Outcomes for patients with non–small-cell lung cancer (NSCLC) have gradually improved over the past 2 decades. Brain metastasis has increasingly become a major problem in the treatment of patients with NSCLC for two reasons: (1) The routine use of magenetic resonance imaging (MRI) for staging purposes, even in asymptomatic patients with metastatic NSCLC, has resulted in the identification of small asymptomatic lesions that would otherwise have gone unnoticed for some time. (2) With the availability of more effective systemic therapy for patients with resected NSCLC and locally advanced NSCLC, the brain as a single site or as the first site of relapse is becoming more common. It is not surprising, therefore, that the rate of brain metastasis in NSCLC is now approximately 30%.[1-5]
Historically, patients with NSCLC and brain metastases have been excluded from many clinical trials evaluating the efficacy of systemic therapies for a variety of reasons. Patients with brain metastases are generally thought to have a very poor prognosis. Moreover, there are concerns that the systemic agents under investigation are unlikely to cross the blood-brain barrier and therefore may not be effective simply because of delivery issues. Finally, the possibility of hemorrhagic conversion of a bland necrosing metastasis in response to therapy, resulting in devastating debilitation (or, worse, death), has tempered enthusiasm about including these patients in prospective clinical trials. This issue, of course, is of paramount importance currently, with increasing interest in antiangiogenic agents. But are the issues real and should they restrict us from including this subset of patients with NSCLC?
Drs. Oh and Stewart address these timely issues head-on and present compelling arguments for including these patients in clinical trials. They correctly point out that the false notion of an intact blood-brain barrier in established brain metastasis persists despite a wealth of information to the contrary. The tumor vasculature associated with established and visible brain metastases is quite distinct from the normal vasculature of the brain with its intact blood-brain barrier. Pharmacologic studies (many from Dr Stewart's own work) have shown that the uptake of pharmaceuticals in lesions metastatic to the brain is similar to that in primary tumors. In addition, multiple clinical studies have confirmed that responses to systemic agents in the brain parallel those at other sites (as outlined in an elegant table accompanying the Oh and Stewart's paper), further reinforcing the point that drugs do indeed reach metastatic lesions in the brain unencumbered by the blood-brain barrier.
The authors also argue that, although historically these patients have had poorer outcomes as compared to other lung cancer patients, with earlier detection of disease by MRI and the availability of stereotactic radiosurgery (SRS), results are beginning to change for the better. While we tend to agree with this view, this issue needs to be studied more carefully in the coming years. There are clearly different subsets of patients with brain metastases who have different outcomes. Finally, although intracranial hemorrhage is a concern, the authors present data showing that in patients treated with whole-brain radiotherapy (WBRT), with or without SRS, the incidence of this adverse event is approximately 1% to 2%—not worrisome enough to exclude patients from clinical trials.
These are persuasive arguments to include patients with brain metastases in clinical trials of advanced NSCLC, and a quick review of ongoing phase III clinical trials in advanced NSCLC (as of January 22, 2008) listed on the National Institutes of Health website, www.clinicaltrials.gov, suggests that this is indeed beginning to happen. We identified 28 trials of systemic therapy in NSCLC, 23 of which stated whether or not patients with brain metastases were eligible (Table 1). Remarkably, 70% of the ongoing phase III trials (for which we have information) do include patients with known brain metastases.
While it is encouraging that patients with brain metastases are being included more frequently, it is important to learn how best to identify the appropriate patients with brain metastases for clinical trials. At the very least, patients should be neurologically stable and must have recovered from the side effects of cranial irradiation before beginning systemic therapy. However, we believe that these criteria are not sufficiently restrictive to identify those who are unlikely to benefit from systemic therapy. We need to exclude those with poor outcomes, as this might adversely affect the development of newer, often only marginally superior, regimens. Readily used exclusion criteria (poor performance status, serum albumin, organ function, etc) may not accurately identify those who have poor outcomes mainly because of intracranial disease.
It is important to carefully study some of the prognostic indices for possible incorporation in clinical trials, either for determining eligibility or for stratification. The Radiation Therapy Oncology Group (RTOG) recursive partitioning analysis (RPA) groups patients with brain metastases into three classes based on clinical criteria with differing outcomes.[6,7] Patients with a controlled primary tumor, no other metastatic sites, age under 65, and a Karnofsky performance status (KPS) > 70 (RPA class I) have the best outcomes (median survival of 6–7 months). However, most patients (65%–75%) fall into RPA class II, and have median survivals of 3.8 to 4.5 months.[6-8]
A new proposed prognostic index, the graded prognostic assessment (GPA), also considers age, other metastases, and KPS, but adds the number of central nervous system metastases—which has been shown to be an independent predictor of survival—as an additional prognostic factor.[8,9] Four score groupings are determined, with median survivals of 11, 6.9, 3.8, and 2.6 months. The majority of patients in this system (62%) fall into the group with a median survival of 3.8 months.
Understanding the Risks
Emerging data suggest that even with the perceived risk of bleeding, antiangiogenic agents may be safe in patients with brain metastases from lung cancer. The potential safety in the setting of treated brain metastases is now being evaluated in the PASSPORT trial, which seeks to enroll 100 patients who have received radiotherapy for brain metastasis and are subsequently receiving bevacizumab(Drug information on bevacizumab) (Avastin) therapy. The preliminary data from two ongoing studies reported no intracranial bleeding in 26 patients with NSCLC and brain metastases treated with bevacizumab.
Only limited data are available with regard to the safety of vascular endothelial growth factor (VEGF) tyrosine kinase inhibitors in patients with brain metastases from lung cancer. One patient treated with sunitinib (Sutent) developed a fatal intracranial hemorrhage and was subsequently found to have a brain metastasis. An episode of grade 4 cerebral hemorrhage was observed in the phase I studies of AZD2171. Patients with brain metastases (following radiation) were subsequently enrolled in clinical trials evaluating sorafenib(Drug information on sorafenib) (Nexavar) and vandetanib (Zactima) independently, with no intracranial hemorrhages reported so far, to the best of our knowledge.[13-15] Although these data are reassuring, clearly more patients need to be studied in controlled clinical trials before antiangiogenic agents would be deemed safe for use in the setting of brain metastases outside the context of clinical trials.
Many anticonvulsants, including phenytoin, carbamazepine(Drug information on carbamazepine), and phenobarbitol (which are commonly used in these patients), induce cytochrome P450 enzymes, leading to increased metabolism of chemotherapeutic agents and corticosteroids. Newer medications, such as levetiracetam(Drug information on levetiracetam) (Keppra), are not enzyme inducers and should be used preferentially in practice. Careful attention should be paid with regard to concurrent anticonvulsant medications when administering systemic therapy to patients with brain metastases.
It is critical to understand the biologic basis of metastases in general and organ-specific metastases in particular. Preclinical studies suggest that VEGF may play an important role in the eventual development of brain metastases.[17,18] Expression of EphA2, a receptor tyrosine kinase, has been found to correlate with the development of brain metastasis in one study. Identifying patients at highest risk for brain metastasis is important, so that strategies such as prophylactic cranial irradiation (PCI) can be targeted for the molecularly defined high-risk population. On a more ambitious note, a better understanding of the molecular basis of brain metastases could lead to the development of novel therapeutic agents.
—Boone Goodgame, MD
—Ramaswamy Govindan, MD
Financial Disclosure: Dr. Govindan is a consultant for and has received honoraria and research support from Genentech. He has also received research support from Pfizer.