NEW YORK (GenomeWeb) – Startup AccuraGen is aiming to tap into the liquid biopsy field, initially in China, with a 61-gene assay it is developing based on creating DNA concatemers that increase the ability to identify low-frequency mutations in cell-free DNA.
Grace Zhao, director of research and a co-founder of the company, spoke at last week's Molecular Medicine Tri-Conference in San Francisco. Following the conference, she told GenomeWeb that the firm, which is based in both Menlo Park, California and Shanghai, China, had raised $6.9 million in a Series A financing round in January 2015 and currently consists of a 12-person R&D team based in the US and a 40-person operational team in China, where it also has six ongoing clinical collaborations.
AccuraGen's other cofounders include Chief Technology Officer Shengrong Lin, Chief Information Officer Johnny Sun, and Director of Genomics Paul Tang. Kang Ying from Fudan University is the company's CEO, and recently, Malek Faham, who was formerly chief scientific officer at immuno-sequencing firm Sequenta before it was acquired by Adaptive Biotechnologies, joined as AccuraGen's CSO.
AccuraGen is developing a next-generation sequencing-based technology to analyze cell-free DNA from plasma. The technology is based on rolling circle amplification. Cell-free DNA from plasma is denatured, creating single-stranded pieces of DNA. That DNA is then amplified using a technique called rolling circle amplification, in which the short piece of DNA is circularized via a ligation reaction. The circularized DNA is then amplified to create a concatemer consisting of numerous tandem repeats.
After rolling circle amplification, the researchers perform target enrichment and sequencing.
Rolling circle amplification increases the overall efficiency of sequencing cell-free DNA while also reducing the number of technical errors introduced, Zhao said.
Sequencing circulating tumor DNA has three main challenges — the extremely small amounts of ctDNA present in a plasma sample, the low signal-to-noise ratio, and the inherent error rate of any sequencing platform, Zhao said.
Other research groups and companies are addressing these issues in a variety of ways, some of whom also presented at last week's Tri-Conference meeting. Many employ some form of molecular barcoding on the front end as well as bioinformatics on the back end to enhance the signal-to-noise ratio and distinguish true variant calls from sequencing error.
AccuraGen is taking a slightly different approach in that it uses rolling circle amplification instead of molecular barcoding. Both techniques help to distinguish low-frequency true variants from sequencing or PCR-induced errors. With molecular barcoding, variants called can be traced back to the fragment of origin in such a way that multiple copies of a DNA fragment with the same molecular barcode will have the same variant. Similarly, with rolling circle amplification, true variants will be present in the same location multiple times in the circularized piece of DNA. If errors are caused by amplification or sequencing, the chances are slim that the same error will occur at the same genomic location multiple times, Zhao said.
One potential disadvantage of molecular barcoding, however, is that attaching barcodes is not very efficient, Zhao said. Barcodes are ligated to each strand of a double-stranded DNA molecule, so the overall techniques have efficiencies of between 10 percent and 30 percent, Zhao said. "If in the first step, you don't have a high conversion efficiency, the whole sensitivity is compromised," Zhao said.
It is especially problematic given that cell-free DNA is already so scarce. By contrast, Zhao said, rolling circle amplification can achieve efficiencies of around 90 percent, increasing the sensitivity and enabling lower input requirements.
The firm has performed analytical validation of its method and is currently conducting six clinical research studies with collaborators in China.
In its analytical validation on cell lines, the team demonstrated that starting from between 10 and 20 ng of cell-free DNA, it could reach 100 percent sensitivity for mutations present at 0.25 percent frequency with no false positives. Going down to 0.1 percent frequency and 0.05 percent frequency mutations, sensitivity fell to 80 percent and 47 percent, respectively, while still not calling false positives.
Zhao said that the 10 ng to 20 ng of cell-free DNA is a "realistic input," given what can typically be obtained from a 5 to 10 mL blood draw.
The firm's clinical collaborations include one with Hongwu Wang, director of endoscope clinics at Beijing Meitan Hospital, to evaluate the assay for different subtypes of lung cancer from patients at different stages of the disease. Already, the researchers have collected over 2,000 samples and sequenced 100.
The researchers are comparing archival tumor tissue to 1 mL of archival plasma in each patient. Zhao highlighted a couple of specific examples. In one patient, the Accu-Act assay reflected the shrinkage of a patient's tumor post-therapy. In a second patient, the Accu-Act showed a dramatic decrease in the allele frequency of two mutations while a patient was being treated. But concurrently, the assay also identified two mutations whose frequency was rising throughout treatment. Zhao said that two months after the last blood draw was taken, the patient developed resistance to the treatment.
In a separate study, AccuraGen is collaborating with Gang Jin, director of the pancreatic cancer center at Changhai Hospital in Shanghai. In that study, the researchers are evaluating the assay in stage I and II pancreatic cancer patients with resectable tumors. The goal is to see whether the ctDNA Accu-Act assay can detect evidence of tumors even in the earliest stages of disease. Zhao reported preliminary results from 80 patients showing that the assay only found ctDNA in about one-third of the patients prior to surgery. However, she noted, detection of ctDNA in these patients also correlated with poor survival rates, "suggesting that it may provide somewhat of a predictive marker for patients before surgery." Zhao said that such results would need to be validated, and noted that for early-stage detection, the firm would have to improve the sensitivity of its assay further.
Shengrong Lin, AccuraGen's CTO, told GenomeWeb that the firm is in the process of transferring its protocols to China and setting up a clinical lab there to validate the Accu-Act test. He said the company plans to launch it there this summer initially as an alternative to tissue biopsy-based tests for patients from whom obtaining a biopsy is not possible.
In addition, the firm plans to collaborate with medical research centers in China to validate the test for monitoring treatment response and evaluating residual disease and expects to extend its indications to those areas in 2017.
Zhao added that the firm is also interested in pursuing collaborations with US-based partners.