The management of patients with leptomeningeal metastases (LM) is multifaceted and complex. Even with an aggressive approach, therapeutic outcomes are uniformly disappointing. This is because of the relentless growth of the central nervous system (CNS) and/or the systemic cancers, or their lethal complications. Advances in the understanding of the homing of cancer cells to the CNS, and of cancer metastasis in general, and more effective anticancer drugs that are adequately delivered to the CNS and cerebrospinal fluid (CSF) are needed to improve outcomes for patients with LM. These advances may lead to better treatments for this disease and, ultimately, its prevention.
Despite the passing of nearly 140 years since its original description, little progress has been made in improving survival for patients with LM. Hematologic malignancies can result in LM in up to 24% of patients. The most common solid tumors causing LM are breast cancer, lung cancer, and melanoma,1with incidences ranging between 5% and 23%. Leptomeningeal metastases risk rises with longer cancer survival.
In the United States, standard treatment for LM includes CSF diversion when indicated, radiation, local intrathecal (refers to intraventricular administration, unless otherwise specified) chemotherapy, and systemic chemotherapy. Recent reviews discuss management of LM.2,3With standard interventions, median survival for patients with LM ranges from 8 to 16 weeks. Roughly 24% to 34% die of LM alone, 22% to 25% die of LM simultaneously with progressive systemic cancer, 19% to 44% die of systemic disease progression, and up to 10% die of other causes.4- 6
The incidence of brain metastases (BM) and LM may increase in the near future for at least 2 reasons: (1) longer control of non-CNS cancers may allow for more time for the development of CNS metastases and (2) the use of large-molecule antineoplastic agents with limited CNS and CSF penetration may control systemic disease but leave LM unaffected behind the blood-brain barrier (BBB) and blood-CSF barrier (BCSFB). Paradoxically, the newer large-molecule therapies may improve overall cancer survival while increasing the incidence of CNS metastases. Longer survival and exposure of tumor cells to genotoxic chemotherapy may select for increasingly chemoresistant cell clones, making LM even more resistant to therapy over time.
Once we understand the biology behind LM, we can develop therapies targeting these biological changes. Metastasis is a cascade of events with multiple cellular and molecular changes.7For LM to develop, tumor cells must detach from the primary site, invade a blood or lymphatic vessel, survive vascular transit, adhere to host organ endothelium,8invade the host organ, proliferate, and develop a blood supply.7Molecular factors implicated in CNS metastases and LM include E-cadherin–catenin complexes, plasmin, urokinase-type plasminogen activator, metalloproteinases, tissue inhibitors of metalloproteinases (associated with brain invasion),9and activated integrin αvβ3.10Metalloproteinases can degrade endothelial tight junction proteins at the BBB. Along with vascular endothelial growth factor (VEGF)11and stromal-derived factor 1,12metalloproteinases may allow for transendothelial migration of tumor cells.13Other mechanisms that may contribute to the development of LM include an ectodermal origin of the primary tumor, which may allow for advantageous cell-cell interactions between the metastatic tumor cells and native brain cells.14,15Further, tumor cell surface markers, such as the extracellular domain of the epidermal growth factor receptor 2 protein, are associated with a higher risk of BM in breast cancer16and possibly LM.
The delivery of many anticancer drugs from blood to the brain17,18and from blood to CSF is restricted, for multiple reasons. Drug-related factors include the drug's level of protein binding and its molecular weight, polarity, and lipid solubility.19Physical factors include the architectural properties of the BBB and BCSFB, such as the tight junctions between the endothelial cells of brain capillaries and the epithelial cells of the choroid plexuses, limiting paracellular diffusion of polar compounds. Further, adenosine triphosphate–dependent pumps such as the P-glycoprotein system, multidrug resistance proteins, and organic and inorganic ion transporters can mediate efflux of some anticancer drugs away from the brain.19- 25
The BBB and BCSFB are not identical. The BCSFB has a more relaxed tight junction architecture that correlates with differential diffusion capacities between it and the BBB.26Recent work has investigated the effect of P-glycoprotein–modulating drugs on the CSF penetration of some chemotherapies. Tamoxifen, a P-glycoprotein inhibitor, decreased the CSF penetration of paclitaxel, supporting the concept that the pumping direction of P-glycoprotein at the choroid plexus is in the opposite direction to the BBB. The P-glycoprotein system appears to direct natural product toxins away from the brain into the CSF, and when inhibited, lower CSF drug levels are found.27Animal models reveal similar findings with the tyrosine kinase inhibitor gefitinib. Its administration results in lower CSF levels and higher brain parenchymal levels of the topoisomerase I inhibitor topotecan.28
Table 1shows the CSF:plasma ratios for some of the drugs studied in humans and rhesus monkeys. A CSF:plasma ratio lower than 0.05 signifies nonspecific leakage of drug. Table 1shows that many drugs normally achieve CSF:plasma ratios lower than 0.05. However, once LM has arisen, or if radiation is directed to the CNS, leakage of the BCSFB develops, and larger molecules can leak from blood into CSF.68Further, some drugs have an intrinsically high CSF:plasma ratio, suggesting their possible utility in treating LM. As the understanding of the BBB and BCSFB advance, we may ultimately be able to facilitate the CNS and CSF penetration of therapeutic molecules, which are now excluded.
Intrathecal chemotherapies typically used in LM include methotrexate, cytarabine, liposomal cytarabine, and thiotepa.3Systemic therapies are usually chosen based on tumor histology, drug penetration into the CSF, and a patient's prior drug exposure.
Even though intrathecal chemotherapy is widely used in the United States for solid-tumor LM, proof of its benefit has not been established in randomized controlled trials.69Randomized controlled trials do suggest modest improvements with long-acting over standard intrathecal chemotherapies,70,71and some retrospective studies suggest intrathecal chemotherapy prolongs survival,72but there exists contrary evidence.73,74A recently begun randomized controlled trial, European Organization for Research and Treatment of Cancer 26051, is testing intrathecal liposomal cytarabine vs supportive care in solid-tumor LM.
Intraventricular (as opposed to intralumbar) chemotherapy delivery results in improved CSF drug levels and less interpatient variability of drug distribution. This form of regional chemotherapy has led to effective treatment of occult and overt meningeal leukemia in humans, and based on this success, investigators continue its evaluation in patients with solid tumor, hoping for similar outcomes. Pharmacokinetics of commonly used intrathecal anticancer agents shows that high drug concentrations can be achieved in the CSF and leptomeninges but not deep into the brain. Because of this, intrathecal administration is not effective for bulky disease in the meninges.75Further, in solid-tumor LM, intraventricular administration, or the use of sustained-release chemotherapeutic agents if the lumbar route is used, appears to improve treatment outcome.76
Experimental and Newer Agents.The most promising recently tested cytotoxic and radiotherapeutic agents are presented in Table 2. The topoisomerase inhibitors appear as effective as traditionally used intrathecal agents, and both etoposide and topotecan hydrochloride have little toxicity, so may be useful in combination with other agents or as prophylaxis.77,78,87A concentration × time study of intrathecal topotecan is open and accruing patients within the Pediatric Brain Tumor Consortium. Because of pain associated with intrathecal administration, mafosfamide requires slow delivery and premedication with steroids and narcotics but may be useful in childhood CNS malignancies to help delay or avoid radiation exposure.79,80Because of almost no toxicity and some efficacy (29% CSF clearance), sodium iodide I 131 (131I) will be further studied in a phase 2 trial with a higher-dose, multiday schedule.84For similar reasons, phase 2 studies evaluating serial intrathecal injections of the GD2-targeted monoclonal antibody 131I-3F8, are under way.85Early data suggest efficacy in childhood primitive neuroectodermal tumors and neuroblastoma.
Noncytotoxic Intrathecal Therapies.Immunotherapies.The CSF space may be excluded from the benefits of the systemic antitumor effects of the immune system, so immunotherapeutic approaches to the treatment of LM are theoretically attractive. Unfortunately, immune responses are frequently associated with inflammation. Intrathecal administration of interleukin 2 or interferon alfa both resulted in responses in patients with LM but were also fairly toxic, limiting enthusiasm for further development.88,89
Rituximab.Rituximab is a humanized monoclonal antibody against the CD-20 antigen expressed on most B-cell lymphomas. It has been used intravenously since 1997. Cerebrospinal fluid levels of this large molecule (146 kDa) are only 0.1% of the serum level after intravenous administration.57Several case reports demonstrating safety and possible benefits of intrathecal administration of rituximab led to a recently reported phase 1 study.90In this study, the maximum tolerated dose of intrathecal rituximab was 25 mg twice weekly (9 doses total). Mean peak CSF concentration 1 hour postdose rose to 472 μg/mL and estimated half-life averaged 34.9 hours. Cytologic responses were seen in 6 of 10 patients; 4 patients experienced a complete response; 2 patients experienced improvement in intraocular lymphoma; and 1 patient's intraparenchymal lymphoma improved. Toxic reactions were limited. Further studies developing this promising therapy are under way. Additionally, a study of intrathecal rituximab combined with intrathecal methotrexate91for patients with intraocular or LM lymphoma has been initiated.
Trastuzumab.Trastuzumab is a humanized monoclonal antibody that binds to the epidermal growth factor receptor 2 protein, which is overexpressed in 30% of primary breast cancers, as well as some other tumors. Trastuzumab inhibits the growth of tumor cells and mediates antibody-dependent cellular cytotoxicity. A recent study showed that the CSF:serum trastuzumab ratio increased from 0.0023 prior to brain radiotherapy to 0.013 after completion of radiotherapy and was as high as 0.02 with concomitant LM after radiotherapy, revealing that CSF trastuzumab levels are low but can increase if BBB function is impaired.68
Promising results from a pilot study using intrathecal trastuzumab in patients with LM due to breast cancer, medulloblastoma, or glioblastoma were recently presented.92In this report, 16 patients with LM (11 glioblastoma multiforme, 4 breast cancer, 1 medulloblastoma) were treated with intrathecal trastuzumab (20-60 mg per dose, either weekly or every other week) for 4 treatments. Stable patients continued every-other-week therapy until neurologic progression. Two patients with breast cancer, 7 with glioblastoma multiforme, and the one with medulloblastoma responded without reported adverse events; the epidermal growth factor receptor 2 protein status appeared to be predictive of response. Based on these results, further study of intrathecal trastuzumab is warranted.
Numerous reports suggest that systemic therapy improves survival for patients with LM.72,93- 100Some authors feel systemic therapy is the most important part of the treatment of LM73,74and exclude intrathecal therapy in patients with responsive cancers.94,95,97,101Agents capable of producing adequate CSF concentrations following systemic administration may benefit patients with LM.
Methotrexate.Methotrexate inhibits dihydrofolate reductase and the synthesis of purine nucleotides and thymidylate, interfering with DNA synthesis and repair. At high doses, methotrexate has favorable CSF penetration. A prospective, nonrandomized study comparing intrathecal methotrexate (n = 15) vs high-dose systemic methotrexate (n = 16) in patients with LM produced provocative results. High-dose methotrexate (8 g/m2over 4 hours) resulted in a mean peak concentration of 17.1 μmol/L in the CSF; cytotoxic CSF methotrexate levels remained measurable much longer than with intrathecal dosing. Furthermore, there was higher CSF tumor cell clearance and survival was longer (13.8 months vs 2.3 months, P = .003) in the systemic methotrexate-treated cohort.102Because of the favorable pharmacokinetics of high-dose methotrexate, further studies in patients with LM are warranted, possibly in combination with other agents.
Capecitabine.Capecitabine is a fluoropyrimidine carbamate designed as an oral alternative to 5-fluorouracil. Capecitabine is enzymatically converted to 5-fluorouracil at the tumor site. The increased drug concentration at the tumor site may enhance its antitumor activity and reduce systemic toxicity. Although there is no formal pharmacokinetic data regarding capecitabine's behavior in the CNS, there are empirical observations of responses to the drug in patients with BM and LM.103,104Capecitabine has also resulted in responses in a few patients with recurrent BM or LM even after previous capecitabine exposure.104,105Based on the existing reports, capecitabine is now frequently being used in patients with LM or BM secondary to breast cancer, and further prospective study is under way.
Temozolomide.Temozolomide is an orally bioavailable alkylator that reaches CSF levels roughly 20% of those in the serum.31In a pilot study of oral temozolomide in 10 patients with LM, the drug was well tolerated, although no responses were seen. Two patients had stable disease through 2 courses (6 weeks receiving therapy, 4 weeks not receiving therapy) but progressed while not receiving treatment,106suggesting that continuous treatment might be more efficacious.
There are several case reports of a beneficial contribution of hormonal therapy for patients with LM with hormone-sensitive tumors (breast and prostate cancer). Responses are reported lasting more than 12 months. For patients with LM from hormone-sensitive cancers, hormonal treatment is reasonable to continue or initiate and may provide some activity against the LM.98- 100
Pemetrexed.Pemetrexed, a chemotherapeutic molecule similar to methotrexate, is approved for mesothelioma and non–small cell lung cancer and is active in methotrexate-resistant malignancies. The CSF penetration of pemetrexed was low in an animal model50; however, the CSF pharmacokinetics of systemically administered pemetrexed are being evaluated in an ongoing study in patients with LM. The drug is unique from methotrexate in that it is a “multitargeted” antifolate compound acting through several enzyme systems involved in folate metabolism. Pemetrexed gains intracellular access via at least 4 mechanisms, which may increase its activity over methotrexate. Early results demonstrate CSF responses in patients with breast cancer with LM (J. Raizer, MD, written communication, June 22, 2009).
Bevacizumab.Bevacizumab is a systemically administered monoclonal antibody directed against VEGF. Bevacizumab is approved for use in colorectal, breast, and non–small cell lung cancers and glioblastoma multiforme. Recent reports have identified elevated VEGF levels in the CSF of the majority of patients with LM due to melanoma or breast or lung cancer.107- 109Preliminary data suggest that in LM responders CSF VEGF levels fall and correlate with response.109The degree to which bevacizumab penetrates the CSF is unknown but is likely limited. Testing is under way at MD Anderson Cancer Center in patients with LM due to breast and lung cancer and melanoma to determine if systemically administered bevacizumab can affect CSF VEGF levels or impact tumor cells in the CSF.
Gefitinib.Gefitinib is a small-molecule tyrosine kinase inhibitor with activity against lung cancers that contain mutations of the epidermal growth factor receptor. Case reports have shown responses in patients with LM from non–small cell lung cancer.110,111A prospective study evaluating high-dose gefitinib (up to 1250 mg/d) in patients with LM with non–small cell lung cancer and sensitizing epidermal growth factor receptor mutations was recently closed. High doses of gefitinib were used (standard dose, 250 mg/d) attempting to increase CNS and CSF drug levels and improve anticancer effects.112Early reports of the clinical, CSF, and imaging outcomes were promising; final results are forthcoming (D. Jackman, MD, written communication, June 22, 2009).
Most reports of intrathecal LM treatments include patients who simultaneously receive systemic agents,5,6,97and many investigators feel combination intrathecal and systemic therapy improves outcomes.72,97Several planned studies will evaluate the concept of combination therapy, prospectively. Some clinical trials in development include a phase 2 study of intrathecal thiotepa for patients with LM due to primary brain tumors, a phase 2 study of lomustine plus cisplatin plus vincristine sulfate and intrathecal liposomal cytarabine for adults with medulloblastoma and CSF positive for tumor cells, and a phase 1/2 study of oral capecitabine plus liposomal cytarabine in patients with breast cancer with LM (R. Soffietti, MD, written communication, June 27, 2009).
Because of a paucity of available patients, LM studies often accrue multiple primary histologies. This heterogeneity obscures potential efficacy signals. Investigators, with respect to rituximab, gefitinib, and bevacizumab, as noted earlier, are beginning to design LM trials with specific histologies in mind.
Prevention strategies similar to those used for children with acute lymphoblastic leukemia113or in patients with aggressive lymphoma114may become feasible if genetic markers identifying tumors with a propensity to invade the CNS can be identified. High positive predictive value plasma or CSF biomarkers could allow for earlier treatment of LM, possibly affording better tumor control. Early studies suggest CSF VEGF may be useful as a biomarker,107but further research is warranted. Until prevention is feasible, or biomarker use is validated, unique clinical scenarios may still hold opportunities for earlier treatment and better outcomes.
Patients with BM may be at increased risk of developing LM, especially if the BM are located in the posterior fossa (BMPF). Among patients undergoing craniotomy for BMPF, estimates of the risk of developing LM are reported as high as 67%.115- 117Recent reports have begun to dissect out the details on the risk of CSF seeding after craniotomy. In a review of 379 patients with BMPF who were treated with either surgical resection or stereotactic radiation, 8.7% developed LM. But, there was a significantly higher risk of LM (14%; rate ratio, 2.45; P = .02) in those patients having a piecemeal resection of their BMPF when compared with either stereotactic radiation or en bloc resection.118A follow-up study of 827 patients undergoing craniotomy for supratentorial BM found a similar result, with a hazard ratio of 5.8 (P = .002) comparing piecemeal resection vs stereotactic radiation and a hazard ratio of 2.7 (P = .009) comparing piecemeal resection vs en bloc resection.119Patients with BM who undergo piecemeal tumor resection may be a good population in which to test biomarker-based or prophylactic interventions against LM.
Leptomeningeal metastases remain a neurologically devastating and fatal late complication of cancer. The molecular biology underpinning the development of LM is slowly being unraveled. To be effective, new treatments for LM need to reach the meninges and CSF and interact with relevant molecular targets. Since only about one-third of patients with LM die solely of LM, therapies that effectively address the systemic cancer and the LM are necessary for major improvements in survival. Progress is slowly being made with the testing of newer targeted agents and combination treatments, but obviously, there is much work to be done to improve outcomes for patients with LM.
Correspondence:Morris D. Groves, MD, UT MD Anderson Cancer Center, Department of Neuro-Oncology, 1400 Holcombe Blvd, Unit 431, Houston, TX 77030 (firstname.lastname@example.org).
Accepted for Publication:July 19, 2009.
Financial Disclosure:Dr Groves received research funding from Genentech, Enzon Pharmaceuticals, and Schering-Plough Research Institute and has been on the speakers' bureau or received honoraria from Enzon Pharmaceuticals and Schering-Plough Research Institute.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Neurology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 20
Customize your page view by dragging & repositioning the boxes below.
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.