Table 3.
Summary of advantages and disadvantages of common methods for detection of gene fusions.
| Method | Advantages | Disadvantages | References |
|---|---|---|---|
| FISH | Diagnostic gold standard, high sensitivity and specificity | Laborious. Visual quantification, difficult to automate, requires pre-defined knowledge of fusions. Low resolution, problematic detection of fusions within the same chromosome. Cannot detect fusions with novel partners | Thompson et al.,176 Nguyen et al.,177 Ali et al.,178 Cruz-Rico et al.,180 De Luca et al.,184 |
| IHC | Common detection method | Detects protein overexpression, but not fusions | Thorne-Nuzzo et al.,185 Boyle et al.,186 Shia.256 Zhang.257 |
| RT-PCR methods | Diagnostic gold standard, excellent sensitivity and specificity, widely accepted and cheap. Easy to multiplex and automate. Straightforward interpretation, best suited for clinical laboratory analysis | Requires information about location of fusion breakpoint. Cannot detect novel fusions | Ali et al.,178 Cruz-Rico et al.,180 Lyu et al.,192 Sorber et al.,194 Abbou et al.,195 Shelton et al.196 |
| DNA sequencing of tumors | Allows unbiased detection of mutations and fusions. Covers both transcribed and non-coding regulatory elements of the genome. For example, allows detection of an active promoter – oncogenic transcript fusions | All NGS methods require bioinformatics support. Analytical methods are not standardized yet. Detects expressed and non-expressed fusions. The whole genome deep sequencing is still expensive. Lower sensitivity compared to RNA-seq-based assays | Gerstung et al.,258 ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium,72 Zhang et al.259 |
| Circulating cell-free DNA sequencing | Minimally invasive analysis of tumor DNA for monitoring of the disease. Straightforward DNA extraction from serum | Relatively high frequency of false-negative results. Accurate sample preparation is needed. Plus, the above disadvantages of DNA sequencing methods | Hofman,212 Dagogo-Jack et al.,162 Blaquier et al.,213 Leighl et al.214 |
| Targeted DNA/RNA sequencing | Relatively high-throughput with high coverage of target fusion sites | Collects only information about tested regions and/or fusion types | Heydt et al.,208 Gasc et al.,219 Zheng et al.220 |
| Bulk RNA-seq | Allows detection of known and novel fusions and the presence of an ORF for possible protein expression. Filters out passenger mutations. Permit analysis of archived FFPE samples. Relatively cheap and achieve high coverage. Allows high throughput. Simultaneously measures level of gene expression | May require further standardization. Can be technically challenging | Sorokin et al.,5 Sorokin et al.,197 Winters et al.,225 Peng et al.240 |
| Single-cell RNA-seq | Measures clonality of tumors. Independently measures gene expression and T-cell receptor heterogeneity of immune cells and quantity and gene expression of other non-cancerous cells | Requires additional equipment for single-cell library preparation. Relatively expensive and requires extensive bioinformatics support. Low coverage in terms of reads per single cell, so very low sensitivity of fusion detection. Clinical benefit is questionable | Rozenblatt-Rosen et al.,241 Zeng et al.,242 Nieto et al.,243 Amir et al.,260 Becht et al.,261 Maynard et al.244 |
FFPE, formalin-fixed paraffin-embedded; NGS, next-generation sequencing; ORF, open reading frame.