Development of clinical-radiological prediction models to prioritize ALK, ROS1, and RET fusion testing in major driver alteration-negative non-small cell lung cancer

Background: The aim of this study was to explore the association between combining radiological and clinical features and anaplastic lymphoma kinase (ALK), c-ros oncogene 1 (ROS1) and rearranged during transfection (RET) fusion in non-small-cell lung cancer (NSCLC).

Methods: A retrospective analysis was carried out on 130 patients diagnosed with fusion gene-positive NSCLC. Of these patients, 85 were ALK-positive, 23 were ROS1-positive, and 22 were RET-positive. Furthermore, 318 cases that tested major driver alteration-negative controls were included. Clinical, serological, and radiological features were compared. Univariate and multivariate logistic regression analyses were performed, and models were internally validated using 1,000-bootstrap resampling.

Results: Multivariate logistic regression analysis revealed that among ALK fusion gene-positive patients, utilizing a combination of age, carbohydrate antigen 153 (CA153), pericardial effusion, and vertebral metastasis resulted in an area under the curve (AUC) of 0.803 (95% confidence interval (CI) 0.748-0.858,P <0.001). For ROS1 fusion gene-positive patients, combining smoking status, thrombotic events, D-dimer (DD2), supraclavicular lymph node metastasis, and vertebral metastasis led to an AUC of 0.894 (95% CI 0.839-0.949, P<0.001). The RET model (thrombosis and esophageal invasion) demonstrated poor discriminative performance (AUC 0.617, 95% CI 0.479–0.756), precluding reliable clinical application.

Conclusions: In major driver alteration-negative lung adenocarcinoma, particularly among younger adults, elevated DD2 and specific metastatic patterns (vertebral metastasis, pericardial effusion, supraclavicular lymph node involvement) enable risk-stratified prioritization of ALK and ROS1 fusion testing. This exploratory triage strategy may guide timely utilization of immunohistochemistry (IHC), polymerase chain reaction (PCR), or fluorescence in situ hybridization (FISH) when comprehensive (NGS) is delayed or unavailable, optimizing treatment initiation in advanced disease. However, these models require prospective external validation before clinical implementation and should not replace guideline-recommended comprehensive molecular testing. The RET findings are hypothesis-generating only and cannot support clinical decision-making.