Continuous Glucose Monitoring System in Acromegalic Patients: Possible Role in the Assessment of Glycemia Control

Abstract

Background:

Acromegaly is characterized by an insulin resistance condition. There is a significant difference between the different types of therapy in relation to the glycometabolic framework. The blinded continuous glucose monitoring system (CGMS), throughout a period of maximum 6 days for a total of 288 glycemic registrations per day, identifies glycemic excursions and could constitute a valid device to understand the 24-hour glycemic profiles.

Aim of the study:

To compare the oral glucose tolerance test (OGTT) and CGMS methods in acromegalic patients to evaluate their glycemic profiles, in relation to different treatments for acromegaly.

Methods:

Thirty-five acromegalic patients were divided into 18 somatostatin analogs (SSA), 9 pegvisomant, and 8 successfully surgically treated. A 72-hour CGM was performed and, immediately after, an OGTT.

Results:

Results obtained from OGTT: 11/35 impaired fasting glucose, 6/35 impaired glucose tolerance, and 4/35 diabetes mellitus. A positive significant correlation was demonstrated between the OGTT peak and CGM peak in all of the patients, CGM peak of patients treated with SSA and those surgically treated, OGTT average and CGM area under concentration–time curve (AUC) for hyperglycemia of patients treated with SSA and those surgically treated, and CGM AUC for hyperglycemia of patients treated with SSA and those surgically treated.

Conclusions:

Our results show a significantly higher response in terms of mean and peak OGTT in patients treated with SSA, both compared to the CGM study, and compared to the group of patients receiving pegvisomant. The CGM system could represent an instrument for the evaluation of the glycemic trend of acromegalic patients.

Introduction

Growth hormone-secreting pituitary adenomas are characterized by excessive growth hormone (GH) and insulin-like growth factor-1 (IGF-1) secretion, which consequently results in a series of metabolic disorders.^1,2 Glucose metabolism alterations, including diabetes mellitus (DM), impaired fasting glucose (IFG), and/or impaired glucose tolerance (IGT), are recognized as one of the most frequent acromegaly complications with a prevalence of 25%-50%, according to the different study groups. The prevalence is related to familiarity, GH concentrations, and duration of disease. Glucose intolerance further contributes to increased cardiovascular risk and mortality.^1,2

Acromegaly is characterized by an insulin resistance condition, due to the GH hypersecretion, which is associated with increased lipolysis and reduced total body fat.^3,4

Glycometabolic homeostasis is controlled by several neurohormonal factors; among these, the somatotropic hormone plays an important role: it seems to have a double effect over time on the glucose metabolism. Immediately after the administration of GH, there is a decrease in glycemic levels following an increased peripheral use of glucose; during this short first phase, GH performs an insulin-like action. At the same time there is an increased esterification of free plasma fatty acids on adipose tissue (with a decrease in circulating free fatty acids [FFAs]).^3,4

GH has two principal metabolic effects: (1) a direct stimulatory action on the islet beta cells; (2) an increased peripheral glucose utilization, secondary to a greater permeability of the cell membranes of the target tissues. Subsequently, however, there is an increase in blood glucose levels linked to a decreased peripheral use of glucose; in this second phase, of longer duration, GH exerts an anti-insulin action.^3,4 Acromegalic patients show hepatic and peripheral insulin resistance and hyperinsulinism. In addition, the excess of GH inhibits the hepatic suppression of gluconeogenesis induced by insulin.

The excess of IGF-1 seems to be involved in the onset of hyperglycemia in acromegalic patients, promoting insulin resistance by lipolysis, stimulating the gluconeogenesis, and blocking the signaling of insulin mediators like p85 alpha and insulin receptor substrate.^5,6

Transsphenoidal adenectomy is the first-line treatment for acromegaly. GH and IGF-1 levels decline rapidly and sharply after successful surgery, which normalizes the glucose metabolism in a lot of patients with preoperative diabetes. Glucose metabolism improves in almost half of acromegaly patients with glucose intolerance. Successful surgical removal of somatotroph tumors improves hyperglycemia, and glucose metabolism has been reported to normalize in 23%-58% of people with preoperative diabetes, after surgical cure of acromegaly.^7-9 Somatostatin analogs (SSA) are first-line drug therapy in patients with macroadenoma and/or cardiovascular or respiratory complications. They act by inhibiting the excessive production of GH and IGF-1 and by binding themselves to somatostatin receptors, which are present on cardiac cells (myocytes and fibroblasts).^10,11

In recent years, pharmaceutical research has led to the development of molecules capable of expanding the receptor spectrum of action. Pasireotide LAR (PAS) is a universal somatostatin receptor ligand (SRL) capable of binding receptor subtypes 1, 2, 3, and 5 with high affinity. It is also able to act on tumors that do not express or have low SST2 receptor expression and are resistant to octreotide. Furthermore, by binding to multiple receptors, PAS could facilitate a receptor cooperation process with enhancement of biological effects. It is well tolerated and has a safety profile comparable to that of first-generation SRLs, except for a greater frequency and degree of hyperglycemia-related adverse events. Pasireotide inhibits both insulin secretion and incretin hormone responses and modestly suppresses glucagon secretion. The effects of pasireotide on insulin, glucagon, and incretin secretion can be explained from its SSTR binding profile. SSTR5 is known to be expressed on pancreatic b-cells, which mediate insulin secretion, but also on enteroendocrine L-cells, which produce glucagon-like peptide 1. ¹²

Pegvisomant (PEG), a GH receptor antagonist, blocks the conformational change required for dimerization by binding itself to the GH receptor; the internalization of the ligand still takes place, but there is no signal transduction and the hepatic production of IGF-1 is inhibited. It is indicated in acromegalic patients nonresponsive to SSA.^13,14 Furthermore, there is a significant difference between the different types of therapy in relation to the glycometabolic framework: SSA therapy would seem to lead to an impairment in glucose regulation since the drug inhibits the secretion of insulin and glucagon, in addition to that of GH. ¹¹ Hyperglycemia seems to be more frequently associated with the use of pasireotide than other analogs. This still represents a significant problem in the therapeutic choice. ¹² Therefore, patients with a glycemic control considered optimal, assessed exclusively with blood fasting glucose and glycosylated hemoglobin (HbA1c), could instead undergo high glycemic peaks and/or hypoglycemias.

It has been recently demonstrated that the continuous glucose monitoring system (CGMS) is a useful and valid tool in defining glucose metabolism in patients with diabetes, in particular with early glucose derangements. Indeed, the CGMS allows monitoring of glycemic profiles throughout a period of maximum 6 days for a total of 288 glycemic registrations per day. It identifies glycemic excursions and constitutes a valid device to understand the 24-hour glycemic trend and profiles.^15,16

The aims of the present study were to compare the oral glucose load (OGTT) and CGMS methods in the diagnosis of altered glucose metabolism in acromegalic patients with a good biochemical control of GH secretion, and to evaluate their glycemic profiles, in relation to different treatments for acromegaly: SSA (lanreotide, octreotide, and pasireotide), PEG, and neurosurgery. The evaluation was performed using CGMS (iPro2 Professional CGM, Medtronic, Northridge, CA, USA) with the intent to determine appropriately the impact of the different drugs on the glycemic metabolism. Moreover, it was our interest to investigate the presence of any hypoglycemic episodes in SSA-treated patients, since the glucagon secretion would also appear to be affected with consequent loss of the counter-regulatory effect.

Patients and Methods

Thirty-five acromegalic patients who regularly attended the Department of Experimental Medicine, Endocrinology, “Sapienza” University of Rome were enrolled.

The diagnosis of acromegaly was performed on the latest available guidelines, which, in patients with elevated or equivocal serum IGF-1 concentrations, recommend confirmation of the diagnosis by finding lack of suppression of GH to <0.4 μg/L, following documented hyperglycemia during an OGTT (2 hours after 75 g of oral glucose).^17,18

Biochemical control of disease was defined on the determination of IGF-1, age-related and expressed in ng/mL.^17,18 The GH dosage was not used as a disease control parameter since there were patients being treated with PEG.

In relation to the therapy subgroups, the population of selected patients was homogeneous concerning the dosage of the administered therapy and the interval of administration.

The selected patients were all in biochemical control for acromegalic disease, nondiabetic, homogeneous for body mass index (BMI), duration of disease, and diagnostic latency. Patients aged over 70 or under 20, with previous diagnosis of DM, uncontrolled acromegaly, and concomitant therapies that could interfere with the glycemic profile, were not included in the study.

In patients treated with SSA, a 72-hour CGM was performed at the fourth week after the administration of the drug and, immediately after the glycemic monitoring, they simultaneously underwent OGTT.

The CGMS includes a sterile disposable subcutaneous sensor, an external electrical transmitter, and a connection cable to a communication device that allows the downloading of the data on the computer. The transmitter analyzes the data every 10 seconds and register an average value every 5 minutes. The communication sensor consists of a platinum microelectrode with a thin coating of glucose oxidase under several layers of biocompatible membrane. The sensor was inserted into the subcutaneous tissue of the anterior abdominal wall, using a spring device and an introducer needle. An electric current is generated by glucose oxidase such as to catalyze the oxidation of glucose in the interstitial fluid. The data are stored as an electronic signal by the transmitter and the signal strength is proportional to glucose levels. Capillary blood glucose measurements (a minimum of two measurements per monitoring day) were obtained with a precision glucometer and were used to calibrate the sensor readings. The data were then downloaded using Carelink iPro software. ¹⁹

From the CGM, the following data were extrapolated for all the patients:

• fasting blood glucose value at 6 o’clock on the first day of monitoring;

• hyperglycemic peak value;

• total average blood glucose monitoring value.

The OGTT was carried out with a 75-g glucose load and blood samples were collected at 30, 60, 90, and 120 minutes for the measurement of glucose and insulin (Trinder method).

The HbA1c was also measured (high-performance liquid chromatography method).

The glycometabolic derangements have been defined as IFG, IGT, and DM, using the current criteria. ²⁰

Statistical Analysis

Descriptive analysis was performed using absolute and relative frequencies to describe qualitative variables while mean and standard deviation were calculated for quantitative variables. A univariate analysis was successively carried out using the following tests: χ² to compare proportions and Student’s t, or analogous nonparametric tests Mann–Whitney and Wilcoxon, to assess differences between quantitative variables, as appropriate.

Quantitative paired data were analyzed by analysis of variance for repeated measures to evaluate differences between different therapeutic groups.

A two-tailed P-value of .05 was considered significant for all analyses, which were performed using Stata 15.1 (Stata Corporation, College Station, TX, USA, 2017).

Results

The 35 acromegalic patients enrolled (17 males and 18 females; mean age 56 ± 13.1 years) were well selected for BMI, duration of disease, and diagnostic latency and were divided according to the type of therapy: 18 SSA (15 octreotide/lanreotide, 3 pasireotide), 9 PEG, and 8 successfully surgically treated (Table 1).

Table 1.

Clinical and Anamnestic Data.

Clinical/anamnestic data	Study population
Male/Female	17/18
Age (years)	56 ± 13.1
BMI (mean) (kg/m2)	28 ± 2.7
HOMA-IR (mean)	2.2 ± 1.2
IGF-1 (mean) (ng/mL)	256 ± 36
Duration of disease (years)	12 ± 9.7
Diagnostic latency (years)	7 ± 5.3
SSA:
OCTR/LANR	15
PAS	3
PEG	9
Neurosurgery	8

BMI, body mass index; HOMA-IR, Homeostatic Model Assessment of Insulin Resistance; IGF-1, insulin-like growth factor-1; PAS, pasireotide; PEG, pegvisomant; SSA, somatostatin analogs (OCTR/LANR, octreotide/lanreotide).

In the subgroups of patients included in the study, all with a good biochemical control of acromegaly, after OGTT we found that (Table 2):

Table 2.

Glycometabolic Profile of the Population Examined.

Subgroups of therapy	Diagnosis (OGTT)
SSA:
OCTR/LANR	7 IFG, 4 IGT, 1 DM, 3 NG
PAS	1 IGT, 2 DM
PEG	2 IFG, 1 DM, 6 NG
Neurosurgery	2 IFG, 1 IGT, 5 NG

DM, diabetes mellitus; IFG, impaired fasting glucose; IGT, impaired glucose tolerance; NG, normoglycemic; OGTT, oral glucose load; PAS, pasireotide; PEG, pegvisomant; SSA, somatostatin analogs (OCTR/LANR, octreotide/lanreotide).

• Eleven out of 35 patients showed an IFG. Of these, seven were in therapy with SSA, two with PEG, and two without therapy;

• Six out of 35 patients showed an IGT. Of these, five were in therapy with SSA (four octreotide/lanreotide and one pasireotide) and one without therapy;

• Four out of 35 patients had a diagnosis of DM. Of these, three were in therapy with SSA (one octreotide/lanreotide and two pasireotide), one on PEG.

We have tried to use CGM data in order to classify patients as those with and without IFG, IGT, or DM: the parameters we chose for this classification are derived from recent scientific literature^21,22 and from our clinical experience (data not validated):

• fasting glucose (as derived by CGM profiles between 6 and 7 AM) (IFG > 100 mg/dL);

• the mean of the individual average 24-hour glucose <104 mg/dL in people >60 years (as our patients);

• the median percentage of time spent between 0 and 140 mg/dL (<93% in people >60 years).

No statistically significant differences were found with respect to the diagnoses obtained through the OGTT. Different data could derive from an extension of the sample examined.

The average blood glucose of the OGTT was significantly higher than that of the CGM, in all of the patients (P = .0051) and in the comparison between the subgroup of patients treated with SSA and that of patients treated with PEG (P = .0019) (Figure 1).

Figure 1.

Comparison of the methods used.

CGM, continuous glucose monitoring; OGTT, oral glucose load; PEG, pegvisomant; SSA, somatostatin analogs (OCTR/LANR, octreotide/lanreotide).

A positive significant correlation was demonstrated between the:

• OGTT peak and CGM peak in all of the patients (P = .0009) (Figure 2);

• CGM peak of patients treated with SSA and those surgically treated (P = .0021);

• OGTT average and CGM AUC for hyperglycemia of patients treated with SSA and those surgically treated (P = .0016);

• CGM AUC for hyperglycemia of patients treated with SSA and those surgically treated (P = .05).

Figure 2.

Direct correlation between OGTT peak and CGM peak in all patients.

CGM, continuous glucose monitoring; OGTT, oral glucose load.

No significant difference was demonstrated between the average HbA1c values comparing the different therapeutic subgroups.

Discussion

Our results confirm that SSA interfere with insulin response to OGTT: the mean blood glucose concentration during glucose tolerance test is higher than during CGMS in acromegalic patients treated with SSA. ²³ This drug has a double effect on the glucose metabolism: it reduces insulin resistance and glucose production by decreasing GH concentration; on the other hand, it exerts an inhibitory action against both insulin and glucagon secretion, interfering with the glycemic response in relation to a stimulus quite impressive, such as the OGTT. ²³

This does not emerge from the CGMS, which represents a more physiological method to study the glycemic trend.^24-26

In this regard, the data in the literature are still conflicting: some researchers show a modest worsening of the glycemic compensation, since the reduced insulin secretion is offset by the improvement in insulin sensitivity; others show a deterioration in the glycemic compensation during therapy with analogs.^27,28 The data of our study suggest that the improvement of insulin sensitivity leads to a stability of fasting blood glucose values, counterbalancing its reduced secretion. We have excluded the presence of fasting hypoglycemia in SSA-treated patients, that we hypothesized could occur in these patients for a reduction in glucagon concentrations with a consequent deficit of the counter-regulation.^3,4,11

Although a limitation of our study is the lack of an adequate number of patients on pasireotide therapy, in order to compare homogeneous therapeutic subgroups, our results nevertheless clearly show a significantly higher response in terms of mean and peak OGTT in patients treated with SSA, both compared to the CGM study, and compared to the group of patients receiving PEG.

An improvement in carbohydrate homeostasis in acromegalic patients treated with GH receptor antagonist is known in the literature and it is related to the improvement in hepatic insulin sensitivity induced by PEG, resulting in improved glucidic metabolism and unaltered FFA kinetics and lipid oxidation. ²⁹ We can suggest that PEG seems to assume a protective role against the counter-regulating physiological action of GH by interfering with its receptor.

A similar observation was provided also by Urbani et al, who showed that the introduction of PEG, as compared with SSA, ameliorates glucose metabolism in partially controlled acromegalic patients. ³⁰

Therefore, the CGM system, combined with traditional methods (OGTT and HbA1c), could represent a valid instrument for the evaluation of the glycemic trend of acromegalic patients.

This new setting seems to offer a safer glycemic monitoring strategy especially in those patients affected by resistant acromegaly who benefit, in terms of biochemical control of acromegalic disease, from the use of pasireotide, a drug described as a deteriorating factor of glycemic homeostasis in over 57% of cases.^12,28

Conclusion

The diagnostic value of OGTT in acromegalic patients is confirmed by this study but a further important use of CGMS has been demonstrated, as a new glycemic monitoring device for clinical therapeutic evaluation in acromegalic patients.

CGM technology usually helps clinicians to obtain more complete insights into patients’ glucose profiles. When used for driving therapeutic decisions or for lifestyle recommendations in patients with diabetes, CGM data can lead to improved glucose control, while if a diagnostic use is hypothesized in patients without diabetes, it can facilitate early identification of glucose profile derangements. It identifies glucose excursions and constitutes a valid device to understand the 24-hour glycemic trend and profile.

The use of CGM in acromegalic patients could represent a valid support for the specialist in the choice of the drug, as reaching the therapeutic goal in the shortest time, without fear of underestimating a metabolic failure, is essential to prevent systemic complications.