Using a Precision Approach to Optimize the Drug Therapy of Patients with Acromegaly Syndrome

Abstract

Modern problems of acromegaly treatment are associated with the heterogeneous composition of somatotrophic tumors, differing in clinical course and sensitivity to the proposed therapy. Under these conditions, the achievement of acromegaly control depends on the stratification of clinical, laboratory and instrumental data in order to identify significant biomarkers that allow predicting the receptor phenotype and biological behavior of the tumor, the tendency to relapse and the long-term effectiveness of drug therapy. The review discusses modern predictor models reflecting the radicality of surgical treatment, the risk of the continued growth of a resident tumor, the long-term results of clinical use of first-generation somatostatin receptor ligands (fg-SRLs), as well as the possibilities of therapeutic maneuver. It is proposed to use pharmacotherapeutic testing to evaluate the receptor expression of tumor cells and predict the effectiveness of long-term treatment of fg-SRLs. Summary data characterizing various morphotypes of somatotrophic tumors are presented. It is shown that the use of a precision approach can significantly accelerate the time to achieve control and improve the quality of the treatment aid in patients with acromegaly syndrome.

1. Introduction

Acromegaly is an insidious chronic disease with a great burden resulting from: insufficient medical awareness, delayed diagnosis and undifferentiated formal therapy. The cumulative effect of excessive secretion of growth hormone (GH) and insulin-like growth factor − 1 (IGF-1) contributes to the formation of specific somatic, systemic and metabolic disorders in the body, often of an irreversible nature. In the absence of adequate treatment, the disease is characterized by high rates of morbidity and premature mortality associated with the development of cardiovascular, respiratory, metabolic and neoplastic disorders. A direct correlation between the duration of uncontrolled disease and the severity of combined life-threatening complications has been proven. Independent predictors of premature death include: cardiovascular disorders; arterial hypertension; diabetes mellitus; high levels of GH and IGF-I persisting, despite ongoing treatment; long duration of the active stage; elderly age of patients [1, 2, 3].

According to epidemiological data, the prevalence of acromegaly in terms of circulation is 28–137 cases and the incidence is 2–11 cases per 1 million inhabitants [4]. An important problem is the delayed diagnosis of acromegaly, which is associated with the slow course of the disease, insufficient awareness of this pathology among practitioners, as well as a low level of dispensary supervision. It is shown that in 54% of patients, the prediagnostic period is more than 10 years, which contributes to the development of irreversible multiple organ complications and negatively affects the quality and life expectancy of patients. At the time of detection, most of the tumors are macroadenomas possibly relating to diagnostic delays and posing challenges in surgical management. Currently, the direction of early diagnosis of the disease is actively developing, including various methods of mass and selective screening, proving that the actual prevalence of acromegaly significantly exceeds the existing indicators for the treatment [5, 6, 7].

2. Acromegaly syndrome

Contrary to previous ideas, acromegaly is not a monomorphic disease but is a syndrome that unites a group of morphologically and functionally different formations. In the 5th Edition of the 2022 WHO Classification of Head and Neck Neuroendocrine Neoplasms, pituitary somatotrophic adenomas are referred to as somatotrophic pituitary neuroendocrine tumors (PitNET). There are at least six distinct genetically determined or sporadic morphologic types PitNET, which, despite belonging to a single PIT1-dependent family and a common acidophilic cell line, nevertheless differ in histological structure, hormonal and proliferative activity, aggressiveness of intracranial growth and sensitivity to the treatment presented [8, 9].

From differentiated histological subtypes, pure—Densely Granulated and Sparsely Granulated Somatotroph Tumors, Mammosomatotroph Tumors and mixed—somatotroph–lactotroph tumors are distinguished. Low-differentiated morphotypes (Mature Plurihormonal Tumors of PIT1-Lineage, Immature PIT1-Lineage Tumors, Acidophil Stem Cell Tumors), characterized by low secretory activity but accelerated invasive growth are also described [10, 11].

The most common variant, occurring in 30–50% of cases, are sporadic Densely Granulated Somatotroph Tumors (DGST), formed during 4–6 decades of life and characterized by preserved species specialization, slow growth and high secretory activity. In addition to GH, DGST secretes an α-subunit (α-SU), the determination of which may have differential diagnostic value. This form of acromegaly is characterized by late manifestation, latent nature of the course and poorly expressed orofacial changes, which is manifested, as a rule, by delayed diagnosis and a wide range of disabling multiple organ and metabolic disorders that negatively affect the quality and life expectancy of patients. Nevertheless, the preservation of the receptor phenotype with the dominant expression of the 2nd subtype of somatostatin receptors (SSTR2) in DGST cells is manifested by good sensitivity to first-generation somatostatin receptor ligands (fg-SRLs) [12, 13].

In 15–35% of cases, the cause of acromegaly is less differentiated Sparsely Granulated Somatotroph Tumors (SGST), manifested by accelerated extracellular and invasive growth, leading to the development of intracranial compression, visual and neurovascular disorders. These tumors are typically invasive macrotumors with extrasellar invasion at the time of diagnosis. In SGST cells, there is a relatively lower density of SSTR2, which causes resistance to fg-SRLs. Increased proliferative index Ki-67 (>3%) and low content of the adhesive protein E-cadherin contribute to the active and invasive character of tumor growth. This subtype of somatotrophic adenomas, which is the most problematic for curation, manifests at a young age and is characterized by a pronounced mass effect, a reviving course and resistance to radiation and drug effects. Due to the high risk of malignancy, SGST is included in the group of aggressive neuroendocrine pituitary tumors requiring active combined treatment and lifelong dynamic control [10, 14].

The biological behavior of mixed somato-lactotrophic tumors and their sensitivity to treatment are determined by the cytological composition of the somatotrophic component of a bicellular tumor and, in the presence of rarely granular cells, these tumors also differ in a negative therapeutic prognosis. Mammosomatotrophic tumors belong to monocellular adenomas, in which each cell secretes GH and prolactin, and are characterized by hereditary predisposition, early manifestation (active linear growth) and high secretory activity. Since these tumors consist of densely granular cells, they differ in sensitivity to fg-SRLs and low signal intensity on T2-weighted MR images. In addition, rare PIT1-line tumors (plurihormonal and poorly differentiated neoplasms, as well as acidophilic stem cell tumors) are isolated, which are characterized by invasive growth and a negative prognosis.

Most of the GH-secreting tumors are sporadic. In 40% of cases, a somatic mutation of the gene encoding the synthesis of the alpha subunit of the receptor G protein (GNAS1) is observed, contributing to an increase in the secretory and mitogenic activity of cells. Approximately 4–12% of patients have a congenital mutation of the gene responsible for the synthesis of aryl hydrocarbon receptor-interacting protein (AIP). This form occurs in young patients and is manifested by accelerated and invasive growth of the pituitary tumor. In about 5% of cases, clinical symptom complexes are observed due to congenital genetic disorders (MEN syndromes of types 1 and 4, Carney complex, McCune-Albright syndrome, familial isolated pituitary adenoma, X-linked acrogiantism, etc.). These tumors manifest at an early age, are clinically manifested by gigantism, and are characterized by rapid growth, aggressive course and resistance to treatment. The extrahypophysial causes of acromegaly, detected in 1% of cases, include ectopic neuroendocrine tumors secreting GH or somatoliberin, the visualization and treatment of which are often a clinical problem [13, 14, 15].

3. Tumor-oriented diagnosis and treatment

Thus, the existing intra-group differences in the biological behavior of GH-secreting adenomas, their ability to residual growth, and susceptibility to drugs dictate the need for a tumor-oriented differential diagnostic search to establish a specific immunohistochemical subtype of a somatotrophic tumor as an indispensable condition for the development of an effective therapeutic aid. Unfortunately, the neglect of the syndrome approach observed in clinical practice manifests itself in the unified pharmacotherapy schemes of the acromegaly method “a trial-and-error” with the “blind” selection of a suitable drug, which, in case of an incorrect decision, is fraught with further progression of the disease and shortening of the survival period. The precision approach in endocrinological practice involves the use of clinical, humoral, imaging and pathomorphological predictors for preliminary stratification of clinically homogeneous patients with a clear clinical prognosis and response to treatment [16, 17].

Currently, individual biomarkers capable of predicting the possibility of postoperative remission, the risk of recurrence (or continued growth of a residual tumor), as well as the long-term effectiveness of drug therapy (DT), taking into account the immunological phenotype and clinical characteristics of PitNET, are widely discussed in the scientific literature. For example, it has been established that the intensity of the tumor signal on T2-weighted magnetic imaging can indirectly indicate the morphological variant of somatotrophic tumors and, accordingly, predict sensitivity to SRL treatment. Thus, DGST is manifested by a low signal intensity compared to surrounding formations (the intact part of the pituitary gland, gray or white matter of the cerebral cortex), whereas SGST is characterized by a high signal intensity. The introduction of this diagnostic feature into domestic clinical practice will make it easier to choose a treatment strategy [18, 19, 20].

It is proved that persistent biochemical control is the main factor determining the quality of life and survival of patients with acromegaly. The modern medical manual includes methods of surgical treatment, drug therapy and radiation exposure. Each of these options has variable effectiveness and specific side effects that must be taken into account when planning a medical program. The criteria for biochemical remission in the treatment of acromegaly are: the age-appropriate level of IGF-1, sporadic level of GH <1 mcg/l and the value of GH nadir when using OGTT with 75 g of glucose <0.4 mcg/l (when using a highly sensitive method for determining GH). Taking into account age and laboratory fluctuations of the normal values of the IGF-1, it is recommended to use a unified indicator of the IGF-1 index, reflecting the amount of excess of the IGF-1 above the upper limit of age norm (IGF-1 index = IGF-1/ULN), the target value of which should be <1. Unfortunately, in a number of national registries and retrospective studies, free handling of the remission target is practiced, in which an increase in the value of the IGF-1 index to 1.2 and 1.3 is allowed; which is not consistent with consensus agreements. Such artificial improvement of statistical indicators increases the rating of the register but negatively affects the therapeutic prognosis, since it eliminates the need for therapeutic correction. This remark also applies to the interpretation of cut-off points. It is clear that at different target values of acromegaly control (for example, at values IGF-1 index <1 or < 1.3), different indicators of cut-off points of prognostic independent variables are assumed, which complicates the development of a consolidated prognostic protocol. Therefore, when assessing the predictive power of the proposed predictors (or thresholds), it is necessary to additionally indicate the remission targets IGF-1 index used in a specific retrospective study [21, 22].

The purpose of this work is to discuss the priority of clinical and laboratory markers and the validation of independent variables that determine the prognosis of the effectiveness of therapeutic measures, including surgical aid and long-term use of first-generation somatostatin receptor ligands (fg-SRLs).

4. Predictors of the prospects and adequacy of surgical aid

According to international recommendations, transsphenoidal adenomectomy of somatotroph pituitary neuroendocrine tumor (PitNET) is the first-line treatment of acromegaly with a high chance of complete cure of the majority of patients with acromegaly. Pharmacological treatment is recommended if surgery is contraindicated or did not lead to disease remission. The choice of treatment best fitting each patient should be based on a thorough investigation of patients’ characteristics.

4.1 Predictors of postoperative remission

The indicators of postoperative (p/o) biochemical remission vary from 32 to 85% depending on the size of the tumor, the severity of invasion of the cavernous sinus, the qualifications of the neurosurgeon and the control criteria used. With delayed diagnosis and large tumor sizes with suprasellar and parasellar growth, radical adenomectomy is possible only in 40–60% of cases. The ambiguity of surgical outcomes contributed to the development of a prognostic direction that determines the effectiveness of the planned surgical aid and the likelihood of adjuvant treatment. In addition to the dynamics of p/o hormonal parameters, among the clinical biomarkers that allow predicting the likelihood of remission, there are: initial indicators of hormonal and proliferative activity of tumor, its volume and the nature of extracellular spread, as well as radiological signs of invasive growth. Later, radiometric and immunophenotypic characteristics of somatotrophic tumors were added to the general list [23, 24].

Taweesomboonyat et al. showed that the remission rate after transsphenoidal removal of micro- and macroadenomas is 100 and 44%, respectively. The conducted multifactorial analysis showed that the combination of a high preoperative value of IGF-1 index (> 2.5) and the presence of 3–4 degrees of adenoma enlargement on the Knosp grade are unfavorable prognostic factors regarding the radicality of the upcoming surgery and indicate the need for subsequent adjuvant treatment [25].

In a retrospective study by Cardinal et al., it was shown that the preoperative level of IGF-1 and the value of GH on the first day after surgery are inversely correlated with hormonal remission. A drop in the concentration of GH in the first 1–2 days after surgery <1.55 ng/ml is a predictor of its success. At the same time, the sensitivity of this indicator, according to the authors, was 75, the specificity was 59%. The average sporadic GH level in patients who achieved remission was 1.6 versus 9.5 ng/ml in patients with persistent activity. In a similar work by Mohyeldin et al., it was found that the value of GH-nadir during OGTT 24–48 hours after surgery <1.15 ng/ml is the best predictor of remission with a sensitivity of 73% and specificity of 85%. As a result of the multivariate logistic regression analysis Tomasik et al. have established that such indicators, as a sporadic concentration of GH < 8.63, the maximum diameter < 15.5 mm, the presence of normoprolactinemia, DGST, positive staining on α-SU на are independent predictors of surgical remission [26, 27, 28].

Since no single marker is able to independently predict the postoperative outcome, prognostic models and nomograms are usually used, including a complex of diverse predictors and allowing to significantly increase their sensitivity and specificity. In the work of Agrawal et al., a mathematical analysis of retrospective studies with the identification of factors affecting the effectiveness of the surgical intervention is presented. It was noted that the young age of patients, a large tumor volume, initially high levels of GH and IGF-1 and the presence of signs of invasion of the cavernous sinus may be predictors of non-radical adenomectomy [29].

Heng et al. proposed a predictive model that allows predicting the presence of SGST with a high degree of probability, characterized by aggressive behavior and a tendency to relapse. The study included 44 patients with DGST and 39 patients with SGST. In the course of comparing the data of the preoperative examination with the results of the morphological diagnosis, it was found that patients with SGST were distinguished by a young age, large adenoma sizes, a high degree of invasion and low sensitivity to octreotide treatment. Thus, the size of the tumor, the degree of invasion on the Knosp scale, the value of the GH index and the percentage of decrease in GH against the background of octreotide administration are independent variables, the combination of which made it possible to compile a graduated scale for predicting SGST with an AUC value of 0.84 and high sensitivity and specificity. According to Swanson et al., the absence of invasion and lower pre-operative IGF-1 index were the only significant predictors of post-surgical remission in this cohort [30, 31].

4.2 Predictors of recurrence or continued growth of a residual tumor

A separate topic is the prediction of the risk of the continued growth of a residual tumor and/or relapse, which seems to be important for determining the strategy of active dynamic control. According to the data presented by Lucas et al., continued growth is observed in 12–58% of patients with residual tumors. Even with radical adenomectomy, 10–20% of operated patients have a relapse of the disease over the next 5–10 years. It is noted that the very fact of the presence of residual tissue is the leading condition for continued growth. According to observations, the overall recurrence rate after 5, 10 and 15 years was 25%, 43% and 61%, respectively. At the same time, an inverse correlation was found between the patient’s age and the risk of further growth of residual tissue. It was also shown that the risk of p/o growth of the remaining tissue outside the sella turcica was 3.7 times higher than that of resident tumors confined to the sella turcica area. Therefore, more active treatment strategies should be used for patients with extrasellar tumor tissue residue [32]. In the process of conducting logistic regression analysis, Mohyeldin et al. identified three statistically significant independent variables indicating a high risk of invasion of the medial wall of the cavernous sinus: the degree of tumor enlargement on the Knosp scale >2; male sex; the presence of an aggressive form of somatotrophic tumor [27].

A clinical study by Freda et al. is of interest, according to the results of which it is concluded that the risk of p/o relapse increases in persons with achieved normalization of IGF-1, but impaired suppression of GH against the background of OGTT. Thus, the pathological value of the GH-nadir may indicate the presence of an autonomous formation [33].

In the prognostic model of continued tumor growth L. Lu et al suggest including markers such as the young age of patients; high pre-operative level of secretory activity; signs of invasive growth; volume and localization of residual tissue; preservation of pseudocapsules; presence of aggressive morphological subtype (SGST) [23].

In the work of Wang et al., 178 patients with pituitary adenomas were retrospectively analyzed with the release of predictors indicating a high level of Ki-67 (>3%). In the process of multivariate regression analysis, it was shown that young age, the abundant blood supply to the tumor and erosion of the back of the turcica sella are independent markers of a high value of the proliferative index Ki-67, and, consequently, the risk of residual growth. In a study by Chen et al., it was noted that independent factors such as BMI (> 25 kg/m2), the 4th degree of increase on the Knosp scale, partial resection, Ki-67 (>3%) can serve as predictors of postoperative progression of tumor growth or recurrence of pituitary adenomas [34, 35].

On the contrary, the predictor of a low risk of relapse is: older age, small tumor, low level of secretory activity, benign morphological subtype, postoperative radiation therapy. According to Lu et al., the recurrence rate in patients who did not receive radiation therapy increased progressively during the follow-up period, reaching 72% after 15 years. On the contrary, in patients who received timely radiotherapy, the recurrence rate stabilized at 9% after 10 years [23].

Thus, these data indicate the practical importance of improving prognostic models that allow determining the postoperative scenario of the disease and the need for adjuvant therapy.

5. Predictors of sensitivity to somatostatin receptor ligands

The existing problem of drug therapy (DT) acromegaly is the lack of a differentiated approach when choosing a targeted drug, taking into account the receptor phenotype of a somatotrophic tumor. Its solution is connected with the identification and verification of clinical and biological predictors that allow identifying groups of patients sensitive to the planned treatment [36]. Current pharmacological options for patients with acromegaly include somatostatin receptor ligands (SRLs), growth hormone receptor agonists, and dopamine agonists (DA). Somatostatin receptor ligands (SRLs) have been used in clinical practice for more than 30 years and play a leading role in the treatment of acromegaly. Currently, three extended forms of SRLs are widely used worldwide (octreotide, lanreotide and pasireotide), which occupy an important place in the algorithm of treatment of acromegaly. It is known that the cells of somatotropic adenomas mainly contain the 2nd and 5th subtypes of somatostatin receptors (SSTRs), the intensity of membrane expression of which is a key factor in the implementation of the suppressive action of SRLs. By acting on the somatostatin receptors of adenomatous cells, SRLs provide a dose-dependent blockade of the rhythmic secretion of GH and inhibition of cell growth.

The binding of SRLs to SSTRs promotes the dissociation of the Gi protein receptor complex, followed by a decrease in the activity of adenylate cyclase, leading to the opening of potassium channels, hyperpolarization of the cell membrane and closure of calcium channels. The changes in ion permeability resulting from the activation of the receptor lead to a critical decrease in the intracellular concentration of calcium ions and cAMP, which blocks the secretion of GH and the proliferation of somatotrophs. Activation of SSTRs and specific proteins involved in intracellular signal transmission (AIP, ZAC1, RKIP, E-cadherin, β-arrestin) promotes suppression of GH secretion, inhibition of proliferation, cell migration and angiogenesis [16, 19, 37].

The prevalent expression of SSTR2 in GH-secreting adenomas makes it a prime target for treating acromegaly. The first-generation SRLs (fg-SRLs) (octreotide LAR and lanreotide autogel) preferentially bind to SSTR2, thereby suppressing GH expression. The introduction of these drugs into wide clinical practice has significantly increased the effectiveness of acromegaly treatment and the quality of life of patients. Comparative results of the use of prolonged forms of octreotide and lanreotide in patients with acromegaly showed similar characteristics. In the work of Albarel et al., a meta-analysis of the effectiveness of treatment of fg-SRLs in 4464 patients with acromegaly was carried out, during which normalization of levels of GH and IGF-1 was noted in 56 and 55% of patients, respectively [18, 38, 39].

The second important option of the therapeutic effect of SRLs is their antitumor effect due to the development of apoptosis and necrosis of tumor cells. Confirmation of the anti-proliferative effect of SRLs is in vivo studies, which revealed lower indicators of the proliferative index Ki-67 in somatotrophic tumor tissue obtained from patients who were previously treated with octreotide, compared with naive patients [40]. The presence of a direct correlation between the expression of the SSTR2 in adenomatous cells, on the one hand, and the degree of decrease in tumor volume, as well as the severity of biochemical sensitivity to the LAR octreotide, on the other hand, was revealed. According to A. Colao et al., the greatest decrease in volume (about 50% of the initial one) is observed during the 1st year of treatment. During the PRIMARY clinical trial, it was noted that 63% of patients receiving lanreotide Autogel achieved a significant (> 20%) reduction in tumor volume within 48 weeks. At the same time, the magnitude of the decrease in the level of IGF-1 during treatment was the most significant predictor of a decrease in tumor mass. A decrease in the volume of the tumor by more than 20% from the initial one was observed in 66% of patients receiving octreotide LAR (an average decrease of 51%) and in 63% of patients receiving Lanreotide Autogel (an average decrease of 27%). The study of Albarel et al. included a cohort of 1685 patients. Meta-analysis of the results of 41 clinical studies revealed a decrease in the volume of tumors (by an average of 50.6%) in 66% of patients with primary DT fg-SRLs [39, 41, 42]. Biomarkers a guiding a more precise therapy.

Clinical practice has shown that fg-SRLs have a good safety profile, with the most common side effects being mild gastrointestinal symptoms observed in 30% of patients and rarely leading to discontinuation of the drug. Mild hyperglycemia was observed in 15% of patients, and cholelithiasis was observed in 35% of patients treated with fg-SRLs for more than 3–6 months [19, 39].

5.1 The phenomenon of resistance to fg-SRLs

Nevertheless, the results of a long-term study showed that the effectiveness of using fg-SRLs in a non-selective sample, according to various authors, does not exceed 50–55%. In a prospective study evaluating the effectiveness of lanreotide (120 mg every 28 days) as a first-line therapy, it was shown that the normalization of GH levels and IGF-I after 24 and 48 weeks of treatment were observed in 23.4 and 30.6% of patients. Thus, a significant part of patients receiving fg-SRLs has partial or complete resistance, which is a consequence of the internal heterogeneity of somatotropin in relation to the receptor expression of SSTs and other target molecules affecting the therapeutic response [16, 17, 18].

It should be noted that, according to consensus agreements, prolonged forms of fg-SRLs are the starting drug therapy, while other medicinal preparations are prescribed with fatal delay and only after realizing the obvious ineffectiveness of long-term pharmacotherapy. Modest results of fg-SRLs treatment are the expected result of a non-personalized approach, when with the help of a single drug the doctor tries to cure all morphotypes of somatotrophic tumors, including those with a different receptor phenotype. The traditional answer to the detected low sensitivity to fg-SRLs is an escalation of the dose of the drug (or a reduction in injection intervals), followed by the connection of a 2nd-line drug—a selective dopamine agonist (cabergoline). Switching to taking drugs with a different mechanism of action is recommended to be done no earlier than after 12 months of DT. According to Kasuki et al., despite the maximum doses of the drug and the long duration of treatment of fg-SRLs, 20–25% of patients retain the activity of the disease, fraught with the progressive development of multiple organ disorders [43, 44].

A more successful therapeutic practice in detecting resistance to fg-SRLs is switching to a multiligand analog of somatostatin of the 2nd generation of long-acting—pasyreotide LAR, which, unlike octreotide, has half the affinity for the SSTR2 and a higher (5 and 40 times) affinity for the SSTR3 and SSTR5. It shows more efficacious in patients resistant to fg-SRLs.

It is noted that in real clinical practice, biochemical remission against the background of treatment with pasireotide LAR is achieved in approximately 54% of patients resistant to fg-SRLs. A decrease in tumor volume (≥ 20%) was observed in 80.8% of pasireotide LAR and in 77.4% of patients receiving octreotide LAR. At the same time, there is a correlation between the sensitivity to pasireotide and the increased intensity of the MR signal in T2-weighted images [45, 46, 47].

Unfortunately, despite the high antisecretory and antitumor activity, taking this drug in about 60% of cases is accompanied by a violation of carbohydrate metabolism, which limits its clinical use outside the resistant group. Further development of multiligand drugs will allow to unify pharmacotherapy of various subtypes of somatotrophic tumors. In this regard, great hopes are pinned on the drug Veldoreotide (Somatropim) undergoing clinical trials, which has a lower hyperglycemic effect [19, 37].

A selective agonist of dopaminergic receptors, cabergoline (CAB) represents another therapeutic option in acromegaly, but their efficacy to control GH secretion is not related to the degree of DRD2 expression in somatotrophic tumors. Pharmacological niches of cabergoline use are: a/small disease activity (IGF-1 index<2), b/presence of mixed somato-lactotrophic tumors, c/detection of partial resistance to fg-SRLs as a 2nd-line drug. As monotherapy, cabergoline is effective in 20–30% of cases. In case of partial resistance to SRLs, the addition of CAB to SRLs is an option and can lead to IGF-1 normalization in 50% of additional patients [37, 48, 49].

Another pharmacological option for detecting resistance (or non-tolerance) of fg-SRLs is the appointment of a GH receptor antagonist—a pegvisomant, that blocks the biological effect of GH in peripheral tissues (primarily in the liver). The drug does not affect tumor activity and blocks the biological effect of GH, steadfastly normalizing the level of IGF-1 in the blood in 63–97% of cases. Pegvisomant (Somavert) can be used as monotherapy (for small adenoma sizes) or in combination with SRLs (or CAB). The 14-year observation program ACROSTUDY, which brought together 2221 patients from different countries, showed the presence of persistent biochemical control in 72% of patients with acromegaly. The continued growth of residual adenoma was observed in 3.7% of patients. For patients with initially high hormonal activity (IGF-1 index >2.7), diabetes mellitus, increased body mass index (>30 kg/m2), higher starting doses and accelerated titration of the preparation are required to normalize IGF-1 [50, 51].

The indication for the appointment of pegvisomant is the preservation of the activity of the disease after nonradical adenomectomy and ineffective secondary DT. The combination with pegvisomant in patients with partial response to fg-SRLs allows disease control in 80% of patients. The advantages of combination therapy include higher efficiency with lower doses of pegvisomant, as well as stabilization of the size of residual tissue. Thus, thanks to the appearance of pasireotide and pegvisomant, an additional therapeutic maneuver became possible to achieve control, but fg-SRLs, as before, occupy leading positions in the drug treatment of acromegaly [19, 21, 52].

The gold standard for determining the drug strategy of secondary DT is the molecular phenotyping of fragments of a removed tumor using immunohistochemical (IHC) analysis). Depending on the morphotype of the somatotrophic tumor, there is a characteristic difference in receptor expression. Thus, DGST predominantly expresses the SSTR2, whereas SGST, certain types of somato-lactotrophic tumors and low—differentiated forms—mainly the SSTR5, which indicates the possibility of differentiated DT acromegaly taking into account the receptor phenotype.

5.2 Predictors of sensitivity to fg-SRLs

The recognition of the fact of the multiplicity of pathomorphological variants of somatotrophic tumors contributed to the development of tumor-oriented diagnostics by determining predictors indicating the presence of a certain morphotype and a feature of receptor expression, including evaluation samples with octreotide and signifor, secretory and morphological characteristics of the tumor, the results of IHC analysis and the intensity of the tumor signal on T2-weighted MR images. Among the predictors of sensitivity to fg-SRLs are: female sex, older age, a slight excess of the levels of GH and IGF-1, the presence of GNAS mutation and DGST, high expression of the SSTR2, a low value of the proliferative index Ki-67, pronounced response to an acute test with octreotide, as well as the detection of hypointensive adenoma on T2-weighted MR images. On the contrary, in young male patients with a high level of GH at diagnosis, the presence of AIP mutation, SGST and hyperintensive tumor signal on T2-weighted images are associated with low sensitivity to fg-SRLs. It is also noted that miRNAs might influence tumor proliferation, invasion, size, and response to fg-SRLs [18, 19, 24, 34, 53, 54].

Although not routinely evaluated by clinical pathologists, SST expression may be a valuable molecular marker for precision-based acromegaly treatment, as a decision for repeat surgery, radiotherapy or choice of a specific SRL could be better guided with a knowledge of SSTR2 and SSTR5 expression According to Störmann et al. the best results of 12-month treatment of fg-SRLs were observed in patients older than 53 years, female, with high expression of the SSTR2. In the study of Durmus et al., it was shown that older passport age, low initial levels of GH and IGF-1 at the time of diagnosis, small tumor volume, hypo-intensity signal on T2-weighted MR images and the presence of densely granular adenoma allow predicting a good biochemical response to fg-SRLs [16, 55, 56].

Puig-Domingo et al. suggest that when choosing a treatment strategy, focus on the data of the intensity of the tumor signal on T2-weighted MR images in combination with the results of acute tests with octreotide and pasireotide, which makes it more likely to determine the most promising candidate. The author believes that the development of an acute pasireotide test will be of much aid in the context of personalized medicine [57].

Coopmans et al. multivariable regression models for predicting the biochemical response to fg-SRLs was proposed. In this study, the biochemical response was categorized as follows: biochemical response (IGF-I ≤ 1.3 ULN), partial response (>20% relative IGF-I reduction without normalization) and nonresponse (≤ 20% relative IGF-I reduction). As a result of a meta-analysis of 622 patients, it was shown that low baseline IGF-I levels and low body weight were the best independent predictors of the biochemical response to first-generation SRLs [AUC 0.77 (95% CI, 0.72–0.81)] [58].

In a retrospective analysis of the treatment results of 153 acromegaly patients by Tomasik et al., it was shown that the older age of diagnosis, male gender, low concentrations of GH, IGF-1 and prolactin, small tumor size, as well as the presence of a-SU, low index Ki-67 and DGST are the best independent predictors of good response to fg-SRLs. The cut-off value of fasting GH concentration predicting good response to fg-SRLs was estimated as GH < 36.6 μg/L with a sensitivity of 45.5% and specificity of 80.0%. The percentage of correctly classified patients was 52.4%. The AUC value was 0,788 [28].

The predictor algorithm of resistance to octreotide (and sensitivity to pasireotide) proposed by a number of authors includes: young age, male sex, large tumor size, high secretory activity, presence of SGST, high membrane expression SSTR5, low ratio between the SSTR2 and SSTR5, as well as low expression of zinc finger protein (ZAC1) regulating apoptosis and cell cycle arrest. Thus, molecular genotyping and identification of clinical and pathological markers are necessary tools for selecting the optimal therapeutic drug [59, 60, 61].

Kasuki et al. noted that high Ki-67 values (2.3% cut-off point) correlate with low sensitivity of fg-SRLs. The Ki-67 level was higher in SGST compared to DGST (p = 0.047). It was also shown that higher expression of the SSTR2 and the presence of DGST significantly correlated with sensitivity to fg-SRLs. The authors conclude that the value of Ki-67, as well as the expression of the SSTR2, can be independent predictors of the effectiveness of using fg-SRLs. During the regression analysis conducted by Gill et al., it was noted that the most significant prognostic biomarker determining sensitivity to fg-SRLs is E-cadherin, whose sensitivity was 73%. The second most important independent factor was the expression of the SSTR2. The combination of these 2 predictors increased the accuracy of the forecast by up to 80%. Thus, low expression of E-cadherin is a marker not only of aggressive behavior of the tumor but also of resistance to fg-SRLs. In a study by Puig-Domingo et al., the predictive value of E-cadherine, SSTR2 and Ki-67 was also confirmed [18, 62, 63].

5.3 Pharmacotherapeutic testing

It should be noted that in modern consensus agreements, fg-SRLs are recommended to all patients (without taking into account morphological diagnosis) as secondary or primary pharmacotherapy by “trial and error”. Since sensitivity to fg-SRLs, according to various authors, is observed in 40–50% of cases, it is not surprising that in some patients this appointment will be obviously ineffective due to a fundamentally different receptor phenotype of tumor cells. Hence, the question arises, what should be the optimal duration of trial treatment of fg-SRLs during traditional therapy in order to predict their long-term effectiveness?

To answer this question, we conducted a comparative analysis of the effectiveness of long-term primary (1) and secondary (2) DT fg-SRLs among 587 patients included in the Moscow Registry of Patients with Acromegaly, during which it was shown that in the selective group (with a decrease in IGF-1 of more than 70% of the baseline level after 3 months of treatment) biochemical remission was observed in 72 and 80% of cases compared with 47 and 51% in the non-selective group (p = 0.0002). The correlation coefficients between the percentage of reduction of IGF-1 after 3 months of treatment and dynamic indicators are presented in Table 1 (p = 0.000). These data indicate the practical importance of the selective selection of patients for the treatment of fg-SRLs [64].

		IGF-1 after 12 months (ng/ml)	IGF-1 at the end of treatment (ng/ml)	Duration of effective treatment (months)
% reduction of IGF-1 after 3 months of treatment fg-SRLs	1 DT	−0,57	−0,61	0,52
	2 DT	−0,51	−0,43	0,43

Table 1.

Correlation analysis of the long-term effectiveness of fg-SRLs depending on the severity of the decrease in IGF-1 3 months after the start of treatment.

In the subsequent work, a retrospective clinical and morphological comparison was carried out, which included 33 (12 men) patients with DGST and 47 (17 men) patients with SGST. The age of diagnosis was 48.4 ± 11.4 vs. 39.4 ± 12.7 years (p = 0.0027), the volume of the residual tumor was 1.6 ± 3.5 vs. 2.7 ± 4.8 cm³ (p = 0.2), the value of IGF-1 index before DT was 2.8 ± 0.8 vs. 2.7 + 0.9 (p = 0.6), respectively, [M(s)]. In the treatment, prolonged forms of lanreotide (Somatulin Autogel, s/c, 120 mg/28–56 days) or octreotide (Octreotide Depot, Octreotide Long, i/m at a dose of 10–30 mg/28 days) were used. The duration of DT was 21.5 ± 21.8 months. The adequacy of pharmacotherapy was assessed by the value IGF-1 index ≤1,0. The control points of pharmacotherapeutic testing were IGF-1 indicators before DT, as well as after 3, 6, 12 months of treatment and at the last visit.

The analysis confirmed good sensitivity to fg-SRLs in patients with DGST compared to SGST. The final value of IGF-1 index was 0.95 ± 0.27 vs. 1.4 ± 0.64 (p = 0.0002) (Figure 1).

Figure 1.

Dynamics of IGF-1 level against the background of fg-SRLs treatment of patients with DGST and SGST.

It was noted that the percentage of decrease in IGF-1 after 3 and 6 months of fg-SRLs treatment compared to the baseline level directly correlates with the number of SSTR2 according to the IRS (r = 0.44; r = 0.36), as well as the difference and the ratio between the SSTR2 and SSTR5 [r = 0.46; r = 0.46 and r = 0.41; r = 0.43; (p < 0.05)]. Against the background of fg-SRLs treatment, the decrease in the level of IGF-1 in the groups of patients with DGST and SGST after 3 months was 54.8 ± 19.6 vs. 28.4 ± 23.7%, after 6 months—58.4 ± 18.0 vs. 31.6 ± 24.5%, respectively (p = 0.0002). The presence of an inverse correlation was revealed between the indicators of a decrease in the level of IGF-1 after 3 and 6 months of fg-SRLs treatment and the final value of IGF-1 index [r = −0.59; and r = −0.72; (p < 0.001)]. According to the results of the ROC analysis, the AUC was 0.841 and 0.853, respectively, which confirms the good informative value of the selected diagnostic indicators. Cut-off prognostic points of effective DT were a decrease in the level of IGF-1 after 3 and 6 months to 46 and 49%, the sensitivity of these markers was 63 and 75% and specificity—79 and 80% (Figure 2). With a decrease in the level of IGF-1 after 3 and 6 months of less than 50%, the final IGF-1 index was 1.49 ± 0.62, whereas with a decrease of more than 50%, the average value of IGF-1 index in the long-term treatment of fg-SRLs was 0.9 ± 0.2; (p = 0.000). [Later, pegvisomant with a positive effect was added to the treatment of resistant patients with SGST. IGF-1 index decreased to 0.98 + 0.44].

Figure 2.

Significance of prognostic model of the reduction IGF-1 after 3 and 6 months of fg-SRLs treatment.

The data obtained confirm that the nature of the decrease in IGF-1 from the baseline level after 3 and 6 months of treatment reflects the severity of absolute or relative overexpression of the SSTR2 in tumor cells, and also indicates the intact postreceptor suppressive mechanisms. Thus, the results of pharmacotherapeutic testing can be used as an additional predictor of the effectiveness of long-term treatment of fg-SRLs and the determination of optimal therapeutic tactics to achieve stable control of acromegaly [65].

This conclusion is consistent with the data provided by M.R. Gadelha et al., who noted that the expression of mRNA of the SSTR2 correlated with a decrease in the levels of GH and IGF-1 after 3 and 6 months of octreotide treatment, as well as with a decrease in tumor volume. It was also noted that in the presence of DGST, the chances of a good response to fg-SRLs were more than 10 times higher than with SGST [19, 28].

Returning to the topic of precision medicine, it should be noted that the main limitation of the use of biomarkers for predicting the effectiveness of DT is the small number of patients included in the studies and the lack of unified methods for analyzing the expression of SSTR2 and other characteristics of tumor cells, which makes it difficult to interpret the results. According to M.R. Gadelha, it is time to change the paradigm of secondary DT by prescribing differentiated treatment taking into account the postoperative hormonal status and expression of SSTR2, SSTR5, AIP and DR2 in tumor cells [19].

In our opinion, the leading system-forming factor that allows the stratification of various clinical and pathological markers is a specific morphological subtype of somatotrophic adenoma, which should become a predictor of the 1st degree, reflecting the features of the biological functioning of autonomous tumor cells with more or less predictable behavior. The immunotypic characteristics, indicators of mitotic activity and features of the response to targeted therapy revealed in this case make it possible to determine the further treatment strategy and the optimal scheme of DT. Table 2 shows the most significant biomarkers reflecting the characteristics of the most common densely and rarely granular somatotrophic tumors (Table 2).

Predictors	DGST	SGST
Age of debut of acromegaly	more than 40 years	less than 40 years old
Tumor size	micro-, macroadenomas	macro-, giant adenomas
Localization of the tumor	intrasellar location	extrasellar distribution
Invasion of the cavernous sinus	not typical	typical, 3-4 gr Knosp
The tendency to residual growth and recurrence of the tumor	low	high
Intensity of the tumor signal on T2-weighted magnetic imaging	hypointensive	hyperintensive
Expression of the SSTR2	high	low
Expression of the SSTR5	low	high
Evaluation test with octreotide	positive	negative
Sensitivity to fg-SRLs	high	low
Reduction of IGF-1 after 3 and 6 months of treatment fg-SRLs	> 50%	< 50%
Proliferation index Ki-67	low (<3%)	high (>3%)
Expression of the α-subunit	high	absent
Cytokeratin expression	low	high
Expression E-cadherin	high	low
Expression ZAC1
GNAS-mutationis	positive	negative
AIP-mutationis	negative	positive

Table 2.

Clinical and biological predictors of DGST and SGST.

However, with all the apparent evidence of the situation, it is impossible not to mention the works that emphasize that the combination of sign predictors (expression of the SSTR2, the presence of a hypointensive signal on T2-weighted MRI, detection of DGST), although it increases the therapeutic prognosis, does not yet guarantee the effectiveness of long-term treatment of fg-SRLs. As a result of a meta-analysis conducted by S. Ezzat et al., it was shown that, even with good expression of the SSTR2, about half of the positive tumors did not respond clinically to fg-SRLs [66]. Since in this case, we are most likely talking about postreceptor disorders, then, among the predictors of sensitivity to fg-SRLs, the reaction of the adenoma itself to the administration of a targeted drug should become an obvious biomarker. It is known that the sensitivity of cells to fg-SRLs depends both on the severity of the receptor expression of the SSTR2 and on the functional integrity of multiple intracellular components of the postreceptor vector, which eventually blocks the pathological secretion of GH. Therefore, a positive reaction to octreotide with a marked decrease in the level of GH (and, accordingly, IGF-1) allows us to judge the complex integrity of the receptor and postreceptor mechanisms and, accordingly, the prospects of the planned long-term drug therapy.

Precision medicine aims to improve patient outcomes through targeted treatment employing genetic, biomarker, phenotypic, or psychosocial characteristics unique to each patient or the disease process. Of course, radiological and/or IHC confirmation of the presence of DGST is extremely important for determining the benign nature of the clinical course of adenoma but does not guarantee absolute sensitivity to fg-SRLs. Therefore, in our opinion, additional predictors of successful lifelong therapy of fg-SRLs should be the results of pharmacotherapeutic testing determining a more accurate treatment strategy. This tactic also allows timely identification of a group of patients with fg-SRLs resistance and aggressive course of the disease who need fundamentally different treatment.