Causes of Hyperprolactinemia in Acromegalic Patients and Clinical Correlations

Matos T; Gomes V; Faria C; Gomes A; Nobre E; Bugalho MJ

Research Article

Austin J Endocrinol Diabetes. 2021; 8(1): 1080.

Causes of Hyperprolactinemia in Acromegalic Patients and Clinical Correlations

Matos T¹, Gomes V¹, Faria C^2,3, Gomes A^1,4, Nobre E^1,4 and Bugalho MJ^1,4*

¹Endocrinology, Diabetes and Metabolism Department, North Lisbon University Hospital Center, Portugal

²Neurosurgery Department, North Lisbon University Hospital Center, Portugal

³Institute of Molecular Medicine Joao Lobo Antunes, Faculty of Medicine, University of Lisbon, Portugal

⁴Faculty of Medicine, University of Lisbon, Portugal

*Corresponding author: Bugalho MJ, Serviço de Endocrinologia, Diabetes e Metabolismo, Hospital de Santa Maria, Centro Hospitalar Universitário Lisboa Norte, Av. Prof. Egas Moniz MB, 1649-028 Lisboa, Portugal

Received: January 22, 2021; Accepted: February 26, 2021; Published: March 05, 2021

Abstract

Hyperprolactinemia in acromegalic patients may result either from cosecretion of growth hormone and prolactin by the tumour or from pituitary stalk compression. The occurrence of both conditions is possible. This study was designed aiming 1) to estimate the prevalence of each cause of hyperprolactinemia and its respective clinical course; 2) to compare the outcomes of patients with tumours staining only for growth hormone against tumours staining for both growth hormone and prolactin. 75 acromegalic patients submitted to transsphenoidal surgery between 1989 and 2018 were included. Patients were divided based on preoperative prolactin levels and immunostaining pattern. Statistical analysis was performed with SPSS version 23.

Hyperprolactinemia was documented in 22 out of 36 patients (61%). Stalk compression was the only underlying cause of hyperprolactinemia in 45% of cases. The levels of prolactin were not associated with the immunostaining pattern for prolactin. Clinical differences were not observed between hyperprolactinemic and normoprolactinemic patients, except for a higher frequency of cavernous sinus invasion (64% vs 29%, p=0,064), that reached the level of significance for the subgroup with macroadenomas staining exclusively for growth hormone (p=0,031). In the present series, no clinical differences were noticed between patients with tumours staining only for growth hormone or staining for both growth hormone and prolactin.

Hyperprolactinemia resulting from stalk compression is likely to anticipate a less favourable course of disease, since it is associated with larger tumours and a higher frequency of cavernous sinus invasion. On the contrary, positive immunostaining for prolactin was not a marker of worse prognosis.

Keywords: Acromegaly; Hyperprolactinemia; Immunostaining; Growthhormone; Prolactin

Introduction

Acromegaly is a chronic disorder resulting from excessive Growth Hormone (GH) secretion. GH hypersecretion stimulates IGF-1 overproduction which in large part mediates the somatic and metabolic effects of GH, leading to a multisystemic disease characterized by somatic overgrowth, multiple comorbidities, premature mortality, and physical disfigurement [1,2]. More than 90% of acromegaly cases are caused by a GH-secreting pituitary adenoma arising from somatotroph cells [2]. Surgery, preferentially through an endoscopic endonasal approach, is recommended as the first line therapy for most patients. However, the optimal treatment approach should be chosen depending on the size, extension of the pituitary adenoma (particularly to the cavernous sinus) and patient characteristics [1,3]. Surgical remission rates can be greater than 85% for microadenomas, however these figures can decrease to 20-30 % for macroadenomas [4-6]. For patients with disease persistence, medical therapy is often indicated with the first-generation Somatostatin Receptor Ligands (SRLs) octreotide and lanreotide. As second-line therapy, both second-generation SRLs such as pasireotide or the GH receptor antagonist pegvisomant are alternatives. Dopamine agonists have a limited efficacy and are likely to be useful in patients with modest elevations of GH and IGF1. Radiosurgery or stereotactic fractionated radiation therapy can also play a role for selected patients [3,4,7]. In a number of patients, a multimodal approach is the most adequate solution.

Hyperprolactinemia has been reported in about 30 to 40% of acromegalic patients. High levels of Prolactin (PRL) can result from co-secretion of GH and PRL by the tumour, from pituitary stalk compression or even from both conditions [8]. Indeed, about 25% somatotrophic adenomas also secrete PRL. These tumours include dimorphous adenomas (mixed somatotroph-lactotroph adenoma) and monomorphous adenomas (mammosomatotroph adenoma) [9]. The aims of the current study were: 1) to estimate the prevalence of each cause of hyperprolactinemia, among a cohort of acromegalic patients from a single center, and follow its evolution after surgery; 2) to compare the clinical presentation and course of acromegalic patients with tumours staining only for GH against tumours staining for both GH and PRL.

Materials and Methods

Study cohort

We retrospectively reviewed medical records of 75 patients with acromegaly submitted to transsphenoidal surgery between January 1989 and December 2018 (endoscopic approach, since 2011), at Hospital de Santa Maria, in Lisbon, Portugal. Inclusion criteria: pre and post-operative clinical, laboratory and imaging data as well as immunohistochemical study. Exclusion criteria: use of medications likely to induce hyperprolactinemia, radiotherapy prior to surgery and adenomas staining for hormones other than GH and PRL.

Histopathology and laboratory data

Analysis was performed based on the pathology reports included in the patients’ files. GH and IGF-1 were measured using an immunometric chemiluminescence assay. IGF1 results were interpreted according to the reference range defined for age and gender. Serum PRL was measured using an electro-chemiluminescence assay and results interpreted according to a reference range adjusted for gender. As a consequence of the retrospective nature of the study, the reference ranges have changed throughout the years. Criteria for biochemical remission were defined as a normal serum IGF-1 for age and gender, as well as a random GH below 1ng/mL [1].

Neuroradiological studies

Pituitary adenomas were classified according to the tumour diameter described on magnetic resonance imaging and/or computed tomography scan. Tumours with diameter <1 centimetre were defined as microadenomas and ≥1 centimetre as macroadenomas [10]. Additionally, cavernous sinus invasion was evaluated.

Statistical analysis

Statistical analysis was performed with SPSS version 23. Chisquare test was used for comparison of categorical variables, whereas Student’s T and Mann-Whitney tests were used for continuous variables. Differences were considered statistically significant when p-value was <0.05.

Results

Forty, out of 75 patients, met the study inclusion criteria. The ratio Female/Male was 1,9. Macroadenomas were present in 36 patients. Prior to surgery, PRL was evaluated in 36 patients. Hyperprolactinemia was documented in twenty-two patients (61%), with a median level of 40,3ng/mL. Furthermore, 10 patients (28%) despite presenting hyperprolactinemia did not disclose immunostaining for PRL. All patients in the latter group had macroadenomas. On the other hand, 6 patients, corresponding to 33, 3% of those with positive staining for PRL, had normal values of serum PRL, (Table 1).

Table 1: Immunostaining and PRL values.




  
    Hyperprolactinemia 
  
  
    
    GH 
    GH+PRL 
  
  
    Microadenoma 
    0 
    1 
  
  
    Macroadenoma 
    10 
    11 
  
  
    Normoprolactinemia 
  
  
    
    GH 
    GH+PRL 
  
  
    Microadenoma 
    1 
    1 
  
  
    Macroadenoma 
    7 
    5



Table 1: Immunostaining and PRL values.

Comparative analysis of clinical, laboratorial and imagiological characteristics of patients with hyperprolactinemia versus those with normal values of serum PRL is summarized in (Table 2). No significant differences were noticed.

Table 2: General characteristics-comparative analysis between normo and hyperprolactinemic patients.




  
    Patient    Caractheristics1 
    Normoprolactinemia    (n=14) 
    Hyperprolactinemia    (n=22) 
    p-value 
  
  
    Demographics 
    
    
    
  
  
    Age at diagnosis (years) 
    50, 6 (±12,5) 
    42, 9 (±13,3) 
    0, 074 
  
  
    Gender (Female) 
    9 (64, 3%) 
    13 (59, 1%) 
    0, 755 
  
  
    Immunostaining 
    
    
    
  
  
    GH 
    8 (57, 1%) 
    10 (45, 5%) 
    0,494 
  
  
    GH/PRL 
    6 (42, 9%) 
    12 (54, 5%) 
  
  
    Biochemical Profile 
    
    
    
  
  
    Diagnosis 
    
    
    
  
  
    PRL (ng/mL) 
    8, 1 (0, 2-15, 6) 
    40, 3 (18-195, 7) 
    0, 000 
  
  
    GH (ng/mL) 
    8, 6 (3-200) 
    25, 6 (2-200) 
    0, 120 
  
  
    IGF-1 (ng/mL) 
    957 (295-2230) 
    876 (340-1724) 
    0, 330 
  
  
    Tumor    characteristics 
    
    
    
  
  
    Microadenoma 
    2 (14, 3%) 
    1 (4, 5%) 
    0, 303 
  
  
    Macroadenoma 
    12 (85, 7%) 
    21 (95, 5%) 
  
  
    Cavernous sinus invasion 
    4 (28, 6%) 
    14 (63, 6%) 
    0, 064 
  
  
    Remission 
    
    
    
  
  
    3 months 
    4 (28, 6%) 
    1 (4, 8%) (n=20) 
    0, 049 
  
  
    1 Year 
    7 (50%) 
    10 (45, 5%) 
    0, 790 
  
  
    Radiotherapy 
    
    
    
  
  
    Yes 
    2 (14, 3%) 
    10 (45, 5%) 
    0, 053 
  
  
    Medical    Therapy 
    
    
    
  
  
    Yes 
    4 (28, 6%) 
    10 (45, 5%) 
    0, 311 
  
  
    Single 
    4 
    6 
    0, 099 
  
  
    Multiple 
    0 
    4 
  
  
    1Data were presented as mean ± standard deviation or median (range) as    deemed more appropriated.



Table 2: General characteristics-comparative analysis between normo and
hyperprolactinemic patients.

On a different perspective, categorizing patients based on the immunostaining pattern (GH positive versus GH+PRL positive) did not disclose differences in terms of serum PRL levels between these 2 subgroups, (Table 3).

Table 3: General characteristics-comparative analysis between immunostaining profile.




  
    Patient    Caractheristics1 
    GH    (n=22) 
    GH/PRL    (n=18) 
    p-value 
  
  
    Demographics 
    
    
    
  
  
    Age at diagnosis (years) 
    48, 1 (±11,5) 
    44 (±15,1) 
    0, 084 
  
  
    Gender (Female) 
    17 (77, 3%) 
    9 (50%) 
    0, 072 
  
  
    Biochemical    Profile 
    
    
    
  
  
    Diagnosis 
    
    
    
  
  
    PRL (ng/mL) 
    19,4 (0,2–181,2) 
    33,8 (3-195, 7) 
    0, 334 
  
  
    (n=18) 
  
  
    Hyperprolactinemia 
    10 (55, 6%) 
    12 (66, 7%) 
    0, 494 
  
  
    PRL >40ng/mL 
    4 (22, 2%) 
    7 (38, 9%) 
    0, 278 
  
  
    GH (ng/mL) 
    17, 6 (4-200) 
    23,7 (2-200) 
    0, 530 
  
  
    IGF-1 (ng/mL) 
    865, 0 (340-1414) 
    943,0 (295-2230) 
    0, 329 
  
  
    Tumor    characteristics 
    
    
    
  
  
    Microadenoma 
    2 (9, 1%) 
    2 (11, 1%) 
    0, 832 
  
  
    Macroadenoma 
    20 (90, 9%) 
    16 (88, 9%) 
  
  
    Cavernous sinus invasion 
    11 (50%) 
    7 (38, 9%) 
    0, 502 
  
  
    Remission 
    
    
    
  
  
    3 months 
    5 (23, 8%) (n=21) 
    1 (5, 9%) (n=17) 
    0,132 
  
  
    1 Year 
    11 (50%) 
    8 (44, 4%) 
    0,761 
  
  
    PRL normalization rate 
    6 (75%) 
    8 (72, 7%) 
    -- 
  
  
    Radiotherapy 
    
    
    
  
  
    Yes 
    6 (27, 3%) 
    6 (33, 3%) 
    0, 677 
  
  
    Medical    Therapy 
    
    
    
  
  
    Yes 
    7 (31, 8%) 
    7 (38, 9%) 
    0, 870 
  
  
    Single 
    6 
    4 
    0, 185 
  
  
    Multiple 
    1 
    3 
  
  
    1Data were presented as mean ± standard deviation or median (range) as    deemed more appropriated.



Table 3: General characteristics-comparative analysis between immunostaining profile.

There was a higher number of patients with serum PRL levels above 40ng/mL among those with GH/PRL staining than among those without PRL staining, although not statistically significant (GH: 22, 2% vs GH/PRL: 38, 9%, p= 0,278), (Table 3).

There was a trend towards a higher incidence of cavernous sinus invasion among patients with hyperprolactinemia (63, 6% vs 28, 6%, p=0,064), (Table 2); this difference reached the level of significance among patients with hyperprolactinemia and immunostaining positive only for GH (GH: n=9, 90%; GH/PRL: n=5, 46%; p=0,031).

Normalization of PRL was not synonymous of remission. Both events occurred in 4 patients. All 10 patients with PRL normalization, despite acromegaly persistence, presented with a macroadenoma, six of them with cavernous sinus invasion; 5/10 had immunopositivity for GH and PRL.

Prolactin levels did not correlate with the final outcome. Nonetheless, hyperprolactinemic patients met remission criteria later after surgery, based on results at 3 months and 12 months after the procedure, (Table 2). Among those in whom adjuvant therapy postoperatively was required, the need for radiotherapy and medical therapy with more than one agent trended towards a higher frequency in hyperprolactinemic patients, (Table 2).

Discussion and Conclusions

Acromegaly is a slow progressive disease with significant heterogeneity in both presentation and surgical outcomes. Hyperprolactinemia has been reported in around 30% of acromegalic patients and may result either from the co-secretion of GH and PRL by tumour tissue or from the pituitary stalk compression (stalk syndrome) with impaired dopamine delivery. Previous studies suggested a more aggressive clinical course in acromegalic patients with hyperprolactinemia [8,11]. However, whether this is somehow related to the origin of the hyperprolactinemia is not completely understood. Thus, we started by comparing acromegalic patients with and without hyperprolactinemia. We documented PRL levels above the normal range in 61% of the patients. Differences in the cut-off used to define hyperprolactinemia are a possible explanation for a higher prevalence than observed in other series [8]. In what concerns the immunostaining of hyperprolactinemic patients, only 55% of our patients presented GH and PRL staining, lower than reported by others [12-14].

Similarly to other studies, we also found an association between hyperprolactinemia and tumour invasiveness [8,14]. However, when we compared hyperprolactinemic patients with positive immunostaining for PRL with those with negative immunostaining, the association between elevated PRL levels and tumour extension was only significant for the latter group. Moreover, despite the absence of a significant difference on the final remission rates, early remission rate (3 months after surgery) was significantly lower in hyperprolactinemic patients. Furthermore, the need for adjuvant therapy postoperatively with more than one agent, as well as the indication for radiotherapy also trended towards a higher frequency among hyperprolactinemic patients.

Hence, these findings are in favour of an association of elevated PRL levels with a more aggressive clinical course and less favourable outcome, among those patients in whom hyperprolactinemia is likely to result from stalk compression, usually associated with larger and more invasive tumours [15].

In order to assess the impact of hyperprolactinemia related to cosecretion by the pituitary tumour, we compared acromegalic patients staining positive for just GH with the ones with positive staining for both GH and PRL. Around 45% of our patients had positive immunostaining for both GH, PRL, a number in line with the one reported by Liang et al., although higher than the one reported by Rick et al and Varlamov et al. [16-19].

Comparative analysis of these subgroups did not disclose significant differences. There was a trend towards a lower age and higher levels of GH and IGF-1 in the subgroup staining for GH and PRL. In terms of serum PRL levels no significant differences were noticed. Immunostaining for PRL did not always translate into hyperprolactinemia as observed in 33% of cases. Indeed, when normoprolactinemic patients with and without PRL immunostaining were compared, no differences concerning tumour presentation or final outcome were found. The retrospective nature of our study did not allow to categorize tumours staining for GH and PRL as mammosomatotroph adenomas or somatotroph-lactotroph adenomas and look for differences between these subtypes as performed by others [17,18].

Looking to those patients in whom hyperprolactinemia coexisted with macroadenoma and immunostaining positive both for GH and PRL, 8/11 normalized serum PRL after surgery. However, only 3 reached remission criteria. In the remaining 5, with persistence of disease, debulking of tumour was enough for PRL normalization reinforcing the dominant role of stalk compression for the occurrence of hyperprolactinemia.

The main constraints of this study are its retrospective nature and the number of patients included. Whether a more detailed characterization of the exact lineage profile of the dual staining tumours, namely whether they were monomorphous or dimorphous adenomas, could add any contribute to final results was far beyond the scope of the study. In conclusion, our results favour a poor outcome for hyperprolactinemic patients, when the underlying cause is stalk compression or, in other words, when the tumours are larger and likely to be more aggressive.

References

Download PDF

Citation: Matos T, Gomes V, Faria C, Gomes A, Nobre E and Bugalho MJ. Causes of Hyperprolactinemia in Acromegalic Patients and Clinical Correlations. Austin J Endocrinol Diabetes. 2021; 8(1): 1080.

Home

Journal Scope

Editorial Board

Instruction for Authors

Submit Your Article