Key words
prolactinoma - aggressive - invasive - malignant - molecular endocrinology
Introduction
Approximately 10–15% of all manifest intracranial tumors originate
from the hypophysis (pituitary gland) [1].
Prolactinomas are benign neoplasms (adenoma) that produce prolactin [2]. Prolactin can be secreted by various
pituitary neuroendocrine tumors deriving either from somatotrophs or
mammo-somatotroph cells besides the lactotroph cells [3]. Similarly, densely or sparsely granulated
lactotroph tumors, poorly differentiated Pit1-lineage tumors and acidophil stem cell
tumors can also cause hyperprolactinemia [3].
Prolactinomas – originating from lactotroph cells – are the most
frequently hormone‑secreting pituitary tumors (approximately 30–40%
of all pituitary tumors), and its prevalence is about 94 per 100 000 inhabitants
[2]. Prolactinomas are much more
frequently diagnosed in female population with a sex ratio around 10:1 before the
fifth decade, but after this age, the prolactinoma frequency is about the same in
both gender [3]. This shows that female
hormones may propagate prolactinoma growth; and indeed, there exist substantial
evidence supporting this hypothesis, which will be discussed below. But seemingly
paradoxical at the first glance, prolactinomas appearing in boys, at a young age
(<20 years), and/or with accompaniment of a genetic predisposition
have worse prognosis [4]. The most frequent
clinical signs of prolactinomas are gonadal and sexual dysfunction and subsequent
infertility in both sexes [2]. Dopamine
agonists, such as bromocriptine and cabergoline, are employed for treatment in
prolactinomas since prolactinomas express high levels of dopamine receptors (D2R)
[2]. When the patients cannot tolerate
dopamine agonists or are refractory to medical therapy, they undergo surgical
treatment and/or radiotherapy [5]
[6]. Ten to 15% of patients are
resistant to medical treatment. Twenty-five percent of the patients receiving
bromocriptine treatment fail to normalize prolactin whereas it is only
10–15% in those receiving cabergoline [6]. Almost half of tumor volume reduction is
encountered in 33% percent of those receiving bromocriptine and
10–15% of those treated with cabergoline [6]. Resistance to cabergoline is defined by the
absence of prolactin normalization or lack of the tumor to reduce in size by
50%, and these refractory prolactinomas tend to exert higher angiogenesis,
cell proliferation and atypia and invasiveness [5]. Quinagolide is a non-ergot-derived D2R-selective agonist, which could
provide significant reduction in tumor size and prolactin levels in around
90% of patients [3]. Quinagolide is
usually given at the average dose of 150 –
300 μg/day and compared to bromocriptine, more prominent
reduction in dizziness, nausea, vomiting, and drowsiness is generally witnessed
[3].
Most patients who respond well to dopamine agonists with reduction of prolactin
levels exert evident decreases in tumor volumes, but not all do. On the contrary,
some patients may experience almost total reductions in tumor volumes unaccompanied
with normalizing prolactin levels [6].
Resistance mechanisms include lowered D2R (dopamine receptor D2) gene transcription,
lowered receptor activity which regulate D2R expression, and lowered inhibitory G
protein-expression which couples the D2R to adenylyl cyclase [5]
[6].
Prolactinomas are classified and treated according to their size, that is,
microprolactinomas (<1 cm) do not generally invade the adjacent
anatomic structures, while macroprolactinomas (>1 cm) tend to
locally invade and compress surroundings. Surgical cure rates for invasive
macroprolactinomas are meager, and even if resected, larger prolactinomas tend to
recur [2]. Here, data regarding molecular
endocrinology of treatment resistant and aggressive prolactinomas are reviewed.
Prolactinomas developing at a young age are correlated to higher proliferation and
invasion. Other features, including the expression of growth factors such as VEGF
(vascular endothelial growth factor) and EGF (epidermal growth factor), adhesion
molecules (E-cadherin), the genes regulating proliferation, invasion and
differentiation, matrix metalloproteinase 9, and chromosome abnormalities
(chromosomes 1, 11, and 19) also correlate with aggressiveness [4].
Evidence Acquisition and Synthesis
Evidence Acquisition and Synthesis
To obtain data on the pathogenesis and treatment of aggressive prolactinomas, the
following keywords were searched in PubMed database: prolactinoma AND (invasive OR
aggressive OR malignant) AND (chemotherapy OR radiotherapy OR molecular OR gene OR
genetic), which yielded 2288 results (Last Check: 11th May, 2021). Based on the
obtained results from PubMed data search, we mostly focused on molecular
endocrinological features of these tumors since they are clinically targetable. The
role of conventional treatments including temozolomide chemotherapy and radiotherapy
is also analyzed. At first, we will define the entities of
“invasive”, “aggressive” and
“malignant” prolactinomas and pituitary carcinomas, their anatomical
spread mechanisms and conventional treatment modalities. Then, we will discuss
molecular endocrinological features of these tumors as potential treatment
targets.
Are the Invasive Prolactinomas, Aggressive Prolactinomas, Malignant
Prolactinomas, and Pituitary Carcinomas Different Entities? Anatomy and
Immunohistopathology
Invasive prolactinomas are prolactin secreting tumors invading adjacent
anatomical structures, while aggressivity determines not only the invasion but
also medical treatment-resistance and high tendency for recurrence. It is
recommended the term “aggressive” should not be used
synonymously with “invasive” as aggressivity encompasses a
broader biologically ominious behavior including drug resistance and
occasionally atypical histology besides invasiveness [5]. In general, invasiveness is mostly a
radiological and aggressiveness a clinical definition [7]. Invasive prolactinomas are tumors with
proven invasion into adjacent structures, including cavernous and sphenoid
sinuses and bony structures, which can be defined radiologically with
preoperative MR investigations, during operation, or with histopathological
demonstration of tumor spread to the bone, dura, or nasal mucosa [7]. A relatively recent study demonstrated
that prolactinomas constitute most invasive pituitary tumors [7]. In our index case, which will be
discussed below, extensive cavernous sinus, prepontine area and pontocerebellar
edge invasion and encasement of internal carotid artery, CV, CVII, and CVIII
nerves were detected. According to World Health Organization’s (WHO)
recent criteria (2016), increased mitotic index, Ki67 labeling index greater
than 3% percent and robust p53 expression indicate the aggressivity of
pituitary adenomas [8]. But it shall be
also emphasized that there exist conflicting results between Ki67 indices with
invasiveness of pituitary tumors [7]. The
tumor suppressor TP53 gene encodes the transcription factor p53, which is
a genomic “gate keeper” that initially blocks cell cycle and
succeedingly induces DNA repair following genomic damage. However, when the DNA
injury is irreversible, it induces apoptosis of the cells, thereby eliminating
the spread of mutations into the cell progeny [9]. Wild type p53 protein exists in benign cells, but due to its very
relatively half-life, p53 is present in diminutive amounts, even undetectable by
immunohistochemistry (IHC), whereas accumulated mutant p53 protein are
detectable by IHC. Therefore, positivity of p53 expression is associated with
p53 mutations [9]. It was shown that
non-invasive and invasive adenomas and pituitary carcinomas exerted expression
of p53 in 0, 15.2, and 100% of cases, respectively, which correlated
with invasion in a recent study [7]. In
addition, it was shown that non-functioning pituitary adenomas are non-reactive
for p53, while functional pituitary adenomas express p53, and functioning
adenomas have more aggressive features than nonfunctioning adenomas including
immunoreactivity for p53, S100, prolactin and MGMT (methylguanine
methyltransferase) [7]. Basaran et al.
defined the fraction of MGMT-immunopositive tumor cells among pituitary adenomas
according to the following score: –, no positive tumor cells;
+ , < 10% positive tumor cells;
++ , 10–50% positive tumor cells;
+++ , > 50% positive
tumor cells, regardless of intensity and showed that MGMT expression correlated
with invasiveness [7].
There exists a spectrum of medication resistance and it is considered that most
treatment-resistant prolactinomas are not carcinomas [5]. The single criterion to define a
prolactinoma as “malignant” is the presence of distance
metastasis according to the WHO 2016 classification. Peculiarly, it is not
necessary that “malignant prolactinomas” exert histopathological
features associated with malignancy (high mitotic rates, necrosis etc.) and
prolactinomas with completely benign morphological appearance can also
metastasize [10]. Therefore, the
definition for malignancy of pituitary tumors is a debated issue as histological
features, immunohistochemistry, or even electron microscopical features cannot
distinguish a malignant pituitary adenoma unless metastases develop [10]. Nonetheless, some researchers suggest
that p53 immunopositivity indicating TP53 gene mutation tend to be more
frequent among malignant prolactinomas [11]. Malignant prolactinomas may metastasize to bony skeleton, lymph
nodes, lung, liver, and ovaries [11].
Pituitary carcinomas invade of adjacent structures and exhibit prominent cell
proliferation and are defined by the existence of craniospinal and/or
systemic metastases. Moreover, besides these features, they also exert
histopathological features of malignancy. Pituitary carcinomas account nearly
0.1–0.2% of all pituitary tumors with an average survival period
less than 4 years [8]. Pituitary
carcinomas possess a higher Ki-67 labeling index (higher than 11%).
Unexpectedly, there are reports of pituitary carcinomas with lower Ki67 indices,
indicating that there exist certainly other factors that contribute to malignant
potential features [5]. The existence of
nuclear pleomorphism and high rates of mitosis should raise suspicion for
pituitary carcinoma [5].
Anatomical Mechanisms of Prolactinoma Spread
Anatomical Mechanisms of Prolactinoma Spread
For the anatomical mechanisms of metastasis, several mechanisms are considered:
spread through dural venous channels, metastases through blood and lymphatic
metastases [10]. Subarachnoid dissemination
during surgery or spontaneously are also presumed as responsible mechanisms. Some
authors prefer the term “metastases” only for extracranial spread,
while they accept the term “seeding deposits” for cerebrospinal
metastases [10]. Nonetheless, some early
studies suggested that the disease course does not support such a distinction,
because hematogenous and cerebrospinal fluidborne metastases exert equally malignant
behavior [10]. It is also assumed that
intraoperative dispersion of neoplastic cells into the subarachnoid space may occur
in patients operated by transcranial or even transsphenoidal approach for pituitary
macroadenoma. Nonetheless, metastases develop rarely and approximately 50%
of cases with systemic metastases had no surgical treatment and the same may be
relevant for intraoperative dispersion of tumor cells via the blood stream [10]. Taking the high percentage of locally
invasive tumors invading venous structure bone and dura (ranging between 10 to
42% percent) into account, it might be considered that tumor cells enter the
circulation in more patients than in those who do develop metastases [10]. Conceivably, it is more logical to assume
that factors that enhance survival and implantation of the tumor cells play a
greater role in metastatic spread [10].
Classical Treatment Modalities of Dopamine Agonist-Resistant
Prolactinomas
Classical Treatment Modalities of Dopamine Agonist-Resistant
Prolactinomas
Surgical Treatment
Before the development of dopamine agonists, surgery was the main treatment for
the management of a prolactinoma [3].
Currently, the main indications of surgery are resistance to medical treatment,
pituitary apoplexy, and the patients personal choice. Even in patients with
visual defects due to a macroprolactinoma, dopamine agonists constitute the
standard of care as compared to surgery [3]. In highly trained hands, adenomectomy normalizes prolactin levels
in 75–90% of microprolactinomas. In invasive macroprolactinomas,
the surgical cure is achieved in about 40% of cases with recurrence
rates about 20% for 10 years [3].
Even in an expert pituitary center, surgery for macroprolactinomas can be
complicated by anterior pituitary deficiencies in 1–15%,
cerebrospinal fluid leak in 2–10%, and diabetes insipidus in
around 5% of cases, respectively [3].
Radiotherapy
Radiotherapy is generally employed in conjunction with surgical treatment to
manage recurring and/or aggressive pituitary adenomas [12]. Radiotherapy aims to slower or block
growth of tumors and normalize prolactin levels which include stereotactic
radiosurgery (SRS) and external beam radiation therapy (EBRT) [12]. Both approaches are almost equivalent
in reducing prolactin levels (34.1% for EBRT vs. 31.4% for SRS).
SRS is preferably applied as its three dimensional approach enables faster
correction of prolactin oversecretion and reduced risks of carotid stenosis and
radiotherapy-associated secondary malignancies [3]. Amongst SRS, GKRS (gamma knife radiosurgery) provides a highly
selective conformal intervention in a single application using a linear particle
accelerator or multiheaded cobalt unit and performed with image guidance [3]. On the contrary, the conformal
radiotherapy is applied with several fractions over time (general administration
on a daily basis) [3]. According to some
observers, prolactinomas’ radiation-sensitivity is not high, with one
study demonstrating only an 18% remission rate at 4 years for
prolactinomas treated with SRS [12]. On
the other hand, another study revealed that Gamma Knife SRS normalized
hyperprolactinemia in 50% of cases with medication-refractory
prolactinomas, yet cavernous sinus invasion was a predictor of treatment failure
to normalize prolactin levels [12]. A more
recent study reported good outcomes with GKRS in treatment of prolactinomas
[13]. Indications for GKRS were (i)
dopamine agonist resistance (17 patients), (ii) intolerance to dopamine agonists
(5 patients), or (iii) attempts to reduce the length of treatment and/or
the dosage of dopamine-agonist treatment (6 patients). After GKRS, normal
prolactin level was achieved in about 82% of patients, out of which
hormonal remission (normal prolactin levels after discontinuation of dopamine
agonists) was achieved in 13 (46.4%), and endocrine control (normal
prolactin levels while taking dopamine agonists) in 10 (35.7%) patients
[13]. GKRS blocked adenoma growth or
reduced adenoma size in all cases.
Temozolomide
Temozolomide is an alkylating chemotherapeutic, which is a dacarbazine derivative
with lipophilic properties enhancing its traversal through the blood-brain
barrier [14]. Temozolomide efficacy was
first documented in glioblastomas before its employment in treatment of
neuroendocrine neoplasias and melanomas. Chen et al. described a 17-year-old
male patient admitted with an aggressive prolactinoma that progressed despite
surgery, gamma-knife, and dopamine agonists, which responded well to
temozolomide treatment with a marked reduction of tumor mass, decrease of
prolactin secretion and progressive clinical improvement [8]. The authors also underlined the
presence of other aggressive prolactinoma cases which responded to temozolomide
treatment and indicated that MGMT expression status is important in response to
temozolomide like the situation encountered for high grade glial tumors [8]. In our case, which will be discussed
below, MGMT expression was encountered which may associate with temozolomide
resistance. Tang et al. reported a prolactinoma patient who was refractory to
cabergoline treatment even at high doses, exerting a continuous enhancement in
both the prolactin levels and the tumor volume [15]. The patient was treated with two consecutive transsphenoidal
surgeries and the pathological examination revealed that the Ki67 index
increased from 3% to 30%, and the expression levels of DRD2 and
MGMT were low. The increase of Ki67 index is exactly similar to what we have
observed in our index case discussed below. Following six cycles of temozolomide
chemotherapy, the tumor first shrank and then vanished completely. During the
6-month follow-up, the tumor did not recur, and the prolactin level did not rise
[15]. Halevy and Whitelaw reported
that temozolomide might be a suitable option in aggressive pituitary adenomas
and carcinomas [16]. They reviewed the
published case series and concluded that 42% of patient responded on
radiographs, and 27% of patients stabilized succeeding temozolomide.
Prolactinomas and corticotroph adenomas responded to temozolomide with an
approximately a 50% response rate, but non-functioning tumors responded
only half as frequently [16].
Strowd et al. reported a prolactinoma case, which showed clinical and
radiographic progression, despite treatment with bromocriptine, transphenoidal
surgical resection, radiation therapy, and cabergoline [1]. Temozolomide treatment was began 6
years after diagnosis and after three cycles of treatment, dramatic radiological
and clinical responses were witnessed with 99.3% reduction in prolactin
levels [1]. After 3 years of follow-up,
the patient again developed radiological progression and increase of prolactin
levels. The patient was rechallenged with temozolomide, and following four
cycles, radiological, hormonal, and clinical responses were observed with a
92.2% decline in prolactin levels [1]. Development of temozolomide resistance was defined during
transition of an atypical prolactinoma to a prolactin-producing pituitary
carcinoma, which was associated with loss of DNA mismatch repair protein MSH6 in
carcinoma [17]. European Society of
Endocrinology Clinical Practice Guidelines advised the employment of
temozolomide as first-line chemotherapy for aggressive pituitary adenomas and
carcinomas in 2017 [18]. However, it is
also possible that pituitary adenomas develop acquired temozolomide resistance
following initial responses to temozolomide treatment [18]. We witnessed a similar situation in
our index case which will be mentioned below.
Peptide Receptor Radionuclide Therapy
Little data exists on peptide receptor radionuclide therapy (PRRT) for the
management of aggressive pituitary tumors [19]. Giuffrida et al. analyzed the safety, efficacy and long-term
outcome of PRRT in three patients with aggressive pituitary tumors and also
reviewed the available literature [19].
First patient (female, giant prolactinoma) was treated with five cycles of
111In-DTPA (diethylenetriamine pentaacetate)octreotide (total
dose 37 GBq) for 23 months, following inefficient surgery and long-term
treatment with dopamine-agonist [19].
Second patient (male, giant prolactinoma) was treated with two cycles of
177Lu-DOTATOC
(DOTA0-Phe1-Tyr3)octreotide (12.6 GBq)
following multiple surgeries, radiosurgery and temozolomide [19]. In the third patient (female,
non-functioning pituitary tumor), five cycles of 177Lu-DOTATOC
(29.8 GBq) was applied after five surgeries, radiation treatment and
temozolomide. First patients tumor shrank and neurological and visual
amelioration was witnessed over 8-year follow-up, while the other two pituitary
tumors continuously grew causing amaurosis and neuro-cognitive disorders [19]. PRRT was not associated with any
adverse systemic effects. The investigators found eleven other cases of
PRRT-treated aggressive pituitary tumors from the literature. When they included
the patients from the literature, 4 of 13 patients exerted tumor shrinkage and
biochemical or clinical amelioration of symptoms after PRRT. Responses did not
associate with age or gender, neither with the employed
peptide/radionuclide, but PRRT failure was associated with failure of
previous temozolomide chemotherapy [19].
Adverse effects were noted only in two patients. The authors concluded that PRRT
is a safe option following failure of multimodal treatment [19].
Molecular Pathogenesis of Prolactinomas Important to Design Future
Treatments
Molecular Pathogenesis of Prolactinomas Important to Design Future
Treatments
Estrogen
The major regulators of lactotroph functions are estradiol and dopamine (DA)
which interact in controlling cell proliferation and prolactin secretion [20]. In prolactinomas, the main related
transcription factors are estrogen receptor-α (ERα) and
pituitary transcription factor 1 (PIT1) [21]. Estradiol exerts its cellular functions via specific nuclear
receptors, ERα and ERβ, via genomic and non-genomic pathways
[21]. ERα66 is the most
commonly encountered type of Erα and a variant of ERα, known as
ERα36, is formed from the promoter residing within the first intron of
ERα66. Unlike ERα66, ERα36 does not harbor AF1 and AF2
transcriptional activation domains for binding to estradiol and coactivators
[21]. Dimerization and DNA-binding
domains and a certain part of the ligand-binding domain still exists in
ERα36. In opposite to ERα66, ERα36 mainly localizes in
both the cytoplasm and plasma membrane and exerts nongenomic estrogen signaling
[21]. Mahboobifard examined the
expression of Ki67, p53, ERα36 and ERα66 by immunohistochemistry
in 62 patients with prolactinoma patients and in normal pituitaries [21]. A salient expression of ERα36
was determined in normal pituitaries. The median scores of ERα66 and
ERα36 expression were 6 and 8 in normal pituitaries and 0 and 4 in
tumors, respectively. Low expression of ERα36 was associated with higher
Ki67 indices and more prominent tumor invasion [21]. Low ERα66 expression was also associated with tumor
invasion, increased tumor volumes and dopamine-agonist resistance. After
controlling for sex, the low ERα36/low ERα66 phenotype
was 6.24 times more prevalent in invasive prolactinomas than in noninvasive
prolactinomas [21]. The authors have
underlined that these associations are relevant mostly for macroprolactinomas
and could differ for microprolactinomas. They attributed the downregulation of
estrogen receptors (ERs) to two possible and opposite mechanisms. At first,
increased estrogen activity may subsequently downregulate ERs; second, decreased
ERs may associate with reduced estrogen-induced apoptosis in lactotroph cells
[21]. Li et al. defined a somatic
mutation in SF3B1R625H (splicing factor 3 subunit B1) in about
20% of prolactinomas which associates with higher prolactin levels and
shorter progression free survival [22].
Importantly, SF3B1R625H mutation leads to erroneous splicing of
estrogen related receptor gamma (ESRRG), a steroid hormone receptor which binds
to tamoxifen metabolite 4-hydroxytamoxifen (an estrogen antagonist) and
diethylstilbestrol (nonsteroidal estrogen) [22]. It remains to be elucidated whether estrogenic pathways
associate with perturbed signaling cascades induced by SF3B1 mutations.
Dopamine inhibits proliferation of lactotroph cells, prolactin synthesis and
release, acting via the D2R expressed in lactotroph cells [20]. Estrogen hormones may contribute to
pathogenesis of prolactinomas, as these tumors grow faster during pregnancy and
develop in transsexual men under estrogen treatment [23]
[24]. Estradiol and estrogens in general decrease the effects of
dopamine agonists, induce prolactin gene transcription and indirectly lower
dopamine synthesis from the hypothalamus [6]
[20]
[25]. Estrogens also directly stimulate
mitotic activity, suppress lactotroph cell apoptosis and modulate the blockage
of dopamine on prolactin gene transcription via a decrease of D2Rs on the
lactotroph cell membrane [6]. In mouse
models, estrogen interacts with bone morphogenic protein 4 (BMP-4) and Smad4 to
trigger enhanced cell growth [5]. In
parallel, high levels of estrogens during gestation cause lactotroph
hyperplasia, hyperprolactinemia, and growth of tumors, nonetheless; usually in
the lack of simultaneous treatment with dopamine agonists [6]. In cell culture models of
prolactinomas, 17β-estradiol induces calbindin-D9k (CaBP-9k) expression,
proliferation, and apoptosis inhibition via interacting with ERα which
can be inhibited with ERα inhibitor AZD9496 [26]. Nonetheless, it shall be also noted
that prolactinomas in men exert lower estrogen receptor alpha (ERα)
expression related to treatment resistance, higher tumor grades and worse
prognosis [27]. Moreover, even in
prolactinomas induced by estrogen treatment, pharmacological high-dose levels of
estrogen can induce tumor regression, suggesting a dichotomic effect of
estrogens in prolactinoma growth [4].
Increased aromatase expression (which catalyzes synthesis of estrogen from
testosterone) was revealed in invasive prolactinomas in post-menopausal women,
in comparison to its expression in noninvasive prolactinomas [28]. ER-β level was also
significantly higher in patients resistant to bromocriptine [28]. For medically refractory
prolactinomas, lowering endogenous estrogen with aromatase inhibitors or
employment of selective estrogen receptor modulators (SERMs) are treatment
candidates [6]. Indeed, there exist some
anecdotal reports of prolactin lowering with aromatase inhibitors anastrozole
and letrozole [27]. In experimental
models, tamoxifen hinders the estrogen-stimulation of prolactin secretion and
blocks growth of prolactinoma in vitro and in vivo with minimal interaction with
bromocriptine. Moreover, few patients with
‘‘bromocriptine-resistant’’ invasive
macroprolactinomas respond with reductions in tumor volumes and prolactin levels
with tamoxifen treatment [6]. Raloxifene,
a SERM employed to treat osteoporosis, decreased prolactin levels significantly
in comparison to placebo in a pilot study conducted on healthy postmenopausal
women. Hence, reducing estrogen signaling with such agents may be employed in
future [6]. In another study, raloxifene
decreased prolactin levels with a mean percentage of 25.9% (8 –
55%) in 10/14 (71%) patients with prolactinoma who were
treated with stable doses of dopamine agonists including 2 cases (14%)
who achieved normoprolactinemia [29].
Another evidence indicating the importance of estrogenic pathways comes from the
studies which employed fulvestrant in treatment of prolactinomas in rats.
Fulvestrant is a selective estrogen receptor degrader (SERD) which binds and
destabilizes the estrogen receptor, leading the cell's inherent protein
degradation cascades to degrade it. Cao et al. established prolactinomas in rats
with estrogen and when they treated these rats with fulvestrant, they revealed
that tumor volumes, weight and blood prolactin levels were substantially reduced
in time- and dose-dependent manners [30].
Fulvestrant also downregulates Pyruvate Kinase-M2 and inhibits glycolytic energy
production in prolactinoma cells and induces their apoptosis in accompaniment
with reductions in XBP1, IRE1 and GRP78 levels [31]
[32]. There also exists
evidence that estrogen and mTOR pathways may synergize in development of
prolactinomas. In rats, estrogen treatment induced prolactinomas with
concomitant increase of mammalian target of rapamycin (mTOR) signaling, which
was blocked by mTOR inhibitor rapamycin [33]. Pituitary knockout of either mTOR negative regulators Tsc1 or
PTEN also led development of prolactinomas, which were blocked by rapamycin
[33]. Hence, it would not be illogical
to presume that estrogen antagonists and mTOR inhibitors may act synergistic in
inhibition of prolactinoma growth. However, it shall be also noted that clinical
data regarding employment of estrogen modulators in treatment of aggressive
prolactinomas is limited [27].
Intersecting Cascades of TGFβ, Progesterone, and Chorionic
Gonadotrophin
Intersecting Cascades of TGFβ, Progesterone, and Chorionic
Gonadotrophin
TGFβ1 inhibits proliferation of lactotroph cells and secretion of prolactin,
and it partly mediates the inhibitory action of dopamine [34]. TGFβ1 is secreted to the
extracellular millieu as an inert complex, and its bioavailability is tightly
regulated by diverse elements of the TGF-β1 system which involve latent
binding proteins (LTBPs), local activating factors (matrix metalloproteases,
integrins, Thrombospondin and others), and TGFβ receptors [34]. Activity of pituitary TGFβ1 and
the expression of varying parts of the TGFβ1 cascades, are controlled by
estradiol and dopamine. According to some investigators, prolactinomas (both in
animals and humans) harbor lower TGFβ1 activity as well as decreased
expression of diverse constituents of the TGFβ1 system [19]. Hence, it is assumed that the reactivation
of TGFβ1 inhibitory potential would provide a new therapeutic approach to
bypass medication resistance in prolactinomas [34]. On the other hand, there also exists data showing higher levels
TGF-β1/Smad3 signaling pathway-related proteins in dopamine
agonist-resistant prolactinoma specimens exerting high fibrosis; and the reversal of
fibrotic and drug resistance pathways with TGFβ1/Smad3 signaling
inhibitor SB431542 [35]. Female mice
transgenically overexpressing the human chorionic gonadotrophin β subunit
(hCGβ+) develop prolactinomas, whereas hCGβ+male
mice do not [20]. Faraoni et al. revealed
lower TGF-β1 levels, lower expression of TGFβ1 receptors and its
target genes including Ltbp1, Smad4, and Smad7 in
hCGβ+female pituitary tissues [20]. Nonetheless, no differences were detected between the wild-type and
transgenic male pituitaries, which were attributed to the fact that the lowered
pituitary TGF-β1 activity in hCGβ+females cause the
development of prolactinomas. Indeed, they also showed that an in vivo treatment
employed to increase pituitary TGFβ1 activity succeeded in decreasing the
development of prolactinomas and hyperprolactinemia in hCGβ+females
[20]. Another finding was that the high
amounts of hCG in circulation triggered luteinization in the ovary of
β-hCG+females, and progesterone became the main steroid synthesized,
but estradiol levels remained at normal ranges [20]. These findings concurred to the previous observations of Ahtiainen
et al. [25]. The authors analyzed the
endocrine pathogenesis of prolactinomas in female transgenic mice expressing the
β-hCG. The LH/CG levels were increased in the mice, with subsequent
stimulation of progesterone synthesis in ovaries [25]. Despite normal levels of estrogen, these mice developed large
prolactinomas and progesterone involvement in prolactinoma pathogenesis was shown
with several lines of evidence. The progesterone-antagonist mifepristone blocked
prolactinoma growth and postgonadectomy estradiol + progesterone
combined treatment was more potent than single estrogen in stimulation of
prolactinoma tumorigenesis [25]. Although
estrogen was not increased in hCGβ-transgenic mice model, these findings
highly suggest that both female hormones may accelerate prolactinoma tumorigenesis.
Further evidence for direct growth-stimulatory effect of progesterone was
encountered in primary mouse pituitary cell and rat somatomammotroph GH3 cell
cultures [25]. In cultured cells and tumors of
the mice, progesterone stimulated the cyclin D1/cyclin-dependent
kinase-4/retinoblastoma protein/transcription factor E2F1 cascades
[25].
Janus Faces of Progesterone in Prolactinoma Growth. Role of Membrane
Progesterone Receptors (mPRs)
Above, we indicated the studies which showed progesterone stimulation of
prolactinoma growth. Nonetheless, there also exists paradoxical opposite data.
Membrane Progesterone Receptor-α (mPRα) is highly expressed in
the rat hypophysis and primarily in lactotroph cells, mediating
progesterone’s inhibitory actions on secretion of prolactin [36]. Importantly, expression of nuclear PRs
and mPRs is significantly lesser in tumoral pituitary tissues in comparison to
benign ones and the relative proportion of mPRα and mPRβ is
significantly higher in prolactinomas. A selective mPR agonist (Org OD 02-0)
significantly blocked prolactinoma release in both tumoral and normal hypophysis
explants, exerting a more salient efficacy in tumoral pituitary tissues [36]. As progesterone also controls
prolactinoma secretion indirectly via dopaminergic neurons, mPR involvement in
this action was also studied. Noteworthy, the hypothalamus highly expresses mPRs
and both OrgOD 02-0 and progesterone enhanced dopamine release in hypothalamus
explants. Moreover, in an in vivo experiment, the mPR agonist robustly lowered
the hyperprolactinemia in transgenic females harboring prolactinoma [36]. Therefore, progesterone may exert
dichotomic effects on prolactinoma growth. Hence, if progesterone antagonists or
agonists will be employed in treatment of aggressive or malignant prolactinomas,
each of these agents shall be tested in primary cultures of aggressive
prolactinoma tissues, whether they exert risks of tumor propagation or whether
they potently inhibit tumor growth.
GnRH (Gonadotropin Releasing Hormone)-Signaling Pathway
Overexpression of proto‑oncogenes encoding proteins driving cell
cycle‑progression, growth factors or receptors were detected in prolactinomas
including high mobility group A2 gene and FGF receptor‑4 [5]. Zhao et al. analyzed gene expression
profiling to determine differentially expressed genes (DEGs) between
prolactinoma (n = 4) and normal (n = 3) samples
[2]. They revealed that the DEGs were
enriched in 15 Gene Ontology (GO) categories in which the GO category
“developmental process” ranked first and “system
development” ranked second. Two pathways including gonadotropin
‑releasing hormone (GnRH)‑signaling pathway and neuroactive ligand‑receptor
interaction was determined. β-LHB (luteinizing hormone β
subunit) and β-FSH (follicle stimulating hormone β subunit) were
downregulated in prolactinoma, which both involve in the GnRH (Gonadotropin
Releasing Hormone)-signaling pathway [2].
GnRH is a tropic hormone produced and released from GnRH neurons within the
hypothalamus which is responsible for the release of luteinizing hormone (LH)
and follicle-stimulating hormone (FSH).
Somatostatin
Immunohistochemical analysis of somatostatin receptors (SSTR) demonstrated that
all SSTR types exist in prolactinomas; SSTR5 were mostly frequent, followed by
SSTR2A and SSTR1
[27]. The somatostatin receptor subtype 5 (SSTR5) is the most
regulator subtype in modifying prolactin secretion and among the FDA-approved
somatostatin analogs, only pasireotide has prominent activity at SSTR5 [6]. In cell culture, three
medication-resistant prolactinomas responded to pasireotide; and hence,
pasireotide application might be tested in a patient failing all other medical
modalities. There also exist anecdotal case reports demonstrating the efficacy
of pasireotide to achieve reductions of tumor volumes and prolactin levels in
resistant prolactinomas, sometimes with dramatic responses [27]
[37]
[38].
Epidermal Growth Factor Receptor (EGFR) Dependent Cascades
The epidermal growth factor receptor (EGFR) family comprises transmembrane
tyrosine kinase receptors including EGFR (ErbB1, HER1), p185ErbB2/neu
(ErbB2, HER2), ErbB3 (HER3), and ErbB4 (HER4) [39]. Ligand binding triggers the assembly of receptor homo- and
heterodimers, activation of the intrinsic kinase domain and subsequent
intracellular signaling [39]. Enhanced
expression of ErbB receptors has been demonstrated in aggressive pituitary
tumors and carcinomas. EGFR, ErbB receptors, p185her2/neu, ErbB3, and
ErbB4 were shown to associate with tumor progression and an enhanced
de-differentiated state in prolactinomas. ErbB receptors and ligands are also
synthesized by nontumoral lactotrophs and stimulate prolactin secretion [39]. Mixed lacto-somatotroph tumors express
ErbB receptors and ligands, and interfering with these pathways modifies
prolactin secretion, and tumor size [39].
Lactotroph tumor cascades blocked by tyrosine kinase inhibitors (TKI) provide
evidence that ErbB receptors stimulate prolactin secretion and growth of
lactotroph cells [39]. Gefitinib binds the
ATP-binding site of EGFR leading selective blockage of the EGFR activity,
lapatinib binds the ATP-binding of the EGFR/HER2 protein kinase domain
and hinders their activation and canertinib is a pan ErbB-inhibitor [39]. Gefitinib reduced proliferation of rat
GH3 somato-lactotroph cells and prolactin mRNA expression in vitro and xenograft
tumor volume and prolactin release in vivo [40]. Importantly, lapatinib reduced prolactin mRNA and protein
secretion from human prolactinoma cells in vitro. Cooper et al. comprehensively
assessed expression of ErB receptors in prolactinomas and evaluated their
association with their clinical features [39]. Expression of EGFR was detected in 82% of adenomas,
ErbB2 in 92%, ErbB3 in 25%, and ErbB4 in 71% [39]. Enhanced ErbB3 expression associated
with optic chiasm compression, suprasellar extension and carotid artery
encasement. Yet, medication-response rates were significantly higher in tumors
with higher expression of ErbB3. Cooper et al. treated two subjects with
aggressive resistant prolactinoma with lapatinib 1250 mg daily for 6
months [39]. Tumor sizes and prolactin
levels were reduced with lapatinib treatment and the authors also suggested the
possibility that D2R-targeting classical medications may synergy with lapatinib
[39]. In 2021, the same group
published their experience on lapatinib, which was applied to 4 patients with
drug resistant prolactinoma; they witnessed that 3 patients had disease
stabilization, with 2 exerting a 6% enhancement and 1 exerting a
16.8% reduction in tumor diameter [41].
PRDM2 (PR/SET Domain 2 / RIZ1-Retinoblastoma Interacting Zinc
Finger protein-1)
PRDM2 is a tumor suppressor gene which is a member of the nuclear histone
methyltransferase superfamily and encodes a zinc finger protein, which binds to
ER, retinoblastoma protein, and the TPA-responsive element of the
heme-oxygenase-1 gene [42]. Gao et al.
analyzed the exomes of six drug-responsive prolactinomas and six drug-resistant
prolactinomas by whole-exome sequencing [23]. They identified ten somatic variants that regulated metabolic
cascades and DNA repair including PRDM2. Quantitative gene expression analysis
with RT-qPCR revealed that PRDM2 mRNA levels were about five-fold lower in
drug-refractory prolactinomas [43]. More
importantly, restoration of PRDM2 expression increased D2DR levels, showed a
synergistic action with bromocriptine to reduce prolactin secretion and MMQ
prolactinoma cell line viability [43].
Hence, PRDM2 seems like an interesting candidate in prolactinoma treatment by
being a member of the main gene-gene interaction hubs including ER.
miRNA Based Analyses Reveal the Potential for Drug Repurposal for Aggressive
Prolactinomas
Aydin et al. analyzed the transcriptomic features of prolactinoma through mRNA
and miRNA level data integration and repurposed novel drugs based on this
integration [44]. They repurposed 7 drugs
including 5-flourocytosine (an antifungal agent), nortriptyline (an
antidepressant), neratinib (an antineoplastic used for breast cancer), puromycin
(an aminonucleoside antibiotic), taxifolin (an anticancer flavonolol),
vorinostat (an antineoplastic histone deacetylase inhibitor), and zileuton (an
anti-asthmatic 5-lipoxygenase inhibitor) for the treatment of resistant
prolactinoma [44]. They also analyzed
effects of these drugs MMQ cell vitality via investigating
PI3-Kinase/Akt signaling pathway and arrest of cell cycle via western
blotting and flow cytometry [44].
Other Treatment Possibilities
There exist anecdotal reports that metformin, an oral antidiabetic could
normalize prolactin levels when it was used in combination with bromocriptine in
bromocriptine-resistant prolactinomas [3].
There also exist in vitro and in vivo evidence that metformin – in
combination with brocriptine – inhibits growth of prolactinomas [3]. mTOR signaling pathway is a promotor of
tumor formation in pituitary tumors, more specifically in prolactinomas [45]. In vitro studies revealed that mTOR
inhibitor rapamycin, could act effective in the management of prolactinomas
[46]. There also exist clinical
reports that mTOR/akt pathway inhibitor everolimus could normalize
prolactin levels and reduce tumor volumes in dopamine agonist-resistant
prolactinoma, which harbors high levels of p-AKT, p70S6K and p4EBP1 [47]. Everolimus can also inhibit growth of
prolactinoma cells carrying prolactin receptor variants which overactivate akt
pathway [48].
Clinical and Radiological Description of the Index Case
[Figures 1]
[2], and [
3] represent pathological features of the case and [Fig. 4] represents the radiological
features of the case. A 51-years-old Caucasian female patient suffering from
amenorrhea and galactorrhea was found to have macroprolactinemia in 2013
following examinations in an external center. Neurological examination was
normal in her admission. Patient’s laboratory and radiological
examinations revealed prolactin levels above 200 ng/ml and the
presence of a central/right paracentral-localizing pituitary adenoma
with a size of 21 × 12 mm. Cabergoline treatment was
initiated. During a 5 months follow-up there was a progressive increase in
prolactin levels and cabergoline dose was substantially increased from
1 mg/week to 3.5 mg/week. Despite this
treatment, prolactin levels elevated up to 5061 ng/ml and
radiological investigations revealed an increased mass lesion reaching
3×2×4.5 cm. At this stage, left eye ptosis developed due
to involvement of the CIII. In the January of 2014, a transsphenoidal pituitary
adenomectomy was performed in an external center due to resistance to medical
treatment and only 50% of the lesion could be surgically removed.
Pathological examination revealed a prolactinoma (lactotroph adenoma) with a
Ki67 labeling index of 5–6% and p53 staining ratio of 3%
and no signs of anaplasia. Cabergoline treatment was continued. In May of 2014,
gamma-knife radiosurgery was applied to the residual lesion at a dose of
19 Gy in one fraction.
Fig. 1 Pathological features of the aggressive prolactinoma.
a: The normal acinar structure of the pituitary gland is
distorted and neoplastic cells reveal papillary configuration with
prominent atypia, pleomorphism, and distinct nucleoli (100×).
b: Neoplastic cells show a heterogeneous staining pattern
with prolactin antibody (200×). Other hormonal
immunohistochemical stains were negative. c: The Ki67 staining
rate was as high as 25% (200×). d: Some of the
cells reveal faint p53 positivity (200×).
Fig. 2 Pathological features of the aggressive prolactinoma.
a: The tumor is synaptophysin positive; suggesting a
pituitary adenoma (100×). b: The tumor shows cytokeratin
positivity (100×). c: There was partial expression of
MGMT (200×).
Fig. 4 Radiological features of the aggressive
prolactinoma.a: T1-weighted axial view of the MRI performed
in August of 2016. Hyperintense mass lesion with high heterogeneous
contrast uptake filling the cella at the right paramedian area,
embracing the right cavernous sinus, extending to the prepontine cistern
at the medial of the 5th cranial nerve in the posterior intracranial
area.b: In T1-weighted coronal sections, extension of the
lesion to the sphenoclival area and extension of the mass lesion to the
base of the 3th ventricle and sphenoclival area is observed.c:
T1-weighted axial and d: coronal view of the postoperative MRI
performed in September of 2016 presented the internal decompression of
the mass lesion. e: T1-weighted axial view of the MRI performed
in February of 2019. Progression of the lesion with intense contrast
uptake invading cavernous sinus, prepontine area and pontocerebellar
edge. f: T1-weighted coronal view of the MRI performed in
February of 2019. A very extensive invasion of the lesion was observed
involving the cavernous sinus, prepontine and petroclival areas, 5th,
7th, and 8th cranial nerves with intense contrast uptake. g:
T1-weighted axial view of the MRI performed in July 2020. Regression of
the mass lesion in prepontine and pontocerebellar edge. h:
T1-weighted coronal view in July 2020. Regression of the mass lesion in
prepontine and petroclival areas.
Following these treatments, signs of the CIII involvement alleviated. In October
of 2014, prolactin levels declined to 400 ng/dl. In her MR scans
obtained in January and April of 2015, a regressing mass was observed when
compared to initial lesion detected one year ago. Thereafter, the prolactin
levels began to rise and cabergoline dose was gradually increased up to
8 mg/week. In June of 2015, the patient developed a traumatic
intracranial hemorrhage following a fall at home. The patient was followed up
conservatively. Yet the prolactin levels progressively increased to
2634 ng/dl, 3267 ng/dl and 6599 ng/dl. At this stage,
cabergoline treatment was stopped and bromocriptin treatment was began, but the
patient did not tolerate bromocriptine and cabergoline treatment was restarted.
Cabergoline treatment was increased to 10 mg/week. Due to the
aggressiveness of the lesion, no surgical options were advised by consulted
centers. The patients vision deteriorated because of the right eye
ophthalmoplegia due to tumor expansion, that was accompanied by high prolactin
levels despite medical treatment. At this stage, the patient admitted to our
neurosurgical clinic and a second transsphenoidal pituitary adenoma resection
was performed in September 2016 by our neurosurgical team. Pathology revealed a
sparsely granulated prolactinoma with a Ki-67 labeling index of 26%, p53
level of 2% and positive MGMT immunoreactivity. Postoperative prolactin
level was 2541 ng/ml. Gamma-knife radiosurgery was applied to the right
side at a dose level of 15 Gy in one fraction in October 2016.
The patient’s general condition was moderate and kept under cabergoline
treatment until 2018 till gait problems and blurred vision appeared. Prolactin
levels progressively increased during this period, and 5 cycles of temozolomide
treatment were applied which started on January of 2018. Until May of 2018, the
lesion was stable and prolactin levels did not rise under temozolomide
treatment. However, massive progression was detected in cranial MRI examination
performed in August 2018. The patient’s general condition deteriorated.
Difficulty in swallowing, intense pain during chewing, tingling in tongue,
difficulty in speech and enhanced balance disorder developed. MRI revealed a
residual mass lesion of 2 cm size on the right side of pituitary
invading the cavernous sinus and the ICA (internal carotid artery). It was
accepted as a de novo macroadenoma of 3 cm in size at the clivus
extending to prepontine region and accompanied with another mass lesion of
11 cm size on the left ICA surrounding the cavernous sinus. The patient
was discussed in a special tumor council for alternative treatment
modalities.
Following explaining the possible risks to the patient and obtaining her consent,
VMAT (Volumetric Arc Therapy) was applied to the tumor lodge at a level of 45
Gray at 25 fractions in September of 2018. The prolactin level declined from 14
000 ng/dl to 4800 ng/dl following radiotherapy. MRI revealed
regression of total tumor mass two months after VMAT and the patient was
followed under cabergoline treatment at a dose of 6 mg/week. In January
of 2019, the prolactin level rose to 1 605 671 ng/dl, MRI revealed
reprogression of old lesions besides development of novel lesions. A PET-CT
analysis did not demonstrate systemic metastasis. The patient’s
treatment was planned as 30 mg cabergoline per 28 days by the medical
oncology department, her general neurological condition improved, and the pain
level decreased in February of 2019 following this treatment. In June of 2019,
octreotide treatment was started due to minimal progression of the lesions in MR
investigations. She was again referred to radiation oncology department due to
prominent progression in February of 2020 and she received another series of
Cyberknife stereotactic radiosurgical treatment (3500 cGy/10 fractions)
to the tumor lodge which progressed. Regression of the mass lesion in
prepontine, pontocerebellar edge and petroclival areas was observed ([Fig. 4]). The patient is still under
follow-up by neurosurgery and medical oncology departments.
Pathological Features of the Prolactinoma
Pathological Features of the Prolactinoma
In the specimen obtained from the second operation in September 2016, the following
features are observed, that is, normal acinar structure of pituitary gland is
distorted and neoplastic cells reveal papillary configuration with prominent
cellular atypia, pleomorphism and distinct nucleoli (400×) ([Fig. 1a]). Synaptophysin [Biocare (27612)] and
pancytokeratin [Scytek (5d3lp34)] stainings were positive ([Fig. 2a, b], respectively). All neuroendocrine
hormonal immunoassays (GH, ACTH, LH/FSH, TSH) were negative but neoplastic
cells show disperse staining pattern with prolactin antibody [Genetex (b109.1)]
(200×) depicting a sparsely granulated prolactinoma ([Fig. 1b]). Ki67 [DAKO (MIB-1)] labeling index
was prominently high as 25% (200×) ([Fig. 1c]). Some of the cells showed faint
reactivity to p53 [Scytek (do/7)] positivity (200×) ([Fig. 1d]). Disperse MGMT [Novus (mt 23,2)]
immunoreactivity was detected ([Fig. 2c]).
Estrogen receptor [ER (Leica 6F11)] was faintly positive in some cells ([Fig. 3a]), while Progesteron receptor (PR) was
negative ([Fig. 3b]). c-ErB-B2 [Fig. 3b] (Ventana Her-2/neu 4B5)] and
human-Chorionic Gonadotropin [h-CG (Biocare)] ([Fig. 3d]) stains were also negative. Temporal changes in the radiological
characteristics of the tumor are shown in [Fig.
4] and its respective features are described in the associated figure
legend.
Fig. 3 Pathological features of the aggressive prolactinoma. a:
The tumor was faintly positive for ER (400×). b: The tumor
was negative for PR (200×). c: The tumor was negative for
ErbB2 (200×). d: The tumor was negative for β-HCG
(200×).
Dramatically, Ki67 indices of the tumor were calculated as 5%-6% and
25% in consecutive specimens obtained from surgeries performed in January of
2014 and in September 2016, respectively. Hence, a transformation involving loss of
cell cycle checkpoint and/or DNA repair genes may contribute to this bizarre
phenomenon and such a high level of Ki67 may inherently associate with the
aggressive tumor biology in this index case considering that even the Ki67 threshold
for pituitary carcinomas is 11% [4].
In our case, the recurred tumor did not show an extensive p53 expression. In our
current case, more than 10% of tumor cells stained with MGMT. As MGMT
involves in the repair of temozolomide-induced DNA damage, the temozolomide
resistance in our case may associate with the positive selection of MGMT-expressing
tumor cells in a time dependent manner. The tumor stained very faintly with ER in
very few cells; thus, an estrogen-antagonist treatment was not considered. The tumor
did not stain either with b-HCG or PR; hence, a progesterone-antagonist (such as
mifepristone) was not employed for treatment.
Conclusions
Despite most of the prolactinomas are typically benign, some of these tumors may
follow a poor course like a malignant tumor. Prominent cellular atypia,
pleomorphism, and distinct nucleoli may be encountered in aggressive prolactinomas,
whereas even metastatic malignant prolactinomas may have a paradoxically benign
appearance in histopathologic examination. Both aggressive, yet still histologically
benign prolactinomas and malignant prolactinomas are challenging pathologies not
only for clinicians but also for pathologists. They must be managed by a
professional and specified team consisted of neurosurgeons,
medical/radiation oncologists, and endocrinologists. It would not be wrong
to envisage that future teams dealing with these tumors would also include molecular
pathologists and clinical geneticist. Only after such an extensive collaboration,
patient-tailored novel treatments targeting aggressive tumors may provide remissions
and even cures in these patients.
Compliance with Ethical Standards
Compliance with Ethical Standards
Ethical Approval
For this study, only retrospective analysis of the already present pathological
specimens and radiological imaging pictures were evaluated and no further
laboratorial tests or invasive procedures were performed for research purposes.
The patient’s identity was not disclosed. Under these circumstances, the
local ethical committee did not request an ethical approval process. The patient
and a witness signed an informed consent form approving the report of their
pathological, radiological, and clinical condition under these conditions.
