Kenneth R. Butler

Lipid metabolism is regarded as a hallmark of cancer. Sphingolipids, a family of structural and signaling lipids, have antiproliferative (ceramides) or prosurvival [sphingosine 1-phosphate (S1P)] functions. The positive association between statin use and bladder cancer, unlike other cancers, suggests that lipid metabolism not affected by HMG-CoA reductase inhibition may contribute uniquely to this cancer. We conducted a hypothesis-generating study investigating associations between plasma sphingolipid metabolic enzymes and bladder cancer incidence and mortality in the Atherosclerosis Risk in Communities study. Among 10,129 men and women with proteomic profiling of plasma collected in 1993 to 1995, 158 incident bladder cancer cases and 47 bladder cancer deaths were ascertained over a median of 20 years of follow-up (164,013 person-years). Cox regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for protein tertiles adjusted for age, race, sex, and known cancer risk factors, including smoking status, pack-years smoked, and cholesterol-lowering medication use. Of the six sphingolipid enzymes, sphingosine kinase 1 (SPHK1), involved in S1P synthesis, was associated with an increased risk of bladder cancer incidence [tertile 3 (T3) vs. T1: HR, 1.74; 95% CI, 1.15–2.63; P trend = 0.008] and bladder cancer mortality (T3 vs. T1: HR, 2.39; 95% CI, 1.02–5.61; P trend = 0.06). Associations for incidence were attenuated in lagged analyses, suggesting potential reverse causation, whereas associations for mortality were unchanged. Our findings suggest that sphingolipid metabolism, specifically enzymes involved in S1P synthesis, may contribute to bladder cancer etiology, with stronger associations in subgroups with lower cholesterol. These findings motivate confirmatory investigation of plasma S1P in relation to bladder cancer outcomes. Prevention Relevance:We report a consistent positive association for SPHK1, involved in S1P metabolism, and bladder cancer risk and mortality in our prospective study. Stronger positive associations for SPHK1 were noted in subgroups with lower cholesterol, reflecting a shift in structural and/or signaling lipids to influence bladder cancer initiation.

Table S1. List of sphingolipid metabolic enzymes and signaling proteins

Table S10. Analysis of mean value of biomarkers measured at visits 2 and 3

Table S4. Sex stratified analyses

Table S2. Correlation coefficients between sphingolipid metabolic enzymes, cholesterol, and signaling proteins

Table S3. Minimally adjusted associations between sphingolipid metabolic enzymes and bladder cancer incidence and mortality

Table S5. Race stratified analyses

Table S7. Signaling proteins and bladder caner risk and mortality

Table S8. Joint categories of sphingolipid metabolic enzymes and signaling proteins and bladder risk and mortality

Table S6. Cholesterol stratified analyses