Blog | Nov. 30, 2022
Liquid biopsies: from here to beyond
There is undisputed potential for liquid biopsies to transform the treatment trajectory of patients with cancer. This is due to its flexibility and utility for treatment selection at diagnosis, monitoring of therapeutic response, and early detection of resistance to therapy. Analysis of circulating tumor DNA (ctDNA) in plasma reduces the risk of not detecting clinically-actionable mutations in cancer patients due to tumor heterogeneity. Moreover, as blood draws for liquid biopsy are minimally invasive there is little potential for either inadequate material for analysis or complications arising from obtaining the needed sample.
An additional benefit of liquid biopsy is the turn-around-time to test result: typically 5 to 7 days whereas for tissue biopsy the time to test result is 15-30 days. This reduced time to return a result enables clinicians to make more timely decisions to put the patient on an effective therapy trajectory earlier, and this alone may greatly benefit progression-free and overall survival. While liquid biopsy testing has clear advantages for improving patient management and transforming clinical trial development for novel therapies, not all liquid biopsy approaches demonstrate the same level of performance. The utility of any liquid biopsy for a specific clinical intended use must be considered within the context of an assay’s analytical and clinical performance in order to truly improve patient outcomes and expedite clinical trial development.
Table 1: Copies of mutant ctDNA per mL of plasma, data adapted from Bettegowda et al.1 across multiple tumor types.
In an influential report published in Science Translational Medicine in 2014 (“Detection of circulating tumor DNA in early- and late-stage human malignancies”)1, Plasma-Safe-SeqS and BEAMing technologies were used to measure ctDNA in a wide range of tumor types from both localized and metastatic disease. As one can see from Table 1, the range of copies of ctDNA in plasma vary significantly, and the Plasma-Safe-SeqS method approaches the sensitivity-level of single-molecule detection.
As Sysmex Inostics only requires 2mL of plasma, the range of copies of mutant DNA per 2mL have been calculated in the last column of Table 1. With 2mL of plasma for Plasma-Safe-SeqS circulating tumor DNA analysis, a 0.05% mutant allele frequency sensitivity is achieved. This Limit of Detection (LoD95) means the analyte is detected 95% of the time.
A recent publication compared four liquid biopsy service providers using identical plasma samples matched to tissue. These four tests were sensitive and specific to only 1% mutant-allele frequency (MAF), regardless of their claims, with almost all of the positive calls less than 1% identified to be false positives.2
With a highly sensitive technology like BEAMing and Plasma-Safe-SeqS it becomes clear that nearly 50% of patients have ctDNA detected below 1% MAF. A survey of plasma samples from patients with advanced, treated and/or recurrent disease shows the importance of sensitivity. In metastatic colorectal cancer (mCRC) a full 48% of patients have mutated RAS <1% MAF. For EGFR-mutated T790M in non-small cell lung carcinoma (NSCLC) the percentage of samples with <1% MAF was 42%. And for estrogen-receptor positive, HER2 negative (ER+/HER2-) breast cancer 45% of samples had mutated PIK3CA <1% MAF. This is illustrated in Figure 1.3
Figure 1: Percentage of patients with <1% mutant allele frequencies of RAS, EGFR or PIK3CA
Assuming a conventional limit of 1% MAF for liquid biopsy, what are the economic consequences of using less-sensitive technologies? For head and neck squamous cell carcinoma (HNSCC) where a low-frequency HRAS mutation status (5% of the target population) is used for a clinical trial, a full 89% of the mutant HRAS patients have MAF less than 1%.4 (See Figure 2.) Therefore to obtain 50 patients with mutant HRAS, the difference between having a sensitivity of 1% versus a 0.05% sensitivity for Plasma-Safe-SeqS is an approximately nine-fold difference in screening populations – from over 9,000 to only 1,000 patients needing testing.
Figure 2: The expensive consequences of a low-sensitivity liquid biopsy assay, using head and neck squamous cell carcinoma and mutant HRAS screening status
Several circulating tumor DNA tests on the market claim limits of detection (LoD) that are well-below 1%, often down to 0.1%. Yet for several NGS-based methods, an examination of their technical specifications or their analytical validation publications reveal that their analytical sensitivity drops precipitously at less than 0.5% MAF (see Figure 3). For example, ‘Competitor F’ has a 98.9% analytical sensitivity at >0.5%, but drops to only 67.3% at a target MAF of 0.1 to 0.49% (see Table 2).6
Both OncoBEAM and Plasma-Safe-SeqS technologies have demonstrated sensitivities down to 0.03% and 0.05% MAF respectively, and do not have this kind of drop-off in analytical sensitivity below 0.5% MAF.
Table 2: The target allele frequency, analytical sensitivity and detection limit across four competing ctDNA tests
Notes: For Competitor R, LoD95 is reported on a per-mutation basis as copies/mL. Assuming an average of 5 ng DNA/mL plasma and 3.3 pg/genomic equivalent (GE) = ~1500 genomic equivalents per mL; 25 mutant copies/1500 GE = 1.7% MAF; 100 mutant copies/1500 GE = 6.7% MAF. OncoBEAM Lung: DNA input in validation studies was ≥40 mutant molecules in order to minimize random sampling error. Analytical sensitivity and CI were calculated for LoD samples according to CLSI EP12-A2a. LoD95 was calculated according to CLSI EP17-A2.9
With a publication record stretching back to 2003, the Sysmex Inostics OncoBEAM enhanced digital PCR technology has the longest and largest set of clinical publications for any liquid biopsy method available today. One publication from 2018 compared Sysmex Inostics OncoBEAM technology to droplet digital PCR (from Bio-Rad) and to an NGS method (56G Oncology Panel from Swift Biosciences) for both mCRC and NSCLC with both matched tissue and plasma samples to do the comparison. The authors state in their abstract:10
“Excellent matches between cfDNA/FFPE mutation profiles were observed. Detection thresholds were between 0.5–1% for cfDNA samples examined using ddPCR and NGS, and 0.03% with BEAMing. This high level of sensitivity enabled the detection of KRAS mutations in 5/19 CRC patients with negative FFPE profiles.”
In a recent publication11, investigators compared the performance of OncoBEAM RAS testing to Foundation’s liquid biopsy NGS method in patients with hepatocellular carcinoma, a tumor that is challenging to biopsy and has been shown to exhibit a lower ctDNA rate than other cancers. Investigators found that RAS mutations were not detected by NGS in over 60% of the samples with a MAF between 0.02% and 0.1% as determined by BEAMing. Above 0.1% MAF the plasma NGS method confirmed RAS mutational status in only 44% of patients (44.4%).11
In a study of ESR1 mutations in estrogen receptor-positive (ER+) metastatic breast cancer (mBC) patients receiving selective estrogen receptor degrador (SERD) therapy, OncoBEAM technology was used to test for hotspot mutations in the ESR1 and PIK3CA genes across 153 patients enrolled in a Phase II clinical study.12
The measured allele frequency distribution for the ESR1 mutations in the target patient population is striking (Figure 3). The median MAF is 0.45% with a large number of patients showing MAF of <0.25%. Indeed, of the 153 total samples analyzed, a full 31 mutations in either ESR1 or PIK3CA were measured using OncoBEAM at less than 0.1% (range between 0.020% and 0.096%) and their distribution is shown in Figure 4.
Figure 3: The mutant allele frequency of ESR1 mutations across 153 patient samples
Figure 4: The distribution of measured MAF of ESR1 and PIK3CA mutations from 31 cell-free DNA samples less than 0.1%. Derived from the Supplement12
Recently Sysmex Inostics has launched a next-generation sequencing (NGS)-based liquid biopsy technology called Plasma-Safe-SeqS, with several disease-specific panels available.
One panel called the Plasma-Safe-SeqS Head and Neck Cancer Panel analyses typically low-frequency mutations in the potential therapeutic targets HRAS and PIK3CA, in addition to truncal mutations (with the potential to add additional information to support a true negative finding) in the TP53 and CDKN2A genes. Analysis with reference standard materials show a 95% detection rate down to 0.05% MAF (Figure 5).
Figure 5: Sensitivity (LoD95) is established using SeraCare Seraseq ctDNA Mutation Mix v2 (6 mutations) with the Plasma-Safe-SeqS Head and Neck Cancer Panel. The indicated number of target mutant molecules is in a background of 10K wildtype GE, thus 5 mutant molecules has a corresponding MAF of 0.05%.
In another study, the Plasma-Safe-SeqS ER+/HER2- Breast Cancer Panel was used to analyze both clinical and contrived samples for ESR1, PIK3CA and AKT1 on both the NGS-based Plasma-Safe-SeqS method and the orthogonal OncoBEAM enhanced digital PCR technology. Across 35 clinical ER+/HER2- breast cancer plasma specimens and several reference material samples at a variety of allele frequencies, the measured R-squared value of 0.97 is as shown in Figure 613.
Figure 6: Clinical data derived from 35 samples of ER+/HER2- breast cancer specimens. For the reference material, data points represent averages for replicate testing at different DNA input levels and mutant allele frequency tiers.
Thanks to the ultra-high sensitivity of both the OncoBEAM and Plasma-Safe-SeqS technologies, new applications for liquid biopsy are possible.
For example, in the setting of breast cancer serial monitoring, utilization of a highly sensitive assay like Plasma-Safe-SeqS to detect disease recurrence after administration of adjuvant therapy may allow change in therapeutic decision-making (Figure 7).
Figure 7: Clinical development example using liquid biopsy for detectioin of disease recurrence after administration of adjuvant therapy
To learn more about OncoBEAM and Plasma-Safe-SeqS or other purpose-designed clinical oncology tests from Sysmex Inostics, please contact us today.
Role of PD-1/PD-L1 pathway in the development of Head and neck squamous cell carcinoma (HNSCC)
Key differences between HPV-positive and HPV-negative head and neck squamous cell carcinomas (HNSCC)