Proceedings Article | 9 October 2021
KEYWORDS: Principal component analysis, Prostate cancer, Raman spectroscopy, Surface enhanced Raman spectroscopy, Diagnostics, Prostate, Statistical analysis, Cancer, Silver, Biopsy
Prostate cancer (PCa) is a main cause of cancer-related death among men aged 50 years and above in the world. To the date, prostate-specific antigen (PSA) is a representational tumour marker which has been widely used in the early diagnosis of PCa. However, in realistic clinical tests, high serum levels of PSA show a high probability for false-positive results, leading to misdiagnoses. The main clinical manifestations of benign prostatic hyperplasia (BPH) is histological hyperplasia of the interstitial and glandular components of the prostate, caused the elderly male are painful to urinate, renal function decreased and so on. Prostate adenocarcinoma of which is very similar to BPH can be expressed in a variety of clinical forms. The difference between the two diseases is that BPH as a benign disease can be cured with proper treatment, however PCa as a malignant tumor and major contributor to cancer-related deaths which there is is difficult to cure so far, so it is important to diagnose and treat it at an early stage. By and large, there are several ways to distinguish between BPH and PCa, including rectal touch, prostate biopsy, PSA testing. These screening tools may give rise to unnecessary biopsies and hurt physical and mental health of patients. To avoid unnecessary invasive biopsy, a new technology namely Metabolomics was introduced. It can closely reflect the content changes of various substances in that the metabolites are expressed in the downstream of genome, transcriptome and proteome. Metabolites are not only excreted in the blood, but also excreted in sweat, respiration, urine, feces, etc. This provides a possibility and a novel idea for noninvasive diagnosis. Urine, as a kind of metabolite, can be used to realize non-invasive cancer detection and diagnosis of other diseases. This non-invasive method detecting diseases by used urine analysis has great potential in cancer diagnosis, monitoring and screening. In addition, many metabolites that may reflect cancer specificity are concentrated in and excreted through urine, which further verifies the logicality of this method. In this study, urine samples were collected from patients who were histologically diagnosed with benign prostatic hyperplasia (BPH) or prostate cancer (PCa). By using surface-enhanced Raman spectroscopy (SERS) with silver nanoparticles, a total of 60 SERS spectra 400~1800 cm-1 were collected from 10 PCa subjects and 10 BPH subjects. Difference spectrum analysis combined with the assignment of Raman bands state clearly that there were obvious changes between prostate cancer and benign prostatic hyperplasia, which could be related to the special changes of xanthopterin, inosine, hypoxanthine, xanthine, uric acid, and urea during pathological changes. In order to distinguish PCa and BPH further, Multivariate statistical techniques, including principal component analysis (PCA) and linear discriminant analysis (LDA) diagnostic algorithms, was employed to analyze the spectra with the result of the diagnostic sensitivity and specificity of 90% and 80%, respectively. It was indicated that the method, using SERS of urine sample combined with PCA-LDA, exhibited good classifications of PCa and BPH. In addition, the effectiveness of PCA-LDA diagnostic algorithm was verified by receiver operating characteristic (ROC) curves and area under the curve (AUC) value was 0.980 (AUC value between 0.85 and 0.95 represents that the effect is very good). Overall, the results of test exhibited the great potential of surface-enhanced Raman spectroscopy analysis of urine combined with PCA-LDA for noninvasive screening and distinguishing PCa from BPH.
