NEW YORK – The Multiple Myeloma Research Foundation (MMRF) has launched the online CureCloud Initiative, a project that will use results from a novel liquid biopsy assay to analyze mutations in the multiple myeloma patient population.
The foundation is working with academic and commercial groups for the project to aggregate patient health data, including patient-specific mutations and drug responses, to help oncologists determine the best course of action for targeted treatment.
In multiple myeloma, a patient's plasma cells divide uncontrollably in the bone marrow. They eventually spread from the marrow onto a patient's bone, forming tumors and producing an abnormal antibody called monoclonal immunoglobulin. The American Cancer Society estimates that about 32,270 new cases will be diagnosed in 2020, and about 12,830 deaths will occur over the year.
Daniel Auclair, CSO of the MMRF, explained that while prior nationwide genome sequencing initiatives like The Cancer Genome Atlas characterized the genome of several different cancer types, they neglected to include multiple myeloma. His team therefore launched initiatives to compensate for the lack of multiple myeloma genomic research and has so far identified over a dozen genomic subtypes of the disease, each with their own mutational drivers, using data from over 2,000 patients.
While CureCloud was initially announced last summer for a group of small pilot studies, Michael Andreini, chief operating officer of the MMRF, said that his team has now fully launched the CureCloud program as a direct-to-patient effort. Starting in August, researchers will begin sequencing submitted patient blood samples at the Broad Institute using a newly-developed multiple myeloma hybrid selection assay.
The MMRF has partnered with several commercial and academic groups as part of CureCloud's launch in addition to the Broad Institute, including the Dana-Farber Cancer Institute, My Gene Counsel, COTA, EMSI Health, Prometheus Research.
Auclair noted that both clinical and commercial groups have developed bone marrow aspirate (BMA)-specific sequencing panels. The Mayo Clinic has developed its Oncospire Genomics assay, while the University of Michigan offers its Oncoseq platform. On the commercial side, he said that Foundation Medicine's FoundationOne Heme and Tempus' xT panel are the most ordered tests for BMA testing and cover between 400 and 500 genes.
However, BMA extraction is often invasive and can lead to health risks such as bacterial infection.
Auclair emphasized that the new assay — developed as part of a joint effort by the MMRF, research teams led by Irene Ghobrial and Keith Ligon at the DFCI, and the Broad Institute's Genomic Platform — is the first liquid biopsy panel to specifically target multiple myeloma. While other groups, such as Foundation Medicine, Tempus, and Memorial Sloan Kettering Cancer Center offer liquid biopsy tests that include multiple myeloma, he argued that the assays lack many of the relevant genes in myeloma.
"Myeloma is the second most prevalent blood cancer, but represents less than 2 percent of all cancers," Auclair noted. "It is difficult to convince diagnostic test developers, especially commercial ones, to advance a CLIA test for that specific small cancer patient population."
The research group's assay analyzes up to 70 commonly altered genes in a patient's blood sample on an Illumina next-generation sequencing (NGS) instrument. The assay performs deep sequencing to capture somatic point mutations and indels present in a patient's cell-free circulating tumor DNA (ctDNA) and circulating tumor cells.
"From CoMMpass, we identified a number of genes that are most frequently altered in multiple myeloma, [and therefore] built a targeted panel using UMI-based duplex technology to assess the mutations," Auclair said. "The probes on the assay also represent Clonal Hematopoiesis of Indeterminate Provenance (CHIP), which we believe are predictive of blood cancers and cardiovascular disease."
Auclair noted that the 70-gene assay has a sensitivity greater than 90 percent for events at a 1 percent variant allele frequency, with a false positive rate "well below 1 per Mb."
"The idea is [to make] this a small panel of relevant genes, allowing us to get deep coverage that allows the strength of the assay," Auclair said.
The academic collaborators presented the initial results of an internal validation study at an American Association of Cancer Research Liquid Biopsy meeting in January, using the assay to compare blood samples containing ctDNA and corresponding BMA samples from patients. Importantly, the researchers found that they required about 50,000-fold coverage before reaching a plateau and having sufficient raw sequencing depth to maximize sensitivity for duplex sequencing.
Noting that additional results should be available at the American Society of Hematology Annual meeting in December, Auclair anticipates that the study will be published in early 2021.
CureCloud workflow
Multiple myeloma patients interested in participating in the CureCloud initiative can fill out a consent form online. The MMRF-led team then aggregates and curates the patient's EHR information that the individual provides by naming health institutions who have provided care for their multiple myeloma.
The group then orders a blood draw from the patient's treating physician, who must sign off on the order. It then sends out a mobile phlebotomist to a patient's home to collect a 10 ml blood sample, which is mailed as a kit to the Broad Institute for genomic and immune profiling.
When the kit arrives at the Broad's CLIA-certified, CAP-accredited lab, researchers begin by extracting ctDNA from the blood using magnetic bead-based chemistry. The team then performs NGS to identify patient-specific mutations. Auclair said that the steps performed at the Broad require around 10 days to produce primary sequence data.
Broad researchers then transfer the raw sequencing results to the DFCI's molecular pathology lab, where Ghobrial and Ligon's teams perform final mutational calls and complete a patient's diagnostic report.
Andreini explained that the researchers generate two reports, one for the patient and one for the treating physician. They work with digital health company My Gene Counsel to generate a patient-friendly report, which identifies any actionable patient-specific mutations identified through sequencing.
While the overall workflow currently requires about four to six weeks to incorporate patient data into the CureCloud database, Auclair's team hopes to shorten the process to two to three weeks.
Patients will have access to an online portal as soon as they consent to the research project, which can help them track where they are in the research process from a logistics perspective. A dashboard summarizes key information from both genome sequencing and electronic health records, including data on a patient's cytogenetics, gene expression profiling, and a multiple myeloma prognosis staging system.
The patient's dashboard also contains treatment-related information, such as which therapies the patient has been exposed to in their clinical journey. The patient can view potential outcome information — such as their progression-free survival and overall survival — as well as adverse events.
Andreini said that the research team behind CureCloud is also fleshing out out two "core visualizations" for the online portal. The group expects to complete these visuals within the next few weeks.
One interactive visual will show a "more holistic view" of a patient's journey to help visualize various biomarkers that have changed over the course of the disease, which Andreini noted will be based on how and when they've been exposed to and responded to treatment. The second visualization will serve as a larger data exploratory tool for multiple myeloma patients, where they can examine what clinical journeys other patients with potentially similar characteristics have undergone over time.
"We also plan to build out the physician-facing portal, where it'll be a similar visualization tool and potentially help facilitate the patient-doctor interaction," Andreini said. "For example, [they could] evaluate a specific set of characteristics in a patient, spot what other patients were like this one, what were their outcomes, and hopefully make better treatment decisions for their specific patient."
While the initial pilot cohorts collected information from a little fewer than 200 patients, Andreini's team expects to now collect sequencing and EHR data from about 5,000 multiple myeloma patients as part of the CureCloud Initiative. He highlighted that 100 new patients have enrolled in the study since it was officially launched earlier this month.
"The test might [also] help determine if you need to remove the tumor before it progresses further," Auclair added. "Even though we've had 12 new drugs [approved] in the last decade, this is still an incurable disease, and we have patients running out of options."
Challenges
However, Andreini acknowledged that the collaborators encountered several challenges while launching the CureCloud initiative. He pointed out that the team had to navigate regulatory and legal considerations while delivering CLIA-grade testing kits across the US.
"Given the breadth of information we're pooling together, and our effort to lower the number of barriers to allow patients to participate, we had to work with partners on the project to cover all the aspects of this initiative," Andreini said.
Auclair and his colleagues also dealt with multiple issues while developing the 70-gene panel. Initially selecting a panel of 400 genes, the group was not thrilled with the sequencing coverage that the large number of genes provided in order to establish the necessary sensitivity. The researchers therefore shifted to a smaller multiple myeloma-specific 70-gene panel, allowing them to start optimizing for a CLIA-certified test.
Auclair pointed out that his team also needed to identify the "perfect" samples to validate the assay. This required both enough ctDNA for multiple rounds of CLIA-validation testing for sensitivity and specificity, as well as corresponding marrows biopsies that had been CLIA-sequenced on orthogonal platforms.
"We [wanted to be] sure that the information we got from the blood represented with very high fidelity the genomic information in the primary tumor, the myeloma cells in the marrow," Auclair said. However, "assembling that collection of samples took some time."
While CureCloud will focus on applying the 70-gene assay to identify patient-specific mutations in a blood sample, Auclair believes his team of researchers could potentially apply the CureCloud platform and assay for other multiple myeloma applications, such as helping to prevent the transition from "smoldering" multiple myeloma to the actual cancer.