ROS1-Rearranged Non-Small Cell Lung Cancer

1. Disease Information

Overview

ROS1-rearranged NSCLC is a molecular subtype of non-small cell lung cancer driven by chromosomal rearrangements involving the ROS1 proto-oncogene (chromosome 6q22). These rearrangements produce chimeric fusion proteins with constitutive tyrosine kinase activity, functioning as oncogenic drivers. The disease was first recognized as a distinct targetable entity following the identification of ROS1 fusions in lung cancer in 2007, and the subsequent demonstration that ROS1 TKIs could produce durable clinical responses.

"ROS1 rearrangements define a molecular subset of non-small cell lung cancer (NSCLC) by accounting for 1%-2% of cases" (PMID: 40171848). "ROS1 fusion-positive non-small cell lung cancer (NSCLC) represents a rare but clinically important subset, occurring in 1-2% of patients and often associated with younger, never-smoker populations" (PMID: 41548253).

Key Identifiers

Synonyms and Alternative Names

• ROS1-positive NSCLC / ROS1+ NSCLC
• ROS1 fusion-positive non-small cell lung cancer
• ROS1-rearranged lung adenocarcinoma
• c-ros oncogene 1 rearranged NSCLC

Information Source

This report is derived from aggregated disease-level resources including published clinical trials, molecular biology studies, population-based registries (SEER), and curated databases (COSMIC, ClinVar, UniProt, PDB), supplemented by individual patient case reports documenting resistance mechanisms and treatment sequences.

2. Etiology

Disease Causal Factors

ROS1+ NSCLC is caused by somatic chromosomal rearrangements that fuse the 3' kinase domain of the ROS1 gene to the 5' portion of various partner genes, producing constitutively active fusion kinases. These are acquired somatic events, not germline inherited mutations. The rearrangements are typically interchromosomal or intrachromosomal inversions/translocations.

The most common fusion partners include: - CD74-ROS1 (~35–44% of cases): the most frequent partner - EZR-ROS1: ezrin gene on 6q25 - SLC34A2-ROS1: solute carrier on 4p15 - SDC4-ROS1: syndecan-4 on 20q12 - TPM3-ROS1: tropomyosin-3 on 1q21 - GOPC-ROS1 (also known as FIG-ROS1): on 6q22 (intrachromosomal) - CLIP1-ROS1, KIF21A-ROS1, and other rare partners

{{figure:ros1_protein_and_partners.png|caption=ROS1 protein domain architecture, fusion breakpoints, resistance mutation sites, and landscape of fusion partners with approximate frequencies}}

Risk Factors

Genetic risk factors: - ROS1 rearrangements are somatic, not inherited; no germline susceptibility loci have been identified - Co-occurring TP53 mutations are frequent and associated with worse prognosis (PMID: 37261522) - CDKN2A/B copy number loss co-occurs in ~15% of fusion-positive patients

Environmental risk factors: - Unlike most NSCLC, ROS1+ disease is not associated with tobacco smoking—75.7% of patients are never-smokers (PMID: 29883837) - The etiology of the chromosomal rearrangement is unknown in most cases - No occupational, dietary, or environmental exposures have been definitively linked

Demographic associations: - Age: Younger patients (median 50–56 years vs. ~65 years for general NSCLC) - Sex: Female predominance (58.9% female in one cohort; fusion frequency 3.71% in women vs. 1.81% in men, p < 0.01) (PMID: 30468296) - Ethnicity: Reported across all ethnicities; Indian populations show frequencies of 3.5–4.1% (PMID: 35634796); Hispanic/Latino ~2% (PMID: 37729688)

Protective Factors

No specific genetic or environmental protective factors have been identified for ROS1+ NSCLC. General lung cancer protective factors (avoidance of tobacco, radon mitigation) apply but are less relevant given the never-smoker predominance of this subtype.

Gene-Environment Interactions

There is no established gene-environment interaction for ROS1 rearrangements. The disease appears to arise from stochastic somatic rearrangement events rather than environmental mutagen exposure.

3. Phenotypes

Clinical Presentation

Phenotype Characteristics

• Age of onset: Adult-onset, typically 5th–6th decade (median 50–56 years), significantly younger than general NSCLC population. "ROS1 fusion-positive patients were significantly younger (55.68 +/- 11.34 vs. negative 61.02 +/- 10.44 years; P < 0.01)" (PMID: 30468296)
• Progression: Progressive without treatment; variable with TKI therapy (initial response followed by eventual progression)
• CNS tropism: High rate of brain metastases (~22–36% at diagnosis), a hallmark of the disease and a key clinical challenge

Histopathological Phenotype

ROS1+ NSCLC shows characteristic histological features: "Histologically, the carcinoma was an adenocarcinoma with a predominant acinar pattern; notably, a mucinous cribriform pattern and a solid signet-ring cell pattern were also observed" (PMID: 23877438). Adenocarcinoma accounts for 98.1% of ROS1+ cases; rare pleomorphic carcinoma has been reported (PMID: 29883837).

Quality of Life

Entrectinib treatment maintains stable global health status and quality of life as measured by EORTC QLQ-C30, QLQ-LC13, and EQ-5D-3L instruments in the STARTRK-2 trial (PMID: 33930659). Post-TKI chemo-immunotherapy carries substantially higher toxicity burden with grade 3–4 adverse events in 63.9–83.8% of patients (PMID: 38830303).

4. Genetic/Molecular Information

Causal Gene

ROS1 (ROS proto-oncogene 1, receptor tyrosine kinase) - Chromosome: 6q22.1 - HGNC ID: HGNC:10261 - OMIM: 165020 - UniProt: P08922 - Protein: 2,347 amino acids; type I transmembrane receptor tyrosine kinase - Normal function: Epithelial cell differentiation, regionalization of proximal epididymal epithelium via NELL2-mediated lumicrine signaling; activates PI3K-mTOR, STAT3, VAV3 pathways (PMID: 38169386)

Pathogenic Variants

Fusion rearrangements (somatic):

All fusions retain the ROS1 kinase domain (exons 34–43) and are classified as gain-of-function (constitutive kinase activation). The fusion partner contributes the promoter (driving expression) and often a dimerization domain (facilitating ligand-independent activation).

Resistance mutations (acquired, somatic):

"ROS1 L2086F mutant kinase is resistant to type I TKI including crizotinib, entrectinib, lorlatinib, repotrectinib, taletrectinib, while the type II TKI cabozantinib and merestinib retain activity" (PMID: 38293020).

Chromosomal Abnormalities

The defining event is a chromosomal rearrangement (translocation, inversion, or deletion) producing the ROS1 fusion gene. These are detected by FISH (break-apart probes), RT-PCR, or next-generation sequencing (RNA-based panels preferred).

Modifier Genes

• TP53 co-mutations: Most frequent co-alteration (31% in fusion-positive patients); associated with worse prognosis and shorter duration of TKI response (PMID: 37261522)
• CDKN2A/B copy number loss: Co-occurs in ~15%; combined with TP53 creates immunosuppressive microenvironment
• SPP1 overexpression: Associated with poor outcomes in ALK/ROS1 fusion-positive cancers not receiving targeted therapy (PMID: 34234236)

Epigenetic Information

Limited data specific to ROS1+ NSCLC. General NSCLC epigenetic alterations (promoter methylation of tumor suppressors) likely apply. No ROS1-specific methylation biomarkers established.

5. Environmental Information

Environmental Factors

No specific environmental toxins, radiation exposures, or occupational hazards have been causally linked to ROS1 rearrangements. This contrasts sharply with general NSCLC, where tobacco smoke and radon are dominant risk factors.

Lifestyle Factors

• Smoking: ROS1+ NSCLC is enriched in never-smokers (75.7%); smoking does not appear to be a causative factor (PMID: 29883837)
• Diet, exercise, alcohol: No established associations

Infectious Agents

No infectious agents have been implicated in ROS1+ NSCLC.

6. Mechanism / Pathophysiology

Molecular Pathways

ROS1 fusion proteins activate multiple oncogenic signaling cascades:

{{figure:ros1_signaling_resistance.png|caption=ROS1 fusion signaling pathways showing convergence on STAT3 and SHP2, downstream MAPK and PI3K-AKT-mTOR effectors, and acquired resistance mechanisms}}

Primary signaling axes:

1. STAT3 pathway: "All studied fusions converge on STAT3 activation" (PMID: 41790556)
2. SHP2 (PTPN11) pathway: ROS1 fusions directly interact with and phosphorylate SHP2 "to a greater extent than ALK fusions, and analyses of downstream pathways suggest MAPK-independent, non-canonical SHP2-driven functions" (PMID: 41790556)
3. RAS-MAPK cascade: ERK1/2 activation via SOS1/GRB2 and SHP2
4. PI3K-AKT-mTOR pathway: Cell survival, growth, and metabolism
5. PLCgamma pathway: Calcium signaling and PKC activation

GO terms for biological processes: - GO:0007169 (transmembrane receptor protein tyrosine kinase signaling pathway) - GO:0070372 (regulation of ERK1 and ERK2 cascade) - GO:0032006 (regulation of TOR signaling) - GO:0030154 (cell differentiation) - GO:0046777 (protein autophosphorylation) - GO:0006468 (protein phosphorylation)

Causal Chain

Chromosomal rearrangement (somatic)
    |
    v
ROS1 fusion gene (e.g., CD74-ROS1)
    |
    v
Constitutively active ROS1 kinase (ligand-independent dimerization)
    |
    v
Phosphorylation of SHP2/STAT3/PLCgamma/IRS1
    |
    v
Activation of MAPK, PI3K-AKT-mTOR, STAT3 cascades
    |
    v
Uncontrolled proliferation, survival, migration
    |
    v
Lung adenocarcinoma (mucinous cribriform/signet ring features)
    |
    v
Metastasis (high CNS tropism)

Cellular Processes

• Proliferation: Constitutive MAPK/ERK activation drives cell cycle progression
• Anti-apoptosis: PI3K-AKT-mTOR and STAT3 signaling promote cell survival
• Migration/invasion: Fusion partner-specific effects (e.g., CLIP1-ROS1 increases cell motility via microtubule interactions) (PMID: 41790556)
• Angiogenesis: Downstream VEGF pathway activation

Cell types involved: - CL:0000066 (epithelial cell) — specifically pulmonary type II pneumocytes/Clara cells - CL:0002063 (type II pneumocyte)

Protein Dysfunction

The ROS1 fusion protein represents a gain-of-function alteration. The fusion partner provides: 1. A constitutively active promoter (driving expression) 2. A coiled-coil or dimerization domain (enabling ligand-independent activation) 3. Subcellular localization signals (e.g., GOPC-ROS1 localizes to Golgi)

Five crystal structures of the human ROS1 kinase domain are available in the PDB:

Additional structures include the full-length extracellular domain (9PVP, 4.57 A) and the ROS1-NELL2 complex (10FT, 3.21 A).

Immune System Involvement

"Patients with co-occurring TP53/CDKN2A/B variations and ALK/RET/ROS1 rearrangements are associated with high TMB, more neoantigens, an immunosuppressive microenvironment and a worse prognosis" (PMID: 37261522). Despite higher TMB and neoantigen burden, these tumors show lower CD8+ T-cell infiltration, creating a paradoxical immune phenotype. PD-L1 positivity is not significantly associated with ROS1 fusion status (PMID: 34600407).

Metabolic Changes

Pemetrexed (an antifolate) shows superior efficacy in ROS1+ NSCLC compared to non-pemetrexed chemotherapy (PFS 179 vs. 110 days, p = 0.0107), with low thymidylate synthase (TS) expression associated with better outcomes (PFS 184 vs. 110 days, p = 0.0105) (PMID: 27738334). This suggests altered folate metabolism in ROS1+ tumors.

SHP2 as a Therapeutic Vulnerability

SHP2 inhibition enhances the effects of TKIs in preclinical models of treatment-naive ALK/ROS1-fusion NSCLC (PMID: 34158345). Given that all ROS1 fusions converge on strong SHP2 activation via direct interaction (PMID: 41790556), combination of ROS1 TKIs with SHP2 inhibitors represents a mechanistically rational therapeutic strategy.

Resistance Mechanisms

On-target (kinase domain mutations): - G2032R (solvent front) — most common - L2026M (gatekeeper) - L2086F (xDFG motif) — pan-type I TKI resistant - D2033N (solvent front)

Off-target/bypass mechanisms: - MET amplification: "acquired MET amplification as a resistance driver in a ROS1-rearranged lung adenocarcinoma after sequential treatment with ROS1 inhibitors" (PMID: 37923925) - HGF-mediated MET activation: "HGF most potently induced entrectinib resistance in KM12SM and HCC78 cells by activating its receptor MET" (PMID: 36416133) - KRAS mutations - RET rearrangements (RUFY1-RET reported) (PMID: 39907599) - FGFR3 amplification

7. Anatomical Structures Affected

Organ Level

Primary organ: Lung (UBERON:0002048) - Predominantly affects the peripheral lung parenchyma - Adenocarcinoma histology typically arising in distal airways

Secondary organ involvement (metastatic sites): - Brain/CNS (UBERON:0000955) — high tropism, 22–36% at diagnosis - Bone (UBERON:0002481) - Liver (UBERON:0002107) - Adrenal glands (UBERON:0002369) - Pleura (UBERON:0000977) — pleural effusion common - Lymph nodes (UBERON:0000029)

Body systems: Respiratory system (primary), nervous system (CNS metastases), skeletal system, lymphatic system

Tissue and Cell Level

• Tissue: Pulmonary epithelium (UBERON:0000115)
• Cell types: Type II pneumocytes (CL:0002063), Clara/Club cells (CL:0000158)
• Histological pattern: Adenocarcinoma with mucinous cribriform pattern, solid signet ring cell pattern, acinar pattern

Subcellular Level

• Cell membrane (GO:0005886): ROS1 receptor localization
• Cytoplasm (GO:0005737): Signaling cascade components
• Nucleus (GO:0005634): STAT3 transcription factor translocation
• Golgi apparatus (GO:0005794): GOPC-ROS1 fusion specifically localizes here

Localization

• Typically peripheral lung nodules/masses
• No lateralization preference (bilateral involvement possible in advanced disease)
• Leptomeningeal carcinomatosis can occur as a specific pattern of CNS spread

8. Temporal Development

Onset

• Typical age: 5th–6th decade (median 50–56 years)
• Onset pattern: Insidious; often diagnosed at advanced stage (stage IIIB/IV) due to nonspecific symptoms
• 42.9% of patients present with stage IV disease in surgical series (PMID: 34234236)

Progression

Staging: Standard AJCC/TNM staging for NSCLC applies (stages I–IV)

Natural history without targeted therapy: - Rapid progression; platinum-based chemotherapy yields median PFS of ~6–7 months - Pemetrexed-based chemotherapy: ORR 40.8%, median PFS 179 days (PMID: 27738334)

With targeted therapy: - First-line crizotinib: median PFS ~12–19 months - First-line repotrectinib: median PFS 35.7 months (PMID: 39402859) - Eventual resistance is near-universal; sequential TKI strategies can extend disease control

Disease course: Progressive; no spontaneous remission. Treatment-induced responses are common but resistance develops.

Critical Periods

• Diagnosis: Molecular testing at diagnosis is essential for treatment selection
• CNS monitoring: Regular brain imaging due to high CNS metastasis risk
• Resistance emergence: Molecular profiling at progression guides sequential therapy

{{figure:ros1_timeline.png|caption=Timeline of key milestones in ROS1-rearranged NSCLC research and treatment, from initial discovery through next-generation TKI approvals}}

9. Inheritance and Population

Epidemiology

Inheritance Pattern

• Not inherited: ROS1 rearrangements are somatic events with no Mendelian inheritance pattern
• No germline ROS1 mutations predisposing to cancer have been established
• Not applicable: penetrance, expressivity, anticipation, carrier frequency

Population Demographics

• Sex ratio: Female predominance (~59% female; fusion frequency 3.71% in women vs. 1.81% in men)
• Age distribution: Younger than general NSCLC (median ~50–56 years); enriched in patients <40 years (ROS1 fusions found in 7% of young-onset NSCLC in India) (PMID: 40122770)
• Ethnic distribution: Present across all populations; potentially higher frequency in East Asian and South Asian populations
• Smoking status: 75% or more are never-smokers

10. Diagnostics

Molecular Diagnostic Testing

ROS1 rearrangement detection is mandatory before initiating TKI therapy. Multiple methods are available:

Recommended approach: IHC screening followed by FISH/NGS confirmation, or upfront comprehensive NGS (RNA-based preferred). "Targeted RNA NGS was confirmed to be the most efficient technique for gene fusion identification in clinical practice" (PMID: 37190044). "FISH should not be dismissed, as they can crucially contribute to the completion of the molecular characterization" (PMID: 37190044).

Imaging Studies

• CT chest/abdomen: Standard staging
• Brain MRI: Mandatory at diagnosis and surveillance (high CNS tropism)
• PET-CT: For comprehensive staging
• Bone scan: If bone metastases suspected

Biopsy and Pathology

• Tissue biopsy preferred; adequate material for molecular testing essential
• IHC: ROS1 protein expression (D4D6 clone, SP384 clone)
• Characteristic: adenocarcinoma with mucinous cribriform pattern, signet ring cells
• ROSE (Rapid On-Site Evaluation) improves specimen adequacy for biomarker testing (PMID: 37805343)

Liquid Biopsy

• ctDNA-based NGS panels can detect ROS1 fusions in plasma
• CSF ctDNA superior to plasma for CNS disease monitoring (PMID: 32838487)
• Useful for resistance mutation detection at progression

Differential Diagnosis

• Other oncogene-driven NSCLC (ALK, RET, NTRK fusions — mutually exclusive with ROS1)
• EGFR-mutant NSCLC (also younger, never-smokers, adenocarcinoma)
• KRAS-mutant NSCLC
• Driver-negative NSCLC

11. Outcome/Prognosis

Survival and Mortality

{{figure:ros1_treatment_landscape.png|caption=Comparison of ROS1 TKI efficacy across generations showing objective response rates (ORR) and median progression-free survival (PFS)}}

"In the TKI-naive cohort (n = 71), 79% of patients achieved an objective response, with a median progression-free survival (PFS) of 35.7 months, surpassing all previously approved ROS1 TKIs" (PMID: 39402859).

"Taletrectinib demonstrated high objective response rates in both TKI-naive (88.8%) and TKI-pretreated (55.8%) patients, including robust intracranial activity and efficacy against the G2032R mutation" (PMID: 41548253).

Prognostic Factors

Favorable: - ROS1 fusion-positive status (vs. driver-negative NSCLC) - Absence of TP53 co-mutation - Absence of CNS metastases at diagnosis - Access to appropriate TKI therapy

Unfavorable: - TP53/CDKN2A/B co-mutations (immunosuppressive microenvironment, worse prognosis) - CNS metastases at diagnosis - G2032R resistance mutation emergence - Lack of access to molecular testing or targeted therapy

Complications

• CNS metastases / leptomeningeal carcinomatosis
• Pleural effusion / respiratory failure
• Bone metastases with pathological fractures
• Hepatic metastases
• Treatment-related toxicity (see Treatment section)

12. Treatment

First-Line Targeted Therapy

MAXO:0000058 (pharmacotherapy)

Second-Line and Beyond

• Lorlatinib: Third-gen ALK/ROS1 TKI; good CNS penetration; used post-crizotinib; limited by neurocognitive side effects (PMID: 38201357)
• Zidesamtinib (NVL-520): ROS1-selective, TRK-sparing; ORR 73% post-crizotinib, 38% post-repotrectinib (PMID: 40118657); crystal structure of G2032R-zidesamtinib complex available (PDB: 9QEK)
• Cabozantinib: Type II multi-kinase inhibitor; retains activity against L2086F mutation (PMID: 38293020, PMID: 40826797)
• Brigatinib: ORR 71.4% in TKI-naive, 31.6% post-crizotinib (PMID: 39018589)

Chemotherapy

MAXO:0000647 (chemotherapy)

Pemetrexed-based platinum doublet is the preferred chemotherapy regimen: "crizotinib-treated group had a higher overall response rate (ORR, 80.0%), disease control rate (DCR, 90.0%) and longer progression-free survival (PFS, 294 days) compared with the rates in pemetrexed-treated group (ORR, 40.8%; DCR, 71.4%; PFS, 179 days) and non-pemetrexed-treated group (ORR, 25.0%; DCR, 47.7%; PFS, 110 days)" (PMID: 27738334).

Immunotherapy

MAXO:0001298 (immunotherapy)

Immune checkpoint inhibitors have limited single-agent activity in ROS1+ NSCLC: - PD-L1 expression is not significantly associated with ROS1 fusion status (PMID: 34600407) - TP53/CDKN2A/B co-mutation creates an immunosuppressive microenvironment despite high TMB (PMID: 37261522) - NCCN guidelines recommend against first-line immunotherapy in ROS1+ NSCLC - May have a role in combination with chemotherapy after TKI exhaustion

Surgical Interventions

MAXO:0000004 (surgical procedure)

• Surgical resection is standard for early-stage disease (stage I–IIIA)
• Limited data on adjuvant targeted therapy post-resection in ROS1+ patients
• ROS1 appears less frequent in early-stage resected disease (0.6% vs 1–2% in advanced) (PMID: 37237384)

Emerging Combination Strategies

• ROS1 TKI + SHP2 inhibitor: Preclinical data supports enhanced efficacy (PMID: 34158345); mechanistic rationale from convergent SHP2 activation across all fusions (PMID: 41790556)
• ROS1 TKI + MET inhibitor: For MET amplification-driven resistance; lorlatinib + capmatinib combination reported (PMID: 37923925)
• APG-2449: FAK + ALK/ROS1 TKI; CNS-penetrant (CSF:plasma ratio 0.65–1.66); phase I data available (PMID: 41146927)

Treatment Algorithm

Diagnosis of advanced ROS1+ NSCLC
    |
    v
1st Line: Next-gen ROS1 TKI (repotrectinib or taletrectinib preferred)
         Consider CNS status when selecting agent
    | (progression)
    v
Rebiopsy / liquid biopsy for resistance mechanism
    |
    v
On-target resistance (e.g., G2032R):
  -> Switch to TKI with activity against specific mutation
  -> G2032R: taletrectinib, zidesamtinib
  -> L2086F: cabozantinib/merestinib (type II TKI)
    |
    v
Off-target/bypass resistance (e.g., MET amp):
  -> Combination therapy (ROS1 TKI + pathway-specific inhibitor)
  -> Consider clinical trials
    |
    v
Post-TKI exhaustion:
  -> Pemetrexed-based chemotherapy +/- immunotherapy
  -> Clinical trials (SHP2 inhibitor combos, novel agents)

13. Prevention

Primary Prevention

• Smoking cessation: While ROS1+ NSCLC is not smoking-related, general lung cancer prevention includes tobacco avoidance
• No specific primary prevention strategies exist for ROS1 rearrangements (stochastic somatic events)

Secondary Prevention (Screening and Early Detection)

• Low-dose CT screening: USPSTF-recommended for high-risk individuals (age 50–80 with 20+ pack-year history); less applicable to the ROS1+ demographic (younger never-smokers)
• Molecular testing at diagnosis: Universal molecular profiling of all advanced non-squamous NSCLC ensures detection of ROS1 fusions
• Guidelines mandate testing for EGFR, ALK, ROS1, BRAF, RET, MET, NTRK, KRAS in advanced NSCLC (PMID: 37455124)

Tertiary Prevention

• Regular surveillance imaging (CT, brain MRI) during and after TKI therapy
• Molecular profiling at disease progression to identify resistance mechanisms and guide sequential therapy
• Supportive care and symptom management

Genetic Counseling

Not typically indicated as ROS1 rearrangements are somatic. No hereditary cancer syndromes are associated with ROS1 fusions (unlike the rare pediatric Li-Fraumeni-associated angiosarcoma with ROS1 rearrangement) (PMID: 36307212).

14. Other Species / Natural Disease

Comparative Biology

ROS1 orthologs:

Animal Models — Normal ROS1 Function

c-ros knockout mice demonstrate the normal physiological role of ROS1: "Transgenic mice with male infertility, the c-ros knockout (KO) and GPX5-Tag2 transgenic mouse models... exhibit severely angulated sperm flagella explaining the infertility" (PMID: 15109745). This results from defective differentiation of the initial segment epithelium of the epididymis.

The NELL2-ROS1 lumicrine signaling axis is conserved in humans: "There was a significant correlation between the epididymal expressions of mouse genes upregulated by the trans-luminal signaling and those of their human orthologs, as evaluated by the correlation coefficient of 0.604" (PMID: 38169386).

Naturally Occurring ROS1 Fusions in Animals

No naturally occurring ROS1 fusion-driven lung cancers have been reported in companion animals or wildlife. Canine and feline lung tumors are rare and not characterized for ROS1 status.

15. Model Organisms

In Vivo Models

Cell Line Models

HCC78 cells are the most widely used model for studying ROS1 TKI efficacy and resistance. HGF-mediated resistance was characterized using both HCC78 and KM12SM cells (PMID: 36416133).

Model Characteristics

• Ba/F3 system: Excellent for rapid assessment of fusion oncogenicity and drug sensitivity; used to characterize L2086F pan-type I TKI resistance (PMID: 38293020)
• PDX models: Best recapitulation of human tumor biology; used for preclinical TKI evaluation
• Limitations: No single model fully captures the human tumor microenvironment, immune interactions, or CNS tropism

Key Findings Summary

Finding 1: ROS1 Rearrangements Define a Rare Molecular Subset

ROS1 fusions occur in 1–2% of NSCLC cases globally, with variation by population (2.59% in China, 3.5–4.1% in India, ~2% in Hispanic/Latino populations). The disease is enriched in younger, female, never-smoking patients with adenocarcinoma.

Finding 2: Distinctive Clinical-Demographic Profile

Median age 50–56 years, 58.9% female, 75.7% never-smokers, 98.1% adenocarcinoma. Statistically significant differences versus ROS1-negative NSCLC in age (p < 0.01), sex (p < 0.01), and smoking status (p < 0.01).

Finding 3: Rapidly Evolving TKI Landscape

First-generation TKIs (crizotinib, entrectinib) achieve ORR 60–80% with PFS 12–19 months. Next-generation agents have dramatically improved outcomes: repotrectinib (ORR 79%, PFS 35.7 months), taletrectinib (ORR 89%), unecritinib (ORR 80.2%).

Finding 4: Complex Resistance Mechanisms

On-target mutations (G2032R, L2086F, L2026M) and off-target bypass pathways (MET amplification, HGF-mediated resistance, KRAS mutations, RET rearrangements) necessitate serial molecular profiling and rational TKI sequencing.

Finding 5: Convergent STAT3/SHP2 Signaling

All ROS1 fusions converge on STAT3 activation and directly interact with SHP2, with significantly greater SHP2 phosphorylation than ALK fusions. This distinguishes ROS1 from ALK biology and identifies SHP2 as a therapeutic vulnerability.

Finding 6: Normal ROS1 Biology — NELL2-Mediated Lumicrine Signaling

ROS1 normally functions in NELL2-mediated epithelial differentiation; c-ros KO mice are infertile. This pathway is conserved in humans (correlation r = 0.604 between mouse and human ortholog expression).

Finding 7: Immunosuppressive Co-Mutation Microenvironment

TP53/CDKN2A/B co-mutations (present in ~31% and ~15% respectively) create a paradoxical phenotype: high TMB and neoantigens but low CD8+ T-cell infiltration, resulting in worse prognosis.

Finding 8: Pemetrexed Biomarker Association

Pemetrexed-based chemotherapy shows superior efficacy in ROS1+ NSCLC (PFS 179 vs. 110 days for non-pemetrexed, p = 0.0107), with low thymidylate synthase expression predicting benefit.

Finding 9: Structural Biology Resources

Five crystal structures of the ROS1 kinase domain are available, including the G2032R mutant-zidesamtinib complex (PDB: 9QEK), enabling structure-based drug design for next-generation inhibitors.

Finding 10: Characteristic Histopathology

ROS1+ adenocarcinoma shows distinctive mucinous cribriform and signet ring cell patterns on histology, which can serve as morphological clues for molecular testing.

Finding 11: Conserved Normal Biology

The NELL2-ROS1 lumicrine signaling axis essential for male fertility is conserved from rodents to humans, providing insights into normal receptor function and potential off-target effects of ROS1 inhibitors.

Finding 12: Quality of Life with Targeted Therapy

Entrectinib maintains stable quality of life (STARTRK-2 PROs), while post-TKI chemo-immunotherapy carries substantially higher toxicity, underscoring the importance of maintaining patients on targeted therapy as long as possible.

Finding 13: SHP2 Inhibition as Combination Strategy

Preclinical data demonstrates that SHP2 inhibition enhances ROS1 TKI efficacy, supported by the finding that all ROS1 fusions uniquely converge on SHP2 activation, providing a mechanistic rationale for clinical combination trials.

Mechanistic Model / Interpretation

The pathophysiology of ROS1+ NSCLC can be understood as a multi-level cascade:

Level 1 — Genomic initiation: A stochastic somatic chromosomal rearrangement fuses the ROS1 kinase domain to a partner gene. The partner provides constitutive expression and a dimerization interface, converting ROS1 from a ligand-dependent receptor (normally activated by NELL2 in epididymal epithelium) to a ligand-independent oncogene.

Level 2 — Signaling amplification: The constitutively active ROS1 fusion kinase phosphorylates SHP2 (more strongly than ALK fusions do) and activates STAT3. These two nodes then fan out to the canonical MAPK and PI3K-AKT-mTOR cascades, plus non-canonical SHP2-dependent pathways that are MAPK-independent.

Level 3 — Phenotypic consequences: Sustained signaling drives proliferation (MAPK/ERK), survival (AKT/mTOR), and migration (fusion partner-specific, e.g., CLIP1 microtubule effects). The result is adenocarcinoma with mucinous/signet ring features and a notable tropism for the CNS.

Level 4 — Therapeutic response and resistance: ROS1 TKIs block the fusion kinase, collapsing the signaling cascade and producing dramatic tumor regression (ORR 70–90%). However, tumor evolution under selective pressure leads to resistance through on-target mutations (G2032R, L2086F) or bypass pathways (MET, RET, KRAS). The type of resistance mutation dictates the next therapeutic strategy: type II TKIs for L2086F, taletrectinib/zidesamtinib for G2032R, MET inhibitor combinations for MET amplification.

Level 5 — Microenvironment modulation: Co-occurring TP53/CDKN2A/B mutations reshape the immune microenvironment, creating an immunosuppressive milieu despite increased neoantigen load. This partially explains the limited efficacy of checkpoint inhibitors and suggests that combination strategies (TKI + SHP2 inhibitor or TKI + immunotherapy) may be needed for durable disease control.

Evidence Base

Landmark Studies

Limitations and Knowledge Gaps

1. Rarity limits clinical trial power: With only 1–2% prevalence, most clinical trials are single-arm; head-to-head comparisons between ROS1 TKIs are lacking and rely on indirect comparisons (MAICs)
2. Optimal TKI sequencing undefined: No randomized data guide the choice between first-line repotrectinib vs. taletrectinib, or optimal sequencing at resistance
3. Immunotherapy role unclear: While immune checkpoint inhibitors are largely ineffective as monotherapy, their role in combination with TKIs or chemotherapy post-TKI exhaustion requires further study
4. Early-stage targeted therapy: No data exist for adjuvant/neoadjuvant ROS1 TKIs in resectable disease
5. Long-term survival data: 5-year and 10-year overall survival with modern TKIs are not yet mature
6. Fusion partner-specific biology: While fusion partners confer different subcellular localizations and additional functions (e.g., CLIP1 motility effects), clinical significance remains uncertain
7. Biomarker-guided immunotherapy: Whether specific co-mutation profiles (high TMB, PD-L1+) predict immunotherapy benefit in the subset of ROS1+ patients remains unknown
8. Liquid biopsy standardization: ctDNA-based monitoring for ROS1 fusions and resistance mutations needs prospective validation
9. Epigenetic landscape: ROS1-specific epigenetic alterations are poorly characterized
10. Pediatric/young adult: Limited data on ROS1+ NSCLC in patients under 30 years

Proposed Follow-up Experiments/Actions

Clinical Priorities

1. Randomized trials comparing next-gen ROS1 TKIs: Head-to-head comparison of repotrectinib vs. taletrectinib in treatment-naive patients to establish optimal first-line therapy
2. Adjuvant ROS1 TKI trials: Phase III studies of post-surgical TKI therapy in resected ROS1+ NSCLC (analogous to ADAURA for EGFR)
3. SHP2 inhibitor combination trials: Based on preclinical evidence of enhanced TKI efficacy with SHP2 inhibition, phase I/II trials combining ROS1 TKIs with SHP2 inhibitors (e.g., TNO155, RMC-4630) are warranted
4. Biomarker-stratified immunotherapy: Trials evaluating checkpoint inhibitors in the TP53-co-mutated, high-TMB subset of ROS1+ patients

Research Priorities

1. Single-cell and spatial transcriptomics: Characterize the tumor microenvironment of ROS1+ NSCLC to understand immune evasion mechanisms and identify combination therapy targets
2. Resistance monitoring with ctDNA: Prospective validation of liquid biopsy-guided adaptive therapy (switching TKIs at molecular progression before radiographic progression)
3. Fusion partner functional studies: Systematic comparison of signaling, metastatic potential, and drug sensitivity across different fusion partners using isogenic models
4. CNS tropism mechanisms: Investigate why ROS1+ NSCLC has high brain metastasis rates; identify targetable pathways driving CNS invasion
5. Type II TKI development: Expand the arsenal of type II ROS1 TKIs to address L2086F and compound mutations
6. Real-world outcomes registry: Multi-national ROS1+ NSCLC registry to capture treatment patterns, sequential therapy outcomes, and long-term survival with modern TKIs

Ontology Term Summary

Report generated from autonomous scientific discovery analysis of 93 published studies with 13 confirmed findings across 5 investigation iterations. Last updated: 2026-05-06.