Definition
- Tumours that originate outside the CNS and spread via the haematogenous route to the CNS or (less frequently) directly invade the CNS from adjacent anatomical structures.
Numbers
- Occur in about 25% of patients who die of cancer
- 1/2 brain tumours
- Most common brain tumours seen
- 14 cases per 100 000 population
- 15% of patients with cancer develop cerebral mets
- Location
- 80% occur in cerebral hemispheres,
- 15% in posterior fossa
- But the commonest tumour in the infratentorial region is mets
- 5% in the brainstem
- 15% of patients with no history of Ca present as cerebral mets
- The majority of BMs present metachronously
- Arise within 2 years of primary cancer diagnosis.
- inc. incidence of mets
- inc length of survival of Ca patients, due to better treatment
- Better imaging
- Many chemo do not cross BBB → safe haven for Mets
- Some chemo transiently weaken BBB → better access for mets
- Multiple lesion Norden et al., 2005
- 1/3 solitary lesion
- 1/3 oligometastatic (2– 3 lesions)
- 1/3 polymetastatic (>4 lesions)
- Time to diagnosis
- The median time between diagnosis of the primary cancer to presentation of brain metastasis is approximately 12 months but ranges from 3 (NSCLC) to 46 months (ovarian carcinoma) (Nussbaum et al., 1996; Kolomainen et al., 2002).
- Spinal mets
- Spinal epidural metastases
- 5-10% of pts
- Spinal leptomeningeal metastases
- Spinal intramedullary metastases
- Age
- CNS metastases are the most common CNS neoplasms in adults,
- CNS metastases account for only about 2% of all paediatric CNS tumours
- 6% of paeds CNS tumour will develop CNS cerebral mets
- 30% of adults and 6-10% of children with cancer develop brain metastases.
- Gender
- Relative proportions of various primary tumours are different between the two sexes,
- Sex has no significant independent effect on the occurrence of CNS metastasis for most tumour types
- Increasing incidence, due to
- Increased detection with improved imaging,
- An increase in the incidence of tumours with a predilection for brain involvement (e.g. lung cancer)
- Introduction of new therapeutic agents that prolong life and are relatively efficacious overall but ineffective at preventing or treating metastatic disease within the CNS
Origin of CNS metastases
Adults
- Most common source for cranial mets
- Incidence of brain metastases highest to lowest
- Primary lung cancer (age of 40-49 years)
- Primary melanoma (50-59 years)
- Renal cancer (50-59 years)
- Colorectal cancer (50-59 years)
- Breast cancer (20-39 years)
- Higher risk of brain metastasis (Pestalozzi et al., 2006):
- Triple negative (oestrogen receptor, progesterone receptor, and HER2 receptor negative)
- HER receptor positive subtypes
- Difference between propensity to spread to brain and actual percentage of brain mets as if you have a high propensity for brain mets but low frequency of tumour spread, for this specific tumour the percentage of brain mets will be low.
- Frequency of mets cancer and relationship to occurrence of brain mets
- Propensity for spread to brain parenchyma:
- High
- Melanoma,
- Small cell lung cancer,
- Choriocarcinoma,
- Other germ cell tumors;
- Intermediate
- Breast cancer,
- Non-small cell lung cancer (adenocarcinoma > squamous cell),
- Renal cell carcinoma;
- Low
- Prostate
- Colorectal
- Ovarian carcinoma,
- Thyroid cancer
- Sarcomas.
Lung cancer (especially adenocarcinoma and small cell carcinoma) | 44% |
Breast cancer | 10% |
Renal cell carcinoma | 7% |
Colorectal cancer | 6% |
Melanoma | 3% |
Undetermined, no primary tumour is found at presentation | 10% |
Cancer in Order of Frequency (Overall) | Cancer in Order of Frequency (Males) | Cancer in Order of Frequency (Females) | Propensity for Radiological Brain Metastases at 5 Years Follow Up (Descending Frequency) | Contribution to Total Brain Metastases Diagnosed (in Descending Frequency) |
(1) Breast (2) Prostate (3) Lung (4) Colorectal (5) Uterus (6) Malignant melanoma | (1) Prostate (26%) (2) Lung (14%) (3) Colorectal (13%) (4) Bladder (4%) (5) Kidney (4%) (6) NHL (4%) (7) Malignant melanoma (4%) | (1) Breast (31%) (2) Lung (12%) (3) Colorectal (11%) (4) Uterus (5%) (5) Malignant melanoma (4%) (6) Ovarian (4%) | (1) Lung (16–20%; SCLC 29% vs. NSCLC 12%) (2) Melanoma (7%) (3) Renal cell carcinoma (7–10%) (4) Breast (5%) (5) Colorectal (1%) | (1) Lung (2) Breast (3) Melanoma (4) Colon (5) Prostate (6) Liver/pancreas |
- Most common source for spinal mets
- Prostate
- Breast
- Lung cancer
- Non-Hodgkin lymphoma
- Multiple myeloma
- Renal cancer
Children
- Most common sources of CNS metastases
- Leukemia's and lymphomas,
- Non-hematopoietic CNS neoplasms such as
- Rhabdomyosarcoma
- Neuroblastoma,
- Germ cell tumors,
- Osteosarcoma,
- Ewing sarcoma,
Head and neck tumours direct invasion
- Bony destruction
- Along cranial nerves
Localization
Brain mets
- Intraparenchymal (75%)
- 80% in the cerebral hemispheres
- Particularly in arterial border zones and at the junction of cerebral cortex and white matter
- In MCA region: junction of temporal, parietal, and occipital Lobe
- 15% in the cerebellum: in adults consider it is mets until proven otherwise.
- via
- Spinal epidual venous plexus (batson's plexus) and
- Vertebral veins
- In the cerebrum, the metastatic cells preferentially adhere at the narrow vascular branch points found in the watershed zone at the grey white interface
- 5% in the brain stem.
- Extraparenchymal: mainly direct invasion from adjacent structures
- Dura
- Dural metastases are relatively common in cancer of the prostate, breast, and lung, and in haematological malignancies
- 8-9% of patients with advanced cancer
- Leptomeninges (aka meningeal carcinomatosis or carcinomatous meningitis)
- Pial, arachnoid mets
- 4-15% of patients with solid tumours
- Leptomeningeal metastases are more common in lung and breast cancer melanoma, and haematopoietic tumours
- Subarachnoid seeding
- Looks like sugar coating
- Locations for seeding
- Basal cisterns
- Interpeduncular cistern
- Cerebellopontine angle cistern
- Along the course of cranial nerves
- Over the convexities
- Can cause Hydrocephalus and meningeal irritation
CNS Primary | Non-CNS Primary |
Children | ㅤ |
Medulloblastoma (PNET) | Leukemia/lymphoma |
Ependymoma (blastoma=PNET) | Neuroblastoma |
Pineal region tumors | ㅤ |
Malignant astrocytomas | ㅤ |
Retinoblastoma | ㅤ |
Choroid plexus papilloma | ㅤ |
Adults | ㅤ |
Glioblastoma multiforme | Leukemia/lymphoma |
Primary CNS lymphoma | Breast |
Oligodendroglioma | Lung |
ㅤ | Melanoma |
ㅤ | Gastrointestinal |
ㅤ | Genitourinary |
- Metastatic CNS tumours seed along the walls of the ventricles or are located in the pituitary gland or choroid plexus.
Spinal Mets
- Vast majority of metastases affecting the spinal cord expand from the vertebral body or paravertebral tissues into the epidural space
- Spinal epidural metastases common in
- Prostate,
- Breast,
- Lung, and
- Kidney,
- Non-Hodgkin lymphoma, and
- Multiple myeloma.
- Intramedullary spinal cord metastases are more common in
- Small cell lung carcinoma
Tumour to-tumour metastasis
- Rare
- Found when donor (lung and breast cancer) metastasize to (recipient) meningioma
- Meningioma have a very vascular architecture that may be susceptible to deposition of mets from extracranial tumours
Pathology
- The capillaries around brain metastases have an abnormal microstructure and leaky gap junctions which, coupled with the local release of cytokines, make them permeable to large molecules and water then follows by osmosis.
- This intense vasogenic oedema contains few tumour cells, and is reduced with steroids or metastectomy.
Pathogenesis
- Steps to achieve prior to metastasis
- Escape from the primary tumour,
- Entry into and survival in the blood stream,
- Arrest and extravasation in the CNS,
- Survival and growth in the CNS microenvironment
- Direct extension from primary tumours in adjacent anatomical structures (e.g. paranasal sinuses and bone) not formally considered metastases, because they remain in continuity with the primary neoplasm.
- Once in contact with the CSF compartments, cells of those tumours may disseminate (seed) throughout the CNS.
Molecular genetics
- Biomarker tests to be considered for targeted therapies include
- EGFR mutation and ALK rearrangement in non-small cell lung cancer,
- ERBB2 amplification and estrogen and progesterone receptor expression in breast cancer,
- BRAF mutation in melanoma,
- RAS mutation in colorectal cancer, and
- ERBB2 amplification in gastro-oesophageal cancer
- Uggerly et al 2023
- Genomic alterations in RB1 or CTNNB1 were associated with a significantly higher risk of rapid recurrence at the resection site.
Clinical features
Cranial
- Caused by
- Increased intracranial pressure or
- Local effect of the tumour on the adjacent brain tissue
- Gradual presentation
- Headache (80%)
- Altered mental status
- FND (30– 40%)
- Paresis
- Ataxia
- Visual changes (6%),
- Sensory disturbances
- Nausea
- Acute Presentation
- Seizure (15– 20%), infarct, or haemorrhage
- The interval between diagnosis of the primary tumour and the CNS metastasis is
- < 1 year for lung carcinoma,
- Multiple years for breast cancer and melanoma
- Asymptomatic
- One series found 40% of patients with lung cancer and brain metastases were asymptomatic
Spinal
- Compression of the spinal cord or nerve roots
- Cause back pain, weakness of the extremities, sensory disturbances, and incontinence over the course of hours, days, or weeks
Investigation
- Malignant cells in initial CSF is + in 50% patients
- This may increase to > 80% when CSF sampling is repeated and adequate volumes (> 10 mL)
- Bloods
- Tumour markers
- Human chorionic gonadotrophin (ßHCG) for germ cell tumours,
- Alpha fetoprotein (αFP) for hepatocellular carcinoma,
- CEA for gastrointestinal, lung, and breast cancers,
- CA125 for ovarian tumours
- Faecal occult blood
- CT
- Current MRI technology more sensitive than CT and is the preferred imaging of choice.
- CT
- Mass may be isodense, hypodense or hyperdense (classically melanoma) compared to normal brain parenchyma with variable amounts of surrounding vasogenic oedema
- CT+C
- Enhancement is variable and can be intense, punctate, nodular or ring-enhancing if the tumour has out grown its blood supply.
- When to do CT TAP:
- Ali et al 2022
- Most sensitive predictors of metastatic disease as: (PIMS)
- Previous history of cancer
- Infratentorial location
- Multiple lesions
- Size <4cm
- A final protocol identified was found to have a sensitivity of 99.1% (AUC 0.704)
- MRI
- T1
- Typically iso- to hypointense
- Hyperintense
- Haemorrhagic mets (in ther acute stage)
- Non-haemorrhagic melanoma metastases: paramagnetic properties of melanin
- T1+C
- Enhancement pattern can be uniform, punctate, or ring-enhancing, but it is usually intense
- Delayed sequences may show additional lesions, therefore contrast-enhanced MR is the current standard for small metastases detection
- T2
- Hyperintense
- Haemorrhage may alter this
- FLAIR
- Hyperintense
- Hyperintense peri-tumoural oedema of variable amounts
- Edema is generally large for the size of met when compared to primary tumour
- DWI/ADC
- Oedema is out of proportion with tumour size and appears dark on DWI
- ADC demonstrates facilitated diffusion in oedema
- MR spectroscopy
- Intratumoural choline peak with no choline elevation in the peritumoural oedema
- Any tumour necrosis results in a lipid peak
- NAA depleted
- Cranial [18F] FDG- PET
- A sensitive way of detecting brain metastases in patients with known or suspected solid organ cancers,
- Distinguish
- Brain metastases VS lymphoma
- But not brain metastases VS high- grade glioma
- Difference between primary and metastatic brain tumour
Primary | Metastatic |
GBM at white matter tracts | At grey white junction |
Small amount of edema when compared to size of lesion | Large amount of edema when compared to size of lesion |
Rarer for multifocal and multicentric GBMs | More common to have multiple lesions but 1/3 are solitary |
Histology
- Macroscopic
- Grossly circumscribed and rounded,
- Greyish-white or tan masses
- Variable central necrosis
- Variable peritumoural oedema
- Specific findings
- Mucoid material: Adenocarcinomas
- Haemorrhage: choriocarcinoma, melanoma, and clear cell renal cell carcinoma.
- Melanin pigment (brown to black colour): Melanoma
- Leptomeningeal metastasis may produce diffuse opacification of the membranes or present as multiple nodules
- Dural metastases can grow as localized plaques or nodules and as diffuse lesions
- Microscopy
- The histological and immunohistochemical features of secondary CNS tumours are as diverse as those of the primary tumours from which they arise
- Tumour necrosis
- May be extensive,
- Leaving recognizable tumour tissue only at the
- Periphery of the lesion
- Around blood vessels
- Well demarcated and variable perivascular growth (so called vascular cooption) in the adjacent CNS tissue
- In leptomeningeal metastasis, the tumour cells are dispersed in the subarachnoid and Virchow-Robin spaces and may invade the adjacent CNS parenchyma and nerve roots
- Immunophenotype
- Similar to those of the tumours from which they originate.
- Metastatic CNS tumours show variable and often marked mitotic activity
- Proliferation index may be significantly higher than in the primary neoplasm
Frailty index Scoring system
- Risk Analysis Index: Preoperative frailty and 30-day mortality after resection of brain metastases
- Skandalakis 2023 et al
- https://nsgyfrailtyoutcomeslab.shinyapps.io/BrainMetsResection/
- RAI
- Was treated as a quantitative variable (0–81)
- Stratified into visual bins and frailty tiers:
- Robust (0–20),
- Normal (21–30),
- Frail (31– 40),
- Severely frail (≥ 41).
- Comparison of postoperative complications and outcomes in patients with RAI > 40 versus those with RAI ≤ 39, ACS-NSQIP, 2012-2020
Characteristic | All (n = 11,038) | RAI ≤ 39 (n = 10,141) | RAI > 40 (n = 897) | p Value* |
Major complication | 669 (6.1) | 587 (5.8) | 82 (9.1) | <0.001 |
Clavien-Dindo grade IV complication | 4477 (4.3) | 411 (4.1) | 66 (7.4) | <0.001 |
Return to operating room | 504 (4.6) | 456 (4.5) | 48 (5.4) | 0.24 |
eLOS | 2801 (25) | 2425 (24) | 376 (42) | <0.001 |
Length of hospital stay | 5.0 (6.0) | 5.0 (5.0) | 7.0 (8.0) | <0.001 |
Hospital LOS >30 days | 55 (0.5) | 43 (0.4) | 12 (1.4) | 0.001 |
NHD | 2527 (23) | 2137 (21) | 390 (43) | <0.001 |
In-hospital mortality | 156 (1.4) | 116 (1.1) | 40 (4.5) | <0.001 |
Unplanned reop | 478 (4.3) | 432 (4.3) | 46 (5.1) | 0.22 |
Unplanned readmission | 1368 (12) | 1240 (12) | 128 (14) | 0.075 |
Mortality w/in 30 days | 464 (4.2) | 351 (3.5) | 113 (13) | <0.001 |
Values are shown as median (IQR) or number (%) unless indicated otherwise.
*Determined with the Pearson chi-square test or Fisher exact test.
Prognostic factors
- Bindal 1993 Modha 2005
Treatment | Median Survival |
W/o treatment | 4 wks |
High dose Dex | 8 wks |
WBRT | 12-24 wks (3-6 months) |
Surgery/SRS | 60 wks (14 months) |
- Main established prognostic factors for patients with brain metastases are
- Patient age,
- Karnofsky performance status,
- Number of brain metastases
- CHANG 2000.pdf
- Probability of tumour control after surgery + WBRT/SRS
- Status of extracranial disease.
- Other factors of prognostic significance include
- Specific tumour type and
- Molecular drivers involved (e.g. ERBB2 in breast cancer)
- Neuroradiological parameters such as peritumoural brain oedema may also provide prognostic information
Number of intracranial mets | Tumour control |
1 | 64% |
2 | 51% |
3 | 41% |
- The prognosis of patients with brain metastases may be estimated by scoring systems of increasing complexity.
- The Recursive Partitioning Analysis (RPA) paper statistically examined a series of patients treated with radiotherapy and described a three- class system, with no benefit found from treating patients with a poor performance score.
- The graded prognostic assessment (GPA) included patients from various trials treated with different strategies, and split each feature and included the number of metastases, while the diagnosis specific GPA included different primary cancer types.
- The one factor that is common to every scoring system and disease is the Karnofsky performance score.
- Sperduto 2007.pdf
- RPA, recursive partitioning analysis; GPA, graded prognostic assessment; DS, disease specific; KPS, Karnofsky performance score; ECM, extracranial metastases; NSCLC, non- small cell lung cancer; SCLC, small cell lung cancer; GI, gastrointestinal; RCC, renal cell cancer.
- Treatment options include surgery, whole-brain radiation therapy (WBRT), SRS, or some combination thereof.
Q&A
- Most common hemorrhagic metastatic brain tumors?
A: In order of highest occurrence: lung, breast, renal cell carcinoma, choriocarcinoma, and melanoma
Made with Bullet