Common Molecular Cytogenetic Pathway in Papillary Tumors of the Pineal Region (PTPR)

Abstract

Primary papillary tumors of the central nervous system and particularly the pineal region are rare. Papillary tumor of the pineal region (PTPR) is a recently described neoplasm that has been formally recognized in the 2007 World Health Organization Classification of Tumors of the Nervous System. Because of their rarity, further pheno‐ and genotypical observations as well as therapeutic experience are necessary to differentiate PTPR from other primary or secondary papillary tumors of this region. We herein present three cases of PTPR characterized by local recurrence in two of them. Primary and recurrent tumors were analyzed by immunohistochemistry and comparative genomic hybridization (CGH). From our results clonal chromosomal aberrations can be postulated which seem to be a feasible tool to differentiate PTPRs from other primary or secondary papillary tumors of this region.

INTRODUCTION

Tumors of the pineal region are infrequent lesions that account for only 1% of all intracranial tumors (2). They often display papillary features necessitating a differential diagnosis of ependymoma, choroid plexus papilloma, papillary meningioma and metastatic papillary carcinoma (40). Papillary tumor of the pineal region (PTPR) has recently been defined as a distinct pathological entity in the 2007 World Health Organization Classification 30, 43 and is increasingly recognized with up to now 79 reported cases 1, 3, 4, 6, 7, 8, 10, 13, 18, 20, 21, 22, 24, 25, 26, 27, 28, 34, 35, 39, 44, 45, 46, 47, one PTPR reported in a dog (52). The origin remains controversial, but a possible origin from specialized ependymal cells of the subcommisural organ has been suggested (22). There is only one report on cytogenetic characterization of five papillary tumors of the pineal region (18). To the published cases we now add the cytogenetic characteristics of four PTPR of three patients, two of them suffering from local tumor recurrence after irradiation and adjuvant chemotherapy. We show distinct chromosomal aberrations, which can not only contribute to a better understanding of the origin of this entity, but also help to differentiate PTPR from other more frequent papillary entities of this region.

CLINICAL CASE ILLUSTRATIONS

Case 1

In 1996, a 13 8/12‐year‐old boy suddenly became symptomatic with signs of hydrocephalus and diplopia. Magnetic resonance imaging (MRI) revealed a 5‐cm in diameter, cystic enhancing tumor growing in the pineal region and protruding into the third ventricle, causing obstructive hydrocephalus. A ventriculoperitoneal shunt was placed first. After 3 months, occipital craniotomy was performed and complete tumor resection was achieved. Histologically the tumor was classified as an anaplastic plexus papilloma World Health Organization (WHO) III. This interpretation was confirmed by the German national reference center for brain tumor pathology. Polychemotherapy with carboplatin, VP16 and Vincristine as well as irradiation of the craniospinal axis followed. At a follow‐up examination 5 years later in 2001, a local tumor recurrence was detected and resurgery was performed using the same operative strategy. Residual tumor at the vena of Galen was treated by stereotactic irradiation (20 Gy). Again, the tumor was histologically classified as an anaplastic plexus papilloma WHO III, which was confirmed by two second consultations taken. Three years later in 2004, a 0.8‐cm recurrent tumor localized at the left dorsal thalamus was as well irradiated by gamma‐knife (22 Gy). Eight months passing, another 3‐mm tumor occurred within the left ventricle adjacent the Foramen of Monroi. Again, gamma‐knife irradiation (20 Gy) was done. In 2005, the slides of the tumor specimen were re‐examined, as a new tumor entity PTPR had been described (23). After further work up with extended immunohistochemistry the tumor was reclassified as PTPR. The patient outcome is satisfying as he is now working as a computer scientist 15 years after initial surgery without neurological deficits.

Case 2

In 1994, a 26‐year‐old woman suffered from progressive headaches and diplopia MRI scan revealed a 3.5‐cm contrast‐enhancing tumor of the pineal region, causing hydrocephalus by obstructing the third ventricle. Stereotactic biopsy was taken and a ventriculoperitoneal shunt was placed. A diagnosis of pinealoblastoma WHO IV was made, which was responded by interstitial radiotherapy with temporary I‐125 seeds. The tumor was constant in size during follow‐up and regained size 7 years after the first biopsy. In 2001, the tumor was operated through a parietal parasagittal approach and was histologically again classified as a pinealoblastoma WHO IV. Irradiation of the craniospinal axis was performed (46 Gy). Three years later (2004), the local recurrent tumor again was resected, this time via a suboccipital approach. Postsurgery the patient suffered from vertical gaze palsy. Histopathological examination now disclosed features reminiscent of anaplastic ependymoma WHO III, and a second opinion was seeked for both this and the foregoing tumor. The German national reference center for brain tumors considered both lesions in this patient as a PTPR of intermediate dignity. The third recurrence of the PTPR 18 months later (2006) was treated by stereotactic irradiation (30.6 Gy). Another 12 months later (2007) a fourth local recurrence as well as a spinal metastasis at L3 (case published by Hong et al (20)) were surgically removed. The fifth craniotomy was performed because of further local tumor recurrence in the pineal region, 5 months after the spinal surgery. Despite of radiation therapy, the neurological status deteriorated rapidly and the patient died 3 months after the last craniotomy.

Case 3

In 2010 a 53‐year‐old man presented with a 2‐month history of nuchal pain. MRI demonstrated a solid tumor in the pineal region with uniform contrast enhancement, measuring 1 cm in diameter. On examination, he did not have any neurological deficits. Total resection of the tumor via a supracerebellar, infratentorial approach by a median suboccipital craniotomy was achieved. Histologically, PTPR was diagnosed, which was confirmed by the German national reference center for brain tumors. The postoperative course was complicated by subcutaneous CSF effusion, which was treated by insertion of a lumbar drain. In the early postoperative period upward gaze was limited by a vertical supranuclear palsy. During follow‐up the vertical gaze palsy improved. As there was no residual or recurrent tumor in the MRI 3 months postoperatively, it was decided not to administer adjuvant therapy. The patient is alive and doing well at 8 months of follow‐up.

MICROSCOPIC ANALYSIS AND IMMUNOHISTOCHEMISTRY

Histological examination in all cases revealed a tumor of moderate cellularity with solid and papillary formations. The neoplastic cells were of medium size, round to oval in solid areas, and columnar or cuboidal covering papillary formations. Cytoplasm was light eosinophilic or clear, partly vacuolated, the nuclei were monomorphic with finely granular chromatin. Focally, the tumors contained ependymoma‐like pseudorosettes and tube‐like structures. Mitotic activity was generally low, but one of the two recurrent tumors (case 2) revealed a high Ki67‐labeling index of 20%. Necrosis was found in cases 1 and 2. Inflammatory reaction was not present.

Immunohistochemical analysis (Figure 1) revealed a mixed neuroendocrine and epithelial phenotype of tumor cells (Table 1) with strong staining for neuron‐specific enolase (NSE) and S100 and focal positivity of cytokeratin 8/18 (CK8/18) except for recurrent tumor of case 1. Cases 2 and 3 also showed strong expression of vimentin and microtubule‐associated protein 2 (MAP2). Synaptophysin was positive in case 1, while reactivity was negative in case 2 and observed only in some cells of case 3. Epithelial membrane antigen (EMA) was noted in papillary areas of case 1, focal and very discrete at the apical border of tumor cells in case 2, and was observed as intracytoplasmic droplets in case 3. Glial fibrillary acidic protein (GFAP) staining was observed focally in a perivascular localization.

Figure 1.

Immunohistochemical analyses in three patients with papillary tumor of the pineal region (PTPR). Abbreviations: CK8/18 = cytokeratin 8/18; EMA = epithelial membrane antigen; GFAP = glial fibrillary acidic protein; HE = hematoxylin and eosin; MAP2 = microtubule‐associated protein 2; NSE = neuron‐specific enolase.

Table 1.

Immunohistochemical profile in four papillary tumors of the pineal region (PTPR) of three patients.

Case	NSE	S100	Synaptophysin	CK8/18	Vimentin	MAP2	EMA	GFAP
1	+++	+++	++	+	+ (vessels only)	+	+	Perivascular
1 (rec)	+++	+++	++	−	+ (vessels only)	+	+	Perivascular
2 (rec)	+++	+++	−	+	+++	++	+	Perivascular
3	+++	+++	+	+	+++	+++	+	−

Abbreviations: CK8/18 = cytokeratin 8/18; EMA = epithelial membrane antigen; GFAP = glial fibrillary acidic protein; MAP2 = microtubule‐associated protein 2; NSE = neuron‐specific enolase; rec = recurrent tumor.

CYTOGENETIC ANALYSIS

DNA was isolated from formalin‐fixed and paraffin‐embedded specimen (primary and recurrent tumor in case 1; recurrent tumor in case 2; and primary tumor in case 3). By using comparative genomic hybridization (CGH) all four tumors were analyzed (17), providing comprehensive overview of chromosomal gains and losses (Table 2). All tumors revealed gain at 9p. In two of primary cases gains at 4, 5, 12p11‐12q21 and loss at 10 were observed. Primary and recurrent tumor in case 1 shared gains of chromosomes 4, 8, 9p and 12p12q21 and the loss of chromosome 10, proving their clonal chromosomal relation. Additionally, gain of 20q and loss of 6q24qter were only observed in the recurrent tumor in case 1.

Table 2.

Chromosomal gains and losses in nine papillary tumors of the pineal region (PTPR).

Case No	Gender/age (years)	Losses	Gains
1 (own)	Male/ 14	1p32pter, 10, 15q	4, 8, 9p, 11, 12p12q21
1 recurrence (own)	Male/ 19	6q24qter, 10	4p15qter, 8, 9pterq2, 12p12q22, 20
2 recurrence (own)	Female/34	10	3, 8, 9p, 12p11qter
3 (own)	Male/53	17	4, 5, 9p13pter, 13q, 18q),
1*	Unknown	None	None
2*	Unknown	10	4,5p12–24, 5q11.2–23, 7q11.2–31, 13q, 15q, 18
3*	Unknown	10, 22q	4, 8q11.2–24.1, 9p, 12p, 12q12–23
4*	Unknown	10, 13q, 22q	4, 8, 9, 11, 12p, 12q12–22, 17,20
5*	Unknown	10, 22q	4, 6p, 8, 9p, 11, 12, 20p

^* cases from Hasselblatt et al (18).

Bold figures indicate the most common chromosomal aberrations.

DISCUSSION

The differential diagnosis of PTPR is broad and includes all pineal region lesions known to exhibit papillary architecture. These are the more common ependymoma and choroid plexus tumors (Table 3), but as well uncommon tumors like papillary meningioma and metastases.

Table 3.

Differential diagnosis to papillary tumor of the pineal region (PTPR).

	PTPR	Ependymoma	Choroid plexus papilloma/carcinoma
Mean age (years)	31.5	children: 5.1	3.5
		adults: 46.4
Gender	female > male	male ≥ female	male > female
MRI appearance	Cystic	Well delineated, plastic	Well delineated, lobulated
T1	Heterogeneous, hyperintens	Iso‐hypointens	Mottled, iso‐hypointens
Blood	−	+	+
Calcification	−	+	+
Enhancement	Light heterogeneous	Moderate heterogeneous	Intense homogeneous
IH vimentin	+−+++	+++	+++
NSE	+++	−	−
S100	+++	+++	+
MAP2	++	−	+
Cytokeratins	+ CK8/18	−	CK7, CK8/18
EMA	+	+	+
Transthyretin	+	−	++
Kir7.1	(+)	−	++
Stanniocalcin	(+)	?	++
Synaptophysin	−/+	−	−
GFAP	−/+	++	+
Treatment	Surgical resection, stereotactic irradiation	Surgical resection, irradiation	Surgical resection (irradiation)
CGH	+4, +8q, +9p, +12p, −10	+1q, −6q, −9p, −13q	+5, +7, +9p, +12q, −10q

CGH = comparative genomic hybridization; EMA = epithelial membrane antigen; GFAP = glial fibrillary acidic protein; IH = immunohistochemistry; MAP2 = microtubule‐associated protein 2; NSE = neuron‐specific enolase.

It was only in 2003 that Jouvet and coworkers reported on a tumor entity of the pineal region, which was distinct from the so far known neoplasms (22). This entity was subsequently called PTPR and was included in the 2007 WHO Classification of Tumours of the Central Nervous System 29, 30. Characteristic features of PTPR are an epithelial‐like appearance with papillary and solid cellular areas, perivascular pseudorosettes or tubes 13, 22, 55. The tumor cells exhibit light eosinophilic or vacuolated cytoplasm, and mostly show strong immune reactions with antibodies against NSE, S100 and vimentin, variable staining for cytokeratin and EMA, and mostly negative reactions for GFAP 22, 47. Mitotic activity varies, and focal necrosis can be found. Precise histological grading criteria remain to be defined (30).

Histological differentiation of PTPR from other papillary tumors of the pineal region is challenging. Differentiation from ependymomas seems to be most difficult as PTPR display features of ependymal differentiation, eg, tubes and perivascular pseudorosettes as well as EMA reactivity. Furthermore, PTPR is suspected to be derived from specialized ependyma of the subcommissural organ 22, 47. Yet, unlike PTPR, ordinary ependymomas usually do not display a striking epitheloid appearance with vacuolated cells and mostly show strong GFAP reactivity 14, 48. Plexus papillomas can be easily distinguished from PTPR by their morphology, but atypical plexus papillomas (WHO grade II) and plexus carcinomas (WHO grade III) may have a more solid appearance and blurring of the papillary architecture, displaying similarities to PTPR. However, PTPR only rarely express Kir7.1 and stanniocalcin‐1, which are markers of choroid plexus tumors 15, 18, 40. PTPR may show neuroendocrine features with strong NSE positivity, and have to be distinguished from pineal gland tumors, which in contrast to PTPR show strong synaptophysin positivity (29). Papillary meningiomas are characterized by a perivascular orientation of the tumor cells, but can be differentiated from PTPR by their lack of cytokeratin expression (9). Metastatic tumors of the pineal gland are rarely reported in the literature, reaching a 4% incidence in patients with disseminated neoplasia 19, 23, 50. Papillary metastasis may originate from thyroid gland and, unlike PTPR, express thyroid transcription factor (TTF1) and thyreoglobulin while metastatic ovarian cancer can be distinguished by the expression of estrogen and progesterone receptor.

Aforesaid problem of correctly classifying PTPR demands a reliable tool for differentiating PTPR from other papillary tumors. By adding the CGH results from our study to the five tumors presented in the study by Hasselblatt and colleagues, one tumor showed no chromosomal imbalances (18), but in PTPR revealing cytogenetic imbalances, seven of eight (88%) show loss of whole chromosome 10 and gains of 9p, six of eight (75%) show gains of whole chromosome 4 and 12p12q21, five of eight (63%) demonstrate gain of chromosome 8 (8q). Intracranial ependymomas are characterized by gains of 1q, losses on 6q, 9p and 13q 5, 11, 16, 41, 49. In comparison, choroid plexus papillomas typically show hyperdiploidy and most abundant aberrations are gains of 5, 7, 9p and 12q as well as losses of 10q 16, 40. Anaplastic meningiomas reveal a loss of chromosome 1p, 9p and 22q in over 70%, followed by losses of 20q, 12q, 15q, 6q, 10q, 14q, 1q, 9q and 17q 37, 38, 42, 51. Therefore, PTPR share loss of chromosome 10q and gain at 9p with choroid plexus papillomas and anaplastic meningiomas, but show no overlapping chromosomal aberrations to ependymoma. Our results support the observation that PTPRs are genetically separate from ependymomas, as it was assumed before in interpreting microarray analyses in two PTPRs (12).

PTPR are believed to behave different from ependymoma and choroid plexus papilloma (30), but understanding of the clinical and biological behavior of PTPR is still evolving and not understood. The clinical course of PTPR seems to be characterized by local recurrence 13, 22 and sometimes aggressive behavior with leptomeningeal spreading (28). Five‐year disease‐free and overall survival are 73% and 27%, respectively (13). Total resection is probably the only clinical factor to be associated with good survival and absence of recurrence 13, 25, 44. The effect of irradiation and chemotherapy on PTPRs needs to be investigated, but stereotactic irradiation seems to be promising (4), as seen in our patient 1. But because of the low number of reported cases, treatment guidelines for PTPRs have not yet been established, in contrast to the other papillary lesions of the pineal region. For ependymomas, stratification based on age reveals 5‐year survival rates of 76% in adults and 14% in children 32, 33. The National Comprehensive Cancer Network (NCCN) suggests gross total resection, followed by radiation in WHO II and III ependymomas, in grade II tumors also mere observation can be considered (36). Choroid plexus tumors can usually be cured by surgery alone with a 5‐year survival rate of 80% to 100% with occasional recurrences 31, 40, 53, 54. The prognosis of these patients is better when complete tumor removal is achieved 53, 54. The use of chemotherapy or radiation therapy is considered on an individual basis, but might be helpful in patients who progress despite surgical removal (27).

In summary, PTPR has probably been misclassified for years as papillary ependymoma, choroid plexus tumor, papillary meningioma or metastatic papillary carcinoma from an unknown primary. Therefore, cytogenetic analyses in large collectives of papillary brain tumors published before 2003, when the entity of PTPR was not yet defined, have to be interpreted with reservation. Interpreting the results of our study, we postulate that with molecular and cytogenetic studies of histologically clearly defined and larger case series, PTPR might be distinguishable from other tumors of the pineal region.