NEW YORK – A team of German researchers in collaboration with Qiagen has built a workflow to analyze genomic and transcriptomic data in parallel from multiple analytes including circulating tumor cells (CTCs), circulating cell-free tumor DNA (cfDNA), and extracellular vesicles (EVs) from a minimal blood volume sample.
The team has now begun analyzing additional analytes in HER2-positive breast cancer patients and will present preliminary results at the American Association of Cancer Research Meeting in May.
As part of the "Evaluation of Multiple Liquid Biopsy Analytes in Metastatic Breast Cancer Patients From One Blood Sample" (ELIMA) trial started in 2016 by Markus Sprenger-Haussels, head of sample technologies product development at Qiagen, the firm began partnering with researchers at the University Hospital Essen to identify how different liquid biopsy analytes could be used in the clinical space.
"We had never seen [CTCs, cfDNA, and EVs] compared from a single sample, and so we wanted to see if one is better than another, or which would give the biggest value for [therapy management]," Sprenger-Haussels said. "The idea was to start a systematic clinical study on a well-defined cohort of cancer patients, look at them in parallel, compare them, and see if there is any synergetic information we could use."
Qiagen believes that the envisioned workflow will allow clinicians to perform multiple analyses from a single blood sample, such as early detection, measuring minimal residual disease, monitoring relapse, and drug response.
"If you need to take a blood tube for each analyte, that's not easy for patients, and thus saving one, two, or even three blood tubes is an achievement from the patient perspective," Sprenger-Haussels said. "So we wanted to see if our technology, if applied in the clinical cohorts, could generate meaningful results for patients."
Project lead author and University Hospital Essen postdoc Corrina Keup said that because HER2 protein expression is routinely used as a predictive marker for targeted therapy in breast cancer, the research team structured the ELIMA project into five different arms by initially collecting blood samples from multiple cohorts of hormone-receptor positive, HER2-negative (HR+/HER2-) metastatic breast cancer (MBC) patients and comparing messenger RNA (mRNA) and genomic DNA (gDNA) from different analytes.
"We wanted to establish a protocol to isolate gDNA and mRNA while using minimal blood volume," Keup explained. "We also wondered if we could use leftover blood after CTC selection for cell-free DNA analysis."
Since launching ELIMA, Keup's team has published three studies examining the clinical potential of different analytes in matched cancer patients, with two additional studies expected to be published later this year.
In the first study, published in Clinical Chemistry in April 2018, the researchers analyzed mRNA profiles of CTCs and EVs from 35 HR+/HER- MBC patients to estimate their utility in therapy management.
Collecting 5 ml of blood at time of disease progression and at two consecutive staging points from each patient, the group isolated CTCs using positive immunomagnetic selection and then collected EVs from 4 ml of plasma using a membrane affinity-based procedure.
Performing reverse-transcription PCR on purified mRNA using Qiagen's Adna-Test EMT-2/StemCell Detect assay, the researchers analyzed the mRNA for 18 genes and normalized the RNA profiles to healthy donor controls.
Keup's team found that the mRNA profiles of CTCs and EVs differed significantly, only sharing about 5 percent of positive signals between fractions.
"We were surprised that the mRNA profiles of CTCs and matched EVs were very different, as we thought that most of the vesicles would match the tumor cells," Keup noted. "[However], we found a significant correlation between ERBB3 overexpression in CTCs with the ERBB3 signals in EVs, ... revealing the additive value of both blood [analytes] and different transcripts of the patients' epidermal growth factor receptor."
The researchers also found that MTOR signals correlated with therapy responsiveness in CTCs and therapy failure in EVs.
Keup's team therefore believes that an integrated approach analyzing mRNA from CTCs and EVs in MBC may allow detection of drug resistance at different timepoints, and that using ERBB3 transcripts in both fractions may be used as a new blood-based monitoring strategy.
The researchers then aimed to establish a cell-free next-generation sequencing workflow by investigating whether cfDNA variant sequencing from CTC-depleted blood affected the results compared to cfDNA variant sequencing from matched whole blood.
In a second study, published in Cancers in Feb. 2019, Keup's team isolated cell-free DNA using 10 ml of matched whole blood and CTC-depleted plasma from 17 (HR+/HER2-) MBC patients.
The team then built cell-free DNA libraries with integrated unique molecular indices (UMIs) using a custom version of Qiagen's Qiaseq Targeted DNA Panel.
The group then sequenced the libraries from the different sample types and analyzed RNA expression in the CTCs using multimarker-PCR.
By using UMIs, the team found that they improved variant detection by 50 percent. Comparing cfDNA variants from whole blood and matched CTC-depleted blood, however, the team did not find any significant difference in the number of variants in each fraction.
"The number of detected variants per patient and the number of exclusively detected variants per patient in only one cfDNA source did not differ between the two matched cfDNA sources," the study authors noted. "[Cell-free DNA] variants from matched whole and CTC-depleted blood exhibited no relevant differences, and parallel isolation of cfDNA and CTCs... in an 'all from one tube' was feasible."
The team therefore believes that matched cfDNA mutational and CTC transcriptional analysis may allow a comprehensive liquid biopsy to improve detection of actionable targets for patient-specific therapy strategies.
Keup's team then sought to compare circulating cfDNA from whole blood and CTC-depleted plasma.
In a third study, published in June 2019 in Cellular and Molecular Life Sciences, the group collected 9 ml of blood from 40 HR+/HER2- MBC patients to examine the prevalence and relevance of detected variants. Isolating cfDNA from the samples, the team created custom libraries for analyzing 18 genes.
The researchers used a targeted NGS approach to sequence all exonic regions of the 18 genes for CTC gDNA and cfDNA, while applying UMIs to guarantee true positive calls and high coverage to identify variants with low allele frequency.
Keup's team found that each patient displayed at least "one pathogenic or likely pathogenic variant," and the highest frequency of pathogenic errors occurred in the androgen receptor gene.
The group also identified single variants with significant correlation with improved survival after metastasis diagnosis. At the same time, the researchers believe that accretion of multiple pathogenic variants may lead to a survival disadvantage for patients, as they found a significant link between the number of variants and patient survival after metastasis diagnosis.
Keup's team therefore believes that identifying new variants with high prevalence, prognostic value, and dynamics under treatment by deep sequencing of ctDNA may allow sensitive monitoring and personalized therapeutic decisions.
In addition to the published studies, Keup's team has launched two additional study arms of the ELIMA project, with a major focus on comparing circulating cfDNA variants from CTC-depleted plasma to matched CTC gDNA.
Analyzing the CTC gDNA and cfDNA with a customized Qiaseq Targeted DNA Panel and UMIs, the researchers found that most of the variants were detected in either CTCs or cfDNA, while half of the cfDNA variants were recovered in the matched CTC gDNA. The group also saw that pathogenic ESR1 and PIK3CA mutations in cfDNA correlated with therapy failure.
The researchers have also submitted results comparing cfDNA and CTC gDNA variants for publication. However Keup noted that her team is still perfecting a bioinformatic approach to integrate data from the multimodal strategy — which compares all the analytes — before submitting the results for an additional publication.
However, Keup's team envisions offering the multimodality liquid biopsy workflow for use on patient blood samples between initial cancer diagnosis and therapy decision using only 20 ml of a blood sample. By analyzing data from the different analytes through qPCR and NGS, Keup believes doctors could identify therapy recommendations for each metastatic breast cancer patient.
"All analytes have additive value, but if you only have a limited amount of blood available, it depends on the clinical question at hand," Keup said. "For example, if you're asking whether to give endocrine therapy, you might only analyze mutations in cfDNA."
Sprenger-Haussels also believes that the multi-modality workflow that the group has built can be applied to other cancers. While analyzing multiple types of genomic and transcriptomic data may initially appear expensive in the clinical setting, Sprenger-Haussels argued that spending money on early diagnosis and prediction of treatment efficacy would outweigh the overall cost of long-term treatment and hospital stay.
Because of the ELIMA project's preliminary results, the researchers launched a new study in 2019, this time collecting blood samples from HER2+ metastatic breast cancer patients. The group will analyze additional liquid biopsy analytes — including microRNA — and potentially perform methylome analysis on blood samples.
Keup noted that she will present the study's preliminary data at the American Association of Cancer Research in May.