Table 1.
Test characteristics for the detection of leptomeningeal metastases in breast cancer patients by using CSF biomarkers
| Ref. | Method | Number of Patients with LM from BC versus LM from all tumor types | Biomarker | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Definition of LM | Definition Control Group (no. of patients) |
|---|---|---|---|---|---|---|---|---|---|
| Pro-angiogenic proteins | |||||||||
| Stockhammer et al19 | ELISA VEGF | 4 out of 11 | VEGF | NA | NA | NA | NA | Clinical signs, MRI findings, and positive CSF cytology | I: Pts with solid malignancies without LM (16) II: Pts with infectious meningitis (35) III: Pts with nonmalignant and noninfectious diseases (100) |
| Corsini et al20 | ELISA VEGF | 11 out of 18 | VEGF index (cutoff 10) | 83.3 | 88.4 | NA | NA | Positive CSF cytology | Non-malignant neurological diseases: inflammatory (33), neurodegenerative (17) |
| Herrlinger et al21 | ELISA VEGF | 5 out of 37 | VEGF (cutoff 100 pg/mL) VEGF (cutoff 250 pg/mL) | 73 51.4 | 93 98.3 | 77.1 90.5 | 91.5 86.3 | Positive CSF cytology, or contrast-enhancing subarachnoid tumor cell deposits on MRI or both | Non-malignant neurological disease (50), multiple sclerosis (28), presumed CNS infectious disease (37) |
| Reijneveld et al22 | ELISA VEGF | 31 out of 53 | VEGF | NA | NA | NA | NA | Clinical features compatible with LM and positive CSF cytology | I:Pts with malignancies without LM (negative CSF and no signs of LM or brain metastases on imaging) (18) II: No malignancy (no systemic malignancy, no CNS infection, no signs of LM or brain metastases on imaging) (25) |
| Langerijt, van de et al23 | ELISA VEGF, tPA | 9 out of 19 | VEGF index tPA index VEGF + tPA | 54a/62b 85a/85b 100 |
72a/71b 79a/96b 73 |
54 | 100 | Positive CSF cytology and/or enhancement of leptomeninges on MRI | I: Pts with solid malignancy (negative cytology and MRI) (54) = a II: Infectious meningitis: viral (16), bacterial (16), nonmalignant and noninfectious neurologic disorders (27) = b |
| Groves et al24 | ELISA VEGF | 8 out of 22 | VEGF (all tumor types) VEFG (BC) | 71 75 |
97 97 |
88 86 |
91 94 |
Positive CSF cytology | I: BC (33), lung (25), melanoma (9) pts suspected for LM with negative CSF cytology. II: No disease at all (1) |
| CK-BB, LDH, β2-microglobulin | |||||||||
| Bach et al25 | CK-BB bio-luminescence assay | 12 out of 12 | CK-BB (cutoff 0.2 U/l) LDH (cutoff 80 U/l) | 83 50 | 87 77 | 60 38 | 96 92 | Positive CSF cytology or LM at autopsy less than 1 month after assayed CSF | I: BC pts with brain metastases (18). II: BC pts without CNS metastases (34) |
| Twijnstra et al26 | LDH with ACA method | 15 out of 15 | β2-microglobulin (cutoff >2.2 mg/l) LDH (cutoff >26 U/l) |
60 60 | 88 93 | NA NA | NA NA | Positive CSF cytology (13) or LM at autopsy (2) | Neurological disorders without CNS metastases (16) |
| Twijnstra et al27 | LDH with ACA method | 15 out of 34 | LDH (cutoff >26 U/l) | 79 | 83 | 33 | 83 | Positive CSF cytology (32) or LM at autopsy (2) | I: Pts with solid tumors (66), hematological tumors (10), benign primary CNS tumors (7), malignant primary CNS tumors (10), others (112) II: Controls (110) |
| CA15.3, CEA | |||||||||
| Le Rhun et al28 | CSF CA15.3 Automatized immuno-enzymatic technology | 20 out of 20 | CA 15.3 (cutoff 3 UI/mL) |
80 | 70 | NA | NA | Positive CSF cytology or clinical signs in combination with positive MRI | I: LM from other cancers than BC (20) (group 2). II: BC pts with brain metastases only (20) (group 3). III: No malignancy (20) (group 4) |
| Yap et al29 | Triple-isotope double antibody method | 23 out of 23 | CEA (cutoff 1.5 ng/mL) |
70 | 100 | NA | NA | Positive CSF cytology and absence of parenchymal brain metastases on CT | Stage IV BC pts and no evidence of metastatic CNS disease (10) |
| Corsini et al20 | CEA, CA15.3, CA125, CA19.9 (Modular Analytics SWA) | 11 out of 18 | Intrathecal synthesis of CEA/CA 15.3/CA125/ CA19.9 | 100 | 100 | NA | NA | Positive CSF cytology | Nonmalignant neurological diseases: inflammatory (33), neurodegenerative (17) |
| Twijnstra et al26 | CEA | 15 out of 15 | CEA (cutoff >4 ng/mL) | 60 | 93 | - | - | Positive CSF cytology or LM at autopsy | Neurological disorders without CNS metastases (16) |
| Proteomics | |||||||||
| Dekker et al30 | MALDI-TOF MS | 54 out of 54 | 79 | 76 | NA | NA | Positive CSF cytology or a compatible neurological syndrome and diagnostic MRI | I: BC pts with negative cytology and follow- up incompatible with LM (54). II: No cancer and no neurological disease | |
| EpCAM-based assays | |||||||||
| Nayak et al31 | CellSearch | 7 out of 15 | Cutoff: presence of CSFTC | 100 | 97.2 | 93.8 | 100 | Positive CSF cytology or positive MRI findings observed within 1 month of the initial evaluation | I: Pts with clinical suspicion of LM (36). II: Nonmalignant neurologic disease (9) |
| Lee et al32 | CellSearch | 18 out of 18 | Cutoff >1.9 cells/mL | 81 | 85 | NA | NA | Positive CSF cytology | I: Non-malignant neurologic disease (14).II: BC pts with negative cytology (20) |
| Lin et al33 | CellSearch | 14 out of 30 | Cutoff ≥1 cell/mL | 93 | 95 | 90 | 97 | Positive CSF cytology or unequivocal MRI findings performed within a month of the CSF analysis. Unequivocal MRI findings defined as: LM enhancement associated with subarchnoid nodules, basal cistern enhancement, or nerve root enhancement/clumping. | Cancer pts with clinical suspicion of LM without LM confirmation (65) |
| Kerklaan et al34 | EpCAM FCI | 7 out of 13 | Cutoff ≥2 CSFTC/ 5 mL CSF | 100 | 100 | NA | NA | Positive CSF cytology or a MRI with positive MRI or progressive neurological symptoms compatible with LM and exclusion of other causes | Cancer pts with clinical suspicion of LM without LM confirmation (16) |
| Subirá et al35 | EpCAM FCI | 32 out of 49 | Cutoff 10 clustered events | 75.5 | 96.1 | 97.4 | 67.6 | Positive CSF cytology or the combination of clinical signs and either MRI or biochemical CSF findings | Cancer pts with clinical suspicion of LM in whom LM was excluded (26) |
| Subirá et al36 | EpCAM FCI | 39 out of 94 | Cutoff 16 clustered events | 79.8 | 84 | 90.36 | 68.85 | Positive CSF cytology and/or compatible clinical signs plus MRI findings with biochemical CSF abnormalities | Cancer pts with clinical suspicion of LM in whom LM was excluded (50) |
| ctDNA | |||||||||
| Bougel et al37 | Methylation of hTERT with MS-HRM | 8 out of 9 | Methylation hTERT | 92 | 100 | 100 | 95 | Positive CSF cytology | No malignancy (21) |
Abbreviations: MS-HRM = methylation-sensitive high resolution melting; NA = not available; pts = patients.
aSensitivity and specificity to discriminate LM from LM negatives.
bSensitivity and specificity to discriminate LM from control patients.