This article has been updated from a previous version to clarify several comments made by Quantgene CEO Johannes Bhakdi.
NEW YORK – Startup Quantgene is attempting to break into the early cancer detection market with a liquid biopsy sequencing service called Serenity that it currently markets directly to consumers with physician signoff and claims can provide a risk assessment for seven different cancer types.
According to the Santa Monica, California-based company, its service stands out from other tests on the market or currently under development due to its exceptional mutation detection sensitivity and limit of detection, and the fact that it assesses a patient's entire clinical profile to provide relative cancer risk.
However, according to some experts, the firm has not yet provided enough clinical data to support its claims and may run the risk of providing physicians and patients with results that are inaccurate, misleading, or unactionable. In addition, the company's unique service model may attract the scrutiny of regulatory authorities.
After finding out that his mother was diagnosed with stage IV colon cancer, Johannes Bhakdi, Quantgene's cofounder and CEO, wanted to investigate new technologies for early-stage cancer screening and detection. According to Bhakdi, one of his family members had asked him if there was a method that could help identify a random, isolated cell in a patient's liquid sample as a cancer cell.
Bhakdi, who to that point had primarily worked in consulting and investment, visited the University of California, Berkeley around 2015 with the idea of identifying cancer-related targets in a person's blood sample.
Around the same time, Bhakdi cofounded Quantgene with bariatric surgeon and childhood friend Monika Hagen, now the firm's chief medical officer. Bhakdi noted that Hagen is a robotic surgery expert who has previously served in roles at Intuitive Surgical, Verb Surgical, and Johnson & Johnson.
Bhakdi tested his theory by running blood samples that Quantgene had begun collecting as part of a IRB-approved trial on Bio-Rad Droplet Digital PCR, or ddPCR, instruments at the university's DNA Sequencing Facility.
"At the beginning, [Bhakdi] was going to detect cancer markers … by using [prior data] and isolating the fragment DNA, which isn't too hard, and finding the gene," said Hitomi Asahara, director of the facility. "Anyone can pay the fee to use the instrument, but we don't get involved in the experiment."
Asahara explained that ddPCR is a very sensitive but involved process, so people "tend to stay away" and shift to other methods like sequencing to look for disease biomarkers.
"[Bhakdi] was not a trained scientist … and was trying to establish whether his concept worked," Asahara added.
Bhakdi also said that his team worked with researcher Mary West at the high-throughput screening facility of the California Institute for Quantitative Biosciences (QB3) at UC Berkeley.
The team began working with what Bhakdi called an undisclosed Irvine, California-based R&D lab specializing in custom NGS assays for pharma and biotech companies. This helped the company flesh out a next-generation sequencing (NGS)-based, variant-calling assay called "DeepGen" that is part of the firm's Serenity service. Bhakdi emphasized that Serenity is a "solution" rather than a diagnostic test.
Quantgene has raised roughly $20 million in seed funding, Series A financing, and through a recent partnership with Charleston, South Carolina-based molecular diagnostic company Vikor Scientific.
Bhakdi noted that Quantgene has filed for two master patents related to the DeepGen assay and the firm's proprietary "Griffin artificial intelligence and core" platform.
Quantgene's DeepGen assay workflow involves centrifuging and isolating cell-free DNA (cfDNA) from a customer's blood and saliva sample using "customized protocols to maximize cfDNA yield," Bhakdi said.
In purified plasma samples, unique molecular identifiers are then ligated to individual cfDNA strands followed by a PCR enrichment step of cancer-relevant sequences using a customized primer panel, which covers genomic targets across 272 genes.
Enriched libraries in the plasma sample undergo NGS and saliva samples undergo whole-exome sequencing (WES) on an Illumina NovaSeq 6000 instrument.
Bhakdi added that sequencing results are then analyzed with Quantgene's Griffin AI platform, which identifies markers including single-nucleotide polymorphisms, multi-nucleotide variants, insertions, and deletions in a blood sample; and SNPs "and other things" in a saliva sample.
"We are combining … somatic data with germline WES as a way to further contextualize each patient's unique genetic profile," said Bhakdi. "We chose WES instead of [whole-genome sequencing] because the exons are the most information-dense sections of the human genome, containing approximately 85 percent of known disease-causing variants."
Quantgene has built its Griffin AI system using a mixture of large public genetic datasets — including The Cancer Genome Atlas and the Institute for Cancer Genetics and Informatics — and undisclosed private tissue datasets, as well as data from about 1,000 samples from Quantgene's ongoing clinical trial.
"We found that you need to hand-build your software and the statistics around your specific panel to get to single-[molecule] accuracy," Bhakdi said. DeepGen, he noted, includes all US Food and Drug Administration-approved drug targets and many targets that are "not known in the literature to be associated with cancer, that we found to be good markers for detection."
The Griffin system also integrates a "high-level statistical evaluation" of a patient's medical records, including their family genetic and cancer history, Bhakdi said. The system generates cancer risk profiles that tell customers their probability of having specific cancers at that point in time.
While Quantgene designed the DeepGen assay to detect variants associated with 15 cancer types, Bhakdi said that the Serenity service currently identifies markers linked to seven cancers, including pancreatic, breast, colorectal, liver, bladder, lung, and prostate cancer. He added that the firm plans to improve the solution to detect all 15 cancers by mid-2022.
A customer interested in Serenity can contact Quantgene through its website. Bhakdi said the firm then puts the individual in touch with one of its "trusted physicians" to decide whether the customer qualifies for the service.
"Our system is intended to be utilized within a preventative medicine framework, so most patients would qualify to receive testing for Serenity Genome," Bhakdi said. "For liquid, our physicians [tend to] recommend that individuals [of] average risk around 40-plus years start going through this testing annual as preventative care."
After receiving the physician's approval, Quantgene mails the customer a standard saliva collection kit and schedules either a mobile phlebotomy visit or an appointment at a physician's office to draw three 10-ml tubes of blood.
Customers fill out an intake form prior to testing to provide information on their family history, personal medical history, and lifestyle risks, Bhakdi said.
Samples are then sent to Quantgene, where Bhakdi said they are processed using the firm's proprietary methods at a third-party lab.
As of early May, Quantgene had noted on its website that it anticipated CLIA certification of its laboratory "in the next few months." However, Bhakdi later clarified that the firm's lab is currently CLIA-certified, but the company is building up a sequencing team there and is not yet running assays at the facility. In addition, partner Vikor Scientific does have a CLIA-certified, CAP-accredited laboratory, but Shea Harrelson, Vikor's cofounder and CEO, referred questions about the companies' partnership to Quantgene.
Quantgene then generates a report that one of its partnered physicians and genetic counselors discuss with the customer over a telehealth session. The duo establishes the customer's clinical situation, explains the genetic patterns identified by the panel, and advises the customer on downstream diagnostic tests if needed.
Quantgene initially launched Serenity in mid-February. Until earlier this month, a customer could purchase either a standalone test called "Serenity Genome" for $950, or a yearly screening service called "Serenity Complete" for $2,750.
However, the company's website now states that all "Serenity Complete Member spots" are currently full. Customers can still sign up for the service via a virtual waiting list, and Bhakdi said that the firm eventually aims to phase out the standalone Genome test while keeping the yearly service.
Bhakdi noted that Quantgene can return a result in roughly four weeks after the company receives a customer's samples, but he acknowledged that his team is currently facing certain "operational bottlenecks" due to some "initial growing pains" and increased test demand.
"We have tens of thousands of [potential customers], but we put a hold on selling it and are working on some operational things ... [in order] to improve turnaround time," Bhakdi said. "[We are trying] to make this operationally work … in terms of customer experience ... because the system we are building is more complex than other systems, because they integrate exome [and] liquid biopsy, but also genetic counseling, and also medical services."
Bhakdi emphasized that any decisions made on the customer's collected information is dependent on the specific clinical context of the data, rather than looking purely at the statistical results provided as part of the service.
A dearth of data
Harriet Feilotter, a professor of pathology and molecular medicine and cancer biomarker expert at Queen's University in Ontario, Canada, said in an email that Quantgene needs to understand the meaning of mutations it might find with the DeepGen assay. She pointed out that Quantgene must also identify extremely low levels of mutations that may be from early-stage tumors shedding DNA in the blood.
"We know that there are lots and lots of mutations that overlap, and many more that are completely random," said Feilotter, who also serves as an associate of diagnostic development at the Ontario Institute for Cancer Research. "[Quantgene] also needs to distinguish between errors (sequencing platforms make errors at rates higher than the rare mutation that they're going to be looking for) and true mutations."
Feilotter added that Quantgene needs to have completed a massive prospective study "to convince anyone that they can detect the molecules" that predict cancer in an individual who will ultimately develop the cancer that firm predicts with the Serenity service.
To train the DeepGen 272-biomarker panel, Quantgene initially launched its international feasibility study in 2015 to detect early-stage disease in a host of different cancers. Partnering with groups including the University of Geneva, University of Arizona Cancer Center and the Orlando Health UF Health Cancer Center, Quantgene plans to gather blood samples from about 10,000 patients as part of the ongoing trial, Bhakdi said. Although the trial was initially slated to be completed by December 2019, the firm has thus far collected roughly 5,500 samples and processed about 1,000 between the centers.
According to Setsuko Chambers, project collaborator and professor of obstetrics and gynecology at UACC, Quantgene initially approached her with a technology that it said "was highly sensitive in their data for early detection." Chambers said that she sent more than 50 blood samples — mainly from patients with ovarian and endometrial cancers — to Quantgene as part of the trial.
"I was attracted to things that were … related to [Quantgene's] particular technology, because I wouldn't say that I have the ability to distinguish that theirs was [significantly] better than the next, but the deployment data was good," Chambers said. "I do have a much-better-than-superficial understanding of their process, but I am not an expert in technicalities … of machine learning algorithms."
She also noted that Quantgene has not shared any specific results of the study so far, other than "exciting data that was confirmatory" in a colon cancer cohort.
"Quantgene has extensively analyzed seven types of cancer so far … but they did not include the gynecological cancers so far," Chambers added. "I assume … that when they're done analyzing the data and looking at the clinical translation components, that they'll share their data."
Quantgene also published a technical validation study in March in MDPI Genes on DeepGen's performance in reference blood samples. By processing the samples with validated mutations to establish the assay's performance and minimal input requirements, the firm found that DeepGen could distinguish between signal and noise down to a variant allele frequency of 0.09 percent and a limit of detection at 0.18 percent.
"We took several clinical [and contrived] samples … where you have a predetermined allele frequency that is validated on ddPCR," Bhakdi said. "You compare [the samples] with orthogonal methods, for example Roche's Avenio assay, to validate a bunch of variants, then go down with diluted samples that ddPCR confirmed, and show how long you can go with a sequencing panel."
However, Feilotter raised issues with the study regarding the relevance of detected mutations. "[Quantgene] likely [has] a nice assay with decent sensitivity, given the amount of sequencing they are doing … [but] that's all they have," Feilotter said. "This in no way tells us anything about what sequencing at that depth tells you about a person's health or risk."
Feilotter also argued that clinicians may not fully understand the assay's results, despite knowing an individual may carry a mutation, "even if it's real at … 0.18 percent."
"Just because you have a cell that has generated a KRAS mutation, for instance, does that mean you will end up with cancer or that the mutation came from a tumor cell?" Feilotter asked. "I'd like to see the proof for what burden of mutation is needed before there is a reasonable expectation that you have or will have cancer."
Quantgene also has a paper that is currently under peer-review covering results across seven cancers, which Bhakdi said covers a "large cohort" and expects to be published within the next two months.
Companion diagnostic plans
Bhakdi said that Quantgene also expects to offer a blood-based companion diagnostic test later this year out of its CLIA lab. The firm is also building labs in California and on the East Coast and hopes to eventually get CLIA certification later this year.
Bhakdi said that one lab will focus more on research and development and on early detection screening, while the other "will focus on reimbursement and clinical [genetics]."
"We have a whole range of FDA labels that we look at right now, and I think we will go after a couple of them in the next 12 months," Bhakdi said. "[The CDx product will] also be priced about the same amount as Serenity [Complete]."
As part of the envisioned CDx workflow, Quantgene has partnered with San Diego-based CureMatch to combine genomic, cloud, and artificial intelligence technologies with the intent of helping physicians identify optimal combination and monotherapies for their cancer patients.
"Serenity patients who are eventually diagnosed with cancer can be referred to CureMatch, where we can help use their liquid biopsy results and our platform to match that patient to treatments that are most likely to work for them," Navid Alipour, CEO of CureMatch, said in an email. He noted that the firms are also working on additional R&D efforts.
Quantgene has also launched a pilot study to demonstrate the use of the DeepGen panel as a companion diagnostic, which Bhakdi expects to finish in roughly two months. In addition to samples collected from Quantgene's "clinical infrastructure," Bhakdi added that the firm is using samples from customers — who he calls "patient innovators" — that have paid for the firm's early detection service. While Bhakdi declined to reveal how many customers have purchased the early detection solution, he claims that more than 90 percent have agreed to donate their deidentified genetic results to the study.
"[Our] companion diagnostic says you currently … [are diagnosed] with cancer going into adjuvant therapy, and you'd do a liquid biopsy to see [if] you have any informed FDA drug targets," Bhakdi said. "This is a system where building the delivery system is not trivial, because you need enough trained clinicians and genetic counselors, [and] all this needs to be coordinated."
Quantgene is a relatively new player in the early cancer detection space. Several other firms have developed sequencing-based pan-cancer liquid biopsy assays or companion diagnostic tests. These include Grail and its Galleri pan-cancer assay, Thrive Earlier Detection and its CancerSeek panel, as well as Delfi Diagnostics and its unnamed fragmentation-based assay.
Firms like Guardant Health and Roche's Foundation Medicine are also developing or have launched pan-cancer blood-based CDx assays. In addition, Invitae's ArcherDx markets Stratafide, while Personal Genome Diagnostics offers its PGDx Elio Plasma Resolve assays.
However, Bhakdi believes that Quantgene's offering distinguishes the firm from competitors because of its "clinical perspective."
"We have a much deeper connection with medical care, and our team like … Dr. Hagen, who has worked with other players in the clinical space," Bhakdi argued. He also claimed that the panel's ability to detect a mutation allele frequency of around 0.1 percent is advantageous.
Feilotter, however, pointed out that Quantgene has yet to publish any prospective clinical data demonstrating the assay's sensitivity and specificity.
Luis Diaz, head of the division of solid tumor oncology at Memorial Sloan Kettering Cancer Center, noted that commercial liquid biopsy assays often struggle with minimizing false positive results, "which is critical for the space." While he said he believes Quantgene's service is a step in the "right direction," he cannot "decipher what they're really doing" with its technology and had questions about the performance data that the firm has not publicly disclosed.
"The problem is that the product itself seems to be a little bit of a black box, at least based on the website," added Diaz, who also serves as chairperson of PGDx and Thrive. "In a novel area like next-generation molecular sequencing, the devil's in the details because you don't want to be delivering suboptimal information, nor do you want to [deliver] inaccurate information that's false."
Regulatory gray area?
In the past decade, other firms have attempted to offer DTC tests for early-stage cancer detection. Pathway OME, previously Pathway Genomics, at one point marketed its CancerIntercept liquid biopsy assay test, which purportedly detected tumor DNA in high-risk but otherwise healthy patients. In 2015, the FDA issued letters of concern to Pathway about the test, which is no longer sold. Pathway was acquired by one of its investors and rebranded OME Wellness in February 2020, and has since become OmeCare, a company focusing on at-home genetic testing for wellness and medication efficacy.
Cynvenio Biosystems offered a blood-based service to track disease status or monitor for early signs of recurrence in breast cancer patients. While Cynvenio noted in 2017 that it was not working under a DTC consumer model per se, the firm's CEO had said that it would talk to "patients and their doctors in that order." Doctors needed to order the testing service, but customers could express their interest to their physician or seek out an oncologist that would order the test. Cynvenio rebranded as LungLife AI in 2019 and has not issued any updates on its test since October 2020.
While Bhakdi's team had early discussions with the FDA, he noted that the company's physician partners and their use of Quantgene's diagnostics products constitutes a medical service, not a diagnostics device, and therefore has to adhere to American Medical Association (AMA) rather than FDA regulations.
"You can complain with the AMA, but the FDA is just not in charge of interfering with US medical practice," Bhakdi said. While Quantgene runs its assay on a sequencing instrument, he argued that "legally … it's up to the physician to decide if they want to run any kinds of tests."
Rather than allowing customers to directly purchase the early detection service, Bhakdi pointed out that the Quantgene puts the individual in contact with a partnering clinician, who then must decide whether the customer qualifies for the test.
"What these other companies like Pathway Genomics did is they basically sold you a test and gave you a test result," Bhakdi added.
However, Alberto Gutierrez, former director of the FDA's Office of In Vitro Diagnostics and Radiological Health, disagrees with the assertion that the FDA could not regulate Quantgene's service. While Gutierrez acknowledges that Serenity can technically fall into a loophole outside the FDA's jurisdiction, he argues that the method of initial sample collection may be dangerous for Quantgene.
"The only reason that the agency was able to keep DTCs [away] from LDTs is because … if [a company] is taking a saliva collection device, which everybody agrees … is a medical device regulated by the FDA, and if they're sending it [to patients] with their own claims, they are at the very least relabeling it," said Gutierrez, who currently serves as a partner with undisclosed firms in the cancer screening space. "If those claims are not part of what the device is typically cleared for, then the agency has a hook [on the firm]."
While Quantgene owns the DeepGen assay and provides the Serenity infrastructure and software, Bhakdi said that the firm technically licenses the service to physicians, including a group called "Pioneer Preventative Care Group." The roughly 10-member group consists of a physician, a gene counselor, physician assistants, and additional "medical specialists" that Quantgene has partnered with to operate the service.
Bhakdi said that that the group members come from medical backgrounds and are willing to undergo extensive genomics, software, and cancer detection training.
"Normally, these are primary care doctors who were very interested in preventative care and [are] unhappy with what's happening currently in primary care," Bhakdi said. "They undergo training, [where] we show them [internal] cases … and they get extensive case studies to … see how these different interpretations work … [since] the same score can mean different things based on the clinical context."
Bhakdi said that Quantgene pays the members market rates for consultations, which are baked into the price of the Serenity service. He also acknowledged that certain members of this team are part of Quantgene's advisory board.
"The patient basically pays the physician, and the physician then pays for the tests," Bhakdi added. "A single physician can actually have a big number of customers … so we are not dependent, like a normal diagnostics company or device company, [on having] tens of thousands of physicians who sell our stuff to patients."
Bhakdi expects to "ramp up" the number of clinical specialists it works with to meet the current demand for its testing services, with a potential goal of having 14 clinician partners by the end of 2021.
GenomeWeb reached out to seven cancer researchers and surgical oncologists who Quantgene had identified on its website as clinical or strategic advisors. Four declined to comment due to either the lack of publicly available data on the assay, because they were unaware the firm had named them as scientific advisors, or their professional role did not allow them to comment. The remainder did not respond.
One member who wished to remain anonymous noted in an email that they never formally agreed to serve as an advisor and had previously asked Quantgene to remove their name from the firm's site. Another researcher who also did not want to be identified said in an email that they had "considered a relationship" with the firm, "but eventually was just too busy to work with them at the time." They have also contacted the firm to remove their name from the website.
As of this story's publication, Quantgene no longer has the above-mentioned individuals on the list of its clinical advisory board. Bhakdi said in a follow-up email that the firm is "reviewing, adding, or removing members from time to time depending on our needs and changing of the advisors."
Gutierrez emphasized that while the FDA is busy with the COVID-19 pandemic, the agency "takes cancer screening very seriously," especially with firms like Quantgene that do not provide any clear prospective data to back their claims.
He said that the FDA has dealt with several companies attempting to offer pan-cancer screening, but that "no one has the data yet" to back up claims about early-stage cancer screening and detection.
"The FDA went after pharmacogenetics two years ago … and got slapped back," Gutierrez said. "That won't be the case when you have something like early cancer detection, because if you have no data, you're either reassuring people who shouldn't be reassured, or you're just sending people down diagnostic loopholes."
"[Quantgene has] clearly put a model together that they believe the FDA will have trouble saying it's under their jurisdiction, but I don't think the agency is worried," he added. "If I worked [at Quantgene], I would be careful, because they might end up in the agency's crosshairs.