前列腺癌（PC）的诊断困难常常导致对PC的过度诊断和过度治疗，这凸显了对新型分子标记的需求。本综述的目的是概述目前用于PC分子标记的细胞代谢组学。我们在PubMed上进行了全面搜索，检索了2004年1月至2022年8月间发表的关于PC检测、发展、侵袭性、复发和治疗反应的生物标记研究。虽然有潜在的研究报道了能够区分良性和癌性前列腺组织的区别性分子存在。但是，关于与疾病发展、治疗反应或肿瘤复发有关的特征分子的研究很少。这些研究大多使用高维数据集，而潜在的代谢产物研究数量经常超过可用样本的数量。鉴于此，预分析、统计学、方法学和混杂因素，如抗雄激素治疗（NAT），也可能与数据集中PC和相关对照样本之间的化学计量多变量差异相关联。尽管存在方法学和程序上的挑战，一系列方法学群体和过程一致地鉴定出一些特征代谢产物和通路，这些产物和通路似乎对PC细胞代谢组学在肿瘤形成和复发中发挥重要作用。

The difficulty of diagnosing prostate cancer (PC) with the available biomarkers frequently leads to over-diagnosis and overtreatment of PC, underscoring the need for novel molecular signatures. The purpose of this review is to provide a summary of the currently available cellular metabolomics for PC molecular signatures. A comprehensive search on PubMed was conducted to find studies published between January 2004 and August 2022 that reported biomarkers for PC detection, development, aggressiveness, recurrence and treatment response. Although potential studies have reported the presence of distinguishing molecules that can distinguish between benign and cancerous prostate tissue. However, there are few studies looking into signature molecules linked to disease development, therapy response or tumour recurrence. The majority of these studies use high-dimensional datasets, and the number of potential metabolites investigated frequently exceeds the size of the available samples. In light of this, pre-analytical, statistical, methodological and confounding factors such as antiandrogen therapy (NAT) may also be linked to the identified chemometric multivariate differences between PC and relevant control samples in the datasets. Despite the methodological and procedural challenges, a range of methodological groups and processes have consistently identified a number of signature metabolites and pathways that appear to imply a substantial involvement in the cellular metabolomics of PC for tumour formation and recurrence.