The hypothalamic-pituitary axis involves the interplay of tightly regulated hormone action through positive and negative signaling. Somatotroph GH production is attenuated by hypothalamic somatostatin, which suppresses timing and amplitude of GH secretory pulses by binding to somatostatin receptors (SST) (4), a family of G protein-coupled transmembrane receptors (GPCR) with five known subtypes (SST1-5), each with differing binding and hormone modulating capacity. SST2 and SST5 are most abundantly expressed receptors on somatotrophs and are the primary targets of adenoma-directed medical therapy (9, 10).

SST2 is expressed in brain, pituitary, stomach, kidney, and intestines (11). It signals to suppress mainly GH, and less compellingly thyroid-stimulating hormone (TSH), adrenocorticotrophin, and prolactin (12). The prevalent expression of SST2 in GH-secreting adenomas (9) makes it a prime target for treating acromegaly. The SRLs octreotide and lanreotide preferentially bind to SST2, thereby suppressing GH expression (1, 13). Disease control with these agents, however, can be variable with a reported mean response rate of approximately 55% (14).

SST5 is primarily expressed in the pituitary but is found as well in the spleen, intestines, and pancreas (15–17). SST5 regulates GH and TSH (12). Unlike with octreotide and lanreotide, the SRL pasireotide predominantly targets and binds with high affinity to SST5. It has shown to be more efficacious in patients resistant to other SRLs (18, 19), but it is associated with an increased risk of hyperglycemia, likely due to the role of SST5 in regulating glucose homeostasis through paracrine regulation of intestinal GLP-1 (20). Intuitively, lowering circulating GH also enhances insulin sensitivity thereby countering some of these hyperglycemic effects, but the risk remains and potentially limits its use.

Immunohistochemical expression of SST2 and SST5 may predict treatment outcomes, especially as expression is less abundant in more aggressive adenomas that are less prone to treatment responsiveness (21). For example, we showed that SRL treatment efficacy is typically high in patients with low or intermediate tumor aggressiveness, and these patients show concordant abundant SST2 expression. By contrast, SST2 expression is lower in patients with more aggressive and highly unresponsive tumors (21). Others have reported that response rate to SRL may be as high as 81% in the presence of SST2 expression compared with no response observed in the absence of SST2 expression (22, 23). Further, a higher SST5:SST2 ratio correlates with better response to pasireotide and resistance to octreotide/lanreotide (24–26). Of note, epigenetic factors, such as the natural antisense transcript SST5-AS, has been shown to alter SST5 expression, potentially influencing adenoma behavior and treatment response (27).

Although not routinely evaluated by clinical pathologists, SST expression may be a valuable molecular marker for precision-based acromegaly treatment, as decision for repeat surgery, radiotherapy, or choice of a specific SRL could be better guided with knowledge of SST2 and SST5 expression.

Human growth hormone receptor (GHR) is a transmembrane protein structurally related to a family of cytokine receptors (28, 29). Binding of GH leads to GHR dimerization and activation through the JAK/STAT pathway, leading to induction of IGF-1 (30). GHR is the target of the peripheral GHR antagonist pegvisomant, which is highly effective in treating acromegaly both as monotherapy and in combination with SRL (31, 32). A novel drug in development, Cimdelirsen (IONIS-GHR-LRx; ISIS 766720), which is an antisense molecule that acts by reducing GHR synthesis in the liver, has also shown promise in treating acromegaly (33). Polymorphisms in GHR, such as d3-GHR, has been studied in acromegaly but correlation with clinical features or therapeutic outcomes has not been consistent (34).

GHR expression is observed in normal somatotroph cells and to a lesser degree in somatotroph adenomas (35, 36). Pegvisomant may in part affect growth hormone production through direct action on GHR on somatotroph cells without impacting cell proliferation (37). However, the value of GHR as either a peripheral or central molecular marker for predicting treatment response is uncertain.

Dopamine receptors (DR) comprise five GPCR subtypes numbered D1R through D5R and are present abundantly in the central nervous system, and peripherally in the pituitary, kidney, and vasculature (38, 39). D2R is present in two isoforms of equal activity and distribution, and is the predominant DR found in normal pituitary as well as in somatotroph and lactotroph adenomas (25).

The downstream cellular action of activated DR includes inhibition of adenylyl cyclase and activation of potassium channels, leading to inhibitory action of hypothalamic dopamine on prolactin production (40). In somatotroph adenomas, dopamine lowers GH production (41). This is likely the mechanism underlying the modest efficacy of DA monotherapy in the treatment of mild acromegaly (42) as expression of DR on somatotroph adenoma may have some value in predicting clinical response to SRL (25) yet serum prolactin level does not correlate with GH responses to cabergoline (42).The very limited value of using D2R as a therapeutic marker for acromegaly is limited to identifying patients with mild disease who may respond to DA as monotherapy or as an adjunct to SRL therapy.

E-cadherin is a transmembrane adhesion protein found on epithelial cells, and its loss is implicated in invasiveness and metastasis (43). Its expression has been suggested to be functional for somatotroph adenoma growth (44), but this association is modest, likely because other factors that promote cell senescence (45) prevent low expression of E-cadherin from causing transformation of pituitary adenomas to malignant tumors. Decreased E-cadherin expression in somatotroph tumors correlates with larger size, invasiveness, and response to SRL treatment (44). Sparsely granulated somatotroph adenomas, which are more likely to be aggressive (discussed below), also show lower E-cadherin expression compared to densely granulated tumors (46, 47). The association between low E-cadherin expression and poor SRL response seems independent of SST2 expression but may correlate inversely with SST5 expression (48). Such correlations between histological and cell surface markers provide an interesting intersection of molecular markers of tumor aggressiveness. However, direct relationships of these factors are not known, and they may simply represent the phenotypic components of more aggressive tumors.

The EGFR family of tyrosine kinase receptors including EGFR (ErbB1 and HER1), p185^Her2/neu (ErbB2 and HER2), ErbB3 (HER3) and ErbB4 (HER4), and their activating ligands including EGF and transforming growth factor-α are expressed in lacto-somatotroph cells and are potential targets for treatment (49). EGFR tyrosine kinase inhibitors (TKI) such as gefitinib, lapatinib, and canertinib have shown promising results in patients harboring aggressive corticotroph and lactotroph adenomas (50, 51). Although EGFR and EGF are abundantly expressed in somatotroph adenomas, trials evaluating TKI in treatment of aggressive acromegaly are lacking.

β-arrestin binds to GPCR and thereby regulates signaling (52). Results of studies on the role of β-arrestin in responsiveness of somatotroph adenomas to therapy has been mixed. In a study of 31 somatotroph adenomas, lower expression of β-arrestin-1 correlated with improved responsiveness to SRL (53), and a similarly sized study showed an inverse relationship between low β-arrestin-1/2 and SST2 expression, as well as a more favorable long-term response to SRL treatment (54). However, this relationship was not observed in another study of 40 patients, which showed no correlation of β-arrestin with SST2, SST5, or D2 expression, nor an association between β-arrestin and SRL response or tumor invasiveness (55). Therefore, the significance of β-arrestin as a molecular target for acromegaly remains unclear.

Inactivating mutations of the tumor suppressor gene AIP in pituitary tumors is associated with familial acromegaly syndromes and is rarely seen in sporadic acromegaly (7). Familial acromegaly involving AIP mutation is more aggressive and less responsive to SRL (60). AIP expression has been associated with dense granulation (61), and inversely correlates with Ki-67 expression (62), suggesting a pro-proliferative state.

The mechanism of AIP mutation and tumorigenesis likely involves alterations in PDE4 phosphodiesterases as well as defective cAMP signaling. Specific PDE4 isoforms, PDE4A4 and PDE4A8, are under-expressed in normal pituitary and overexpressed in somatotroph adenomas, suggesting that disruptions of PDE4-AIP interaction play a role in tumorigenesis (63). Dysfunction in cAMP signaling through defective and decreased expression of Gai subunits, which help regulate AIP mediated cAMP signaling, also plays a key role (64). The mechanism of interplay between expression of AIP and cell-surface molecular receptors requires further study and may yield novel sub-cellular targets for treatment.

Mutations in GNAS are present as somatic mutations in up to 40% of sporadic somatotroph adenomas and as mosaic mutations in McCune-Albright syndrome, a genetic disorder characterized by skin and bone manifestations as well as increased incidence of acromegaly (65). The impact of GNAS mutation on treatment response is uncertain. Though a meta-analysis showed a greater GH reduction in response to SRL treatment in GNAS mutations (66), a more recent study did not show a significant difference in SRL response and GNAS mutations (67). Pegvisomant is also an effective treatment option for acromegaly in patients with MAS (68); however, whether GNAS mutation influences treatment outcome between different available therapies is not known.

PTTG is a homolog of securin proteins that prevent sister chromatin separation (69), and regulate the cell cycle through interaction with p53 (70). Over 70% of somatotroph adenomas overexpress PTTG, and this expression is an important component of cell senescence (71). PTTG expression may correlate with aggressiveness across different pituitary adenoma types (72), but it’s role as a molecular marker for clinical use in somatotroph adenoma characterization and targeted treatment warrants further investigation.

Zinc finger regulator of apoptosis and cell cycle arrest (ZAC1) is a tumor suppressor that attenuates cell proliferation (73). The antiproliferative property of SRL therapy may be mediated in part through increased ZAC1 gene expression (74), with evidence that suggests increased AIP expression is the linking mechanism (75). The lack of difference in ZAC1 expression between densely and sparsely granulated adenomas implies that ZAC1 does not play a significant role in tumor aggressiveness (57).

In the 2017 WHO Classification of Tumors of the Pituitary Gland, the term or classification of atypical adenoma, which was defined partly by Ki-67 proliferative index, was no longer recommended (76). However, Ki-67 may still hold value in predicting aggressiveness and treatment response in somatotroph adenomas.

Most non-aggressive adenomas show a Ki-67 proliferative index of <3% (21), and low Ki-67 expression has been associated with more favorable SRL response (67). Additionally, Ki-67 index correlates inversely with cavernous sinus invasion, surgical cure, and response to medical therapy (77). SRL treatment may also alter Ki-67 expression as evidenced by lower adenoma Ki-67 values observed in patients treated chronically with octreotide (78).

CDK and their inhibitors may serve as markers of somatotroph adenoma subtypes (4, 79). p21, a CDK inhibitor that plays a key role in cell senescence, is activated to maintain the benign nature of somatotroph adenomas (45, 71). Increased expression of p21 is associated with less aggressive somatotroph adenomas, likely exerting a dampening effect on cell proliferation (21). However, its impact in predicting curative success of surgery may be limited. In a series of 55 patients undergoing surgical adenoma resection, p21 overexpression did not correlate with biochemical remission after surgery (80), likely because the skill and volume of the surgeon, is a primary predictive factor for surgical success, more so than any one particular molecular factor (81).

The value of other CDK inhibitors such as p16 and p27 as molecular drivers in acromegaly are less certain. p16 expression is low or undetectable in all pituitary adenoma types and does not correlate with somatotroph tumor aggressiveness (21, 82). Low or absent p27 expression seems to play an especially important role in aggressive corticotroph adenomas as well as malignant pituitary tumors. Although p27 expression is also lower in somatotroph adenomas, its correlation to tumor aggressiveness or response in acromegaly is less clear (83). Despite limitations, CDK inhibitors may serve as compelling molecular predictors of tumor proliferative growth and as targets for development of novel drugs that target CDK pathways.

p53-mediated tumor suppression in pituitary pathology acts through the p53/p21 senescence pathway (71). However, p53 is not reliable in predicting tumor aggressiveness. Somatotroph adenoma p53 overexpression has not shown significant association with tumor invasiveness or response to treatment (84–86).