Leptomeningeal Metastases Workup

In a patient with known cancer (especially with a lung, breast, lymphoma, or melanoma primary), the index of suspicion for leptomeningeal metastases (LM) should be high. CNS symptoms or signs usually trigger diagnostic imaging, typically by MRI with and without gadolinium contrast, which may include the brain and/or entire spine depending on the individual patient. Cytological evaluation of the CSF is still the "gold" standard for diagnosing LM, and also helps assess treatment response. ^[16]

CSF cytology, when positive for malignant cells, is definitive for disease confirmation. However, false negatives are common and can occur in up to 50% of a single CSF analysis even with a large sample volume (> 10 ml). ^[17]

Diagnosis of leptomeningeal metastases (LM) is typically made with positive CSF cytologic results (the most useful test), subarachnoid metastases identified on MRI, or a history and physical examination suggestive of LM along with abnormal CSF findings (typically mild pleocytosis, elevated protein and decreased glucose). ^[3]

Evaluation for LM is considered for patients presenting with:

• Neurologic signs and symptoms at more than 1 level of the neuraxis (present in 75% of patients with LM).

• Neurologic signs and symptoms consistent with a single lesion but with no mass evident on imaging.

• Neurologic signs and symptoms consistent with inflammatory meningitis but without fever.

• Imaging showing leptomeningeal enhancement or CSF flow obstruction.

• Elevated CSF protein level in a patient with cancer but without known cerebral metastases.

The first step in the diagnostic workup should be gadolinium-enhanced MRI of the area of maximal symptomatology, followed by a lumbar puncture (LP) if the patient has no evidence of increased ICP, repeated as many as 3 times or until findings are positive.

In general, imaging findings are consistent with or suggestive rather than diagnostic of leptomeningeal metastases (LM). They are also useful in detecting secondary complications of LM, such as hydrocephalus, periventricular edema, gyral effacement, and spinal cord compression.

About 50% of patients with LM have abnormal imaging findings, most commonly contrast enhancement of the basilar cisterns, cortical convexities, Sylvian fissure, tentorium, ventricles, or cauda equina, or preence of hydrocephalus even without a discrete obstructive lesion. Identification of enhancing lesions usually follows positive cytologic findings, by up to 6 months.

MRI of the spinal cord involvement can show nerve-root thickening, cord enlargement, intraparenchymal and subarachnoid nodules, or epidural compression.

MRI

Gadolinium-enhanced MRI of the entire CNS is used in patients with cancer and neurologic symptoms to look for metastases and to determine the risk of possibly catastrophic brain herniation from LP. 

MRI (1.5T) has been reported to be similar in sensitivity to CSF cytology in the diagnosis of LM for patients with solid tumors but less sensitive than CSF cytology for patients with LM from leukemia or lymphoma. ^[18]  Normal MR imaging does not exclude the diagnosis of LM.

Meningeal enhancement can also be seen in infections, inflammatory diseases, trauma, or subdural hematoma; after craniotomy; and sometimes after LP.

Delineation of the extent of enhancement on MRI may be used to discriminate LM (in addition to cytology) from metastases and/or epidural compression, and might therefore be used to inform decisions regarding radiation therapy and/or radiosurgery.

CT scan

Brain CT with and without contrast may also be useful to look for metastases, hemorrhage and hydrocephalus, and to determine the risk of herniation from LP. While enhancement of structures may sometimes be seen, this test is even less sensitive (than MRI) for detecting LM itself. As with MRI, normal CT imaging does not exclude the diagnosis of LM.

Myelography

Although seldom indicated, myelography may show nodularities or thickening of the nerve roots in approximately 25% of patients with LM. Myelography can show intra-arachnoid nodular filling defects, longitudinal striations, prominent and crowded nerve roots of the cauda equina, or scalloping of the subarachnoid space.

CSF flow studies

Radionuclide studies using either¹¹¹ indium-diethylenetriamine penta-acetic acid or⁹⁹ Tc macroaggregated albumin can be used to assess CSF flow, which is abnormal in 30–40% of patients with LM. Abnormal CSF flow must be addressed prior to the administration of intrathecal chemotherapy, as it can prevent delivery throughout the CSF space.

Cerebral arteriography, MRA, EEG, and electromyography (EMG) are rarely indicated for leptomeningeal metastases (LM) itself but may be needed according to individual patient circumstances.

Monoclonal antibodies can be useful in diagnosing CSF lymphoma, particularly if cytologic examination cannot distinguish between reactive lymphocytes and malignant lymphocytes.

Hormonal status is an important prognostic factor in patients with breast cancer-related LM. Patients with positive hormone receptor status have been shown to have a longer time from diagnosis to development of LM and a greater chance of survival.

Lumbar puncture and CSF cytology

If safe to perform, lumbar puncture is the most useful test.

Analysis of CSF obtained by lumbar spinal puncture is more accurate than that obtained by using a ventricular catheter, as ventricular fluid usually has higher glucose and lower protein levels and is less likely to yield positive cytologic findings. For this reason, periodic LP is recommended, even in patients with catheters.

Measure the opening pressure (elevated in 50% of patients) and send the CSF for an analysis of cytology, flow cytometry, cell counts, and protein and glucose levels.

Carcinoma cells in the CSF are diagnostic, with the exception of a few false-positive results in patients who have reactive lymphocytes (which are difficult to distinguish from malignant lymphomatous cells) because of an infectious or inflammatory process in the CSF. However, negative cytologic findings do not rule out the diagnosis, as 50% of patients with LC have a negative cytologic result on the first LP. This percentage drops to 20% after 2 high-volume LPs and 15% after 3.

Cytologic findings are more likely to be positive in patients with extensive leptomeningeal involvement than in patients with focal involvement because CSF obtained from a site distant to the pathology is more likely to yield negative pathology.

Other causes of false negatives can include not obtaining CSF from a site of symptomatic or radiographically demonstrated disease, withdrawing < 10.5 mL CSF, delayed processing of samples, and obtaining only 1 sample.

CSF pleocytosis and modest protein elevations are consistent with but not indicative of the diagnosis, but reduced glucose levels usually are seen only with LM (ie, abnormal glucose transport) or infection (ie, increased glucose utilization).

The lymphocyte count is elevated in more than 50% of patients with LC, and the presence of eosinophils should raise the suspicion of lymphomatous infiltration (except patients who are given ibuprofen).

CSF samples in LM patients with solid tumors have a greater number of inflammatory cells and a different leukocyte distribution than CSF samples from patients with lymphomatous LM. CSF polymorphic neutrophils (PMN) are more likely to be present in patients with LC than in patients with LM or patients with brain metastases due to solid tumors without LM. ^[19]

Flow cytometry immunophenotyping (FCI) may be helpful in identifying epithelial cell cancers. Compared with routine cytology, FCI had greater sensitivity (79.79% vs. 50%) and negative predictive value with lower specificity (84% vs. 100%) and positive predictive value. Patients with 8% or more epithelial cell adhesion molecule positive cells had statistically worse survival. ^[20]

Xanthochromia can occur from leptomeningeal bleeding, which is most likely in LM from a melanoma.

Biochemical markers in the CSF

Most biochemical markers in CSF have poor sensitivity and specificity, but when present, levels decline with successful therapy. Their reelevation can thus signal a relapse before any other findings become apparent. Useful markers include carcinoembryonic antigen (CEA) from adenocarcinomas, alpha-fetoprotein and beta-human chorionic gonadotropin from testicular cancers, 5-hydroxyindoleacetic acid (5-HIAA) from carcinoid tumors, and immunoglobulins from multiple myeloma; their presence in CSF is virtually diagnostic. Nonspecific markers such as endothelial growth factor can be strong indirect indicators of LM, but none are sensitive enough to improve the cytological diagnosis.

Epithelial-associated glycoprotein (HMFGI antigen) has been reported to be present in 90% of LM patients.

Cytokeratins measured by tissue polypeptide antigen (TPA) and tissue polypeptide-specific antigen (TPS) have 80% sensitivity to LM from breast cancer.

Neither CEA nor beta-glucuronidase is helpful in detecting solid tumors or metastases, nor are they useful in detecting leptomeningeal lymphomatosis. However, if their levels are elevated, a return to normal levels of both markers signifies successful treatment.

Elevated CSF CEA is specific, unless serum levels are unusually high (ie, >100 ng/mL). The combination of CEA with a second tumor marker CYFRA 21-1 in lung cancer patients increased specficities to 100%, and elevations of either CEA or CYFRA 21-1 were associated with a 100% sensitivity. ^[21]

CSF beta-glucuronidase values are frequently elevated, but wide fluctuations make it unreliable as a marker, and elevations also occur with bacterial, viral, fungal, or tubercular meningitis. In association with elevated lactate dehydrogenase (LDH), however, high CSF beta-glucuronidase levels can indicate LM from a breast primary tumor with a high sensitivity and specificity.

CSF fibronectin values are elevated in LM but also in bacterial meningitis and tick-borne encephalitis.

Myelin basic protein can indicate disease activity, particularly if values are measured longitudinally.

CSF vascular growth factor has recently been suggested as a useful biomarker. ^[22]

Antithrombin III has been suggested as a useful biomarker in patients with primary CNS lymphoma but has not been evaluated in patients with LM.

For lymphoma and leukemia, the weight of the evidence (as well as recent National Comprehensive Cancer Network guidelines) suggests that flow cytometry is more sensitive than cytology and should be used instead. ^{[5, 23]}

Monoclonal antibodies are not more sensitive than cytology but can be used to distinguish between reactive and neoplastic lymphocytes in the case of LM from lymphoma.

Creatine-kinase BB isoenzyme (CK-BB), tissue polypeptide antigen (TPA), b_2- microglobulin, β -glucuronidase, LDH isoenzyme-5, and vascular endothelial growth factor (VEGF) are strong indirect indicators of LM, but are not sensitive enough to improve on cytology.

LDH concentrations are elevated in cases of stroke, bacterial meningitis, CSF pleocytosis, head injury, primary CNS tumors, and some metastases. Levels are also elevated in 80% of patients; therefore, they can be useful in confirming the diagnosis. LDH isoenzyme-5 levels are elevated in LM from breast or lung primary tumors and melanoma, as well as bacterial meningitis, but they are sometimes normal even when cytologic findings are positive

Levels of CSF β ₂ -microglobulin may be useful in detecting LM caused by hematologic spread but not in LM from solid tumors. levels may be elevated after treatment with intrathecal methotrexate (MTX).

Ferritin levels are sensitive to inflammatory changes in the CSF, but they are nonspecific for early LM.

CSF alkaline phosphatase levels may be elevated in an LM from a lung primary tumor.

CSF prostate-specific antigen (PSA) may be elevated in an LM from a prostate primary tumor.

PCR is not useful as the precise genetic alteration of the neoplasia is usually not known.

An NMR metabolomics approach to LM diagnosis has been proposed. In a pilot study, a combination of specific CSF biometabolites was associated with a higher likelihood of LM.

Leptomeningeal biopsy may be necessary if the patient has no evidence of a primary tumor. The findings can be diagnostic if results of all other tests are negative, especially if taken from an enhancing region identified on MRI. Macroscopic pathology shows diffuse fibrotic thickening of the brain and spinal cord, as well as layering of the nerve roots with tumor tissue. Microscopic examination shows local fibrosis with tumor cells covering the blood vessels and nerve roots, either as a single layer or as aggregates.

Staging varies by primary cancer, but LM represents metastatic disease that, by definition, is a stage IV malignancy.