In Reply.—We appreciate the comments concerning our study1 comparing the immunohistochemical (IHC) profiles of giant cell glioblastomas (GCGBMs) and pleomorphic xanthoastrocytomas (PXAs), and we agree with virtually all of them. They focus on the interpretation of results from 1 IHC marker used in our study, class III β-tubulin (βIII).

Studies by Katsetos and colleagues show the utility of βIII as an IHC marker for nonneoplastic and neoplastic cells of neuronal or neuroendocrine lineage (reviewed in Katsetos et al2). However, βIII subsequently proved to be less specific for neuronal differentiation when applied to high-grade gliomas.3,4 In a separate study, we also observed widespread βIII immunoreactivity in high-grade gliomas, including many conventional types of glioblastoma.5 We agree that immunoreactivity for βIII (when staining is used in isolation and not in conjunction with other neuronal markers, such as synaptophysin or neuronal nuclear antigen) is not indicative of neuronal differentiation in high-grade glial neoplasms. What had not been previously reported, to our knowledge, is whether βIII would also be found in the GCGBM, a rare subtype of glioblastoma.

Our recently published study in the Archives was not a specific investigation of βIII expression, but rather was a comparison of IHC profiles (using a panel of antibodies that recognize p53, glial, and neuronal antigens) between 2 rare types of glial tumors, the GCGBM and PXA. As we noted in the article, these 2 tumors have sufficiently overlapping clinicopathologic features, including occurrence in younger patients, gross circumscription, reticulin deposition, lymphocytic infiltrates, and tumor giant cells, to create a potential diagnostic conundrum. To address this possible confusion, we studied reasonable numbers of each tumor (9 each), obtained from the combined files of several large institutions, and found that GCGBM can be reliably distinguished from PXA based on results of an IHC panel rather than a single antibody in isolation. Indeed, we specifically cautioned that the 2 tumors share reactivity for βIII.

Applying this IHC panel has practical implications, since these 2 tumor types have differing prognoses and treatments (PXA, World Health Organization [WHO] grade II, vs GCGBM, WHO grade IV). While we did not study PXA with anaplastic features (WHO grade III), the diagnostic criteria for this tumor have not been well established, and some neuropathologists do not even recognize “anaplastic PXA” as a specific entity. Also, distinguishing between a grade III and grade IV tumor would have less significant therapeutic and prognostic implications for the individual patient than correctly categorizing a tumor as a grade II glioma (PXA) versus a grade IV glioma (GCGBM).

We hope our study will assist practicing pathologists, who see relatively few examples of these rare tumors, in making the correct diagnosis. At the very least, the difference in IHC profile may further dispel any notion that these 2 tumor types are the same biological entity, however many features they have in common.

Martinez-Diaz
,
H.
,
B. K.
Kleinschmidt-DeMasters
,
S. Z.
Powell
, and
A. T.
Yachnis
.
Giant cell glioblastoma and pleomorphic xanthoastrocytoma show different immunohistochemical profiles for neuronal antigens and p53 but share reactivity for class III β-tubulin.
Arch Pathol Lab Med
2003
.
127
:
1187
1191
.
Katsetos
,
C. D.
,
M. M.
Herman
, and
S. J.
Mork
.
Class III β-tubulin in human development and cancer.
Cell Motil Cytoskeleton
2003
.
55
:
77
96
.
Katsetos
,
C. D.
,
L.
Del Valle
, and
J. F.
Geddes
.
et al
.
Aberrant localization of the neuronal class III β-tubulin in astrocytomas: a marker for anaplastic potential.
Arch Pathol Lab Med
2001
.
125
:
613
624
.
Katsetos
,
C. D.
,
L.
Dell Valle
, and
J. F.
Geddes
.
et al
.
Localization of the neuronal class III β-tubulin in oligodendrogliomas: comparison with Ki67 proliferative index and 1p/19q status.
J Neuropathol Exp Neurol
2002
.
61
:
307
332
.
Yachnis
,
A. T.
Class III β tubulin immunoreactivity in gliomas.
J Neuropathol Exp Neurol
2001
.
60
:
517
.