At the 22nd Annual Conference of the National Comprehensive Cancer Network (NCCN), held March 23-25, 2017, in Orlando, Florida, Gregory J. Riely, MD, PhD, Memorial Sloan Kettering Cancer Center, spoke about the what, when, and how of biomarker testing in non-small cell lung cancer.
Biomarker testing is essential in lung cancer, Riely said, and should be done at diagnosis. “Even if it is not done at diagnosis, testing before the choice of second-line therapy is valuable as well.”
Riley showed the scanned image of the lungs of a woman who presented with an adenocarcinoma. Diagnosis of an adenocarcinoma may not be as bad as it was a decade ago, he said, adding, “We need to identify biomarkers for this, which, today, can start with molecular analysis and PD-L1 testing.”
Based on the results of the molecular analysis, the patient may receive:
Targeted therapy if positive for EGFR, ALK, or ROS1 mutations
Platinum-based chemotherapy if PD-L1 expression is <50%
Pembrolizumab if PD-L1 expression is >50%
“Current NCCN Guidelines for a patient who presents with metastatic disease recommend molecular testing to establish histologic subtype, smoking cessation counseling, and palliative care. For all the various histological subtypes, molecular testing is central,” Riley said. The key thing is to conduct these tests as part of a broad molecular testing profile. An important consideration is multiplexing to be able to maximize on the small biopsy sample.
Molecular testing for lung cancer has focused on DNA-based tests like sequencing and fluorescence in situ hybridization, known as FISH
. But over the last few years, protein tests have grown in usage, primarily immunohistochemistry.
Riley then shared phase 3 results from the KEYNOTE-024 trial that were presented at the annual meeting of the European Society for Medical Oncology last year and published in the New England Journal of Medicine
, which showed that pembrolizumab was effective in advanced lung cancer, compared with platinum-based chemotherapy, when PD-L1 was expressed in at least 50% of tumor cells.
“However, PD-L1 tests are all over the map, so how do you choose?” asked Riley. The fact that each PD-1 inhibitor introduced and approved has developed a complementary assay to test PD-L1 expression in the tumor samples makes it challenging to figure out how all these tests compare.
A recently published study in JAMA Oncology
compared 4 such PD-L1 assays, and found that while the tests were analytically interchangeable, they had not been cross validated. Further, only 3 of the 4 assays were concordant and reproducible.
An important point that Riley noted during his presentation was that PD-L1 expression is probably stable and there is no clear benefit to repeat a biopsy unless the prior sample is exhausted.
It’s also important, to customize biomarker testing needs based on the institution.
“Often, institutions use a combination of tactics to achieve comprehensive evaluation in a timely manner,” Riley said.