Effects of Growth Hormone Receptor Antagonism and Somatostatin Analog Administration on Quality of Life in Acromegaly

Laura E. Dichtel; Allison Kimball; Kevin C. J. Yuen; Whitney Woodmansee; Melanie S. Haines; Qiu Xia Guan; Brooke Swearingen; Lisa B. Nachtigall; Nicholas A. Tritos; Julie L. Sharpless; Ursula B. Kaiser; Anu V. Gerweck; Karen K. Miller

Disclosures

Clin Endocrinol. 2021;94(1):58-65. 

In This Article

Results

Clinical Characteristics

Demographics, treatment history, medical comorbidities and laboratory values are reported by group in Table 1. There were no group differences in mean demographic variables between the ACTIVE, SSA and PEG groups, including age, sex, BMI and race. Per study design, mean IGF-1 levels and Z-scores were higher in the ACTIVE group (678 ± 221 ng/mL and 3.9 ± 1.0 SD) than the SSA (174 ± 51 ng/mL and 0.5 ± 0.7 SD, P < .0001) and PEG groups (176 ± 58 ng/mL and 0.5 ± 0.7 SD, P < .0001). Neither mean IGF-1 levels nor IGF-1 Z-scores differed between the SSA and PEG treatment groups. The mean time since diagnosis of acromegaly was significantly shorter in the ACTIVE group (1 ± 3 years) than the SSA (9 ± 7 years) and PEG (14 ± 10 years) groups (P < .0001 ACTIVE vs SSA and PEG). There was a trend towards longer mean time since acromegaly diagnosis in the PEG group than in the SSA group (P = .06). Patients in the SSA (n = 53/55, 96%) and PEG (n = 28/29, 97%) groups had a similar proportion of surgical intervention that was significantly higher than that of the ACTIVE group (n = 4, 13%; P < .0001 for ACTIVE vs SSA and PEG). The proportion of patients who underwent pituitary radiation therapy followed this same pattern (SSA n = 14, 26%; PEG n = 12, 41%; and ACTIVE n = 0, 0%; P < .001 for ACTIVE vs SSA and PEG).

There was a higher prevalence of hypothyroidism in the SSA and PEG groups vs the ACTIVE group (47% and 52% vs 6%, P < .0001). There was a higher incidence of male hypogonadism in the PEG group (88%) compared to the ACTIVE (33%, P = .01) and SSA groups (43%, P = .03). The percentage of premenopausal women did not differ across groups (47% in ACTIVE, 34% in SSA and 62% in PEG, P = .2). Among the premenopausal women, the incidence of hypogonadism was 13% in the ACTIVE group, 64% in the SSA group and 31% in the PEG group (P = .08). Gonadal steroid replacement is detailed for men and women in Table 1. There was no difference in incidence of adrenal insufficiency, daily glucocorticoid replacement dose equivalent or diabetes insipidus across groups. Few subjects had diabetes mellitus; there was no statistically significant difference in prevalence between groups (ACTIVE 19%, SSA 11%, PEG 24%; P = .26). There was no difference in mean HbA1c between groups (ACTIVE 6.0 ± 0.9, SSA 6.1 ± 0.9, PEG 5.8 ± 0.6; P = .26). There was no difference in the presence of depression, hypertension, hyperlipidemia or history of malignancy across groups.

QoL Comparisons Across Groups

There were no differences in the four QoL primary end-point measures between groups (Table 2). There were no differences in any QoL domain between the SSA and PEG groups. When compared to all treated subjects (SSA and PEG combined, n = 84), the ACTIVE group had poorer QoL in the AcroQoL Appearance Domain (ACTIVE 38 ± 26 vs SSA/PEG 50 ± 23, uncorrected P = .0005) and the GIQLI Emotions Domain (ACTIVE 2.4 ± 0.9 vs SSA/PEG 2.9 ± 0.9, uncorrected P = .02). The ACTIVE group also had poorer QoL in the AcroQoL Appearance Domain vs the SSA and PEG groups individually (ACTIVE 38 ± 26 vs SSA 51 ± 25 [uncorrected P = .001] and vs PEG 49 ± 19 [uncorrected P = .004]). The ACTIVE group had a poorer QoL score in the GIQLI Emotions Domain compared to the SSA group only (2.4 ± 0.9 vs 3.0 ± 0.8, uncorrected P = .006), but neither the ACTIVE nor SSA group differed from the PEG group (2.6 ± 0.8) in this domain. After correction for multiple comparisons (n = 21), QoL remained poorer in the AcroQoL Appearance Domain in ACTIVE vs combined SSA/PEG (corrected P = .01) and in ACTIVE vs SSA (corrected P = .02), and there was a trend towards poorer QoL in ACTIVE vs PEG (corrected P = .08). None of the GIQLI Emotions Domain comparisons reported above remained significant after correction for multiple comparisons.

Determinants of QoL

On univariate analysis in the entire cohort, several QoL domains were significantly negatively associated with HbA1c (SF-36 Physical Component Summary Score, GIQLI Global Score, and GIQLI Social Function Score), BMI (SF-36 Physical Function Score) and IGF-1 Z-score (AcroQoL Psychological and Appearance Domains) after controlling for multiple comparisons (Table 3). Age, total and free testosterone, and free T4 were not significantly associated with QoL. Individuals with prior radiation treatment did not have worse QoL scores on average than those with no history of radiation. Similarly, individuals with any pituitary hormone deficiency did not have worse QoL scores on average than those without any pituitary hormone deficiencies.

There were no differences in the four primary outcome QoL measures between groups when controlling for HbA1c, IGF-1 Z-score and BMI. The SSA group had worse QoL in the SF-36 Physical Function domain compared with both the PEG group; however, this did not remain significant when controlling for multiple comparisons.

Impact of Glucose Homeostasis on QoL in Participants Without Diabetes Mellitus

A subset of the subjects without diabetes mellitus had fasting insulin and glucose measurements (ACTIVE n = 22, SSA n = 29 and PEG n = 16). HbA1c and fasting glucose were not significantly different across groups in this subset (Table 1); however, fasting insulin and HOMA-IR were lower in the SSA vs PEG and ACTIVE groups, which did not differ from each other. QoL in this subset was negatively correlated with fasting insulin (4 domains), HOMA-IR (2 domains) and fasting glucose (1 domain), but none of these associations remained significant after correction for multiple comparisons.

A smaller subset of individuals without diabetes underwent OGTT testing (ACTIVE n = 10, SSA = 16 and PEG n = 5). There were no differences across groups in glucose or insulin at any time points (0, 30, 60, 90 and 120 minutes) or in insulin or glucose AUC calculations.

processing....