Smok-Kalwat Jolanta, Góźdź Stanisław, Macek Paweł, Kalwat Zuzanna, Khalavka Maryna, Rzad Wioletta, Stepulak Andrzej, Depciuch Joanna
Department of Clinical Oncology, Holy Cross Cancer Centre, 3 Artwinskiego Street, 25-734, Kielce, Poland.
Collegium Medicum, Jan Kochanowski University, 25-317, Kielce, Poland.
Sci Rep. 2024 Dec 30;14(1):31678. doi: 10.1038/s41598-024-81649-8.
Using Fourier Transform Infrared spectroscopy (FTIR), it is possible to show chemical composition of materials and / or profile chemical changes occurring in tissues, cells, and body fluids during onset and progression of diseases. For diagnostic application, the use of blood would be the most appropriate in biospectroscopy studies since, (i) it is easily accessible and, (ii) enables frequent analyses of biochemical changes occurring in pathological states. At present, different studies have investigated potential of serum, plasma and sputum being alternative biofluids for lung cancer detection using FTIR. However, until now, it has not been shown which biofluid; among serum and plasma, that can serve as the best material medium for detecting lung cancer with highest levels of accuracy. In this study, plasma and serum isolated from blood consenting participants without lung cancer symptoms (controls) and lung cancer patients. The samples were measured using FTIR and subsequently analyzed by machine learning (ML) algorithms in order to show which fluids (serum or plasma) would better enhance detection of lung cancer. Higher absorbances values of PO, CH, CH and amides vibrations in FTIR spectra of both serum and plasma samples, collected from lung cancer patients were observed in comparison to individuals without lung cancer symptoms (controls). Principal component analysis (PCA) of FTIR spectra showed plasma and serum samples collected from lung cancer patients and individuals without lung cancer symptoms were better differentiated in fingerprinting region (from 800 to 1800 cm) when compared to lipid region (2800-3000 cm). Moreover, also sensitivity specificity and accuracy calculated by logistic regression (LR) and receive operating characteristic (ROC) showed higher values for fingerprint range (800-1800 cm) in comparison with lipids (2800-3000 cm) one for both, serum and plasma. However, using these methods differences between serum and plasma were not existed. From the all obtained results, it was visible, that both fluids could be used in detected lung cancer using FTIR. Moreover, it was also showed that fingerprint range gave a better distinction between the studied patient groups than the lipid range. This was noticeable for both serum and plasma.
利用傅里叶变换红外光谱(FTIR),可以显示材料的化学成分和/或疾病发生和发展过程中组织、细胞和体液中发生的化学变化特征。对于诊断应用,在生物光谱学研究中使用血液是最合适的,因为:(i)血液易于获取;(ii)能够频繁分析病理状态下发生的生化变化。目前,不同的研究已经调查了使用FTIR检测血清、血浆和痰液作为肺癌检测替代生物流体的潜力。然而,到目前为止,尚未表明在血清和血浆中,哪种生物流体能够作为检测肺癌的最佳物质介质,具有最高的准确度。在本研究中,从没有肺癌症状的血液同意参与者(对照组)和肺癌患者中分离出血浆和血清。使用FTIR测量样品,随后通过机器学习(ML)算法进行分析,以显示哪种流体(血清或血浆)能更好地提高肺癌检测率。与没有肺癌症状的个体(对照组)相比,观察到从肺癌患者收集的血清和血浆样品的FTIR光谱中PO、CH、CH和酰胺振动的吸光度值更高。FTIR光谱的主成分分析(PCA)表明,与脂质区域(2800 - 3000 cm)相比,从肺癌患者和没有肺癌症状的个体收集的血浆和血清样品在指纹区域(800至1800 cm)中能更好地分化。此外,通过逻辑回归(LR)和受试者工作特征(ROC)计算的敏感性、特异性和准确性也表明,与脂质(2800 - 3000 cm)区域相比,指纹范围(800 - 1800 cm)在血清和血浆中均具有更高的值。然而,使用这些方法,血清和血浆之间不存在差异。从所有获得的结果可以看出,两种流体均可用于FTIR检测肺癌。此外,还表明指纹范围比脂质范围在研究的患者组之间给出了更好的区分。这在血清和血浆中均很明显。