Diagnostic Utility of GATA3 and ISL1 in Differentiating Neuroblastoma From Other Pediatric Malignant Small Round Blue Cell Tumors

Abstract

Objectives.

Accurate diagnosis of neuroblastoma may be challenging, especially with limited or inadequate specimen and at the metastatic sites due to overlapping imaging, histopathologic, and immunohistochemical (immunohistochemistry [IHC]; infidelity among various lineage-associated transcription factors eg FLI1, transducin-like enhancer 1, etc) features. GATA3 and ISL1 have recently been described as markers of neuroblastic differentiation. This study aims at determining the diagnostic utility of GATA3 and ISL1 in differentiating neuroblastoma from other pediatric malignant small round blue cell tumors.

Methods.

We evaluated GATA3 and ISL1 expression in 74 pediatric small round blue cell tumors that included 23 NMYC-amplified neuroblastomas, 11 EWSR1-rearranged round cell sarcomas, 7 SYT::SSX1-rearranged synovial sarcomas, 5 embryonal rhabdomyosarcomas, 10 Wilms tumors (nephroblastomas), 7 lymphoblastic lymphoma, 7 medulloblastoma, and 4 desmoplastic small round cell tumor.

Results.

All 23 neuroblastomas (moderate to strong staining in >50% of the tumor cells), 5 T-lymphoblastic lymphomas (moderate to strong staining in 40%-90% of the tumor cells), and 2 desmoplastic small round cell tumors (weak to moderate staining in 20%-30% of the tumor cells) expressed GATA3, while other tumors were negative. ISL1 immunoreactivity was observed in 22 (96%) neuroblastomas (strong staining in in >50% of the tumor cells, n  =  17; moderate to strong staining in 26%-50% of the tumor cells, n  =  5), 3 embryonal rhabdomyosarcoma (moderate to strong staining in 30%-85% of the tumor cells), 1 synovial sarcoma (weak staining in 20% of the tumor cells), and 7 medulloblastoma (strong staining in 60%-90% of the tumor cells). Other tumors were negative. Overall, GATA3 showed 86% specificity, 100% sensitivity, and 90% accuracy for neuroblastoma, with a positive predictive value (PPV) and negative predictive value (NPV) of 77% and 100%, respectively. ISLI showed 72% specificity, 96% sensitivity, and 81% accuracy for neuroblastoma, with a PPV and NPV of 67% and 97%, respectively. After the exclusion of T-lymphoblastic lymphoma and desmoplastic small round cell tumors, GATA3 had 100% specificity, sensitivity, accuracy, and PPV and NPV for neuroblastoma. Similarly, in pediatric small round blue cell tumors, ISL1 had 100% specificity, sensitivity, accuracy, PPV, and NPV for neuroblastoma, after embryonal rhabdomyosarcoma, synovial sarcoma, and medulloblastoma were excluded.

Conclusions

GATA3 and ISL1 may be valuable in the diagnostic work-up of neuroblastoma and may reliably be used to support the neuroblastic lineage of pediatric small round blue cell tumors. Furthermore, dual positivity helps in challenging scenarios, when there is equivocal imaging, overlapping IHC features, limited specimen, and the lack of facility for a molecular work up.