Secondary Logo

Journal Logo

Case Reviews

Pitfalls and Traps in Neuropathology

Smith, Michael Timothy MD; Bruner, Evelyn T. MD

Author Information
doi: 10.1097/PCR.0000000000000365
  • Free

Abstract

The subspecialty of neuropathology, like all subspecialty areas, is rich with possibilities for traps, pitfalls, and errors. The reasons are many to include the low numeric case load of neuropathology specimens in most hospitals and therefore producing pathologist nonfamiliarity. Even in high neuropathology workload environments, the experienced pathologist must be wary. Low-grade gliomas and gliosis still resemble one another at the edges of the tumor and brain interface. Inflammatory processes also can resemble glioma if only by the cellularity. Origin of biopsies may not be revealed by the neurosurgeon and can make a huge difference. Many nonneoplastic conditions that increase cellularity still masquerade as glioma. Such traps and pitfalls were plentiful enough in the hematoxylin and eosin (H&E) era and have not decreased in the molecular era. Molecular techniques have provided specific testing queries and classification changes that clarify some glioma diagnostics and can provide some therapeutic direction. The variation in overlapping pathologies is great, and even more specialized testing will likely be forthcoming. The intraoperative consultation introduces additional variables through the fog of frozen section that must also be considered. The following cases illustrate only some of the myriad of pitfalls, traps, and swindles lurking in neuropathology.

CASE 1

This 79-year-old man came to the emergency room after experiencing progressive right hemiparesis. Imaging showed a left-sided mass lesion with midline shift. Imaging favored glioma, and a biopsy was performed on the lesion. The initial H&E interpretation of that biopsy was erroneously high-grade glioma. Figure 1A shows the increased cellularity of that biopsy, whereas Figure 1B shows another area with large pink gemistocytes. Importantly, the gemistocytes are evenly spaced as expected in a reactive process. Neoplastic astrocytes, on the other hand, are characteristically unevenly spaced, and the nuclei may inappropriately touch one another. The increased cellularity in this case is due to macrophages and lymphocytes identifying this process as nonneoplastic. The macrophages engulf the disintegrating neuropil and fall apart in a mosaic pattern characteristic of infarcts. Figure 1C shows the granulation tissue–like vascular reaction sometimes seen around the edges infarcts.

FIGURE 1
FIGURE 1:
A, This low-power image of the lesion shows hypercellularity that could be misinterpreted as glioma (H&E stain, original magnification ×100). B, Higher magnification shows a variety of cell types including large reactive astrocytes with pink cytoplasm. The majority of nuclei belong to macrophages that have partially replaced the neuropil (the dense matrix of glial fibers, nerve fibers, their branches and synapses). Ischemia has dissolved the neuropil via liquefactive necrosis, and the macrophages have engulfed the debris. The cracks in the section are features of disintegration. Note that nearly all the nuclei are equidistant from one another. Glioma nuclei reliably touch adjacent nuclei, a phenomenon useful for identifying gliomas. The predominance of inflammatory cells militates against a glioma (H&E stain, original magnification ×250). C, Along the viable edges of the infarct, there is capillary proliferation with many macrophages emerging from the capillaries (H&E stain, original magnification ×200). These capillary changes are not the complex thick-walled microvascularity seen in high-grade gliomas.

This case shows the infrequent problem of increased cellularity of an infarct that along with confusing imaging can lead to a misdiagnosis.1 Recognition of the evenly spaced gemistocytes and many macrophages (CD68 positive) is key in recognizing an infarct. The sheets of macrophages somehow masquerade as native, cellular neuropil. The vascular reaction and necrotic tissue in Figure 1C are confirmatory. The vascular changes are nothing like the vascular complexity of glioblastoma. Multiple sclerosis may be in the differential, but large areas of necrosis are not a feature. If perivascular neutrophils are found, then fungi should be sought.

CASE 2

A 35-year-old African American man developed visual symptoms, and imaging showed a 3-cm suprasellar mass that produced pressure on the optic nerves, tracts, and chiasm. The imaging differential included midline germ cell tumor, midline glioma, pituitary neoplasms, unexpected metastasis, and others. The biopsied lesion is shown in Figure 2A, which is obviously granulomatous, raising the differential of fungi, mycobacteria, and sarcoid. Fungal and mycobacterial stains were negative. Cultures were eventually all negative. Because of disease progression and lack of diagnosis, additional tissue was acquired in a second biopsy shown in Figure 2B. The additional tissue displayed the obvious diagnostic features of large germ cells in a sea of lymphocytes. The patient was then appropriately treated.

FIGURE 2
FIGURE 2:
A, The first biopsy shows a typical granuloma with a central multinucleate histiocyte surrounded by lymphocytes, macrophages, fibroblasts, and plasma cells (H&E stain, original magnification ×250). B, The second biopsy possessed the same inflammatory infiltrate but also displayed large malignant epithelioid germ cells of germinoma. The large epithelioid cells marked with germ cell markers (H&E stain, original magnification ×300).

Granulomatous response to seminoma/germinoma has been reported in testis, mediastinum, central nervous system, and elsewhere.2–5 The pitfall occurs when the biopsy contains only the granulomas and not the diagnostic seminoma. So the neoplastic process has produced a nonneoplastic reaction that may totally obscure the malignancy as in the first biopsy, and the pathologist is swindled. Of course, the tumor stains with the usual germ cell markers (OCT3/4, SALL4, CD117, etc), but those are seldom employed when the pathologist is confronted with granulomas. This is a pitfall that can snare the unwary pathologist.

CASE 3

A 20-year-old woman had a posterior fossa lesion that involved the central portion of the cerebellum by imaging. The patient had problems with balance and gait and visual field symptoms. Imaging reasonably suggested medulloblastoma, but a suprasellar component was unusual. A microscopic field from the surgical specimen is shown in Figure 3. The cellularity is increased beyond normal brain, so high-grade glioma was briefly considered until the glial fibrillary acidic protein was negative, as were CD45, synaptophysin, CD34, keratins, progesterone receptor, and others. This tumor was considered to be in cerebellar parenchyma until note was made of the thick-walled collagenized blood vessels that do not occur in brain or in primary brain tumors (unless they have been radiated). Cerebral vessels are not naturally collagenized. Then, the possibility of meningeal origin was considered, but the usual nonspecific stains used for meningioma were negative. Soft tissue–like tumors were finally considered and with a positive STAT6, the diagnosis of solitary fibrous tumor (SFT) was confirmed. Negative CD34 is expected in higher-grade SFTs.

FIGURE 3
FIGURE 3:
The biopsy showed a cellular tumor with a syncytial pattern mimicking a glioma. No distinct architectural pattern is present. There is adequate hyperchromasia and pleomorphism for a malignant glioma, but the collagenized blood vessel identifies the tumor is nonglial (H&E stain, original magnification ×250). Glial fibrillary acidic protein stain was negative. The inset shows positive STAT6 identifying the tumor as an SFT/hemangiopericytoma (STAT6, original magnification ×350).

Historically, the terms solitary fibrous tumor and hemangiopericytoma are tenaciously intertwined both in surgical pathology and in neuropathology. Originally, the lesion in the chest was considered a form of mesothelioma, while the tumor in the cranial vault was called angioblastic meningioma.6,7 That fascinating and outdated ancient history will not be covered here. All of those historical terms are now unified under solitary fibrous tumor, but the morphological patterns are not as unified, showing a spectrum extending from the classic staghorn vesseled hemangiopericytoma to the ovoid/spindle cell patternless-pattern SFT.8 These should but may not have a recognizable collagenized vascular architecture or patternless pattern. The aggressive malignant forms are even more variable in morphology. Mitotic rate greater than 4/10 high-power field as published by Enzinger and Smith9 identifies the malignant tumors. STAT6 is a good marker to identify these tumors, regardless of location, and in one study marked 86% of tumors.10 Other positive stains include CD34, CD99, BCL2, and β-catenin, but staining in the higher-grade tumors is usually unreliable.11,12 In a tiny biopsy, there can be an easy morphological overlap with fibromatosis. The best key to recognition is awareness that it can be a meningeal neoplasm and that it may not stain with CD34.13

CASE 4

A 48-year-old man presented with symptoms of stroke, and imaging showed a large posterior frontal mass with surrounding edema. Peripheral enhancement and slight midline shift were present. Initial biopsy interpretation was oligodendroglioma (Fig. 4). This case illustrates the trap created by neoplastic astrocytes developing perinuclear halos that mimic the classic fried egg appearance of oligodendroglioma. Despite the halos, the chicken-wire vascular pattern is lacking. The shape of the nuclei is mostly oval, not round as for oligodendroglia. An occasional mitosis is allowed for oligodendrogliomas, but this tumor was not an oligodendroglioma but an astrocytoma. In modern diagnostic neuropathology, the diagnosis of oligodendroglioma requires discovery of the 1p 19q codeletion.8(pp60–69) This case lacked that molecular marker, therefore showing its astrocytic lineage, and the diagnosis of astrocytoma was made. The molecular identification is now a key part of making the diagnosis. This is a trap created by knowledge and technology.

FIGURE 4
FIGURE 4:
The image shows a hypercellular tumor with some oligodendroglial features. There are round tumor nuclei with perinuclear halos, but the background of a capillary network is lacking. The case was initially considered an oligodendroglioma, but 1p19q was not codeleted negating the possibility of oligodendroglioma (H&E stain, original magnification ×300).

CASE 5

A 20-year-old woman had a many-year history of visual symptoms, and a lateral sellar mass was discovered. Imaging showed a partially calcified mass, and the tumor was biopsied. Figure 5A shows a field from small needle biopsy. The area shown is composed of elongate fibrillary astrocytes and is slightly hypercellular and punctuated by prominent eosinophilic granular bodies and Rosenthal fibers. The diagnosis of pilocytic astrocytoma was made, but a second biopsy was performed for unclear reasons. Figure 5B shows a portion of that second biopsy. Nests of palisaded epithelial cells with reverse polarization were present in the astrocytic stroma in the pattern of craniopharyngioma. Craniopharyngiomas typically have a prominent surrounding piloid gliosis that is easily mistaken for pilocytic astrocytoma.14 The imaging finding of calcifications could have been a clue, but pilocytic astrocytomas may also be calcified. The pitfall in this case was the small size of the initial biopsy. Small biopsies are threats to pathologists everywhere.

FIGURE 5
FIGURE 5:
A, The first supratentorial biopsy on the left, shows mildy hyperccellular neuropil with abundant Rosenthal fibers. The nuclear features and the overall appearance are a dead ringer for a pilocytic astrocytoma (H&E stain, original magnification ×250). B, Shows an adjacent area with epithelial nests of a craniopharyngioma (H&E stain, original magnification ×300).

DISCUSSION

The foregoing examples are only some of the pitfalls that exist in neuropathology/surgical pathology. All pathologists have experienced some pitfalls, and there is likely an endless supply caused by tiny biopsies, new knowledge, new classifications, and molecular techniques. In case 1, the problem presented to the pathologist is too many nuclei. If the hypercellularity is not correctly identified as secondary to macrophages, the infarct can be called a high-grade malignant neoplasm. The error is made easier by confusing imaging. In case 2, the deceptive tumor itself obfuscated the appearance with granulomas. Here, the pathologist must have knowledge of the possibility to avoid the trap. Case 3 also requires pathologist knowledge of possible occurrence of SFT in the meninges. Additionally, the pathologist must be aware of the possible morphological variations. Similarly in case 4, the pathologist must be aware of the change in diagnostic classification requirements. Many areas of surgical pathology and neuropathology are undergoing classification changes reflected by the publication of new World Health Organization books. Case 5 is another example of a tumor that creates surrounding tissue features that resemble a different process. Again the pathologist must be armed with knowledge of these possibilities.

Another and frequent opportunity for pitfalls is the frozen-section event. Intraoperatively, all pathologists wish to provide the most exact diagnosis possible. The neurosurgeon will push us to provide the best most exact diagnosis possible. Despite this pressure, pathologists must consider realistic capabilities. The most common questions asked by the neurosurgeon will be in the situation of suspected tumor. Studies performed by very capable neuropathologists have shown that glioma (without specific grading) can be accurately diagnosed as low-grade glioma or high-grade glioma.15 The pathologist who tries to assign a specific subtype of glioma or grade intraoperatively will often have the diagnosis change on permanent sections. Indeed, this is partially illustrated in case 4. Some knowledge of clinical expectations and history will also greatly aid the pathologist. If the biopsy is desired for inflammatory disease, myelin disease, or infectious agents, the literature clearly shows that a frozen section is not the route to follow.16 In those instances, the best path is to quote the literature to the neurosurgeon, defer the diagnosis, and solve the problem with permanent sections and staining.

CONCLUSION

Pitfalls and traps in pathology and neuropathology can lead to significant errors. The pitfalls are large in number and create a treacherous minefield that must be navigated. The pitfall problem has not been lessened by reclassifications and molecular studies. The new molecular studies have instead resulted in newer potential traps for the pathologist. Reclassification resulting from molecular parameters establish new criteria and therefore new pitfalls of which the pathologist must be aware. This review has covered some of the more common neuropathology traps but there are many others to learn from.

REFERENCES

1. Allen CM. Differential diagnosis of acute stroke: a review. J R Soc Med 1984;77:878–881.
2. Lee H, Lim JK, Lee SY, et al. Granulomatous reaction of primary mediastinal seminoma leading to diagnostic delay: a case report. J Thorac Dis 2018;10(2):E98–E102.
3. Ye H, Ulbright TM. Difficult differential diagnoses in testicular pathology. Arch Pathol Lab Med 2012;136(4):435–446.
4. Gupta D, Rath A, Rathi KR, et al. Primary thymic mediastinal seminoma with florid granulomatous reaction. Indian J Pathol Microbiol 2016;59:351–354.
5. Massey J, Walker S, Galloway M, et al. Granulomatous CNS inflammation associated with seminoma. J Neurol 2019;266:1389.
6. Folpe AL. “Hey! Whatever happened to hemangiopericytoma and fibrosarcoma?” An update on selected conceptual advances in soft tissue pathology which have occurred over the past 50 years [published online October 24, 2019]. Hum Pathol 2019; pii: S0046-8177(19)30165-0. doi: 10.1016/j.humpath.2019.10.001.
7. Klemperer P, Rabin C. Primary neoplasms of the pleura: a report of five cases. Arch Pathol 1931;11:385–412.
8. Louis DN, Ohgaki H, Wiestler OD, et al. WHO Classification of Tumours of the Central Nervous System. Revised 4th ed. Lyon: IARC; 2016:249–254.
9. Enzinger FM, Smith BH. Hemangiopericytoma. An analysis of 106 cases. Hum Pathol 1976;7:61–82.
10. Demicco EG, Harms PW, Patel RM, et al. Extensive survey of STAT6 expression in a large series of mesenchymal tumors. Am J Clin Pathol 2015;143(5):672–682.
11. Mosquera JM, Fletcher CD. Expanding the spectrum of malignant progression in solitary fibrous tumors: a study of 8 cases with a discrete anaplastic component—is this dedifferentiated SFT? Am J Surg Pathol 2009;33:1314–1321.
12. Geramizadeh B, Marzban M, Churg A. Role of immunohistochemistry in the diagnosis of solitary fibrous tumor, a review. Iran J Pathol 2016;11:195–203.
13. Chaubal A, Paetau A, Zoltick P, et al. CD34 immunoreactivity in nervous system tumors. Acta Neuropathol 1994;88:454–458.
14. Kasai H, Hirano A, Llena JF, et al. A histopathological study of craniopharyngioma with special reference to its stroma and surrounding tissue. Brain Tumor Pathol 1997;14:41–45.
15. Plesec TP, Prayson RA. Frozen section discrepancy in the evaluation of central nervous system tumors. Arch Pathol Lab Med 2007;131:1532–1540.
16. Plesec TP, Prayson RA. Frozen section discrepancy in the evaluation of nonneoplastic central nervous system samples. Ann Diagn Pathol 2009;13:359–366.
Keywords:

glioma; hemangiopericytoma; oligodendroglioma; craniopharyngioma; gliosis

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.