Gene fusion detection: Liquid biopsies and classification of co-mutations

Andrea Boscolo, from the Veneto Institute of Oncology IOV – IRCCS, showcased the use of VarSome Premium in the recent paper; NGS detection of gene rearrangements and METexon14 mutations in liquid biopsy of advanced NSCLC patients: A study of two Italian centers.

In the paper, a collaboration between Veneto Institute of Oncology IOV – IRCCS and the University Hospital of Parma, the authors evaluate the efficiency of liquid biopsy in detecting gene rearrangements and METexon14 mutations in advanced NSCLC patients.

We caught up with Andrea to learn more about the study.

What are your thoughts on the future of liquid biopsies in a world where tissue biopsies are the gold standard? Will they be a complimentary test? Under what circumstances could they eliminate the need for invasive procedures required for a tissue biopsy?

Andrea Boscolo: I think tissue biopsy will remain the gold standard, but in a new way, along with liquid biopsy. It is fundamental for capturing the entire tumor heterogeneity at diagnosis or progression, and tissue biopsy is spatially limited. Tissue and liquid biopsy offer complementary information about primary tumors and metastases. Liquid biopsy is also necessary in all cases in which an invasive procedure is not recommended or impossible.

What were the main factors influencing the concordance rate between tissue and plasma samples in detecting ALK, ROS, RET fusions and METexon14 mutations in your study?

AB: Based on our experience, two factors influence the concordance of fusions/METex14 in plasma/tissue: biological and technical. Some tumors shed low ctDNA, and the NGS kit used is limited in its sensitivity, guaranteeing fusion detection down to a tumor fraction of 1%. Both combined limit our capacity to detect fusions in liquid biopsy. Research is necessary to improve technology for detecting fusions.

What are some challenges you faced in detecting gene rearrangements and METexon14 mutations in liquid biopsy samples?

AB: Samples with low tumor fractions (<1%) represent a limitation for detecting fusions or other types of mutations. Impaired sensitivity is also highlighted by the fact that other pathogenic mutations were found in fusion-negative liquid biopsy samples.

What strategies did your team use to increase the chances of detecting mutations in ctDNA samples? Any unexpected findings during the study?

AB: To increase the chances of detecting mutations in ctDNA, it is fundamental to correct and prevent pre-analytical and analytical factors that may influence cfDNA extraction efficiency and library preparation. We used cell-free DNA blood collection tubes, which prevent cfDNA degradation and wild-type DNA release stabilizing white blood cells. In the case of EDTA tubes, blood must be processed within 2 hours from the blood draw. The maximum cfDNA input available and supported by the protocol must be used to obtain high-quality and reliable NGS libraries. To obtain a sufficient quantity of cfDNA, we recommend performing extraction from at least 4 ml of plasma, if available. The detection rate of RET fusions (80%) is a surprising and unexpected result since it is not described in literature to the best of our knowledge, and it is probably related to a more shedding phenotype and the presence of extra-thoracic disease in this type of tumor.

You mentioned that you used VarSome Premium for manual variant classification. How did it help you in the accuracy and reliability of your results?

AB: Variant classification is a common problem in the field and needs more standardization; VarSome addresses this obstacle by aggregating precious information from various databases. We used VarSome Premium to classify co-mutations found along with fusions/METex14 in plasma. To assess if a co-mutation may affect prognosis negatively, we selected only those variants classified as pathogenic, likely pathogenic or variants of unknown significance as proposed by the ACMG rules. In addition, we also considered the AMP rules (we kept Tier I, II, and III). Excluding benign and likely benign variants (Tier IV) was fundamental because they could cause noise in the analysis and mask the potentially detrimental effect of other non-benign mutations.

In your research, you mentioned the need for different/more sensitive approaches in detecting gene fusions and METexon14 mutations. Could you discuss any future advancements or emerging technologies that may address these challenges?

AB: Several strategies were identified to enhance the detection of gene fusion, such as increasing the probe coverage of the most common fusion partner genes, including extensive tiling of introns, and creating a more optimized bioinformatics pipeline for calling fusions. Another chance is to integrate cfDNA and cfRNA sequencing in a combined approach.

How do you believe your findings contribute to our understanding in Non-Small Cell Lung Cancer and its treatment?

AB: Our findings suggest that RET mutated disease is easier to detect in liquid biopsy (80%), probably due to a more shedding tumor. In comparison, ROS1 is difficult (18%) due to the peculiar gene structure and breakpoint positions. ALK fusions detection rate is intermediate (38%). In addition, the presence of hepatic metastasis favors fusion detection chances. The presence of other co-mutations, along with fusions such as KRAS and NRAS, is a negative prognostic factor. These co-mutations might be targeted along with the original fusion in a combined approach.

Is there anything else you would like to mention?

AB: This study was a collaboration between Veneto Institute of Oncology IOV – IRCCS and the University Hospital of Parma. I want to thank Dr. Michela Verzè, who co-authored this study and the team from the Hospital of Parma, particularly Professor Marcello Tiseo and Dr. Roberta Minari. In this case, collaboration was essential to get the numbers sufficient to make a relevant statistic since patients with gene fusions are a minority but not less important than others. Fundamental was also our internal collaboration with Medical Oncology 2, particularly with Professor Giulia Pasello and Dr. Daniela Scattolin. Finally, I would like to thank my team from the Basic and Translational Oncology Unit, particularly my mentor, Professor Stefano Indraccolo.

Andrea Boscolo, Researcher, Veneto Institute of Oncology IOV – IRCCS

Andrea Boscolo is a biologist with a PhD in Oncology. He is now a researcher at the Veneto Institute of Oncology IOV – IRCCS, Padua, Italy, in the Basic and Translational Oncology Unit. His work, both in the diagnostic and research field, is focused on liquid biopsy in cancers and Next Generation Sequencing technology, ddPCR, and other molecular techniques.

VarSome in Action

This blog series shows how VarSome supports clinicians and researchers around the world. If you would like to have your publication featured, or if you would like a demo of VarSome Clinical or VarSome Premium, let us know at sales@VarSome.com.