Chicago, IL (UroToday.com) Cell-free tumor-derived DNA (ctDNA) has enabled non-invasive detection and tracking of various cancers. However, the utility of ctDNA in renal cell carcinoma (RCC) has not been well established. In this presentation, Dr. Stewart presented the DIAMOND and MonReC studies, which were designed to undertake a detailed analysis of baseline and longitudinal ctDNA in RCC tumors. This was done by characterizing the detection rates, levels and composition of ctDNA in plasma and urine of RCC patients. The DIAMOND and MonRec study designs are shown in figure 1.
Figure 1 - The DIAMOND and MonRec study designs:
Untargeted sequencing methods revealed that ctDNA levels are lower in RCC patient than other cancers of the similar stage, detecting ctDNA in the plasma and urine of 3/47 (6%) and 4/21 (19%) urine samples, respectively. With size selection, the ctDNA was identified in 34% of patients. Tumors were significantly larger amongst patients with detectable ctDNA as compared to those without. A fragmentation feature based random forest model was capable of triaging patients with detectable ctDNA levels from their sWGS ctDNA profile.
Further interrogation of those patients with detectable ctDNA revealed, for the first time, that urine ctDNA is capable of overcoming genetic heterogeneity and offers information that is complementary to that provided by plasma. Furthermore, INVAR-TAPAS analysis had been shown to improve the signal. Confirming ctDNA detection in RCC patients is challenging, and it does not correlate with tumor size. Longitudinal sampling of >200 plasma and urine samples revealed that in a subset of patients, ctDNA could be used to monitor disease course, and show clonal evolution of the growing lesion.
Dr. Stewart concluded his talk, stating that using the most sensitive methods, ctDNA is present in at least 50% of RCC patients (and up to 85%), but more likely to be found in larger and locally advanced disease, with the highest tumor fractions in metastatic RCC patients. CtDNA detection is challenging in RCC, but this study helps us gain knowledge on what is required to optimize:
- A need for both urine and plasma sampling
- Triage of patients using the random forest model
Presented by: Grant Stewart, University Lecturer; Honorary Consultant Urological Surgeon, The University of Cambridge
Written by: Hanan Goldberg, MD, Urologic Oncology Fellow (SUO), University of Toronto, Princess Margaret Cancer Centre, @GoldbergHanan at the American Urological Association's 2019 Annual Meeting (AUA 2019), May 3 – 6, 2019 in Chicago, Illinois