Reliable identification of prostate cancer using mass spectrometry metabolomic imaging in needle core biopsies
Reliable identification of prostate cancer using mass spectrometry metabolomic imaging in needle core biopsies
复制标题
DOI：
10.1038/s41374-019-0265-2
复制
发表时间：
2019-10-01
影响因子：
5
通讯作者：
Berman, David M.
中科院分区：
文献类型：
--
作者：
Morse, Nicole;Jamaspishvili, Tamara;Berman, David M.
代谢组学分析可以帮助了解癌症发展和进展中的关键生物学过程，还可以产生诊断生物标志物。解吸电喷雾电离耦合质谱成像（ESI-MSI）已被提出作为一个潜在的辅助诊断外科病理学，特别是前列腺癌。然而，由于低分辨率采样，质谱数量少，验证很少，已发表的研究尚未测试该方法是否足够稳健，值得临床翻译。我们使用了超过900个空间分辨的MSI光谱来建立一个准确的，高分辨率的前列腺癌代谢谱。我们鉴定了25种不同丰度的代谢物，癌组织显示脂肪酸（FA）和磷脂增加，沿着三羧酸循环的利用，良性组织显示溶血磷脂酰乙醇胺（PE）水平增加。此外，我们首次鉴定了两种lyso-PE，其丰度随癌症等级而降低，两种磷脂酰胆碱（PCh），其丰度随癌症等级的增加而增加。重要的是，我们开发并内部验证了前列腺癌的多变量代谢组学分类器，在训练队列中使用534个空间感兴趣区域（ROI），在测试队列中使用430个ROI。具有出色的统计功效，训练队列达到了97%的平衡准确度，测试数据集的验证显示了85%的平衡准确度。鉴于该分类器的验证准确性和差异丰富的代谢物与前列腺癌细胞代谢的既定模式的相关性，我们得出结论，MSI是表征前列腺癌代谢的有效工具，具有临床翻译的潜力。
Metabolomic profiling can aid in understanding crucial biological processes in cancer development and progression and can also yield diagnostic biomarkers. Desorption electrospray ionization coupled to mass spectrometry imaging (DESI-MSI) has been proposed as a potential adjunct to diagnostic surgical pathology, particularly for prostate cancer. However, due to low resolution sampling, small numbers of mass spectra, and little validation, published studies have yet to test whether this method is sufficiently robust to merit clinical translation. We used over 900 spatially resolved DESI-MSI spectra to establish an accurate, high-resolution metabolic profile of prostate cancer. We identified 25 differentially abundant metabolites, with cancer tissue showing increased fatty acids (FAs) and phospholipids, along with utilization of the Krebs cycle, and benign tissue showing increased levels of lyso-phosphatidylethanolamine (PE). Additionally, we identified, for the first time, two lyso-PEs with abundance that decreased with cancer grade and two phosphatidylcholines (PChs) with increased abundance with increasing cancer grade. Importantly, we developed and internally validated a multivariate metabolomic classifier for prostate cancer using 534 spatial regions of interest (ROIs) in the training cohort and 430 ROIs in the test cohort. With excellent statistical power, the training cohort achieved a balanced accuracy of 97% and validation on testing data set demonstrated 85% balanced accuracy. Given the validated accuracy of this classifier and the correlation of differentially abundant metabolites with established patterns of prostate cancer cell metabolism, we conclude that DESI-MSI is an effective tool for characterizing prostate cancer metabolism with the potential for clinical translation.