Plasma Polyamine Biomarker Panels: Agmatine in Support of Prostate Cancer Diagnosis

Abstract

Prostate cancer is the most frequent malignant tumour among males (19%), often clinically silent and of difficult prognosis. Although several studies have highlighted the diagnostic and prognostic role of circulating biomarkers, such as PSA, their measurement does not necessarily allow the detection of the disease. Within this context, many authors suggest that the evaluation of circulating polyamines could represent a valuable tool, although several analytical problems still counteract their clinical practice. In particular, agmatine seems particularly intriguing, being a potential inhibitor of polyamines commonly derived from arginine. The aim of the present work was to evaluate the potential role of agmatine as a suitable biomarker for the identification of different classes of patients with prostate cancer (PC). For this reason, three groups of human patients-benign prostatic hyperplasia (BPH), precancerous lesion (PL), and prostate cancer (PC)-were recruited from a cohort of patients with suspected prostate cancer (n = 170), and obtained plasma was tested using the LC-HRMS method. Statistics on the receiver operating characteristics curve (ROC), and multivariate analysis were used to examine the predictive value of markers for discrimination among the three patient groups. Statistical analysis models revealed good discrimination using polyamine levels to distinguish the three classes of patients. AUC above 0.8, sensitivity ranging from 67% to 89%, specificity ranging from 74% to 89% and accuracy from 73% to 86%, considering the validation set, were achieved. Agmatine plasma levels were measured in PC (39.9 ± 12.06 ng/mL), BPH (77.62 ± 15.05 ng/mL), and PL (53.31 ± 15.27 ng/mL) patients. ROC analysis of the agmatine panel showed an AUC of 0.959 and p ≤ 0.001. These results could represent a future tool able to discriminate patients belonging to the three different clinical groups.

Keywords: LC-HRMS; agmatine; biomarkers; prostate cancer.

Figure 1

Multivariate analysis using PLS-DA method. PLS-DA loading plot shows a good dis-crimination of the three groups of samples, PC (GREEN), the PL group (RED) and the BPH group (BLUE).

Figure 2

PLS-DA score plots derived from the LC-HRMS spectra of plasma and corresponding coefficient loading plots obtained from the three groups—PC (GREEN), PL (RED), and BPH (BLUE).

Figure 3

Contribution plot from model including total VIP; peaks with positive contribution scores correspond to metabolites with higher levels. The Y-axis indicates the VIP scores corresponding to each variable on the X-axis.

Figure 4

Factors with the highest VIP scores and contributory variables in class discrimination in the PLS-DA model.

Figure 5

Validation method: 300-fold cross permutation validation plot. The Y-axis represents R2 (triangles) and Q2 (circles) for the model, and the X-axis designates the correlation coefficient between original and permuted response data.

Figure 6

(A) ROC curve diagram of agmatine sensitivity and specificity in prediction of pathological conditions in PC and BPH patients. (B) Boxplots showing the distribution of agmatine among subjects with prostate cancer (PC, n = 92), precancerous lesions (PL, n = 26), or benign prostatic hyperplasia (BPH, n = 49). The centre line of the boxplots indicates the median (limits of the box indicate the 25th and 75th percentile). The whiskers represent either 1.5 times the interquartile range (IQR) or the maximum/minimum data point if they are within 1.5 times the IQR. Wilcoxon’s test was used to compare mean agmatine levels among groups.

Figure 7

Arginine decarboxylase (ADC) quantification by ELISA. The concentration of ADC was measured in the plasma of patients from each group. Data are expressed as mean ± SD with reference to the control (mean ± SD; n = 170) (* p ≤ 0.05).