Newer Concepts in the Classification of Pituitary Adenomas
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.287667
Source of Support: None, Conflict of Interest: None
Keywords: Pituitary adenoma, Transcription factors, PitNETKey Message: This review provides an insight into the current classification scheme which utilizes transcription factors to categorise pituitary adenomas based on their lineage.
The classification of pituitary adenomas has evolved from one based on tinctorial properties to that based on adenohypophyseal cell lineage with particular emphasis on pituitary specific transcription factors., The last few years have seen a flurry of activities with discussions on nomenclature.,, The 4th edition of the World Health Organization (WHO) on Tumours of Endocrine Organs, 2017, witnessed several changes in the classification, however, the debate on nomenclature is still active.,,,,, This review highlights the novel aspects of the WHO 2017 classification of pituitary adenomas.
Pituitary adenomas were classified as acidophil, basophil, and chromophobe adenomas in the early part of the last century based on cytoplasmic staining with hematoxylin and eosin stains. The classification then evolved in the '80s and '90s to one based on immunohistochemical staining for the various pituitary hormones.
Horvath and Kovacs further refined the classification introducing electron microscopic features. In the four authored WHO classification of 2000, a five-tier classification was proposed that included clinical data, histological appearance, immunohistochemistry (IHC), and electron microscopy. In 2004, the principles of the classification remained focussed on immunoprofile and ultrastructural features but added granulation pattern of tumor cells into the subtyping of individual tumors.
The role of pituitary-specific transcription factors in clarifying cytogenesis of pituitary adenomas was acknowledged but not included in the WHO 2004 classification and it was only in 2017 that their role in the tumorigenesis of pituitary adenomas was taken into account.,
The new classification is based on IHC for pituitary hormones, transcription factors, and other markers used in routine pathology practice, like low molecular weight cytokeratin (LMWCK) and estrogen-receptor alpha (ERα). The classification is practical and eliminates electron microscopy. In the new classification, the term “hormone-producing” has been changed to “troph” to underscore the role of transcription factors in cell differentiation and regulation of specific pituitary anterior hormones., For example, growth hormone-producing adenomas are now termed somatotroph adenomas.
The three main pathways involved in cell differentiation in the adenohypophysis are as follows:
Immunohistochemical markers are available for each of these transcription factors but have not been fully validated. Pituitary hormone expression by IHC remains the first line of investigation in subtyping of pituitary adenomas. Transcription factors are particularly important to ascertain lineage in those tumors that have minimal (<5%) to absent immunopositivity with the various pituitary hormones.
In addition to transcription factors, IHC for LMWCK, ERα, and GATA transcription factor 2 (GATA2) also serve a role in the subtyping of some of these tumors. The LMWCK recommended is CAM 5.2 or AE1/AE3 pankeratin cocktail. LMWCK is an important immunohistochemical marker in the distinction of sparsely granulated and densely granulated somatotroph adenoma. The former show paranuclear cytokeratin positive “fibrous bodies” in more than 70% of tumor cells. Intense cytoplasmic staining for cytokeratin is seen in densely granulated somatotroph and corticotroph adenomas. Crooke cell adenomas and Crooke's hyaline of non-tumorous corticotrophs show a perinuclear ring of positivity for cytokeratin. Acidophil stem cell adenomas and plurihormonal PIT1 positive (silent subtype 3) adenomas have also been reported to contain “fibrous bodies.” ERαthe expression is seen in lactotroph adenomas and gonadotrophs. while GATA2 expression is seen in gonadotrophs and thyrotrophs. GATA3 has been suggested as an immunohistochemical marker complementary to SF1 in the diagnosis of gonadotroph adenomas.
There has been some reshuffling of entities in the WHO 2017 classification [Table 1]. Mammosomatotroph adenomas and mixed somatotroph and lactotroph adenomas are now classified under somatotroph adenomas. The additional expression of the alpha subunit and estrogen receptor alpha by these two tumors differentiates them from somatotroph adenomas that may on occasion express prolactin (PRL) (<10% of cells). Mammosomatotroph adenomas express growth hormone (GH) and PRL (>10%) in the same cell while mixed somatotroph and lactotroph adenomas express them in different cells.
Acidophil stem cell adenoma was classified under both somatotroph and lactotroph adenomas in the 2004 classification. Although this entity is mentioned in the WHO 2017 in both the chapters on somatotroph adenomas and lactotroph adenomas, in the consolidated classification, it is classified as a lactotroph adenoma., These tumors are chromophobic with oncocytic change and intracytoplasmic vacuoles. The latter correlates with abnormal accumulation of degenerative (dilated or giant) mitochondria and differs from pure lactotroph adenomas in that they additionally show focal expression of GH in the background of diffuse immunoreactivity for PRL and often contain small intracytoplasmic fibrous bodies that are highlighted by LMWCK.,,,
Silent adenomas are those that are clinically non-functional but show immunoreactivity for the various pituitary hormones. Silent subtypes are described amongst somatotrophs, lactotrophs, corticotrophs, and thyrotrophs. Gonadotrophs are by nature clinically silent.
The definition of pituitary carcinomas remains unchanged. It is defined as a tumor of adenohypophyseal cells that metastasizes craniospinally or is associated with systemic metastasis.
Until 2017, clinically silent pituitary adenomas that were negative for all the pituitary hormones were termed null cell adenomas. The advent of transcription factors led to a redefinition of null cell adenoma and plurihormonal PIT1 positive adenomas [Table 2].
Kovacs et al. while defining null cell adenomas had predicted that with the emergence of a better understanding of the cytodifferentiation of pituitary adenomas, it was likely that the term “null cell adenomas” would become obsolete. The prevalence of null cell adenomas in that series was 16.5%. The WHO 2004 classification defined null cell adenomas as “adenomas that have no hormonal immunoreactivity and no other immunohistochemical or ultrastructural markers of specific adenohypophyseal cell differentiation.” Null cell adenomas are defined in the WHO 2017 as non-functional tumors that are immunonegative for all hormones and pituitary-specific transcription factors. The use of SF1 IHC in non-functional, hormone negative tumors have predictably helped in identifying “silent” gonadotroph tumors which, in turn, has led to shrinkage in the proportion of null cell tumors. In addition, PIT1 recognizes the clinically silent plurihormonal adenomas and TPIT IHC identifies the silent corticotroph adenomas that are immunonegative for ACTH. True null cell adenomas are, therefore, tumors that are immunonegative for both hormones and transcription factors. Once a sellar tumor is found to be hormone and transcription factor negative, it is important to exclude other entities such as paraganglioma, before ascribing it to the term null cell adenoma. Sellar paragangliomas are usually negative for cytokeratins and are immunopositive for synaptophysin and chromogranin. However, some pituitary adenomas such as gonadotroph and null cell adenomas have a similar immunoprofile. Sellar paragangliomas show nuclear positivity for GATA3 and also express tyrosine hydroxylase. Preliminary data has now reported the prevalence of null cell adenomas to be <1% of all adenomas.
The silent subtype 3 was earlier considered a subtype of the silent corticotroph group of tumors based on immunoreactivity for POMC-derived peptides. However, the lack of consistent immunoreactivity for POMC-derived peptides and its unique ultrastructural characteristics led to the word “corticotroph” being dropped and the tumors being named silent subtype 3 adenomas. Subtypes 1 and 2 are silent subtypes of corticotroph adenomas, the former being densely granulated and the latter a sparsely granulated corticotroph adenoma. The expression of PIT1 by silent subtype 3 together with variable immunoreactivity for pituitary hormones led to them being classified in WHO 2017 as plurihormonal PIT1 positive adenomas. Asa, however, sought to clarify that amongst the plurihormonal PIT1 lineage tumors, are some hormonally active tumors morphologically resembling densely granulated somatotroph or mammosomatotroph adenomas that should be classified as a subtype of their functional counterparts. On the other hand, the clinically silent PIT1 positive tumors have a distinct morphology with elongated cells with scant secretory granules and have prominent nuclear spheridia on ultrastructure. These tumors, apart from being clinically silent, are highly invasive and resistant to medical therapy. Asa argues that the term “poorly differentiated tumor of Pit-1 lineage” should be ascribed to this subset of clinically silent PIT1 positive tumors. These tumors on IHC may express GH and to a lesser extent PRL and may be positive for ER-α, alpha subunit and TSH. Intracytoplasmic fibrous bodies may be identified by LMWCK/CAM5.2.
The most controversial aspect of the 2004 classification was the inclusion of the term “atypical pituitary adenomas” for those invasive tumors with an elevated mitotic index, Ki67/MIB-1 labeling index of ≥3% and extensive nuclear immunostaining for p53. Several clinical studies found these criteria to be unreliable both in terms of reproducibility and in the prediction of aggressive behavior in pituitary adenomas.,,, Atypical adenomas were, therefore, removed in the 2017 classification and this entity is now obsolete. WHO 2017 recommends that in addition to the sub-typing of the tumor by histochemical and immunohistochemical methods, accurate assessment of MIB-1 labeling index and clinical features such as tumor invasion are needed to identify clinically aggressive tumors. The latter is referred to as “high-risk” adenomas in WHO 2017 and include:
Furthermore, pituitary adenomas with germline AIP mutations are also refractory to medical and surgical treatment and tend to recur,
Oncocytomas with a prevalence of 13.4% were included as a subset of null cell adenomas in the WHO 2004., This is now accepted as a phenotypic variant with abundant mitochondria seen in any subtype, but more frequently seen in gonadotroph adenomas.
Adenomas with unusual immunohistochemical combinations
These are plurihormonal tumors that are polymorphous and composed of an admixture of cells of different lineages that express more than one adenohyophyseal hormone and their corresponding transcription factors. Synchronous GH and PRL, or synchronous β-follicle-stimulating hormone (β-FSH) and β-luteinizing hormone (β-LH) are not considered plurihormonal adenomas. Care must be taken not to misinterpret trapped non-tumorous tissue or cross-reactivities of antisera as plurihormonal tumors.
Pituitary tumors that show geographically distinct cell populations are seen rarely. The distinct cell lineage can now be confirmed on IHC for pituitary specific transcription factors. These components may be clinically silent or functional., In autopsy series, silent lactotroph adenomas with other adenoma types are the most frequent. On the other hand, in surgical series hormonally active somatotroph adenomas are more common.
Pituitary blastomas, regarded as a separate entity in the WHO 2017 classification, are rare primitive malignant neoplasms of the pituitary gland. They occur exclusively in infants and children younger than 24 months (median 8 months). There is a slight female predominance. Cushing's syndrome followed by ophthalmoplegia are reported in about two-thirds of cases and is associated with DICER1 mutation. Histologically, they are composed of three main components, including epithelial glands with rosette-like arrangement resembling immature Rathke epithelium, small primitive-looking cells with a blastema-like appearance, and large secretory cells resembling adenohypophyseal cells. The tumors are positive for ACTH.
Although pituitary adenomas are benign a significant number behave aggressively with local invasion, repeated recurrences, significant morbidity, and persistent symptoms requiring long-term medication and on occasion radiation therapy for disease control. To address this conundrum the Pituitary Pathology Club in its 14th meeting in Annecy, France in November 2016, proposed that pituitary adenomas be renamed as “Pituitary Neuroendocrine Tumors” with the acronym PitNET. They argued that the term “tumor” instead of “adenoma' permitted for the variability in biological behavior and aligned these tumors with the terminology used for other neuroendocrine neoplasms (NENs). Those against this change in terminology argue that replacing the word “adenoma” with “tumor” is misleading as it gives pituitary adenomas a sinister connotation when the majority are not aggressive. The grouping with neuroendocrine tumors implies that there is no distinction between an endocrine cell and a neuroendocrine cell and does not address the issue of whether this term should apply to tumors of the thyroid and adrenal. Since the change would not affect decision making regarding management they argue that there is no compelling case to change the name. The counter-argument by the proponents of a change in terminology is that the main function of the hormone-secreting cells of the adenohypophysis, unlike that of thyroid follicular epithelium and adrenal cortical cells, is the uptake of amines that are processed into peptide hormones which are stored in membrane-bound neurosecretory granules before secretion. This functional feature together with its expression of neuroendocrine markers synaptophysin and chromogranin support their neuroendocrine nature. They feel that the change emphasizes the use of an integrated approach in management like that for NETs including using the novel treatment protocols that have emerged and are common to NETs at all sites. The common classification framework for neuroendocrine neoplasms suggested by the International Agency for Research on Cancer (IARC) following a consensus conference in November 2017, in consultation with additional experts endorsed the proposal of the Pituitary Pathology Club that pituitary tumors be classified as NETs, i.e., pituitary NETs (PitNETs) rather than adenomas or carcinomas. These tumors would be considered as well-differentiated NETS based on criteria used in other organs. They also stated that poorly differentiated neuroendocrine carcinomas (NECs) do not occur in the pituitary and further genetic studies are required to ascertain if NECs exist. Furthermore, they clarified that paragangliomas arising in and around the sella turcica are to be classified as a separate family distinct from PitNETs.
A multidisciplinary approach including endocrinologists, neuroradiologists, neurosurgeons, neuropathologists, geneticists, and oncologists has always been advocated for the management of pituitary adenomas. The baseline information must include a detailed assessment of the endocrine status, preoperative pharmacological treatment, family history, tumor size, presence or absence of invasion into surrounding structures, and whether or not the tumor is recurrent. The histopathological examination would require, apart from the routine hematoxylin and eosin stain, the use of a reticulin stain or Collagen IV by IHC to identify normal adenohypophysis. The latter is particularly relevant in identifying a corticotroph microadenomas within the adenohypophysis. Although not indicators of aggressiveness, microscopic invasion of the dura, bone, or respiratory mucosa should be documented. The mitotic activity should be looked for and counted per mm2. Taking into account variations between microscopes five fields at a magnification of 40× equates to about 1 mm2. Amitotic count of >2/10 HPF indicates an aggressive tumor. Hot spots seen on MIB-1/Ki-67 should be reexamined for mitosis. The European Pituitary Pathology Group (EPPG) recommends counting the MIB-1 positive cells in two hotspots and counting of 500–1000 cells per hotspot. An index of ≥3% is of prognostic importance. WHO 2017, however, does not give a precise cutoff and only recommends estimation of the proliferative index.
The EPPG further recommends that the hormonal profile by IHC should include an approximate percentage of positive cells. The role of LMWCK and other markers such as GATA2 and ERα have already been alluded to and can be utilized when required. Some authors recommend the use of transcription factors as the first line of immunohistochemical testing as a resource-saving measure. However, TPIT has only recently become commercially available and requires validation., It is premature to move to an approach that uses these markers exclusively in the classification of pituitary adenomas., The main role of transcription factor IHC at present is in further classification of the hormone negative cases to identify the silent gonadotrophs, the silent corticotrophs, and the plurihormonal PIT1 positive tumors.
The evolution in our understanding of pituitary adenomas has been fascinating. More than a decade elapsed before the 2004 classification of pituitary adenomas was modified., There is no doubt that the 2017 WHO classification will evolve further with our improving understanding of the biology and genetics of these tumors together with improvements in treatment strategies. While the technique of IHC for classification of pituitary adenomas can be performed in any histopathology laboratory, the financial implications of an expanding panel of markers need to be balanced with its utility in guiding management and prognosis.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]