Fusions can be difficult to detect from DNA alone. RNA can reveal expressed transcripts that may otherwise stay hidden. This lesson has been established in tissue. In blood, the evidence is now catching up.

An estimated 120,000 cancer patients in the United States may have actionable gene fusions each year*. [1, 2]

That accounts for a significant number of patients whose cancers may carry a target that can be matched to a specific therapy, an approach that has improved outcomes in selected molecularly defined cancers. [3]

But there is a caveat: the target has to be found.

Gene fusions are one class of cancer-driving alterations. They occur when parts of two genes become abnormally joined, creating a new hybrid gene. In some cancers, that rearrangement becomes the driver mutation, the genetic alteration that fuels cancer growth.

In non-small cell lung cancer (NSCLC), fusions involving ALK, ROS1, RET, NTRK, NRG1, and other genes can define the disease at a molecular level. Many of these fusions now have matched targeted therapies or emerging therapeutic strategies. This is one of the central shifts in modern oncology: treatment is increasingly guided not only by where the tumor started, but by what is driving it.

The question now is no longer whether fusions matter. It is how to ensure we see these fusions.

The Cleaner Signal in RNA

DNA-based next-generation sequencing (NGS), a method that reads many genes at once to map cancer alterations, has become central to cancer profiling. It detects point mutations (single-base changes in DNA), insertions and deletions, copy number changes, resistance alterations, and many structural events such as rearrangements. But fusions behave differently.

At the DNA level, the breakpoint, where one gene meets another in a fusion, may sit inside a long intronic region. Introns are stretches of DNA between the protein-coding pieces of a gene. They can be large, repetitive, and difficult for targeted panels to cover without compromising on depth. As a result, some fusions can be missed by DNA analysis alone.

In contrast, RNA gives a much cleaner view. When a gene is transcribed into RNA form, introns are spliced out, leaving only the coding pieces joined together. The resulting fusion transcript is therefore shorter and more directly readable. For that reason, professional guidelines increasingly support RNA-based NGS for fusion and splice-event detection, particularly when DNA testing does not identify a driver. [4]

Current evidence in tissue is strong. In a cohort of 5,570 patients with advanced NSCLC, analyzing both RNA and DNA side by side identified 15.3% more patients with actionable structural variants than DNA-NGS alone. [5] These included ALK, RET, ROS1, and NTRK1/2/3 fusions, as well as MET exon 14 skipping alterations. [5] 

DNA shows the rearrangement while RNA shows the expressed event. Both are important in understanding the disease. 

The Same Problem Exists in Blood

We have come to expect RNA in tissue profiling. The question is why we should expect less from liquid biopsy? 

In advanced NSCLC, up to 1 in 5 tissue biopsies may be insufficient for broad molecular testing across approved biomarkers. [6] Liquid biopsy offers a faster, safer alternative. 

Liquid biopsy is often used when tissue is insufficient, unsafe to obtain, or no longer representative of the current disease. In advanced lung cancer, plasma testing is increasingly ordered alongside tissue testing because waiting for tissue alone can delay molecular answers. However, many liquid biopsy conversations still revolve solely around ctDNA. 

There's no denying the clinical utility of ctDNA, it remains the core source of information from plasma NGS. But for detection of fusions, it can face the same limitations as tissue DNA: complex breakpoints, low tumor fraction, incomplete capture of rearranged regions.

The same rationale that supports RNA in tissue now has growing support in liquid biopsies. The BRIGHTSTAR ALK-positive cohort reported a 35.7% increase in ALK fusion detection. [7] A Japanese RNA-based NGS cohort detected ALK and ROS1 fusions in 77.8% of fusion-positive samples versus 33.3% by DNA-NGS. [8] And a pan-cancer analysis of 1,007 plasma samples found a 36.7% increase in actionable fusion yield with ctRNA, with five of seven patients with ctRNA-only fusions in RET, ROS1, or ALK responding to matched targeted treatment. [9]

These studies may differ in design, cohort size, assay, and endpoint. Nonetheless, they all support that ctRNA can identify clinically relevant fusions that ctDNA alone may miss.

The Answer Lies in Another Analyte

A reported case illustrates the problem clearly.

A 50-year-old woman with lung cancer was mechanically ventilated in an intensive care unit. Tissue from a cervical lymph node biopsy was inadequate for molecular profiling. A plasma ctDNA-only test did not identify a treatment target.

A combined ctDNA and ctRNA liquid biopsy detected an actionable ALK fusion in ctRNA within 7 days. She was treated with an ALK-targeting tyrosine kinase inhibitor, a targeted drug treatment that blocks the signal driving an ALK-fusion cancer. Subsequently, she was able to be weaned off the ventilator and discharged. At one-year follow-up, no disease progression was reported. [10]

A single case should not be overstated. Individual outcomes vary and molecular testing does not treat cancer by itself, but this case shows a clear clinical gap.

Tissue was inadequate, ctDNA was uninformative, and ctRNA found the expressed fusion, a clear indication of real-world utility.

Not More Testing. Better Signal.

The lesson from tissue NGS is already clear. For fusions, DNA alone does not always tell the whole story. 

ctDNA remains essential. It carries much of the mutation information that liquid biopsy depends on. But when the clinical question is fusion detection, ctRNA can add a layer of signal that DNA alone may miss.

RNA can reveal the expressed transcript, the signal that shows the fusion is not only present but being read by the cancer cell. In tissue, this has become part of how clinicians think about fusion detection. Liquid biopsy is now reaching the same point.

Not more testing for its own sake. Better matching of signal to biology.

The goal is simple: avoid missing a target that could guide the next clinical step.

*This figure is based on extrapolation from a 9,624-sample cohort and approximately 2 million new cancer cases annually.

References

1. Gao Q, Liang WW, Foltz SM, et al. Driver fusions and their implications in the development and treatment of human cancers. CellRep. 2018; 23(1): 227–238.e3. 

2. National Cancer Institute. Cancer Statistics. Accessed August 15, 2025. 

3. Peters S, et al. Alectinib versus crizotinib in previously untreated ALK-positive advanced non-small cell lung cancer: final overallsurvival analysis of the phase III ALEX study. Ann Oncol. 2026;37(1):92-103. 

4. Ettinger DS, Wood DE, Aisner DL, et al. NCCN Clinical Practice Guidelines in Oncology: Non-Small Cell Lung Cancer. NationalComprehensive Cancer Network. Accessed 2025. 

5. Owen D, Ben-Shachar R, Feliciano J, et al. Concurrent DNA- and RNA-based next-generation sequencing in advanced non-smallcell lung cancer. JAMA Netw Open. 2024; 7(11): e2442970. 

6. Wu CH, et al. Plasma-based next generation sequencing in advanced non-small cell lung cancer: significance in diagnosis andtreatment in Asian patients. BMC Cancer. 2025. 

7. Heeke S, Gandhi S, Tran HT, et al. Plasma ctRNA improves ALK fusion detection in BRIGHTSTAR. JTO Clin Res Rep. 2025; 6: 100795. 

8. Hasegawa N, Kohsaka S, Kurokawa K, et al. RNA-NGS for detection of ALK and ROS1 fusions in non-small cell lung cancer. CancerSci. 2021; 112(10): 4393–4403. 

9. Poh J, Pek M, Petro D, et al. Clinical utility of ctRNA in a combined ctDNA and ctRNA next-generation sequencing pan-cancerliquid biopsy assay. J Clin Oncol. 2025; 43(16_suppl): 3016. 

10. Chan D. Paper presented at: ASCO Annual Meeting; June 2, 2025; Chicago, IL.