Basic Science of Brain Metastases

This chapter includes an accompanying lecture presentation that has been prepared by the authors: .

Key Concepts

Melanomas and lung, breast, colorectal, and renal cancers show the most proclivity for the brain, followed more uncommonly by thyroid, gastrointestinal (GI), and prostate cancers. Lung cancer is the primary tumor in about 40% to 50% of patients diagnosed with brain metastases (BMs). Breast cancer is unique in its longer latency to forming BMs, and this has been associated with the acquisition of neuronal characteristics.
The microenvironment is a prominent player in metastatic outgrowth, underscoring the key principle that for tumor cells to become established in a specific region, they must possess a certain phenotype that imbues them with the ability to survive within that region. It is becoming increasingly clear that cancer cells are able to exploit the expression of CNS-specific genes to ease their way through the BBB into the brain metastatic microenvironment.
There is now evidence to suggest that reciprocating communication between cancer cells and the tumor microenvironment orchestrates homing, dormancy, proliferation, formation of micrometastases, and tumor growth. This communication may involve any number of molecules such as neurotransmitters, growth factors, cytokines, and many other contributors that facilitate reciprocating communication between tumor cell and host stroma.
The current standard of care for BMs includes surgical excision, whole-brain radiotherapy (WBRT), and stereotactic radiosurgery (SRS) combined with steroids. Targeted molecular therapies show little or no effect mainly because of their inability to penetrate the elusive blood-brain barrier (BBB).

Brain metastases (BMs) are the most common cause of intracranial neoplasms in adults with invasive cancers; 20% to 45% of cancer patients are diagnosed with BMs in their lifetime. Metastatic brain tumors occur at a much higher rate than both adult and pediatric primary brain tumors, and they are the major cause of mortality from malignant brain disease. ^, Advances in treating primary cancers have led to patients living longer, therefore they are more likely to experience brain metastasis complications. Multiple BMs make their prognosis challenging and worsen the long-term rate of survival. Intracranial metastasis is associated with poorer prognosis (<1 year), moderate to severe neurodegeneration, and overall reduction in quality of life. Of those who develop BMs, 60% to 75% of patients become symptomatic.1 At diagnosis, symptomatic patients usually present with headaches, seizures, motor weakness, and dysphasia. The incidence and severity of brain metastatic disease varies according to the origin of the primary tumor and the treatment strategy followed for the patient. Melanomas and lung, breast, colorectal, and renal cancers show the most proclivity for the brain, followed more uncommonly by thyroid, gastrointestinal (GI), and prostate cancers.

Intracranial metastases were usually considered end-stage, and patients were only subjected to palliative therapy. However, patient survival has increased as a result of advances in control of primary tumors and extracranial metastases as well as superior methods of early detection. This has led to increasing numbers of patients being diagnosed with BMs, with or without concomitant extracranial disease. Tumors originating from different tissues show varying latency to metastasize to the brain. This can be explained by the aggressiveness of the tumor type, modes of dissemination, development of resistance to therapy, or molecular affinity for the neuronal niche. Despite considerable advances in elucidating the cellular and molecular events underway in metastasis, few treatments have been realized, and as such prognoses remains poor. The current standard of care for BMs includes surgical excision, whole-brain radiotherapy (WBRT), and stereotactic radiosurgery (SRS) combined with steroids. Targeted molecular therapies show little or no effect mainly because of their inability to penetrate the elusive blood-brain barrier (BBB). Thus, with limited options available, malignant tumors take refuge in the brain and escape most forms of intervention, contributing to patient mortality. The management of BMs is an urgent unmet clinical need and warrants immediate attention and investigation. This chapter summarizes and provides perspective on the current understanding of brain malignancies in the context of the tumor; their clinical biology, diagnosis, and management; and the neuronal microenvironment. We hope this knowledge is translated into innovative therapeutic interventions that will enhance the quality and expectancy of life for affected patients.

Diagnosis of Brain Metastases

Sixty percent to 75% of BMs are symptomatic, while a smaller number of patients harbor CNS metastases without any neurological signs. Imaging is crucial in the detection and diagnosis of BMs. It is used to confirm previously undiagnosed CNS metastases in patients with neurological symptoms, to confirm brain involvement in patients with systemic metastatic disease, and to stage and monitor BMs over the course of therapy. Imaging is also essential before surgery to plan safe excision of the tumor from the brain.

BMs can present as solitary or multiple lesions. Most BMs are solitary, with 20% of diagnoses having less than or equal to two lesions and 30% having greater than or equal to three lesions. Lung cancers and melanomas primarily lead to multiple BMs, while breast, renal, and colon cancers usually present as single lesions. Contrast-enhanced MRI is the preferred method of detection, but nonenhanced computed tomography (NECT) is often used for initial screening purposes, especially for naïve patients who present with new neurological symptoms. For certain cancers like small cell lung cancers (SCLCs), contrast-enhanced CT (CECT) is equivalent to detection by MRI because no survival benefit is offered by MRI versus CT. On MRI, metastases appear hypointense on T1-weighted imaging and hyperintense on T2-weighted imaging. Gadolinium contrast MRI is important in detection of small multiple lesions, and its use improves diagnostic confidence.

Diagnosis of nonparenchymal BMs can be challenging. Pachymeningeal (dural) metastases can be hard to distinguish from meningiomas. For this, information about prior history of systemic or CNS metastasis in the patient is key. Detection/diagnosis of leptomeningeal metastases is also difficult by CT, and therefore evaluation of the CSF in suspected cases is essential for confirmation.

Cancer Type and Propensity of Brain Metastasis

Metastasis is a carefully orchestrated process involving breakage of cells from the primary tumor, gain of invasive properties, and interaction with the microenvironmental niche to establish tumors at distant sites. Clinically, different kinds of tumors show varying proclivities in their ability to successfully metastasize within the CNS. The most common primary sites are lung cancer, breast cancer, and melanoma (in that order). These are followed more uncommonly by thyroid, GI, and prostate cancers. Propensity of occurrence of brain metastasis can also be further ranked by clinical subtypes of solid cancers.

Lung Cancer

Lung cancer is the primary tumor in about 40% to 50% of patients diagnosed with BMs. Fifty percent of lung cancer BMs occur at disease presentation/diagnosis, and sometimes the CNS is the only location of dissemination. This highlights the aggressive nature of the primary tumor and the short latency period seen for lung-to-brain metastasis. Because of a higher incidence, non–small cell lung cancers (NSCLCs) compose a higher percentage of BMs than SCLCs, but recent studies show that SCLCs have a higher biologic propensity for CNS spread. About 7% of patients with NSCLC present with BMs at diagnosis, and about 25% to 30% develop BMs over the course of their disease. Surgical resection has been shown to improve disease control in patients with solitary resectable lung-to-brain metastasis (LBM). A combined approach of surgery and WBRT has also improved intracranial disease control for such patients, although the efficacy of this combined approach has not yet been unequivocally verified in patients with multiple or advanced LBMs. Targeted therapy is recommended for NSCLC patients with genetic mutations, but in cases of mass effect or impending brain herniation, surgery and subsequent systemic therapy are preferred. Brain metastasis is seen at diagnosis in 23.8% of ALK -rearranged tumors and in 24.4% of EGFR -mutated NSCLCs. Patients who harbor mutations in such oncogenic drivers may be treated using targeted therapies. ALK gene fusions have been reported in 2% to 7% of patients with NSCLC, but the percentage may be higher in select groups. Crizotinib was used as a first-generation ALK inhibitor for systemic treatment, but patients developed CNS metastases within 1 year of starting therapy. Crizotinib also has low CNS penetrance and is a substrate for P-glycoprotein efflux transporters, but it showed CNS disease control rates comparable to systemic disease control rates. Current trials are investigating methods to increase CNS availability of crizotinib. Ceritinib, alectinib, brigatinib, and lorlatinib are second-generation ALK tyrosine kinase inhibitors that have shown improved CNS efficacy and increased intracranial response rates in ALK ⁺ NSCLC patients with BMs. ^, ^, These drugs are more effective partly because of their increased CNS availability and BBB penetration.

Ten percent to 20% of patients with lung adenocarcinoma exhibit EGFR mutations. The first- and second-generation EGFR inhibitors erlotinib and afatinib, respectively, can be used as standalone treatments (without chemotherapy) for EGFR -mutated NSCLC. These drugs have moderate CNS penetrance and have shown substantial intracranial response rate (RR). ^, However, secondary EGFR T790M mutations have arisen as a mechanism of resistance to these drugs. Osimertinib is a third-generation EGFR -mutant inhibitor effective against EGFR T790M that is approved by the US Food and Drug Administration. Osimertinib has BBB permeability greater than older EGFR inhibitors ^, and was shown to be effective in a phase 1 trial in leptomeningeal disease in EGFR -mutant NSCLC.

The median overall survival (OS) is 4.9 months for SCLC patients with BMs and 1.9 to 2.4 months for patients with leptomeningeal metastases. SCLC usually manifests as multiple brain lesions, thus surgery is not a common therapeutic avenue for patients. Therapeutic WBRT is the standard of care for SCLC patients with BMs who have never undergone prophylactic cranial irradiation (PCI). SRS is only performed in cases of recurrent BMs to avoid exacerbation of cognitive decline by repeated WBRT. Systemic chemotherapy is not routinely used for SCLC patients with newly diagnosed or recurrent BMs. Several studies have investigated the use of systemic nontargeted cytotoxic drugs (e.g., cisplatin, temozolomide, etoposide) along with WBRT and have shown some CNS RR, but OS still remains dismal. There are very few indications of SCLC-specific targets, unlike NSCLC, but immunotherapy trials with antibodies targeting programmed cell death (nivolumab) and CTLA4 (ipilimumab) are underway.

Breast Cancer

Up to 30% of all breast cancer patients are diagnosed with BMs within their lifetime. Out of these patients, those diagnosed with the triple-negative breast cancer (TNBC) subtype show increased risk for brain metastasis, followed by the HER2 ⁺ and hormone-positive (ER ⁺ /PR ⁺ ) subtypes. ^, Successful control of extracranial breast carcinoma and the emergence of advanced diagnostics has increased the incidence and diagnosis of breast-to-brain metastases (BBMs). Breast cancer cells that escape chemotherapy or surgical extraction can persist in the patient’s body and eventually lead to BBMs. Breast cancer is unique in its longer latency in forming BMs, and this has been associated with acquisition of neuronal characteristics. ^, A recent clinical study also substantiated that BBMs are able to acquire mutations in clinically targetable genes (HER2) and that about 20% of HER2 ^– -diagnosed breast tumors showed a switch to HER2 positivity in the brain. This highlights the need for molecular characterization of highly invasive breast tumors for patients over the course of their disease to avoid nonrelevant therapeutic interventions.

Currently, there are no breast cancer–specific treatments for BMs in particular, and a multivariate approach with surgery and radiotherapy is used. For a limited number of brain-metastatic lesions (1 to 4), surgical resection followed by SRS or WBRT is performed. SRS is preferred over WBRT to avoid the associated neurocognitive decline, similar to treatment in other cancers with BMs. For multiple metastatic lesions (>4) or when surgery or SRS is not feasible, WBRT is performed, with a shift toward hippocampal-sparing WBRT. This approach is reasonable given the low incidence of BBMs occurring in or around (5-mm margin) the hippocampus. Currently, systemic drugs such as lapatinib are being investigated as radiosensitizers. Recent trials also investigated the inclusion of prophylactic memantine (an oral NMDAR antagonist), which may delay the loss of cognitive capabilities after WBRT. Accordingly, newer regimens with the combined use of memantine and hippocampal-sparing WBRT are being tested in patients.

Systemic targeted therapies, especially for HER2 ⁺ breast cancer (trastuzumab, lapatinib) have limited CNS permeability and are not protective against BMs. Lapatinib as monotherapy showed very meager CNS response but showed improved efficacy in BM patients when combined with capectabine. ^, Targeted therapies like neratinib and tucatinib are now being investigated for CNS response and efficacy. PI3K and cell cycle (cyclin-dependent kinase [CDK]) inhibitors in the context of hormone-positive breast cancers are in trials for CNS response in patients with BMs. For TNBC patients, poly ADP ribose polymerase (PARP) inhibitors and microtubule inhibitors are currently in trial for treatment of CNS metastases.

Melanoma

BMs are a significant complication in patients diagnosed with advanced melanoma. About 20% of these patients present with BMs at diagnosis, and about 40% to 50% develop them during the course of the disease. Melanoma brain metastases (MBMs) can present as single or multiple intracranial lesions, with a more favorable prognostic index for patients presenting with a single BM at melanoma diagnosis. The median survival for melanoma patients with BMs is 3 months (without treatment) to approximately 9 months (with treatment).

Therapy for MBMs has been controversial and is evolving as more postdiagnostic and post-therapeutic data are collected. Surgery is preferred for patients with solitary MBM and has an OS benefit when compared with patients who receive radiotherapy alone. Surgery has an advantage of symptomatic relief and procurement of tumor tissue for molecular characterization. WBRT has been shown to confer survival benefit to melanoma patients, with a favorable prognostic profile according to Karnofsky Performance Score (KPS), age, and number of extracranial metastases. WBRT after surgery or SRS, however, does not improve survival but leads to better intracranial disease control. Because of observed neurocognitive decline after WBRT, hippocampal-sparing WBRT and inclusion of prophylactic memantine are in trials for MBM therapy. SRS is used to achieve local control in patients with small (<3 cm) and fewer than three brain lesions. SRS use is dependent on a variety of factors including accessibility of the lesion, proximity to eloquent brain regions, and suitable candidacy of the patient. About 50% of patients with advanced melanoma exhibit BRAF mutations. Inclusion of BRAF -targeted small molecules in metastatic melanoma therapy has shown significant results and is being tested for intracranial disease. A retrospective study showed that patients with MBMs who were treated with the BRAF inhibitor vemurafenib showed 71% control of intracranial disease and 50% control of extracranial disease. Immunotherapy, especially immune checkpoint inhibitors like CTLA4, PD-1, and PDL-1 antibodies, have shown promise in metastatic melanoma. A phase 2 trial with ipilimumab in patients with small and asymptomatic MBMs showed that 24% of patients showed a partial response or stable disease.

Theories of Metastasis

The Paget Seed and Soil Hypothesis

In 1889 Stephen Paget observed the preferential migration of breast cancers to bone and postulated that interactions between tumor cells (“seeds”) and host tissues (“soil”) were instrumental in modulating the metastatic process. Studying 650 autopsy specimens in patients with breast cancer, Paget commented on the extraordinarily high rates of secondary tumor growth in the distal ends of femurs and skulls. These striking observations led him to formulate the seed and soil hypothesis: “when a plant goes to seed, its seeds are carried in all directions: but they can only live and grow if they fall on congenial soil.” This paved the way for our current mode of thinking, which figures the microenvironment as a prominent player in metastatic outgrowth, underscoring the key principle that for tumor cells to become established in a specific region, they must possess a certain phenotype that imbues them with the ability to survive within that region. ^, The seed and soil theory is a gross simplification of the metastatic cascade, but it highlights a fundamental principle in cancer biology: that growth of secondary tumors requires distinct cues to flourish, failure of which precludes their formation to begin with.

Cloned Evolution Theory Versus Cancer Stem Cell Theory

Two major theories for breast cancer tumorigenesis and dissemination have been posited: the cloned evolution theory and the cancer stem cell theory. In cloned evolution theory, the hypothesis is that all cells are created equal and thus have the same potential for tumor induction and progression. Cancer stem cell theory, however, states that solid tumors are created from and maintained by a subset of self-renewing cells termed cancer stem cells. The cancer stem cell theory is the predominating model, with data supporting its validity coming from work showing that only a small population of tumor cells is capable of initiating osseous spread and disease establishment. ^, In this regard, there is work showing a relationship between epithelial-mesenchymal transition (EMT) and acquisition of stem cell phenotype by cancer cells. ^, Indeed, experiments on human tumors reveal a high number of stem cells and changes in mesenchymal markers and cellular architecture (mamospheres) that may facilitate disease spread. CD44 ^low CD24 ^high breast cancer cells were shown to be capable of tumorigenesis.

Epithelial-Mesenchymal Transition Theory

Type 1 EMT is a normal physiologic process that contributes to embryonic gastrulation during development. Type 2 EMT is evident in tissue regeneration and organ fibrosis. Type 3 EMT is the oncogenic version implicated in cancer pathways. Spread of epithelial cancers like breast and prostate cancers requires alterations in tumor phenotype whereby a transition occurs from a sessile epithelial state to a motile mesenchymal state, the so-called epithelial-mesenchymal transition. Subsequent seeding of these cells into the circulation is the basis for metastasis to distant sites. Once circulating tumor cells (CTCs) lodge in a distant organ, a reversal of fate is then necessary for tumor growth such that the newly motile mesenchymal cells revert to the sessile epithelial state. EMT phenotype reversal is essential for colonization to occur. There is now data to suggest that PRX1 and TWIST1 work together to initiate EMT while suppressing stem cell– like characteristics. Eventually, suppression of the two reconstitutes stem cell features, permitting tumor establishment.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here