Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Urine as an Alternative to Blood for Cancer Liquid Biopsy and Precision Medicine

Author(s): Zhang A, Lee TJ, Jain S, Su YH

Publication: IEEE International Conference on Bioinformatics and Biomedicine, 2018, Vol. , Page 2815-2820

Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared levels of cell-free DNA (cfDNA) in plasma and urine and investigated if the detection of four variants in plasma or urine can distinguish patients with hepatocellular carcinoma (HCC) from those with cirrhosis or hepatitis. The authors also provide information on how cfDNA levels may be affected by patients consuming fluid (1 L) in the 2 hours preceding urine collection; additional details on methodology were not provided. Deidentified frozen plasma and urine (10 mM EDTA) were obtained from 97 patients with HCC (25% female, mean age: 64.4 years) and 112 non-HCC patients (31% female, mean age: 56.5 years). Total DNA concentration in plasma and urine were compared in urine and plasma specimens collected from 56 patients with HCC. The detection of the HCC markers TP53 249T and CTNNB1 32-37 were compared using plasma and urine from 32 HCC patients. Analysis of hTERT 124 and mRASSF1A was initially conducted using 44 plasma and urine from patients with HCC (16 with >200 ng/mL AFP, 9 with 20-200 ng/mL AFP and 19 with <20 ng/mL AFP). The performance of the 4 ctDNA biomarkers (TP53 249T, CTNNB1 32-37, hTERT 124, and mRASSF1A) was investigated in plasma and urine from 112 HCC patients: 38 patients with hepatitis and 34 patients with cirrhosis. DNA was isolated and fractionated to obtain DNA <400 bp by an unspecified method. Levels of TP53, CTNNB1, hTERT 124, and mRASSF1A were quantified by real-time PCR.  To quantify variants in TP53, CTNNB1, and hTERT 124, an oligonucleotide that inhibited the amplification of wild-type DNA and a variant-specific probe were used in the real-time PCR assays. The methylated mRASSF1A was detected by real-time PCR after bisulfite conversion using the EZ DNA Methylation-Lightning Kit.

   Summary of Findings:

   TP53 and CTNNB1 were both detected in the urine of HCC patients, albeit at significantly lower levels than in matched plasma (P=0.007 and P=0.003, respectively). There was a high degree of variability in TP53 and CTNNB1 levels among urine specimens;  some specimens had a concentration of DNA below the threshold of detectability.  The authors note that TP53 concentration in urine was lower when patients consumed 1 L of liquid in the 2 hours preceding urine collection, although details relating to controls, experimental design, and results were not presented.  While TP53 249T was detected in more urine specimens than plasma specimens (6 versus 0), CTNNB1 32-37 detection was more prevalent in plasma than urine (7 versus 4). Nevertheless, the sensitivity of detection of TP53 249T and/or CTNNB1 32-37 for HCC was higher for urine than plasma (31% versus 22%); urine specimens were also more sensitive at detecting HCC when the markers in question were combined with a minimum detection threshold of >20 ng/mL AFP (sensitivity of 70% for urine and 60% for plasma). Sensitivity for HCC detection was not further increased when plasma detection of TP53 249T and/or CTNNB1 32-37 was combined with urine detection and the >20 ng/mL AFP threshold. The markers hTERT 124 and/or mRASSF1A were detected in more plasma than urine specimens (39 versus 33 of 44), but the difference was not significant.  Combined detection of the markers hTERT 124 and/or mRASSF1A in both plasma and urine led to HCC detection in 43 of 44 (98%) HCC specimens. Importantly, using a combination of hTERT 124 and/or mRASSF1A in plasma and/or urine allowed for HCC detection in 18 of 19 AFP negative specimens where there were no other clinically useful markers. In the larger cohort, the authors confirmed that TP53 249T detection was more prevalent in urine than plasma specimens (14% vs 4%), but hTERT 124 detection was more prevalent in plasma than urine (56% vs. 34%) and that a higher percentage of HCC urine and plasma specimens had detectable hTERT 124 (34% and 56%, respectively) or mRASSF1A (68% and 75%, respectively) than detectable TP53 249T (14% and 4%, respectively) or CTNNB1 32-37 (18%, both). All four markers were generally found to have significantly higher detection in specimens from HCC patients than non-HCC patients (cirrhosis or hepatitis), but the difference was not significant for CTNNB1 32-37 in urine or TP53 249T in plasma. Finally, the sensitivity and specificity for HCC detection using the four markers in plasma and urine combined with AFP were 90.5% and 81%, respectively, although this was the same detection rate observed for the markers in plasma in conjunction with AFP. However, detection of the four markers in urine and plasma had a higher sensitivity for HCC detection than AFP alone (81% versus 71.4%) or detection of the markers only in plasma (81% versus 76.2%).

   Biospecimens

   • Bodily Fluid - Urine
   • Bodily Fluid - Plasma

   Preservative Types

   • Frozen

   Diagnoses:

   • Cirrhosis
   • Neoplastic - Carcinoma
   • Hepatitis

   Platform:

   Analyte	Technology Platform
   DNA	Bisulfite conversion assay
   DNA	Real-time qPCR

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Preaquisition	Diagnosis/ patient condition	HCC Cirrhosis Hepatitis
   Biospecimen Acquisition	Biospecimen location	Plasma Urine
   Biospecimen Acquisition	Time of biospecimen collection	Within 2 h of consumption of 1 L of liquid Assumed not within 2 h of consumption of 1 L of liquid

   View more