TL;DR
This paper introduces SwissMTB, a novel workflow for comprehensive molecular cancer diagnostics in Swiss clinics. By integrating high-throughput sequencing (WES, WGS, RNA-seq) with a multidisciplinary molecular tumor board, SwissMTB aims to identify actionable genetic variants and guide personalized therapy decisions for late-stage cancer patients, paving the way for more widespread adoption of precision oncology in Switzerland.
What Problem Does This Paper Address?
Precision oncology, which tailors cancer treatment based on a patient's molecular profile, holds immense promise but faces significant integration challenges in clinical practice. These include the need for rapid analysis turnaround times, interpretation of complex genetic interactions, standardization of workflows, and addressing ethical, legal, and awareness barriers. Current molecular diagnostics in Swiss hospitals often rely on targeted gene panels, which may miss actionable targets or resistance mechanisms. This study addresses the need for a comprehensive, clinic-ready workflow that can analyze a broader genomic landscape and translate these findings into actionable therapeutic recommendations for patients, particularly those with limited treatment options.
Methods
The SwissMTB workflow is a collaborative effort between bioinformaticians and clinicians. It involves:
- Sample Extraction and Sequencing: Tumor biopsies (and matched normal samples where available) undergo DNA and RNA extraction. Sequencing is performed using Whole Exome Sequencing (WES), Whole Genome Sequencing (WGS), and RNA sequencing (RNA-seq), or targeted gene panels.
- Bioinformatics Analysis: This is a largely automated pipeline that includes:
- DNA Analysis: Mapping reads to the human reference genome, followed by somatic variant calling (SNVs, InDels) using multiple callers (MuTect, VarScan2, Strelka) for increased confidence. Copy Number Variant (CNV) calling is performed using BIC-seq2 (for WGS) or EXCAVATOR (for WES). For panel data, platform-specific callers (e.g., Ion Reporter) are used.
- RNA Analysis: RNA-seq data is processed to determine gene expression levels, which are compared to reference cohorts (e.g., TCGA) to identify over- and underexpressed genes.
- Annotation and Prioritization: Identified variants are annotated with information from various databases (COSMIC, dbSNP, ClinVar) and analyzed for their functional impact, associated pathways, and incidence across cancer types. This step helps prioritize variants for clinical relevance.
- Clinical Report Generation: A concise report is generated summarizing actionable therapies and clinical trial opportunities. Therapies are categorized by confidence level (A-D) based on evidence and regulatory approval.
- Molecular Tumor Board Discussion: The report is discussed in an interdisciplinary tumor board to support therapy decisions.
Datasets
The study utilized data from patients treated at the University Hospitals in Zurich and Basel. For the analysis, publicly available reference cohorts were used, specifically the TCGA Skin Cutaneous Melanoma (SKCM) and TCGA Uveal Melanoma (UVM) cohorts for gene expression comparisons.
Key Insights and Ideas
- Comprehensive Sequencing is Crucial: The study highlights that WES and WGS, by covering a broader genomic landscape, identify more actionable variants and potential therapies compared to targeted gene panels. This is particularly important for identifying novel treatment options and understanding resistance mechanisms.
- RNA-seq Adds Value: Integrating RNA-seq data proved beneficial, as gene expression results affected therapy recommendations in a significant portion of the analyzed cases. This suggests that understanding gene activity alongside genetic alterations provides a more complete picture.
- Molecular Tumor Boards are Essential: The workflow emphasizes the importance of a multidisciplinary tumor board to interpret complex molecular data and translate it into clinically relevant recommendations.
- Addressing Resistance is Key: Identifying resistance-associated variants is as critical as finding actionable targets, as it can prevent the use of ineffective therapies and guide alternative treatment strategies.
- Swiss-Specific Workflow: The SwissMTB workflow is designed to adhere to Swiss regulatory guidelines, making it directly applicable within the Swiss healthcare system, unlike approaches developed in other countries.
Key Results
- The SwissMTB workflow successfully identified actionable targets in 86% of the 22 patients included in the pilot study.
- For 19 out of 22 patients, actionable targets were identified, exceeding the generally reported success rate of 60% in the literature.
- In five out of eleven patients analyzed before treatment, the SwissMTB diagnostics influenced the treatment decision. Four of these patients received the recommended therapy, with two showing a positive response.
- The study identified resistance variants for treatments that had proven ineffective, suggesting that SwissMTB could have prevented non-beneficial therapies in retrospect for two patients.
- In cases where RNA-seq data was available (five patients), gene expression results influenced therapy recommendations in three of them.
Limitations and Future Work
- Small Pilot Study Size: The study was a pilot project with a small number of patients, necessitating larger clinical trials to definitively prove clinical benefit.
- Turnaround Time: While bioinformatics analysis is automated, the manual generation of the clinical report remains a bottleneck, impacting the overall turnaround time.
- Patient Drop-out and Deterioration: Rapid health deterioration in late-stage patients limited the number of cases where recommendations could be discussed in the tumor board or acted upon.
- Future Work: The authors suggest extending the workflow to include other molecular profiling technologies like proteomics and single-cell sequencing. They also emphasize the need for further automation of variant prioritization and literature search to reduce report generation time. The increasing feasibility of high-coverage WGS will likely shift the standard from panel sequencing to whole exome and genome analysis.
Summary updated