UPDATES IN AGGRESSIVE PITUITARY TUMORS

Abstract

Aggressive pituitary tumors lie between pituitary adenomas and carcinomas, displaying a particular behavior, with invasion, resistance to conventional therapy and early recurrence. The radiological grading, along with prognostic markers such as Ki-67 proliferation index, p53, MGMT and transcription factors are important factors in establishing the benign, aggressive, or malignant nature of pituitary tumors, with a more accurate treatment strategy.

In this article, we report the novelties in defining, classifying, and managing aggressive pituitary tumors and their malignant potential, focusing on clinicopathological, histological, molecular and radiological data.

INTRODUCTION

Most of the sellar masses that arise from intrapituitary tissue are benign pituitary adenomas. They can cause hormone hypersecretion or compression of the surrounding structures and some exhibit an aggressive behavior. In rare cases, pituitary tumors are defined as carcinomas (0.2%), presenting mandatory with craniospinal or systemic metastatic dissemination (1,2).

The clinical phenotype of pituitary tumors can vary significantly, with an extended cycle, from “totally” silent and clinically silent to “whispering” and functioning adenomas (3). In a large cohort of patients with aggressive tumors, most were clinically functional (69%), and from the clinically silent diagnosed tumors, 25% became functional (4).

Pituitary adenomas can be sporadic or familial, situation when they arrive as part of a syndrome (multiple endocrine neoplasia type 1 or 4, McCune-Albright, Carney complex or DICER1). They appear at a younger age, with an aggressive behavior and resistance to conventional treatment strategies, as well as AIP-related ones, which expose an aggressive behavior through the alteration of the cytoskeleton organization. AIP and GPR101 are the two genes responsible for FIPA (familial isolated pituitary adenomas). Considering the aggressive behavior of SDHx-related pituitary tumors, patients with these mutations should be screened for this kind of tumors (5). In a clinical case of a patient with acromegaly due to a giant pituitary macroadenoma, with double surgical transsphenoidal intervention and high voltage radiotherapy, he did not achieve biochemical control even with combined therapy with maximum doses of somatostatin analogue and growth hormone receptor antagonist. He has a confirmed AIP mutation (Fig. 1).

Figure 1.

Direct sequencing of PCR products for AIP gene - ht c.940C>T, p.R314W </i>(courtesy of Dr. I. Baciu, Dr. S. Găloiu).

The terminology of pituitary neuroendocrine tumor (PitNET) has been proposed to recognize the complexity and heterogeneity of these tumors, with a wide potential of invasion and proliferation (1).

In 2017, the World Health Organization (WHO) published the 4^th edition of the classification of endocrine tumors (WHO Classification of Tumours of Endocrine Organs – IARC, 2017). The most common neuroendocrine tumors of the pituitary, pituitary adenomas, are classified utilizing immunohistochemistry (IHC) as the main tool, according to cell lineages and essential transcription factors (TFs) for differentiation and maturation of pituitary cells: PIT-1 (pituitary specific POU-class homedomain transcription factor) for acidophilic lineage (somatotrophs, lactotrophs and thyrotrophs), T-PIT (T-box family member TBX19) for corticotrophs and SF-1 (steroidogenic factor 1) for gonadotrophs (7, 8).

Although the definition of aggressiveness varies widely in literature, the most recent guideline of the European Society of Endocrinology for the management of aggressive pituitary tumors states that they are large, invasive in surrounding structures, grow more rapidly than non-aggressive ones and have clinically relevant growth despite conventional therapies, such as surgery, radiotherapy or classical medical treatment (9). In a recent publication (2), important details are offered regarding this definition. An invasive tumor has a Knosp grade 3 or 4, invades the sphenoid sinus and associates an unusually rapid growth, over 20% and at least 2 mm in 6 months. A progression over 20% despite optimal treatment is required to state that the tumor growth is clinically relevant (2).

The likelihood for recurrence is higher in adenomas with an elevated proliferative activity and in some special variants of tumors, such as sparsely granulated somatotroph adenoma, lactotroph adenoma in men, silent corticotroph adenoma, Crooke cell adenoma and plurihormonal PIT-1 positive adenoma (7).

Examining twelve clinical cases, the authors proposed the term “refractory pituitary adenoma” using the following criteria: tumor infiltrates adjacent structures, a Ki-67 index greater than 3% and growth velocity above 2% monthly, on which current treatments fail to control tumor growth and/or hormonal hypersecretion and tumor recurrence occurs within 6 months after surgery (10).

Pituitary carcinomas usually originate from prolactinomas and corticotropinomas, with a high frequency of initially silent and later evolution to functioning behavior (2).

The versatile behavior of aggressive pituitary tumors can be highlighted by another clinical case from our archive (Fig. 2). At initial presentation, the female patient, 17 years old, had a 30 mm pituitary tumor, with suprasellar extension and optic chiasm compression. She had clinical signs of GH hypersecretion and gonadotroph and thyrotroph deficiency. She had double pituitary transsphenoidal surgery and gamma-knife radiosurgery after second surgical intervention, with no hormonal control. IGF-1 levels normalized only after adding pegvisomant to octreotide LAR. Yet, few months after, the patient presented for diplopia, exophthalmos, palpebral ptosis and spontaneous reappearance of menses, probably in the context of apoplexy.

Figure 2.

Biochemical and tumor size evolution on imaging after multimodal treatment (courtesy of Prof. Dr. C. Poiană, Dr. A. Dumitrascu).

Epidemiology

Although considered rare in the past, pituitary adenomas are now a common pathology, with a prevalence of 80-100/100.000. Pituitary adenomas that are clinically significant affect 1 in 1000 people (9,11). Yet the prevalence of aggressive pituitary adenomas is not clear.

The most commonly pediatric aggressive pituitary adenomas are somatotropinomas and giant prolactinomas, with implication of MEN1 and AIP mutations in up to 25% of patients. Because of the decreased cell proliferation in the elderly, the incidence of aggressive pituitary adenomas tends to be lower (12) and are most clinically non-functioning (9).

A recent publication states that less than 0.1% of detected pituitary tumors are malignant, about 0.5% are aggressive and, considering all pituitary tumors, the percentage is lower than that (13).

Immunohistochemistry and molecular biology

As mentioned earlier, clinical markers used to diagnose aggressive pituitary tumors are size, invasiveness and growth rate.

Regarding tumor functionality, from a total of 49 cases in a clinical study, 26 were identified as clinically aggressive and, out of them, 46.2% were more likely to be functional, being detected in males (65.4%) and, in multivariate analysis, gender, functional status and Ki67 index were independent predictors of clinical aggressiveness (14).

Despite several debates from literature concerning its cut-off and true prognostic value, the ESE guideline recommends the dosage of proliferation index Ki67 and immunodetection of pituitary hormones as a minimal starting evaluation. As concluded in a recent review, a Ki67 ≥10% can be seen as prognostic marker for aggressivity, yet taken together with the other markers (15).

PIT-1 is associated with estrogen receptor (ER) and GATA transcription factor 2 (GATA2), TPIT and NeuroD1 are required for corticotroph differentiation and SF1 and GATA2 for gonadotroph lineage, as well as GATA3 suggested in a recent study (16,17). Using immunohistochemistry, the detection of TFs is essential in immunonegative, plurihormonal, somatotrophs or silent corticotroph tumors and in those with ≤5% of immunopositive cells, but the diagnosis cannot be primarily or exclusively made only on TFs detection (18).

A five-tiered classification conceived by the European Pituitary Pathology Group, which classifies PitNETs into 7 main morphological and functional types, uses tumor diameter, type and grading, the last based on invasion (MRI) and proliferation (Ki-67 index, mitotic count and p53) and its prognostic value has been validated on 1470 patients in 4 studies (18). Although the new WHO classification does not plead precisely in favor of p53 expression as a marker of aggressiveness, it may help in differentiating some cases and also in identifying patients with high risk of early recurrence and progression (19) and the detection is also recommended by the ESE guideline, when Ki67 index is ≥3% (7,9). Besides hormonal pituitary panel and proliferation markers, in their IHC analysis and histological report, the French group can also quantify TFs, LMWCK (low molecular weight cytokeratins) for stratification of somatotroph and identification of corticotroph tumors and chromogranin A. If required, ERα detection for lactotroph tumors, SSTRs (somatostatin receptors), MGMT (O(6)-methylguanine-DNA methyltransferase) and TIFF-1 for differential diagnosis between metastases and non-endocrine sellar masses are also identified (20).

A study on 1055 resected pituitary tumors supports the idea that the tumor classification and radiological quantification of the extent are much better prognostic and aggressivity predictors than the proliferation markers (21).

Pituitary carcinomas have always been a controversy because of different theories regarding their starting point: either there is a transformation from a pre-existing pituitary adenoma or they arise “de novo” (22).

Upregulation of cyclin D1, VEGF (vascular endothelial growth factor), MMP-9 (matrix metallo peptidase 9), miRNAs, and p21Cip1 showed a specific relation with progression, vasculogenesis, metastasis and invasion, and can contribute to the transformation from adenomas to carcinomas. Tumor suppression and temozolomide (TMZ) sensitivity are influenced by downregulated factors such as MGMT, p16Ink4A, p27Kip1 and MT3 (metallothionein isoform 3). These findings by Yang et al. suggest that the pathogenesis of pituitary carcinomas has many similarities to other neoplasia (23). The same review states that the average latency period for a pituitary adenoma to become a carcinoma is of 9.5 years for ACTH-secretin carcinomas and of 4.7 years for prolactin secreting carcinomas, based on case reports from the literature.

Neuroimaging

Imaging information plays an important role in defining and classifying aggressive pituitary tumors. Aggressiveness, as defined radiologically, means invasion and rapid growth. Invasion implies tumor extension beyond bone or duramater. In some cases, invasion of the cavernous sinus means aggressiveness, although in some prolactinomas with good response to dopamine agonist treatment, this is not always true (24). The Knosp criteria are largely used for imaging classification and invasion. In a study by Xu et al., the classification had a sensitivity of 47% and a specificity of 91.1% in predicting the aggressiveness of Cushing’s disease (25).

In a multicentric retrospective study with 297 acromegaly cases, T2-hypointensity on MRI was associated with a lower rate of optic chiasm compression, a less frequently invasion of the cavernous sinus and higher IGF-1 levels. Also, these tumors were smaller than their T2-hyperintense or isointense counterparts (26).

To precisely identify tumor volume changes, MRI evaluation should be performed every 6 to 12 months. Bonneville et al. proposes a comparison of diameters on coronal sections, using a subcalossal line as a fixed anatomical structure (24).

Very useful tools for a more precise postsurgical follow-up are machine learning models that correlate pituitary tumors preoperative T2-weighted (T2W) MRI images with Ki-67 proliferation index found at immunohistochemistry. In a recent study, the authors reported an accuracy of 91.37% of correctly classified patients (27). Another retrospective study performed on 400 patients with pituitary adenomas reported an accuracy of 87% in predicting early postoperative outcomes (28).

In a predictive model used to identify surgical remission in acromegaly, the best features were older age, Knosp grade 0-2 and lower presurgical GH values, with a sensitivity of 93.8% (29).

Regarding the common sites for metastases of pituitary carcinomas (spinal cord, lungs, pancreas, liver, kidneys and bones), the imaging diagnosis is a key part in identifying the type of tumors (24). In defining tumor remnants after surgery or to screen for metastatic disease in case of suspicion of a pituitary carcinoma, a 68Ga-DOTA-TOC can be more precisely than MRI alone, as seen in several clinical cases (30).

Results from different studies emphasize the role of PET-MRI and 11C-methionine PET imaging, which are very useful in the accurate diagnosis of tumor target in recurrent disease and can guide a successful surgical reintervention (31).

Treatment

Therapeutic options for aggressive pituitary tumors include surgery (performed by a high-volume pituitary neurosurgeon), radiotherapy, standard medical treatments in functioning tumors, chemotherapy (temozolomide) and loco-regional therapies (9).

Besides the clinical and histological characteristics of aggressive pituitary tumors, another special feature is the lack of responsiveness to common treatment steps.

Surgery is the first step recommended in the management of aggressive pituitary tumors and the option to repeat surgery before considering other treatments (9).

Both types of radiotherapy have an elevated efficacy in controlling tumor growth, with a more rapid antisecretory effect in the case of Gamma-Knife (GK). Although the timing of radiotherapy is debatable, data suggests that it is also effective in controlling tumor volume (32).

Temozolomide, first line chemotherapy, is used in aggressive tumors that grow in spite of classical therapy and can reduce up to 40% of tumor size (33). The ESE guideline recommends a first radiological evaluation after 3 months of treatment and offers the possibility to combine it with radiotherapy in rapid growing tumors (Stupp protocol) (9). Data from literature support the implication of HIF1α pathway in pituitary tumorigenesis, its inhibition being demonstrated to increase sensitivity to TMZ (5).

Low tumoral MGMT expression predicts longer survival and a better response to treatment and a high expression may confer resistance to TMZ (34). It is also recommended to evaluate the MGMT status by IHC as a predictor of response to TMZ. Other predictors are DNA mismatch repair proteins (MMR) such as negative MSH6, observed to occur in case of progression to carcinoma and resistance to TMZ (22).

In a few reported cases, TMZ has been associated with other drugs, such as capecitabine, thalidomide or BCNU and also with novel targeted therapies, but these associations are not recommended prior to the 3-months monotherapy with TMZ (34).

PRRT (peptide receptor radionuclide therapy), a therapeutic option for carcinoid and neuroendocrine tumors, has rarely been used in pituitary pathology. In a clinical series, six out of twelve patients responded to PRRT with disease stabilization or partial tumor remission (median follow-up of 44 months) (30).

Because many vascular growth factors are involved in pituitary tumorigenesis (35), anti-VEGF therapy can be useful in stabilizing the disease, as proven by treatment with bevacizumab in some clinical cases of corticotroph carcinomas (36, 37). In pituitary tumors, gefitinib, lapatinib and canertinib are the tyrosine-kinase inhibitors evaluated for their treatment. Also, targeting the ErbB pathway proved to be useful in aggressive lactotroph and corticotroph tumors, which have a higher EGFR and ErbB2 protein activity and expression (38).

In conclusion, aggressive pituitary adenomas are unique for many reasons: they can have a silent behavior; they can have a rapid invasive evolution, or they can suffer transformation to pituitary carcinomas. Whatever their features may be, the management must be supervised by a multidisciplinary team, for best diagnosis and therapeutic decision. Having the WHO 2017 classification of endocrine tumors and the ESE 2018 Guideline for management of aggressive pituitary tumors, the way for a good clinical practice for these tumors tends to be facilitated. The nature of pituitary adenomas can be predicted with the help of immunohistochemistry and specific markers such as Ki67, p53, MGMT or by detecting transcription factors. This helps in better patient management and can predict the outcome, recurrences, and prognosis.

Conflict of interest

The authors declare that they have no conflict of interest.