ROS1 Fusion Test: Lung Cancer, Targeted Therapy, and Results

A ROS1 fusion test looks for a rearrangement that joins part of the ROS1 gene to another gene and creates an abnormal growth signal in cancer cells. The finding is most important in non-small cell lung cancer, especially lung adenocarcinoma, because a confirmed ROS1 fusion can make the tumor eligible for a ROS1-targeted tyrosine kinase inhibitor. The alteration is usually acquired by the tumor rather than inherited, so a positive result normally does not mean relatives have the same cancer risk.

Testing may use tumor tissue, cytology material, or circulating tumor DNA from blood. RNA-based next-generation sequencing is often especially effective because it detects the expressed fusion transcript, while DNA sequencing, fluorescence in situ hybridization, and immunohistochemistry each have different strengths and limitations. A negative result is reliable only when the assay covered ROS1 well and the sample contained enough tumor. The complete report should be reviewed with the cancer type, stage, prior treatment, brain imaging, and other molecular findings.

• A confirmed ROS1 fusion is an actionable tumor driver in a small subset of non-small cell lung cancers.
• ROS1 fusions are usually somatic changes in cancer cells, not inherited variants found throughout the body.
• RNA-based testing can detect diverse fusion partners that limited DNA panels or single-gene assays may miss.
• A negative blood test does not exclude a ROS1 fusion when little tumor DNA is circulating.
• Treatment selection depends on the exact result, disease setting, prior therapy, brain involvement, local approvals, and patient factors.

Table of Contents

• What a ROS1 Fusion Means
• Who Should Have ROS1 Testing
• Samples and Testing Methods
• Understanding the Report
• Targeted Therapy Decisions
• Resistance and Repeat Testing
• Limitations and Common Pitfalls
• Next Steps After Testing

What a ROS1 Fusion Means

ROS1 encodes a receptor tyrosine kinase, a protein that can transmit signals telling a cell to grow and survive. In a ROS1 fusion, a chromosome break joins the kinase portion of ROS1 to part of another gene. The partner gene often supplies a structure that keeps the new fusion protein switched on. The result is continuous signaling even when the normal cellular controls are absent.

Several partners have been described, including CD74, EZR, SDC4, SLC34A2, TPM3, and others. The exact partner can help confirm that the event is biologically plausible, but most established in-frame ROS1 fusions are treated as the same actionable class. An in-frame fusion preserves the genetic reading frame needed to make a functional protein. A laboratory may also check that the ROS1 kinase domain remains intact and that RNA evidence shows the transcript is expressed.

ROS1 fusions are best known in non-small cell lung cancer, where they occur in roughly 1% to 2% of cases and are enriched in adenocarcinoma. They can occur in people with any smoking history, although they are more often identified in younger people and never-smokers or light smokers. Clinical features alone cannot identify the alteration reliably; broad biomarker testing is needed.

A fusion is different from a ROS1 point mutation, amplification, or overexpression. Those findings do not automatically predict response to a ROS1 inhibitor. It is also different from a germline result. The typical ROS1 fusion is present only in the tumor and is therefore part of somatic cancer testing. Routine testing of relatives is not indicated solely because a lung tumor has a ROS1 fusion.

A ROS1 fusion often acts as the dominant oncogenic driver. It is commonly mutually exclusive with other strong drivers such as classic EGFR mutations, ALK fusions, and some KRAS alterations, although rare co-alterations occur. When two competing drivers are reported, pathology review and confirmation may be needed to determine whether both are real, whether they occur in separate tumor populations, or whether one is an assay artifact.

The result does not describe the stage or aggressiveness of the cancer by itself. A small resectable tumor and a widely metastatic tumor can carry the same fusion. Staging still depends on imaging, lymph-node assessment, and clinical evaluation. The fusion mainly provides a treatment target and a marker that can be followed when repeat molecular testing is clinically useful.

Who Should Have ROS1 Testing

Current lung-cancer biomarker practice supports ROS1 testing for patients with advanced or metastatic nonsquamous non-small cell lung cancer. Many centers also test selected squamous cancers when the patient is young, has little or no smoking exposure, or when the biopsy is small and an adenocarcinoma component cannot be excluded. Testing increasingly occurs through a broad panel rather than as a stand-alone order because tissue is limited and several actionable genes must be assessed at once.

Testing may also be appropriate in earlier-stage disease. The immediate treatment may be surgery, radiation, or chemotherapy rather than a ROS1 inhibitor, but knowing the molecular profile can help with clinical-trial eligibility and future planning if the cancer recurs. Practices vary by stage, country, and guideline, so the oncology team should explain why testing is being ordered now.

ROS1 testing is especially important when a lung adenocarcinoma has no result yet for other established drivers. It should not be withheld because the patient has a smoking history, is older, or has high PD-L1 expression. Those features change the probability but do not rule out a fusion. Missing the alteration can lead to a treatment sequence that is less effective or more toxic than a targeted approach.

The test may be ordered at diagnosis, at recurrence, or when disease becomes metastatic. A prior negative result deserves review when:

• The old assay tested only a small set of genes or common fusion partners.
• The report used DNA sequencing with limited coverage of large ROS1 introns.
• The specimen had low tumor content or poor nucleic-acid quality.
• Only immunohistochemistry was performed without confirmatory testing.
• The disease has changed and new tissue or plasma is now available.
• The result was “quantity not sufficient,” “indeterminate,” or “no result” rather than truly negative.

A ROS1 fusion can occur in cancers other than lung cancer, but evidence and approvals are not identical across tumor types. Broad tumor genomic testing may discover ROS1 fusions in rare tumors. The treating team should then assess whether the fusion is a recognized driver in that cancer, whether a tumor-agnostic trial accepts it, and whether another alteration better explains the disease.

Testing should be coordinated with pathology before tissue is exhausted. Small biopsies must supply material for diagnosis, immunostains, and multiple biomarkers. Reflex panel testing initiated by the pathology laboratory can shorten delays and reduce repeated cutting of the tissue block. When a repeat biopsy is planned, the interventional team should know that molecular testing is needed so enough viable tumor can be collected and preserved.

Samples and Testing Methods

ROS1 can be assessed with RNA next-generation sequencing, DNA next-generation sequencing, fluorescence in situ hybridization, reverse-transcription PCR, or immunohistochemistry. No method is perfect. The most suitable approach depends on specimen type, available tissue, turnaround time, and whether the laboratory needs to test many biomarkers together.

RNA-based next-generation sequencing

RNA sequencing reads the expressed fusion transcript. It can identify the partner and show whether the fusion is in-frame, making it a strong method for confirming functional rearrangements. Anchored or partner-agnostic designs can detect novel partners without requiring a separate probe for each one.

RNA is more fragile than DNA. Formalin fixation, old tissue, decalcification, and very small specimens can degrade it. A failed RNA assay is not a negative result. The report should state quality metrics and whether a different method is recommended.

DNA-based next-generation sequencing

DNA panels can assess mutations, copy-number changes, and rearrangements in one test. They are convenient when tissue is scarce. However, ROS1 introns can be large and repetitive, and not every panel captures all possible breakpoints. A DNA rearrangement may also require RNA confirmation to prove that a functional transcript is produced.

A broad solid tumor NGS panel should list whether it validates ROS1 fusions and whether the assay uses DNA, RNA, or both. The number of genes on the panel matters less than the quality of its fusion design and validation.

Fluorescence in situ hybridization

Break-apart FISH uses colored probes on either side of ROS1. Separation of the signals suggests a rearrangement. FISH can work on limited formalin-fixed tissue and does not require knowledge of the partner. It does not usually identify the partner or confirm expression, and borderline signal patterns can be difficult to interpret. Complex genomic events may produce false-positive or false-negative patterns.

Immunohistochemistry

Immunohistochemistry detects ROS1 protein in tissue. It is relatively fast and tissue-sparing, so some laboratories use it as a screening test. Expression patterns and intensity can vary, and non-rearranged tumors may stain. A positive or equivocal screen generally needs confirmation with a molecular method before targeted therapy.

Plasma circulating tumor DNA

A blood-based liquid biopsy can provide a rapid option when tissue is unavailable, unsafe to obtain, or too limited. It can also reveal resistance alterations from several metastatic sites. The main limitation is tumor shedding. Some cancers release little DNA into plasma, especially when disease volume is low or confined to the chest or brain. A positive plasma fusion can be highly useful, while a negative plasma result often requires tissue testing if feasible.

Bone specimens need special attention. Strong acid decalcification can damage DNA and RNA. A non-bone lesion, cell block, pleural fluid, or a decalcification method compatible with molecular testing may provide a better sample. The pathology report should document tumor percentage and specimen processing because these details affect confidence in the result.

Understanding the Report

A complete positive report usually names both genes, describes the breakpoint or exons, states whether the event is in-frame, identifies the method, and gives an interpretation of clinical significance. It may also list therapies, guideline associations, clinical trials, and evidence levels. Drug lists generated by software require clinician review because approvals change and may depend on cancer type, stage, and prior treatment.

Fusion detected

A clearly established, in-frame ROS1 fusion that retains the kinase domain is generally considered oncogenic and actionable in advanced non-small cell lung cancer. The report should be reconciled with histology and other drivers. If RNA and DNA agree, confidence is high. If only one method is positive, the laboratory may still consider the result definitive, but unusual findings deserve discussion with a molecular pathologist.

The variant allele fraction on a DNA report is not a measure of how strongly the tumor depends on ROS1 and is not used like a medication level. It is influenced by tumor purity, copy number, normal-cell contamination, and whether the event is easy to quantify. RNA read counts are also assay-specific and should not be compared across laboratories.

No fusion detected

A negative result means the assay did not detect a reportable ROS1 fusion in that specimen. It does not always prove absence. Confidence depends on adequate tumor content, successful quality control, validated breakpoint coverage, and whether RNA was analyzed. The report may be truly negative, technically limited, or noninformative.

Questions to ask after a negative result include:

• Did the sample pass all quality checks?
• What percentage of nucleated cells were tumor?
• Did the test evaluate RNA fusions or only selected DNA regions?
• Was the specimen decalcified or otherwise damaged?
• Was the result negative, indeterminate, or insufficient?
• Would another block, cytology specimen, tissue biopsy, or plasma test add information?

Rearrangement detected but significance uncertain

A test can show a ROS1 break without establishing an expressed, functional fusion. Examples include an unusual FISH pattern, a DNA breakpoint with no RNA support, or an out-of-frame transcript. Reflex RNA testing or an orthogonal method can clarify the event. Treatment should not rest on the word “rearrangement” alone when the laboratory labels its significance uncertain.

ROS1 mutation, amplification, or protein expression

These are not interchangeable with a fusion. A missense mutation may be benign, uncertain, or related to resistance, depending on the exact amino-acid change and clinical setting. Amplification means extra copies of the gene but is not a standard predictive marker for ROS1 inhibitors in lung cancer. Protein staining alone usually serves as a screen rather than final proof.

The report may also include PD-L1, tumor mutational burden, microsatellite status, and other drivers. These results should be integrated rather than ranked mechanically. In an oncogene-driven lung cancer, a high PD-L1 percentage does not erase the significance of a targetable fusion.

Targeted Therapy Decisions

ROS1 inhibitors block the abnormal kinase produced by the fusion. Several agents have shown substantial response rates in advanced ROS1-positive non-small cell lung cancer, but their indications, line of therapy, age restrictions, and availability differ by jurisdiction and change over time. The oncology team should use the current prescribing information and local guideline rather than an old report’s therapy list.

Drug selection considers prior ROS1 therapy, brain metastases, expected central nervous system penetration, resistance profile, other medical conditions, interactions, dosing schedule, access, and likely adverse effects. Newer agents were designed in part to improve brain activity and address some resistance mutations. There is no single best choice for every patient.

Brain imaging is important because ROS1-positive lung cancer can spread to the central nervous system. A drug with intracranial activity may control both systemic and brain disease, although symptomatic or threatening lesions can still require radiation or surgery. The care plan should be coordinated among medical oncology, radiation oncology, and neurosurgery when appropriate.

Common monitoring domains include liver enzymes, kidney function, blood counts, heart rate and rhythm, blood pressure, neurologic symptoms, gastrointestinal effects, edema, vision changes, and medication interactions. The exact profile differs among drugs. Dose interruption or reduction can often manage toxicity while preserving benefit. Patients should not stop or restart a kinase inhibitor without oncology guidance.

The treatment response is evaluated with symptoms, examination, and serial imaging. A dramatic early scan response is encouraging but does not guarantee permanent control. Conversely, minor measurement variation does not always mean resistance. Radiologists and oncologists consider the pattern of change, new lesions, and the possibility of treatment-related effects.

A ROS1 inhibitor can be used in a treatment sequence that includes surgery, radiation, chemotherapy, or clinical trials. Immunotherapy may still have a role, but single-agent immune checkpoint inhibition has often been less effective in oncogene-driven lung cancers than PD-L1 expression alone might suggest. Prior immunotherapy can also affect the timing and toxicity of later targeted therapy. Sequencing should be individualized rather than based on one biomarker in isolation.

The fusion result does not automatically make every ROS1 inhibitor medically appropriate. The cancer must fit the evidence or approved indication, the patient must be able to take the drug safely, and the expected benefit must outweigh risks. A molecular tumor board can help when the cancer type is unusual, the partner is novel, or access requires an evidence-based appeal.

Resistance and Repeat Testing

Most advanced cancers eventually develop resistance even after a strong initial response. Resistance may arise within ROS1 or through another pathway that allows the cell to grow without relying on ROS1. The pattern can differ between metastatic sites, so one biopsy may not represent every resistant clone.

On-target resistance changes occur in the ROS1 kinase domain. The G2032R solvent-front mutation is a well-known example, but other substitutions can interfere with drug binding. Off-target resistance can involve activation of MET, KRAS, or other signaling pathways, changes in cell lineage, or histologic transformation. Some mechanisms are detectable in plasma; others require tissue and microscopic review.

Repeat testing is most useful when progression could lead to a different drug or clinical trial. Options include:

1. Plasma ctDNA, which is minimally invasive and may capture alterations from several lesions.
2. Tissue biopsy, which confirms histology and can identify transformation or a resistance mechanism in a selected site.
3. Both methods, which may provide complementary information when safe and feasible.

A negative plasma result at progression is not proof that no resistance mechanism exists. The sample may have a low tumor fraction. A tissue biopsy may be particularly valuable when progression is limited to the brain because central nervous system disease can shed little DNA into peripheral blood.

Progression is not always all-or-none. If one or two sites grow while most disease remains controlled, local treatment to the progressing sites may allow continuation of the ROS1 inhibitor. Widespread or symptomatic progression more often prompts a systemic change. This approach depends on lesion location, pace, prior treatment, and patient preference.

Repeat testing should distinguish a persistent original fusion from newly acquired resistance. The original ROS1 fusion often remains detectable and confirms tumor DNA is present, but it does not by itself explain why treatment stopped working. The report must be compared with the pretreatment profile.

Limitations and Common Pitfalls

The largest testing pitfall is treating “not detected” as equally reliable across all assays. A high-quality combined DNA/RNA panel on an adequate tumor specimen provides stronger exclusion than a failed RNA assay or a small DNA panel with limited intronic coverage. The laboratory’s method and quality statement deserve as much attention as the result line.

Other common problems include:

• Ordering sequential single-gene tests until the tissue is exhausted.
• Using immunohistochemistry alone as the final evidence for treatment.
• Assuming every ROS1 alteration is a targetable fusion.
• Overlooking a positive fusion because PD-L1 is high.
• Calling an insufficient plasma result a true negative.
• Using a therapy annotation that is outdated or belongs to another cancer type.
• Forgetting to assess brain disease and central nervous system activity of the chosen drug.
• Failing to retest when an old assay did not evaluate RNA fusions adequately.

Preanalytic handling can determine success. Cold ischemia, fixation time, tumor cellularity, necrosis, and decalcification affect nucleic-acid quality. A pathologist may enrich the tumor by marking an area for macrodissection. For pleural fluid or fine-needle aspirates, a well-prepared cell block can be suitable even when no large tissue biopsy exists.

Tumor heterogeneity creates another limitation. A fusion detected in one lesion is usually a truncal driver, but resistance alterations can differ across sites. Blood testing may broaden the view, while tissue preserves spatial and histologic context. Neither method replaces clinical judgment.

Cost and turnaround time also matter. Broad testing can take days to several weeks. When the patient is acutely ill, clinicians may need to stabilize symptoms while awaiting results. Starting an empiric regimen should not prevent reviewing the molecular result as soon as it arrives and changing course when appropriate.

Next Steps After Testing

After a positive result, the patient should receive a copy of the complete molecular report, not only a note saying “ROS1 positive.” The report should identify the sample date, method, fusion partner, interpretation, and quality metrics. Keeping the original document makes future treatment and trial screening easier.

A focused post-test discussion should cover:

1. Whether the fusion is definitively established and expressed.
2. The cancer stage and current treatment objective.
3. Which ROS1-targeted options are current in the patient’s location.
4. Baseline brain imaging and whether central nervous system disease is present.
5. Medication interactions, organ function, and monitoring requirements.
6. The planned imaging interval and criteria for response or progression.
7. What symptoms require urgent contact, such as severe shortness of breath, fainting, new neurologic deficits, jaundice, or uncontrolled vomiting.
8. Whether repeat tissue or plasma testing would be considered at progression.

After a negative result, verify that the broader molecular workup is complete. Advanced nonsquamous lung cancer commonly requires assessment of multiple drivers and immune biomarkers. A “negative ROS1” line is only one part of the profile. If no actionable alteration is found, confirm that the panel had suitable breadth and that the sample was adequate before concluding the tumor is driver-negative.

A genetics referral is not routinely needed for a conventional tumor-only ROS1 fusion. Referral may still be appropriate when the patient has a separate strong hereditary cancer history or when the report identifies an unrelated possible germline variant. Tumor profiling can occasionally uncover findings that need confirmation in blood or another normal sample, but the ROS1 fusion itself usually does not create a familial testing obligation.

Molecular information should remain connected to the treatment record. The oncology team should document the drug, start date, dose changes, best response, progression pattern, and any resistance findings. This longitudinal record helps later clinicians interpret why a therapy was chosen and which options remain.

References

• Advances and future directions in ROS1 fusion-positive lung cancer 2024 (Review)
• Targeting ROS1 Rearrangements in Non–Small Cell Lung Cancer 2025 (Review)
• Evolving Therapeutic Landscape of ROS1-Positive Non-Small Cell Lung Cancer 2025 (Review)
• ROS1-positive non-small cell lung cancer: from genomics to treatment decisions 2026 (Review)
• Testing for ROS1 in non-small cell lung cancer: a review with recommendations 2016 (Review)

Disclaimer

ROS1 test results and treatment choices require interpretation by an oncology team familiar with the assay, cancer stage, and current drug approvals. This information is educational and does not replace personalized medical advice, urgent evaluation of new symptoms, or the current prescribing information for a specific therapy.