Clinical utility of urine tumor DNA in bladder cancer patients.
Hailong Hu,Yuanjie Niu,HaiTao Wang,Zhaoxia Guo,Huanhuan Liu,Chong Shen,La Da,Gangjian Zhao,Lili Wang,Zhouliang Wu,Zhe Zhang,Shanbo Cao,Feng Lou,Huina Wang
DOI: https://doi.org/10.1200/jco.2022.40.16_suppl.e16560
IF: 45.3
2022-06-01
Journal of Clinical Oncology
Abstract:e16560 Background: Concordance between tumor DNA (tDNA) and urine tumor DNA (utDNA) profiles using paired samples was analyzed. Genomic analysis of utDNA and tDNA showed a high consistency, indicating that urine is an optional sample for noninvasive detection and response evaluation of bladder cancer (BC). Methods: Methods and results data were analyzed from an ongoing study, collecting all cases from Second Hospital Affiliated to Tianjin Medical University. Cases with pathologically confirmed BC and matching tissue/urine/whole blood before treatments were recruited. In 85 patients, 48 with paired samples were enrolled in this analysis after excluding 37 paired samples failed in sample QC or sample collection. WES (Whole Exome Sequence) was applied to tDNA and paired white blood cell (WBC) DNA. An Acornmed 808-Gene panel was applied to utDNA and paired WBC DNA. Somatic events, including SNV (single nucleotide variant), InDel (insertion and deletion) of hot genes were analyzed. Consistency of tDNA and utDNA were identified and analyzed in the intersection set between the two chips, and tDNA profiling was used as the“golden standard”. Results: Among 48 patients, 24(50.0%) were MIBC, 21(43.8%) were NMIBC and 41 (81.5%) were male. With neoadjuvant therapy, 23(47.9%) are identified as responders (CR or PR), 7(14.6%) identified as non-responders (SD or PD), 18(37.5%) are still following up. Overall, 75.4% (214/284) tDNA mutations were found in utDNA, and 75.1% (214/285) utDNA mutations were found in tDNA. The top20 genes’ mutation landscape of utDNA highly resembled that of tDNA. The top20 genes could cover 97.9% (47/48) tissue samples and 89.6% (43/48) urine samples. Also, top20 mutated genes showed a strong correlation between tDNA and utDNA(r = 0.7371, p = 0.0002). TP53, KMT2D, ARIDA1A, CREBBP, ERBB2, ERCC2 and KDM6A are the most frequent abnormalities captured in tDNA and utDNA. Interestingly, we also found that the consistency of top20 genes in responders (n = 23, r = 0.7015, p = 0.0006) are higher than non-responders (n = 7, r = 0.4118, p = 0.1853). As a consequence, 70.1% (157/224), 61.8% (21/34) tDNA mutations were found in utDNA, and 72.0% (157/218), 25.0% (21/84) utDNA mutations were found in tDNA in responders and in non-responders, respectively. Conclusions: In brief, utDNA was robustly consistent and highly associated with tDNA, with 61.8%̃75.4% tDNA variations detected in utDNA target sequecing by acornmed 808 panel. The consistency would be higher if TERT promoter included in WES panel. While in non-responders, only 25.0% utDNA mutations were found in tDNA, indicating a spatial and temporal tumor heterogeneity. Our data provided initial prospective evidence on further developing urine as an optional sample for noninvasive detection and response evaluation of BC. The large prospective cohort study is still ongoing.
oncology