Background: There is an urgent need for the detection of aggressive prostate cancer. Glycoproteins play essential roles in cancer development, while urine is a noninvasive and easily obtainable biological fluid that contains secretory glycoproteins from the urogenital system. Therefore, here we aimed to identify urinary glycoproteins that are capable of differentiating aggressive from non-aggressive prostate cancer./nMethods: Quantitative mass spectrometry data of glycopeptides from a discovery cohort comprised of 74 aggressive (Gleason score ≥8) and 68 non-aggressive (Gleason score = 6) prostate cancer urine specimens were acquired via a data independent acquisition approach. The glycopeptides showing distinct expression profiles in aggressive relative to non-aggressive prostate cancer were further evaluated for their performance in distinguishing the two groups either individually or in combination with others using repeated 5-fold cross validation with logistic regression to build predictive models. Predictive models showing good performance from the discovery cohort were further evaluated using a validation cohort./nResults: Among the 20 candidate glycoproteins, urinary ACPP outperformed the other candidates. Urinary ACPP can also serve as an adjunct to serum PSA to further improve the discrimination power for aggressive prostate cancer (AUC= 0.82, 95% confidence interval 0.75 to 0.89). A three-signature panel including urinary ACPP, urinary CLU, and serum PSA displayed the ability to distinguish aggressive prostate cancer from non-aggressive prostate cancer with an AUC of 0.86 (95% confidence interval 0.8 to 0.92). Another three-signature panel containing urinary ACPP, urinary LOX, and serum PSA also demonstrated its ability in recognizing aggressive prostate cancer (AUC=0.82, 95% confidence interval 0.75 to 0.9). Moreover, consistent performance was observed from each panel when evaluated using a validation cohort./nConclusion: We have identified glycopeptides of urinary glycoproteins associated with aggressive prostate cancer using a quantitative mass spectrometry-based glycoproteomic approach and demonstrated their potential to serve as noninvasive urinary glycoprotein biomarkers worthy of further validation by a multi-center study.

中文翻译：

---

与侵袭性前列腺癌相关的尿糖蛋白

背景：迫切需要检测侵袭性前列腺癌。糖蛋白在癌症的发展中起着至关重要的作用，而尿液是一种无创且容易获得的生物液体，含有来自泌尿生殖系统的分泌糖蛋白。因此，我们在这里旨在鉴定能够区分侵袭性和非侵袭性前列腺癌的尿糖蛋白。/n方法：糖肽的定量质谱数据来自一个发现队列，该队列由 74 个侵袭性（格里森评分 ≥ 8）和 68 个非侵袭性（格里森评分 = 6）前列腺癌尿液标本组成，这些数据是通过数据独立采集方法获得的。使用重复的 5 倍交叉验证和逻辑回归建立预测模型，进一步评估了在侵袭性相对于非侵袭性前列腺癌中显示出不同表达谱的糖肽在区分两组单独或与其他组的组合方面的性能。使用验证队列进一步评估了从发现队列中显示出良好性能的预测模型。/n结果：在 20 种候选糖蛋白中，尿 ACPP 优于其他候选糖蛋白。尿液 ACPP 还可作为血清 PSA 的辅助手段，以进一步提高对侵袭性前列腺癌的鉴别能力（AUC=0.82，95% 置信区间 0.75 至 0.89）。包括尿 ACPP、尿 CLU 和血清 PSA 在内的三特征组显示出区分侵袭性前列腺癌和非侵袭性前列腺癌的能力，AUC 为 0.86（95% 置信区间 0.8 至 0.92）。另一个包含尿 ACPP、尿 LOX 和血清 PSA 的三特征组也证明了其识别侵袭性前列腺癌的能力（AUC=0.82，95% 置信区间 0.75 至 0.9）。此外，当使用验证队列进行评估时，每个小组都观察到了一致的表现。/n结论： 我们已经使用基于定量质谱的糖蛋白组学方法鉴定了与侵袭性前列腺癌相关的尿糖蛋白糖肽，并证明了它们作为非侵入性尿糖蛋白生物标志物的潜力，值得通过多中心研究进一步验证。