Date of Award
Doctor of Philosophy
Based on the traditional serological uses to obtain diagnoses of cancer and biopsy techniques, common cancer detection could be not only invasive but expensive and some tests are also unreliable. Currently, urine is one of the most frequently used and collected specimens in the clinical diagnoses. While the areas of urinalysis and metabolomic profiling has received interest in the top clinical research, there are limited components to the validity, specificity, as well as sensitivity of endogenous urine substrates to detect early stages of cancers. Although there is research showcasing that urine is impacted by age, cancer type, geographical location, and diet; there is a substantial gap in knowledge as to how these variables affect the metabolic excretion process, as well as the urine’s specific profile used for cancer diagnosis. The use of urine as a noninvasive detection method is of great importance; currently, renal cell carcinoma (RCC) lacks early detection and is more frequently found incidentally within populations. Kidney cancers’ lifetime risk of diagnosis has been accounted to 1 in 46 males and 1 in 80 women. In 2023, it was estimated that 81,800 new cases will arise in the US, resulting in over 14,890 deaths: with an 85% increased incidence in underserved communities. These experimental studies aimed to evaluate the volatile organic compound (VOCs) profiles between healthy and renal cancer patients, to study the change in biomarkers amongst populations and biospecimens to discover reliable VOCs for renal cancer diagnosis. To validate the urinary VOC biomarker panel, matched urine, blood, and tissue samples will be analyzed in tandem, thereby creating an overlapping liquid biopsy profiles. Objectives of this study included: 1) investigation of the impact of sample conditions on urine metabolites; 2) development of a urinary VOC-based diagnostic model for RCC; and 3) evaluation of potential RCC biomarkers via profile comparisons in urine, blood, and tissue. Our findings revealed that urine storage temperature, storage duration, sample amount, and fasting/non-fasting sample collection did not significantly impact urinary metabolite profiles. This suggests flexibility in urine sample collection conditions, enabling individuals to contribute samples under varying circumstances. However, the influence of freeze-thaw cycles was evident, as VOC profiles exhibited distinct clustering patterns based on increased cycles. This emphasizes the effect of freeze-thaw cycles on the integrity of urinary profiles. Using Wilcoxon Rank Sum Test in screening urine bivariate association with clear cell renal cell carcinoma (ccRCC), 56 VOCs were found to be increased in urine samples of ccRCC-positive patients while 227 VOCs corresponded to the control patients (p < 0.05). A 24 VOC diagnostic logistic model was developed for in the RCC. Based on predicted probabilities from the model via cross-validation, the area under the receiver operating characteristic curve (AUC-ROC) was 0.98, with 99% sensitivity and 97% specificity. The VOC-based RCC diagnostic model was then validated within a testing group. These results showed a promising diagnostic power of this VOC-based model in ccRCC screening with a 0.94 AUC, 86% sensitivity, and 92% specificity. A total of 5,155 VOCs were found in the matched urine, plasma, serum, and tissue samples. Using a 5% occurrence screening, 174 VOCs were found to overlap within all four biomatrices. Of the 174 VOCs screened and validated with our urinary VOCs model, four were identified as being present in all four biomatrices: 1-Hexanol, 2-ethyl-; Heptadecanolide; cis-Vaccenic acid; and carbonic acid, decyl nonyl ester. Our cross-validation using urine as the comparison base indicated that there is a 26% overlap of VOC profiles between all four biospecimens. Further statistical analyses using partial least squares-discriminant analysis (PLS-DA) identified 25 additional VOCs for biomatrices overlap comparison for an additional ccRCC-specific urinary panel. These results showed a promising biomarker diagnostic ability for ccRCC screening by using urine as the primary noninvasive collection and detection matrix. Evolving from this noninvasive diagnostic method of collection and analyses, this research presented a more feasible and accurate point of care method that can be utilized by every individual. This research intended to fill in the gap between the lack of RCC early detection and the current late-stage diagnoses methods, utilize statistical approaches within RCC diagnosis and to bring more awareness and screening participation within communities.
Recieved from ProQuest
Holbrook, Kiana, "A Noninvasive Urine-Based Method For Kidney Cancer Early Detection" (2023). Open Access Theses & Dissertations. 3983.