Clinical Professor of Medicine
Lung Cancer Unit Leader
Division of Hematology-Oncology
Department of Medicine
University of California, San Diego
San Diego, California
Lyudmila Bazhenova, MD, has disclosed that she has received consulting fees from AstraZeneca, BeyondSpring, Blueprint, Boehringer Ingelheim, G1 Therapeutics, Genentech, Loxo Oncology, and Takeda and has ownership interest in Epic Sciences.
Effectively managing central nervous system (CNS) metastasis in our patients with and without oncogene-driven non-small-cell lung cancer (NSCLC) remains an ongoing challenge. In general, CNS metastases are more common in oncogene-driven disease, being present at diagnosis in 28% to 65% of patients with EGFR mutations and 34% of those with ALK aberrations. Even among patients without CNS disease at baseline, the prolonged survival afforded by targeted therapies leads to many developing CNS metastasis or, in up to 5% of patients, leptomeningeal carcinomatosis at some point during their illness. What factors should we consider when treating CNS disease?
Things I Consider When Deciding How to Manage CNS Metastasis in My Patients With NSCLC
At Time of New Diagnosis
I ask myself 3 key questions when a patient has just been diagnosed with CNS metastasis. First, how symptomatic is the patient? Is this an emergency requiring immediate treatment, or can treatment wait? In these patients, I obtain a neurosurgical consult to try to resect the tumor. This is because if the patient is highly symptomatic, we cannot wait for radiation or systemic therapy to have an effect. After the surgery, we perform resection cavity radiation followed by systemic therapy.
Second, where is the CNS metastasis? Is it located in a critical region such as the brain stem, where even a millimeter of growth will cause life-altering symptoms? Because I am not an expert in interpreting brain imaging, I always make this determination in collaboration with a CNS radiation oncologist and CNS radiologist, and together we decide whether we can afford to wait to radiate or not. Although many of our newer targeted therapies have good CNS activity, it takes time for the metastases to decrease. In addition, even with the most effective later-generation TKIs, the intracranial response rates are high but not 100%. The intracranial ORR and CR rate with first-line osimertinib for EGFR-mutated disease is 77% and 18%, respectively; 55% and 20% with entrectinib for ROS1-altered disease; 81% and 38% with first-line alectinib in ALK-altered disease; 78% and 11% with first-line brigatinib in ALK-altered disease; and finally, in progressive ALK-altered disease, 42.0% and 7.7% with brigatinib, and 60% and 20% with lorlatinib. If the lesion is located where even minimal growth will affect a patient’s quality of life, radiating it is the best option.
Third, and of most importance, can I offer an intracranially active systemic therapy to the patient and avoid brain radiation? Here, I also consider whether the patient would receive stereotactic radiation, which has fewer adverse events, or whole‑brain radiation for extensive metastases, which has more complications over time. I tend to lean more toward systemic therapy if proposed radiation includes whole-brain radiotherapy.
We currently lack randomized clinical trial data on optimal sequencing of brain radiation and TKIs in oncogene-driven NSCLC with CNS metastasis. When faced with a patient presenting with CNS metastases, should I give those TKIs known to have good CNS penetration upfront and then only use radiation as salvage if my chosen systemic treatment does not work in the CNS? Or should I radiate upfront and start on a TKI simultaneously or immediately after. Retrospective analyses suggest that giving radiation upfront to patients with oncogene‑driven disease may result in better outcomes, but these studies were done in the era of older, less CNS-penetrant TKIs, and lack of randomization allows for selection bias.
The ongoing phase II OUTRUN trial in Australia and Singapore comparing intracranial PFS in patients with EGFR-mutated NSCLC randomized to osimertinib vs osimertinib plus stereotactic radiosurgery should shed light on this open question and help guide treatment decisions in the setting of CNS metastases in the near future (NCT03497767).
There are several key points regarding surveillance and evaluation of treatment response in these patients. If your treatment for CNS metastases was just systemic therapy, it is important to obtain follow-up MRI relatively soon to ensure that you do not miss a progressing tumor. In our clinic, if the decision was to postpone radiation, I generally conduct an MRI 6-8 weeks after starting systemic therapy. If the brain metastases are responding, then I continue with just systemic therapy. If there is no change, I lean toward adding radiation to the current systemic therapy the patient is receiving. However, given the lack of informative data in this setting, I make these decisions through a multidisciplinary discussion with my radiation oncologist.
I monitor these patients regularly with brain MRIs—every 3 months. In patients without a history of CNS metastases but with high risk of CNS involvement, such as patients with oncogenic-driven tumors, I tend to perform brain MRI every 6 months in addition to symptom screening at every clinic visit. With this approach, I am rarely surprised by a large, symptomatic CNS tumor.
Both patients who had no CNS metastasis at diagnosis and those who had their initial CNS metastasis successfully treated can potentially develop CNS metastasis as the sign of first recurrence. When I detect a CNS metastasis as recurrence, I consider the same factors outlined above when making my treatment decisions. Where is the brain metastasis? How symptomatic is the patient? Do I have an active systemic therapy to give the patient? The one slight difference in this setting is the need to understand the status of systemic disease control. If CNS recurrence is the only site of progression, one can consider radiating just the CNS metastasis, especially if it is solitary, and continue the patient on the same systemic therapy. Monitoring is also important here, as discussed.
However, if the patient is progressing both in the brain and systemically, I switch the systemic therapy and make further decisions based on CNS penetrability of my new systemic agent. For example, in a patient with ALK-altered NSCLC who has been receiving alectinib or brigatinib and then develops both systemic and asymptomatic CNS recurrence, I would choose lorlatinib because of its systemic and CNS activity in patients who failed a second-generation ALK TKI. In a patient experiencing systemic and CNS progression on lorlatinib, my next choice is chemotherapy, which has mediocre CNS activity. Thus, I would first radiate the CNS and then switch to the chemotherapy.
Considerations for Patients With Advanced NSCLC and Leptomeningeal Carcinomatosis
We currently lack effective therapies to manage meningeal involvement in our patients with advanced NSCLC; most studies assess CNS penetration into brain parenchyma, rather than penetration into the cerebrospinal fluid. However, there are prospective trial data that at least one of our new targeted therapies is active against leptomeningeal metastases. In the phase I BLOOM trial, patients with EGFR-mutated NSCLC and symptomatic leptomeningeal disease experienced symptom improvement when given osimertinib 160 mg.
Furthermore, a recent retrospective analysis of the AURA program reported similar clinical benefit with the standard osimertinib dose of 80 mg. If there is an active systemic therapy known to have potential CNS penetration available, I will offer it to my patients with leptomeningeal disease. The alternative would be intrathecal chemotherapy (eg, with methotrexate), but I reserve this for when targeted therapy is not an option.Get up to date on the latest developments in lung cancer biomarkers and associated targeted therapies by continuing to follow CCO and this educational program! Visit here to learn more about leveraging targeted agents in lung cancer care with downloadable slides and on-demand Webcasts from live Webinars, a biomarker summary handout, and a treatment decision support tool (all forthcoming in 2020).
You are accessing CCO's educational content today as a Guest user.
If you would like to continue with free, full access to the CCO Web sites, including free CME/CE credits, please click the button below.