Brain metastases

PRIMER Brain metastases Achal Singh Achrol1*, Robert C. Rennert2*, Carey Anders3, Riccardo Soffietti4, Manmeet S. Ahluwalia5, Lakshmi Nayak6, Solange Peters7, Nils D. Arvold8, Griffith R. Harsh9, Patricia S. Steeg10 and Steven D. Chang9* Abstract | An estimated 20% of all patients with cancer will develop brain metastases, with the majority of brain metastases occurring in those with lung, breast and colorectal cancers, melanoma or renal cell carcinoma Brain metastases are thought to occur via seeding of circulating tumour cells into the brain microvasculature; within this unique microenvironment, tumour growth is promoted and the penetration of systemic medical therapies is limited Development of brain metastases remains a substantial contributor to overall cancer mortality in patients with advanced-​stage cancer because prognosis remains poor despite multimodal treatments and advances in systemic therapies, which include a combination of surgery , radiotherapy , chemotherapy , immunotherapy and targeted therapies Thus, interest abounds in understanding the mechanisms that drive brain metastases so that they can be targeted with preventive therapeutic strategies and in understanding the molecular characteristics of brain metastases relative to the primary tumour so that they can inform targeted therapy selection Increased molecular understanding of the disease will also drive continued development of novel immunotherapies and targeted therapies that have higher bioavailability beyond the blood–tumour barrier and drive advances in radiotherapies and minimally invasive surgical techniques As these discoveries and innovations move from the realm of basic science to preclinical and clinical applications, future outcomes for patients with brain metastases are almost certain to improve *e-​mail: achrol@jwci.org; rrennert@ucsd.edu; sdchang@stanford.edu https://doi.org/10.1038/ s41572-018-0055-y Brain metastases develop following the spread of cells from a primary tumour through the blood (haemato­ genous seeding) to the brain microvasculature, although seeding from established metastases might also occur Complex microenvironmental niche–tumour inter­ actions, neuroinflammatory cascades and, possibly, neo­ vascularization (Fig. 1) are involved in establishing a new metastasis (colonization) Brain metastases ultimately manifest clinically with mass effect on the brain and from the consequences of their treatment They eventually progress despite multimodal treatments Brain metastases can often be the cause of the initial presenting symptoms in patients with otherwise previ­ ously undiagnosed advanced-​stage cancer The majority of brain metastases result from lung, breast and colorec­ tal cancer (CRC), melanoma and renal cell carcinoma (RCC), although brain metastases from other tumour types (including haematological cancers) have been reported Upon diagnosis, brain metastases are com­ monly treated with multimodal therapies that can include a combination of surgery, radiotherapy, chemotherapy, immunotherapy and targeted therapies However, prog­ nosis after the development of brain metastases from NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); most cancer types remains poor, with overall 2-year survival percentages in the single digits1 Underlying heterogeneity between and within brain metastases, as well as clonally selected molecular differences relative to the primary tumour site, are likely contributory factors underlying poor outcomes As the fundamental biologi­ cal differences across cancer types and between brain metastases and their parent tumours become better understood, it is expected that newer mechanism-​based therapies will play a greater part in treatment Indeed, development of treatments with enhanced blood–tumour barrier (BTB) penetration is also expected to improve clinical outcomes Transforming these discoveries into new therapies, and integrating them with changing para­ digms in surgery and radiosurgery, will be among the key advances necessary for improving the prognosis for patients with brain metastases In this Primer, we discuss the epidemiology, patho­ physiology, diagnosis, management and quality of life (QOL) impact of brain metastases in people with cancer We conclude with a discussion of the future outlook in the field of brain metastases based on the most recent clinical and research advances Primer Epidemiology An estimated 20% of patients with cancer will develop brain metastases2–4 However, the true incidence is likely higher as such estimates are often limited to patients who are considered for treatment; guidelines for the majority of solid tumours not recommend routine brain MRI screening in patients who not display neurological symptoms Additionally, many studies report only the presence or absence of brain metastases at the time of initial diagnosis but not provide further informa­ tion on the disease course or subsequent sites of meta­ static involvement Accordingly, autopsy studies have suggested higher incidences (up to 40%) of intracranial metastases in patients with cancer5–7 As overall survival continues to improve following initial cancer diagno­ ses8, and clinical trial enrolment increases (often with concomitant requirements for brain MRI screening), the actual and reported incidence of brain metastases is likely to increase Although any type of cancer can metastasize to the brain, the three most common primary tumours associ­ ated with brain metastases are lung (20–56% of patients), breast (5–20%) and melanoma (7–16%)2,4,9,10 The preva­ lence of brain metastases in patients with RCC and CRC is also significant and increasing11 Lung cancer is the most frequent to metastasize to the brain irrespective of patient sex and is the most common brain metastasis occurring in men In women, breast cancer is the most commonly occurring brain metastasis The molecular subtype of the primary tumour can further influence the risk of developing brain metastases For example, women with breast cancers characterized by human epidermal growth factor receptor (ERBB2; also known as HER2) amplification or triple-​negative hormone receptor status (that is, oestrogen receptor (ER)-negative, progesterone receptor (PR)-negative and normal HER2 levels) have a higher risk of developing brain metastases than those with ER-​positive and/or PR-​positive breast cancer12 In fact, brain metastases in these subtypes account for nearly one-​quarter of all breast cancer metastases in newly diagnosed patients12 Another illustration of molecular subtype influencing metastatic patterns is the high propensity of ALK-​rearranged non-​small-cell lung cancer (NSCLC) to specifically metastasize to the brain13 Author addresses Department of Neurosurgery and Neurosciences, John Wayne Cancer Institute and Pacific Neuroscience Institute, Santa Monica, CA, USA Department of Neurosurgery, University of California–San Diego, San Diego, CA, USA Division of Hematology/Oncology, Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC, USA Department of Neuro-​Oncology, University of Turin, Turin, Italy Burkhardt Brain Tumor and Neuro-​Oncology Center, Cleveland Clinic, Cleveland, OH, USA Center for Neuro-​Oncology, Dana-​Farber Cancer Institute, Boston, MA, USA Medical Oncology, Lausanne University Hospital, Lausanne, Switzerland Department of Radiation Oncology, St Luke’s Cancer Center, Duluth, MN, USA Department of Neurosurgery, University of California–Davis, School of Medicine, Sacramento, CA, USA 10 Women’s Malignancies Branch, Center for Cancer Research, National Cancer Center, Bethesda, MD, USA | Article citation ID: The risk of developing brain metastases also increases with more advanced primary disease14,15, although whether contemporary trends towards earlier cancer detection and treatment have an effect on the epi­ demiology of brain metastases remains to be determined In one study comparing epidemiological differences between two cohorts of patients treated ~20 years apart, brain metastases in the more recent cohort were more likely to be identified at the time of an initial cancer diag­ nosis and more likely to be associated with other areas of extracranial disease, such as metastatic involvement of the liver or lungs11 However, in patients without brain metastases at initial diagnosis, the median time from ini­ tial cancer diagnosis to the development of brain metas­ tases was increasing overall and highly influenced by the primary cancer type11 In another study, the median time for patients with breast cancer to develop brain metasta­ ses was 44 months, compared with 11 months for lung cancer10 Although some of these differences may be due to variations in screening and the stage at which the primary tumour is diagnosed, incompletely understood molecular differences across tumour types are also likely to contribute In addition to sex, tumour source and molecular sub­ type, other factors associated with the development of brain metastases include ethnicity, age and geographic location In one study, African Americans with lung cancer, melanoma and breast cancer demonstrated a higher incidence of brain metastases than other ethnic groups with the same cancer types4 Conversely, the inci­ dence proportions for RCC brain metastases in African Americans were lower than in white patients and were similar to white patients for CRC Although it is possi­ ble that socio-​economic factors leading to diagnosis at a later stage partially contributed to these findings, the variations across primary tumour types suggest the pres­ ence of additional contributing mechanisms that warrant further study The risk of brain metastases also varies with age, although its effect is similarly influenced by primary tumour type4,16 For example, in one study, risk of breast cancer brain metastases was highest in younger patients (20–39 years of age), whereas risk of lung cancer brain metastases was highest in middle age (40–49 years of age); melanoma, RCC and CRC brain metastases occurred with highest risk a decade later (50–59 years of age) Although molecular subtypes vary in occurrence with age and may partially influence these trends, at least with regards to breast cancer, the increased risk of brain metastases for patients 35 years of age, triple-​negative or HER2-enriched breast cancer subtypes were associated with a higher brain metastasis risk Finally, although considerable geographic variabil­ ity in cancer patterns is evident17,18, whether this trans­ lates to variability in the incidence of brain metastases remains unclear Efforts to accurately prognosticate patients have included key factors such as age, extent of primary disease control, presence of extracranial metastases or leptomeningeal disease, Karnofsky Performance Status (KPS; in which a score of 0–100 is given whereby a higher score means the patient is better able to carry out www.nature.com/nrdp (2019) 5:5 0123456789(); Primer Metastasis Primary tumour Colonization Micrometastasis Invasion into the stromal environment Cell death Dormancy Proliferation Extravasation Travel through the circulation Intravasation into the vasculature Circulating tumour cell Adhesion Fig | Cancer cell metastatic dissemination Brain metastases develop following the haematogenous (that is, through the blood) spread of cells from a primary tumour to the brain microvasculature, with subsequent tumour growth involving microenvironmental niche–tumour interactions, neuroinflammatory cascades and neovascularization Initially , tumour cells break away from the primary bulk tumour and invade the surrounding tissues, venules, capillaries and lymphatic system (intravasation) Tumour cell interactions with immune cells promote cell motility via clearance of extracellular matrix Once in the circulation, these circulating tumour cells begin the process of metastatic extravasation from the vasculature, which is facilitated by these cells undergoing adhesive (circulatory) arrest Brain metastases tend to occur at the junction between grey and white matter and watershed areas between vascular territories where it is postulated that circulating tumour cells benefit from longer relative mean transit times of blood flow , enabling more time for the cells to overcome the blood–brain barrier and successfully egress from the vasculature After arresting and extravasating, most tumour cells die rather than form metastases whereas others can lie dormant at metastatic sites for extended periods daily activities) and treatment status, including history of surgical resection, radiotherapy and chemo­therapy1,19–21 Data-​driven prognostication tools for patients with brain metastases include the recursive partitioning analysis (RPA) score (derived from patient age, KPS and tumour status) and Graded Prognostic Assessment (GPA), which provides histology-​specific information19,22,23 (see below) The focus on ‘big data’ with these tools is emphasized owing to considerable heterogeneity in terms of primary tumour histology and subtype9,24,25 For example, in patients with NSCLC brain metastases, the presence or absence of mutations in EGFR (encod­ ing epidermal growth factor receptor) and translocations of ALK can influence survival26–28 Tumour subtype can also substantially affect prognosis for patients with breast cancer who have brain meta­stases24,25 However, to date, the clinical value of these prognostic factors have been limited — even with favourable prognostic factors, the diagnosis of brain metastases portends poor sur­ vival, with overall 2-year and 5-year survival estimates across all primary tumour types being 8.1% and 2.4%, res­pectively1 Highlighting their clinical importance, neuro­logical disease is the cause of death in up to 52% of patients with brain metastases29,30 Mechanisms/pathophysiology Addressing the poor prognosis of those with brain metastases requires an understanding of the disease complexity at a molecular level Although multiple hypotheses have been proposed to explain the unique NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); metastatic patterns of different primary cancers, includ­ ing the importance of ‘seed’ (or cancer cell) and ‘soil’ (or microenvironment of the receiving organ) factors or the variations in circulatory patterns between com­ mon primary and metastatic sites, evidence supports a dynamic interplay between metastatic cells and the tumour microenvironment that is crucial for growth after cell seeding To be successful, effective treatments for these heterogeneous diseases will need to adapt to the unique biological susceptibilities of each tumour type and to the molecular differences between the pri­ mary tumour and its metastases In addition, the tumour microenvironmental factors that limit the efficacy of treatments — irrespective of the relative sensitivity of the tumour cell itself — must be recognized The BBB and blood–CSF barrier The central nervous system (CNS) is protected by several functional barriers including the blood–brain barrier (BBB) and blood–cerebrospinal fluid (CSF) barrier The BBB has evolved as protective insulation for neuronal signalling This structure consists of endothelial cells with low transcytosis rates and high expression of efflux pumps that are connected by continuous tight junctions (Fig. 2) In addition, two basement membranes (endothe­ lial and astrocytic (parenchymal)), embedded pericytes and astrocytic terminal processes (also known as end­ feet) all further contribute to BBB functions Specific astrocyte functions critical to maintaining the BBB include coupling between endothelial cells and pericytes Primer a Endothelial basement Endothelial membrane cell Parenchymal basement membrane Perivascular space Postcapillary venule Capillary Astrocyte Pericyte Antigen-presenting cells CNS parenchyma b CNS parenchyma Ependymal cell Epithelial basal membrane Choroid plexus epithelium c Lymphatic vessel Dura Ventricle Arachnoid CNS parenchyma Pia RBC Stromal myeloid cells Choroid plexus stroma Interstitial fluid movement Perivascular space Fenestrated microvessel Fig | Central nervous system barriers a | The blood–brain barrier has evolved as protective insulation for neuronal signalling and comprises endothelial cells characterized by a low transcytosis rate and high expression of efflux pumps that are connected by continuous tight junctions In addition, two basement membranes (endothelial and astrocytic (parenchymal)), embedded pericytes and the endfeet of astrocytes all further contribute to its barrier functions At the level of postcapillary venules, there is a potential space between the basement membranes containing sparse antigen-​presenting cells b | The blood–cerebrospinal fluid (CSF) junction is formed by choroid plexus epithelial cells that are connected via tight junctions; the choroid plexus capillaries have fenestrations and intercellular gaps that enable the movement of molecules Circulating leukocytes rarely enter the brain parenchyma under homeostatic conditions, although endogenous parenchymal immune cells are present c | The brain lymphatic system provides a route for immune cells and proteins from the CSF to drain through lymphatic channels in the meninges to deep cervical lymph nodes CNS, central nervous system; RBC, red blood cell to neurons to form a neurovascular unit, ionic regula­ tion via endfeet ion and aquaporin channel expression and harbouring of protein transporters for glucose and glycoproteins Astrocytes also communicate directly via gap junctions and secrete growth factors such as vascu­ lar endothelial growth factor (VEGF) to promote endo­ thelial cell barrier function31 As a result of this barrier, without a specific transporter, only small uncharged compounds can diffuse through the BBB Despite these BBB restrictions, cell migration into the CNS can still occur, although the mechanisms behind this are not well understood In the past decade, however, studies have elucidated the ability of circulating monocytes to cross the BBB in response to homeostatic perturbations in the brain32, a mechanism that circulating tumour cells may exploit to gain entry into the brain The blood–CSF barrier, by comparison, is formed by choroid plexus epithelial cells that are connected via | Article citation ID: tight junctions, with choroid plexus capillaries having fenestrations and intercellular gaps that enable the free movement of molecules between these compartments33 Notably, expression of the complement protein (C3) by primary cancer cells can act to disrupt the blood–CSF barrier, enabling mitogens to enter the CSF34 This process provides an explanation for tumour cell egress and subsequent growth in cases of leptomeningeal metastases Seeding All metastases arise from the intravasation of tumour cells into the circulation, either directly or via the lym­ phatic system Tumour cell interactions with immune cells such as macrophages result in the formation of actin-​rich degradative protrusions on tumour cells, which promote cell motility via clearance of extracel­ lular matrix35 Once in the circulation, these circulating www.nature.com/nrdp (2019) 5:5 0123456789(); Primer tumour cells need to survive, undergo adhesive (circula­ tory) arrest within the vasculature and extravasate, either as single cells or emboli The dense microcapillary net­ work of the brain and high proportional blood flow leads to an increased exposure to circulating tumour cells Circulatory arrest is thought to be promoted by slowed movement at branch points in capillaries (~3–7 μm in diameter), as well as the larger size of tumour cells (up to 20 μm) compared with deformable red blood cells (7 μm)36,37 Although single-​cell studies indicate that the majority of arrested cells will not go on to form metastases, tumour cell clusters within the microvascu­ lature might act synergistically to promote metastatic growth38 Specific interactions between tumour cells and brain endothelia can also increase tumour cell adhesion and promote circulatory arrest For example, upregula­ tion of the membrane glycosyltransferase ST6GALNAC5 was shown to specifically mediate circulating breast cancer cell adhesion to the brain endothelium39 Once arrested, upregulation of genes encoding mitogenesis-​associated enzymes and growth factors, including COX2 and HBEGF (encoding a ligand for EGFR) in metastatic breast cancer cells, has also been linked to cell migration across the BBB (as well as metas­ tasis formation within the lung, presumably via a simi­ lar mechanistic pathway)39 Additionally, tumour cell and immune cell-​derived VEGF, which also increases endothelial permeability, works in combination with matrix metalloproteinases (MMPs) to promote vas­ cular growth and extracellular matrix destruction, key elements for metastatic tumour extravasation, seeding and micrometastasis formation40–44 Common signal­ ling pathways in tumour cells, involving, for example, the proto-​oncogene tyrosine-​protein kinase Src, also activate proteases to permeabilize the BBB45 and can be augmented by pathologically activated astrocyte release of pro-​angiogenic and growth-​promoting molecules such as MMP9 once in the brain microenvironment46 The proteinase cathepsin S, produced by macrophages and tumour cells, also provides proteolysis of junctional adhesion molecules between endothelial cells that assists with extravasation and colonization47 Once behind the BBB, tumour cells use many routes to spread, including migration along the outside of lepto­ meningeal and parenchymal blood vessels, between surfaces separating brain compartments and within the CSF48 Extravasated tumour cells can also lie dormant at metastatic sites for extended periods by entering a self-​ imposed slow cell-​cycling state and through expression of stemness-​associated transcription factors such as SOX2 and SOX9 (refs38,49) Rather than form metastases, circulating tumour cells can also self-​seed back into their tumour of origin, after being attracted by tumour-​derived cytokines such as IL-6 and IL-8 (ref.50) Most circulating tumour cells nonetheless likely die, even after successfully arresting and extravasating in an end organ38 The BTB After a metastasis forms, the remnants of the BBB are termed a BTB In model systems, permeability of the BTB to drugs or contrast agents was higher, but more heterogeneous, than that of the BBB, with only a minor NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); proportion of lesions having sufficient permeability to permit drug responses51,52 Similar findings have been reported using MRI53, which document that not all experimental lesions are permeable to the contrast agent gadolinium54 These data are supported by the observed heterogeneous uptake of preoperatively administered chemotherapeutic agents (capecitabine and lapatinib) into surgically resected brain metastases55 and dis­ appointing responses of brain metastases to systemically administered drugs in multiple clinical trials BTB permeability is a result of multiple processes Paracellular changes, including the loss of continuous endothelial cell adhesions by downregulation of pro­ teins, such as zona occludens (ZO1) and vascular-​ endothelial cell adhesion molecule (VE-​CAM), create abnormal molecular permeation channels In haemato­ genous models of breast cancer brain metastases, the BTB was also found to differ from the BBB by: the presence of swollen capillary endothelial cells; abundant VEGF; a prominent local neuroinflammatory milieu consist­ ing of activated microglia (resident monocytic cells of the brain that provide innate immunity), astrocytes and immune cells; an altered basement membrane compo­ sition; shifts in subpopulations of pericytes; and loss of astrocyte endfoot polarization of aquaporin channels56 Although the exact molecular cascades underlying each of these findings remain unclear, shifts in expression of the intermediate filament desmin and reduced expres­ sion of CD13 within pericyte subpopulations were found to be key markers for relative permeability changes within the BTB Other molecular alterations reported in BTB endothelial cells that contribute to increased permeability include altered expression of membrane transporters57, cytokine receptors such as the recep­ tor for tumour necrosis factor (TNF) and membrane proteins and growth factors including claudin and angiopoietin (refs58–60) Interestingly, in patients with breast cancer, tumour subtype-​specific variations in BTB endothelial cell function were also observed57, providing one potential explanation for the variable clinical course and responsiveness to treatments of breast cancer brain metastases from different subtypes Abnormal signal­ ling by other cells, such as expression of the receptor for the angiogenesis-​related molecule lysophospholipid sphingosine phosphate (S1P3) on neuroinflammatory astrocytes, has also been reported as a contributory fac­ tor to BTB permeability61 However, of all these findings, only alterations in the astrocytic basement membrane, pericyte subpopulations and S1P3 expression were appar­ ent between low-​permeability and high-​permeability experimental brain metastases56,61, supporting their importance in abnormal BTB permeability Colonization Tumour cell interactions with the brain microenviron­ ment govern the outgrowth of extravasated tumour cells as well as cancer-​specific signalling pathways that enable autonomous growth Brain metastatic coloniza­ tion remains incompletely understood, but preclinical models have provided insights Once on the abluminal side of the endothelium, the upregulation of integrins on tumour cells, such as αvβ3 integrin and β1 integrin, Primer promotes cell adhesion to the endothelium and pro­ vides an initial microenvironment for colonization37,44 Metastases, even at the micrometastatic stage, are sur­ rounded by a prominent neuroinflammatory response consisting of activated astrocytes (distinct from those whose endfeet contribute to the BBB), microglia and immune cells46,62 Interactions of tumour cells with neuroinflammatory cells and healthy brain parenchy­ mal components to promote metastasis are involved in colonization, accomplished by diverse pathways Activated astrocytes are an important partner For example, the formation of tumour cell–astrocyte gap junctions has been demonstrated in murine models of breast and lung cancer brain metastases; these struc­ tures enable transfer of metabolites, such as the second messenger cGAMP from tumour cells to astrocytes63 Ensuing astrocytic secretion of the inflammatory chemokines interferon-​α (IFNα) and TNF supports tumour growth and chemoresistance via activation of transcription and cell survival pathways63 In vitro studies corroborating the pro-​survival effects of the astrocyte– tumour cell connections have further identified upreg­ ulation of the survival genes GSTA5, BCL2L1 and TWIST1 as key mediators64 In a breast cancer model, ER-​expressing astrocytes have also been shown to be stimulated by premenopausal hormones to secrete tumour colonization-​promoting chemokines that upreg­ ulate EGFR signalling in tumour cells, ultimately leading to increased expression of S100 calcium-​binding protein A4 (S100A4), a promoter of cell motility and invasion65 The potential importance of ER signalling within reac­ tive astrocytes for the formation of brain metastases from other cancers, such as lung cancer, has also been demonstrated66 Additionally, in a tripartite interaction, astrocytes have been shown to secrete exosomes con­ taining microRNAs (miRNAs) that inhibit brain metas­ tasis expression of the tumour suppressor phosphatase and tensin homologue (PTEN), leading to increased tumour cell chemokine secretion and recruitment and activation of tumour-​promoting brain-​derived myeloid cells67 Alterations in microglia also occur in the tumour microenvironment, with tumour cells blocking micro­ glial cytotoxic activity by secretion of the signalling protein neurotrophin (ref.68) Given the presence of resident microglia and the lymphatic system, the brain is no longer thought to be immune-​privileged Additionally, peritumoural and intratumoural infiltration of antigen-​presenting micro­ glia and macrophages (expressing HLA-​DR), as well as rare cytotoxic CD8+ T cells have been described in cra­ niotomy specimens from patients with various primary tumours69; lineage tracing techniques confirmed the bone marrow origin of some of this infiltrate70 Tumour-​ infiltrating lymphocytes (TILs) have been described in brain metastases in stromal, diffuse and peritumoural patterns, with the highest infiltration of TILs (defined as CD3+ and CD8+) being in melanoma and RCC brain metastases71 T cell infiltration was further associated with white matter tract disruption and improved patient survival72 However, evidence of an immunosuppressive brain–tumour microenvironment includes reduced T cell heterogeneity (determined by sequencing of the | Article citation ID: T cell receptor-​β complementarity-​determining region 3) in lung cancer brain metastases compared with primary tumours despite a higher metastatic mutational bur­ den and a lack of preserved clonality between paired lesions73 This idea is supported by the expression of immunosuppressive checkpoint proteins such as pro­ grammed cell death (PD-1) and programmed cell death ligand (PD-​L1) in resected brain metasta­ ses from patients with various primary tumours71,74–77 Similarly, in animal models, immune checkpoint thera­ pies for brain metastases have been shown to depend on initial activation in extracranial disease78 Clinical data have also demonstrated that immune checkpoint therapeutic efficacy can be augmented by systemic immunological priming79 Additional tumour cell predispositions or adapta­ tions that assist with brain colonization include meta­ static cell metabolic enhancements in glucose oxidation, with concomitant increases in activation of the pentose phosphate pathway and the glutathione system that lead to a reduced production of reactive oxygen species80,81 The brain also has the highest vascular density in the body, and many reports using preclinical models sup­ port tumour cell co-​optation of existing vasculature for subsequent growth41,82,83, although contributions from de novo angiogenesis cannot be excluded Leptomeningeal metastases Leptomeningeal metastases represent a subset of brain metastases that grow in the lining of the brain or spine and/or in the CSF Lung and breast cancers and mela­ noma have the highest incidences of leptomeningeal metastases, which can occur with or without brain parenchymal metastases Prognosis of patients with lep­ tomeningeal involvement is very poor Mechanistically, in preclinical model systems both the COX2 and TGFβ signalling pathways (involved in cell proliferation, differ­ entiation and mitogenesis) were drivers of brain meta­ static tumour cell escape into the CSF84,85, suggesting a contributory role for these pathways in leptomeningeal metastasis formation Molecular signatures of brain metastases Whether brain metastases are molecularly similar to the primary tumours from which they arise was unknown until recently Next-​generation sequencing techniques have increased our understanding of the dynamic inter­ play of tumour cells and their microenvironment, with transcriptomic data from cross-​species microarray hybridization demonstrating a reprograming of meta­ static tumour cells to gain neuronal cell characteristics after growing in the brain microenvironment86 Another study demonstrated (through whole-​exome sequen­ cing of matched primary tumours and brain metas­ tases from a variety of solid tumours) that although clonally related primary tumour and brain metastasis pairs shared a common ancestor, a distinct evolution pattern occurred at the metastatic site87 This finding led to the identification of clinically relevant mutations in the brain metastases that were not detected in the primary tumours, including alterations in molecules and pathways involved in cell proliferation, growth www.nature.com/nrdp (2019) 5:5 0123456789(); Primer and survival, and the introduction of the idea that brain metastases may not share the driver mutations of the primary tumour For example, distinct mutations in CDKN2A and PIK3CA, loss of PTEN, amplifications of ERBB2 and activating proto-​oncogene KRAS mutations have been noted in metastases but not their respective primary tumours (Table 1), all of which have implications for treatment87 Additionally, brain metastases in a patient were more closely related to one another than to their matched primary tumours or to other extracranial meta­ stases These data suggest the importance of personal­ ized, brain metastasis-​specific therapies that address this unique molecular milieu Accordingly, an early-​ phase clinical trial across multiple solid tumours utilizing genomic information from both brain meta­ stases and the primary tumour to rationally guide tar­ geted therapy is in development through the Alliance in Oncology Similar unmatched analyses of brain metastases from multiple primary tumour types (including lung, breast and oesophageal cancer and melanoma) detected 26 significant gene mutations in brain metastases, pro­ viding insights into shared gene alterations common across brain metastases88 Identified genes included the tumour suppressor TP53, KRAS and DSC2, which encodes a cadherin involved in epithelial cell connectiv­ ity88 Gene network analyses have identified pathways involved in axonal guidance and angiogenesis signalling (that is, genes of the Slit–Robo pathway and the netrin, semaphorin and ephrin families) and cell metabolism (for example, genes along the EGFR and HER pathways) that are repeatedly altered in brain metastases88 In addi­ tion, although expression of the growth factor receptor HER3 was found to be high in brain metastases, tumour cell expression of its ligand neuregulin was low88 However, as the brain microenvironment is rich in neuregulin 1, this finding highlights specific adapta­ tions to the brain microenvironment in the tumour cell clones that successfully grow and proliferate to form the resulting brain metastasis Other studies have illus­ trated similar relationships between brain metastasis cells and the brain microenvironment, including epi­ genetic adaptations in metabolic processes83,89,90 These data highlight the importance of the brain–tumour microenvironment in metastatic tumour biology as well as important inherent molecular differences that exist between brain metastases and their source primary tumours (Table 1) Lung cancer brain metastases Genetic analyses have linked unique driver mutations with the development of lung adenocarcinoma, and gene expression profil­ ing has further identified distinct molecular subtypes, namely, bronchoid, squamoid and magnoid cancers, each with unique stage-​specific survival patterns and sites of metastases91,92 Among these drivers, mutations in EGFR and ALK gene rearrangement are of particu­ larly high importance owing to the high prevalence of brain dissemination of these entities and the clini­ cal avail­ability of targeted therapeutic inhibitors93,94 Table | Summary of molecular alterations in primary tumours and brain metastases Tumour typea Chromosomal alterations Genes amplified, overexpressed or activated Genes deleted, underexpressed or inactivated Primary NSCLC Deletions in 3p, 9p, and 17p; trisomy 7; and isochromosomes i(5)(p10) and i(8)(q10) AKT1, ALK, ERBB2, MET, FGFR1, PDGFRA, KRAS and EGFR PTEN and TP53 NSCLC brain metastases No data available EGFR and KRAS LKB1 Primary SCLC Gains on 3q, 1p, 1q and 14q; FGFR1, MYC and MYCN deletions in 3p, 5q, 10, 16q and 17p NOTCH family , TP53 and RB1 SCLC brain metastases No data available ANGPT4 and PDGFRB TGFB1 Primary breast cancer Gains in 1q and 14q; deletions in 1p, 1q, 3p, 11p, 13q, 17p and 22q AKT1, AKT2, CCND1, ERBB2, FGFR1, PDK1 and PIK3CA BRCA1, BRCA2, CDH1, CDKN2A, CDKN2B, LKB1, PTEN and TP53 Breast cancer brain metastases Gains in 1q, 5p, 8q, 11q and 20q; deletions in 8p, 10q, 17p, 21p and Xq AKT1, ATAD2, BRAF, DERL1, DNMT3B, EGFR, KRT5, MYC, PROM1, NEK2A and NES ATM, CDKN2A, CDKN2B, CRYAB, HSPB2 and PTEN Primary melanoma Gains in 1q, 6p, 7p, 7q, 8q, 11q and AKT1, BRAF, CCND1, CDK4, 17q; deletions in 3q, 4q, 6q, 8p, 9p, KRAS and PIK3CA 9q, 10p and 10q CDKN2A and PTEN Melanoma brain metastases No data available AKT1 and TBX2 PTEN, CDKN2A, CDH13 and PLEKHA5 Primary RCC Deletions in 3p, 6, 8, and 14 No data available BAP1, CDKN2A, MTOR, PBRM1, TP53 and VHL RCC brain metastases No data available PIK3CA CDKN2A and PTEN Primary CRC Deletions in 18q BRAF, KRAS and PIK3CA APC, TGFBR2 and TP53 CRC brain metastases No data available KRAS, NRAS and PIK3CA No data available CRC, colorectal cancer ; NSCLC, non-​small-cell lung cancer ; RCC, renal cell carcinoma; SCLC, small-​cell lung cancer aData are scant on subtype-​specific features across cancer types, except for SCLC and NSCLC NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); Primer (especially new tyrosine kinase inhibitors (TKIs) charac­ terized by a substantial penetration and activity in the CNS95,96) Molecular analyses have also identified genes not only associated with lung cancer formation and growth but also specific to the risk of developing brain metastases, such as the tumour suppressor LKB1 and KRAS97 Hyperactivity within the WNT signalling pathway (culminating in T cell factor signalling) has also been linked to lung adenocarcinoma brain (and bone) metastatic formation, acting through the tran­ scription factor mediators HOXB9 and LEF1 to stimu­ late tumour cell invasion and proliferation98 Analyses of small-​cell lung cancer (SCLC) brain metastases have demonstrated both unique (including upregulation of genes related to angiogenesis such as ANGPT4 and PDGFRB) and shared genetic alterations with NSCLC99 A topic of clinical interest is whether these pathways offer effective therapeutic targets for prophylactic treat­ ments against the risk of developing brain metastases following an initial diagnosis of non-​disseminated lung cancer Breast cancer brain metastases Gene expression analy­ ses have identified distinct molecular subtypes of breast cancer (luminal A, luminal B, HER2 positive and tri­ ple negative), each with unique metastatic patterns100; HER2-positive and triple-​negative tumours are most likely to develop brain metastases Important driver mutations in breast cancer include genes related to cell growth and proliferation (ERBB2, CCND1, CDKN2A and CDKN2B) as well as tumour suppressors (PTEN and TP53)101 Genomic profiling has further demonstrated unique chromosomal aberrations, copy number alter­ ations and methylation patterns in a variety of genes associated with breast cancer brain metastases, including known proto-​oncogenes such as BRAF102 The alterations vary by subtype, which may explain clinical differences in risk of brain metastasis For exam­ ple, although overall methylation patterns were increased across breast cancer brain metastases compared with primary tumours, lower methylation levels were observed in brain metastases from the triple-negative subtype100,103 A potential link between hypomethylation and metastatic invasiveness has been demonstrated in other cancer types such as CRC104 Similarly, genetic profiling of primary tumours in patients with symp­ tomatic brain metastases early in their disease course from HER2-positive breast cancer revealed a potential profile of 13 associated genes (refined to a three-gene classi­f ier of RAD51, HDGF and TPR — molecules involved in DNA repair and cell proliferation) that was predictive of brain metastatic risk105 The immune response and alterations in the cell proliferation phosphoinositide 3-kinase (PI3K)–AKT pathway were also shown to be significantly associated with overall survival outcomes in patients with breast cancer brain meta­stases74,106 Accordingly, increased microenviron­ mental expression of the membrane protein HER3 (with HER2–HER3 heterodimers being a strong inducer of the PI3K–AKT pathway) has been found in HER2amplified brain metastases models and may underlie resistance to PI3K inhibitors in this tumour subtype107 | Article citation ID: Improved understanding of these immune micro­ environment differences, as well as the inherent molec­ ular differences between brain metastasis subtypes, is already being leveraged for the development of subtype-​ specific therapies108,109 and will be of importance in developing novel targeted therapies Melanoma brain metastases Key driver mutations have been identified in the development of melanoma and involve CDKN2A, BRAF, NRAS and KIT110 Molecular profiling of both extracranial and intracranial mela­ noma metastases has identified similar mutational hot spots in BRAF and NRAS and in molecules involved in cell–cell adhesion and gene transcription, such as CTNNB1 (refs111,112) Therapies addressing these tumour-​specific mutations have already shown clinical promise, for example, in patients with BRAF-​mutant melanoma brain metastases113 Specific biomarkers of the PI3K–AKT pathway have also been shown to be preferentially increased in melanoma brain meta­stases, and loss of the pathway inhibitor PTEN has been found to correlate with shorter time to brain metastasis for­ mation and overall survival114 Overexpression of the membrane-​associated protein PLEKHA5 is another possible mediator of melanoma brain metastases, act­ ing potentially via increased cellular BBB transmigra­ tion and invasion and/or PI3K–AKT signalling115 These findings carry potential therapeutic implications given that brain-​permeable PI3K inhibitors are currently in clinical development116 The immune response to melanoma metastases is also an area of particular interest given that higher peritumoural immune infiltrate (particularly CD3+ and CD8+ T cells) and expression of immune-​related genes, such as those related to T cell receptor pathways, have been linked to improved survival117 At baseline, the overall immune response to brain metastases is low com­ pared with other metastatic sites118, providing additional opportunities for future studies of immunotherapies and immune-​modulation treatments in patients with melanoma brain metastases RCC brain metastases RCC is subdivided into three main histological cell types that include clear cell, papillary and chromophobe subtypes119 Mutational drivers are best characterized for the clear cell subtype, with inactivation of the tumour suppressor gene VHL being the most common mutation in both familial and sporadic clear cell RCC In addition, molecular subtypes associated with different clinical prognoses have been identified in those with clear cell RCC120 Although future studies are needed to characterize the papillary and chromophobe subtypes in more detail, mutations in the tyrosine kinase gene MET and TP53 have been described Comparisons of RCC brain metastases and primary tumours have identified mutations in the genes PTEN, PIK3CA and CDKN2A that are unique to brain metastases121 Whether these pathways are potential therapeutic targets for the goal of developing effective prophylactic treatments against the development of brain metastases warrants further study www.nature.com/nrdp (2019) 5:5 0123456789(); Primer a b c d e f * Fig | Imaging characteristics of metastatic brain lesions Neuroimaging of a patient reveals several of the typical radiographic features of metastatic brain lesions a | Multiple contrast-​enhancing mass lesions of varying sizes (white arrows) are shown on MRI throughout both hemispheres of the brain at grey–white and watershed locations with fairly well-​encapsulated appearances b | Evidence of significant surrounding vasogenic oedema (yellow arrows) and associated mass effect, brain compression and midline shift are evident on MRI c | Heterogeneous T2 characteristics including microcysts are shown on MRI d | Contrast-​enhanced MRI perfusion imaging indicates high perfusion (green–red) in areas of viable tumour compared with baseline brain perfusion (black–blue) e | Diffusion-​weighted MRI showing minimal diffusion restriction compared with the hyperintense signal typically seen in infectious and inflammatory processes f | Non-​contrast CT image showing the hyperattenuating mass lesions as hyperintensities (red arrows) relative to brain parenchyma and the hypoattenuating vasogenic oedema as hypointensities relative to the brain parenchyma (asterisk) This patient was subsequently discovered to have metastatic melanoma as the primary underlying diagnosis CRC brain metastases Extensive molecular analyses have also been performed on CRCs, resulting in four consensus molecular subtypes: CMS1 characterized by microsatellite instability and strong immune activation; CMS2 possessing WNT and MYC upregulation; CMS3 notable for metabolic dysregulation; and CMS4 defined by TGFβ activation, cellular invasion and angiogen­ esis122 Mutations in PIK3CA, KRAS and NRAS have also been associated with CRC brain metastases, with the RAS family mutations having added clinical importance owing to their association with resistance to anti-​EGFR therapies (such as cetuximab)121,123 Diagnosis, screening and prevention The clinical presentation of brain metastases is similar to the presentation of any space-​occupying intracra­ nial lesion associated with brain compression and mass effect124 Headache, which might be mild, is a presenting symptom in up to 50% of patients and is more common with patients with multiple or posterior fossa metas­ tases Papilloedema (optic disc swelling) is associated with headaches in 15–25% of patients Up to 40% of patients with brain metastases present with focal neuro­ logical deficits, such as weakness or numbness, and seizures occur in 15–20% of patients Another 5–10% of patients have an acute ‘stroke-​like’ onset of symptoms NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); owing to intratumoural haemorrhage (commonly in tumour types with higher risk of haemorrhage such as melanoma and RCC) Altered mental status or impaired cognition are also frequently experienced by patients with multiple metastases and/or increased intracranial pressure from mass effect, vasogenic oedema or alter­ ations in CSF flow Nevertheless, the signs and symptoms of brain metastases at presentation can often be subtle As a general rule, brain metastases should be suspected in any patient with known systemic cancer in whom neurological findings develop Advanced neuroimaging techniques can be used to identify brain lesions but are not specific enough for definitive diagnosis125 and, as a result, histopatho­ logical analysis of tissue harvested at surgical resection remains the gold standard for diagnosis On the basis of the authors’ experience, minimally invasive surgical resection of accessible lesions should be considered in patients with either an unknown primary tumour or well-​controlled systemic cancer, especially if the initial cancer diagnosis was remote (in the authors’ experience, ‘remote’ is >5 years after the appearance of the brain lesion) A high index of suspicion for brain metastases should nonetheless be maintained in the presence of an active or recently diagnosed primary cancer As such, clinicians may elect to proceed with treating a lesion in the brain as a presumptive metastasis in certain clinical settings without requiring a histological verification Neuroimaging Brain MRI with or without intravenous gadolinium con­ trast is the method of choice for the assessment of all brain metastases and in particular has a better sensitivity over contrast-​enhanced CT for lesions in the posterior fossa where bone artefact can occur, for multiple punc­ tate metastases and for leptomeningeal metastases126,127 Double or triple doses of gadolinium-​based contrast agent are superior to single doses, but increasing the dose may lead to an increase in false-​p ositive find­ ings128 Common radiographic characteristics of brain metastases on contrast-​enhanced MRI (Fig. 3) often include ring enhancement with prominent peritumoural oedema, watershed or grey–white junctional location, fairly spherical well-​encapsulated shape and presence of multiple lesions, although these findings are not specific to metastases As such, the differential diagnosis of brain metastases includes primary brain tumours (especially malignant gliomas and lymphomas) and non-​neoplastic conditions (abscesses, infections, demyelinating diseases and vascular lesions) Additionally, some brain metasta­ ses can display particular imaging characteristics that can complicate radiographic assessment For example, muci­ nous metastases can show low T2 signal intensities, and metastases from melanoma can demonstrate an abnor­ mally high signal on non-​contrast T1 images129; how­ ever, such signal characteristics may vary substantially and change over time owing to haemorrhage or accu­ mulation of melanin and paramagnetic ions Primary tumour subtype can also affect the imaging character­ istics of brain metastases; for example, triple-​negative breast cancer brain metastases can be substantially more necrotic and cystic than other subtypes130 Primer Advanced neuroimaging techniques are increasingly being used in the clinical evaluation of brain metasta­ ses Diffusion-​weighted (DW)-MRI can be helpful in evaluation of ring-​enhancing cerebral lesions, as diffu­ sion is commonly clearly restricted in abscesses but less restricted in brain metastases, although hypercellular tumour regions can demonstrate some diffusion restric­ tion compared with healthy brain tissue131 Additionally, restricted diffusion patterns can vary between brain metastases from different primary tumours and both restricted and unrestricted diffusion can coexist in fungal abscesses; accordingly, this technique is nonspe­ cific DW-​MRI has also been used to differentiate brain metastases from high-​grade gliomas (HGGs), although results for this application have been contradictory132 T2-weighted and fluid attenuation inversion recov­ ery (FLAIR) MRI sequences can also be used to help identify brain metastases as they are weighted to show vasogenic oedema as areas of increased signal intensity, but not all metastatic lesions have sufficient oedema to be identified The diagnostic utility of other imaging techniques focusing on peritumoural changes in perfusion and metabolism have also been explored Perfusion MRI may be able to detect lower cerebral blood volumes (CBVs) in peritumoural regions in brain metastases than in glioblastomas133–135 Conversely, the enhancing rim of brain metastases demonstrates a higher CBV than that of pyogenic abscesses136 The intravoxel incoherent motion (IVIM) technique estimates parameter values for diffusion and perfusion separately and, therefore, can potentially provide more accurate information for tumour diffusion characteristics, especially in hypervas­ cular brain tumours137 Similarly, lower choline:creatinine ratios have been reported on magnetic resonance spectroscopy in brain metastases than in HGGs138,139 Other metabolism-​based imaging techniques, such as 18F-​f luorodeoxyglucose (FDG)–PET and amino acid–PET (in particular with the tracer fluoroethyl​l-tyrosine), not provide sufficient differentiation between metastases and HGGs140,141 Scant data support the use of PET with radiolabelled choline for imaging of brain metastases142 identified or accessible extracranial sites have had a specimen collected87,143 Once the surgical specimen has been obtained, rou­ tine haematoxylin and eosin staining of formalin-​fixed, paraffin-​embedded tissue is performed for pathological review to confirm the neoplastic nature of the lesion and to make distinctions between metastases, malignant gliomas, meningiomas, lymphomas and other rare enti­ ties Further immunohistochemical diagnostic mark­ ers can aid in the detailed characterization of tumours and are particularly useful among poorly differentiated subtypes that are difficult to diagnose without molecu­ lar charac­terization In patients with a known primary tumour who have undergone brain biopsy, the histol­ ogy and marker profile of the primary tumour and the cerebral metastasis will usually show enough similari­ ties to provide a confirmatory diagnosis A histological comparison between these specimens is also potentially helpful in patients who have more than one cancer history, in which the metastasis could have originated from multiple potential primary sources Additionally, molecular analy­ses of metastases in relation to known primary tumours, when conducted, can aid in the iden­ tification of lineage markers and biomarkers with poten­ tial implications for treatment selection and clinical trial eligibility144,145 In patients with an unknown primary tumour, determining the lineage of the metastasis is impor­ tant for tailoring therapies of both the metastases and the primary tumour and begins with a basic morpho­ logical analysis (for example, differentiation between a carcinoma, lymphoma or melanoma) In addition, immunohistochemical profiles of metastases can indicate the site and lineage of the primary tumour146 These markers can overlap, and most are not specific to one tumour type For example, in cerebral adeno­ carcinoma of unknown primary, TTF1 (transcription termination factor 1) positivity is strongly associated with NSCLC and cancer of the thyroid, whereas nega­ tivity for the keratin CK7 and positivity for CK20 hints at CRC146 Neuroendocrine differentiation is assessed with staining for chromogranin, synaptophysin and antibodies directed against specific hormones (such as insulin, gastrin, glucagon, serotonin and somatostatin) Histopathology Further, immunohistochemical panels are available A tissue diagnosis of a brain metastasis should be pur­ that include cytoskeletal markers (such as vimentin, sued in patients in whom the primary tumour cannot desmin and S100) to define mesenchymal tumours be identified and should be considered in patients with Attempts to identify unknown primary tumours from a long-​term well-​controlled systemic cancer or active their metastases using RNA expression profiles are also systemic cancer and cerebral lesions with atypical ongoing and rely on our increasing understanding of radiographic characteristics (such as significant DW-​ the molecular differences in primary tumours and their MRI restriction)143 Given the availability of minimally cerebral metastases invasive surgery and stereotactic biopsies, there are Serum and CSF sampling are not routinely used in the fewer clinical scenarios in well-​resourced settings that histopathological diagnosis of brain metastases, Recently, justify irradiating ‘presumed brain metastases’ with­ however, a role for liquid biopsy of CSF has been sug­ out a histological diagnosis of the underlying cancer gested as a minimally invasive means to assess the pres­ In addition, owing to the increasing understanding of ence of key genetic alterations of cerebral metastases molecular differences between brain metastases and (including EGFR mutation in NSCLC, HER2 amplifica­ their extracranial paired samples, it will become more tion in breast cancer and BRAF mutation in melanoma) common to obtain surgical tissue from brain metastases as well as drug resistance mutations in patients with pro­ to gather the molecular data to guide clinical decision-​ gressive CNS disease despite treatment147 Prospective making, even when the primary tumour has been studies are needed to validate this technology 10 | Article citation ID: www.nature.com/nrdp (2019) 5:5 0123456789(); Primer Additionally, high interstitial fluid pressures and abnor­ mal local perfusion affect convective drug delivery to brain metastases174 Experimental approaches addressing these issues include convection-​enhanced delivery that Management uses stereotactically placed catheters and hydrostatic Current treatment of brain metastases typically includes pressure to deliver antitumour agents directly into the some combination of surgery (for tissue diagnosis, cere­ tumour bed The anti-​VEGF antibody bevacizumab bral decompression and prolonged survival when paired has also been used in combination with the HER2 with adjuvant radiotherapy in select cases) versus stand-​ inhibi­tors trastuzumab and lapatinib in a murine proof-​ alone radiotherapy and/or systemic medical therapies, of-concept model of HER2-amplified breast cancer with the overall goal of selecting the optimum treatment brain metastases to synergistically inhibit microvascu­ or treatments for an individual patient to maximize lar growth and enhance the direct cytotoxic effects of QOL and overall survival Although curative treat­ the anti-​HER2 agents, presumably via increased tumour ments remain elusive in most cases following a diagno­ penetration175, with this approach assessed in subsequent sis of brain metastases, increasing consideration is being clinical trials176 Multiple standard chemotherapeutic regimens have paid to anticipated patient survival, competing risks and long-​term toxicities when choosing among available also been evaluated for the most common cancers that cause brain metastases Cytotoxic agents that have treatment options demonstrated modest efficacy for NSCLC brain metas­ Medical treatment tases include combined cisplatin and pemetrexed177, Medical therapies for brain metastases can be divided cisplatin and vinorelbine178, paclitaxel and cisplatin179 into two broad categories of symptomatic manage­ and pemetrexed and cisplatin180 Temozolomide, an ment and tumour-​targeting therapies Corticosteroids, oral alkylating agent that has proven efficacy in gliomas, such as dexamethasone, represent the main sympto­ also showed initial promise in two small randomized matic treatment in addition to pain medications and trials181,182; however, further studies failed to demonstrate are often prescribed in response to signs of increased a significant benefit183–185 SCLC, with its propensity to intracranial pressure or symptomatic peritumoural metastasize to the brain, is associated with metastases oedema171 However, the beneficial effects of steroids that are more sensitive to standard chemotherapeutic are not permanent and a rapid taper is typically recom­ agents when combined with radiotherapy186; however, mended after symptoms are controlled to minimize such treatments not affect overall survival owing to drug-related adverse effects, including weight gain, the aggressive nature of the cancer For patients with breast cancer brain metastases, mood changes, slow wound healing and hyperglycae­ mia In addition, with increased understanding of the molecular subtypes influence systemic therapy choices role of immunosuppression in the pathophysiology For example, systemic therapies are a mainstay treatment of metastatic disease, the potential harms of steroid-​ for triple-​negative breast cancers187, with specific regi­ associated immunosuppression are increasingly impor­ mens tested for patients with breast metastases including tant; accordingly, efforts to minimize steroid exposure temozolomide (as a single agent)188,189 and capecitabine and consider alternatives to steroid therapy (for exam­ (a thymidylate synthase inhibitor)190 Studies of cispla­ ple, VEGF inhibition) are encouraged Prophylactic tin either with temozolomide188 or etoposide191 have also anti-​epileptics drugs (AEDs) such as phenytoin or shown promise Additionally, novel systemic agents such levetiracetam are often prescribed as an additional as sagopilone (an analogue to the microtubule inhibitor symptomatic agent in an attempt to decrease the sei­ epothilone B) and ANG-1005 (a peptide-​drug conjugate zure risk in patients with brain metastases However, of the chemotherapy paclitaxel and the cell-​penetrating recent meta-​analysis data not support this prac­ peptide angiopep-2) are also being developed192–194 By tice owing to the questionable preventive efficacy of contrast, for patients with ER-​positive or PR-​positive AEDs172 The remaining majority of medical therapies breast cancer brain metastases, hormonal therapies for brain metastases are tumour-​targeting treatments, such as tamoxifen, letrozole, anastrozole and megestrol which include both systemic chemotherapies as well as acetate are the therapeutic focus, with some evidence of molecular and immunotherapeutic approaches aimed clinical efficacy195–197 For patients with melanoma brain metastases, tradi­ at shrinking tumours or slowing their growth and pre­ venting or delaying tumour-​related symptoms Despite tional systemic therapeutic options are especially poor the promise of newer molecular therapies and immuno­ Specifically, multiple alkylating agents with CNS pene­ therapies, practical challenges in tissue sampling in tration have been assessed, such as fotemustine198 or temo­ patients with small, deep lesions or in those unable to zolomide (alone or in combination with lomustine)199,200, tolerate general anaesthesia must be considered when but all have demonstrated only limited clinical benefit choosing a therapeutic approach Targeted therapies Although the initial clinical testing Systemic chemotherapies Traditional systemic chemo­ of targeted therapies was largely performed in late-​ therapeutic agents have a limited role in the manage­ stage and/or pretreated patients and often excluded ment of brain metastases owing to the BBB, the BTB those with brain metastases, subsequent trials have and highly specialized transmembrane efflux pumps started to elucidate the potential utility of these agents that export these drugs from the brain 44,173 (Fig.  2) for patients with brain metastases (Table  2) From brain metastases undergoing radiosurgery, considering factors such as age, KPS, extracranial disease status and number of brain lesions169,170 12 | Article citation ID: www.nature.com/nrdp (2019) 5:5 0123456789(); Primer Table | Key clinical trials of targeted therapy in brain metastases by primary tumour Drug Trial arms (n) Intracranial response rate (%)a PFSb Overall survival 58 5.6 months 10.8 months Asymptomatic, BRAF V600E mutation, prior brain metastasis-​directed therapy (16) 56 7.2 months 24.3 months Asymptomatic, BRAF V600D, V600K or V600R mutation, ± prior brain metastasis-​directed therapy (16) 44 4.2 months 10.1 months Symptomatic, BRAF V600D, V600K or V600R mutation, ± prior brain metastases-​directed therapy (17) 59 5.5 months 11.5 months Asymptomatic, BRAF V600E mutation, no prior brain metastasis-​directed therapy (74) 39.2 16.1 weeks 33.1 weeks Asymptomatic, BRAF V600K mutation, no prior brain metastasis-​directed therapy (15) 6.7 8.1 weeks 31.4 weeks Asymptomatic, BRAF V600E mutation, prior brain metastasis-​directed therapy (65) 30.8 16.6 weeks 16.3 weeks Asymptomatic, BRAF V600K mutation, prior brain metastasis-​directed therapy (18) 22.2 15.9 weeks 21.9 weeks Ref Melanoma Dabrafenib Asymptomatic, BRAF V600E mutation, no prior brain metastasis-​directed plus trametinib therapy (76) Dabrafenib 228 113 Vemurafenib Symptomatic, BRAF V600 mutation, prior brain metastasis-​directed therapy (24) 37 4.4 months 5.3 months 310 Vemurafenib Asymptomatic and/or symptomatic, BRAF V600 mutation, no prior brain metastasis-​directed therapy (90) 18 3.7 months months 227 Asymptomatic and/or symptomatic, BRAF V600 mutation, prior brain metastases-​directed therapy (56) 20 3.9 months 9.5 months Asymptomatic and/or symptomatic, no prior radiotherapy ; WBRT + SRS (44) NR 8.1 monthsc 13.4 months Asymptomatic and/or symptomatic, no prior radiotherapy ; WBRT + SRS + temozolomide (40) NR c 4.6 months 6.3 months Asymptomatic and/or symptomatic, no prior radiotherapy ; WBRT + SRS + erlotinib (41) NR 4.8 monthsc 6.1 months Erlotinib Asymptomatic and/or symptomatic, ± prior surgery and/or radiotherapy ; WBRT + erlotinib (40) 86 NR 11.8 months 201 Gefitinib Asymptomatic and/or symptomatic, ± prior radiotherapy and/or chemotherapy (41)d 27 months months 202 Osimertinib Asymptomatic, EGFR mutation, no prior brain metastasis-​directed therapy (279) 80 18.9 months 83%e 205 Crizotinib Asymptomatic, ALK-​positive, ± prior radiotherapy , no prior TKI treatment (151) 50 10.4 months 84%f 95 Crizotinib Asymptomatic and/or symptomatic, ALK-​positive, ± prior radiotherapy , no prior TKI treatment (138) 29 9.8 months 86% 212 Alectinib Asymptomatic, ALK-​positive, ± prior radiotherapy and/or chemotherapy , prior crizotinib (50) 64 10.8 months NR Alectinib Asymptomatic and/or symptomatic, ALK-​positive, ± prior radiotherapy , no prior TKI treatment (152) 81 25.7 months 82%f 95 Brigatinib Asymptomatic and/or symptomatic, ALK-​positive, ± prior radiotherapy , no prior TKI treatment (137) 78 Not reached 85%f 212 Lapatinib plus capecitabine Asymptomatic and/or symptomatic, ERBB2 mutation, no prior brain metastasis-​directed therapy (45) 65.9 5.5 months 17 months 222 Neratinib Asymptomatic and/or symptomatic, ERBB2 mutation, prior brain metastasis-​ directed therapy (40) 1.9 months 8.7 months 223 Neratinib plus capecitabine Asymptomatic and/or symptomatic, ERBB2 mutation, prior brain metastasis-​ directed therapy (39) 49h NR 63%f 224 NSCLC Erlotinib f 183 311, g Breast cancer A selected list in which only trials with n ≥ 20 have been included ±, with or without; NR , not reported; NSCLC, non-​small-cell lung cancer ; PFS, progression-​free survival; SRS, stereotactic radiosurgery ; TKI, tyrosine kinase inhibitor ; WBRT, whole-​brain radiotherapy aDefined as the percentage of patients with a confirmed complete (disappearance) or partial (≥30% decrease in sum of diameters of target lesions) response assessed using RECIST criteria312 bProgression defined as ≥20% increase in sum of diameters of target lesions assessed using RECIST criteria312 cProgression defined as an increase in perpendicular bidimensional tumour area (>50% for lesions 1,000 patients with brain metastases from ALK-​ positive NSCLC, comparable intracranial response rates to ALK inhibitors were seen regardless of previous radio­ therapy and/or chemotherapy, or previous treatment with ALK inhibitors214 This last finding is especially relevant given the common clinical scenario of initial NSCLC treatment with crizotinib followed by devel­ opment of brain metastases However, in the context of diseases (such as NSCLC) with a high propensity of 14 | Article citation ID: brain dissemination and increasing survival from cancer, which increases the cumulative risk of brain metastasis, new standards of frontline care are needed to prevent or delay this clinically challenging scenario Of great inter­ est, through this use of highly active targeted therapies, WBRT can be safely postponed in asymptomatic EGFR-​ mutated or ALK-​positive NSCLC with brain metastases, diminishing the related toxic effects that can negatively affect QOL For patients with brain metastases from breast can­ cer, most of the targeted therapy trials have focused on HER2-positive patients215 Nonetheless, trastuzumab, a commonly used anti-​HER2 monoclonal antibody, has limited BBB penetration216 Although smaller molecules such as neratinib (a dual HER2 and EGFR inhibitor) are more likely to cross the BBB than trastuzumab, a dis­ cordant intracranial and extracranial drug response observed with both drugs suggests other drug resistance mechanisms might be responsible for lower intracranial activity217–219 The TKI lapatinib also has limited activity as a single agent and has, therefore, mainly been used in combination with capecitabine220–222 In phase II studies, this treatment regimen demonstrated response rates of 20% in patients who progressed after radiotherapy and 66% in radiotherapy-​naive patients220–222 Although neratinib demonstrated limited efficacy as a single agent for the treatment of HER2-positive brain metastases223, more promising results were obtained when combined with capecitabine224 ONT-380 and tesevatinib are other anti-​HER2 agents that are being studied in early-​phase trials225,226 For patients with ER-​positive or PR-​positive breast cancer brain metastases, preliminary studies have reported adequate CNS penetration and potential efficacy of oral selective inhibitors of cell cycle proteins CDK4 and CDK6 (such as abemaciclib)215 Targeted therapies have also shown promise for the treatment of melanoma For patients with brain meta­ stases from BRAF-​mutant melanoma, use of the BRAF inhibitors vemurafenib and dabrafenib as single agents has demonstrated response rates between 20% and 38%113,227, with higher responses in radiotherapy-​naive patients Combination molecular therapies, such as dab­ rafenib and the MAPK pathway inhibitor trametinib, have also been explored and demonstrated even higher intracranial response rates (55%) in BRAF-​mutant brain metastases228 Despite these high initial response rates, the main current limitation of these agents is a fairly short response duration229, with PFS often 12 months)236 In the second-​line, phase III NSCLC OAK trial com­ paring atezolizumab with docetaxel, the magnitude of the PFS benefit of atezolizumab in patients with asymptomatic controlled brain metastases was similar to the group without brain metastases as well as to the intention-​to-treat population237 Combined regimens of ipilimumab and nivolumab in patients with mela­ noma brain metastases have demonstrated even higher intracranial response rates approaching 60%238–240 In the most recent of these trials, the intracranial res­ ponse rates with this dual regimen were comparable to extracranial response rates and sustained for a period of at least months240 Overall survival in this trial was 81.5% at year and ~70% at years — representing a drastic improvement over the median overall survival of 4–5 months in patients with melanoma brain metastases from the pre-immunotherapy era Inhibitors of PD-​L1 (for example, MEDI4736) have also been developed and are being tested in clinical trials for patients with brain metastases (for example, NCT02669914) Combined regimens of SRS (which pre­ cisely focuses radiation around the lesion from different angles to avoid damage to nearby healthy brain tissue) and/or WBRT with immunotherapies have also been explored and seem to display synergism in response rates to the dual therapies241,242 Although this approach was found to be safe overall in small retrospective studies, 15 Primer the risk of radiation necrosis — especially in patients with prolonged survival — will need to be assessed in future studies241,242 Interestingly, the efficacy of immu­ notherapies in a murine model of melanoma brain metastases seems to be linked to the presence of extrac­ ranial tumours, which are postulated to overcome the decreased antigen response to CNS triggers and lead to peripheral expansion of effector CD8+ T cells and their increased recruitment to intracranial tumour sites78 Systemic delivery of oncolytic reovirus has also been shown to immunologically prime brain meta­stases by upregulation of the PD-1–PD-​L1 axis, ultimately leading to enhanced efficacy of PD-1 immunotherapy in a murine model79 Clinical correlation of these data is nonetheless needed, especially in light of the recent high intracranial response rates observed with dual checkpoint inhibitor therapies240 A better understanding of the regulatory mechanisms underlying immune infiltration and activation will likely lead to improved incorporation of immune therapies into the therapeutic landscape for brain metastases For example, regarding immune checkpoint inhibitors, it is not well characterized whether the antibodies penetrate the lesions or reprogramme immune cells systemically243 Another issue that requires resolution is that of pseudo­ progression, a potentially fatal intense inflammatory response that can mimic rapid tumour progression244 Although the optimal diagnosis and management strat­ egies for pseudoprogression remain active areas of study, immune depression with a course of steroids is often pursued Systemic toxicity with immunotherapies is also a concern, with mild toxicity rates in phase II studies as high as 50%, although rates of grade or toxicities were largely 2 cm, performed in combination with adjuvant radiother­ apy252 Critically, even in contemporary series, surgery alone has been shown to be insufficient for local tumour control253, and the use of adjuvant radiotherapy, and in particular SRS to the resection cavity bed, is of impor­ tance (see below) Additionally, it is increasingly neces­ sary to obtain surgical tissue from brain metastases to obtain molecular data to guide clinical decision-​making, even when the primary tumour has been identified or accessible extracranial sites have been sampled, owing to molecular differences between primary and metastatic lesions Furthermore, as newer targeted therapies and immunotherapies are developed, continual reassess­ ment of the advantages of surgical resection is needed as www.nature.com/nrdp (2019) 5:5 0123456789(); Primer patients now benefit from increasing medical treatment options and have longer overall expected survival Continued innovation in neurosurgical techniques to minimize incisions, reduce discomfort and expedite recovery times will likely further increase the observed benefits of surgical resection Newer techniques are also being explored that not require a craniotomy, includ­ ing stereotactic laser ablation (which involves inserting a small laser catheter through a burr hole), convection-​ enhanced delivery (which involves placement of cathe­ ters through burr holes for intratumoural therapy) and focused ultrasound (which involves no surgery at all but performs a noninvasive transcranial focused stereotactic ablation) Stereotactic laser ablation, or laser interstitial thermal therapy (LITT), has already been used to safely treat deep lesions or tumours that have recurred follow­ ing SRS254–257 Future indications for patients with brain metastases are likely to increase Radiotherapy In addition to the roles of surgery and systemic therapies in the management of select patients with brain metas­ tases, radiotherapy remains an important cornerstone of treatment in most patients (Fig. 5) Historically, WBRT was the standard treatment for most patients with brain metastases because it could be initiated quickly, was widely available, provided symptom palliation and treated both visible and occult lesions Although WBRT still remains the most commonly used treatment for patients with brain metastases258, especially among those with multiple lesions, radiotherapy has increas­ ingly moved towards focused radiation techniques, such as SRS, in recent years The continued use of WBRT is likely due in part to an early phase III trial that found WBRT following surgery superior to surgery alone in terms of local tumour control and preventing neuro­ logical death (although no overall survival benefit was demonstrated)259 Over the past two decades, however, as the cognitive adverse effects of WBRT have become increasingly recognized (including fatigue, somnolence and learning and memory impairments), SRS alone has become the favoured treatment approach for patients with a limited number of brain metastases The tech­ nique can deliver targeted, conformal treatment to the metastasis alone in a single session SRS also results in substantially less cognitive dysfunction and fatigue260 and can be delivered with minimal delay in secondary systemic therapies261 For patients with a single brain metastasis, level I data suggest that adding SRS to WBRT is associated with a significant survival advantage over WBRT alone262 The addition of SRS to WBRT has also been suggested to con­ fer a survival benefit for patients with a good prognosis (GPA 3.5–4.0) and up to three brain metastases263 Three landmark randomized trials compared outcomes of patients with a limited number of brain metastases (four or fewer) who received SRS with or without WBRT and found remarkably similar findings264–266 Importantly, adding WBRT to SRS does not improve patient survival, and the risk of cognitive dysfunction is lowered substan­ tially when WBRT is withheld However, higher rates of new brain metastases (that is, distant brain failure) NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); a b Fig | Minimally invasive neurosurgical techniques Minimally invasive ‘keyhole’ techniques have developed from advances in neuronavigation (Box 1) and provide a precise and personalized surgical approach Traditional neurosurgical incisions are large (black dotted line in panel a) and follow stereotyped patterns that enable wide craniotomies (black dotted line in panel b) and brain exposure Since the advent of neuronavigation, newer keyhole incisions (green dotted line in panel a) can be used that are a fraction of the length and require only a small keyhole craniotomy (green dotted line in (panel b) These techniques minimize central nervous system exposure and can accelerate postoperative recovery , shorten length of stay and enable faster initiation of adjuvant therapies are observed among patients receiving SRS alone, and although local control of individual brain metastases is high (~75%) after SRS alone, it improves further with the addition of WBRT On the basis of a thorough anal­ ysis of these data, US and European guidelines have recommended against the addition of WBRT to SRS for patients with limited brain metastases143,267, largely in an effort to avoid the learning and memory deficits associated with WBRT266 Evidence also suggests that SRS alone versus SRS and WBRT for patients with up to three brain metastases is cost-​effective268 Whether there may be some scenarios in which combination treatment could be beneficial warrants further study269,270, such as when frequent salvage therapy for new brain metasta­ ses after SRS alone is anticipated, which could increase treatment-​related adverse effects and/or costs from additional imaging and therapies271 Patient involve­ ment in decision-​making regarding treatment selection is important to help understand their priorities, with one survey study finding that a majority of patients (>90%) preferred to receive SRS alone if it was an option on the basis of individual valuations of QOL, ability to maintain functional independence and the overall influence of the treatment options on their survival272 There are two groups of patients who merit addi­ tional consideration First, for patients with very poor prognoses from brain metastases, a large randomized trial of patients with NSCLC demonstrated that WBRT was not associated with an improvement in overall sur­ vival compared with dexamethasone alone273 Second, in patients with SCLC with limited but stable or treatment-​ responsive extracranial disease, an overall survival bene­ fit from prophylactic cranial irradiation (PCI) has been demonstrated in early studies274,275 However, more 17 Primer a b c High dose Low dose d e Isodoses (cGy) 1,800 1,700 1,600 1,400 1,200 800 500 Fig | Evolution of radiation-​based treatments of brain metastases Advances in treatment delivery platforms from whole-​brain radiotherapy (panel a) and 3D conformal radiotherapy (in which the radiation beams are delivered to match the shape of the tumour ; panel b) to stereotactic radiosurgery (SRS; panels c–e) have made treatment for high numbers of brain metastases more technically feasible SRS involves the precise focusing of radiation around the lesion from different angles to avoid healthy tissues and provides a more confined area of irradiation than 3D conformal radiotherapy Panel d highlights the sample beam arrangement, whereas panels c and e demonstrate the resulting focal, high-​dose radiation delivered with SRS Hypofractionated SRS, which typically entails 3–5 treatments, is increasingly being used for brain metastases near tissues more sensitive to radiotherapy , such as the optic nerves and chiasm and the brainstem Early evidence suggests that this approach provides good tumour control with a favourable toxicity profile recent data have called the benefits of PCI into question in patients with extensive-​stage disease and the absence of disease progression upon initial chemotherapy, with one randomized trial finding no survival benefit for PCI compared with MRI surveillance276 Although these two strategies in extensive SCLC disease might never be for­ mally compared in the future, an open discussion with patients is needed in this clinical setting Strategies min­ imizing adverse effects of PCI, including hippocampal sparing, are under prospective evaluation (for example, NCT02635009) The survival benefit of PCI versus active surveillance for patients with limited-​stage SCLC is also currently unknown and would require a dedicated study Recent advances in radiotherapy for brain metasta­ ses include an expanding role for SRS as well as meth­ ods to reduce the risk of the neurocognitive sequelae of WBRT A prospective observational Japanese study among nearly 1,200 patients with ≤10 newly diagnosed brain metastases treated with SRS alone found that median survival was similar in patients with 5–10 versus 2–4 brain metastases (10.8 months in both groups)277 Similar toxicity rates, neurocognitive function, neuro­ logical death, new brain metastases and need for sal­ vage WBRT were noted between the groups, with only modestly higher risk of leptomeningeal dissemination among patients with 5–10 brain metastases277 Other 18 | Article citation ID: studies have shown variable effects of SRS on the devel­ opment of leptomeningeal disease253,278 Ongoing trials are now examining outcomes after SRS alone for patients with four or more brain metastases in a randomized fashion (NCT01592968 and NCT02353000) Advances in treatment delivery platforms have expanded the application of SRS to patients with large numbers of brain metastases279 Hypofractionated SRS, which typically entails 3–5 treatments, is increasingly being used for brain metastases near tissues more sen­ sitive to radiotherapy, such as the optic nerves and chi­ asm and the brainstem (Fig. 5) Early evidence suggests this approach has a favourable toxicity profile, with only 5% of patients treated with fractionated SRS devel­ oping radiation necrosis280, which is lower than might be anticipated with single-​fraction SRS for large lesions, although prospective data are lacking In addition, headframe-​based SRS is increasingly being replaced by frameless image-​guided SRS, facilitating greater patient comfort and flexibility of scheduling Importantly, no differences in adverse effects, treatment outcomes or accuracy have been documented between frameless and frame-​based systems281 In an effort to avoid postresection WBRT, data from randomized studies have demonstrated improved neuro­cognitive preservation with SRS of the resection www.nature.com/nrdp (2019) 5:5 0123456789(); Primer cavity versus postoperative WBRT282 However, local control at the resection cavity is ~20% lower after cavity SRS than after WBRT, potentially owing to inconsist­ encies in delineating the postresection target for SRS, leading to recent consensus contouring guidelines for SRS in this setting283 Specifically, these guidelines rec­ ommend that clinical target volumes for the resec­ tion cavity for postoperative SRS include the entire gadolinium-​enhanced surgical cavity, the surgical tract and tailored dural, bone flap and venous sinus mar­ gins on the basis of preoperative tumour contacts283 Other strategies to reduce the risk of neurocognitive dysfunction after radiotherapy include using intensity-​ modulated radiotherapy, which uses linear accelerators to contour radiation to a targeted area while minimiz­ ing the dose to surrounding healthy tissue, to spare the hippocampus during WBRT treatments This approach has shown encouraging phase II data with less neuro­ cognitive decline at months when compared with a historical control group284 and is currently being evalu­ ated in randomized trials (for example, NCT02753790) Another large randomized controlled trial has reported that patients starting WBRT who received the N-​methyl-d​-aspartate glutamate receptor blocker memantine (shown to be effective in treating vascular dementia from glutamate excitotoxicity) for months had a modestly longer time to cognitive decline (HR 0.78, P = 0.01) than patients on placebo285 Memantine is now being used clinically in some patients receiv­ ing WBRT in an attempt to protect against potential glutamate-​induced excitotoxicity from radiation Quality of life Cancer that has metastasized to the brain exerts a tremen­ dous disease burden on affected patients Receiving a diagnosis of brain metastases can negatively affect a patient’s QOL independent of any subsequent treat­ ments they receive The adverse effects associated with treatment for brain metastases may also seriously affect a patient’s life by limiting their ability to perform daily activities, including work, and by altering neurocogni­ tive function This loss of ability to continue with routine activities, coupled with treatment-​related adverse effects and an overall poor prognosis, substantially reduces QOL for both the patient and their caregivers Despite much effort in assessing the survival ben­ efits associated with various treatment modalities for patients with brain metastases, QOL data on this topic are limited No study has looked exclusively at QOL in patients with brain metastases outside of the context of treatment Furthermore, most of the data assessing QOL effects of various treatments for brain metastases come from cohort studies using heterogeneous patient populations, including a mixture of primary cancers and treatment modalities, and using variable QOL assessment scales and methodologies, with limited follow-up periods286–295 Of the nonrandomized studies examining WBRT monotherapy for brain metastases286–293, only one found a significant improvement in QOL following WBRT, with improvement in 3-month post-​treatment Spitzer Quality of Life Index (SQOLI) for daily living (P = 0.0029), NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); health (P 50% of patients (P = 0.0001 and P = 0.004, respectively) In the other studies, which used alternative QOL assessments including the Functional Assessment of Cancer Therapy–Brain score286,287, the Edmonton Symptom Assessment System293, the Brain Symptom and Impact Questionnaire293, serial KPS295 and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire QLQ-C15 PAL289,294, among others, QOL was reproducibly worse following WBRT Major parameters negatively affected by WBRT in these studies included deterioration of physical or cognitive function, fatigue, appetite loss, drowsiness, hair loss and weakness One phase III randomized trial also recently evalu­ ated the effect of WBRT for brain metastases on QOL273 In this work, patients with NSCLC brain metastases who were unsuitable for surgical resection or SRS were ran­ domly assigned to receive supportive care including dexa­ methasone (a corticosteroid) with or without WBRT The primary outcome measure was quality-​adjusted life-​years (QALYs), generated from overall survival and patients’ weekly completion of the EuroQOL five dimen­ sions questionnaire Similar to the above cohort data, this study reported a negligible difference in QALYs and no significant difference in survival or QOL between the two groups, providing further evidence that WBRT monotherapy has limited QOL benefit Given the negative neurocognitive effects of WBRT, others have examined the impact of SRS alone, SRS versus WBRT or SRS versus combined SRS and WBRT regimens on QOL In one randomized controlled trial involving patients with 1–3 brain metastases, SRS versus SRS and WBRT resulted in less cognitive deterioration and a higher QOL (as measured by the Functional Assessment of Cancer Therapy– Brain tool) at months260 These findings are supported by data from cohort studies in which SRS alone or in combination with WBRT was found to have a positive impact on QOL (as determined by KPS and SQOLI) when compared with WBRT alone295,296 Importantly, combined SRS and WBRT regimens in these works did not significantly improve survival or QOL compared with SRS alone260,295 These data highlight the superior­ ity of SRS for neurocognition and QOL compared with WBRT as well as the importance of QOL standardization in future works Providing insight into the potential importance of patient-​related variables on QOL, one study assessing the physical activity of patients undergoing WBRT for brain metastases demonstrated that sedentary patients had a worse QOL and were more depressed than more active patients297 An interesting publication also demonstrated that in patients with EGFR-​mutated NSCLC treated with first-​generation TKIs, those with brain metastases experienced more symptoms, including fatigue, nausea and vomiting Similarly, the presence of brain metastases resulted in significant increases in total medical costs In another trial focusing on ALK-​positive NSCLC, brain metastases were associated with a substantial economic burden However, the use of brain-​penetrating TKIs, 19 Primer Immune checkpoint inhibitors (NCT02621515 and others) Pathways associated with trials Oestrogen Macitentan Endothelin receptor Translational pathways Microglia Mannosylated clodronate liposomes Infiltrating lymphocytes Meclofenamate (NCT02429570) BDNF Reactive astrocyte Tumour cell Astrocyte–carcinoma gap junctions IL-6 TrkB Astrocyte Basement membrane foot Text Pericyte Text Vesicles Blood Gap junction Angiogenesis Transcytosis Angiopoetin Anti-angiogenesis drugs (for example, sorafenib (NCT01724606) and bevacizumab (NCT02185352)) Text Integrins Endothelial cell Efflux pump Receptor Paracellular permeability Reactive oxygen species RRx-001 + radiotherapy (NCT02215512) Lapatinib loaded BSA-coated nanoparticles LRP-coated liposomes Integrin-binding tumour-homing peptides GRN1005 (NCT01497665) Polar drugs active in tumour pathways Nonpolar drug diffusion Nanoparticles Certain PI3K inhibitors Certain cMET inhibitors Temozolomide  T-DM1 (NCT03190967) Fig | Selected potential targets in the blood–tumour microenvironment for future therapies Some approaches have progressed to clinical trials (yellow boxes) whereas others are still in preclinical development (green boxes) Clinical trials for anti-​angiogenesis agents include NCT02185352 and NCT01724606 Tumour-​homing peptides include iRGD (sequence: CRGDKGPDC), which contains the integrin-​binding RGD motif RRx-001 is a radiosensitizer being developed BDNF, brain-​derived neurotrophic factor ; BSA , bovine serum albumin; LRP, low-​density lipoprotein-​receptor related protein; PI3K , phosphoinositide 3-kinase; TrkB, tropomyosin receptor kinase B namely, alectinib, was estimated to significantly reduce related costs by preventing or delaying the occurrence of brain metastases compared with crizotinib Both trials strongly suggest using the most potent CNS drugs as soon as possible in the course of the disease298,299 The landscape of cancer treatments has dramatically changed Therapeutic strategies including personalized targeted or immunotherapeutic agents are available, and patients are living longer Under these circumstances, management of brain metastases has become increas­ ingly important; this is translating into clinical studies becoming more inclusive to patients with brain metas­ tases Nonetheless, major gaps in the literature remain and future studies are needed to assess QOL with newer treatment protocols including minimally invasive sur­ gery, radiosurgery, targeted therapies and immuno­ therapies for patients with CNS dissemination — across diseases as well as for every cancer subtype Outlook Despite the currently poor prognosis associated with brain metastases, our fundamental understanding of the biology of this disease is rapidly increasing As these dis­ coveries move from the realm of basic science to preclini­ cal and clinical applications, future outcomes for patients with brain metastases are almost certain to improve 20 | Article citation ID: Although the prevalence of brain metastases is con­ centrated in patients with lung cancer, breast cancer and melanoma, the incidence from other cancers is expected to continue to increase as upfront treatment options improve and patients are living longer after their initial cancer diagnosis From a clinical standpoint, a multi­ disciplinary team-​based approach is essential, including the consistent tracking of multiple types of data, such as longitudinal molecular and tumour microenvironment data (when feasible), as well as neurocognitive and QOL outcomes and overall survival data Multiple strategies are being pursued that leverage our increasing understanding of brain metastasis biology (Fig. 6) To address the low and heterogeneous permeabil­ ity of brain metastases to therapeutic agents due to the BTB, targeted carrier systems that selectively increase drug penetration are being pursued; additionally, iden­ tification and testing of novel agents with increased BTB permeability are being examined300 One example of this work is the second-​generation ALK inhibitor ceritinib, which is used in NSCLC and achieves increased brain permeability; preliminary clinical testing has shown rates of response and disease progression that are com­ parable to those at systemic sites301 The emerging field of theranostic nanomedicine, which describes the develop­ ment of nanoparticle platforms that combine diagnostic www.nature.com/nrdp (2019) 5:5 0123456789(); Primer and therapeutic capacities within a single system, also has promise for the future treatment of brain meta­stases In one pertinent study, semiconductor nanocrystals, which also function as luminescence probes, were com­ plexed with inhibitory RNA to modulate the perme­ ability of the BBB302 Bioactivation and cell targeting of nanocrystals are also possible303, raising the possibility of an externally controllable, dual-​function drug delivery and diagnostic platform Research into the modulation of other areas of brain metastasis biology is also occurring, including the brain–tumour microenvironment For example, strate­ gies focusing on the brain–tumour microenvironment include targeting various steps in metastatic cell coloniza­ tion and early tumour growth such as integrin, MMP and VEGF function304 Additionally, intriguing new mecha­ nistic pathways are also being explored for therapeutic purposes, including oestrogen blockade to decrease the formation of triple-​negative breast cancer brain metas­ tases through the inhibition of ER-​positive astrocytes in the brain microenvironment65 In preclinical models, these therapies act to prevent new metastasis formation, an appealing strategy that goes beyond just treatment of the existing lesions Aside from the obvious appeal of prevention versus treatment, this approach bypasses many of the real-​world challenges of molecular targeting strategies for brain metastases that relate to the difficulty of tissue sampling in patients with small or deep lesions or those too sick to tolerate general anaesthesia Despite these challenges, incorporation of molecular subtyping of brain metastases is likely to lead to impor­ tant future therapeutic developments given the varied clinical behaviour of brain metastases As the mutational signature of a cancer can differ between the primary and brain metastatic site, and can evolve over time in response to treatment, it will be important to continue to safely obtain specimens in a minimally invasive man­ ner to better understand the genetics of brain metasta­ ses over time and to inform clinical decisions regarding selection of targeted therapies Progress in molecular imaging and ‘liquid biopsies’ (using peripheral blood or CSF)305 would be an important advance that also bypasses many of the challenges associated with direct tissue sampling Similarly, advances in the contouring and hypofractionation of radiotherapy will be especially helpful in patients unable to be treated with molecular-​ based therapies, as will future randomized data on the efficacy of SRS versus WBRT for patients with four or more brain metastases306 Hall, W A., Djalilian, H R., Nussbaum, E S & Cho, K H Long-​term survival with metastatic cancer to the brain Med Oncol 17, 279–286 (2000) Nayak, L., Lee, E Q & Wen, P Y Epidemiology of brain metastases Curr Oncol Rep 14, 48–54 (2012) Tabouret, E et al Recent trends in epidemiology of brain metastases: an overview Anticancer Res 32, 4655–4662 (2012) Barnholtz-​Sloan, J S et al Incidence proportions of brain metastases in patients diagnosed (1973 to 2001) in the Metropolitan Detroit Cancer Surveillance System J Clin Oncol 22, 2865–2872 (2004) Tsukada, Y., Fouad, A., Pickren, J W & Lane, W W Central nervous system metastasis from breast carcinoma Autopsy study Cancer 52, 2349–2354 (1983) Recent standard-​of-care treatments for brain metas­ tases have evolved rapidly, particularly relating to the decline of WBRT and the rise of fractionated SRS and increased use of targeted therapies and immunother­ apies However, more data on patient outcomes using these newer care regimens, stratified by cancer and molecular subtypes, are first needed to be able to for­ mulate end points in next-​generation clinical trials Future trial designs and end points will also need to evolve to keep up with the moving target of outcomes for brain metastasis therapies307 To guide future trials, the Response Assessment in Neuro-​Oncology (RANO) working group for brain metastases (RANO-​BM) and leptomeningeal metastases (RANO-​LM) published guidelines for the assessment of both systemic and local therapies for brain and leptomeningeal metasta­ ses150,308,309 These works provide recommendations for tailored end point selection for, for example, early-​phase trials identifying a therapeutic signal to aid future devel­ opments and late-​phase trials confirming a therapeutic impact, and for how to optimally include patients in early-​phase systemic agent trials on the basis of the likelihood of CNS activity of the studied agent Although few drugs to date have shown the ability to shrink established brain metastases in trials, immuno­ therapies have recently shown promise for this target As therapies targeting microenvironmental modulation enter the clinical pipeline, the prevention of brain metas­ tases will also need to be considered as a primary end point Although this rapid evolution may seem daunting for trial design, data acquisition on new targets can be collected in phases For example, to begin to assess brain metastasis prevention as a primary outcome, second­ ary prevention trials can first be performed, whereby the prevention of additional brain metastases in patients with limited, treated brain lesions is tracked, providing an initial signal of efficacy before moving to larger, more expensive primary prevention trials Future clinical trials focusing on tumour subtype will also need to meticu­ lously record and report molecularly stratified outcomes as these data will be critical to the development of indi­ vidualized therapies in patients in whom molecular analy­ ses are possible As these discoveries move from the realm of basic science to preclinical and clinical appli­ cations, with improved clinical trial designs, the future outcomes for patients with brain metastases are almost certain to improve Published online xx xx xxxx Posner, J B & Chernik, N L Intracranial metastases from systemic cancer Adv Neurol 19, 579–592 (1978) Percy, A K Neoplasms of the central nervous system: epidemiologic considerations Neurology 20, 398–399 (1970) Siegel, R L., Miller, K D & Jemal, A Cancer Statistics, 2017 CA Cancer J Clin 67, 7–30 (2017) Sperduto, P W et al Diagnosis-​specific prognostic factors, indexes, and treatment outcomes for patients with newly diagnosed brain metastases: a multi-​ institutional analysis of 4,259 patients Int J Radiat Oncol Biol Phys 77, 655–661 (2010) 10 Berghoff, A S et al Descriptive statistical analysis of a real life cohort of 2419 patients with brain metastases of solid cancers ESMO Open 1, e000024 (2016) NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); 11 Nieder, C., Spanne, O., Mehta, M P., Grosu, A L & Geinitz, H Presentation, patterns of care, and survival in patients with brain metastases: what has changed in the last 20 years? Cancer 117, 2505–2512 (2011) 12 Martin, A M et al Brain metastases in newly diagnosed breast cancer: a population-​based study JAMA Oncol 3, 1069–1077 (2017) 13 Toyokawa, G., Seto, T., Takenoyama, M & Ichinose, Y Insights into brain metastasis in patients with ALK+ lung cancer: is the brain truly a sanctuary? Cancer Metastasis Rev 34, 797–805 (2015) 14 Yamanaka, R Medical management of brain metastases from lung cancer (review) Oncol Rep 22, 1269–1276 (2009) 15 Steeg, P S., Camphausen, K A & Smith, Q R Brain metastases as preventive and therapeutic targets Nat Rev Cancer 11, 352–363 (2011) 21 Primer 16 Hung, M H et al Effect of age and biological subtype on the risk and timing of brain metastasis in breast cancer patients PLOS ONE 9, e89389 (2014) 17 Znaor, A., Lortet-​Tieulent, J., Laversanne, M., Jemal, A & Bray, F International variations and trends in renal cell carcinoma incidence and mortality Eur Urol 67, 519–530 (2015) 18 Torre, L A., Siegel, R L., Ward, E M & Jemal, A International variation in lung cancer mortality rates and trends among women Cancer Epidemiol Biomarkers Prev 23, 1025–1036 (2014) 19 Gaspar, L et al Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials Int J Radiat Oncol Biol Phys 37, 745–751 (1997) 20 Nieder, C., Grosu, A L & Gaspar, L E Stereotactic radiosurgery (SRS) for brain metastases: a systematic review Radiat Oncol 9, 155 (2014) 21 Ekici, K et al Survival and prognostic factors in patients with brain metastasis: single center experience J BUON 21, 958–963 (2016) 22 Gaspar, L E., Scott, C., Murray, K & Curran, W Validation of the RTOG recursive partitioning analysis (RPA) classification for brain metastases Int J Radiat Oncol Biol Phys 47, 1001–1006 (2000) 23 Sperduto, P W., Berkey, B., Gaspar, L E., Mehta, M & Curran, W A new prognostic index and comparison to three other indices for patients with brain metastases: an analysis of 1,960 patients in the RTOG database Int J Radiat Oncol Biol Phys 70, 510–514 (2008) 24 Sperduto, P W et al Summary report on the graded prognostic assessment: an accurate and facile diagnosis-​specific tool to estimate survival for patients with brain metastases J Clin Oncol 30, 419–425 (2012) 25 Sperduto, P W et al Effect of tumor subtype on survival and the graded prognostic assessment for patients with breast cancer and brain metastases Int J Radiat Oncol Biol Phys 82, 2111–2117 (2012) 26 Sperduto, P W et al Estimating survival in patients with lung cancer and brain metastases: an update of the graded prognostic assessment for lung cancer using molecular markers (Lung-​molGPA) JAMA Oncol 3, 827–831 (2016) This article provides an updated prognostic index for patients with lung cancer brain metastases that reflects the importance of molecular markers as well as clinical factors in overall prognosis 27 Balasubramanian, S K et al Impact of EGFR and ALK mutaiton on the outcomes of non-​small cell lung cancer (NSCLC) patients with brain metastases [abstract] J Clin Oncol 34 (Suppl 15), 2005 (2016) 28 Chen, L F., Patel, J D & Lukas, R V Advances in brain metastases presented at the American Society of Clinical Oncology 2016 annual meeting: part I Future Oncol 12, 2535–2538 (2016) 29 McTyre, E R et al Predictors of neurologic and nonneurologic death in patients with brain metastasis initially treated with upfront stereotactic radiosurgery without whole-​brain radiation therapy Neuro Oncol 19, 558–566 (2017) 30 Neal, M T et al Predictors of survival, neurologic death, local failure, and distant failure after gamma knife radiosurgery for melanoma brain metastases World Neurosurg 82, 1250–1255 (2014) 31 Cabezas, R et al Astrocytic modulation of blood brain barrier: perspectives on Parkinson’s disease Front Cell Neurosci 8, 211 (2014) 32 Shi, C & Pamer, E G Monocyte recruitment during infection and inflammation Nat Rev Immunol 11, 762–774 (2011) 33 Johanson, C E., Stopa, E G & McMillan, P N The blood-​cerebrospinal fluid barrier: structure and functional significance Methods Mol Biol 686, 101–131 (2011) 34 Boire, A et al Complement component adapts the cerebrospinal fluid for leptomeningeal metastasis Cell 168, 1101–1113 (2017) 35 Berghoff, A S & Preusser, M The inflammatory microenvironment in brain metastases: potential treatment target? Chin Clin Oncol 4, 21 (2015) 36 Chambers, A F., Groom, A C & MacDonald, I C Dissemination and growth of cancer cells in metastatic sites Nat Rev Cancer 2, 563–572 (2002) 37 Kienast, Y et al Real-​time imaging reveals the single steps of brain metastasis formation Nat Med 16, 116–122 (2010) 38 Heyn, C et al In vivo MRI of cancer cell fate at the single-​cell level in a mouse model of breast cancer 22 | Article citation ID: 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 metastasis to the brain Magn Reson Med 56, 1001–1010 (2006) Bos, P et al Genes that mediate breast cancer metastasis to the brain Nature 459, 1005–1010 (2009) Fan, J et al Integrin β4 signaling promotes mammary tumor cell adhesion to brain microvascular endothelium by inducing ErbB2-mediated secretion of VEGF Ann Biomed Eng 39, 2223–2241 (2011) Küsters, B et al Vascular endothelial growth factor-​A165 induces progression of melanoma brain metastases without induction of sprouting angiogenesis Cancer Res 62, 341–345 (2002) Izraely, S et al The metastatic microenvironment: brain-​residing melanoma metastasis and dormant micrometastasis Int J Cancer 131, 1071–1082 (2012) Gorantla, V., Kirkwood, J M & Tawbi, H A Melanoma brain metastases: an unmet challenge in the era of active therapy Curr Oncol Rep 15, 483–491 (2013) Eichler, A F et al The biology of brain metastases-​ translation to new therapies Nat Rev Clin Oncol 8, 344–356 (2011) Zhang, S Y et al Src family kinases as novel therapeutic targets to treat breast cancer brain metastases Cancer Res 73, 5764–5774 (2013) Lorger, M & Felding-​Habermann, B Capturing changes in the brain microenvironment during initial steps of breast cancer brain metastasis Am J Pathol 176, 2958–2971 (2010) Sevenich, L et al Analysis of tumour-​and stroma-​supplied proteolytic networks reveals a brain-​metastasis-promoting role for cathepsin S Nat Cell Biol 16, 876–888 (2014) Simonsen, T G., Gaustad, J V & Rofstad, E K Intracranial tumor cell migration and the development of multiple brain metastases in malignant melanoma Transl Oncol 9, 211–218 (2016) Malladi, S et al Metastatic latency and immune evasion through autocrine inhibition of WNT Cell 165, 45–60 (2016) Kim, M Y et al Tumor self-​seeding by circulating cancer cells Cell 139, 1315–1326 (2009) Lockman, P R et al Heterogeneous blood-​tumor barrier permeability determines drug efficacy in experimental brain metastases of breast cancer Clin Cancer Res 16, 5664–5678 (2010) Nduom, E K., Yang, C Z., Merrill, M J., Zhuang, Z P & Lonser, R R Characterization of the blood-​brain barrier of metastatic and primary malignant neoplasms laboratory investigation J Neurosurg 119, 427–433 (2013) Murrell, D H et al Understanding heterogeneity and permeability of brain metastases in murine models of HER2-positive breast cancer through magnetic resonance imaging: implications for detection and therapy Transl Oncol 8, 176–184 (2015) Henry, M N., Chen, Y H., McFadden, C D., Simedrea, F C & Foster, P J In-​vivo longitudinal MRI study: an assessment of melanoma brain metastases in a clinically relevant mouse model Melanoma Res 25, 127–137 (2015) Morikawa, A et al Capecitabine and lapatinib uptake in surgically resected brain metastases from metastatic breast cancer patients: a prospective study Neuro-​oncology 17, 289–295 (2015) Lyle, L et al Alterations in pericyte subpopulations are associated with elevated blood-​tumor barrier permeability in experimental brain metastasis of breast cancer Clin Cancer Res 22, 5287–5299 (2016) Yonemori, K et al Disruption of the blood brain barrier by brain metastases of triple-​negative and basal-​type breast cancer but not HER2/neu-​positive breast cancer Cancer 116, 302–308 (2010) Connell, J J et al Selective permeabilization of the blood-​brain barrier at sites of metastasis J Natl Cancer Inst 105, 1634–1643 (2013) Avraham, H K et al Angiopoietin-2 mediates blood-​brain barrier impairment and colonization of triple-​negative breast cancer cells in brain J Pathol 232, 369–381 (2014) Ma, S C et al Claudin-5 regulates blood-​brain barrier permeability by modifying brain microvascular endothelial cell proliferation, migration, and adhesion to prevent lung cancer metastasis CNS Neurosci Ther 23, 947–960 (2017) Gril, B et al Reactive astrocytic S1P3 signaling modulates the blood-​tumor barrier in brain metastases Nat Commun 9, 2705 (2018) 62 Fitzgerald, D et al Reactive glia are recruited by highly proliferative brain metastases of breast cancer and promote tumor cell colonization Clin Exp Metastasis 25, 799–810 (2008) 63 Chen, Q et al Carcinoma-​astrocyte gap junctions promote brain metastasis by cGAMP transfer Nature 533, 493–498 (2016) 64 Kim, S J et al Astrocytes upregulate survival genes in tumor cells and induce protection from chemotherapy Neoplasia 13, 286–298 (2011) 65 Sartorius, C A et al Estrogen promotes the brain metastatic colonization of triple negative breast cancer cells via an astrocyte-​mediated paracrine mechanism Oncogene 35, 2881–2892 (2016) This study demonstrates that oestrogen blockade can decrease the formation of triple-​negative breast cancer brain metastases through inhibition of ER-​positive astrocytes in the brain microenvironment, providing initial data for prevention as a strategy against brain metastases 66 Xu, L., Gao, G., Ren, J., Su, F & Zhang, W Estrogen receptor β of host promotes the progression of lung cancer brain metastasis of an orthotopic mouse model J Cancer Ther 3, 352–358 (2012) 67 Zhang, L et al Microenvironment-​induced PTEN loss by exosomal microRNA primes brain metastasis outgrowth Nature 527, 100–104 (2015) 68 Louie, E et al Neurotrophin-3 modulates breast cancer cells and the microenvironment to promote the growth of breast cancer brain metastasis Oncogene 32, 4064–4077 (2013) 69 Berghoff, A S., Lassmann, H., Preusser, M & Hoftberger, R Characterization of the inflammatory response to solid cancer metastases in the human brain Clin Exp Metastasis 30, 69–81 (2013) 70 Bowman, R L et al Macrophage ontogeny underlies differences in tumor-​specific education in brain malignancies Cell Rep 17, 2445–2459 (2016) 71 Harter, P N et al Distribution and prognostic relevance of tumor-​infiltrating lymphocytes (TILs) and PD-1/PD-​L1 immune checkpoints in human brain metastases Oncotarget 6, 40836–40849 (2015) 72 Zakaria, R et al T-​cell densities in brain metastases are associated with patient survival times and diffusion tensor MRI changes Cancer Res 78, 610–616 (2018) 73 Mansfield, A S et al Contraction of T cell richness in lung cancer brain metastases Sci Rep 8, 2171 (2018) 74 Duchnowska, R et al Immune response in breast cancer brain metastases and their microenvironment: the role of the PD-1/PD-​L axis Breast Cancer Res 18, 43 (2016) 75 Berghoff, A S et al Tumour-​infiltrating lymphocytes and expression of programmed death ligand (PD-L1) in melanoma brain metastases Histopathology 66, 289–299 (2015) 76 Berghoff, A S et al Tumor infiltrating lymphocytes and PD-​L1 expression in brain metastases of small cell lung cancer (SCLC) J Neurooncol 130, 19–29 (2016) 77 Teglasi, V et al Evaluating the significance of density, localization, and PD-1/PD-​L1 immunopositivity of mononuclear cells in the clinical course of lung adenocarcinoma patients with brain metastasis Neuro-​oncology 19, 1058–1067 (2017) 78 Taggart, D et al Anti-​PD-1/anti-​CTLA-4 efficacy in melanoma brain metastases depends on extracranial disease and augmentation of CD8+ T cell trafficking Proc Natl Acad Sci USA 115, E1540–E1549 (2018) 79 Samson, A et al Intravenous delivery of oncolytic reovirus to brain tumor patients immunologically primes for subsequent checkpoint blockade Sci Transl Med 10, eaam7577 (2018) 80 Chen, E I et al Adaptation of energy metabolism in breast cancer brain metastases Cancer Res 67, 1472–1486 (2007) 81 Chen, J et al Gain of glucose-​independent growth upon metastasis of breast cancer cells to the brain Cancer Res 75, 554–565 (2015) 82 Leenders, W P J et al Antiangiogenic therapy of cerebral melanoma metastases results in sustained tumor progression via vessel co-​option Clin Cancer Res 10, 6222–6230 (2004) 83 Valiente, M et al Serpins promote cancer cell survival and vascular co-​option in brain metastasis Cell 156, 1002–1016 (2014) 84 Allen, J E et al COX-2 drives metastatic breast cells from brain lesions into the cerebrospinal fluid and systemic circulation Cancer Res 74, 2385–2390 (2014) www.nature.com/nrdp (2019) 5:5 0123456789(); Primer 85 Zhang, C Y., Zhang, F H., Tsan, R & Fidler, I J Transforming growth factor-​β2 is a molecular determinant for site-​specific melanoma metastasis in the brain Cancer Res 69, 828–835 (2009) 86 Park, E S et al Cross-​species hybridization of microarrays for studying tumor transcriptome of brain metastasis Proc Natl Acad Sci USA 108, 17456–17461 (2011) This study demonstrates that the brain microenvironment (‘soil’) can reprogramme metastatic tumour cells (‘seed’) to gain neuronal cell characteristics 87 Brastianos, P K et al Genomic characterization of brain metastases reveals branched evolution and potential therapeutic targets Cancer Discov 5, 1164–1177 (2015) This study demonstrates distinct evolution patterns within brain metastases compared with their primary tumours 88 Saunus, J M et al Integrated genomic and transcriptomic analysis of human brain metastases identifies alterations of potential clinical significance J Pathol 237, 363–378 (2015) 89 Wang, L et al Astrocytes directly influence tumor cell invasion and metastasis in vivo PLOS ONE 8, e80933 (2013) 90 Neman, J et al Human breast cancer metastases to the brain display GABAergic properties in the neural niche Proc Natl Acad Sci USA 111, 984–989 (2014) 91 Hayes, D N et al Gene expression profiling reveals reproducible human lung adenocarcinoma subtypes in multiple independent patient cohorts J Clin Oncol 24, 5079–5090 (2006) 92 The Cancer Genome Atlas Research Network Comprehensive molecular profiling of lung adenocarcinoma Nature 511, 543–550 (2014) 93 Ansari, J., Palmer, D H., Rea, D W & Hussain, S A Role of tyrosine kinase inhibitors in lung cancer Anticancer Agents Med Chem 9, 569–575 (2009) 94 Venur, V A & Ahluwalia, M S Targeted therapy in brain metastases: ready for primetime? Am Soc Clin Oncol Educ Book 35, e123–130 (2016) 95 Peters, S et al Alectinib versus crizotinib in untreated ALK-​positive non-​small-cell lung cancer N Engl J Med 377, 829–838 (2017) 96 Mok, T S et al Osimertinib or platinum-​pemetrexed in EGFR T790M-​positive lung cancer N Engl J Med 376, 629–640 (2017) 97 Zhao, N et al Alterations of LKB1 and KRAS and risk of brain metastasis: comprehensive characterization by mutation analysis, copy number, and gene expression in non-​small-cell lung carcinoma Lung Cancer 86, 255–261 (2014) 98 Nguyen, D X et al WNT/TCF signaling through LEF1 and HOXB9 mediates lung adenocarcinoma metastasis Cell 138, 51–62 (2009) 99 Ilhan-​Mutlu, A et al Expression profiling of angiogenesis-​related genes in brain metastases of lung cancer and melanoma Tumour Biol 37, 1173–1182 (2016) 100 Kennecke, H et al Metastatic behavior of breast cancer subtypes J Clin Oncol 28, 3271–3277 (2010) 101 Nik-​Zainal, S & Morganella, S Mutational signatures in breast cancer: the problem at the DNA level Clin Cancer Res 23, 2617–2629 (2017) 102 Salhia, B et al Integrated genomic and epigenomic analysis of breast cancer brain metastasis PLOS ONE 9, e85448 (2014) 103 Harrell, J C et al Genomic analysis identifies unique signatures predictive of brain, lung, and liver relapse Breast Cancer Res Treat 132, 523–535 (2012) 104 Jin, H et al Methylation status of T-​lymphoma invasion and metastasis promoter and its overexpression in colorectal cancer Hum Pathol 42, 541–551 (2011) 105 Duchnowska, R et al Predicting early brain metastases based on clinicopathological factors and gene expression analysis in advanced HER2-positive breast cancer patients J Neurooncol 122, 205–216 (2015) 106 Adamo, B et al Phosphatidylinositol 3-kinase pathway activation in breast cancer brain metastases Breast Cancer Res 13, R125 (2011) 107 Kodack, D P et al The brain microenvironment mediates resistance in luminal breast cancer to PI3K inhibition through HER3 activation Sci Transl Med 9, eaal4682 (2017) 108 Phillips, G D et al Dual targeting of HER2-positive cancer with trastuzumab emtansine and pertuzumab: critical role for neuregulin blockade in antitumor response to combination therapy Clin Cancer Res 20, 456–468 (2014) 109 Askoxylakis, V et al Preclinical efficacy of ado-​ trastuzumab emtansine in the brain microenvironment J Natl Cancer Inst 108, djv313 (2016) 110 Hodis, E et al A landscape of driver mutations in melanoma Cell 150, 251–263 (2012) 111 Chen, G et al Molecular profiling of patient-​matched brain and extracranial melanoma metastases implicates the PI3K pathway as a therapeutic target Clin Cancer Res 20, 5537–5546 (2014) 112 Jakob, J A et al NRAS mutation status is an independent prognostic factor in metastatic melanoma Cancer 118, 4014–4023 (2012) 113 Long, G V et al Dabrafenib in patients with Val600Glu or Val600Lys BRAF-​mutant melanoma metastatic to the brain (BREAK-​MB): a multicentre, open-​label, phase trial Lancet Oncol 13, 1087–1095 (2012) 114 Bucheit, A D et al Complete loss of PTEN protein expression correlates with shorter time to brain metastasis and survival in stage IIIB/C melanoma patients with BRAFV600 mutations Clin Cancer Res 20, 5527–5536 (2014) 115 Jilaveanu, L B et al PLEKHA5 as a biomarker and potential mediator of melanoma brain metastasis Clin Cancer Res 21, 2138–2147 (2015) 116 Niessner, H et al Targeting hyperactivation of the AKT survival pathway to overcome therapy resistance of melanoma brain metastases Cancer Med 2, 76–85 (2013) 117 Hamilton, R et al Pathologic and gene expression features of metastatic melanomas to the brain Cancer 119, 2737–2746 (2013) 118 Kluger, H M et al Characterization of PD-​L1 expression and associated T-​cell infiltrates in metastatic melanoma samples from variable anatomic sites Clin Cancer Res 21, 3052–3060 (2015) 119 Cairns, P Renal cell carcinoma Cancer Biomark 9, 461–473 (2010) 120 Takahashi, M et al Gene expression profiling of clear cell renal cell carcinoma: gene identification and prognostic classification Proc Natl Acad Sci USA 98, 9754–9759 (2001) 121 Han, C H & Brastianos, P K Genetic characterization of brain metastases in the era of targeted therapy Front Oncol 7, 230 (2017) 122 Guinney, J et al The consensus molecular subtypes of colorectal cancer Nat Med 21, 1350–1356 (2015) 123 Yaeger, R et al RAS mutations affect pattern of metastatic spread and increase propensity for brain metastasis in colorectal cancer Cancer 121, 1195–1203 (2015) 124 Soffietti, R., Ducati, A & Rudà, R Brain metastases Handb Clin Neurol 105, 747–755 (2012) 125 Usinskiene, J et al Optimal differentiation of highand low-​grade glioma and metastasis: a meta-​analysis of perfusion, diffusion, and spectroscopy metrics Neuroradiology 58, 339–350 (2016) 126 Sze, G., Milano, E., Johnson, C & Heier, L Detection of brain metastases: comparison of contrast-​enhanced MR with unenhanced MR and enhanced CT AJNR Am J Neuroradiol 11, 785–791 (1990) 127 Schellinger, P D., Meinck, H M & Thron, A Diagnostic accuracy of MRI compared to CCT in patients with brain metastases J Neurooncol 44, 275–281 (1999) 128 Sze, G et al Comparison of single- and triple-​dose contrast material in the MR screening of brain metastases AJNR Am J Neuroradiol 19, 821–828 (1998) 129 Mills, S J., Thompson, G & Jackson, A Advanced magnetic resonance imaging biomarkers of cerebral metastases Cancer Imag 12, 245–252 (2012) 130 Yeh, R H et al Distinct MR imaging features of triple-​negative breast cancer with brain metastasis J Neuroimag 25, 474–481 (2015) 131 Desprechins, B et al Use of diffusion-​weighted MR imaging in differential diagnosis between intracerebral necrotic tumors and cerebral abscesses AJNR Am J Neuroradiol 20, 1252–1257 (1999) 132 Sternberg, E J., Lipton, M L & Burns, J Utility of diffusion tensor imaging in evaluation of the peritumoral region in patients with primary and metastatic brain tumors AJNR Am J Neuroradiol 35, 439–444 (2014) 133 Bulakbasi, N et al Assessment of diagnostic accuracy of perfusion MR imaging in primary and metastatic solitary malignant brain tumors AJNR Am J Neuroradiol 26, 2187–2199 (2005) 134 Law, M et al High-​grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MR imaging Radiology 222, 715–721 (2002) NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); 135 Sparacia, G., Gadde, J A., Iaia, A., Sparacia, B & Midiri, M Usefulness of quantitative peritumoural perfusion and proton spectroscopic magnetic resonance imaging evaluation in differentiating brain gliomas from solitary brain metastases Neuroradiol J 29, 160–167 (2016) 136 Toh, C H et al Differentiation of brain abscesses from glioblastomas and metastatic brain tumors: comparisons of diagnostic performance of dynamic susceptibility contrast-​enhanced perfusion MR imaging before and after mathematic contrast leakage correction PLOS ONE 9, e109172 (2014) 137 Shim, W H., Kim, H S., Choi, C G & Kim, S J Comparison of apparent diffusion coefficient and intravoxel incoherent motion for differentiating among glioblastoma, metastasis, and lymphoma focusing on diffusion-​related parameter PLOS ONE 10, e0134761 (2015) 138 Chiang, I C et al Distinction between high-​grade gliomas and solitary metastases using peritumoral 3-T magnetic resonance spectroscopy, diffusion, and perfusion imagings Neuroradiology 46, 619–627 (2004) 139 Server, A et al Proton magnetic resonance spectroscopy in the distinction of high-​grade cerebral gliomas from single metastatic brain tumors Acta Radiol 51, 316–325 (2010) 140 Kosaka, N et al 18F-​FDG PET of common enhancing malignant brain tumors AJR Am J Roentgenol 190, W365–W369 (2008) 141 Hutterer, M et al [18F]-fluoro-​ethyl-L-​tyrosine PET: a valuable diagnostic tool in neuro-​oncology, but not all that glitters is glioma Neuro-​oncology 15, 341–351 (2013) 142 Calabria, F F., Barbarisi, M., Gangemi, V., Grillea, G & Cascini, G L Molecular imaging of brain tumors with radiolabeled choline PET Neurosurg Rev 41, 67–76 (2018) 143 Soffietti, R et al Diagnosis and treatment of brain metastases from solid tumors: guidelines from the European Association of Neuro-​Oncology (EANO) Neuro-​oncology 19, 162–174 (2017) 144 Berghoff, A S et al Predictive molecular markers in metastases to the central nervous system: recent advances and future avenues Acta Neuropathol 128, 879–891 (2014) 145 Shen, Q et al Breast cancer with brain metastases: clinicopathologic features, survival, and paired biomarker analysis Oncologist 20, 466–473 (2015) 146 Monzon, F A & Koen, T J Diagnosis of metastatic neoplasms: molecular approaches for identification of tissue of origin Arch Pathol Lab Med 134, 216–224 (2010) 147 Pentsova, E I et al Evaluating cancer of the central nervous system through next-​generation sequencing of cerebrospinal fluid J Clin Oncol 34, 2404–2415 (2016) 148 Grossman, S A & Krabak, M J Leptomeningeal carcinomatosis Cancer Treat Rev 25, 103–119 (1999) 149 Chamberlain, M et al Leptomeningeal metastasis: a response assessment in neuro-​oncology critical review of endpoints and response criteria of published randomized clinical trials Neuro-​oncology 16, 1176–1185 (2014) 150 Chamberlain, M et al Leptomeningeal metastases: a RANO proposal for response criteria Neuro-oncology 19, 484–492 (2017) 151 Niwinska, A., Tacikowska, M & Murawska, M The effect of early detection of occult brain metastases in HER2-positive breast cancer patients on survival and cause of death Int J Radiat Oncol Biol Phys 77, 1134–1139 (2010) 152 Zhu, H et al Risk factors for brain metastases in completely resected small cell lung cancer: a retrospective study to identify patients most likely to benefit from prophylactic cranial irradiation Radiat Oncol 9, 216 (2014) 153 Koiso, T et al A case-​matched study of stereotactic radiosurgery for patients with brain metastases: comparing treatment results for those with versus without neurological symptoms J Neurooncol 130, 581–590 (2016) 154 Wolf, A et al Toward the complete control of brain metastases using surveillance screening and stereotactic radiosurgery J Neurosurg 128, 23–31 (2017) 155 Le Pechoux, C et al Clinical neurological outcome and quality of life among patients with limited small-​cell cancer treated with two different doses of prophylactic cranial irradiation in the intergroup phase III trial 23 Primer (PCI99-01, EORTC 22003-08004, RTOG 0212 and IFCT 99-01) Ann Oncol 22, 1154–1163 (2011) 156 Boggs, D H et al Strategies to prevent brain metastasis in high-​risk non-​small-cell lung cancer: lessons learned from a randomized study of maintenance temozolomide versus observation Clin Lung Cancer 15, 433–440 (2014) 157 Li, N et al Randomized phase III trial of prophylactic cranial irradiation versus observation in patients with fully resected stage IIIA-​N2 nonsmall-​cell lung cancer and high risk of cerebral metastases after adjuvant chemotherapy Ann Oncol 26, 504–509 (2015) 158 National Comprehensive Cancer Network NCCN guidelines NCCN https://www.nccn.org/professionals/ physician_gls/default.aspx (2018) 159 Ramakrishna, N et al Recommendations on disease management for patients with advanced human epidermal growth factor receptor 2-positive breast cancer and brain metastases: ASCO Clinical Practice Guideline Update J Clin Oncol 36, 2804–2807 (2018) 160 Taylor, M B., Bromham, N R & Arnold, S E Carcinoma of unknown primary: key radiological issues from the recent National Institute for Health and Clinical Excellence guidelines Br J Radiol 85, 661–671 (2012) 161 Klee, B., Law, I., Hojgaard, L & Kosteljanetz, M Detection of unknown primary tumours in patients with cerebral metastases using whole-​body 18F-​ flouorodeoxyglucose positron emission tomography Eur J Neurol 9, 657–662 (2002) 162 Almuhaideb, A., Papathanasiou, N & Bomanji, J 18 F-​FDG PET/CT imaging in oncology Ann Saudi Med 31, 3–13 (2011) 163 Siravegna, G et al Genotyping tumour DNA in cerebrospinal fluid and plasma of a HER2-positive breast cancer patient with brain metastases ESMO Open 2, e000253 (2017) 164 Rudà, R., Borgognone, M., Benech, F., Vasario, E & Soffietti, R Brain metastases from unknown primary tumour: a prospective study J Neurol 248, 394–398 (2001) 165 Nguyen, L N., Maor, M H & Oswald, M J Brain metastases as the only manifestation of an undetected primary tumor Cancer 83, 2181–2184 (1998) 166 Lagerwaard, F J et al Identification of prognostic factors in patients with brain metastases: a review of 1292 patients Int J Radiat Oncol Biol Phys 43, 795–803 (1999) 167 Subbiah, I M et al Validation and development of a modified breast graded prognostic assessment as a tool for survival in patients with breast cancer and brain metastases J Clin Oncol 33, 2239–2245 (2015) 168 Barnholtz-​Sloan, J S et al A nomogram for individualized estimation of survival among patients with brain metastasis Neuro-​oncology 14, 910–918 (2012) 169 Lorenzoni, J et al Radiosurgery for treatment of brain metastases: estimation of patient eligibility using three stratification systems Int J Radiat Oncol Biol Phys 60, 218–224 (2004) 170 Weltman, E et al Radiosurgery for brain metastases: a score index for predicting prognosis Int J Radiat Oncol Biol Phys 46, 1155–1161 (2000) 171 Ryken, T C et al The role of steroids in the management of brain metastases: a systematic review and evidence-​based clinical practice guideline J Neurooncol 96, 103–114 (2010) 172 Kong, X et al A meta-​analysis: prophylactic antiepileptic drugs in patients with brain tumors decrease the incidence of seizures? Clin Neurol Neurosurg 134, 98–103 (2015) 173 Fortin, D The blood-​brain barrier: its influence in the treatment of brain tumors metastases Curr Cancer Drug Targets 12, 247–259 (2012) 174 Askoxylakis, V., Arvanitis, C D., Wong, C S F., Ferraro, G B & Jain, R K Emerging strategies for delivering antiangiogenic therapies to primary and metastatic brain tumors Adv Drug Deliv Rev 119, 159–174 (2017) 175 Kodack, D P et al Combined targeting of HER2 and VEGFR2 for effective treatment of HER2-amplified breast cancer brain metastases Proc Natl Acad Sci USA 109, E3119–E3127 (2012) 176 Falchook, G S et al A phase I trial of combination trastuzumab, lapatinib, and bevacizumab in patients with advanced cancer Invest New Drugs 33, 177–186 (2015) 177 Barlesi, F et al Pemetrexed and cisplatin as first-​line chemotherapy for advanced non-​small-cell lung cancer (NSCLC) with asymptomatic inoperable brain 24 | Article citation ID: metastases: a multicenter phase II trial (GFPC 07-01) Ann Oncol 22, 2466–2470 (2011) 178 Robinet, G et al Results of a phase III study of early versus delayed whole brain radiotherapy with concurrent cisplatin and vinorelbine combination in inoperable brain metastasis of non-​small-cell lung cancer: Groupe Francais de Pneumo-​Cancerologie (GFPC) protocol 95-1 Ann Oncol 12, 59–67 (2001) 179 Cortes, J et al Front-​line paclitaxel/cisplatin-​based chemotherapy in brain metastases from non-​small-cell lung cancer Oncology 64, 28–35 (2003) 180 Dinglin, X X et al Pemetrexed and cisplatin combination with concurrent whole brain radiotherapy in patients with brain metastases of lung adenocarcinoma: a single-​arm phase II clinical trial J Neurooncol 112, 461–466 (2013) 181 Antonadou, D et al Phase II randomized trial of temozolomide and concurrent radiotherapy in patients with brain metastases J Clin Oncol 20, 3644–3650 (2002) 182 Verger, E et al Temozolomide and concomitant whole brain radiotherapy in patients with brain metastases: a phase II randomized trial Int J Radiat Oncol Biol Phys 61, 185–191 (2005) 183 Sperduto, P W et al A phase trial of whole brain radiation therapy and stereotactic radiosurgery alone versus WBRT and SRS with temozolomide or erlotinib for non-​small cell lung cancer and to brain metastases: Radiation Therapy Oncology Group 0320 Int J Radiat Oncol Biol Phys 85, 1312–1318 (2013) 184 Chua, D et al Whole-​brain radiation therapy plus concomitant temozolomide for the treatment of brain metastases from non-​small-cell lung cancer: a randomized, open-​label phase II study Clin Lung Cancer 11, 176–181 (2010) 185 Robins, H I., O’Neill, A., Mehta, M & Grossman, S A phase trial of whole brain radiation therapy and stereotactic radiosurgery alone versus WBRT & SRS with temozolomide or erlotinib for non-​small cell lung cancer and to brain metastases: Radiation Therapy Oncology Group 0320: in regard to Sperduto et al Int J Radiat Oncol Biol Phys 86, 809–810 (2013) 186 Lekic, M et al Outcome of small cell lung cancer (SCLC) patients with brain metastases in a routine clinical setting Radiol Oncol 46, 54–59 (2012) 187 Isakoff, S J Triple-​negative breast cancer: role of specific chemotherapy agents Cancer J 16, 53–61 (2010) 188 Christodoulou, C et al Temozolomide (TMZ) combined with cisplatin (CDDP) in patients with brain metastases from solid tumors: a Hellenic Cooperative Oncology Group (HeCOG) phase II study J Neurooncol 71, 61–65 (2005) 189 Trudeau, M E et al Temozolomide in metastatic breast cancer (MBC): a phase II trial of the National Cancer Institute of Canada - Clinical Trials Group (NCIC-​CTG) Ann Oncol 17, 952–956 (2006) 190 Ekenel, M., Hormigo, A M., Peak, S., Deangelis, L M & Abrey, L E Capecitabine therapy of central nervous system metastases from breast cancer J Neurooncol 85, 223–227 (2007) 191 Nieder, C Front-​line chemotherapy with cisplatin and etoposide for patients with brain metastases from breast carcinoma, nonsmall cell lung carcinoma, or malignant melanoma A prospective study Cancer 86, 900–903 (1999) 192 Silvani, A et al Systemic sagopilone (ZK-​EPO) treatment of patients with recurrent malignant gliomas J Neurooncol 95, 61–64 (2009) 193 Kumthekar, P et al ANG1005, a novel brain-​ penetrant taxane derivative, for the treatment of recurrent brain metastases and leptomeningeal carcinomatosis from breast cancer [abstract] J Clin Oncol 34 (Suppl 15), 2004 (2016) 194 Tang, S C et al ANG1005, a novel peptide-​paclitaxel conjugate crosses the BBB and shows activity in patients with recurrent CNS metastasis from breast cancer, results from a phase II clinical study [abstract] Ann Oncol 27 (Suppl 6), 324O (2016) 195 Eichler, A F et al Survival in patients with brain metastases from breast cancer: the importance of HER-2 status Cancer 112, 2359–2367 (2008) 196 Stewart, D J & Dahrouge, S Response of brain metastases from breast cancer to megestrol acetate: a case report J Neurooncol 24, 299–301 (1995) 197 Salvati, M., Cervoni, L., Innocenzi, G & Bardella, L Prolonged stabilization of multiple and single brain metastases from breast cancer with tamoxifen Report of three cases Tumori 79, 359–362 (1993) 198 Mornex, F et al A prospective randomized multicentre phase III trial of fotemustine plus whole brain irradiation versus fotemustine alone in cerebral metastases of malignant melanoma Melanoma Res 13, 97–103 (2003) 199 Agarwala, S S et al Temozolomide for the treatment of brain metastases associated with metastatic melanoma: a phase II study J Clin Oncol 22, 2101–2107 (2004) 200 Larkin, J M et al A phase I/II study of lomustine and temozolomide in patients with cerebral metastases from malignant melanoma Br J Cancer 96, 44–48 (2007) 201 Welsh, J W et al Phase II trial of erlotinib plus concurrent whole-​brain radiation therapy for patients with brain metastases from non-​small-cell lung cancer J Clin Oncol 31, 895–902 (2013) 202 Ceresoli, G L et al Gefitinib in patients with brain metastases from non-​small-cell lung cancer: a prospective trial Ann Oncol 15, 1042–1047 (2004) 203 Grommes, C et al “Pulsatile” high-​dose weekly erlotinib for CNS metastases from EGFR mutant non-​ small cell lung cancer Neuro Oncol 13, 1364–1369 (2011) 204 How, J., Mann, J., Laczniak, A N & Baggstrom, M Q Pulsatile erlotinib in EGFR-​positive non-​small-cell lung cancer patients with leptomeningeal and brain metastases: review of the literature Clin Lung Cancer 18, 354–363 (2017) 205 Soria, J C et al Osimertinib in untreated EGFR-​ mutated advanced non-​small-cell lung cancer N Engl J Med 378, 113–125 (2018) 206 Ahn, M et al Phase I study (BLOOM) of AZD3759, a BBB penetrable EGFR inhibitor, in patients with TKI-​naïve, EGFRm NSCLC with CNS metastases [abstract] J Clin Oncol 35 (Suppl 15), 2006 (2017) 207 Wang, H et al The ability of avitinib to penetrate the blood brain barrier and its control of intra-/extracranial disease in patients of non-​small cell lung cancer (NSCLC) harboring EGFR T790M mutation[abstract] J Clin Oncol 35 (Suppl 15), e20613 (2017) 208 Li, D et al BIBW2992, an irreversible EGFR/HER2 inhibitor highly effective in preclinical lung cancer models Oncogene 27, 4702–4711 (2008) 209 Economopoulou, P & Mountzios, G Non-​small cell lung cancer (NSCLC) and central nervous system (CNS) metastases: role of tyrosine kinase inhibitors (TKIs) and evidence in favor or against their use with concurrent cranial radiotherapy Transl Lung Cancer Res 5, 588–598 (2016) 210 Soria, J C et al First-​line ceritinib versus platinum-​ based chemotherapy in advanced ALK-​rearranged non-​small-cell lung cancer (ASCEND-4): a randomised, open-​label, phase study Lancet 389, 917–929 (2017) 211 Gadgeel, S et al Alectinib versus crizotinib in treatment-​naïve anaplastic lymphoma kinase-​positive (ALK+) non-​small-cell lung cancer: CNS efficacy results from the ALEX study Ann Oncol 29, 2214–2222 (2018) 212 Camidge, D R et al Brigatinib versus crizotinib in ALK-​positive non-​small-cell lung cancer N Engl J Med 379, 2027–2039 (2018) 213 Shaw, A T et al Lorlatinib in non-​small-cell lung cancer with ALK or ROS1 rearrangement: an international, multicentre, open-​label, single-​arm first-​in-man phase trial Lancet Oncol 18, 1590–1599 (2017) 214 Petrelli, F et al Efficacy of ALK inhibitors on NSCLC brain metastases: a systematic review and pooled analysis of 21 studies PLOS ONE 13, e0201425 (2018) 215 Venur, V A & Leone, J P Targeted therapies for brain metastases from breast cancer Int J Mol Sci 17, 1543 (2016) 216 Lin, N U., Amiri-​Kordestani, L., Palmieri, D., Liewehr, D J & Steeg, P S CNS metastases in breast cancer: old challenge, new frontiers Clin Cancer Res 19, 6404–6418 (2013) 217 Kabraji, S et al Drug resistance in HER2-positive breast cancer brain metastases: blame the barrier or the brain? Clin Cancer Res 24, 1795–1804 (2018) 218 Askoxylakis, V., Kodack, D P., Ferraro, G B & Jain, R K Antibody-​based therapies for the treatment of brain metastases from HER2-positive breast cancer: time to rethink the importance of the BBB? Breast Cancer Res Treat 165, 467–468 (2017) 219 Lewis Phillips, G D et al Trastuzumab uptake and its relation to efficacy in an animal model of HER2positive breast cancer brain metastasis Breast Cancer Res Treat 164, 581–591 (2017) www.nature.com/nrdp (2019) 5:5 0123456789(); Primer 220 Lin, N U et al Multicenter phase II study of lapatinib in patients with brain metastases from HER2-positive breast cancer Clin Cancer Res 15, 1452–1459 (2009) 221 Lin, N U et al Phase II trial of lapatinib for brain metastases in patients with human epidermal growth factor receptor 2-positive breast cancer J Clin Oncol 26, 1993–1999 (2008) 222 Bachelot, T et al Lapatinib plus capecitabine in patients with previously untreated brain metastases from HER2-positive metastatic breast cancer (LANDSCAPE): a single-​group phase study Lancet Oncol 14, 64–71 (2013) 223 Freedman, R A et al Translational breast cancer research consortium (TBCRC) 022: a phase II trial of neratinib for patients with human epidermal growth factor receptor 2-positive breast cancer and brain metastases J Clin Oncol 34, 945–952 (2016) 224 Freedman, R et al TBCRC 022: phase II trial of neratinib + capecitabine for patients (Pts) with human epidermal growth factor receptor (HER2+) breast cancer brain metastases (BCBM) [abstract] J Clin Oncol 35 (Suppl 15), 1005 (2017) 225 Borges, V F et al Efficacy results of a phase 1b study of ONT-380, a CNS-​penetrant TKI, in combination with T-​DM1 in HER2+ metastatic breast cancer (MBC), including patients (pts) with brain metastases [abstract] J Clin Oncol 34 (Suppl 15), 513 (2016) 226 Lin, N U et al Determination of the maximum tolerated dose (MTD) of the CNS penetrant tyrosine kinase inhibitor (TKI) tesevatinib administered in combination with trastuzumab in HER2+ patients with metastatic breast cancer (BC) [abstract] J Clin Oncol 34 (Suppl 15), 514 (2016) 227 McArthur, G A et al Vemurafenib in metastatic melanoma patients with brain metastases: an open-​label, single-​arm, phase 2, multicentre study Ann Oncol 28, 634–641 (2017) 228 Davies, M A et al Dabrafenib plus trametinib in patients with BRAFV600-mutant melanoma brain metastases (COMBI-​MB): a multicentre, multicohort, open-​label, phase trial Lancet Oncol 18, 863–873 (2017) This phase II clinical trial demonstrates intracranial response rates >50% using targeted molecular therapies on the BRAF and MAPK pathways combined 229 Shonka, N., Venur, V A & Ahluwalia, M S Targeted treatment of brain metastases Curr Neurol Neurosci Rep 17, 37 (2017) 230 Obenauf, A C et al Therapy-​induced tumour secretomes promote resistance and tumour progression Nature 520, 368–372 (2015) 231 Hirata, E et al Intravital imaging reveals how BRAF inhibition generates drug-​tolerant microenvironments with high integrin β1/FAK signaling Cancer Cell 27, 574–588 (2015) 232 Boogerd, W., Dalesio, O., Bais, E M & van der Sande, J J Response of brain metastases from breast cancer to systemic chemotherapy Cancer 69, 972–980 (1992) 233 Valpione, S et al Rechallenge with BRAF-​directed treatment in metastatic melanoma: a multi-​ institutional retrospective study Eur J Cancer 91, 116–124 (2018) This study demonstrates that rechallenge BRAF-​ targeted molecular therapies after initial failure can have significant response rates (>40%) in patients with melanoma brain metastases 234 Berghoff, A S., Venur, V A., Preusser, M & Ahluwalia, M S Immune checkpoint inhibitors in brain metastases: from biology to treatment Am Soc Clin Oncol Educ Book 35, e116–e122 (2016) 235 Margolin, K et al Ipilimumab in patients with melanoma and brain metastases: an open-​label, phase trial Lancet Oncol 13, 459–465 (2012) 236 Goldberg, S B et al Pembrolizumab for patients with melanoma or non-​small-cell lung cancer and untreated brain metastases: early analysis of a non-​randomised, open-​label, phase trial Lancet Oncol 17, 976–983 (2016) 237 Gadgeel, S et al OAK, a randomized Ph III study of atezolizumab vs docetaxel in patients with advanced NSCLC: results from subgroup analyses [abstract PL04a.02] J Thorac Oncol 12 (Suppl.), S9–S10 (2017) 238 Tawbi, H et al Efficacy and safety of nivolumab (NIVO) plus ipilimumab (IPI) in patients with melanoma (MEL) metastatic to the brain: results of the phase II study CheckMate 204 [abstract] J Clin Oncol 35 (Suppl 15), 9507 (2017) 239 Long, G et al A randomized phase II study of nivolumab or nivolumab combined with ipilimumab in patients (pts) with melanoma brain metastases (mets): the Anti-​PD1 Brain Collaboration (ABC) [abstract] J Clin Oncol 35 (Suppl 15), 9508 (2017) 240 Tawbi, H A et al Combined nivolumab and ipilimumab in melanoma metastatic to the brain N Engl J Med 379, 722–730 (2018) This phase II trial demonstrates that combined nivolumab and ipilimumab have a 56% intracranial response rate (comparable to the extracranial response) and result in an overall survival of 81.5% at year and ~70% at years in patients with melanoma brain metastases 241 Anderson, E S et al Melanoma brain metastases treated with stereotactic radiosurgery and concurrent pembrolizumab display marked regression; efficacy and safety of combined treatment J Immunother Cancer 5, 76 (2017) 242 Nardin, C et al Tolerance and outcomes of stereotactic radiosurgery combined with anti-​programmed cell death-1 (pembrolizumab) for melanoma brain metastases Melanoma Res 28, 111–119 (2018) 243 Quail, D F & Joyce, J A The microenvironmental landscape of brain tumors Cancer Cell 31, 326–341 (2017) 244 Cohen, J V et al Melanoma brain metastasis pseudoprogression after pembrolizumab treatment Cancer Immunol Res 4, 179–182 (2016) 245 Patchell, R A et al A randomized trial of surgery in the treatment of single metastases to the brain N Engl J Med 322, 494–500 (1990) 246 Vecht, C J et al Treatment of single brain metastasis: radiotherapy alone or combined with neurosurgery? Ann Neurol 33, 583–590 (1993) 247 Paek, S H., Audu, P B., Sperling, M R., Cho, J & Andrews, D W Reevaluation of surgery for the treatment of brain metastases: review of 208 patients with single or multiple brain metastases treated at one institution with modern neurosurgical techniques Neurosurgery 56, 1021–1034 (2005) 248 Stark, A M., Tscheslog, H., Buhl, R., Held-​Feindt, J & Mehdorn, H M Surgical treatment for brain metastases: prognostic factors and survival in 177 patients Neurosurg Rev 28, 115–119 (2005) 249 Bindal, R K., Sawaya, R., Leavens, M E & Lee, J J Surgical treatment of multiple brain metastases J Neurosurg 79, 210–216 (1993) 250 Wron´ski, M., Arbit, E., McCormick, B & Wrónski, M Surgical treatment of 70 patients with brain metastases from breast carcinoma Cancer 80, 1746–1754 (1997) 251 Hatiboglu, M A., Wildrick, D M & Sawaya, R The role of surgical resection in patients with brain metastases Ecancermedicalscience 7, 308 (2013) 252 Iwadate, Y., Namba, H & Yamaura, A Significance of surgical resection for the treatment of multiple brain metastases Anticancer Res 20, 573–577 (2000) This article provides evidence that surgical resection of tumours >2 cm have improved outcomes after adjuvant radiotherapy including overall survival, even in the presence of multiple brain metastases 253 Mahajan, A et al Post-​operative stereotactic radiosurgery versus observation for completely resected brain metastases: a single-​centre, randomised, controlled, phase trial Lancet Oncol 18, 1040–1048 (2017) 254 Ali, M A et al Stereotactic laser ablation as treatment for brain metastases that recur after stereotactic radiosurgery: a multiinstitutional experience Neurosurg Focus 41, E11 (2016) 255 Ahluwalia, M et al Laser ablation after stereotactic radiosurgery: a multicenter prospective study in patients with metastatic brain tumors and radiation necrosis J Neurosurg https://doi.org/10.3171/ 2017.11.JNS171273 (2018) 256 Rennert, R C et al Patterns of clinical use of stereotactic laser ablation: analysis of a multicenter prospective registry World Neurosurg 116, e566–e570 (2018) 257 Beechar, V B et al Volumetric response of progressing post-​SRS lesions treated with laser interstitial thermal therapy J Neurooncol 137, 57–65 (2018) 258 Kann, B H., Park, H S., Johnson, S B., Chiang, V L & Yu, J B Radiosurgery for brain metastases: changing practice patterns and disparities in the United States J Natl Compr Canc Netw 15, 1494–1502 (2017) NATURe RevIeWS | DISEASE PRIMERS | Article citation ID: (2019) 5:5 0123456789(); 259 Patchell, R A et al Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial JAMA 280, 1485–1489 (1998) 260 Brown, P D et al Effect of radiosurgery alone vs radiosurgery with whole brain radiation therapy on cognitive function in patients with to brain metastases: a randomized clinical trial JAMA 316, 401–409 (2016) 261 Hardee, M E & Formenti, S C Combining stereotactic radiosurgery and systemic therapy for brain metastases: a potential role for temozolomide Front Oncol 2, 99 (2012) 262 Andrews, D W et al Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial Lancet 363, 1665–1672 (2004) 263 Sperduto, P W et al Secondary analysis of RTOG 9508, a phase randomized trial of whole-​brain radiation therapy versus WBRT plus stereotactic radiosurgery in patients with 1-3 brain metastases; poststratified by the graded prognostic assessment (GPA) Int J Radiat Oncol Biol Phys 90, 526–531 (2014) 264 Aoyama, H et al Stereotactic radiosurgery plus whole-​brain radiation therapy vs stereotactic radiosurgery alone for treatment of brain metastases: a randomized controlled trial JAMA 295, 2483–2491 (2006) 265 Kocher, M et al Adjuvant whole-​brain radiotherapy versus observation after radiosurgery or surgical resection of one to three cerebral metastases: results of the EORTC 22952-26001 study J Clin Oncol 29, 134–141 (2011) 266 Chang, E L et al Neurocognition in patients with brain metastases treated with radiosurgery or radiosurgery plus whole-​brain irradiation: a randomised controlled trial Lancet Oncol 10, 1037–1044 (2009) 267 American Society for Radiation Oncology 2014 choosing wisely list ASTRO https://www.astro.org/ Patient-​Care-and-​Research/Patient-​Education/ 2014-Choosing-​Wisely-List (2015) 268 Kim, H., Rajagopalan, M S., Beriwal, S & Smith, K J Cost-​effectiveness analysis of stereotactic radiosurgery alone versus stereotactic radiosurgery with upfront whole brain radiation therapy for brain metastases Clin Oncol (R Coll Radiol) 29, e157–e164 (2017) 269 Sahgal, A et al Phase trials of stereotactic radiosurgery with or without whole-​brain radiation therapy for to brain metastases: individual patient data meta-​analysis Int J Radiat Oncol Biol Phys 91, 710–717 (2015) 270 Aoyama, H., Tago, M., Shirato, H & Japanese Radiation Oncology Study Group 99-1 (JROSG 99-1) Investigators Stereotactic radiosurgery with or without whole-​brain radiotherapy for brain metastases: secondary analysis of the JROSG 99-1 randomized clinical trial JAMA Oncol 1, 457–464 (2015) 271 Liu, Y et al Salvage whole brain radiotherapy or stereotactic radiosurgery after initial stereotactic radiosurgery for 1-4 brain metastases J Neurooncol 124, 429–437 (2015) 272 Zeng, K L et al Patient preference for stereotactic radiosurgery plus or minus whole brain radiotherapy for the treatment of brain metastases Ann Palliat Med 6, S155–S160 (2017) 273 Mulvenna, P et al Dexamethasone and supportive care with or without whole brain radiotherapy in treating patients with non-​small cell lung cancer with brain metastases unsuitable for resection or stereotactic radiotherapy (QUARTZ): results from a phase 3, non-​inferiority, randomised trial Lancet 388, 2004–2014 (2016) 274 Slotman, B et al Prophylactic cranial irradiation in extensive small-​cell lung cancer N Engl J Med 357, 664–672 (2007) 275 Aupérin, A et al Prophylactic cranial irradiation for patients with small-​cell lung cancer in complete remission N Engl J Med 341, 476–484 (1999) 276 Takahashi, T et al Prophylactic cranial irradiation versus observation in patients with extensive-​disease small-​cell lung cancer: a multicentre, randomised, open-​label, phase trial Lancet Oncol 18, 663–671 (2017) 277 Yamamoto, M et al Stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901): a multi-​institutional prospective observational study Lancet Oncol 15, 387–395 (2014) This prospective observational study of nearly 1,200 patients demonstrated that SRS without WBRT may be a suitable alternative for patients with up to ten brain metastases 25 Primer 278 Johnson, M D et al Surgical resection of brain metastases and the risk of leptomeningeal recurrence in patients treated with stereotactic radiosurgery Int J Radiat Oncol Biol Phys 94, 537–543 (2016) 279 Thomas, E M et al Comparison of plan quality and delivery time between volumetric arc therapy (RapidArc) and gamma knife radiosurgery for multiple cranial metastases Neurosurgery 75, 409–417 (2014) 280 Rajakesari, S et al Local control after fractionated stereotactic radiation therapy for brain metastases J Neurooncol 120, 339–346 (2014) 281 Muacevic, A., Kufeld, M., Wowra, B., Kreth, F W & Tonn, J C Feasibility, safety, and outcome of frameless image-​guided robotic radiosurgery for brain metastases J Neurooncol 97, 267–274 (2010) 282 Brown, P D et al Postoperative stereotactic radiosurgery compared with whole brain radiotherapy for resected metastatic brain disease (NCCTG N107C/ CEC·3): a multicentre, randomised, controlled, phase trial Lancet Oncol 18, (1049–1060 (2017) This randomized controlled trial demonstrates less frequent declines in cognitive function with postoperative SRS versus WBRT in patients with brain metastases, with no difference in overall survival 283 Soliman, H et al Consensus contouring guidelines for postoperative completely resected cavity stereotactic radiosurgery for brain metastases Int J Radiat Oncol Biol Phys 100, 436–442 (2018) 284 Gondi, V et al Preservation of memory with conformal avoidance of the hippocampal neural stem-​cell compartment during whole-​brain radiotherapy for brain metastases (RTOG 0933): a phase II multi-​ institutional trial J Clin Oncol 32, 3810–3816 (2014) 285 Brown, P D et al Memantine for the prevention of cognitive dysfunction in patients receiving whole-​brain radiotherapy: a randomized, double-​blind, placebo-​controlled trial Neuro-​oncology 15, 1429–1437 (2013) 286 Bezjak, A et al Symptom response after palliative radiotherapy for patients with brain metastases Eur J Cancer 38, 487–496 (2002) 287 Doyle, M et al Quality of life in patients with brain metastases treated with a palliative course of whole-brain radiotherapy J Palliat Med 10, 367–374 (2007) 288 Gerrard, G E et al Investigating the palliative efficacy of whole-​brain radiotherapy for patients with multiple-brain metastases and poor prognostic features Clin Oncol (R Coll Radiol) 15, 422–428 (2003) 289 Steinmann, D et al Prospective evaluation of quality of life effects in patients undergoing palliative radiotherapy for brain metastases BMC Cancer 12, 283 (2012) 290 Wong, J et al Symptoms and quality of life in cancer patients with brain metastases following palliative radiotherapy Int J Radiat Oncol Biol Phys 75, 1125–1131 (2009) 291 Chow, E., Davis, L., Holden, L., Tsao, M & Danjoux, C Prospective assessment of patient-​rated symptoms following whole brain radiotherapy for brain metastases J Pain Symptom Manage 30, 18–23 (2005) 292 Fernandez, G., Pocinho, R., Travancinha, C., Netto, E & Roldão, M Quality of life and radiotherapy in brain metastasis patients Rep Pract Oncol Radiother 17, 281–287 (2012) 26 | Article citation ID: 293 Pulenzas, N et al Fatigue scores in patients with brain metastases receiving whole brain radiotherapy Support Care Cancer 22, 1757–1763 (2014) 294 Caissie, A et al Quality of life in patients with brain metastases using the EORTC QLQ-​BN20+2 and QLQ-C15-PAL Int J Radiat Oncol Biol Phys 83, 1238–1245 (2012) 295 Li, B et al Comparison of three treatment options for single brain metastasis from lung cancer Int J Cancer 90, 37–45 (2000) 296 DiBiase, S J., Chin, L S & Ma, L Influence of gamma knife radiosurgery on the quality of life in patients with brain metastases Am J Clin Oncol 25, 131–134 (2002) 297 Lowe, S S et al Associations between objectively measured physical activity and quality of life in cancer patients with brain metastases J Pain Symptom Manage 48, 322–332 (2014) 298 Fernandes, A W., Wu, B & Turner, R M Brain metastases in non-​small cell lung cancer patients on epidermal growth factor receptor tyrosine kinase inhibitors: symptom and economic burden J Med Econ 20, 1136–1147 (2017) 299 Burudpakdee, C et al Economic impact of preventing brain metastases with alectinib in ALK-​positive non-small cell lung cancer Lung Cancer 119, 103–111 (2018) 300 El-​Habashy, S E et al Novel treatment strategies for brain tumors and metastases Pharm Pat Anal 3, 279–296 (2014) 301 Crino, L et al Multicenter phase II study of whole-body and intracranial activity with ceritinib in patients with ALK-​rearranged non-​small-cell lung cancer previously treated with chemotherapy and crizotinib: results from ASCEND-2 J Clin Oncol 34, 2866–2873 (2016) 302 Bonoiu, A et al MMP-9 gene silencing by a quantum dot-​siRNA nanoplex delivery to maintain the integrity of the blood brain barrier Brain Res 1282, 142–155 (2009) 303 Pinaud, F., King, D., Moore, H P & Weiss, S Bioactivation and cell targeting of semiconductor CdSe/ZnS nanocrystals with phytochelatin-​related peptides J Am Chem Soc 126, 6115–6123 (2004) 304 Kodack, D P., Askoxylakis, V., Ferraro, G B., Fukumura, D & Jain, R K Emerging strategies for treating brain metastases from breast cancer Cancer Cell 27, 163–175 (2015) 305 Siravegna, G., Marsoni, S., Siena, S & Bardelli, A Integrating liquid biopsies into the management of cancer Nat Rev Clin Oncol 14, 531–548 (2017) 306 Sahgal, A et al Stereotactic radiosurgery alone for multiple brain metastases? A review of clinical and technical issues Neuro-​oncology 19 (Suppl 2), ii2–ii15 (2017) 307 Lin, N U et al Response assessment criteria for brain metastases: proposal from the RANO group Lancet Oncol 16, e270–e278 (2015) 308 Camidge, D R et al Clinical trial design for systemic agents in patients with brain metastases from solid tumours: a guideline by the Response Assessment in Neuro-​Oncology Brain Metastases working group Lancet Oncol 19, e20–e32 (2018) 309 Alexander, B M et al Clinical trial design for local therapies for brain metastases: a guideline by the Response Assessment in Neuro-​Oncology Brain Metastases working group Lancet Oncol 19, e33–e42 (2018) 310 Dummer, R et al Vemurafenib in patients with BRAF(V600) mutation-​positive melanoma with symptomatic brain metastases: final results of an open-​ label pilot study Eur J Cancer 50, 611–621 (2014) 311 Gadgeel, S M et al Pooled analysis of CNS response to alectinib in two studies of pretreated patients with ALK-​positive non–small-​cell lung cancer J Clin Oncol 34, 4079–4085 (2016) 312 Eisenhauer, E A et al New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1) Eur J Cancer 45, 228–247 (2009) 313 Di Giacomo, A M et al Three-​year follow-​up of advanced melanoma patients who received ipilimumab plus fotemustine in the Italian Network for Tumor Biotherapy (NIBIT)-M1 phase II study Ann Oncol 26, 798–803 (2015) 314 Wolchok, J D et al Guidelines for the evaluation of immune therapy activity in solid tumors: immune-related response criteria Clin Cancer Res 15, 7412–7420 (2009) Author contributions Introduction (A.S.A., R.C.R and S.D.C.); Epidemiology (L.N.); Mechanisms/pathophysiology (C.A and P.S.S.); Diagnosis, screening and prevention (R.S.); Management (M.S.A., A.S.A., R.C.R., S.D.C., G.R.H and N.D.A.); Quality of life (S.P.); Outlook (S.P.); Overview of the Primer (A.S.A, R.C.R and S.D.C.) A.S.A and R.C.R contributed equally to the manuscript Competing interests C.A has received funding support from Novartis, Merrimack, PUMA, Lilly, Merck, Cascadian/Seattle Genetics, Nektar, Tesaro and G1-Therapuetics; she has held uncompensated advisory roles with Novartis, Merrimack, Lilly, Genentech, Nektar and Cascadian/Seattle Genetics; she has held compensated advisory roles with PUMA, Merck and Eisai; and she has received royalties from UpToDate, Jones and Bartlett M.S.A has stock options in MimiVax and Doctible and has received grants and/or personal fees from Monteris Medical, AbbVie, BMS, AstraZeneca, Datar Genetics, CBT Pharmaceuticals, Kadmon Pharmaceuticals, Elsevier, NovoCure, Novartis, Incyte, Pharmacyclics, Tracon Pharmaceuticals, Prime Oncology, Flatiron, Merck, Bayer, Varian Medical Systems, VBI Vaccines and Caris Lifesciences S.P receives honoraria or consultation fees from AbbVie, Amgen, AstraZeneca, Bayer, Biocartis, Boehringer-​Ingelheim, Bristol-​Myers Squibb, Clovis, Daiichi Sankyo, Debiopharm, Eli Lilly, F Hoffmann-​La Roche, Foundation Medicine, Illumina, Janssen, Merck Sharp and Dohme, Merck Serono, Merrimack, Novartis, Pharma Mar, Pfizer, Regeneron, Sanofi, Seattle Genetics and Takeda; she has given talks in an organized public event for AstraZeneca, Boehringer-​I ngelheim, Bristol-​M yers Squibb, Eli Lilly, F Hoffmann-​La Roche, Merck Sharp and Dohme, Novartis, Pfizer and Takeda; and she is a (sub)investigator in trials (institutional financial support for clinical trials) sponsored by Amgen, AstraZeneca, Boehringer-​Ingelheim, Bristol-​Myers Squibb, Clovis, F Hoffmann-​La Roche, Illumina, Merck Sharp and Dohme, Merck Serono, Novartis, MedImmune and Pfizer P.S.S receives research funding from MedImmune The remaining authors declare no competing interests Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Reviewer information Nature Reviews Disease Primers thanks M Davies, G Rao, J Saunus, and other anonymous reviewer(s), for their contribution to the peer review of this work www.nature.com/nrdp (2019) 5:5 0123456789(); ... Pembrolizumab Asymptomatic melanoma brain metastases, ± prior brain metastasis-directed therapy or immunotherapy (18) Asymptomatic NSCLC brain metastases, ± prior brain metastasis-directed therapy ,... angiogenesis cannot be excluded Leptomeningeal metastases Leptomeningeal metastases represent a subset of brain metastases that grow in the lining of the brain or spine and/or in the CSF Lung and... utility of these agents that export these drugs from the brain 44,173 (Fig.  2) for patients with brain metastases (Table  2) From brain metastases undergoing radiosurgery, considering factors