Immunotherapy shows activity against brain metastases among certain patients with NSCLC
Approximately one-quarter of patients with advanced lung cancer develop brain metastases.
The standard treatment for these patients — radiation — can sometimes result in significant toxicity.
Immunotherapy with checkpoint inhibitors has shown efficacy in various cancers, including lung cancer. However, the activity of immunotherapy in the brain had not been studied extensively.
Sarah B. Goldberg, MD, MPH, associate professor of medicine at Yale Cancer Center, and colleagues conducted an open-label, phase 2 study to test the PD-1 inhibitor pembrolizumab (Keytruda, Merck) among 42 patients with stage IV non-small cell lung cancer and small brain metastases (5 mm to 20 mm) that had not been previously treated or had progressed after radiotherapy.
“Until this trial, it was not really known whether immune therapies had activity in the brain. We knew they could have fantastic activity in the body, but the brain hadn’t been studied,” Goldberg told Healio. “Most other trials have required radiation or local therapy to the brain, so the goal of our study was to try to understand whether the drug can also work in the brain.”
Median follow-up was 8.3 months (interquartile range, 4.5-26.2).
Results showed responses in the brain among nearly one-third of patients whose tumor had PD-L1 expression. The treatment also appeared safe.
Goldberg spoke with Healio about the study’s findings, published in The Lancet Oncology, and what they could mean for future treatment of NSCLC with brain metastases.
Question: Can you describe the study design?
Answer: We treated patients with NSCLC and brain metastases. Most of these patients had tumors with PD-L1 expression, although we included an exploratory cohort of patients who either had PD-L1-negative disease or could not be evaluated for PD-L1. We treated patients with pembrolizumab at a higher dose than is currently FDA approved because we started this study before FDA approval of pembrolizumab. The dose we used was 10 mg/kg every 2 weeks until disease progression or significant adverse effects that required treatment discontinuation.
Among patients who had any PD-L1 expression, we observed a response rate in the brain of approximately 30%. This means that about 30% of these patients had shrinkage of their tumors that reached the level of partial or complete response in the brain. Response in the body was about the same. Interestingly, though, some patients had response in the brain but not in the body, and vice versa. Among the five patients who were PD-L1-negative or unevaluable, we did not see any responses in the brain.
Q: What new information did you learn about PD-L1 levels and immunotherapy?
A: PD-L1 expression is the standard biomarker in NSCLC used to choose patients for therapy. We use it as part of our standard of care because, based on other studies, a higher level of PD-L1 is associated with a better chance for response to PD-1 and PD-L1 inhibitors. However, it isn’t an ideal biomarker. In our study, we looked at other potential biomarkers to try to understand whether we could select patients with brain metastases who could benefit from pembrolizumab. We looked at tumor-infiltrating lymphocytes and found some correlation with benefit, but nothing clear-cut.
Q: Do you think immunotherapy could represent a new treatment option for patients with NSCLC and brain metastases?
A: This is a small phase 2 study, so I don’t think we can say this is the new standard for all patients. However, it shows us that PD-1 inhibitors can have activity in the brain, and a few other subset analyses from larger trials or retrospective studies have supported this, as well. We don’t have randomized data at this point to know if local therapy in the form of radiation is superior, inferior or equivalent to immune therapy alone. However, it can be considered for patients in certain situations. For example, patients sometimes have asymptomatic, small brain metastases but a large bulk of systemic disease, in such cases one could consider starting systemic therapy upfront rather than delaying systemic treatment for local therapy to the brain. When this strategy is used, we need to monitor the brain closely because the response rate is still only 30%. What we’ve learned from our trial is that those patients who do progress in the brain can often go on to receive local therapy and can do quite well. A larger randomized trial will be important to confirm that this is appropriate as a standard for most or all patients. For select patients, in the meantime, I think this strategy can be considered after discussion with a multidisciplinary team of clinicians.
For more information:
Sarah B. Goldberg, MD, MPH, can be reached at firstname.lastname@example.org.