VANCOUVER, British Columbia – Plasma and urine show promise as complementary liquid biopsy samples for molecular profiling of tumors using targeted sequencing, according to early-stage research presented at the Association for Molecular Pathology annual meeting held here last week.
The approach could be particularly useful in triple-negative breast cancer (TNBC), where novel biomarkers for clinical monitoring are urgently needed, said Ivonne Nel, first author of the study and head of the Clinical Gynaeco-oncology Lab at the University Hospital Leipzig, during a corporate workshop sponsored by Qiagen and Element Biosciences.
Nel, whose team conducted the research with scientists from the two companies, also presented the results in a scientific poster at the AMP meeting.
Treatment decisions in breast cancer are generally based on the tumor profile at diagnosis, Nel noted. However, tumors are known to change under therapy, developing targetable mutations that often were not detectable in primary tumor tissue.
In nearly every cancer type, circulating tumor DNA (ctDNA) in patient blood has been heavily explored as a source of biomarkers that might provide a real-time picture of tumor progression and aid in early detection of therapy resistance. However, Nel noted, only a few studies have investigated the use of urine as a liquid biopsy source, despite its potential benefits as a clinical sample — noninvasive sample collection, large volumes, high-frequency serial testing, and the potential for self-sampling.
One such study, published by researchers from China in 2020, found detectable mutations in 38 percent of 300 early-stage breast cancer patients, with the mutations showing 97 percent concordance with matched tissue samples.
Some other small studies have also shown promise, Nel said, but research has shown that only about 0.5 percent to 2.5 percent of cfDNA crosses the kidney barrier, making detection more difficult than in blood. This might be overcome by using larger volumes, Nel said, as well as the fact that cfDNA fragments from tumors are generally shorter than those from normal tissue.
In addition, research has shown that urinary cfDNA of cancer patients is increased compared to healthy individuals, but these cfDNA levels are not necessarily cancer-specific. As such, it is necessary to take a mutational profiling approach to urine-based ctDNA detection, she said.
Building on a small pilot study of 15 patients that Nel and colleagues published in 2022 showing proof of principle that blood and urine taken together could be a rich source of breast cancer-related genetic variants, her research team set out more recently to explore the potential for the discovery of novel biomarkers in variants of unknown significance (VUS).
In their validation study, completed just two weeks ago, Nel's team examined a larger cohort of 150 breast cancer patients and used a commercial analyte stabilizer for urine samples, the lack of which may have been a limitation in the pilot study.
They collected matched plasma and urine samples (4 ml and 10 ml, respectively) from the patients, resulting in 300 samples. To date, they have extracted cfDNA from 138 of these samples using Qiagen's QiaAmp MinElute ccfDNA (circulating cell-free DNA) kit. Notably, the median cfDNA concentrations in the urine samples was much higher than in the plasma samples, demonstrating that their analyte stabilizer worked. However, statistical tests revealed that there was only a weak correlation between plasma-derived and urinary cfDNA concentrations, indicating the complementary nature of the two bodily fluids as sample types.
Then, to validate their prior study, they conducted targeted sequencing on matched plasma-derived and urinary cfDNA samples from 30 TNBC patients (60 total samples).
In their previous work published in 2022, they conducted targeted sequencing using the QiaSeq Human Breast Cancer Panel, which amplifies all coding regions of 93 genes known to be associated with breast cancer, then sequenced them using the Illumina NextSeq 550.
In the study presented at AMP, the researchers used the same Qiagen library prep kit but sequenced samples using Element Biosciences' Aviti system. Last year, Qiagen and Element partnered to offer NGS workflows on the Aviti system, which included validating multiple QiaSeq panels on the sequencer and an integrated bioinformatics solution, including Qiagen's CLC LightSpeed ultrafast secondary analysis and QCI Interpret variant interpretation software.
Nel said the Aviti system provides "industry-leading accuracy" due to both its proprietary base chemistry, which generates so-called "polonies" with rolling circle amplification from circular templates, and "avidite" compounds, which increase the power of multivalent binding to ensure high-fidelity basecalling. Another tangential benefit of this is that lower reagent concentrations are needed, cutting overall sequencing costs, Nel said.
One of the key findings from their work so far was that a pathogenic variant of the NCOR1 (nuclear receptor corepressor 1) gene — which encodes a protein that mediates ligand-independent transcription repression of thyroid-hormone and retinoic-acid receptors — was found in 10 of the 30 urine samples and only four plasma samples.
The QCI software revealed that this is an extremely rare urinary biomarker in the literature, and thus, Nel said, warrants further investigation.
The researchers are still finishing up their clinical validation study, but they noted that the early results demonstrate that both urine and plasma are valuable sample types for cfDNA detection; that the Aviti system is suitable for low-frequency variant detection; and that the combination of Aviti with Qiagen's targeted sequencing panels and informatics have potential for finding clinically relevant variants for tumor progression monitoring and therapeutic decision-making.