Brain Pathology Case of the Month - May 2007



Pleomorphic xanthoastrocytoma (PXA) is an astrocytic neoplasm with a relatively favorable prognosis, typically encountered in children and young adults, with superficial location in the cerebral hemispheres. Characteristic histologic features include (1) fascicular arrangement of the neoplastic astrocytes; (2) pleomorphic astrocytes with bizarre cytologic features; (3) lipidization of the neoplastic astrocytes; (4) perivascular patchy lymphocytic infiltrates; (5) dense reticulin staining around single or grouped tumor cells; and (6) eosinophilic granular bodies. The neoplastic astrocytes in the PXA can show neuronal differentiation. Mitotic index and Ki-67 are important prognostic predictors. World Health Organization (W.H.O.) recommends the designation "PXA with anaplastic features" to denote PXA featuring high mitotic activity with or without accompanying necrosis. The designations "anaplastic PXA" and "PXA grade III or IV" are currently not recommended by W.H.O.

PXA is uncommon, accounting for less than 1% of all astrocytic neoplasms. Two thirds of the PXA reported cases are less than 18 years old. (1) PXA is typically supratentorial, temporal lobe is the classic site with superficial location in the cerebral hemispheres and attachment to the leptomeninges. In many instances, PXAs show slow evolution with long symptom free periods and remain subclinical until adulthood. Radiologically, a PXA classically appears as a cystic mass with a mural nodule usually in the cortical or meningeal side of the cyst wall. The most consistent gross features of PXA are discreteness, and non-infiltrating pattern. Abundance of the lipidized cells within the tumor can sometimes impart a yellow hue. (1, 2)

Microscopically, PXAs have fascicular pattern that often appears as interweaving cell bundles. The storiform pattern of PXA can be so conspicuous to mimic the pattern of a fibrohistiocytic neoplasm, as we have in the current case. The tumor cells usually show marked pleomorphism, which refers to irregular spindle cells, intermingled with mono- or multinucleated giant cells with great variability in size and shape. Many tumor cells exhibit xanthomatous changes, which is caused by intracellular lipid accumulation. (1, 2) The presence of eosinophilic granular bodies (EGB) is an important diagnostic feature. EGB represent lysosomes distended by autophagic debris. Periodic Acid Schiff (PAS) stain facilitates the detection of these structures if they are hard to find. It is important to mention that EGB are not specific to PXA as they can be shared by other tumors, notably pilocytic astrocytoma and ganglioglioma. PXA has striking density of reticulin fibers, which tend to invest individual cells or grouped tumor cells as evidenced by reticulin staining. This is referred to the basement membranes of the tumor cells, which can be recognized ultrastructurally as pericellular basal laminae. (1, 2) This suggests that PXA originates from subpial astrocytes as they are invested by basal laminae in contrast to the astrocytes in other sites. (1) Patchy lymphocytic infiltrates are often seen in PXA and they are typically present perivascularly. Mitotic figures are usually rare or absent. Ki-67 index in PXA is generally very low (0-1%). Immunohistochemically, the astrocytic quality of PXA is confirmed by glial fibrillary acidic protein (GFAP) reactivity. Tumor cells of PXA can exhibit occasional positivity to neuronal markers like neurofilament protein (NFP), as we have in the current case.

This divergent neuronal differentiation may suggest a developmental relationship between PXAs and other neuronal tumors like ganglioglioma and may indicate that these tumors arise from multipotential neuroectodermal precursor cells common to both neurons and astrocytes. (3, 4)

The preoperative differential diagnosis of this case was pilocytic astrocytoma but the characteristics of this tumor didn't fit with this diagnosis as pilocytic astrocytoma has biphasic pattern: microcystic areas alternating with compact areas which are rich in Rosenthal fibers and EGBs. However, they are generally less cellular and pleomorphic than PXAs, and the xanthomatous changes and the abundant reticulin deposition are usually lacking. In addition, Rosenthal fibers are not component of PXA.

Typically, classic PXAs are low grade astrocytomas as they are less aggressive than cytologic features would suggest. Several reported studies showed that mitotic index, necrosis, and Ki-67/MIB-1 index are important outcome predictors. (5, 6)

PXA corresponds to World Health Organization (W.H.O.) grade II. The PXA of the current case showed behavior of grade higher than that of classic PXAs :(1) the overall mitotic index in the current case was high (5 mitoses per ten high power fields while normally it is hard to find mitotic figures), and (2) Ki-67 index was high (Ki-67 overall reactivity was 3%, while normally it is 0-1%). Therefore, we described the PXA in hand as "PXA with anaplastic features". This is because the designation of "PXA with anaplastic features" to denote PXA featuring high mitotic activity ( ≥ 5 mitoses per 10 HPF) is recommended by W.H.O. while the use of term "anaplastic PXA" is currently not preferred. Moreover, upgrading PXA from W.H.O. grade II to W.H.O. grade III or VI in such cases is also objectionable as the criteria used in describing or grading diffuse astrocytomas showing anaplastic features should not be applied to PXA. This is because the use of term "anaplastic" or upgrading to grade III or VI might be found "unfair" when applied to PXA which behaves less aggressively than ordinary anaplastic astrocytoma. In other words, PXA should be distinct from diffuse astrocytoma even if they share the same anaplastic features since the malignant histology and the anaplastic features in PXA do not correlate with prognosis as reliably as in ordinary astrocytoma. Several studies reported that the clinical course of patients with histologically malignant PXAs is often more favorable than that of patients with diffuse astrocytomas showing the same features. Interestingly, the molecular genetic studies showed that the genetic events in PXA formation and malignant progression differ from those involved in diffuse astrocytoma tumorigenesis as PXA usually shows low incidence of TP53 gene mutation which is very common in diffuse astrocytomas. This provides further support for maintaining a separation between PXA and diffuse astrocytoma if both share the same anaplastic features. (6, 7) Based on these findings and recommendations, we used the term "PXA with anaplastic features" in reporting the diagnosis of the current case to indicate correlation with potentially more aggressive clinical behavior and we chose not to use the term "anaplastic PXA" as this designation may provoke inappropriately aggressive treatment. On the other hand, Ki-67 index appears to be predictive of recurrence in PXA. It is worth to mention that in one study, the PXA cases where Ki-67 was higher than 2.2% showed higher incidence of recurrence with anaplastic progression in the recurrent tumors. Therefore, the high Ki-67 index in the current case (3%) not only may indicate more aggressive behavior but also higher incidence of recurrence. (8)

As a discrete and circumscribed tumor, surgical excision is the principal treatment of PXA. (2) The extent of resection is found to be one of the outcome predictors in some studies. (9) Complete resection of the tumor in our case was successfully done.


  1. P. Kleihues, W.K. Cavenee. WHO Classification of Tumours: Pathology and Genetics of Tumors of Nervous System. World Health Organization Blue Book Series, WHO, 2000: 52-54.
  2. Peter C. Burger, Bernd W. Scheithauer, F. Stephen Vogel. Surgical Pathology of the Nervous System and Its Coverings, Churchill Livingstone; 4th edition, 2002: 215:220.
  3. Lach B, Duggal N, DaSilva VF, Benoit BG. Association of pleomorphic xanthoastrocytoma with cortical dysplasia and neuronal tumors. A report of three cases. Cancer. 1996 Dec 15;78(12):2551-2563.
  4. Powell SZ, Yachnis AT, Rorke LB, Rojiani AM, Eskin TA.Divergent differentiation in pleomorphic xanthoastrocytoma. Evidence for a neuronal element and possible relationship to ganglion cell tumors. Am J Surg Pathol. 1996 Jan;20(1):80-85.
  5. Sugita Y, Shigemori M, Okamoto K et al. Clinicopathological study of pleomorphic xanthoastrocytoma: correlation between histological features and prognosis. Pathol Int. 2000 Sep; 50 (9): 703-708.
  6. Giannini C, Scheithauer BW, and Burger PC: Pleomorphic xanthoastrocytoma: what do we really know about it? Cancer. 1999 May 1; 85(9): 2033-2045.
  7. Macaulay RJ, Jay V, Hoffman HJ, et al: Increased mitotic activity as a negative prognostic indicator in pleomorphic xanthoastrocytoma. Case report. J Neurosurg 79:761 768, 1993
  8. Prliferative activity and p53 mutation as prognostic indicators in pleomorphic xanthoastrocytma. A clinicopathologic study of six cases. E.L. Munoz, D.A. Eberhard, M.B.S. Lopes, B.F. Schenider, F. Gonzalez-Fernandez, and S.R. Vandenberg, Journal of Neuropathology & Experiemental Neurology Vol. 55 No. 5 May 1996: 606
  9. Pahapill PA, Ramsay DA, Del Maestro RF: Pleomorphic xanthoastrocytoma: case report and analysis of the literature concerning the efficacy of resection and the significance of necrosis. Neurosurgery 38:822 829, 1996

Contributed by Osama M. Al-Agha, MD

International Society of Neuropathology