Choline-phospholipids inter-conversion is altered in elderly patients with prostate cancer

Multimodal convolutional neural networks based on the Raman spectra of serum and clinical features for the early diagnosis of prostate cancer

We collected surface-enhanced Raman spectroscopy (SERS) data from the serum of 729 patients with prostate cancer or benign prostatic hyperplasia (BPH), corresponding to their pathological results, and built an artificial intelligence-assisted diagnosis model based on a convolutional neural network (CNN). We then evaluated its value in diagnosing prostate cancer and predicting the Gleason score (GS) using a simple cross-validation method. Our CNN model based on the spectral data for prostate cancer diagnosis revealed an accuracy of 85.14 ± 0.39%. After adjusting the model with patient age and prostate specific antigen (PSA), the accuracy of the multimodal CNN was up to 88.55 ± 0.66%. Our multimodal CNN for distinguishing low-GS/high-GS and GS = 3 + 3/GS = 3 + 4 revealed accuracies of 68 ± 0.58% and 77 ± 0.52%, respectively.

High level of phosphatidylcholines/lysophosphatidylcholine ratio in urine is associated with prostate cancer

The altered levels of phospholipids (PLs) and lysophospholipids (LPLs) in prostate cancer (CaP) and benign tissues in our previous findings prompted us to explore PLs and LPLs as potential biomarkers for CaP. Urinary lipidomics has attracted increasing attention in clinical diagnostics and prognostics for CaP. In this study, 31 prostate tissues obtained from radical prostatectomy were assessed using high‐resolution matrix‐assisted laser desorption/ionization imaging mass spectrometry (HR‐MALDI‐IMS). Urine samples were collected after digital rectal examination (DRE), and urinary lipids were extracted using the acidified Bligh‐Dyer method. The discovery set comprised 75 patients with CaP and 44 with benign prostatic hyperplasia (BPH) at Kyoto University Hospital; the validation set comprised 74 patients with CaP and 59 with BPH at Osaka University Hospital. Urinary lipidomic screening was performed using MALDI time‐of‐flight MS (MALDI‐TOF/MS). The levels of urinary lysophosphatidylcholine (LPC) and phosphatidylcholines (PCs) were compared between the CaP and BPH groups. The (PC [34:2] + PC [34:1])/LPC (16:0) ratio was significantly higher (P < .001) in CaP tissues than in benign epithelial tissues. The urinary PCs/LPC ratio was significantly higher (P < .001) in the CaP group than in the BPH group in the discovery and validation sets.