Caris Life Sciences has added next-generation sequencing on the Illumina MiSeq to the array of molecular tumor profiling technologies it uses to help guide cancer therapy, including fluorescence in situ hybridization, PCR, and immunohistochemistry.
For its next-gen sequencing offering, the firm has validated Illumina's TruSeq targeted cancer panel on the MiSeq, which focuses on 44 genes related to cancer, but it plans to also develop its own customized panel with "the key clinically useful mutations and some additional ones that will be relevant for research purposes," Caris' senior vice president and group head of oncology, Tom Spalding, told Clinical Sequencing News.
The next-gen sequencing offering will replace an eight-gene Sanger-based test that it currently uses.
Spalding said that what differentiates Caris' offering from other companies and academic institutions offering similar services is both the comprehensiveness of the services offered and the focus on the end report and the impact to patient care.
"We are drug centered, and start with the drugs in mind," he said. "So we work backwards from the drug and ask what biomarkers are relevant for those drugs and what technologies and platforms support [the identification of] those biomarkers."
The tests are run out of the company's CLIA-certified laboratory in Phoenix, Ariz., which houses three MiSeq instruments. Average turnaround time is 14 days. The company's annual test run rate is currently 15,000 and Spalding expects that to increase significantly. Caris does not disclose pricing for its tests, of which it offers several versions.
All patients' tumors will be sequenced, and depending on the specific cancer, or physician's request, the patient will also receive one or multiple other tests to search for therapy-related biomarkers, unless the physician requests only the next-gen sequencing portion. Physicians can then choose a comprehensive report or a select report, which will only report on a select number of mutations.
The report that is delivered back to the physician, which Caris has dubbed Molecular Intelligence Profile, or MI Profile, contains three different categories of information: drugs that will potentially provide benefit to the patient, drugs that are likely to be ineffective, and clinical trials for which the patient may be eligible.
Between the three categories, Caris identifies an average of 25 reportable biomarkers per patient.
The list of drugs that are likely to be beneficial is then further divided into on-label and off-label drugs. "We think there's a lot of utility and value in bifurcating between the on- and off-label drugs," Spalding said.
The rest of the report then contains all the evidence surrounding the drug list, including the specific mutation, what technology was used to identify the mutation, and the literature that supports the drug match.
"Everything we report out on, we reference the literature that supports that there is a correlation between that biomarker and that drug," Spalding said. "In addition to saying, 'This is a drug that might work or might not work,' physicians want to know the basis for how we came to that."
Caris will likely compete with Foundation Medicine's FoundationOne test, which assesses 182 genes using Illumina's HiSeq 2000. However, Spalding said that even though Foundation's test looks at more genes, Caris' test uses multiple technologies and can identify biomarkers that cannot be assessed with sequencing alone and therefore could identify more actionable results. Foundation reported last year that its test identifies an average of 1.3 actionable results per patient, while Spalding said Caris averages 25 (CSN 6/13/2012).
Additionally, a number of academic institutions have begun offering tumor profiling services, including Johns Hopkins University, which has spun out Personal Genome Diagnostics to commercialize cancer genome analysis tests (CSN 5/15/2012); and the Fox Chase Cancer Institute, which recently launched a test called CancerCode-45 that analyzes tumors using Ion Torrent's AmpliSeq Cancer Panel (CSN 1/16/2013).
Massachusetts General Hospital and MD Anderson are also both implementing sequencing-based tumor profiling assays and the University of Michigan has a program in place, dubbed Mi-OncoSeq, to do whole-exome and transcriptome sequencing of cancer patients (CSN 9/19/2012, CSN 10/10/2012, CSN 12/19/2012).
Aside from having a variety of technology offerings, Spalding said another differentiator between Caris and its competitors is that Caris' staff continuously reviews the literature surrounding the various biomarkers the company reports on, and adjusts the software responsible for the analysis. Additionally, the company also has pathologists and molecular geneticists on its staff.
Caris bills insurance companies for the tests, based on a "marker-by-marker, procedure-by-procedure basis using appropriate CPT codes," Spalding said. The overall price per patient varies greatly because it depends on the number of tests ordered by the physician. Additionally, while the lab's services are "generally reimbursed by the major payors," the "actual reimbursement amounts received varies by payor and may depend on specific factors, including type of payor, existing contractual arrangements, etc.," he added.
The company also maintains a registry of all the biomarkers it analyzes. Spalding said that moving forward, one main function of the next-gen sequencing component will be to help populate the registry and hopefully identify additional mutations with clinical actionability. Additionally, the company will add content to its sequencing panel based on current clinical relevance as well as mutations that could be beneficial for research purposes.
"We want to continue to make sure that our registry is populated with potential emerging markers," Spalding said.
In terms of moving into exome or whole-genome sequencing, Spalding said that in the near term, the company would stick with a more targeted approach. He said that Caris would consider a more comprehensive approach "if and when the science and cost allow that to be feasible," but he doesn't believe whole-genome sequencing has reached that point yet.