背景：

据2020年全球癌症统计数据显示，肺癌新发病例数220万例，死亡人数180万，发病率仅次于乳腺癌，而死亡率居所有恶性肿瘤之首。这主要与肺癌早期症状隐匿，早期诊断困难有关。由于晚期肺癌患者缺乏手术切除等有效治疗手段，导致其生存率低、预后差，IV期患者的5年生存率不足5%。低剂量电脑断层（Low-dose computed tomography，LDCT）的发展为肺癌的早期诊断提供了新手段，降低了肺癌患者的死亡率。但由于影像学技术的局限性，LDCT对微小肺结节（直径＜3cm）良恶性鉴别诊断的准确性有限，导致早期肺癌筛查的假阳性率较高。血清学生物标志物由于无创、简便、经济等特点，对微小肺结节的良恶性鉴别诊断显得尤为重要，这将有效提升临床对肿瘤早期诊治能力。

目前，多种生物标志物（如proGRP、CEA、CYFRA21-1、SCCA、肺癌相关自身抗体检测等）可与LDCT联合用于肺癌诊断，但这些标志物在早期肺癌诊断中存在敏感性不足，特异性差等局限性。CA125作为卵巢癌和肺癌监测的常用肿瘤标志物，其特异性差，在多种良恶性疾病中均可见血清CA125水平升高。前期我们发现并报道了一例合并血清CA125水平升高的肺占位患者，最终诊断为华氏巨球蛋白血症（Waldenström macroglobulinemia，WM）的案例。该临床案例在拓展了CA125应用范围的同时，引发了我们对传统肿瘤标志物诊断特异性不足等问题的关注。基于传统肿瘤标志物特异性不足等局限性，探索新型生物标志物以提升LDCT在早期肺癌诊断和鉴别诊断中的效能具有重要的临床价值。

目的：

本研究以早期（IA期）肺腺癌患者和良性肺结节患者为研究对象，通过分析两组患者血清蛋白质谱，筛选差异蛋白以发现可用于微小肺结节（直径≤3cm）良恶性鉴别诊断的新型血清生物标志物。

方法：

纳入2022年至2024年就诊于北京协和医院胸外科、呼吸内科的227例早期肺腺癌（Lung adenocarcinoma，LUAD）患者和137例良性肺结节的疾病对照者（Disease Control，DC）作为对照组，其中初筛队列包括100例LUAD患者和50例DC患者，验证队列包括127例LUAD患者和87例DC患者。利用非依赖性采集定量质谱（Data-independent acquisition mass spectrometer，DIA-MS）技术对初筛队列血清样本进行蛋白质组学分析，鉴定组间差异表达蛋白，对所鉴定出的差异表达蛋白进行生物信息学分析，筛选出可作为微小肺结节良恶性鉴别诊断的生物标志物，在独立的验证队列中利用酶联免疫吸附实验（enzyme-linked immunosorbent assay，ELISA）验证候选蛋白质。

结果：

通过DIA-MS共鉴定到2745个蛋白质，其中209个蛋白被认为是组间显著差异蛋白质（FC＞1.5或FC＜0.67，q＜0.05），包含95个显著上调的蛋白质和114个显著下调的蛋白质。生物信息学分析表明，差异蛋白质主要参与免疫应答、补体激活等生物过程，主要富集于补体和凝血的级联反应、胆固醇代谢途径。根据生物信息学分析结果筛选出四个显著差异蛋白进行ELISA验证，成功验证到SERPING1、KNG1、EFNA3的表达水平与前期蛋白质组学结果一致（p＜0.05）。我们发现了三个可用于微小肺结节良恶性鉴别诊断的潜在蛋白标志物，其中KNG1与肺腺癌的疾病相关性既往未报道过。

结论：

通过对早期（IA期）肺腺癌患者与良性肺结节患者的血清样本进行DIA-MS分析，我们发现了三个可用于微小肺结节良恶性鉴别诊断的潜在血清生物标志物，可联合低剂量电脑断层（LDCT）对微小肺结节良恶性进行鉴别，辅助诊断早期肺腺癌，提高LDCT筛查的诊断价值和适用性。

Background:

According to the 2020 Global Cancer Statistics, lung cancer accounted for 2.2 million new cases and 1.8 million deaths, ranking second in incidence after breast cancer while having the highest mortality rate among all malignancies. This is primarily attributed to the insidious early symptoms and diagnostic challenges of lung cancer in the initial stages. Due to the lack of effective treatments such as surgical resection for advanced-stage patients, survival rates remain low with poor prognosis, as the 5-year survival rate for stage IV patients is less than 5%. The development of low-dose computed tomography (LDCT) has provided a new approach for early diagnosis, reducing mortality among lung cancer patients. However, due to imaging technology limitations, LDCT has limited accuracy in differentiating benign from malignant small pulmonary nodules (<3cm in diameter), leading to a high false-positive rate in early lung cancer screening. Serological biomarkers, being noninvasive, convenient, and cost-effective, play a crucial role in distinguishing benign from malignant small pulmonary nodules, thereby significantly improving early clinical diagnosis and treatment of tumors.

Currently, a variety of biomarkers (e.g., proGRP, CEA, CYFRA21-1, SCCA, lung cancer-associated autoantibodies, and EarlyCDT-Lung assay) can be used in conjunction with LDCT for lung cancer diagnosis; however, there are limitations, such as insufficient sensitivity and poor specificity of these markers in the diagnosis of early-stage lung cancer. CA125, as a commonly used tumor marker for ovarian and lung cancer monitoring, has poor specificity, and elevated serum CA125 levels can be seen in a variety of benign and malignant diseases. We previously reported a case of pulmonary lesion with elevated serum CA125, ultimately diagnosed as Waldenström macroglobulinemia (WM). This clinical case, while expanding the scope of application of CA125, raises our concern about the lack of diagnostic specificity of traditional tumor markers and other issues. Given the constraints of traditional tumor markers, the exploration of novel biomarkers to enhance the diagnostic efficacy of LDCT in early-stage lung cancer detection and differential diagnosis is of significant clinical importance.

Objective:

This study focuses on patients with early-stage (stage IA) lung adenocarcinoma and those with benign pulmonary nodules. By analyzing the serum proteomic profiles of both groups, we aim to identify differentially expressed proteins and discover novel serum biomarkers for distinguishing benign from malignant pulmonary nodules (≤3cm in diameter).

Methods:

A total of 227 patients with early-stage lung adenocarcinoma (LUAD) and 137 disease controls (DC) with benign lung nodules were included in this study. The patients in this study were selected from those who visited the Department of Thoracic Surgery and Respiratory Medicine at Peking Union Medical College Hospital (PUMCH) between 2022 and 2024. The primary screening cohort consisted of 100 LUAD patients and 50 DC patients, while the validation cohort comprised 127 LUAD patients and 87 DC patients. Proteomic analysis of serum samples from the primary screening cohort was performed using data-independent acquisition quantitative mass spectrometry (DIA-MS) to identify differentially expressed proteins (DEPs) between the two groups. Bioinformatics analysis was then applied to these DEPs to screen for potential biomarkers that could aid in distinguishing benign from malignant small lung nodules. Finally, enzyme-linked immunosorbent assay (ELISA) was used to validate the candidate proteins in the independent validation cohort.

Results:

DIA-MS-based proteomic analysis identified a total of 2,745 proteins, among which 209 demonstrated significant differential expression (FC>1.5 or FC<0.67, q<0.05) between groups. These differentially expressed proteins included 95 significantly upregulated and 114 downregulated proteins. Bioinformatics analysis indicated their primary involvement in immune response and complement activation, with significant enrichment in pathways including complement and coagulation cascades and cholesterol metabolism. Based on the bioinformatics results, four candidate proteins were selected for ELISA validation. The expression levels of SERPING1, KNG1, and EFNA3 were successfully confirmed to align with the proteomics data (p<0.05). We identified three potential protein markers for distinguishing benign from malignant small lung nodules. Notably, the association of KNG1 with lung adenocarcinoma has not been previously reported.

Conclusion:

Using DIA-MS analysis of serum samples from patients with early-stage (stage IA) lung adenocarcinoma and benign lung nodules, we discovered three serum biomarker candidates capable of discriminating between benign and malignant pulmonary nodules with clinical relevance. These biomarkers could be used in combination with low-dose computed tomography (LDCT) to improve the distinction between benign and malignant nodules, aid in the diagnosis of early-stage lung adenocarcinoma, and enhance the diagnostic accuracy and clinical utility of LDCT screening.