NEW YORK – Based on promising results from two proof-of-concept studies presented at the virtual 2020 American Association for Cancer Research meeting earlier this week, Inivata believes that its RaDaR (Residual Disease and Recurrence) assay could be used to help monitor cancer patient relapse following curative treatment.
In one study, the UK-based firm found that RaDaR detected circulating tumor DNA (ctDNA) in lung cancer patients up to a year before standard clinical progression occurred in most of the cases.
Inivata's RaDaR patient-specific assay runs on the company's InVision sequencing platform, which is built on a technology called targeted amplicon sequencing. RaDaR profiles a patient's tumor from a resection surgery and can target up to 48 unique mutations in the patient's cancer.
After performing whole-exome sequencing (WES) on a patient's tumor tissue to identify somatic variants, Inivata ranks and prioritizes the variants based on proprietary algorithms to design a patient-specific panel.
By developing patient-specific primers, Inivata then builds a primer pool covering up to 48 variants. The firm takes a blood draw and sequences both the buffy coat DNA and the cell-free DNA to compare variants found in the patient's blood sample to those found in the tumor tissue. The team then analyzes the sequencing data using the firm's proprietary algorithm.
Karen Howarth, director of cancer genomics at Inivata and co-author on the AACR studies, explained that RaDaR stands out from similar methods of ctDNA detection because it partitions the sample before sequencing. The partitioning step increases the signal-to-noise ratio to detect low-frequency variants prior to high-depth sequencing.
"If you're putting 10,000 copies into a reaction, and you're looking for one mutant molecule, you're looking at one molecule in a background of 10,000," Howarth said. "But, if you split it into four samples [using partitioning], you're [only] looking at one molecule in 2,500 molecules, which pushes up the signal."
In the first study presented at AACR, Howarth and her colleagues at the Cancer Research UK Cambridge Institute demonstrated RaDaR's ability to detect minimal residual disease (MRD) and cancer recurrence using cancer cell lines and tissues from several tumor types.
The team analyzed three cancer cell lines at different dilutions, DNA reference material, seven formalin-fixed paraffin-embedded samples (from breast, colon, and melanoma), as well as 366 lung cancer samples from the Lung cancer-CIrculating tumor DNA (LUCID) study. They used DNA inputs of 20,000 or 4,000 copies.
"[We] tested [RaDaR] using high and low input DNA, so 20,000 copies [was] used where available, but in very early-stage cancers, it is sometimes more realistic to expect lower amounts of DNA," Howarth explained. "The assay was therefore tested at lower input DNA concentrations of 4,000 copies, [and] we looked at about 1,300 datapoints across different cancer cell lines.”
Using 48 mutations, the researchers found that RaDaR had a sensitivity of 97 percent using 20,000 copies, and of 63 percent using 4,000 copies at a variant dilution of 20 parts per million (equivalent to a variant allele frequency of 0.002 percent), with 100 percent specificity. When looking at a subset of 16 mutations, RaDaR produced a sensitivity of 97 percent at 40 parts per million, 75 percent at 20 parts per million, and 38 percent at 10 parts per million, also with 100 percent specificity.
In the second study, Howarth and her colleagues evaluated RaDaR's prognostic value and its ability to detect ctDNA prior to or during relapse in stage I to stage III non-small cell lung cancer (NSCLC) patients treated with curative intent.
Recruiting 90 patients from the LUCID study who had available tumor tissue, the team collected plasma samples from patients every three months for nine months. Patients who underwent surgery (70) also had a plasma sample taken within three days of resection.
The researchers first performed WES to identify mutations in patient tumor tissues, followed by developing a customized RaDaR panel for each patient. They then correlated detection of residual disease with progression-free survival data.
Analyzing 79 baseline blood samples, the team identified ctDNA in a higher percentage of lung squamous cell carcinoma patients (84 percent) than adenocarcinoma patients (34 percent). They also spotted ctDNA in 35 percent of baseline samples from stage I patients, as well as greater than 90 percent of samples from stage II and stage III patients.
The researchers saw that the tumor fraction (TF) detected using RaDaR ranged from 6 parts per million to 20,000 parts per million in baseline samples. Meanwhile, the group saw that the median TF across all 124 tumor-positive samples was 136 parts per million.
While the results are based on provisional clinical data, Howarth and her colleagues saw that RaDaR detected DNA at baseline or follow-up in about 72 percent of the NSCLC patients. RaDaR could also identify ctDNA between six and 12 months before typical disease progression in 60 percent of the lung cancer patients.
The group also found that for patients with samples analyzed between two weeks and four months, the detection of ctDNA was linked to lower progression-free survival, compared to the ctDNA-negative group.
Howarth noted that building a patient-specific primer panel with 48 primer pairs that produce accurate results "[the] first-time, every-time" was one of the major technical challenges her team dealt with in the proof-of-concept studies. To tackle this issue, the researchers spent time developing the technology and improving the algorithms to deliver a high success rate.
Because of the studies' preliminary results, Inivata CEO Clive Morris believes that a patient-specific assay using RaDaR could be used to monitor recurrence in a wide variety of cancers.
"You could detect the DNA in the blood very early on, and potentially intervene with therapy before a patient realizes that there is anything going on," Morris said. "[But] we need to generate the data that shows [RaDaR's] clinical utility, which might vary from one cancer to another."
As part of Inivata's commercial plans to develop patient-specific RaDaR assays, the firm is now investigating additional cancer types in retrospective and prospective studies. While Morris declined to provide additional information about the cancer types, he noted that the firm plans to share results at future academic meetings.
By offering RaDaR as a tool for MRD and cancer recurrence, Inivata believes the technology will also help biopharmaceutical firms spur drug development for recurrent cancer treatment.
"[While] a majority of the patients are cured, you don't know who they are, [and thus drug] development trials have to be very large and long to get over that dilution effect," Morris said. "We can help biopharma solve that [problem] by identifying who has residual disease to select the patients in the trial, but then also use the same test to monitor the patients and detect recurrence early on."
Morris believes that the AACR data will ultimately lend credence to additional prospective studies that Inivata plans to launch with undisclosed clinical and biopharma collaborators in the future.
By signing an agreement with NeoGenomics last month to commercialize its InVisionFirst lung test, Inivata is leveraging both companies' strengths to work with biopharma companies. While the firms are still evaluating both a laboratory-developed-test pathway and a US Food and Drug Administration pathway for RaDaR, Morris believes the eventual product will be designed to meet all requirements to "ensure we maximize the potential for the product."