Altered Plasma, Urine, and Tissue Profiles of Sulfatides and Sphingomyelins in Patients with Renal Cell Carcinoma.

Abstract

Simple SummaryRenal cell carcinoma (RCC) is among the most common cancer types in both men and women, and its early detection significantly improves survival. Minimally-invasive blood- or urine-based tests may increase the RCC detection rate, especially before patients develop symptoms. Here, we report significant changes in concentrations of sulfatides and sphingomyelins in plasma and urine in RCC patients compared to healthy controls. For the first time, we present findings that similar alterations appear in the lipid profiles of body fluids and tissues in patients. We observe gradual changes in sulfatide and sphingomyelin concentrations with increasing tumor stage and grade. We built binary classifiers that detect RCC based on plasma and urine lipidome dysregulations, and we show that the plasma lipidome alterations enable distinguishing between early-stage RCC and controls. Our results demonstrate the considerable potential of lipid screening in biofluids for RCC detection and monitoring in clinical settings.Purpose: RCC, the most common type of kidney cancer, is associated with high mortality. A non-invasive diagnostic test remains unavailable due to the lack of RCC-specific biomarkers in body fluids. We have previously described a significantly altered profile of sulfatides in RCC tumor tissues, motivating us to investigate whether these alterations are reflected in collectible body fluids and whether they can enable RCC detection. Methods: We collected and further analyzed 143 plasma, 100 urine, and 154 tissue samples from 155 kidney cancer patients, together with 207 plasma and 70 urine samples from 214 healthy controls. Results: For the first time, we show elevated concentrations of lactosylsulfatides and decreased levels of sulfatides with hydroxylated fatty acyls in body fluids of RCC patients compared to controls. These alterations are emphasized in patients with the advanced tumor stage. Classification models are able to distinguish between controls and patients with RCC. In the case of all plasma samples, the AUC for the testing set was 0.903 (0.844–0.954), while for urine samples it was 0.867 (0.763–0.953). The models are able to efficiently detect patients with early- and late-stage RCC based on plasma samples as well. The test set sensitivities were 80.6% and 90%, and AUC values were 0.899 (0.832–0.952) and 0.981 (0.956–0.998), respectively. Conclusion: Similar trends in body fluids and tissues indicate that RCC influences lipid metabolism, and highlight the potential of the studied lipids for minimally-invasive cancer detection, including patients with early tumor stages, as demonstrated by the predictive ability of the applied classification models.