The clonal relation of primary upper urinary tract urothelial carcinoma and paired urothelial carcinoma of the bladder

Abstract The risk of developing urothelial carcinoma of the bladder (UCB) in patients treated by radical nephroureterectomy (RNU) for an upper urinary tract urothelial carcinoma (UTUC) is 22% to 47% in the 2 years after surgery. Subject of debate remains whether UTUC and the subsequent UCB are clonally related or represent separate origins. To investigate the clonal relationship between both entities, we performed targeted DNA sequencing of a panel of 41 genes on matched normal and tumor tissue of 15 primary UTUC patients treated by RNU who later developed 19 UCBs. Based on the detected tumor‐specific DNA aberrations, the paired UTUC and UCB(s) of 11 patients (73.3%) showed a clonal relation, whereas in four patients the molecular results did not indicate a clear clonal relationship. Our results support the hypothesis that UCBs following a primary surgically resected UTUC are predominantly clonally derived recurrences and not separate entities.


| DNA extraction
Tumor hematoxylin and eosin slides were reviewed by an expert genitourinary pathologist (GvL) and regions containing ≥50% tumor cells were selected for DNA isolation (Supplementary Table 2). Tumor and corresponding normal tissue sections were manually microdissected in 5% Chelex 100 Resin (Bio-Rad, Hercules, CA) cell lysis solution (Promega, Madison, WI). DNA was extracted by proteinase K (Roche, Mannheim, Germany) digestion at 56 C. Proteinase K was inactivated for 10 minutes at 95 C after which the samples were centrifuged for 5 minutes at 14000 rpm to collect cell debris and chelexresin. Finally, DNA was collected into new tubes and the concentration was measured by using a Qubit 2.0 fluorometer (Thermo Fisher Scientific, Waltham, MA), as described by the manufacturer.  (Table 2 and Supplementary Table 3). 5-7 NGS was performed with the Ion Torrent platform using supplier's materials and protocols (Thermo Fisher Scientific). Median coverage depths were 1994x for UTUC, 1712x for UCB and 1914x for the adjacent normal tissue. Libraries were made using the Ion AmpliSeq Library Kit plus-384 LV, template was prepared with the Ion 510/520/530 Chef kit and sequencing was performed on a 530-chip using the Ion S5 system. Data were analyzed using SeqPilot (JSI medical systems). To correct for potential germline mutations, NGS was also performed on DNA isolated from matched nonmalignant kidney tissue. The final tumor cell percentage was calculated based on the DNA quality and quantity and the results of the NGS.

| Genomic alterations
A visual inspection by an experienced technician (ICM) and clinical scientist (HJD) in molecular pathology making use of Torrent Variant Caller and SNPitty was carried out to identify the genomic alterations. 6 These genomic alterations were stored in VCF format. 6,8 Figure 1 summarizes all detected genomic alterations; single-nucleotide variants (SNVs), indels, allelic imbalance (AI), amplifications and homozygous deletions. For AI analysis, single nucleotide polymorphisms with a total coverage of >100 reads were included. For any informative SNP without AI, a variant allele frequency (VAF) of 0.5 was expected.
With a VAF of <0.5 (relative loss of variant allele) or a VAF of >0.5 (relative loss of reference allele), AI was indicated. 9 What's new?  threshold value was increased to 0.30 to discard most of the false positives with very low VAF T .
The probability of a clonal relationship between UTUC and UCB samples from the same patient was evaluated following the clonality test approach developed by Ostrovnaya et al. 11 The test was performed on all SNVs and indels. As described by Mauguen et al, the clonality test based on SNVs and indels was performed using the mutation reference data set for bladder cancer from the TCGA study. 12

| RESULTS
In total, 15 patients with primary UTUC, treated by RNU, who subsequently developed 19 UCBs, treated by transurethral resection of the bladder, were included. Patient, treatment and tumor characteristics of the study population are listed in Table 1 and Supplementary Table 1. between both tumors. However, as this mutation only occurs in less than 1% of urothelial carcinoma, a clonal relationship remained statistically significant (P Adj = .025). Patients II and XV also exhibited only a single-shared mutation between both tumors, but as these alterations are common hotspot mutations in urothelial carcinoma, the presence in both entities did not unambiguously reflect a clonal relation. In Patients I and VI, we did not observe any shared somatic mutations, so could not support a clonal relationship.  16 Notwithstanding these limitations, our observation that almost 75% of the paired tumors were clonally related strongly suggests that seeding of tumor cells from the upper urinary tract to the bladder represents the most important mechanism of UCB development following RNU.

| DISCUSSION
Importantly, three patients in our cohort developed multiple subsequent UCBs, and all tumors were clonally related to the primary UTUC, which further supports the mechanism of seeding of tumor cells.

| CONCLUSIONS
The results of our study underscore the rationale to (a) minimalize the