FDA Grants Breakthrough Device Designation to HLA-LOH Companion Diagnostic Test

Author(s):

The FDA has granted a breakthrough device designation to HLA-LOH as a companion diagnostic test, according to an announcement from Tempus.

FDA

The FDA has granted a breakthrough device designation to HLA-LOH as a companion diagnostic test, according to an announcement from Tempus.¹

The test features a machine-learning model designed to analyze data from the next-generation sequencing (NGS)–based xT CDx assay, which is a 648-gene NGS assay for solid tumor profiling and a companion diagnostic for patients with colorectal cancer that received FDA approval in May 2023.²

The assay is designed to identify patients with solid tumors who may benefit from specific targeted therapies if their tumors underwent allele-specific loss of heterozygosity (LOH) for specific HLA class I alleles.¹

“HLA-LOH provides a clear molecular distinction between cancer and non-cancer cells and is a potential biomarker for immune therapy resistance. The Tempus HLA-LOH test is intended to measure this biomarker and better understand which patients may respond to new therapies. This breakthrough device designation from the FDA recognizes the novelty and potential clinical impact of our HLA-LOH test for this promising biomarker,” Kate Sasser, PhD, chief scientific officer of Tempus, stated in a news release.

“HLA-LOH is of special interest for the application of cell therapy to treat solid tumors, but also has broader potential for other precision medicine approaches in oncology, including in combination with other established biomarkers,” Sasser added. “The Tempus test is being developed to identify HLA-LOH and may help optimize existing therapies and facilitate the advancement and implementation of novel and transformative treatments.”

xT CDx is a qualitative NGS-based in vitro diagnostic device designed to identify substitutions such as single nucleotide and multinucleotide variants, as well as insertion and deletion alterations in 648 genes. Additionally, it is capable of detecting microsatellite instability status through DNA isolated from formalin-fixed paraffin embedded tumor tissue specimens and DNA isolated from matched normal blood or saliva from patients with previously diagnosed solid malignant neoplasms.²

For patients with CRC, xT CDx can be utilized as a companion diagnostic to identify those with KRAS wild-type disease (absence of mutations in codons 12 or 13) who could benefit from treatment with cetuximab (Erbitux), or those who could be eligible for treatment with panitumumab (Vectibix) through the detection of KRAS wild-type (absence of mutations in exons 2, 3, and 4) and NRAS wild-type (absence of mutations in exons 2, 3, or 4) disease.³

The detection of alterations by xT CDx was compared with data from an externally validated orthogonal method (OM), with 114 overlapping genes between the 2 assays. Samples of 416 patients across 31 different tumor types were used to compare the detections of mutations, insertions, and deletions between the 2 methods.

The samples were derived from patients with CRC (n = 69), breast cancer (n = 44), ovarian cancer (n = 38), glioblastoma (n = 34), non–small cell lung cancer (n = 29), endometrial cancer (n = 26), clear cell renal cell carcinoma (n = 22), bladder cancer (n = 18), melanoma (n = 17), pancreatic cancer (n = 14), thyroid cancer (n = 12), low-grade glioma (n = 12), sarcoma (n = 10), tumor of unknown origin (n = 8), meningioma (n = 7), prostate cancer (n = 7), gastrointestinal stromal tumor (n = 7), endocrine tumor (n = 6), gastric cancer (n = 5), head and neck squamous cell carcinoma (n = 4), kidney cancer (n = 3), brain cancer (n = 3), small cell lung cancer (n = 3), biliary cancer (n = 3), cervical cancer (n = 3), esophageal cancer (n = 3), oropharyngeal cancer (n = 2), liver cancer (n = 2), head and neck cancer (n = 2), mesothelioma (n = 2), and adrenal cancer (n = 1).

Concordance in hotspot and non-hotspot regions were identified, and positive and negative percent agreements (PPA, NPA) for each type of variant between the 2 assays were used to assess the accuracy of xT CDx. Findings showed that 148 variants reported as somatic through the OM were detected as germline by xT CDx.

A hotspot concordance analysis with the OM examined variants reported in hotspot regions that overlapped with OM targeted regions. Of the 416 samples, 164 featured at least 1 reported variant in an overlapping hotspot region, and 214 base pairs intersected at the defined hotspot regions of both the xT CDx and OM targeted regions.

Investigators assessed 192 reported variants in hotspots from both assays, including 187 substitutions across 10 genes and 5 INDELs spanning 4 genes. They calculated the PPA and NPA for substitutions and INDELS in hotspot regions through the total variant counts of each classification across tumor samples to assess the accuracy of xT CDx.

Among the 69 CRC samples tested with the OM, the detection of specific KRAS and NRAS CDx variants was examined. Among 31 CDx variants identified through OM, 31 were also detected by xT CDx, resulting in a PPA of 100% (95% CI, 88.8%-100.0%). Additionally, 649 variants were negative by the OM, and 648 were negative by xT CDx, leading to a NPA of 99.8% (95% CI, 99.1%-100.0%).