In:
American Journal of Surgical Pathology, Ovid Technologies (Wolters Kluwer Health), Vol. 43, No. 1 ( 2019-01), p. 56-63
Abstract:
Ependymomas (EPs) are tumors of the brain and spinal cord constituting ∼10% of the childhood central nervous system neoplasms and about 30% in children aged 〈 3 years. Their anatomic distribution varies according to the age, with those arising in the supratentorial (ST) compartment, spinal cord being more common in older children and adults, and those at the infratentorial location are more common and occurring more frequently in infants and children. Recently, molecular classification of EP subgroups has been proposed and a supratentorial ependymoma subgroup characterized by RELA-fusion genes (ST-EP-RELA) has been established. It would be useful to define a standardized, robust method for the diagnosis of these relevant fusion genes. We used real-time polymerase chain reaction, conventional real-time polymerase chain reaction, and Sanger sequencing to characterize RELA fusion status in formalin-fixed paraffin-embedded samples from 42 ST-EPs (12 adults and 30 pediatric). We tested p65/RELA and L1CAM protein immunohistochemistry for their ability to predict RELA-fusion status. We reviewed clinical data to assess significant associations in this anatomic subgroup. Of the 42 patients, we identified RELA-fusion genes in 17 cases. L1CAM immunostaining displayed 94% sensitivity, 76% specificity, 73% positive predictive value (PPV), 95% negative predictive value (NPV). The p65/RELA immunostaining displayed 100% sensitivity, 92% specificity, 89.5% PPV, 100% NPV. Concordant double immunostaining improves PPV to 92.5% and maintains 100% NPV. Immunohistochemistry using both p65/RELA and L1CAM antibodies is valuable for ST-EP-RELA diagnosis: the negativity with both antibodies consistently predicts the absence of RELA fusions, whereas verification of fusion transcripts by molecular analyses is warranted only in single-positive or double-positive staining cases.
Type of Medium:
Online Resource
ISSN:
0147-5185
DOI:
10.1097/PAS.0000000000000979
Language:
English
Publisher:
Ovid Technologies (Wolters Kluwer Health)
Publication Date:
2019
detail.hit.zdb_id:
2029143-7
