Prospective serum metabolomic profiling of lethal prostate cancer

Impaired metabolism may play an important role in the pathogenesis of lethal prostate cancer, yet there is a paucity of evidence regarding the association. We conducted a large prospective serum metabolomic analysis of lethal prostate cancer in 523 cases and 523 matched controls nested within the Alpha‐Tocopherol, Beta‐Carotene Cancer Prevention (ATBC) Study. Median time from baseline fasting serum collection to prostate cancer death was 18 years (maximum 30 years). We identified 860 known biochemicals through an ultrahigh‐performance LC–MS/MS platform. Conditional logistic regression models estimated odds ratios (OR) and 95% confidence intervals of risk associated with 1‐standard deviation (s.d.) increases in log‐metabolite signals. We identified 34 metabolites associated with lethal prostate cancer with a false discovery rate (FDR) < 0.15. Notably, higher serum thioproline, and thioproline combined with two other cysteine‐related amino acids and redox metabolites, cystine and cysteine, were associated with reduced risk (1‐s.d. OR = 0.75 and 0.71, respectively; p  ≤ 8.2 × 10 −5 ). By contrast, the dipeptide leucylglycine (OR = 1.36, p  = 8.2 × 10 −5 ), and three gamma‐glutamyl amino acids (OR = 1.28–1.30, p  ≤ 4.6 × 10 −4 ) were associated with increased risk of lethal prostate cancer. Cases with metastatic disease at diagnosis ( n  = 179) showed elevated risk for several lipids, including especially the ketone body 3‐hydroxybutyrate (BHBA), acyl carnitines, and dicarboxylic fatty acids (1.37 ≤ OR ≤ 1.49, FDR < 0.15). These findings provide a prospective metabolomic profile of lethal prostate cancer characterized by altered biochemicals in the redox, dipeptide, pyrimidine, and gamma‐glutamyl amino acid pathways, whereas ketone bodies and fatty acids were associated specifically with metastatic disease.