ROS1 Rearrangements in Lung Cancer

by Matthew Cecchini, MD PhD FRCPC
March 20, 2026

ROS1 is a gene that encodes a receptor protein involved in signalling pathways that regulate cell growth and survival. In healthy adult lung tissue, ROS1 is essentially inactive. In approximately 1–2% of non-small cell lung cancers, a chromosomal rearrangement fuses the ROS1 gene to a partner gene, creating an abnormal fusion protein that is permanently switched on and continuously drives cancer cell growth. Although ROS1 rearrangements are relatively uncommon, they are one of the most important alterations to identify in lung cancer, because patients whose tumours harbour a ROS1 rearrangement can achieve exceptional and durable responses to ROS1-targeted drugs. The characteristics of patients who develop ROS1-rearranged lung cancer — typically younger, often never-smokers, frequently with lung adenocarcinoma — make accurate identification particularly meaningful, as many of these patients have long lives ahead of them and benefit greatly from effective, well-tolerated targeted therapy over the long term.

What the test looks for

The ROS1 gene sits on chromosome 6. In a small subset of lung cancers, a structural rearrangement — a break in the chromosome that causes ROS1 to fuse with a different gene — creates a hybrid fusion gene. The resulting fusion protein retains the signalling domain of ROS1 but is now constitutively active, meaning it continuously transmits growth signals regardless of whether the cell actually needs to divide.

More than 20 different fusion partners for ROS1 have been identified in lung cancer, including CD74, SLC34A2, EZR, TPM3, SDC4, and others. The specific fusion partner influences the biology of the rearrangement to some degree — particularly the likelihood of brain metastases — but all ROS1 fusions share the same fundamental mechanism of constitutive kinase activation and, crucially, all predict sensitivity to ROS1-targeted drugs.

ROS1 is structurally and functionally similar to ALK, and this similarity has important practical consequences: some ALK inhibitors — particularly crizotinib and lorlatinib — also inhibit ROS1, and are approved for use in ROS1-rearranged lung cancer. This structural overlap also means that the testing platforms used for ALK and ROS1 are broadly similar.

Why is the test done

To determine eligibility for ROS1-targeted therapy. Crizotinib (Xalkori), entrectinib (Rozlytrek), and lorlatinib (Lorbrena) are approved for ROS1-rearranged NSCLC. These drugs achieve high response rates and durable disease control in this population. Without testing, patients who would benefit from these drugs would be incorrectly treated with chemotherapy alone.
To avoid less effective treatment. As with other driver-positive lung cancers, patients with ROS1 rearrangements respond far better to targeted therapy than to standard chemotherapy as initial treatment. Identifying the rearrangement ensures the most effective option is selected from the start.
To guide drug selection. Multiple ROS1 inhibitors are now available with different potency profiles, central nervous system (CNS) penetration, and resistance profiles. Knowing that a ROS1 rearrangement is present — and in some settings, the specific fusion variant — guides the choice of drug and the strategy for managing progression.
To identify patients at higher risk of brain metastases. ROS1-rearranged lung cancers have a high propensity for CNS spread, both at diagnosis and over the course of treatment. This influences initial staging, monitoring, and drug selection toward agents with strong brain penetration.
To support clinical trial eligibility. The ROS1 field continues to evolve. New inhibitors and combination strategies are being investigated, and knowledge of ROS1 status opens up potential clinical trial options.

Who should be tested

Current guidelines recommend ROS1 rearrangement testing for:

All patients with advanced or metastatic non-small cell lung cancer, as part of comprehensive molecular profiling at diagnosis, regardless of histological subtype, though ROS1 rearrangements are most common in adenocarcinoma.
Never-smokers or light smokers with NSCLC, in whom ROS1 rearrangements are proportionally more prevalent.
Younger patients with NSCLC without an obvious smoking-related history are more likely to have a driver rearrangement such as ROS1 than older, heavy smokers.

In practice, ROS1 testing is performed simultaneously with all other major lung cancer biomarker tests as part of a comprehensive NGS panel at the time of diagnosis. It is not performed in isolation at most major cancer centres.

How the test is performed

ROS1 rearrangement testing can be performed using several methods, and the choice depends on the laboratory’s platform and the amount of tissue available.

Next-generation sequencing (NGS)

RNA-based next-generation sequencing (NGS) is currently the preferred method for detecting ROS1 fusions at most major cancer centres. By sequencing the messenger RNA produced by tumour cells, RNA-based NGS can directly identify the fusion transcript, confirm which partner gene is involved, and characterise the specific breakpoint. This is the most comprehensive and sensitive approach and simultaneously assesses all other clinically relevant lung cancer genes in the same run. DNA-based NGS panels can also detect ROS1 rearrangements, though RNA-based panels are generally more sensitive for fusion detection.

Fluorescence in situ hybridization (FISH)

Fluorescence in situ hybridization (FISH) was historically the most commonly used method for ROS1 testing and remains the FDA-approved companion diagnostic for crizotinib in this setting. FISH uses fluorescent probes flanking the ROS1 gene; separation of the probe signals (a “split signal”) indicates a rearrangement has occurred. FISH is highly specific but labour-intensive, cannot identify the fusion partner, and requires careful interpretation because the ROS1 locus can show complex patterns, including isolated 5′ probe loss, which requires a pathologist’s expertise to interpret correctly.

Immunohistochemistry (IHC)

Immunohistochemistry (IHC) using antibodies against the ROS1 protein (most commonly the D4D6 clone) can detect abnormal ROS1 protein expression in tumour cells. Because normal lung tissue expresses little or no ROS1, strong positive staining is suspicious for a rearrangement. IHC is fast, inexpensive, and widely available, and it performs well as a screening tool with high sensitivity. However, its specificity is lower than that of FISH or NGS — some IHC-positive cases do not have a confirmed rearrangement on molecular testing — so a positive IHC result is ideally confirmed by molecular testing before treatment is initiated. IHC is most useful as a rapid screening step, particularly when tissue is limited.

Liquid biopsy

Cell-free circulating tumour DNA in blood can be tested for ROS1 rearrangements. As with ALK fusions, structural rearrangements are more challenging to detect reliably in cell-free DNA than point mutations, so sensitivity is lower than for tissue-based methods. Liquid biopsy may be used when tissue is unavailable or insufficient, or to monitor disease during treatment and at progression. A negative liquid biopsy does not exclude a ROS1 rearrangement, and tissue testing should follow when the result is negative but clinical suspicion remains.

How results are reported

ROS1 results are reported as positive (rearrangement detected) or negative (no rearrangement detected), with notation of the testing method and, for NGS results, the specific fusion partner and variant. A typical positive NGS report might read: “ROS1-CD74 fusion detected” or “ROS1-EZR fusion, exon 34 breakpoint confirmed.” A FISH report will note the percentage of cells showing split signals and whether the result exceeds the laboratory’s positivity threshold (typically 15% or more of cells showing a split signal).

Some reports may note a fusion involving ROS1 with an uncharacterised or novel partner gene. The clinical significance should be discussed with a thoracic oncologist, as most ROS1 fusions — regardless of partner — predict sensitivity to ROS1-targeted drugs, but confirmation may be warranted.

What the result means

ROS1 rearrangement positive. A ROS1 fusion gene is present in the cancer cells. The tumour is driven at least in part by the constitutively active ROS1 fusion protein and is expected to respond to ROS1 inhibitor therapy. This is a highly actionable finding. Current approved options for advanced ROS1-rearranged NSCLC include crizotinib (Xalkori), entrectinib (Rozlytrek), and lorlatinib (Lorbrena). The choice among these depends on several factors, including CNS involvement, prior treatment history, and local drug availability and approval status. Entrectinib and lorlatinib have demonstrated strong CNS activity and are generally preferred when brain metastases are present or at high risk. Response rates to ROS1 inhibitors are among the highest seen with any targeted therapy in lung cancer — consistently above 70% in clinical trials — and progression-free survival can extend to several years in responding patients.
ROS1 rearrangement negative. No ROS1 fusion was detected in the regions assessed. ROS1 inhibitor therapy is not indicated. Other molecular findings, PD-L1 expression, and disease stage will guide treatment. If all driver mutations are absent (EGFR wild-type, ALK-negative, ROS1-negative, and so on), treatment typically centres on immunotherapy with or without chemotherapy, based on PD-L1 expression.
Equivocal or borderline result. This can occur with IHC (weak or heterogeneous staining) or with FISH (a borderline proportion of split-signal cells). Confirmatory testing with an alternative method — preferably RNA-based NGS — should be performed before treatment decisions are made. An equivocal result does not mean the cancer is ROS1-rearranged; it means the question is not yet definitively answered.

ROS1 inhibitors and resistance

The treatment of ROS1-rearranged lung cancer has evolved considerably since crizotinib was first approved in 2016, and choosing the right drug — particularly in the context of CNS disease and anticipated resistance patterns — is now an active area of clinical decision-making.

Crizotinib (Xalkori). The first approved ROS1 inhibitor has well-established efficacy in ROS1-rearranged NSCLC. However, crizotinib has limited CNS penetration, and brain metastases are a common site of progression. It is less commonly recommended as first-line therapy at centres where entrectinib or lorlatinib are available, particularly for patients with existing brain involvement.
Entrectinib (Rozlytrek). A second-generation inhibitor with meaningful CNS penetration and activity against brain metastases. It also targets NTRK fusions (relevant for a different set of lung and other cancers) and ALK. Clinical trial data showed durable responses and intracranial activity, making it a preferred first-line option for patients with brain metastases or at risk of CNS spread.
Lorlatinib (Lorbrena). Originally developed as a third-generation ALK inhibitor, lorlatinib also has potent ROS1 activity and exceptional CNS penetration. It is active against several resistance mutations that emerge after crizotinib or entrectinib, making it valuable both as a first-line option in high CNS-risk patients and as a later-line agent at progression. Lorlatinib is increasingly used as first-line therapy for ROS1-rearranged NSCLC at centres with experience in this area, based on its broad resistance coverage and brain penetration.

When a ROS1-rearranged cancer progresses on ROS1 inhibitor therapy, resistance can arise through mutations within the ROS1 kinase domain itself (on-target resistance, such as the G2032R mutation) or through activation of bypass signalling pathways. Repeat molecular testing at progression — typically with both liquid biopsy and tissue biopsy — is recommended to identify the resistance mechanism, because the choice of subsequent therapy differs depending on whether resistance is on-target or off-target.

ROS1 rearrangements and brain metastases

ROS1-rearranged lung cancers have a high rate of brain metastases — higher than most other NSCLC subtypes — both at initial diagnosis and as a site of progression during treatment. This is one of the defining clinical features of this molecular subtype. In one series, approximately 35% of patients with ROS1-rearranged NSCLC had brain metastases at presentation, and cumulative rates increased substantially over the course of the disease.

The high rate of CNS metastasis makes the choice of a ROS1 inhibitor particularly important. Agents with strong CNS penetration — entrectinib and lorlatinib — are generally favoured over crizotinib when brain involvement is present or when patients are at high risk. Brain MRI is typically included in the initial staging workup and in ongoing monitoring for all patients with ROS1-rearranged lung cancer.

ROS1 rearrangements: germline vs. somatic

ROS1 rearrangements found in lung cancer are somatic — they arise within the cancer cells during the patient’s lifetime and are not inherited. There is no known hereditary cancer syndrome associated with germline ROS1 mutations. Patients do not need to worry that their ROS1 rearrangement can be passed to their children, and family members do not require ROS1 screening on this basis.

What happens next

If a ROS1 rearrangement is found, your oncologist will recommend ROS1 inhibitor therapy. Brain imaging — typically MRI — will be performed to assess for CNS metastases before or at the start of treatment, as this influences drug selection. Your oncologist will discuss which drug is most appropriate for your specific situation. Treatment is taken as a daily oral tablet. Side effects vary by drug and will be reviewed with you before starting.
If the result is negative, ROS1 inhibitor therapy is not indicated. The full molecular profile will guide treatment, including assessment of other targetable alterations and PD-L1 expression.
If the result is equivocal, Confirmatory testing with an alternative platform will be arranged. Treatment planning will proceed once a definitive result is available.
At progression on ROS1 inhibitor therapy, repeat molecular testing with both liquid biopsy and tissue biopsy is recommended to identify the resistance mechanism. The choice of subsequent therapy depends on whether an on-target ROS1 mutation, a bypass pathway, or transformation to a different histology drives resistance.

Questions to ask your doctor

Has my tumour been tested for ROS1 rearrangements, and what method was used?
If my ROS1 test is positive, which ROS1 inhibitor is recommended for me, and why?
Do I have brain metastases, and how does that affect the choice of ROS1 inhibitor?
Will I need a brain MRI as part of my initial staging?
What side effects should I expect from ROS1 inhibitor therapy, and how are they managed?
If my cancer progresses on a ROS1 inhibitor, will re-testing be done to identify the resistance mechanism?
Are there clinical trials studying new ROS1-targeted treatments that I might qualify for?
What other biomarkers have been tested in my tumour, and were any other actionable alterations found?