More Patients With Advanced Tumors May Benefit From Genomic Profiling Testing Than Previously Thought

Use of an optimized polymerase chain reaction–based comprehensive genomic profiling (PCR-CGP) test coupled with additional testing for samples not fitting inclusion criteria yielded meaningful results for more than 94% of samples, including 81% of samples that normally do not fit inclusion criteria, according to a recent study.

The investigators said that the study, published in JCO Precision Oncology, suggested that a greater number of patients may benefit from undergoing tissue CGP than previously thought.

“An optimized PCR-CGP test and testing of exception samples may increase the proportion of patients who can undergo tissue CGP to guide biomarker-based therapies,” wrote the investigators.

Next-generation sequencing (NGS) is used to assess relevant biomarkers from formalin-fixed paraffin-embedded tumor tissue or circulating cell-free DNA liquid biopsy samples. CMS has said that tissue-based CGP using NGS is medically necessary for patients with advanced solid tumors. However, CGP applicability can be challenging because many specimens are small and the nucleic acid yield may be considered low quality.

Additionally, CGP tests require information on tumor size, tumor content (TC), and nucleic acid and carry a failure rate between 30% and 50% in clinical trial cohorts, where testing is attempted on all samples. According to the investigators, this suggests “that current CGP approaches, which are largely based on hybrid capture library preparation, may only be applicable for a subset of specimens.”

The investigators analyzed samples collected between February 13, 2017, and June 25, 2020 for the Strata Trial, an observational study of 100,000 patients with advanced solid tumors. The samples were tested using StrataNGS, a PCR-CGP test that is covered for Medicare beneficiaries.

Tests with at least 1 reported prioritized alteration or that met TC of 20% or greater and sequencing quality control assessments were considered successful. Tests for prostate carcinoma and non–small cell lung cancer (NSCLC) adenocarcinoma that reported at least 1 actionable or informative alteration or met TC of 20% or greater and sequencing quality control assessments were deemed actionable.

Across 28 US health systems, 31,165 consecutive solid-tumor samples collected from 30,565 patients were tested. Overall, 10.7% of the samples had a final TC of less than 20%, which is a minimum TC requirement for many CGP tests. Most (57.2%) of the samples were from biopsies.

Additionally, 59.2% of the samples were considered small, having a tumor surface area (TSA) of less than 25 mm². TSA of 25 mm² is the minimum TSA required for many leading commercial hybrid capture CGP tests, including the only FDA-approved tissue CGP companion diagnostic device (FoundationOne CDx).

Among the 31,101 samples that were evaluable for input requirements, 8089 (26.0%) did not meet the required TSA for many CGP tests. However, 94.2% of the evaluable samples were successfully reported after CGP testing, including 80.5% of the samples that did not meet CGP testing requirements and were tested as exceptions.

The positive predictive value of the PCR-CGP test for ERBB2 amplification in exception samples and/or sequencing breast cancer samples that failed quality control analysis was 96.7%.

For prostate cancer, 84.0% of the tested samples produced results that could inform treatment selection. Similarly, 87.9% of the NSCLC adenocarcinoma samples that were tested yielded information that could aid providers in selecting treatments.

The investigators listed several study limitations including that the impact of their clinical approach is unclear and that there was a lack of head-to-head testing with hybrid capture–based tissue tests and liquid biopsy testing. They said that comparison of testing methods should be investigated in future studies. Also, patients with noninformative PCR-CGP testing were not followed to determine whether rebiopsy or liquid biopsy was pursued or had an impact.

Reference

Tomlins SA, Hovelson DH, Suga JM, et al. Real-world performance of a comprehensive genomic profiling test optimized for small tumor samples. JCO Precis Oncol. 2021;5:1312-1324. doi:10.1200/PO.20.00472