Background:

Bladder cancer is the ninth most common cancer worldwide and has high morbidity and mortality. We aimed to search for potential serum peptide biomarkers and establish a diagnostic model for early bladder cancer.

Methods:

A total of 67 bladder cancer patients and 64 healthy volunteers were randomly divided into a training set and testing set 1. There were 30 hematuria patients used as testing set 2. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry based on weak cation exchange magnetic beads was used to obtain and analyze the serum peptide profiles between bladder cancer patients and healthy volunteers in the training set. Serum peptide diagnostic model through a k-nearest neighbor algorithm, was established and validated, and significantly differentially expressed protein biomarkers were ultimately identified.

Results:

We constructed a diagnostic model containing five peptides (m/z 1954.9, m/z 2081.0, m/z 3938.3, m/z 3946.5, and m/z 4268.8). In the training set, the area under the curve (AUC) value of the five-peptide model was 0.923, and the sensitivity and specificity was 93.75% and 96.77%, respectively. In testing set 1, the sensitivity and specificity was 91.43% and 90.91%, respectively, and the specificity of testing set 2 was 73.33%. For early-stage bladder cancer, the sensitivity and specificity was 92.31% and 93.75%, respectively; the sensitivity of early low-grade bladder cancer was 90.00%; and the AUC value was 0.944.

Conclusion:

The five-peptide diagnostic model established in this study had high sensitivity and specificity, especially in the diagnosis of early bladder cancer, and could differentiate between healthy volunteers and hematuria patients.

中文翻译：

---

MALDI-TOF-MS揭示的新型血清肽模型及其在早期膀胱癌中的诊断价值。

背景：

膀胱癌是全球第九大最常见的癌症，发病率和死亡率都很高。我们旨在寻找潜在的血清肽生物标志物并建立早期膀胱癌的诊断模型。

方法：

将 67 名膀胱癌患者和 64 名健康志愿者随机分为训练集和测试集 1。血尿患者 30 名作为测试集 2。基质辅助激光解吸/电离飞行时间质谱基于弱阳离子交换磁珠用于获取和分析训练集中膀胱癌患者和健康志愿者之间的血清肽谱。通过k-最近邻算法建立并验证了血清肽诊断模型，最终鉴定出显着差异表达的蛋白质生物标志物。

结果：

我们构建了一个包含五种肽（m/z 1954.9、m/z 2081.0、m/z 3938.3、m/z 3946.5 和 m/z 4268.8）的诊断模型。在训练集中，五肽模型的曲线下面积（AUC）值为0.923，敏感性和特异性分别为93.75%和96.77%。在测试集 1 中，敏感性和特异性分别为 91.43% 和 90.91%，测试集 2 的特异性为 73.33%。对于早期膀胱癌，敏感性和特异性分别为92.31%和93.75%；早期低级别膀胱癌的敏感性为90.00%；AUC值为0.944。

结论：

本研究建立的五肽诊断模型具有较高的敏感性和特异性，特别是在早期膀胱癌的诊断中，可以区分健康志愿者和血尿患者。