Molecular Profiling Nuances in ROS1-Positive Non-Small Cell Lung Cancer: PD-L1, TMB, and Testing Workflow

The panel addresses scenarios where patients arrive on treatments other than the panel's preferred agent. Dr. Rodriguez notes she has seen patients who received immunotherapy based on high PD-L1 expression (including cases with TPS of 95%) before ROS1 testing was complete, describing this as a missed opportunity that may harm patients and limit future options. The pathologist clarifies that patients with oncogenic driver mutations typically do not respond well to immunotherapy, likely related to lower tumor mutational burden (TMB) and the biology of tumors arising in never-smokers without a smoking mutagenesis signature.

Dr. Riedlinger explains why high PD-L1 can co-occur with driver fusions: different mechanisms regulate PD-L1 upregulation, and the tumor microenvironment in ROS1-positive (and other fusion-positive) cancers tends to have low TMB despite elevated PD-L1 expression, creating a scenario where immune checkpoint inhibitor monotherapy is ineffective despite a seemingly favorable biomarker. Additionally, blood-based TMB calculation artifacts may produce spuriously elevated results that do not correlate with tissue TMB, particularly at the 10 mutations per megabase threshold; cutoffs of 16 to 20 mutations per megabase have been proposed for blood-based TMB but are not yet harmonized.

The panel uses a restaurant analogy: PD-L1 is the breadbasket, which arrives quickly but shouldn't be confused for the main course; NGS is the steak, the information clinicians and patients truly need. Some laboratories withhold PD-L1 results until the full molecular panel is complete, whereas others release it immediately, a practice the panel endorses to prevent premature immunotherapy initiation.