Efficient distinction between good and malignant lung nodules, this new method may be tried

*For medical professionals for reading reference

Eliminate CT scanning high -fake positive, come to hear what Professor Liang Wenhua says?

With the popularization of low -dose CT scanning, more and more people have "discovered" one or more nodules on their lungs. The early manifestations of lung cancer are lung nodules. In the context of increasing the incidence of lung cancer year by year, many People do n’t think about tea, and the rice is not fragrant. Although CT scanning can reduce the mortality of lung cancer by 20%, the false positive rate of CT scanning is as high as 90%and the excessive resection rate reaches 24%. How to distinguish between good and accurate distinction and malignant lung nodules is currently clinically solved.

At the European Cancer Internal Science Society (ESMO) in 2022, Professor Liang Wenhua shared a model of models that integrate DNA methylation, clinical characteristics and imaging characteristics to accurately perform the classification of lung nodule [1]. Professor Liang Wenhua, the First Affiliated Hospital of the Medical University University, interpreted the accurate lung nodule and improved the diagnosis of lung cancer.

New combination screening,

Cancer CT scan fake positive risks

Lung cancer is one of the most common tumors. Early diagnosis is an effective way to reduce the mortality of lung cancer. Low -dose CT scanning can effectively identify malignant nodules through the size of the lung nodule and the image characteristics of lung cancer such as burrs and pleuramus. It is an important means of early screening of lung cancer.

Although studies have shown that low -dose CT scans can find early lung cancer more effectively, thereby reducing the mortality of lung cancer by 20%[2]. However, Professor Liang pointed out that there is a certain limitations of CT scanning -high -fake positive rate. For patients with lung cancer, patients with lung cancer will cause excessive treatment, increase patient anxiety, and affect the quality of life.

It is to avoid the limitations of CT scanning that Professor Liang's team has conducted research on high -throughput DNA methylation detection models. The accuracy of the model of this model for good and malignant lung nodules is as high as 85%[2]. Although the model accuracy is high, the pulmonary nodules with inert lung nodules and CTDNA non -falling lesions are still unable to detect.

Professor Liang preached that considering that the image features have better sensitivity to CTDNA non -falling lesions, CTDNA biomarkers are more special. I hope that CTDNA methylation, clinical characteristics and image characteristics can be combined (Pulmoseek Plus model) The effect of 1+1> 2 (Figure 1).

Figure 1 Professor Liang Wenhua ESMO explained PPT

Combine multi -dimensional biomarkers,

Accurately diagnose lung nodule classification

In order to verify the accuracy of the Pulmoseek Plus model, Professor Liang's team was incorporated into 1,380 patients with isolation pulmonary nodules in the early stage through the collection and retrospective blind evaluation of forward -looking specimens, and for clinical and video student material logo (CIBM models) and and The establishment and verification of CTDNA methylation (Pulmoseek model) combined models (Figure 2).

Figure 2 Pulmoseek Plus model research solution design

Professor Liang introduced that there were two important discoveries in the study. First: The combination of clinical and imaging student material logos and CTDNA methylation play a 1+1> 2 effect, which improves the accuracy of lung nodule diagnosis. The under area (AUC) of the Pulmoseek Plus model is 0.91, which is significantly better than the CIBM model (0.85) and the Pulmoseek model (0.85).

Second: The combination of clinical and imaging student material logo and CTDNA methylation improve the specificity and sensitivity of pulmonary nodule diagnosis. The Pulmoseek Plus model shows a better performance than Pulmoseak and CIBM separate models, and the sensitivity is increased by 5%-6%. And when the specificity is 50%, the general sensitivity of the Pulmoseek Plus model is 98%, and the sensitivity to different lung nodules is more than 98%(Figure 3).

Figure 3 Pulmoseek Plus, Pulmoseak and CIBM model research results

Professor Liang preached that this shows that the Pulmoseek Plus model is designed to be designed to not miss any malignant nodules, and clinical application is higher. At the same time, the specificity is 50%, which can ensure that patients with 50%of benign lung nodules are not treated. In the nodules that are difficult to judge in existing models, this ratio is as high as 73%.

New combination screening model, clinical usability is higher

In addition to judging the accuracy and sensitivity of the Pulmoseek Plus model, Professor Liang's team also compared the existing pulmonary nodule diagnostic model, such as: Mayo model and Brock model. The results showed that the Pulmoseek Plus model AUC was significantly higher than the existing model, and in the Mayo model, the risk of viciousness was 5%-65%of the lung nodules. The Pulmoseek Plus model could further determine the risk of malignant. The Pulmoseek Plus model applies high -sensitivity low -cut values ​​and high -specific high -cut values, which can divide the uncertain lung nodules into low, medium and high risk groups.

At the same time, the decision curve displayed the Pulmoseek Plus model is the best clinical benefits of the five existing models. By reorganizing, the Pulmoseek Plus model can reduce unnecessary surgery and 89%delayed therapy by 73%(Figure 4).

Figure 4 Pulmoseek Plus model clinical net benefit research results. Finally, Professor Liang pointed out: "In this study, DNA methylation, clinical characteristics, and imaging characteristics multi -dimensional biomarkers combined with good diagnosis and malignant lung nodules, not only high and accurate accuracy Sex can also reduce the burden on patients and deserves to be used in subsequent clinical decisions. "

Expert Introduction

Liang Wenhua

Professor/blog director/deputy chief physician

Department of chest oncology at the First Affiliated Hospital of Guangzhou Medical University

National Youqing Fund winner (oncology), young Pearl River scholar, Guangdong outstanding youth medical talent

Assistant Dean of the Guangzhou Institute of Respiratory Health, Director of the Office of the National Respiratory Medical Center

Deputy Leader of the National Key Laboratory of Respiratory Diseases National Key Laboratory

Chairman of the Commissioner of Immunotherapy for Chest Diseases in Guangdong Province

Deputy Chairman of the Guangdong Medical Association Precision Medicine and Molecular Diagnostic Branch

Youth Member of the China Clinical Oncology Society (CSCO), member of the NSCLC Special Committee, and a member of the expert group of lung cancer guidelines

Deputy editor -in -chief of Transl LUNG CANCER Res, J Thorac Dis and Ann Transl Med

Mainly engaged in comprehensive diagnosis and clinical transformation of lung cancer. So far, more than 200 SCI papers have been published, 12 points above 30 points, 33 more than 10 points, the total impact factors of 2000+, total number of quotes 27000+, H index 34, including 160 first/communication authors, including: J clin oncol (IF 50.7, two articles), nejm, bmj (IF 93.3), Lancet Oncol, Cell Res, J Thorac Oncol, J Clin Invest, Chest and other magazines.

Won ASCO Merit Awards, the Name Doctor of the People's Daily Online, CSCO, 35 most potential youth oncologists in the country, and 2020 Alidhara Hospital Green Orange Award (first medicine). , Innovation team, and national innovation contest.

Literature reference:

[1]https://oncologypro.esmo.org/meeting-resources/esmo-congress/synergistic-combination-of-clinical-imaging-and-dna-methylation-biomarkers-improves-the-classification-of-pulmonary-nodules

[2] OSTROWSKI M, Marjanski T, RZyman W.low-Dose Computed Tomography Screening Reduces Lung Cancer Mortality.advances in Medical Sciences.2018; 63: 230-6.

The first release of this article: the medical world tumor channel

Author of this article: ： 本

Editor in charge: Sweet

- END -