GMS) in England may limit the applicability of the model to other healthcare settings. The relatively small sample size may limit the statistical power of some of the findings. The study did not compare its results directly with fully automated methods, only against their initial standard method.