NEW YORK – Cancer detection and monitoring firm C2i this week shared a suite of new study results on its platform for minimal residual disease detection and monitoring using AI and whole-genome sequencing.
The company, which launched as a joint venture between New York Genome Center (NYGC) and Weill Cornell Medicine in August 2019, has set itself apart from other MRD vendors in focusing on developing its platform for distributed use in interested labs, rather than operating solely as a single laboratory-developed test provider.
The company received a CE mark for its C2inform test last year and has had ongoing collaborations with several European institutions, including Aarhus University in Denmark.
At the American Association for Cancer Research annual meeting this week, investigators from that group shared data from a cohort of 146 stage III CRC patients for whom plasma samples had been collected serially for up to three years. Using C2i's AI-based error suppression platform, investigators derived a personalized mutational signature for each patient based on analysis tissue and blood-based WGS data, which they could then apply to detecting circulating tumor DNA (ctDNA).
Patients who tested positive using these panels either immediately after treatment or during subsequent surveillance had much poorer outcomes than those who did not with hazard ratios of 7.1 and 9.5 respectively. Overall, the testing was able to identify recurrence (indicated by positive MRD test result) up to 21 months before detection by radiologic imaging, and about nine months earlier on average.
Examining serially collected plasma samples for novel genomic changes, the researchers could also study the genomic evolution of patient tumors, which they confirmed by WGS of metastatic tissue samples.
C2i CEO Asaf Zviran said that a notable added feature of the Danish study was the ability to examine reproducibility of its system across different labs, which he reiterated is central to the company's vision.
"The ability to run the same assay in different labs across the globe with very high reproducibility and performance [is] part of our mission of trying to democratize this method, so [we performed] a cross-lab validation between the US and in Denmark," he said.
Testing paired samples in Denmark and the US with identical C2i-based bioinformatic processing, the team saw about 99 percent concordance.
Investigators from two US studies also shared new data at AACR. Most notably, a team from Memorial Sloan Kettering presented results from a prospective analysis that was part of the Organ Preservation for Rectal Adenocarcinoma (OPRA) trial.
Investigators used C2inform in 31 patients with locally advanced, stage 2 or 3 rectal cancer undergoing neoadjuvant chemotherapy and radiation, hypothesizing that the assay could serve as prognostic biomarker to guide the course of treatment toward either an organ-sparing watch-and-wait protocol or surgical resection. Patients without a clinical complete response underwent resection while patients with a clinical complete response were enrolled in a watch-and-wait protocol.
According to the authors, all but one patient had a positive ctDNA test at baseline, before therapy. Positive tests during treatment were associated with a lower rate of complete response and shorter time to recurrence, and at a three- to six-month follow-up all five patients who eventually developed clinical recurrence had positive ctDNA results, a finding that the authors said suggests potential clinical utility for treatment de-escalation in the context of organ preservation.
Overall, individuals who had a complete response to neoadjuvant treatment showed clearance of ctDNA down to an undetectable level, while non-responders exhibited either increasing or lingering evidence of cancer.
C2i said that the trial results indicate that surgery, in this case a total mesorectal excision, can be deferred — with equivalent survival outcomes — for the approximately 50 percent of LARC patients that obtain a complete response following neoadjuvant therapy.
The plan now is to expand on these results with a dedicated prospective study that C2i has started in partnership with the Spier Family Foundation and Massachusetts General Hospital, Zviran said.
He added that organ preservation in rectal cancer is also a hot area for the company in terms of pharma partnerships. Zviran didn't name the company but said that C2i recently signed a service agreement with a major firm. "We are really seeing that is a very important focus," he said.
Although colorectal cancer has been the lead indication for MRD technologies in the clinical market, including tests from companies like Natera, Guardant Health, NeoGenomics, and others, it is not the only tumor type where these technologies might have clinical value.
C2i's third study at AACR, with investigators from NYU Langone Health, profiled 50 stage 1 or 2 lung adenocarcinomas. Using its platform to derive a personalized assay for each patient, the company tested blood samples for ctDNA at up to 18 available time points during therapy or follow-up, and investigators quantified patients' ctDNA tumor fraction, comparing it to their clinical status and time to recurrence.
"These early-stage patients are going for surgery and not getting any adjuvant treatment and then being followed, but still there is still something like 20 percent that recur," Zviran said.
Indeed, 33 patients in the cohort showed no recurrence and 12 recurred during the study period. The researchers calculated a recurrence prediction sensitivity of 75 percent and specificity of 82 percent with a positive predictive value of 60 percent and negative predictive value of 90 percent.
Evidence of ctDNA predicted recurrence with a median lead time of 508 days before imaging or clinical symptoms, and in one case, the team was able to identify the emergence of a second primary cancer, not a true recurrence, by studying the mutational signature of the emerging ctDNA.
Zviran said that this was a great illustration of the value of C2i's whole-genome sequencing approach in both tissue and blood, compared to other MRD approaches that monitor ctDNA using small fixed-signature panels.
"With small panels they're basically monitoring a very, very small subset of what is already known in the primary, and so they cannot evaluate tumor evolution and so on. We can basically monitor the entire clonal dynamic and even detect new things," he said.
According to the authors, the results suggest that C2inform could be used both to identify lung cancer patients with lingering disease who could benefit from adjuvant treatment and catch recurrences much earlier than currently possible.
Although the field is still awaiting hard evidence that earlier intervention in cases of lung cancer recurrence can improve patients' overall outcomes, tools like C2inform are making possible, for the first time, the studies that could provide that data.
Zviran said that a clinical network in Europe has recently begun to accept its first out-of-pocket patient samples in a soft launch, and the company expects that to expand later this year with a more formal announcement. C2i also plans to offer testing in the US, but progress on that front is still in the works.