Gu Xiaolian, Coates Philip, Wang Lixiao, Erdogan Baris, Salehi Amir, Sgaramella Nicola, Zborayova Katarina, Nylander Karin
Department of Medical Biosciences, Umeå University, Umeå, Sweden.
Research Centre for Applied Molecular Oncology, Masaryk Memorial Cancer Institute, Brno, Czechia.
Front Oncol. 2021 Nov 23;11:753699. doi: 10.3389/fonc.2021.753699. eCollection 2021.
As early detection is crucial for improvement of cancer prognosis, we searched for biomarkers in plasma from individuals who later developed squamous cell carcinoma of the oral tongue (SCCOT) as well as in patients with an already established SCCOT. Levels of 261 proteins related to inflammation and/or tumor processes were measured using the proximity extension assay (PEA) in 179 plasma samples (42 collected before diagnosis of SCCOT with 81 matched controls; 28 collected at diagnosis of SCCOT with 28 matched controls). Statistical modeling tools principal component analysis (PCA) and orthogonal partial least square - discriminant analysis (OPLS-DA) were applied to provide insights into separations between groups. PCA models failed to achieve group separation of SCCOT patients from controls based on protein levels in samples taken prior to diagnosis or at the time of diagnosis. For pre-diagnostic samples and their controls, no significant OPLS-DA model was identified. Potentials for separating pre-diagnostic samples collected up to five years before diagnosis ( = 15) from matched controls ( = 28) were seen in four proteins. For diagnostic samples and controls, the OPLS-DA model indicated that 21 proteins were important for group separation. TNF receptor associated factor 2 (TRAF2), decreased in pre-diagnostic plasma (< 5 years) but increased at diagnosis, was the only protein showing altered levels before and at diagnosis of SCCOT (-value < 0.05). Taken together, changes in plasma protein profiles at diagnosis were evident, but not reliably detectable in pre-diagnostic samples taken before clinical signs of tumor development. Variation in protein levels during cancer development poses a challenge for the identification of biomarkers that could predict SCCOT development.
由于早期检测对于改善癌症预后至关重要，我们在后来发展为口腔舌鳞状细胞癌（SCCOT）的个体以及已确诊SCCOT的患者的血浆中寻找生物标志物。使用邻位延伸分析（PEA）在179份血浆样本中测量了261种与炎症和/或肿瘤过程相关的蛋白质水平（42份在SCCOT诊断前采集，有81份匹配对照；28份在SCCOT诊断时采集，有28份匹配对照）。应用统计建模工具主成分分析（PCA）和正交偏最小二乘判别分析（OPLS-DA）来深入了解各组之间的分离情况。基于诊断前或诊断时采集的样本中的蛋白质水平，PCA模型未能实现SCCOT患者与对照的组间分离。对于诊断前样本及其对照，未识别出显著的OPLS-DA模型。在四种蛋白质中发现了将诊断前长达五年采集的样本（n = 15）与匹配对照（n = 28）分离的潜力。对于诊断样本和对照，OPLS-DA模型表明21种蛋白质对于组间分离很重要。肿瘤坏死因子受体相关因子2（TRAF2）在诊断前血浆中降低（<5年）但在诊断时升高，是唯一在SCCOT诊断前和诊断时水平发生变化的蛋白质（p值<0.05）。综上所述，诊断时血浆蛋白质谱的变化很明显，但在肿瘤发展临床体征出现前采集的诊断前样本中无法可靠检测到。癌症发展过程中蛋白质水平的变化对识别可预测SCCOT发展的生物标志物构成了挑战。
