Expert perspectives on the significance and limitations of PD-L1 testing in non-small cell lung cancer and the importance of integrating genomic characterization for informed treatment choices.
62-year-old woman presented to the ED with vague complaints of voice changes and cough.
Past Medical, Family, and Social History
Final Pathology: consistent with squamous cell carcinoma; metastatic stage IV
PD-L1 expression by IHC: 0%
NGS: No actionable mutations
Sandip P. Patel, MD: PD-L1 testing, which at the current time is done on tissue, is a key biomarker we look at in non–small cell lung cancer, both in the metastatic setting [and] increasingly in the adjuvant postsurgical setting as well. Typically, a PD-L1 test is most commonly a DAKO 22C3, which is the companion diagnostic for pembrolizumab, or VENTANA SP263, which has been a companion diagnostic for multiple anti–PD-1 and PD-L1 antibodies. Either [is] reasonable, and there’s a reasonably high incidence of concordance looking at PD-L1 expression on tumor membrane on these tissue biopsies. For those patients with PD-L1 scores greater than 50%, the opportunity for immuno-monotherapy is reasonable, meaning anti–PD-1 or anti–PD-L1 directed approaches without chemotherapy are a potential option for a subgroup of those patients.
However, some limitations are that PD-L1 in and of itself as an expression biomarker only predicts response about 47% of the time. In fact, many times patients can have high PD-L1 and in fact harbor a driver mutation for which their best therapy would actually be targeted therapy. An example would be a patient with an EGFR mutation exon 19 deletion but a PD-L1 score of 80%. The PD-L1 score will come back sooner because it’s immunohistochemistry. The next-generation sequencing can take anywhere from 2 to 3 weeks. Acting on the PD-L1 in isolation can make you want to try to treat a patient with anti–PD-1–directed monotherapy. But in fact, their best treatment is targeted therapy, which is a pill. If you try to transition from an immunotherapy to an EGFR inhibitor, there’s a risk of potentially fatal pneumonitis. And so, PD-L1 remains an imperfect biomarker. But I think it’s informative, especially for those patients who are trying to avoid chemotherapy, especially if their PD-L1 score is greater than 50%. But I think it’s important to emphasize that we should not act on PD-L1 in isolation. We also need to have the patient’s genomic characterization before we treat a patient with immunotherapy.
At the current time, the decision between single-agent immunotherapy, typically an anti–PD-1 or anti–PD-L1 agent, vs dual immunotherapy, meaning adding an anti–CTLA-4 targeting immunotherapy to an anti–PD-1 or PD-L1 backbone, is really driven by the potential net benefit of anti–CTLA-4. While we don’t have prospective randomized control studies showing a benefit in a subgroup of patients, across multiple trials, there’s been a trend that patients who receive anti–CTLA-4–based dual immunotherapy regimens who have high-risk features such as PD-L1–negative tumors, squamous histology, brain metastases, or STK11/KEAP1 mutations may preferentially benefit from introducing that anti–CTLA-4–blocking agent in addition to anti–PD-1 agent many times in combination with chemotherapy. Those are the 4 subgroups of patients I tend to consider using anti–CTLA-4 blockade with. Those patients [who] are PD-L1 negative, squamous histology, brain metastases, or STK11/KEAP1 comutation.
Transcript is AI-generated and edited for readability.