Genomic Complexity of Urothelial Bladder Cancer Best Detected With Urinary Cell-Free DNA Samples

January 22, 2016

Urothelial bladder cancers (UBC) may display complex and heterogeneous features, driven in part by a diverse genomic profile.

1,2

Findings published in theEuropean Journal of Human Geneticsby Togneri et al have shown that cell-free DNA (cfDNA) present in the supernatant of urine samples, allows for an even greater ability to detect genomic tumor biomarkers than does DNA extracted from pelleted cellular material in urine or classically obtained tumor tissue samples.3The findings have important implications for expanding the use of whole genome tumor profiling, disease monitoring, and clinical decision-making in UBC, without the need for invasive biopsies.

The authors highlight several challenging clinical features of UBC, including a need for long-term surveillance, and a variable risk of recurrence, progression, and death, which cannot, at present, be assessed with proven prognostic biomarkers.3They also note that the clinical characteristics of UBC are heterogeneous, and this is driven, in part, by their genomic complexity, which has historically been assessed with the use of formalin-fixed paraffin embedded (FFPE) or fresh-frozen tumor tissue samples.1,2,3

Obtaining such tissue generally involves invasive clinical procedures, which may not always be possible in all patients, nor feasible for ongoing treatment and surveillance. As such, methodologies have been developed to allow for evaluating tumor genomics in UBC though urine sampling; these methods have utilized either DNA obtained from pelleted tumor cells in the urine, or so called cell-free DNA (cfDNA) that remains in the supernatant after the sample has been centrifuged.3The former method may be limited, in part, by the quantity of exfoliated cells that are obtained, and hence the amount of DNA available for analysis.

It has previously been suggested that urinary cfDNA may have the benefit of being highly enriched with tumor cell DNA, relative to the germline DNA of normal tissue that may also be present in the urine sample.4Thus, the goal of the current study was to compare genomic profiling results in UBC as assessed using DNA derived from FFPE tumor tissue, urinary cell pellets, and cfDNA present in the supernatant.3

Material for the study was obtained from a biospecimen repository, and each sample was required to have matched FFPE tissue, urine cell pellets, and more than 5 mL of urine supernatant available for analysis; patients in the study (n = 23) had initial cystoscopic evidence of primary UBC. The investigators were able to confirm, using next-generation sequencing approaches, all mutations that were identified in the FFPE specimens with the urine-obtained DNAs, with only 1 exception.3

Of note, with further comparative analyses, the researchers found a greater tumor genome burden, with a lower germline DNA contamination, in the urinary cfDNA compared with the cellular DNA (P<.001).3There was also a lower rate of detection of targetable mutations in the cellular DNA obtained from the urine pellets compared with the cfDNA in the supernatant. Limitations of the cellular DNA was also evidenced by the fact that there was insufficient cellular material obtained from 6 of the 23 samples to compare with the FFPE assay, whereas 22 of 23 of the urine supernatants provided sufficient cfDNA to accurately characterize the genomic abnormalities in the samples.3

Additional experiments showed a potential benefit of the cfDNA over the FFPE material, as evidenced by a greater aberrant tumor genome load, and in one patient only the cfDNA sample was able to detect the tumor’s genome, as there was insufficient DNA obtained from the FFPE material for the analysis. The average analytical sensitivity was estimated to be ~90% with cfDNA for the detection of key genomic biomarkers identified with FFPE material (with 17 informative samples). By comparison, the corresponding average analytical sensitivity of the cellular DNA samples from urine was 61% (with 15 informative samples), and the difference in analytical sensitivity between these two DNA sources was significant (P<.04).3

Both FFPE and cfDNA samples also revealed an average of 2.3 ‘Tumor Alterations Relevant for GEnomics driven Therapy’ (TARGET) aberrations (22 samples informative), whereas samples derived from the cellular DNA source showed an average of 1.3 TARGET aberrations (18 samples informative). In a blinded analysis using urinary cfDNAs obtained from non-UBC patients, a single patient, who was found later to have prostatic duct carcinoma, had copy number alterations consistent with malignancy, whereas 11 of 12 remaining samples showed a genomics profile that was consistent with the germline DNA only.3

Taken together, their data show that, in comparison to urinary cellular or FFPE sources, urinary cfDNA provides a more highly representative sampling of the tumor genome of patients with UBC, and the investigators hypothesize that an increased rate of necrosis of tumor cells, relative to the normal epithelial cells, may account for the increased representation of tumor genome in the urinary cfDNAs of patients with UBC. Actionable biomarkers that were present in the FFPE specimens at diagnosis could be robustly recapitulated in the urinary cfDNA samples, with high degree of sensitivity.3

Their additional finding of prostate carcinoma in a non-UBC patient further suggests that the use of the urinary cfDNA genomic analysis may not be limited to UBC. The results indicate that this type of cfDNA analyses may be useful as a means for tumor diagnosis and surveillance, and for the identification of relevant prognostic and predictive treatment biomarkers, with the added, highly desirable benefit of being completely noninvasive and applicable for a wide range of patients with UBC.

References

  1. Knowles MA, Hurst CD. Molecular biology of bladder cancer: new insights into pathogenesis and clinical diversity.Nat Rev Cancer. 2015; 15(1): 25-41.
  2. The Cancer Genome Atlas Research Network: Comprehensive molecular characterization of urothelial bladder carcinoma.Nature. 2014; 507: 315-322.
  3. Togneri FS, Ward DG, Foster JM, et al. Genomic complexity of urothelial bladder cancer revealed in urinary cfDNA.Eur J Hum Genet. 2016. doi: 10.1038/ejhg.2015.281. [Epub ahead of print]
  4. Szarvas T, Kovalszky I, Bedi K, Szendroi A, Majoros A, Riesz P et al: Deletion analysis of tumor and urinary DNA to detect bladder cancer: urine supernatant versus urine sediment.Oncol Rep. 2007; 18: 405-409.