Nantong University Affiliated Hospital Licenses Exosomal circRNA-Based Early Lung Adenocarcinoma Diagnostic Assessment Technology for RMB 50,000

Recently, the Affiliated Hospital of Nantong University issued a public notice regarding the transfer of patent rights, in which the hospital proposed to transfer an invention patent.“Evaluation Method for the Efficiency of Exosomal circRNAs in Diagnosing Early-Stage Lung Adenocarcinoma”Transferred to Chengdu Quanyi Intellectual Property Operation Co., Ltd., with a proposed transfer amount of RMB50,000 yuan.

The inventor of this patented technology isZhang Haijian and His R&D Team at the Affiliated Hospital of Nantong University.

Core team members are all from the Affiliated Hospital of Nantong University, with a long-term focus on early diagnosis of lung cancer and research on molecular biomarkers, having accumulated substantial technical expertise in the fields of exosomes and circular RNAs (circRNAs).

This invention providesA Method for Evaluating the Diagnostic Efficiency of Exosomal circRNAs in Early-Stage Lung Adenocarcinoma.This method involves collecting serum exosomes and performing transcriptome sequencing, then employing machine learning techniques such as support vector machine recursive feature elimination (SVM-RFE) and Lasso-logistic regression to screen for key molecular signatures. Subsequently, a scoring system is constructed to systematically evaluate its efficacy in the early diagnosis of lung adenocarcinoma.

Lung adenocarcinoma remains a malignant tumor that seriously threatens human health, and achieving early diagnosis and intervention is crucial for controlling disease progression and improving prognosis. Liquid biopsy is an important in vitro diagnostic technology in this field, which diagnoses diseases by detecting biomarkers in body fluids such as blood, saliva, and urine.

Compared with traditional tissue biopsy,Liquid BiopsyIt offers advantages such as being non-invasive, repeatable, and convenient for dynamic monitoring, demonstrating significant potential in early tumor screening, subtyping, treatment efficacy assessment, and prognosis monitoring.

Currently, common liquid biopsies for tumors mainly focus on circulating tumor cells and circulating tumor nucleic acids. However, these two biomarkers still have certain limitations in practical applications, such as low detection rates, limited specificity, and significant influence from individual differences, which restricts their widespread adoption in clinical practice.

In recent years,Exosomes and Their Carried Molecular Biomarkershas gradually emerged as a new research direction in liquid biopsy. Exosomes are nanoscale vesicles actively secreted by cells, capable of carrying proteins, nucleic acids, and other substances, and participating in intercellular communication.

Among these, circular RNAs (circRNAs) are a class of non-coding RNA molecules characterized by a closed loop structure, exhibiting high abundance and stability within exosomes. Studies have shown that exosomal circRNAs can mediate intercellular communication between tumor cells and other cells in the tumor microenvironment (such as immune cells and fibroblasts), thereby regulating key processes in cancer progression, including immune evasion, angiogenesis, drug resistance, and metastasis.

Given that circRNAs exhibit high stability, widespread distribution, and favorable expression abundance in exosomes, they are considered highly promising liquid biopsy biomarkers for tumor diagnosis and prognostic assessment.

However, to date, no clinically well-validated tumor diagnostic assay based on peripheral blood-derived exosomal circRNAs has been established. To address this technological gap, systematic and effective solutions remain lacking.

To address the technical challenges in the diagnosis of early-stage lung adenocarcinoma, the present invention proposesAn Efficiency Evaluation Method Based on Exosomal circRNAs, aiming to overcome the limitations of existing technical solutions.

The specific technical solutions provided by the present invention are as follows:First, serum samples were collected from patients with early-stage lung adenocarcinoma and healthy volunteers, and exosomes were extracted therefrom.Exosomes are microvesicles secreted by cells that carry various biomolecules, including circular RNAs (circRNAs). Following circRNA transcriptome sequencing of the extracted exosomes, the samples were randomly divided into a training cohort for model construction and a validation cohort for performance verification.

Next, a variety of machine learning and statistical models were integrated to screen for key molecular signatures in the training cohort.In the first step, the Support Vector Machine Recursive Feature Elimination (SVM-RFE) method was employed. This approach evaluates feature importance by iteratively removing features and identifies the feature subset with the lowest classification error rate and highest accuracy, thereby preliminarily selecting 32 candidate circRNAs.

In the second step, a LASSO-logistic regression model was applied. By incorporating an L1 regularization penalty term into logistic regression, this model automatically performs feature selection and prevents overfitting, thereby identifying 22 characteristic circRNAs. In the third step, based on the receiver operating characteristic (ROC) curve and its area under the curve (AUC), a threshold of AUC > 0.7 was set to screen out 21 circRNAs with superior diagnostic performance. Finally, by constructing a Venn diagram to identify the intersection of the results from the aforementioned three methods, seven core circRNAs were ultimately determined as effective molecular biomarkers for the early diagnosis of lung adenocarcinoma.

Then, a comprehensive scoring system was constructed based on the expression levels of these seven molecular markers.A risk score was obtained by calculating the weighted sum of the expression values of each circRNA using a specific formula.

By calculating the sensitivity and specificity of the scoring results and identifying the cutoff value that maximizes the Youden index (sensitivity + specificity − 1), this value was established as the optimal threshold for distinguishing between high-risk and low-risk groups. In the training cohort, the diagnostic efficiency of the scoring system was evaluated, and risk heatmaps were used to visually illustrate the risk distribution across different samples.

Finally, validation was performed in an independent validation cohort using the same molecular labels and scoring system.The area under the curve was recalculated to assess diagnostic accuracy, and decision curve analysis (DCA) was employed to evaluate the clinical net benefit of the model across different decision thresholds, thereby comprehensively demonstrating the reliability and practicality of this molecular signature panel for diagnosis.

The beneficial effects of the present invention lie in systematically screening a panel of circRNA biomarkers with high specificity and high sensitivity from peripheral blood exosomes by integrating high-throughput sequencing with multiple machine learning algorithms.

The constructed scoring model demonstrated excellent diagnostic performance in both the training and validation cohorts (e.g., a combined diagnostic accuracy of 97.4%), providing a non-invasive and precise auxiliary diagnostic tool for early-stage lung adenocarcinoma. This approach not only offers a scientific basis for early intervention but also lays a solid technical foundation for achieving individualized precision diagnosis and treatment.

For the non-invasive diagnostic market for early-stage lung adenocarcinoma targeted by this technology, although detection methods based on a single type of biomarker have made significant progress, clinical practice still calls for comprehensive solutions with higher sensitivity, stronger specificity, and the ability to achieve multi-dimensional assessment.

To this end, numerous enterprises and research institutions both domestically and internationally are actively advancing in the field of liquid biopsy, dedicated to exploring various molecular entities beyond circRNAs within exosomes, and promoting their translational applications in early screening, treatment response monitoring, and prognostic assessment across different cancer types.

Fudan University Shanghai Cancer Center TeamIt was found that under hypoxic conditions, lung cancer cells induce the production of exosomal circPLEKHM1, which drives metastasis in non-small cell lung cancer (NSCLC) by polarizing macrophages toward the M2 phenotype. The underlying mechanism involves the uptake of exosomal circPLEKHM1 by macrophages, where it promotes the interaction between PABPC1 and eIF4G, thereby enhancing the translation of the oncostatin M receptor (OSMR). This ultimately induces M2 polarization of macrophages and accelerates cancer metastasis.

Furthermore, the team employed antisense oligonucleotides (ASOs) targeting circPLEKHM1 for therapeutic intervention, which significantly inhibited NSCLC metastasis in vivo, thereby demonstrating that exosomal circPLEKHM1 can serve as a therapeutic target for lung cancer metastasis.

Research on this target has completed multiple key stages, including target discovery, transcriptional regulation studies, functional validation, therapeutic potential validation, and clinical cohort validation.

Research teams from institutions such as Hospital Universitario Miguel Servet and CIBER Enfermedades Respiratorias in SpainThe LUCEx study has been initiated, with current achievements primarily reflected in the finalized study design and the commencement of research activities.

This is a 5-year prospective, non-interventional longitudinal study planning to enroll 600 patients who undergo radical surgery for CT-suspected malignant lung lesions (postoperatively stratified into cancer and non-cancer groups based on pathological results) and 50 controls with CT false-positive findings. Participants will be followed up for at least 5 years. The study will collect preoperative peripheral blood samples and intraoperative samples of normal lung tissue distal to the tumor. Exosomes will be isolated using commercial precipitation kits and characterized by transmission electron microscopy (morphological characterization), quantified by nanoparticle tracking analysis (NTA), and analyzed for protein markers (e.g., CD63, CD81, EPCAM) by Western blot or ELISA. Small RNA analysis will include RNA sequencing to explore differentially expressed miRNAs and RT-qPCR for validation. These techniques will be employed to investigate the role of exosomal miRNAs in the diagnosis of lung cancer (differentiating benign from malignant lesions and among different pathological subtypes) and prognosis (monitoring postoperative recurrence).

This study has been approved by the Aragon Clinical Research Ethics Committee and is currently in the phase of patient recruitment, sample collection, and follow-up.

In the future, the application of liquid biopsy technology in the field of early tumor diagnosis will continue to deepen. However, substantial efforts are still required in biomarker validation, standardization of testing procedures, cost control, and large-scale prospective clinical studies to truly translate novel biomarkers, including circRNA, into clinically accessible diagnostic products with reliable results. The further development of the industry will depend on a tighter integration between technological innovation and actual clinical needs.