Immunohistochemical ATRX expression is not a surrogate for 1p19q codeletion.

The IDH-mutant and 1p/19q co-deletion (1p19q codel) provides significant diagnostic and prognostic value in lower-grade gliomas. As ATRX mutation and 1p19q codel are mutually exclusive, ATRX immunohistochemistry (IHC) may substitute for 1p19q codel, but this has not been comprehensively examined. In the current study, we performed ATRX-IHC in 78 gliomas whose ATRX statuses were comprehensively determined by whole exome sequencing. Among the 60 IHC-positive and 18 IHC-negative cases, 86.7 and 77.8% were ATRX-wildtype and ATRX-mutant, respectively. ATRX mutational patterns were not consistent with ATRX-IHC. If our cohort had only used IDH status and IHC-based ATRX expression for diagnosis, 78 tumors would have been subtyped as 48 oligodendroglial tumors, 16 IDH-mutant astrocytic tumors, and 14 IDH-wildtype astrocytic tumors. However, when the 1p19q codel test was performed following ATRX-IHC, 8 of 48 ATRX-IHC-positive tumors were classified as "1p19q non-codel" and 3 of 16 ATRX-IHC-negative tumors were classified as "1p19q codel"; a total of 11 tumors (14%) were incorrectly classified. In summary, we observed dissociation between ATRX-IHC and actual 1p19q codel in 11 of 64 IDH-mutant LGGs. In describing the complex IHC expression of ATRX somatic mutations, our results indicate the need for caution when using ATRX-IHC as a surrogate of 1p19q status.