RET Fusion Testing in Non–Small Cell Lung Cancer: A Predictive Biomarker for Selective Inhibition

RET fusion testing has become a standard part of the molecular workup for advanced non–small cell lung cancer (NSCLC), and for good reason: it identifies a small but therapeutically important subset of patients whose tumors depend on a druggable oncogenic driver. The purpose here is predictive, not prognostic — the test tells us whether a patient's tumor is likely to respond to a specific class of targeted agents. Getting this determination right, on the right specimen, with the right assay, is where the diagnostic craft lives.

What the Test Measures

RET (rearranged during transfection) encodes a receptor tyrosine kinase that, under normal circumstances, participates in cell growth and differentiation signaling. In a subset of lung adenocarcinomas, a chromosomal rearrangement fuses the kinase domain of RET to an upstream partner gene. The fusion product drives constitutive, ligand-independent kinase activity — in effect, the cell's growth switch is jammed in the "on" position [1].

What matters diagnostically is that the fusion is the oncogenic driver. That's why a well-designed selective inhibitor can be so effective: it targets the very lesion the tumor cannot do without. RET fusions are generally mutually exclusive with other canonical NSCLC drivers, which reinforces the biological logic of testing for them as part of a broad panel rather than in isolation.

How It's Tested

Two principal approaches are in use, and they're not interchangeable. Next-generation sequencing (NGS) — particularly RNA-based or fusion-optimized DNA panels — is the workhorse. It can detect a wide range of fusion partners, including novel ones, and it does so alongside the other drivers a comprehensive panel interrogates. Fluorescence in situ hybridization (FISH), using break-apart probes, detects the physical rearrangement at the RET locus regardless of partner, and it remains a useful orthogonal method.

The specimen is typically formalin-fixed, paraffin-embedded (FFPE) tumor tissue, though circulating tumor DNA (ctDNA) from plasma offers a liquid-biopsy alternative when tissue is scant or the patient can't tolerate a repeat procedure. Preanalytics deserve attention. FFPE fixation degrades nucleic acids, and RNA-based fusion detection is especially sensitive to overfixation, decalcification, and prolonged block storage — factors that can quietly undermine an otherwise elegant assay. Adequate tumor cellularity is a prerequisite; a beautifully run test on a paucicellular sample tells you very little.

Scoring is refreshingly binary in concept: fusion present versus absent. There's no percentage threshold, no intensity gradient to argue over. The interpretive difficulty lies not in scoring but in confidence — was the input adequate, and does the platform reliably capture the fusion in question?

What Each Result State Means

A result of RET fusion present identifies a tumor driven by RET signaling. This is the finding that opens a specific therapeutic door.

A result of RET fusion absent means no rearrangement was detected by the method used — and that qualifier matters. A negative NGS result on high-quality tissue carries more weight than a negative ctDNA result, because tumors shed variable amounts of DNA into the blood. In practice, the most common interpretive error is treating a negative plasma result as definitive. It isn't. When clinical suspicion is high and plasma is negative, tissue confirmation is the sensible next step.

What It Determines for Treatment Eligibility

When a RET fusion is present, the result informs eligibility for selective RET inhibitors — a drug class developed specifically against RET-driven malignancy. Regulatory approval of selpercatinib for RET-altered lung and thyroid cancers established this eligibility framework and cemented RET as an actionable target [2]. The clinical evidence supporting this class derives from RET inhibitor trials, and the biomarker functions as the gate: without a documented fusion, the patient falls outside the population in which these agents were studied and approved [1,2].

It's worth emphasizing what this does and doesn't say. The test predicts eligibility for a therapeutic class; it does not dictate any individual patient's treatment plan, which remains a clinical decision integrating stage, performance status, prior therapy, and patient preference.

Caveats and What's Evolving

A few realities keep this from being as tidy as the binary result state suggests. Assay sensitivity varies by platform and specimen, and the two methods answer slightly different questions — FISH sees the physical rearrangement, RNA-based NGS confirms it's transcribed and identifies the partner. When they disagree, the resolution usually turns on specimen quality and the biological plausibility of the partner detected, so it helps to know which platform generated the result before acting on it.

Liquid biopsy adds both reach and ambiguity. ctDNA is genuinely valuable when tissue is exhausted, yet its sensitivity depends on tumor shed, disease burden, and site of metastasis. A negative plasma result in a patient with low-volume or predominantly intracranial disease should be read with real caution, and reflex tissue testing is often warranted. Fusion partner identity and rare novel fusions also continue to be characterized as sequencing panels broaden — meaning today's "absent" occasionally reflects the limits of yesterday's panel design rather than true absence.

The RET story illustrates something larger about molecular diagnostics in NSCLC: a negative result is not the end of a workup but a demand for rigor. It obligates the pathologist to ask whether the specimen was adequate, whether the method could have detected the alteration, and whether an orthogonal test is needed before the door to a targeted class is closed. In driver-oncogene lung cancer, the most consequential mistakes are rarely false positives — they're the actionable fusions we fail to look for, or fail to trust the negative on. Treating each RET result as a claim with a confidence level attached, rather than a verdict, is the habit that serves patients best.

References

1. Review. Lung Cancer: Targeted Therapy in 2025. Curr Oncol. 2025. PMC11941068 (verify).
2. Markham A et al. / FDA. FDA Approval Summary: Selpercatinib for the Treatment of Lung and Thyroid Cancers with RET Gene Mutations or Fusions. Oncologist. 2021. PMID: 33239432.