Real-world evidence of RET fusion prevalence and testing in patients with metastatic non-small cell lung cancer Authors: BN Cajiao Garcia, VD de Jager, CCHJ Epskamp-Kuijpers, AJ van der Wekken, SM Willems, LC van Kempen, E Schuuring Published: 2026 ABSTRACT Highlights•RET fusion prevalence in NSCLC may be overestimated at 1–2%.•RNA-based NGS is more accurate than FISH for detection of RET fusions.•The testing rate for RET fusions is expected to further increase in the Netherlands.•The prevalence of RET fusions is expected to be lower than 1%.Keywords: Non-small cell lung cancer, RET fusion, Predictive biomarker testing, Selpercatinib, Pralsetinib Introduction Lung cancer causes most cancer-related deaths worldwide, with 1.8 million deaths reported in 2020 [1]. The most common type of lung cancer, non-small cell lung cancer (NSCLC), accounts for 85% of all cases [2]. Approximately 70% of NSCLC can be characterized as non-squamous. Often diagnosed at an advanced stage of disease, the five-year overall survival rate of metastatic (stage IV) non-squamous NSCLC is 10–20% in the Netherlands [1,3,4]. Non-squamous NSCLC may harbor specific molecular driver aberrations for which effective targeted therapies are available [5,6]. These targetable molecular aberrations include mutations in epidermal growth factor receptor (EGFR), Kristen rat sarcoma viral oncogene homolog (KRAS) G12C, v-Raf murine sarcoma viral oncogene homolog B1 (BRAF) V600, ErbB-2 receptor tyrosine kinase 2 (ERBB2), and mesenchymal-epithelial transition (MET) exon 14 skipping mutations and fusion of anaplastic lymphoma kinase (ALK), ROS proto-oncogene 1 (ROS1), rearranged during transfection (RET), and neurotrophic receptor tyrosine kinase 1/2/3 (NTRK1/2/3) [[7], [8], [9]]. RET is a proto-oncogene that encodes for a receptor tyrosine kinase involved in the activation of Mitogen-Activated Protein Kinase (MAPK), Phosphoinositide 3-kinase/AKT Serine/Threonine Kinase 1 (PI3K/AKT) and Janus kinase/signal transducer and activator of transcription (JAK-STAT) pathways [10]. In NSCLC, RET fusions have been described with more than forty-five different fusion partners, though most RET fusions constitute a fusion between RET and Kinesin family member 5B (KIF5B) [[11], [12], [13]]. The chimeric fusion protein results in ligand-independent, expression and constitutive activation of RET signaling driving cell proliferation, growth and survival [14]. RET fusions were first identified as oncogenic drivers in non-squamous NSCLC in 2012 with an estimated prevalence of 1% to 2% [[15], [16], [17], [18], [19], [20]]. Patients with NSCLC harboring a RET fusion are characterized by relatively young age (early 60’s), non-smoking history and high rates of brain metastases at diagnosis [18,21,22]. Initially, fluorescent in situ hybridization (FISH) was the commonly used method to detect genomic rearrangement of RET [20,37,38]. While functional RET fusions are relevant for treatment, not all RET rearrangements detected by FISH result in expression of the fusion product and downstream activation of RET-related cascades [23]. As a result, RNA-based RET fusion detection assays are currently recommended [20,38]. Based on the results of the LIBRETTO-001 and ARROW trials, selpercatinib and pralsetinib were approved by the European Medicine Agency (EMA) in 2021 and by the Food and Drug Administration in 2020 [24,25] for their use in patients with RET fusion-positive metastatic NSCLC [[24], [25], [26], [27], [28]]. As a result, the European Society for Medical Oncology (ESMO) and National Comprehensive Cancer Network (NCCN) guidelines now include recommendations for RET fusion testing in patients with NSCLC [29,30]. From 2015 to 2020, the Dutch NSCLC treatment guidelines advised RET fusion testing to be tested sequentially, if EGFR, KRAS, and ALK testing results were negative. In addition, a team of Dutch experts strongly recommended molecular testing for RET fusions in stage IV NSCLC to ensure inclusion of patients in clinical trials [31]. As of 2020, the Dutch NSCLC treatment guidelines include routine RET testing for all patients with non-squamous NSCLC [32]. Despite (inter)national guidelines recommendations for RET fusion testing, the overall testing rate for RET fusions in the period preceding the inclusions in the national guidelines was low, with 20% in Q4 of 2017 in the Netherlands [33] and 27% between 2015 and 2019 in Germany [34]. The aim of the current study was to determine RET fusion testing rates and results in patients with metastatic non-squamous NSCLC in the Netherlands in 2017 and 2019 using real-world data from national databases. Methods Patient selection Data were collected from two databases with nationwide coverage: the Netherlands Cancer Registry (NCR) (managed by Netherlands Comprehensive Cancer Organisation (IKNL)) and the Dutch Nationwide Pathology Databank (Palga). The NCR, a nationwide registry of patients diagnosed with cancer in the Netherlands, was queried to select all adult patients diagnosed with metastatic non-squamous NSCLC in 2017 or 2019. Clinical data from the NCR were then linked to the respective pathology reports of the Palga databank by a trusted third party (ZorgTTP, Houten, the Netherlands). Data on (molecular) testing and testing results were manually extracted from the pathology reports. Patients were excluded if they were diagnosed outside the study period, had an autopsy report only or were diagnosed with squamous NSCLC. This study was approved by the Palga privacy and scientific committee (application number: LZV2018-​199, LZV2021_101) and the ethical committee of the NCR-governing body (application numbers: K18.311, K18311_3). Data were processed in accordance with the General Data Protection and Regulation (GDPR). Data extraction from pathology reports Palga reports were analyzed manually to collect data on the following variables: type of tumor specimen that was tested (histological, cytological or both), histopathological diagnosis and RET testing data, including whether RET testing was performed (≤3 months after diagnosis), testing technique, result of the test and timing of RET testing with respect to EGFR testing (in parallel or sequential). Testing technique was subdivided into DNA fluorescent in situ hybridization (FISH), RNA-based testing or both. Statistical analysis RET fusion testing rates were divided into three subgroups: (1) all patients tested for RET, (2) patients that tested wildtype (wt) for RET and wt for other tested biomarkers (EGFR, KRAS, BRAF V600, ERBB2, ALK, ROS1) and (3) patients tested for RET who also carry a driver gene mutation (EGFR, KRAS, BRAF V600, ERBB2, ALK, ROS1). The frequencies of the testing methods (FISH or RNA-based NGS) that were used were calculated for each study year. Prevalence of RET fusions was calculated for the entire Dutch NSCLC population by extrapolating observed prevalence of RET fusions in tested patients to those that did not carry a driver gene mutation and were not tested for RET fusions. This was performed according to a cross-tabulation and formula adapted from measure of association [35] (supplement I). Difference between the RET tested population and stage IV NSCLC Dutch patients was tested using Chi-square statistical test [36]. Interlaboratory variations were calculated for individual laboratories with a 99% confidence interval. Statistical analysis was performed using IBM SPSS Statistics for Windows, version 28.0.1.0 (SPSS Inc., Chicago, IL, USA) and figures were made using Microsoft Excel (version 2403). Results Patient characteristics A total of 3746 (for 2017) and 4048 patients (for 2019) with metastatic non-squamous NSCLC were retrieved from the National Cancer Registry (NCR) (supplement II). After the exclusion criteria were applied, 3651 and respectively 3934 patients were included. Patient characteristics of the entire cohort, as well as RET fusion-tested patients are presented in Table 1. Significant differences (p < 0.05) were observed between all patients diagnosed with stage IV NSCLC and patients tested for the characteristics: sex, type of laboratory and tissue type. Only diagnosis did not differ significantly. An increase in the population tested for RET from 22.6% to 31.8% was observed from in 2017 to 2019. Table 1Characteristics of patients diagnosed with metastatic non-squamous NSCLC in 2017 and 2019.Table 1 dummy alt text20172019All patients diagnosed with stage IV NSCLC (n = 3651)Patients tested for RET (n = 825)All patients diagnosed with stage IV NSCLC (n = 3934)Patients tested for RET (n = 1251)Mean age in years (±SD)68.0 (±10.2)67.0 (±11.0)68.0 (±10.3)67.0 (±10.4)Sex Male53.4% (n = 1951)59.0% (n = 487)52.7% (n = 2075)57.5% (n = 719) Female46.6% (n = 1700)41.0% (n = 338)47.3% (n = 1859)42.5% (n = 532)Diagnosis Adenocarcinoma86.7% (n = 3166)86.2% (n = 711)87.1% (n = 3426)87.0% (n = 1088) NSCLC – NOS12.7% (n = 464)13.3% (n = 110)12.0% (n = 473)12.2% (n = 153) Adenosquamous carcinoma0.5% (n = 18)0.4% (n = 3)0.6% (n = 23)0.6% (n = 7) NSCLC-neuroendocrine0.1% (n = 3)0.1% (n = 1)0.3% (n = 12)0.2% (n = 3)Type of laboratory Academic10.3% (n = 376)16.5% (n = 136)8.4% (n = 339)12.3% (n = 154) Non-academic89.7% (n = 3275)83.5% (n = 689)91.4% (n = 3595)87.7% (n = 1097)Tissue type Histology (+/- cytology)67.4% (n = 2459)76.0% (n = 627)67.1% (n = 2639)71.9% (n = 899) Cytology32.6% (n = 1192)24.0% (n = 198)32.9% (n = 1295)28.1% (n = 352)SD, standard deviation; NSCLC – NOS, non-small cell lung cancer – not otherwise specified.2017. sex: p < 0.05, χ2 (1, 4476) = 8.5, diagnosis: p > 0.05, χ2 (3, 4476) = 1.0, type of laboratory: p < 0.05, χ2 (1, 4476) = 25.4, tissue type: p < 0.05, χ2 (1, 4476) = 23.5. 2019. sex: p < 0.05, χ2 (1, 5185) = 8.5, diagnosis: p > 0.05, χ2 (3, 5185) = 1.0, type of laboratory: p < 0.05, χ2 (1, 5185) = 15.0, tissue type: p < 0.05, χ2 (1, 5185) = 10.0. Characteristics of patients diagnosed with metastatic non-squamous NSCLC in 2017 and 2019. SD, standard deviation; NSCLC – NOS, non-small cell lung cancer – not otherwise specified. 2017. sex: p < 0.05, χ2 (1, 4476) = 8.5, diagnosis: p > 0.05, χ2 (3, 4476) = 1.0, type of laboratory: p < 0.05, χ2 (1, 4476) = 25.4, tissue type: p < 0.05, χ2 (1, 4476) = 23.5. 2019. sex: p < 0.05, χ2 (1, 5185) = 8.5, diagnosis: p > 0.05, χ2 (3, 5185) = 1.0, type of laboratory: p < 0.05, χ2 (1, 5185) = 15.0, tissue type: p < 0.05, χ2 (1, 5185) = 10.0. RET fusion testing rates The overall testing rates for RET fusions were 23% (825/3651) in 2017 and 32% (1251/3934) in 2019. Testing rates in patients without a reported activating mutation in EGFR, KRAS, BRAF, ERBB2, ALK or ROS1 were 73.9% in 2017 and 78.2% in 2019. In contrast, only 8.8% of patients diagnosed in 2017 with a mutation in one of the other predictive markers were tested for RET fusions, and this increased to 16.8% in 2019 (supplement II). Order of RET fusion testing In 2017, RET was tested in a sequential manner for 73.1% of cases (i.e., RET fusion testing was performed only if EGFR, KRAS, and ALK testing results were negative) (Fig. 1). In 2019, sequential testing of RET decreased to 52.6%, while parallel testing increased to 27.6%. For the remaining cases, testing order was unknown (2.9% in 2017 and 19.8% in 2019). Order of RET fusion testing in patients with metastatic non-squamous NSCLC in 2017 and 2019. Sequential, tumor specimens tested for other predictive biomarkers (at least EGFR) prior to RET fusion testing; parallel, RET and other predictive markers tested in simultaneously; not reported, samples tested for RET fusion but test status or date of test request of other predictive markers is unknown. RET fusion testing assays In 2017, RET FISH was performed in 91.3% of RET-tested patients, while only 6.4% of patients were tested using an RNA-based test. In 2.1% of the patients, test type was either not reported or not retrievable from the available reports. In 2019, the use of FISH decreased to 47.0%, whilst RNA-based testing increased to 46.0%. In 4.6% of the patients, the assay was not reported or not retrievable from the available reports. The remaining 2.7% of patients were tested with both FISH and an RNA-based assay (Fig. 2). In 2017, three patients with RET fusion-positive FISH test results had a co-occurring (driver) mutation (ALK fusion, EGFR exon 19 deletion, and PIK3CA mutation) (Supplement IIIA and IIIB). Techniques used for RET fusion testing in patients with metastatic non-squamous NSCLC in 2017 and 2019. Technique not reported comprises cases for which test technique was not reported in the pathology report. FISH, fluorescent in situ hybridization. Prevalence of RET fusions The prevalence of RET fusions in the RET-tested population was 2.3% in 2017 (19/825) and 1.2% in 2019 (15/1251). Estimated prevalences of RET fusions in the entire patient cohorts were calculated by extrapolating for the patients who did not carry a mutation in another predictive biomarker and who were not tested for RET fusions. The calculated prevalences of RET fusions were 1.3% and 0.7% in 2017 and 2019, respectively (Supplement I). Interlaboratory variation for RET fusion testing Testing rates of individual laboratories were evaluated for the 2017 and 2019 cohorts. Thirteen laboratories scored outside the 99% confidence interval (CI), of which seven below the lower limit of the CI (Figs. 3A and B). In 2019, twelve laboratories were outside the 99% CI, of whom six were below the lower limit of the CI. Funnel plots demonstrating interlaboratory variation in RET testing with 99% confidence intervals in (A) 2017 and (B) 2019. CI, Confidence interval; NTR, National testing rate. Discussion Nationwide real-world data from the NCR and the Palga were used to establish molecular predictive testing rates, assess test techniques and determine the prevalences of RET fusions in patients with metastatic (stage IV) non-squamous NSCLC diagnosed in the Netherlands in 2017 and 2019. The prevalence of RET fusions calculated from these data was 1.3% in 2017 and 0.7% in 2019. This decrease was associated with the reduced use of FISH testing, which has been described to yield higher false positive rates compared to RNA-based fusion testing [23]. Our analysis demonstrates that the previously reported prevalence of RET fusions in NSCLC (1–2%) [[15], [16], [17], [18], [19], [20]] is likely overestimated. In both 2017 and 2019, FISH was commonly used for the detection of RET fusions (91.3% and 47.0%, respectively). However, back in 2019, it was not yet known that FISH was unspecific for detection of clinically relevant RET fusions [23], thereby necessitating orthogonal testing of positive cases [23,37]. The assessment of FISH patterns associated with true rearrangements in tissue sections is challenging and requires specific expertise [38]. In addition, FISH assays do not provide information regarding gene fusion partners nor on whether these genomic RET rearrangements are actually expressed [[39], [40], [41]]. As a result, the use of RET FISH may lead to reporting of non-functional RET fusions [23]. Reporting of false-positive RET fusion testing results inevitably leads to an overestimation of RET fusion prevalence, which is reflected by the findings of our study. In 2017, more RET fusion-positive cases were detected compared to 2019 (19 compared to 15 cases), while RET fusion testing rates were higher in 2019. RNA-based NGS also enables simultaneous analysis of other relevant markers (e.g., ALK, ROS1, MET-skipping and NTRK1/2/3), reducing turnaround time and using tissue more efficiently than sequential application of single gene tests [42,43]. The ESMO 2021 guideline [20] and the Dutch 2023 cieBOD-guideline [44] recommend RNA-based NGS-based testing for the detection of RET fusions for the reason described above. Previous studies have described RET fusion prevalences at 1–2% [[15], [16], [17], [18], [19], [20],45] in patients with NSCLC. This prevalence is usually reported from the observed RET fusion positive patients within the total cohort not taking into account case selection that might affect the prevalence [33,45]. In some studies, it was estimated that in patients with no mutations or fusions in other predictive markers, RET fusion prevalence was 5% [46,47]. However, the latter studies concerned Asian populations, which have been described with different prevalences for other driver mutations (e.g., EGFR) compared to European populations, and thus direct comparison is not appropriate. In our study, the prevalences of RET fusions in the entire patient cohorts were estimated by correcting for case selection. RET fusion prevalence in the RET-tested population was extrapolated to patients that were not tested for RET fusions and had no other reported driver mutations (supplement I). These calculations resulted in estimated population-wide prevalences of 1.3% for the 2017 cohort (vs. 2.3% in the RET-tested population) and 0.7% for the 2019 cohort (vs. 1.2% in the RET-tested population). The 2015 Dutch NSCLC guidelines advised RET fusion testing in patients with advanced-stage NSCLC, if no mutations were detected in EGFR, KRAS and ALK [32]. At the time of these guidelines, treatment of RET fusion-positive NSCLC was limited for inclusion in clinical trials and compassionate use programs. In total, 21 laboratories in 2017 and 20 laboratories in 2019 had a testing rate for RET fusion above the national testing rate (Figs. 3A and B). The patient cohort not tested for RET fusions is likely due to sequential rather than parallel testing, with sequential testing more commonly applied in 2017 compared to 2019. Furthermore, other reasons for not testing involves inadequate tissue quality or poor health status, the patient is unable to undergo a tissue biopsy procedure due to their physical condition. In case no (appropriate) tumor tissue is available, the Dutch committee for the Assessment of Diagnostic (cieBOD) in June 2024 recommended the use of plasma-based circulating cell-free DNA testing using similar NGS-based assays that include the testing for RET rearrangements [48]. By 2023 and 2022, the ESMO and NCCN guidelines also recommend RET fusion testing, respectively [29,30]. An analysis of RET testing rates reported a lower rate between 40%−50% for a Dutch cohort of NSCLC patients that tested negative for EGFR, KRAS, BRAF, MET, ERBB2 in Q4 of 2017. The discrepancy in testing rates is attributed to the sample selection of the NSCLC population and the negative genes [33]. In a similar study conducted in German patients reported a testing rate similar to this study for RET testing, increasing from 20% to 30% between 2016–2019 [34]. The increase in testing rates for RET fusions in the Netherlands could be attributed to availability of EMA targeted therapies selpercatinib and pralsetinib already in 2015, the use of RNA-based NGS and uptake of RET testing recommendations in the national guidelines. Recently, we reported an overview of the development and application of targeted therapies and clinical-relevant predictive biomarker testing in European patients with advanced stage NSCLC in 2023 including RET-fusion testing. Using data from 11 European countries, we found that since 2023 the EMA-approved selpercatinib and pralsetinib are available and recommended in the national guidelines in all these 11 EU countries. Because these RET-directed targeted drugs were included in the recommendations only recently, also in 2023 none of the other 9 countries had information on RET-fusion testing rates or its prevalences. But interestingly, the commonly-used RET-fusion detection assays were FISH and targeted RNA-NGS in 4, and only targeted RNA-NGS in 7 EU countries. For more comprehensive real-world data on RET fusion testing over time, this information might be collected through international External Quality Assessment programs. A limitation of this study is the retrospective method by which data were analyzed. Here we described a group of patients for whom RET fusions were reported based on the FISH result. However, it is not certain whether these patients do harbor a real RET fusion because of the high false-positive rates of RET FISH testing. Based on these findings and the period the study was conducted, a prevalence lower than the one observed in 2019 is expected in the present. Moreover, although this study was conducted for an entire population, the size of this study was relatively small and compared data from 7 to 9 years ago, thus inclusion of more recent years is recommended to more accurately estimate actual prevalence. Moreover, the availability of drugs targeting RET in either regular care or via clinical trials (e.g. compassionate use program), as well as the reimbursement of this therapy, differs between various countries [9], making these findings specific for the Netherlands. Conclusions Our study demonstrates that RET fusions in patients with non-squamous NSCLC may have a lower prevalence than what was previously been reported in literature. The decrease in detected RET fusions between 2017 and 2019 was associated with a decrease in RET FISH testing and increased use of RNA-based NGS assays. These real-world findings align with previous reports that have described RET FISH testing to yield higher false positive rates compared to RNA-based NGS. As over half of RET fusions in 2019 were still reported based only on FISH assay results, our estimated prevalence of 0.7% may yet be an overestimation of the actual prevalence of RET fusions in patients with NSCLC. Nonetheless, the analysis of more recent data is required to asses the impact of shifting testing methods on the prevalence and treatment outcomes of NSCLC patients. CRediT authorship contribution statement B.N. Cajiao Garcia: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. V.D. de Jager: Writing – review & editing, Methodology, Data curation. C.C.H.J. Epskamp-Kuijpers: Writing – review & editing, Resources, Data curation. A.J. van der Wekken: Writing – review & editing, Supervision, Methodology, Formal analysis, Conceptualization. S.M. Willems: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization. L.C. van Kempen: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Conceptualization. E. Schuuring: Writing – review & editing, Writing – original draft, Visualization, Supervision, Investigation, Funding acquisition, Conceptualization. Declaration of competing interest VdJ has received speaker fees received from Uitgeverij Jaap (Roche Diagnostics Netherlands B.V., Janssen-Cilag B.V.) (paid to UMCG account) and Canadian Anatomic and Molecular Pathology – Pathology Oncology Digital Series (Amgen, Bayer, Daiichi-Sankyo, AstraZeneca, Eli Lilly, Roche, Pfizer) (paid to UMCG account). AvdW has received grants or contracts from AstraZeneca, Boehringer Ingelheim, Pfizer, Roche, and Takeda, has received consulting fees from AstraZeneca, Janssen, Eli Lilly, Roche, and Takeda, has received payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Pfizer, and Roche, has a leadership or fiduciary role in oncology section NVALT, guideline committee NSCLC and CUP, dure geneesmiddelen committee NVALT and FMS. LvK has received institutional grants or contracts from Amgen, AstraZeneca, Bayer, Janssen-Cilag, Merck, Roche, and Servier, has received payments or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from AstraZeneca, Bayer, Bristol-Myers Squibb, Eli Lilly, Janssen, Novartis, Pfizer, and Roche (all to institution), has received support for attending meetings and/or travel from Roche and ThermoFisher (all to institution), has participated on a Data Safety Monitoring Board or Advisory Board of Cyclomics, Janssen-Cilag, LOGEX, Merck, Menarini, Protyon, and Roche (all payments to institution), has a leadership or fiduciary role in the EORTC Melanoma Group and the Commission Personalized Medicine – Belgium (all payments to institution), has stock or stock options in Cyclomics (personal). ES has received (unrestricted) grants from Abbott, Biocartis, AstraZeneca, Invitae/Archer, Bayer, Bio-Rad, Roche, Agena Bioscience, CC Diagnositcs, MSD/Merck, and SNN/EFRO (all paid to UMCG institution), has received advisory board (incidental) and travel expenses (honoraria/grant paid to UMCG institution) from MSD/Merck, AstraZeneca, Roche, Novartis, Bayer, BMS, Lilly, Amgen, Illumina, Agena Bioscience, CC Diagnostics, and Janssen Cilag (Johnson&Johnson), has received advisory board (incidental) (honoraria/grant paid to UMCG institution) from Astellas Pharma, GSK, Sinnovisionlab, Sysmex, and Protyon, has received payments or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Bio-Rad (travel expenses and honoraria paid to UMCG institution), Seracare (honoraria paid to UMCG institution), Roche (travel expenses and honoraria paid to UMCG institution), Biocartis (travel expenses and honoraria paid to UMCG institution), Lilly (travel expenses and honoraria paid to UMCG institution), Agena Bioscience (travel expenses and honoraria paid to UMCG institution), and Illumina (honoraria paid to UMCG institution, has received support for attending meetings and/or travel from BioRad (only travel expenses), Biocartis (only travel expenses), Ageno Sciences (only travel expenses), Illumina (travel expenses), Roche/Foundation Medicine (travel/hotel/registration expenses (partially)), and QCMD (travel/hotel/registration expenses), has a leadership or fiduciary role in the Dutch Society of Pathology (board member, unpaid), European Society of Pathology (board member, unpaid), European Liquid Biopsy Society (board member, unpaid), is chairman/member of the Committee for assessment of molecular diagnostics (cieBOD, honoraria paid to UMCG institution), is committee member of the National guideline advisory committee (honoraria paid to UMCG institution), and is board member of the Committee for Clinical Essential Targets (cieKNT) (unpaid). SW has received grants or contracts from Roche, Bayer, Eli Lilly, Pfizer, AstraZeneca, Merck Sharp & Dohme, Amgen, and Novartis (unrestricted research grants), and has a leadership or fiduciary role in the strategic advisory board of Roche. The remaining authors declare no conflict of interest. Acknowledgements This work was financially supported by a shared grant (‘Landelijke implementatie predictieve analyses bij longkanker op regionaal niveau’) from AstraZeneca, Bayer, Pfizer, Lilly, MSD and Roche. None of these sponsors had a role in the design of the study, data collection, data analysis, interpretation of data, manuscript writing, or the decision to publish the results. Footnotes Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.tranon.2026.102854. Supplementary materials References Associated Data Supplementary Materials