Clinical Utility of an In-House Targeted Sequencing Panel Compared to Commercial Comprehensive Genomic Profiling for Solid Tumors
AMP 2025
T. Wang, K. Shenoy, P. Hernandez, E. Duncavage, A. Hughes, L. Sun
Introduction
Somatic tumor profiling through next-generation
sequencing (NGS) has become common practice for prognostication and
therapy selection. Comprehensive genomic profiling (CGP) is an approach
that assesses hundreds of genes and mutational signatures, and may
include both DNA and RNA sequencing. In contrast, targeted NGS panels
often evaluate a subset of these targets. We hypothesized that for some
tumor types, a targeted panel is sufficient to make biomarker-informed
therapy decisions. Methods: We validated an in-house, DNA-only,
targeted NGS panel, GatewaySeq, that includes less than 50 relevant
genes per tumor type. We retrieved retrospective cases from 2023 to
2025 that had both GatewaySeq as well as a CGP assay (Tempus xT or
Caris MiProfile) ordered, and compared concordance of variants detected.
In addition, we evaluated the electronic medical record to determine
whether sequencing results from each assay affected management
decisions. Results: Thirty-three cases were assessed (17 non-small cell
lung cancers, 7 breast carcinomas, 4 endometrial cancers, 3
gastrointestinal stromal tumors, 1 ovarian cancer, and 1 cancer of
unknown primary). Altogether, the CGP assays reported 127 biologically
or clinically significant variants, whereas GatewaySeq reported 91
significant variants, with a concordance rate of 76/142 (53.5%). Fifty
(35.2%) single nucleotide variants (SNVs) and copy number alterations
(CNAs) were reported as significant by CGP, although their corresponding
genes were not included in the GatewaySeq panel; none of these were
variants associated with eligibility for or resistance to approved targetable
therapies. One SNV and 4 CNAs were reported by GatewaySeq but not
by CGP. Only 1 variant, a PTEN loss, was included as a validated target
in both assays but was not detected by GatewaySeq. Six cases were
quantity not sufficient (QNS) by CGP but sufficient by GatewaySeq, even
though 2 of the cases had GatewaySeq ordered on the same block after
CGP was attempted. GatewaySeq reported variants associated with
tumor-specific approved targeted therapies for 4 of the CGP-QNS cases.
In total, 16/33 (48.5%) cases had a variant with tumor-specific therapeutic
implications; in 9 (27.2%) of these cases the NGS result informed the
patient’s subsequent treatment. There were no changes in management
based on the additional variant and mutational signature information from
the CGPs that was not present in GatewaySeq. Conclusions: A small,
targeted NGS panel is sufficient for therapeutic decision-making in lung
cancer, breast cancer, and gastrointestinal stromal tumors. A tiered
approach to somatic tumor profiling could begin with testing by a targeted
panel, followed by CGP if needed for clinical trial eligibility.
Related publications