Preoperative profiles of plasma amino acids and derivatives distinguish periampullary cancer and benign disease

Abstract

Periampullary cancers, including pancreatic ductal adenocarcinoma, ampullary-, cholangio-, and duodenal carcinoma, are frequently diagnosed in an advanced stage and are associated with poor overall survival. They are difficult to differentiate from each other and challenging to distinguish from benign periampullary disease preoperatively. To improve the preoperative diagnostics of periampullary neoplasms, clinical or biological markers are warranted.In this study, 28 blood plasma amino acids and derivatives from preoperative patients with benign (N = 45) and malignant (N = 72) periampullary disease were analyzed by LC-MS/MS.Principal component analysis and consensus clustering both separated the patients with cancer and the patients with benign disease. Glutamic acid had significantly higher plasma expression and 15 other metabolites significantly lower plasma expression in patients with malignant disease compared with patients having benign disease. Phenylalanine was the only metabolite associated with improved overall survival (HR = 0.50, CI 0.30-0.83, P < 0.01).Taken together, plasma metabolite profiles from patients with malignant and benign periampullary disease were significantly different and have the potential to distinguish malignant from benign disease preoperatively.

Keywords: Amino acid; Blood plasma; Metabolite; PDAC; Periampullary cancer.

Fig. 1

Plasma metabolite correlation plots based on Spearman correlation. A Correlation of the metabolites in the benign samples. B Correlation of the metabolites in the malignant samples. C The change in correlation comparing the benign to the malignant correlations. The color of the dot represents the direction of correlation on a continuous scale from blue (positive) to red (negative) correlation. The size of the dot indicates the strength of the correlation and the star in the lower quadrant indicates if the correlation is significant (P < 0.05). In panel C, only changes in correlation > 0.25 are visualized

Fig. 2

Boxplot illustrating the difference in metabolite profiles between samples from patients with benign pancreatic disease (blue, N = 45) and patients with periampullary malignancy (grey, N = 72; including 6 patients with neoadjuvant treatment). On the x-axis, the metabolite concentration (log2-transformed and mean centered). Wilcoxon’s Rank Sum Test was used to calculate the p-values. Multiple testing was performed by the Benjamini-Hochberg method (FDR). Significant difference in expression between the malignant and benign group for each metabolite is marked with stars; *: FDR < = 0.05, **: FDR < = 0.01, ***: FDR < = 0.001, ****: FDR < = 0.0001

Fig. 3

PCA plot of the metabolite profiles of patients with benign and malignant periampullary disease. Each dot represents a sample. Each color represents a given type of sample: Black = malignant (N = 60), Blue = benign (N = 45), Orange = malignant sample with neoadjuvant treatment (N = 6). The shape of the dot represents if the sample was collected before or after surgery: Circle = Before, Triangle = After. One sample had both neoadjuvant chemotherapy and a post-surgery sampling date

Fig. 4

Consensus clustered heatmap of 117 patients and 28 metabolites. Four bars beneath the cluster tree illustrate if the given sample is benign (light blue) or malignant (pink), sampled before (dark blue) or after (red) surgery, if the sampled patient had received neoadjuvant treatment (turquoise) or not (grey), and if plasma was isolated from blood immediately after sampling or if the blood was stored overnight before pre-processing

Fig. 5

Kaplan-Meier curve illustrating overall survival of cancer patients with Phenylalanine levels above and below the mean. Patients that received neoadjuvant chemotherapy (N = 6), one patient with metastasis at the time of diagnosis and one patient with non-standard treatment regime were excluded