NEW YORK – Cancer early detection firm Singlera is pushing forward to commercialize a DNA-methylation-based assay called PanSeer with an immediate focus on colorectal cancer, but potential to expand that to a more pan-cancer setting in the future.
The company, which was founded in 2014 and has offices in California and Shanghai, reached a milestone this month with the publication of promising data from a study of the assay's performance in a longitudinal cohort of cancer-free individuals — the intended use population for this type of screening test.
Reporting in Nature Communications, investigators from the company and collaborating institutions described the application of PanSeer to retrospective samples from a 10-year longitudinal study launched in 2007 by Fudan University in China. Researchers had also previewed some the results previously at a 2018 meeting of the American Society of Human Genetics.
Overall, the group tested blood samples from 605 asymptomatic individuals, 191 of whom were later diagnosed with stomach, esophageal, colorectal, lung or liver cancer — all within four years of the analyzed blood draw. They also analyzed samples from 223 already diagnosed cancer cases, dividing the overall cohort into a training set and an independent validation set.
Authors reported that the test detected cancer in 95 of asymptomatic individuals in the test set who were later diagnosed with cancer.
In addition, the test accurately detected cancer in 88 percent of samples from 113 patients who were already diagnosed with five common cancer types. The test also recognized cancer-free samples 95 percent of the time.
Kun Zhang, who developed the methylation method behind PanSeer and cofounded Singlera, said this week that the commercial strategy for the test has evolved over the last few years to focus on colorectal cancer, at least initially.
The firm announced in May that it had completed a second pre-submission meeting with the US Food and Drug Administration, in which the agency provided "clear guidance on the regulatory pathway and clinical trial plan for the Singlera Genomics ColonES assay."
As numerous methylation-based and other epigenetic technologies have come to dominate a rapidly emerging market of blood-based genomic cancer screening, some companies have maintained a focus on pan-cancer assays, while several have narrowed in on specific cancer types like colorectal cancer.
Joining Singlera in this more focused strategy are companies like Guardant Health and Freenome, which both also utilize DNA methylation in their assays. Other firms, like CellMax Life, are attempting the same with technologies to detect circulating tumor cells.
Singlera said recently that it plans to start the ColonES pivotal study in the United States in the second half of 2020 and has recruited over 100 clinical study sites to participate.
Zhang, who now serves as Singlera's scientific advisor, said the firm has faced some delay due to the Covid-19 pandemic, but is on track for that stated timeline.
The method behind PanSeer relies on identifying methylation haplotypes — instances of co-methylation across a number of different CPG sites — that are specific to a particular tissue or cell type. According to Zhang, the most salient feature of the method compared to other epigenetic strategies is that it requires a relatively small amount of blood, and a relatively limited number of genomic targets, which can then be sequenced more deeply without ballooning assay costs.
The study's authors suggested that the approach may be particularly useful as a first-line, inexpensive screen, with patients testing positive on PanSeer undergoing a "more expensive reflex blood test and/or follow-up imaging to allow tissue of origin mapping."
Rui Liu, a senior author on the paper and co-founder and chief technology officer of Singlera said in a statement that a unique aspect of the study is that it was conducted in a longitudinal cohort with samples obtained from patients up to four years before a cancer diagnosis, as opposed to a case-control design, where an assay might be tested against samples from already diagnosed patients versus normal controls.
The test showed an overall sensitivity of 95 percent in detecting individuals that would later go on to be diagnosed with a cancer. And when the group broke this down into earlier- and later-stage disease, the predictive power remained similar. Analyzing by cancer type, sensitivity ranged from 91 percent in esophageal cancer to 100 percent in liver cancer.
Zhang explained that because of the way this PanSeer method works, it's possible to tune prediction more toward sensitivity or specificity depending on the particular application.
"It uses a proprietary method to do targeted bisulfite sequencing assay of about 500 targets, and at the end you use … [an] algorithm to integrate measurement of these 500 targets into one single score," he said. "Based on that score, you can draw a cutoff [that defines] cancer or no cancer. That's where specificity comes in, because the signal we have ... depending on where you draw the cutoff, you can decide the level of specificity versus sensitivity."
Zhang suggested it boils down to thinking about what kind of people are going to receive the test, what false positive rate is acceptable, and drawing an appropriate cutoff.
The authors wrote that it's not possible to know if the false positive calls in the study were actually misclassifications by PanSeer, or if the patients in question did have cancer present at the time of the blood draw but had not yet developed detectable tumors by the end of the study time period. They decided to be conservative and report them as false positives for the purpose of the study.
In ongoing work using other methylation-based cancer screening assays, some companies have already reported that putative false-negatives have subsequently turned out to be true positives as individuals were later diagnosed.
Similarly, because of the retrospective nature of the analysis, pre-diagnosis false negatives could have been examples of individuals who were cancer-free at the time of their blood draw, but later developed a malignancy. Prospective follow-up studies will help to clarify the test's actual false-negative rate.
One concern about cancer screening and early detection tests now being advanced, especially those that assay epigenetic signals rather than cancer driver mutations, is whether they are actually picking up signs of a cancer, or instead detecting an elevated risk of cancer.
But the Nature Communications authors argued that their data suggests that PanSeer is unlikely to be used to predict whether high-risk patients go on to develop cancer. Instead, it is likely identifying patients who already have cancerous growths but remain asymptomatic and undetectable by current screening methods.
The authors reiterated that further large-scale longitudinal studies, as Singlera is now gearing up for, are needed to confirm the performance they saw in the Fudan University cohort.