On August 7, 2024, Guardant announces publication of its REVEAL test as a colorectal cancer recurrence assay (MRD assay). It's Nakamura et al. at Clinical Cancer Research.
Guardant assessed 1977 samples from 342 patients. High quality MRD surveillance assays require surveilling a large population at a reasonable frequence from many quaters, to asses the performance of the molecular test versus regular imaging (or clinical) surveillance. Large numbers give more precise performance data with more narrow error bars.
Detection of stage II or higher cancers (in a recurrence setting) was 81%, and as high as 100% for detection of new liver mets. Specificity was 98%.
Lead time over clinical or imaging (legacy detection methods) was 5.3 months.
Their webcast today is archived at Guardant's investor events website. They note (slide 10) that data has been submtted for ADLT status by 2025. The new data has also been submitted to Medicare / MolDx for surveillance reimbursement as an MRD test (page 16). Guardant also discussed (their slide 14) a "major upgrade" to the LDT verion of G360, now with 739 genes, See "10X improvement to detect disease burden," and genotype/phenotype information (noting histology, subtype, and prognosis.) (See Q&A about MolDx and ADLT timelines around minute 29 of the online recording, and Q&A specifically on SHIELD pricing ($1495) at minute 36. Discussion of MRD PCR vs NGS methods, around minute 42). (Informal auto transcript here; look for official validated transcripts to rely on).
https://doi.org/10.1158/1078-0432.CCR-24-1651
##
Note that an FDA-approved Guardant test called SHIELD for CRC screening (in health populations) was recently approved.
##
The Guardant REVEAL assay is tissue naive, e.g. does not require a prior exome test of the tumor, and mesures 20,000 epigenomic loci in ctDNA.
