Abstract
Background/Aim: The histopathological variability of each type of pituitary adenoma (PA) that causes growth hormone (GH) excess influences the phenotype, radiological characteristics and therapy response of acromegaly patients. We correlated the immunohistochemical (IHC) features of GH-secreting PAs with their clinical, laboratory and imaging data. Patients and Methods: We included 32 patients with documented acromegaly; tumour specimens were histologically and IHC examined: anterior pituitary hormones, pituitary-specific transcription factor-1 (PIT-1), Ki-67 labelling index were evaluated. Results: Macroadenomas represented 93.75%. Post-surgery disease control negatively correlated with the maximum initial tumour diameter (p=0.04). Ki-67 did not predict remission. No correlation was found between GH serum levels and IHC expression (p=0.45). PIT-1 was positive in all specimens, two had a weak expression. Four were considered PIT-1 positive plurihormonal adenomas and several had unusual IHC combinations. Conclusion: PIT-1 accurately classifies GH-secreting PAs. The IHC classification as well as radiological dimensions and extent influence disease control, probably being the best prognosis factors.
Complex and heterogeneous neoplasm, pituitary adenomas (PAs) present with a varied clinical expression, associated with multiple growth, proliferative and invasive manifestations (1, 2). The clinical manifestations of PAs can be very different, with an extensive course, from ‘completely’ and clinically silent, to ‘whispering’ and functioning PAs (3). Of all PAs, somatotrophs account for 10-15% of all resected tumours (4) and represent the main cause of acromegaly (5). These tumours exhibit nuclear positivity for PIT-1 and cytoplasmic positivity for growth hormone (GH). Pituitary-specific transcription factor 1 (PIT-1) is a nuclear transcription factor involved in the normal differentiation and growth of somatotroph, mammosomatotroph and lactotroph cells, as well as in the abnormal cell proliferation of pituitary adenomas, PIT-1 protein being highly correlated with immunohistochemical (IHC) tumour staining for GH, thyroid stimulating hormone (TSH) and prolactin (PRL) (6).
With a tendency to have larger and more invasive tumours, female patients with acromegaly, especially after menopause, exhibit lower surgical remission rates. Unlike the sex of the patient, age does not seem to be a predictor of surgical remission (7, 8). Associating a good evolution of glucose metabolism, cardiovascular disease and sleep apnea, the biochemical control of acromegaly remains the strongest predictor of the patient’s prognosis and evolution (7). Producing GH and PRL, mamosomatotroph PAs are most frequent in young patients with acromegaly (6). Because of the substantial evolution in the treatment of acromegaly and its complications, the specific mortality seems to be the same as the general population in adequately treated patients (9). Transsphenoidal surgery (endoscopic or microscopic technique) is, usually, the first line treatment, followed, when the surgical outcome is not favourable, by pharmacological (10) and, in selected cases, radiotherapeutic management (the last one as third line in the treatment algorithm) (11). This management strategy has a wide spectrum of side-effects, without necessarily controlling this seeming benign tumour.
According to the most recent WHO classification of PAs (2017), GH-secreting PAs are divided in multiple histological subtypes, which determine their variability in clinical and biological behaviour: densely granulated (most common subtype, DGSA), sparsely granulated (SGSA), intermediate granulated, mammosomatotroph, mixed somatotroph-lactotroph adenomas, acidophil stem cell adenoma, plurihormonal adenomas and silent somatotroph adenomas (12, 13). The EPPG (European Pituitary Pathology Group) proposed the term of somatotroph plurihormonal PIT1-positive tumour for defining lesions with acromegaly that have variable expression of PIT-1, GH, TSH and/or PRL. In the same group, there are discordant opinions that the management of these tumours is different (2, 14). The role of pathology has been demonstrated in the prediction of clinical and biochemical outcome, so it is very important to define the exact subtype of GH-secreting PA by using together histopathology, IHC and the expression of transcription factors (TF) (15) and correlate them all with clinical, biochemical and radiological findings. The same group stated that Ki-67 index ≥3%, p53 expression, and the their five-tiered prognostic classification may be of prognostic significance (2, 14, 16). There is no sufficient evidence to consider that SGSA or silent plurihormonal PIT-1 positive PAs are more aggressive than others (14). When considering the five main PAs types, the tumour type is correlated with the prognosis, but the risk of tumour progression varies from one study to another. Therefore, the histological subtype is a questionable prognostic factor (15, 17).
In this study, we correlated the clinical and biochemical profile of acromegaly patients with the findings from histopathological and IHC staining, as well as with the expression of PIT-1 transcription factor and Ki-67 labelling index. We highlighted the importance of TF to classify PAs (in this case GH-secreting PAs), and targeted to find correlation between laboratory, imaging, and pathology as prognostic factors.
Patients and Methods
In the present study, we included 32 patients with confirmed diagnosis of acromegaly admitted at “C. I. Parhon” National Institute of Endocrinology (Pituitary and Neuroendocrine Pathology Department, Bucharest, Romania), who underwent pituitary surgery at the Neurosurgery Clinic of “Bagdasar Arseni” Emergency Clinical Hospital (Bucharest, Romania). The patient information was collected retrospectively. Using the postoperative tumour blocks, we performed morphological and IHC analysis. The study has the approval of the “C. I. Parhon” National Institute of Endocrinology Ethics Committee.
The biochemical diagnosis was made considering The Endocrine Society clinical practice guideline (11) and the national protocol for diagnosis and therapeutic management of acromegaly, as follows: raised IGF-1 (insulin-like growth factor-1) levels for age and sex and lack of suppression of GH (growth hormone) to <1 ng/ml following documented hyperglycaemia during an oral glucose loading test (OGTT). For the visualization of tumour size and appearance, we used computed tomography (CT). We also performed visual field testing and evaluation of comorbidities: arterial hypertension ([HTA], blood pressure with systolic value over 140 mmHg and/or diastolic value over 90 mmHg), acromegalic cardiomyopathy (diagnosed with echocardiography: concentric hypertrophy, mainly involving the left ventricle, with consequent diastolic and/or systolic dysfunction), glycemic metabolism [using basal serum glycemia, oral glucose tolerance test and glycated haemoglobin (HbA1c) and defined as impaired glucose tolerance (IGT) or diabetes mellitus (DM)], low bone mass [diagnosed with bone densitometry (DXA)], colonic polyps (diagnosed with colonoscopy), sleep apnea (diagnosed with polysomnography) and arthropathy (clinical signs). Patients had a complete pituitary hormone evaluation and biochemical screening for multiple endocrine neoplasia syndromes.
All patients were treated with transsphenoidal surgery. The patients who did not achieved biochemical control after surgery only received one of the following secondary treatments: repeat surgery, radiotherapy, medical or combined therapy. The criteria for postoperative biochemical remission were based on the established guidelines (11): random serum GH <2.5 ng/ml, glucose inhibition test with a nadir GH <1 ng/ml and normalized serum IGF-1for age and sex. Patients with discrepant GH and IGF-1 were considered as having active disease.
All tissue specimens were within the limits of the standard dimensions, having less than 1 cm3. The sampling was followed by all the steps of the primary processing, thus obtaining the paraffin blocks. The histopathological diagnosis was established after routine staining with haematoxylin and eosin (H&E), on 3 µm sections for each case. The quality of the specimens was verified using immunostaining with vimentin (ETU Leica, clone V9, RTU, Figure 1). Morphological staining was performed using Leica Autostainer XL (Leica Biosystem Newcastle Ltd, Balliol Business Park West, Benton Lane, New Castle Upon Tyne NE 12 EW, UK). Microscopic examination was performed with Nikon Eclipse E 600 microscope (Nikon Corporation, Tokyo, Japan).
Vascular network highlighted by IHC staining with vimentin, in normal pituitary gland, as indicated by the blue arrow (A) (×5 magnification) and in a pituitary adenoma (B) (×10 magnification), where the predominance of vessels with definitive lumen is observed (C) (×10 magnification).
After evaluation of the specimens stained with the H&E and vimentin, the cases were evaluated as hormone profile: GH, PRL, TSH, adrenocorticotropic hormone (ACTH), follicle stimulating hormone (FSH), luteinizing hormone (LH), as well as for PIT-and Ki-67 (Table I). Bond Epitope Retrieval Solution 1 and 2 with pH 6 and 9 were used for unmasking (Leica Biosystems, Newcastle Ltd, Newcastle UponTyne, UK) and 3% hydrogen peroxide was used to block endogenous peroxidase for 5 min. The next step was incubation. Secondary and tertiary antibodies were applied for 8 min each. The visualization was made using the Bond Polymer Refine Detection System. Incubation with 3.3 diamino-benzidine chromogen was 10 min. The counter-staining was performed with haematoxylin, applied for 5 min This was followed by the introduction of the sections in absolute alcohol for 5 min, then drying and clarification in benzene for the same period. Mounting has been done automatically with Entellan (Leica CV Mount), a permanent mounting medium type.
IHC staining of GH-secreting Pas.
The IHC reactions were assessed at the cellular level. The IHC expression of GH, PRL, TSH, ACTH, FSH, LH was analysed at the cytoplasmatic level and the expression of Ki-67 and PIT-1 in the nucleus. Stains for the 6 pituitary hormones were scored in a blinded fashion. The proportion score for anterior pituitary hormones was quantified according with the following criteria: score 0 (0-10% positive cells), score 1+ (10-30% positive cells), score 2+ (30-60% positive cells) and score 3+ (>60% positive cells). The intensity scores used were from 0 to 3+ (from absent to strong staining). A staining over 10% was considered positive for the purpose of interpreting the results. The nuclear positive cells for Ki-67 were quantified by bright field microscopy (magnification x200) using QuantCenter software (3DHISTECH, Budapest, Hungary). Staining quantification for PIT-1 was done by three observers under a multi-head microscope, and results were expressed as percentage and intensity of PIT-1-positive cells. The scoring criteria used was, as follows: 0 (0-10%), 1+ (10-30%), 2+ (30-60%), 3+ (60-80%), 4+ (80-100%). Staining over 10% of tumour cells was considered positive, excluding marginal areas adjacent to the adenohypophysis.
Frequencies of the categorical variables (sex, histological type, and adenoma size [macro- or microadenomas]) were shown as percentage. For numerical variables (age at diagnosis, diameter of the lesion), mean±standard deviation and median were used, except GH, IGF-1 and PRL [median (25, 75 percentiles)], Spearman’s coefficient was used to verify correlations between numerical variables. The Mann–Whitney U non-parametric test was used to compare numeric variables between the groups. The level of significance adopted for the statistical tests was 5% (p<0.05). All statistical analyses were performed using Medcalc Software Ltd, version 19.6.4 (Ostend, Belgium).
Results
The study included 32 patients with confirmed acromegaly, with a male-to-female ratio of 1:1. From all cases, 29 had GH hypersecretion, while 3 had co-secretion of GH and PRL. Demographics and clinical presentation of patients are shown in Table II. The age at diagnosis had a negative correlation with the greatest tumour diameter at diagnosis, but with no statistical significance (p=0.22). At diagnosis, the patients presented mostly with typical acral enlargement and course facial features, headache, visual field disturbance or polyarthralgia. Tinnitus, sweating, amenorrhea or galactorrhoea were less often recorded as first symptoms. Complications of GH hypersecretion were also seen at diagnosis, almost half of patients had HTA, associated with acromegalic cardiomyopathy in several cases. IGT or DM were also present at diagnosis in some patients. Low bone mass was not so frequent, yet one patient came with lumbar pain due to multiple fragility vertebral fractures as the first symptom of acromegaly.
Patient demographics and clinical characteristics.
In terms of radiological presentation, macroadenomas were diagnosed in 30 cases (93.75%), while microadenomas were found in 2 cases (6.25%) (Table III). Tumour diameter at diagnosis had a positive correlation with extrasellar extension (p=0.001), the last being also correlated with the cavernous sinus invasion (p<0.0001). The present study shows a correlation between nadir GH in OGTT and IGF-1 levels at diagnosis (p=0.02). Three to six months after surgery, GH nadir in OGTT and random GH had a reduction of 85.34% and 83.82%, respectively, compared to initial values. Postoperative IGF-1 median level had a reduction of 30.82% compared to the diagnostic level. The postoperative median value of PRL was reduced with 49.88% of the initial value. Preoperative pituitary deficiency was found in 11 patients (34.37%), with both gonadotroph and thyreotroph deficiency in 4 cases and with panhypopituitarism in only 1 case, with postoperative persistence. From all patients included in the study, only 2 had disease remission after surgery alone. Besides these cases, it was one case that gained remission after 1 year of post-surgical medical treatment with somatostatin analogue. The greatest residual tumour diameter after surgery had a 54% overall reduction of the initial dimension. The post-surgical serum level of random GH correlated with the maximum diameter of the tumour remnant after TS surgery (p=0.01). Disease control after surgery correlated negatively with the maximum tumour diameter at diagnosis (p=0.04) and no significant correlation with GH and IGF-1 pre-operatory serum levels (p=0.24 and p=0.21, respectively) was found.
Patient tumour characteristics and treatment.
H&E staining revealed that most cases were acidophils, with 2 mixed (acidophils and chromophobe) and only one chromophobe. The main architectural pattern was diffuse/solid (Figure 2, Table IV). We followed the new 2017 WHO classification to group the somatotroph PAs according to their IHC characteristics (Table V). All specimens were stained for GH. Almost all were strongly positive for GH (3+), while one had moderate (2+) and another had weak positivity (1+), the last one having a strong positivity for PIT-1. A number of 23 adenomas were positive for PRL (71.87%). The immunopositivity for TSH, ACTH, FSH and LH was 15.62%, 9.37%, 3.12% and 18.75%, respectively. From all the cases, the tumours composed exclusively of somatotroph cells (6 cases) were divided into densely and sparsely granulated, the first morphological subtype being the most common (4 cases). When the somatotroph tumour showed dominant co-immunoreactivity higher than 10% for PRL, we considered the tumour as a mixed somatotroph-lactotroph PA (14 cases, 43.75%). As well, we found weak staining for GH in one case of acromegaly, with associated positive staining for PRL and strong positive staining for PIT-1.
(A) Acidophil with papillary pattern (longitudinal-sectioned papillae), with connective tissue and centrally arranged blood vessels, surrounded by several rows of tumour cells (×20 magnification); (B) Cytoplasmic acidophilia (×60 magnification); (C) Trabecular pattern with large, chromophobic cells and central nucleus (×60 magnification); (D) Chromophobic cells (×40 magnification); (E) Mixed, with island of chromophobic cells, surrounded by eosinophilic cells (×20 magnification); (F) Alveolar pattern (×40 magnification).
H&E staining of GH-secreting Pas.
IHC Classification of somatotroph adenomas.
Tumours that showed dominant immunoreactivity for more than one pituitary hormone coming from the same pituitary transcription factor (GH, PRL, TSH) were considered as plurihormonal PIT-1 PAs (3 cases co-expressing GH, PRL ang TSH and 1 case with GH and TSH co-expression, all presenting with strong PIT-1 expression) (Figure 3). In 8 cases, the tumour specimens showed immunoreactivity for more than one pituitary hormone coming from different pituitary TF (GH/ACTH, GH/LH) and we considered these as being unusual plurihormonal PAs. We did not find any correlation between IHC tissue expression of GH and GH serum values serum levels (p=0.45), but we found a positive correlation between PRL immunostaining and its serum levels (p=0.005). In this study, PIT-1 was positive in 93.75% of cases (Figure 4), with only 2 negative cases. From the 30 PIT-1 positive acromegaly cases, 19 patients scored 4+, 7 patients scored 3+, 3 patients scored 2+, while one patient scored 1+. Two tumours showed weak/no expression of PIT-1 (0-10%), one with intense positivity for GH, and one with positivity for both GH and PRL (Figure 4). Regarding the Ki-67 labelling index, 81.25% of cases had a value <3%, only 5 PAs scoring ≥3%. No tumour had a Ki-67 index over 10%. Examples of Ki-67 nuclear immunostaining are shown in Figure 5. There was no difference in age, gender, hormonal staining (mixed tumour or pure GH), baseline GH, or IGF-1 levels between patients harbouring somatotropinomas with low or high Ki-67 labelling index (Table VI).
A case of acromegaly due to a GH-secreting PA with positive IHC staining for GH (A), PRL (B) and TSH (C), with intense positivity for PIT-1 (D).
An acromegaly case with intense GH immunostaining (A), moderate PRL positivity (B), but negative for PIT-1 (C).
IHC expression of Ki-67 in two different cases of acromegaly: Ki-67 <3% (A) and Ki-67 >3% (B).
Patient characteristics according to Ki-67 LI. Results are presented as median (min–max).
Discussion
We examined the clinicopathological features of GH-producing pituitary adenomas from a series of acromegaly patients. All adenomas were of the PIT-1 lineage and consisted of monohormonal and plurihormonal PAs. Clinical, neuroimaging, and laboratory findings from our patients were similar to those reported in previous studies (18-21). Additionally, PIT-1 immunostaining pattern was nuclear, whereas GH, PRL, TSH, FSH, LH, ACTH immunostaining results presented diffuse cytoplasmic staining, in accordance with what has been reported by other investigators (18, 22-26).
The correct assessment of the patients for tumour secretory activity and hormonal status before and after treatment requires both GH and IGF-1 measurements (27, 28). The present study shows a correlation between nadir GH in OGTT and IGF-1 levels at diagnosis (p=0.02), yet it is stated that for the same GH levels, IGF-1 can have a wide range of values, with a non-linearity between GH and IGF-1 (27). We also tested for correlations between GH expression and the levels of its functionally associated IGF-1 and found no correlation. Although some reports state that an effective biochemical parameter used for prognosis seems to be the preoperatory GH level (rather than IGF-1), we found no significant correlation between GH and IGF-1 pre-operatory serum levels and disease control. A study that evaluated 163 acromegaly patients after trans-sphenoidal surgery did not find preoperative GH and IGF-I levels as predictors of surgical remission (29), consistent with our results. Another study that evaluated 69 acromegaly patients submitted to surgery did not find any correlation between preoperative GH and IGF-I levels and surgical remission (30), stating only the presence of cavernous sinus invasion as a preoperative characteristic for persistent disease. In our study, the maximum tumour diameter at diagnosis had a positive correlation with extrasellar extension, the last being also correlated with the cavernous sinus invasion. Important differences in extrasellar growth patterns were observed among various histological subtypes of pituitary macroadenomas and, despite smaller tumour diameters than those seen in non-functioning PAs (which exhibit more often a suprasellar extension), GH-macroadenomas demonstrate a proclivity for infrasellar extension (31).
In contrast with prolactinomas (32), clear correlations have not been established between the tumour dimension and serum GH and IGF-1 levels in the cases of GH-producing PAs. In this study, we found a positive correlation between maximum tumour diameter at diagnosis and GH nadir value in OGTT test (p=0.03), but no correlation with IGF-1 levels (p=0.98). Serum PRL level also correlated with tumour size (p=0.0008). We found that the post-surgical serum level of random GH correlated with the maximum diameter of the tumour remnant after TS surgery (p=0.01). In a study of 282 acromegaly patients, maximum tumour diameter at diagnosis was the only determinant of early biochemical remission (33). A larger tumour is associated with a lower chance of early biochemical remission, as also found in the present study, disease control after surgery having a negative correlation with the maximum tumour diameter at diagnosis.
Age at diagnosis is an important marker of tumour size and aggressiveness, with young patients tending to have larger tumours that secrete more GH and are more aggressive, unlike older patients, which may have lower GH levels at diagnosis, smaller tumours and achieve surgical cure more often (34, 35). In a predictive model of surgical remission in acromegaly which included 86 patients, the variables associated with surgical remission were older age, lower basal presurgical GH levels and lower maximum tumour diameter (36).
Almost half of our patients had cardiac involvement and more than a third had disturbance of the glucose metabolism, in the context of acromegaly. Cardiovascular disease represents the most common comorbidity in acromegaly, accounting for up to 80% of complications, followed by respiratory disorders and metabolic complications (37). Although obtaining complete hormone normalization, disease control cannot restore complications in all acromegaly patients. Surgery improves both glucose tolerance and dyslipidemia and medical therapy improves lipid metabolism and diabetes prevalence, this influencing the further chosen therapy (38). One patient in our study had vertebral fractures as diagnosis. In a study that included 31 well-controlled acromegalic patients, the progression of vertebral fractures was high, showing the deleterious effects of GH excess on bone (39).
Besides clinical symptoms, serum hormone levels and neuroimaging examinations, the postoperative pathological diagnosis of pituitary tumours is also important. Published in 2017, the fourth WHO classification of pituitary tumours introduced a more precise cell-line based classification using IHC characteristics regarding hormones and TFs expression (40), excluding the term “atypical PA” and not recommending the expression “hormone-producing PA”. PIT-1 leads the differentiation of somatotroph, lactotroph and thyreotroph cells and is preserved in tumours derived from these cell types (2, 41). In this study, PIT-1 enabled the identification of lesions associated with acromegaly. This observation is in accordance with previous findings, as PIT-1 may only be detected in specific pituitary cell types, such as somatotrophs (21, 42-44). We found two tumours with weak expression of PIT-1 (0-10%), one with intense positivity for GH, and one with positivity for both GH and PRL. In a study of 1,055 PAs, of which 316 were PIT-1 lineage tumours, there were tumours that expressed PIT-1 and no hormones of this lineage, but there were no tumours found to express GH/PRL/TSH without PIT-1 (26).
Tumours with a Ki-67 labelling index ≥3% comprised 15.62% of patients, the majority of patients having a Ki-67 under 3%, as seen in other publications for PIT-1 positive tumours (26). The variable expression of Ki-67 suggests that pituitary hormones expression affects the tumour proliferation rate, a higher Ki-67 being found in PRL-secreting PAs, mainly prolactinomas and mixed GH/PRL secreting PAs (45). We did not find any correlation between Ki-67 index and the size of the tumour at diagnosis or post-surgical control of the disease (p>0.05). Our results support the fact that Ki-67 is not a predictor of aggressive behaviour and our statement is supported by other reports which evaluated Ki-67 as a factor associated to surgical remission and did not find it as a predictor (30, 46). Instead, the tumour classification and radiological dimensions and extent can influence disease control and are probably the best prognosis factors (13, 15). As shown in other studies (47), Ki-67 index did not correlate with age, gender, hormonal staining (mixed/pure GH tumour) or GH and IGF-1 levels.
More than half (56.25%) of GH-producing PAs in the study expressed also PRL and/or TSH. Multiple hormonal staining at IHC is commonly observed in this type of tumours, about half of which are reported to contain PRL (48). We did not find any correlations between IHC tissue expression of GH and GH serum values serum levels (p=0.45), but we found a positive correlation between PRL immunostaining and its serum levels (p=0.005). This might reflect the transitory state of intracellular storage highlighted by the immunohistochemical analysis and not the exact hormonal secretory status (18, 49).
Plurihormonal adenomas are rare tumours that have unusual immunoreactivity for more than one type of pituitary hormone class that are unrelated by normal cytogenesis and development of the anterior pituitary (13). Most plurihormonal PIT-1-positive adenomas are clinically silent, although some patients may present with acromegaly, hyperprolactinemia, or hyperthyroidism. Our data show 4 PIT-1 positive plurihormonal adenomas with the clinical and biochemical picture of acromegaly. Although tumours with other combinations of pituitary hormone expression such as GH-ACTH or PRL-ACTH or ACTH-LH are very rare (50, 51), we diagnosed cases with unusual IHC combinations, which may derive from adenohypophysial stem cells (52). Some reports state the presence of high molecular weight (HMW) ACTH, a biologically inactive form in patients with acromegaly, but with no cushingoid phenotype (53). Our patients did not exhibit signs of another hormone excess rather than GH. In such cases, the correlation between the degree of PA blood vessel maturation and the hormonal status might partially explain the heterogeneity between clinical and biological features (54). Although many papers reported the absence of reticular fibres network on IHC as compulsory for differentiation between normal pituitary and PA (55), in a study including 133 PAs, 13.53% had persistence of reticular fibres, reporting the use of this technique as not being enough to characterise pituitary lesions (56). In the cases with unexpected and unusual hormone co-secretion, identifying the expression of the steroidogenic factor 1 (SF-1) and transcription factor member of the T-box family (TPIT) is useful, in order to exclude the possibility of trapped non-tumoral tissue (55) or collision tumours with different lineages (57), facts that could mislead the diagnosis.
The strength of our study includes blinded IHC review of the full panel of anterior pituitary hormones, PIT-1 transcription factor and Ki-67 labeling index in a series of patients with acromegaly due to GH-secreting pituitary adenomas. We integrated diagnostic methods of phenotype, serum hormone changes, radiology, and pathology, to enhance disease classification. In addition, we captured radiographic aggressiveness and clinical and biochemical follow-up. Although our study comprised 32 cases, our results were similar to those of reports that included larger series of patients concerning the correlations between clinical, biochemical, radiological, pathology parameters and disease control. For the cases with plurihormonality, we will analyze the other pituitary TF (SF-1 and TPIT, respectively), in order to have an accurate morphological assessment of these pituitary tumours.
Study limitations. A limitation of our study is the relatively small number of patients. For pituitary hormones and PIT-1 transcription factor, we did not use the digital image analysis, we used automated scoring only for quantification of Ki-67 labelling index. For a more precise pattern differentiation of SG versus DG tumours, antibodies directed against low molecular weight cytokeratins (LMWCK) should be used. A special staining for reticulin fibres, or the immunoreaction for collagen IV could highlight the residual reticulin fibres and, thus distinguishing plurihormonal PAs from normal pituitary cells, as well as immunostaining for SF-1 and TPIT.
Conclusion
The use of PIT-1, as recommended by the WHO 2017 classification, accurately identifies lesions associated with acromegaly, as revealed by the connection between the transcription factor staining, presurgical clinical phenotype and hormone values of all cases included in this study. More than half of the GH-producing PAs in the study expressed also PRL and/or TSH. Our data showed plurihormonal adenomas with the clinical and biochemical picture of acromegaly and we also diagnosed unusual IHC combinations, these cases being the most suitable for further dosage of SF-1 and TPIT, sustaining the idea of routine multiplex IHC for pituitary TF, along with hormones and sustaining the path for translational research.
A larger GH-secreting tumour is associated with a lower chance of early biochemical remission, as suggested by the correlation between the post-surgical serum level of random GH and the maximum diameter of the tumour remnant after TS surgery. In conclusion, the tumour classification and the dimensions and extent of the PAs influence the disease control and are probably the best prognosis factors.
Acknowledgements
The Authors would like to thank neurosurgeon Vasile Ciubotaru, who performed the tumour removal with high precision and accuracy. They would also like to thank histotechnologists Patricia Berzava and Ciprian Onica, for their technical excellence.
Footnotes
Authors’ Contributions
IFB collected the data, interpreted the data, performed the statistical analysis and wrote the paper. VNN collected the data, interpreted the data and participated to the manuscript writing. AMC, MR and VNN evaluated histopathology and immunohistochemistry of the specimens. ARC performed the immunostaining and participated to the manuscript writing. IB, CC, RD, SG, DN, SR, RT and CP provided the cases. AD provided the imaging evaluation of the tumours. LCC performed the steps of the primary processing for obtaining the paraffin blocks and the initial HP diagnosis. AMC, MR and CP conceived and coordinated the study and wrote the paper.
Conflicts of Interest
No competing interests to declare.
- Received March 7, 2021.
- Revision received March 23, 2021.
- Accepted March 30, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.










