Tumor Mutational Analysis Using Urinary Cell-Free DNA

June 17, 2014
Special Reports, Precision Diagnostics, Volume 1, Issue 1

Mutational analysis is emerging as a powerful tool in oncology; from the identification and validation of a tumor-specific genetic lesion, to the development of therapeutic agents, to the subsequent, longitudinal assessment of acquired mutations.

Mutational analysis is emerging as a powerful tool in oncology; from the identification and validation of a tumor-specific genetic lesion, to the development of therapeutic agents designed to target that lesion, to the subsequent, longitudinal assessment of acquired mutations that may be associated with therapeutic resistance.

To realize the full potential of this technology, however, rapid, minimally invasive, and cost-effective approaches for mutational analysis of tumors are needed not only at surgery and diagnosis, but also throughout an extended course of treatment. While polymerase chain reaction (PCR)-based technologies allow for the amplification and identification of mutant DNA from small quantities of biopsy material, material for analysis cannot always be readily obtained, or obtained in sufficient quantity or quality, for the desired analysis. Furthermore, the invasive and expensive nature of tumor biopsies precludes their routine use to evaluate tumor response or resistance to a given targeted therapy.

Filip Janku, MD, PhD, on the Importance of Urine Tests for Mutation Detection

Janku is an assistant professor at MD Anderson Cancer Center.

A New Method for Mutational Analysis

In a publication for this year’s American Society of Clinical Oncology (ASCO) Annual Meeting, Filip Janku, MD, PhD, and coworkers at the M.D. Anderson Cancer Center in Houston, Texas, described a technique for longitudinal assessment of a specific mutation,BRAF V600E, in patients with metastatic tumors using urinary cell-free DNA (cfDNA).1 Janku noted that mutational analysis is quickly moving toward becoming the standard of care. He emphasized that not only is this important from the standpoint of identifying patients with a specific mutation who are candidates for targeted therapy, but also for excluding those who are not appropriate candidates for a given targeted therapy, and indeed, who may do worse on that therapy. He noted, for example, that while patients with tumors containingBRAFmutations, who are onBRAFinhibitors, have good responses, those without mutations given the same therapies in fact do worse on the therapy; for this reason, the Food and Drug Administration (FDA) now mandates that those patients initiatingBRAFinhibitors have aBRAFmutation test.David Hyman, MD, a medical oncologist at Memorial Sloan Kettering Cancer Center in New York City, noted that tumor genetic profiling can be challenging because of the limited amount of tumor material present in a biopsy, and that the benefit of urine-based cfDNA genetic profiling is that it allows for testing even when adequate tumor material is unavailable. “This avoids the need for additional, and possibly invasive, biopsies,” Hyman said.

In their study, Janku’s group used urine samples collected sequentially at 4 or more week intervals from patients with advanced cancers known to harbor theBRAF V600Emutation, predetermined by tumor tissue assessment from a Clinical Laboratory Improvement Amendments (CLIA)-certified laboratory.1Mutation-containing fragments were enriched within the samples by preamplification of theBRAFgene, andBRAF V600Emutation in urinary cfDNA was quantitatively evaluated using droplet digital PCR methodology.1,2Janku emphasized that there are several benefits to the urine-based approach for obtaining cfDNA; first, urine is easier to obtain, allowing for multiple time points, which (in the case of plasma) would require multiple blood draws. He also noted that, in the clinical trial setting, patients are required to have multiple blood draws for other purposes, and there is a defined limit, according to the Institutional Review Board (IRB)-approved protocol, to the volume of blood that can be drawn. Another benefit is stability; while blood samples need to be carefully and rapidly processed, urine samples, by comparison, are relatively stable and nonperishable. The higher sensitivity of the PCR assay used for the urine-based approach also shows a clear benefit over other methods; Janku noted that normal testing of tumor biopsy material, even with the most sensitive methods, “is capable of detectingBRAF-mutant DNA in the tissue sample, as long as you have at least 5%, maybe even 10%, of mutant alleles in the sample.” With the urine-based cfDNA system, he noted that a much higher sensitivity is needed for advanced cancers.

Filip Janku, MD, PhD

Agreement (Concordance) Between Methods

“You need to find 1 mutant allele in 1000 wild types, at the least,” Janku said. Likewise, Hyman emphasized that the digital droplet PCR technology used here is “very sensitive and capable of detecting even a very low level of mutations.”The results of Janku’s studies demonstrated that, among patients with advanced cancers harboring theV600Emutation in the tumor (n = 17), the same mutation was detected using urinary cfDNA in 15 patients (88%). In previous studies applying the same methodology, agreement betweenV600Emutation as detected by tumor tissue versus urinary cfDNA was 100% (3 of 3 samples) among patients with Erdheim-Chester Disease (ECD), 95% (19/20) among patients with various metastatic cancers, and 88% among 33 patients with advanced cancers or ECD.1,3,4A separate analysis ofKRASG12 mutations showed agreement between the CLIA laboratory assessed- and urinary cfDNA-assessed KRAS G12 mutation in 7 of 7 patients (100%).3

Use in Longitudinal Analyses

Commenting on the overall concordance between tissue-assessed mutational status and the urine-based cfDNA method, Janku said that “when you collect cell-free DNA either in plasma or in urine, at the same time as you acquire the tissue [biopsy sample]… analyzing both samples at the same time, your concordance is really close to 100%, if not 100%.” He also noted that, as was the case in the present study, if the tissue is collected at a different time as part of the standard patient care, the concordance rate goes down a little, but typically, “concordance is roughly 85% plus… which I think is actually acceptable.”Further findings from the study showed a significant correlation (r = 0.69,P= .002) between changes in the amount ofBRAF V600Eas detected by urinary cfDNA and the response to therapy targetingBRAF/MEK, evaluated by percentage change as determined by the Response Evaluation Criteria In Solid Tumors (RECIST 1.1). There was also a longer median time to treatment failure (259 days vs 61 days,P= .002) for patients who experienced a decrease inBRAF V600E.

Looking Ahead

Commenting on the implications of the longitudinal data, Janku noted that, to date, all of the data have been evaluated in a retrospective manner, that is, patients are treated according to standard of care, sampled, and then evaluated for their response. He suggested that the next step in this research would be to have a situation where one would actually adjust treatment based on evaluation of cfDNA mutational status, that is, intervention based on the cell-free DNA test.  Using this method of urinary cfDNA monitoring for mutations in patients with metastatic cancers, one can envision the next step toward individualized cancer care: a strategy whereby tumor burden, response to targeted therapies, and molecular mechanisms underlying the development of resistance to these can be readily evaluated in a simple and completely noninvasive manner. As a final thought, Janku also alluded to some emerging data indicating that the overall mutational load, as assessed by urinary cfDNA, might be additionally useful as a prognostic indicator for patient survival. Hyman also added, “The ability to longitudinally measure and quantify mutations in cfDNA may help revolutionize the way we follow response to targeted therapies in the clinic”, and he further emphasized that it may soon be possible to predict responses before observed computed tomography (CT) changes, and perhaps detect emergent resistance before the patient develops symptoms or the CT worsens.


  1. 2014 ASCO: Janku F, Vibat CRT, Falchook GS, et al. Longitudinal monitoring ofBRAF V600Emutation in urinary cell-free DNA of patients with metastatic cancers.J Clin Oncol. 2014;32(suppl): Abstract e22175).
  2. Janku F, Vibat CRT, Kosco K, et al. BRAF V600E mutations in urine and plasma cell-free DNA from patients with Erdheim-Chester disease.Oncotarget. 2014 [Epub ahead of print] www.impactjournals.com/oncotarget
  3. Janku F, Falchook GS, Piha-Paul SA, et al. Detection and Monitoring of BRAF and KRAS Mutations in Cell-Free Urinary DNA of Metastatic Cancer Patients by Droplet Digital PCR. American Association for Cancer Research (AACR) 2013 Annual Meeting. Available at: http://www.trovagene.com/trovagene/media/Corporate-Media/TROVAGENE-POSTER-AACR-2.pdf. Accessed June 7, 2014.
  4. Janku F, Falchook GS, Piha-Paul SA, et al. Droplet Digital PCR Mutant and Longitudinal Monitoring of BRAF Mutations in Cell-Free Urinary DNA of Patients Metastatic Cancers or Erdheim-Chester Disease. American Association for Cancer Research (AACR) 2014 Annual Meeting. Abstract LB170. Available at: http://www.trovagene.com/trovagene/media/Trovagene/140408-Trovagene_2014-AACR_Annual_Monitoring.pdf. Accessed June 7, 2014.