Pituicytoma—An outlook on possible targeted therapies

Summary

Introduction

Pituicytoma is a rare neoplasm of the sella region. Tumor resection is the primary treatment option, but remains subtotal due to excessive bleeding in many cases. The search for alternative or additional treatment regimens is necessary.

Aims

We aimed to determine the receptor expression of pituicytoma to find alternatives or supplements to surgical therapy in the use of targeted therapies.

Methods

Pituicytoma samples were collected from three institutions between 2006 and 2015 and were stained for vascular endothelial growth factors (VEGF), thyroid transcription factor (TTF1), and somatostatin receptors (SSTR 2/3/5). The stains were classified from 0=no staining to +++=strong staining. A complementary retrospective analysis of the patient charts regarding sex, age, and primary symptoms, pituitary function, and perioperative complications was performed.

Results

Ten samples were analyzed; mean patient age was 57.8 years SD 16.3 years. Seven samples were acquired from male patients (one relapse) and three from female. All tumors stained strongly positive (+++) for VEGF‐R. Six samples stained positive for TTF1. As for somatostatin receptors, three samples were slightly positive for SSTR 2; seven were negative. SSTR 3 was + in one, three were ++, three were +++, and three were 0. SSTR 5 stained +++ in 1, ++ in 5, + in 1, and 0 in three patients.

Conclusions

Pituicytomas were generally positive for VEGFR and showed regular expression of SSTR 3 and 5 indicating a possible treatment option through targeted therapies in cases where resection remains insufficient. Further research is necessary as to whether tumor growth can be inhibited using these pathways.

1. INTRODUCTION

The term pituicytoma was first introduced by Liss in 19581 describing a tumor of the posterior pituitary gland. It is generally thought to be a rare neoplasm of the sellar region (MRI sequence, Figure 1) originating from the cells of the infundibulum or the neurohypophysis,2 although a recent study by Phillips speculates about a possible development from the folliculostellate cells of the adenohypophysis.3 Pituicytomas generally consist of bipolar spindle cells4 with occasional intermediate filaments and rare mitotic figures.3 Immunohistochemical analysis shows a general positivity for glial fibrillary acidic protein (GFAP), thyroid transcription factor 1 (TTF1),5, 6 and S100,3 as well as an occasional positivity for vimentin.7 Recently positivity for vascular endothelial growth factor (VEGF) has been described.7 With the revision of the World Health Organization Classification of Tumors in 2016, these entities are classified as group I of TTF1‐expressing tumors,8 a group also containing spindle cell oncocytoma and granular cell tumor who were historically already discussed to be the same tumor family as pituicytoma.9

Figure 1.

T1‐weighed biplanar MRI with contrast enhancement (Gadolinium). Left: Coronal view. Right: Sagittal view, both with suprasellar pituicytoma as a homogenously enhancing suprasellar mass

So far, only 96 cases have been published10, 11, 12 with the largest series containing 11 patients12 while most descriptions are single case reports or small series of no more than three tumors. Surgery with complete tumor resection is a curative option and therefore considered the primary goal of any treatment. However, due to a tendency toward strong vascularization and adherence to the surrounding tissue, the goal of gross total resection (GTR) is not always reachable. In a recent review of 67 cases by Feng, only 42% of all transsphenoidal approaches and 33% of all craniotomies could reach GTR.10 As resection often remains incomplete, recurrence can be observed in up to 30% after subtotal tumor removal.10 After recurrence, either repeat surgery or radiotherapy has been tried for further treatment. Due to the high rates of recurrence, lesions suspicious of a pituicytoma should be treated only when symptomatic or when radiographic progression is seen in order to omit complications and keep the patient out of harm's way.

Targeted therapies as an alternative treatment method have not yet been described for recurrent pituicytoma but may be a possible treatment option after incomplete resection and radiotherapy. Karamandchani et al. in their study from 2012 described a diffuse positivity for VEGF.7 However, receptors for targeted therapies have to our knowledge not yet been described. In order to establish individual treatment algorithms in surgically difficult cases or recurring pituicytoma, the aim of this study was to assess the presence of the necessary receptor configuration for potential alternative therapies.

2. METHODS

We identified nine patients with histologically confirmed pituicytoma (Table 1), three females and six male with a mean age of 57.6 years, range 22‐79 years. All patients were initially surgically treated between 2004 and 2014 at three major institutions for pituitary surgery.

Table 1.

Patient cohort and clinical findings

ID	Sex	Age	Symptoms	Endocrine dysfunction before surgery							Endocrine dysfunction after surgery							GTR	STR	Bleed	Surgical approach	Recurrence
				GH	TSH	ACTH	Gon	Lacto	ADH	OX	GH	TSH	ACTH	Gon	Lacto	ADH	OX
1	M	79	Weight loss, muscle weakness, visual impairment		X	X	X					X	X	X				X		–	TS	No
2	M	54	Fatigue, moodiness, libido loss		X		X					X		X					X	–	TS	No
3	M	72	Fatigue, moodiness, libido loss	X	X		X				X	X		X					X	X	TS/TC	Deceased surgery related
4	F	43	Amenorrhea					X											X	X	TS	No
5	F	51	Visual impairment					X							X				X	X	TC	No
6	M	68	Headache, libido loss, fatigue, circulatory dysregulation	X	X	X	X	X			X	X	X	X	X				X	–	TS	No
7	F	56	Visual impairment																X	–	TC	No
8	M	74	Collapse during addison crisis	X	X	X	X	X			X	X	X	X	X				X	X	TS	Yes
9	M	22	Epilepsy, impaired memory, headache	X	X		X					X		X			X	X		–	TS	No

The intra‐operative specimens were fixed in buffered formalin for at least 2 hours followed by paraffin embedding of all tissue in every case; 4‐μm paraffin sections were stained for hematoxylin and eosin (HE), periodic‐acid Schiff reaction (PAS), and for reticulin fibers using Gomori′s stain following standard laboratory procedures. Immunohistochemistry for thyroid transcription factor 1 (TTF1; manufacturer: Dako Agilent, Santa Clara, CA, USA, catalogue number: M3575; at a dilution of 1:100), vascular endothelial growth factor (VEGR; Santa Cruz SC‐7269, Biotechnology, Dallas, TX, USA; 1:1000) and its receptor (VEGF‐R; Acris, Herford, Germany 1:100), somatostatin receptor types 2, 3, and 5 (SSTR 2, SSTR 3, SSTR 5; ZytoMed RBK046‐05, Zytomed Systems GmbH, Berlin, Germany, 1:100; Thermo PA3207, Thermo Scientific, Pinneberg, Germany, 1:1000; RBK051‐05, 1:10, Thermo Scientific, Pinneberg, Germany), and proliferation marker Ki67 (MIB‐1; RM‐9106‐S, Thermo Scientific; 1:1000) was performed on an automated Ventana HX IHC system (Ventana‐Roche Medical systems, Tucson, AZ, USA) following the manufacturer′s instructions. Validity of staining results was verified using external positive controls in every staining passage. Specifically, a capillary hemangioblastoma WHO‐grade I was used as a positive control for VEGF and VEGFR, a meningothelial meningioma WHO‐grade I for SSTR 2, normal pituitary gland (adenohypophysis) for SSTR 3 and SSTR 5, and a bronchogenic adenocarcinoma for TTF1. For quantitative assessment of proliferative activity (Ki‐67), the region with the highest density of labeled nuclei was chosen and the percentage of labeled nuclei in relation to all tumor cell nuclei was evaluated in an area of 0.1 mm².

3. ETHICAL CONSENT

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study was conducted in accordance with ethical guidelines of the local review board. This is a retrospective study. For this type of study, formal consent is not required.

4. RESULTS

Nine patients were surgically treated for pituicytoma; one was treated twice by one of the senior surgeons (once for recurrence). GTR was achieved in three cases (33.3%), and no recurrence was observed for GTR. Six patients underwent only partial removal. Two‐thirds (66.6%) of the tumors removed only in part showed excessive bleeding intra‐operatively, three received adjuvant radiation therapy, and one of those showed a recurrent tumor. One patient with incomplete resection and remarkable intra‐operative bleeding was treated with repeat surgery a year later due to tumor progression, and this patient died 3 weeks after surgery due to extensive postoperative hemorrhage at the site of resection (Figure 2).

Figure 2.

Timeline of Patient #3: CT scans at initial presentation (upper left) followed by transsphenoidal biopsy (postbiopsy upper middle), tumor progression after 1 y (upper right) with subsequent transcranial resection attempt, and fatal hemorrhage in sagittal (lower left) and axial (lower right) CT scan

Symptoms and endocrine dysfunction are presented in Table 1; initially, no patient showed deterioration of posterior pituitary function. Only in one patient pituitary function improved after surgery, and the previously observed amenorrhea was completely reversible.

Routine microscopy of the tumors in our series showed the characteristic spindle cell pattern for all tumors with varying degree of storiform to diffuse growth patterns. Ki67 labeling index was low (<1%) in one of 10 cases, medium (1%‐5%), 7 of 10 cases, and high (10%) in two cases. One high Ki67 of 10% was found in patient #8 also presenting recurrent pituicytoma. For the case #3, the Ki67 increased from initial surgery to repeat surgery (<1%‐4.4%). There was a strong positivity for TTF1 in all samples, vimentin was available for 6 of 10 samples and always stained positive. GFAP expression was observed irregularly in 33% of all cases, whereas S100 was found positive in 90% (9 of 10 cases) both markers always stained with high intensity. Synaptophysin was slightly positive in one case (1 of 6 cases) and otherwise negative.

4.1. Potential target receptors

After staining, the ten samples of pituicytoma a uniform positivity for VEGF‐R was observed in all tumor specimens, VEGF ligand was originally stained in four cases, and only one showed mild signs of staining (Figure 3A,B).

Figure 3.

(A) Upper left: negative stain for SSTR 2; Upper right: strong positivity for SSTR 3; Lower left: slight positivity for SSTR 5; Lower right: Strong positivity for VEGFR. (B) Hematoxylin‐eosin stain of pituicytoma; Right: TTF1 positivity in pituicytoma

Somatostatin receptor (SSTR) 2 showed only slight positivity in three samples, and all others were negative. It stained slightly positive for patient #3 initially but was negative on recurrence.

A positivity for SSTR 3 was observed in six cases (three strong, three medium), one showed slight staining, and three were negative. There was an increase in expression for the recurring tumor of patient #3.

SSTR 5 was positive in seven samples (six medium, one slightly) and negative in three samples (Table 2).

Table 2.

Receptor configuration for target therapies: Degree of positivity 0/+/++/+++; n.a.=stain not available

ID	Recurrence	Ki67	SSTR 2	SSTR 3	SSTR 5	VEGFR	VEGF	TTF1	Vimentin	GFAP	S100	Synaptophysin
1	No	11.3%	(+)	+++	++	+++	0	Positive	+++	0	+++	n.a.
2	No	4.1%	0	++	++	+++	n.a.	Positive	+++	0	0	0
3	Yes	<1%	0	+++	++	+++	0	Positive	n.a.	0	+++	n.a.
3 (recurrence)	Yes	4.4%	(+)	++	++	+++	0	Positive	n.a.	0	+++	0
4	No	1.2%	0	+++	++	+++	n.a.	Positive	+++	0	+++	n.a.
5	No	1%	0	0	0	+++	n.a.	Positive	++	+++	+++	0
6	n.a.	1%	0	0	0	+++	n.a.	Positive	n.a.	0	+++	0
7	n.a.	1%	0	(+)	0	+++	n.a.	Positive	+++	+++	+++	+
8	Yes	10%	(+)	0	+	+++	n.a.	Positive	+	0	+++	0
9	No	1.2%	0	++	+++	+++	(+)	Positive	n.a.	+++	+++	n.a.

Controls on healthy neurohypophyseal samples showed no expression of SSTR 2, 3, or 5. VEGF‐R was present only within the capillary tissue of the posterior hypophysis and very few pituicytes.

5. DISCUSSION

In the present study, we analyzed 10 pituicytoma samples of nine patients in order to investigate the presence of receptors for targeted therapies. All tumors showed a strong positivity for VEGF‐R and 67% were positive for both SSTR 3 and 5. All of these receptors have previously been addressed in targeted therapies for other tumor entities and may be a viable therapeutic option in pituicytoma as well.

Alternative therapeutic approaches are required since pituicytoma, a benign tumor by nature, poses some significant challenges in treatment. Due to the highly vascular nature of the stroma and the firmness of the tumor tissue itself, removal may be complex. In a review by Feng et al.10 in 2013, 47 tumors were reviewed specifically, GTR could be accomplished for 49% and a subtotal resection was achieved in 51% of all tumors operated on and a tumor recurrence in 29.2% of all incomplete resections. The current treatment for recurring pituicytoma is repeat surgery in curative intention or radiation therapy10; however as shown in our case study, repeat surgery of vulnerable well‐vascularized lesions poses some significant risks which may even lead to a fatal outcome.

With regard to the high rate of recurrence, our findings of pituicytoma receptor status may be beneficial after partial resection, especially in tumors with strong tissue adhesion and a tendency toward excessive bleeding. The ligand of the VEGF‐R would pose an excellent target for a targeted therapy with, for example, the VEGF antibody bevacizumab as the receptor is so strongly represented in the tumor cell membranes. A major drawback in the treatment of intracranial tumors with bevacizumab has been the reported spontaneous hemorrhage and stroke.13 However, through genetic evaluation of the single nucleotide polymorphism, the patient population at risk of vascular events may be reduced.14 Additionally, in a recent larger study comparable adverse event ratios for standard treatment and additional bevacizumab in human glioblastoma were shown.15 Therefore, bevacizumab may still be a viable therapeutic option after failed resection and radiation therapy.

The somatostatin receptors may also be available for a direct antitumoral treatment regime using somatostatin analogues; however, we found a strong variation in tumor receptor expression. As the samples from our collective were all mechanically stained in one batch, a technical inaccuracy is very unlikely. Therefore, we conclude that SSTR receptor expression seems to be of some variance within the tumor collective of pituicytoma. The receptor may still me a ligand for a targeted therapy, but specific qualitative and quantitative examinations should precede the decision on such a treatment.

Two methodologically different treatment alternatives exist in this sector. Somatostatin receptor analogues are ligands to the receptor inducing cell cycle arrest and a downregulation of protein biosynthesis therefore creating an imbalance toward increased apoptosis within the affected stroma.16, 17 Also as a secondary effect, somatostatin analogues, like VEGF inhibitors, also affect tumor vascularization; this may be of special importance due to the highly vascular nature of the pituicytoma itself.17 As pituicytoma samples in our study stained mainly for SSTR 3 and 5, the use of, for example, pasireotide with a high affinity to those receptors may be a potential option.17 This agent has so far been approved for the use as a second line treatment in Cushing's disease and acromegaly and shown good clinical results including tumor mass reduction,18, 19

An alternative tumor treatment method also employing the somatostatin receptor as a target structure may be the use of radiolabeled somatostatin analogues like DOTATOC or DOTATATE.20, 21 The key principle of this approach is the accumulation of local toxicity through beta radiation of isotopes attached to somatostatin receptor ligands.21 The main target receptor in the approaches for the central nervous system so far has been SSTR 220 because of the high affinity of the available radiotracers toward this structure.22 This is unfortunate for the case of pituicytoma which showed no particular expression of SSTR 2 in our series; the available ligands do, however, bind to SSTR 3 and 5 as well but with lower affinity.22 A therapeutic approach may therefore still be possible, but radiotracers with high affinity for SSTR 3 and 5 and a low affinity for SSTR 2 are necessary; otherwise undesired higher toxicity and adverse effects in other nontumor tissue may result; however these results may be employed in a diagnostic approach prior to surgery, for example, by DOTATOC‐PET.

Regarding tumor diagnostics, we could confirm the findings by previous studies describing multiple pituicytoma histologies.2, 11, 12, 23 Typically, the tumors presented spindle cell architecture with a storiform or interlacing pattern11 and only few mitoses. The Ki67 labeling index showed a stronger variation (1%‐10%) compared to the literature where it has been reported at 1%‐6%,12, 24 especially in our cases of recurring pituicytoma the Ki67 labeling index appeared to be increased. In the case of recurrent pituicytoma with present first and second histology, we could observe an increase of the Ki67 from <1%‐4.4% potentially explaining the sudden clinically apparent tumor growth. However, one patient with 10% Ki67 did not show recurrence probably due to complete resection. Another case with 5% Ki67 was partly resected but received immediate postoperative radiation therapy and may therefore not have shown recurrent growth. Aside from ultrastructural features, we could confirm the regular presence of S100 except for a single case that did not stain accordingly and were able to observe the reported positivity of TTF1 and vimentin12 in all of our samples. GFAP expression was variable and only present in 33% of our cases and synaptophysin only stained positive in one sample also confirming previous investigations.11, 12, 24

6. CONCLUSION

Pituicytoma of the neurohypophysis is a rare but usually benign neoplasia of the central nervous system, which may pose a challenging therapy, due to the difficulty resulting from strong adherence, invasion, and bleeding tendency. As the majority of tumors are not completely resectable and recurrence after partial removal is high, adjuvant therapies move into the therapeutic focus. Targeted therapies may be a possible option in this treatment regime in the future, especially the strong representation of VEGF‐R may be an excellent target for individual treatment approaches. A more practical approach may be SSTR‐based treatments considering the variable SSTR receptor expression as serious adverse events may be reduced.

ACKNOWLEDGMENTS

No funding was received for this work.

DISCLOSURE

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this article.

CONFLICT OF INTEREST

All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or nonfinancial interest in the subject matter or materials discussed in this manuscript.

Mende KC, Matschke J, Burkhardt T, et al. Pituicytoma—An outlook on possible targeted therapies. CNS Neurosci Ther. 2017;23:620‐626. 10.1111/cns.12709