This story was originally published May 18.
Adaptive Biotechnologies, a Fred Hutchinson Cancer Research Center spinout that began offering immune sequencing services in 2010, is aiming by the end of the year to become CLIA certified and to launch a clinical assay to detect residual disease in leukemia patients.
The Seattle-based company, which changed its name from Adaptive TCR at the beginning of the year, is developing a sequencing-based assay to detect minimal residual disease in T-lineage acute lymphoblastic leukemia to help physicians predict a patient's risk of relapse.
Researchers from the company, the University of Washington, and the Fred Hutchinson Cancer Research Center recently detailed the assay in a paper published in Science Translational Medicine, and showed that it was more sensitive than the current standard of care, which is based on flow cytometry, at detecting residual disease in patients post-treatment.
The minimal residual disease, or MRD, assay will be the company's first clinical assay, followed by an expansion of the test to screen for residual disease in B-cell acute lymphoblastic leukemia. Adaptive Biotech is also looking to develop an assay to measure immune reconstitution in transplant patients, and a prognostic test for ovarian and colon cancer patients based on tumor infiltrating lymphocyte count.
Aside from the oncology field, the company also plans to develop tests for hematology, autoimmune disease, and infectious disease.
Currently, the company offers a research-use only assay called ImmunoSeq, which targets key adaptive immune system genes for sequencing and includes bioinformatics software. It markets ImmunoSeq to academic institutions, as well as to biotechnology and pharmaceutical firms that use it for patient stratification and predicting drug response or outcomes.
So far, the company's internal sequencing work is done primarily on the Illumina HiSeq 2000 instrument, but it is currently evaluating both the MiSeq and the Ion Torrent PGM for the clinical test, and it will consider the HiSeq 2500 when it becomes available.
"We're developing the assay to be platform agnostic," CEO Chad Robins told Clinical Sequencing News. He added, however, that "in terms of turnaround time that's necessary to be clinically relevant," the test would need to be on one of the quicker platforms like the MiSeq, PGM, or the HiSeq 2500.
Adaptive Biotech has not yet nailed down a price for the MRD assay, but Robins said it would be comparable to the cost of flow cytometry-based tests.
In the Science Translational Medicine study, the researchers demonstrated the MRD assay on 43 patients with T-ALL and compared it to multiparametric flow cytometry, or mpFC, the standard of care in the US.
The MRD assay uses next-gen sequencing to assess immune cell diversity. In healthy individuals, no particular T-cell clone would be present at greater than a couple percent frequency, while in leukemia patients, a single clone could comprise up to 98 percent of all T-cells, explained Harlan Robins, senior author of the study and associate member in the Public Health Sciences Division at the Hutch as well as a co-founder of Adaptive Biotechnologies and brother of Chad Robins.
The specific clone will vary from patient to patient, he said, but identifying the clone in any patient prior to treatment is relatively straightforward. That T-cell clone then acts as a "genomic tag, so we can track and follow the cancer," he said.
However, because the goal of treatment is to kill the cancer, the clone is greatly reduced following treatment, making residual disease harder to identify among the millions of other T-cell clones.
Identifying the leukemic clone is important for predicting relapse, and previous studies have shown that having a leukemic T-cell clone in 1 out of 100,000 cells is predictive of relapse.
From the 43 samples, which were collected from the University of Washington as part of a pediatric oncology clinical trial, the researchers used their ImmunoSeq assay and HiSeq 2000 to target and sequence the TCRB and TCRG immune genes from 67,000 cells, or 400 nanograms of DNA, from each pre-treatment sample; and 200,000 cells, or 1,200 nanograms of DNA, in each sample 29 days post-treatment.
In 31 cases, they identified a clonal TCRB sequence in the pre-treatment sample, and in an additional four cases, they identified a clonal TCRG sequence. In 12 cases, neither sequencing nor mpFC was able to detect a clonal marker in the pre-treatment sample.
Following treatment, the team re-analyzed those 35 samples with both sequencing and mpFC. In nine cases, neither method identified residual disease. In 12 cases, both methods identified disease, and in 10 cases sequencing identified disease, but mpFC did not.
In the cases where only the sequencing assay was able to detect disease, disease prevalence was 10-fold to 100-fold lower than in the cases that both methods could detect. The sequencing assay was able to detect residual disease down to a frequency of 1 cell in 1,000,000.
While the sequencing method demonstrated increased sensitivity for detecting disease, the clinical implications of this are still unknown. The patients will have to be followed to see if these lower levels of residual disease impact relapse risk.
Harlan Robins said that the limit of detection for sequencing is a function of input material and sequencing coverage, and that, theoretically, the assay could detect one T-cell clone in five million. However, he said, "there's no reason to do that. It requires quite a bit of DNA, and becomes quite expensive."
Additionally, he said, "there's mounting evidence that at the level of one cell in 100,000 … the risk for relapse is way higher." At levels below that, though, "there's no clinical information to determine whether that has value," since no other method has been able to look at that resolution.
The next step to moving this test to the clinic is to demonstrate it on a larger patient population and also to follow patients for a longer period of time to demonstrate a correlation between residual disease detection and survival, said Chad Robins. He added that he expects to have results from these studies in about four to six months.
Additionally, the team is expanding the test to B-cell acute lymphoblastic leukemia.