by Nicholas A Tritos, MD, DSc
NEPTCC Newsletter Volume 21, Issue 1, Spring/Summer 2014 [PDF version]
Growth hormone deficiency (GHD) may eventually occur in the majority of patients with cured acromegaly as a consequence of pituitary surgery and/or radiation therapy (1).
Several studies have suggested that unreplaced GHD in patients with cured acromegaly (acroGHD) is associated with abnormal body composition (increased total and visceral adiposity), abnormal cardiac function, increased serum C reactive protein levels, and impaired quality of life (QoL) in comparison with patients with cured acromegaly who are GH sufficient (2-3). These findings suggest that the phenotype of patients with acroGHD may be similar to that of deficient adults whose GHD is caused by other etiologies and raise the possibility that GH replacement might be of benefit in this population.
The paradigm of replacement therapy in deficient patients with prior history of hormone excess is certainly familiar to endocrinologists, who routinely replace glucocorticoids in patients with Cushing’s disease that were rendered adrenally insufficiently as a result of pituitary surgery and/or radiation therapy. Similarly, levothyroxine replacement is advised in patients with Graves’ disease that developed primary hypothyroidism as a result of thyroid surgery and/or radioiodine therapy. So, should GH replacement be offered to patients with acroGHD?
The effects of growth hormone (GH) replacement in a population of patients with acroGHD have been examined in several studies, which have generally been of limited size and/or duration. Some studies have suggested beneficial effects of GH replacement in patients with acroGHD, including improvements in body composition, serum lipids and QoL (4-9). In contrast, other studies have found little evidence of benefit (10) or raised concerns regarding the cardiovascular safety of this therapy in patients with acroGHD (8).
To further evaluate the effectiveness and safety of GH replacement in patients with acroGHD, a team of investigators has recently analyzed data extracted from a large, international, pharmaco-epidemiological database of adult patients with hypopituitarism of diverse etiologies who have been receiving GH replacement (11).
Computerized searches identified an “effectiveness population”, including 115 adults with acroGHD and a control population of age, gender and body mass index (BMI) - matched patients, consisting of 142 adults with GHD and history of clinically non-functioning pituitary adenoma (NFPA). All patients had stringently defined GHD, based on the diagnostic criteria recommended by the Growth Hormone Research Society (12), and had been receiving GH replacement for at least 75% of the study period. Changes in serum lipids and QoL in response to GH replacement were retrospectively analyzed in this population.
In addition, the safety of GH replacement was assessed in the entire database population of patients with acroGHD on GH replacement (164 adults) in comparison with the entire database population of patients with GHD and NFPA (2469 adults) receiving GH replacement, as well as several external reference groups drawn from the general population [including the World Health Organization “Global Burden of Disease” (WHO GBD), “Cancer Incidence in Five Continents”, the Kronoberg County Study, the US National Health Interview Survey, the KORA S4/F4 and the Bruneck Study, as well as the MONICA Augsburg Cohort Study].
The median GH replacement dose was 0.3 mg daily in both groups (acroGHD and NFPA) at the end of the 5 year observation period and serum insulin-like growth factor 1 (IGF-1) standard deviation scores (SDS) were comparable and physiologic in both groups (Figure 1). There were also comparable decreases in total and low density lipoprotein (LDL) cholesterol levels in both groups (Figure 2A and 2B). Quality of life scores, assessed by the QoL-Assessment of Growth Hormone Deficiency in Adults (QoL-AGHDA) questionnaire, showed comparable improvements in both groups during the same period (Figure 3).
Reassuringly, safety analyses showed that all-cause mortality in patients with acroGHD was similar to that in the general population [ratio between observed/expected (O/E) cases (95% confidence intervals) = 1.32 (0.70 - 2.25)]. All cause mortality was lower in patients with NFPA [O/E ratio = 0.58 (0.48 - 0.70)] than the general population, perhaps reflecting a healthier group selected for GH replacement.
Somewhat less reassuringly, cardiovascular mortality was increased in patients with acroGHD [O/E ratio = 2.89 (1.16 - 5.92)] in comparison with the general population. In contrast, cardiovascular mortality was lower in patients with NFPA [O/E ratio = 0.68 (0.48 - 0.94)] than the general population, yielding a standardized mortality rate (SMR) ratio of 4.23 (1.89 - 9.47) between patients with acroGHD and NFPA (P = 0.0004). These findings remained robust to adjustments for history of hypertension, diabetes mellitus, dyslipidemia and cardiovascular disease at baseline, or after the exclusion of patients with cardiovascular or cerebrovascular disease at study entry. Of note, cerebrovascular mortality was not increased in either group (acroGHD or NFPA) in comparison with the general population.
The incidence of diabetes mellitus was higher in both groups (acroGHD and NFPA) than the general population [acroGHD patients: O/E ratio = 3.84 (2.31 - 5.99); NFPA patients: O/E ratio = 3.86 (3.43 - 4.33)], and increased with higher BMI at study entry. However, there was no difference in the incidence of diabetes mellitus between these two groups. Similarly, there was no difference between patients with acroGHD and NFPA with regards to the incidence of all malignancies combined, benign or malignant brain tumors, incident cardiovascular or cerebrovascular disease.
What can we make of these findings? The data from the present study affirm several possible benefits of GH replacement in patients with acroGHD, including potentially salutary effects on serum lipids and QoL. However, body composition (fat mass, lean body mass and bone mineral density), cardiovascular function or exercise capacity were not assessed in this study, owing to lack of sufficient data in the database.
With the exception of a possible increase in cardiovascular mortality, the findings of this study suggest that the safety of GH replacement in patients with acroGHD is comparable to that in patients with GHD and NFPA.
It should be kept in mind that patients with active GH excess are at increased risk of cardiovascular morbidity and mortality (13-14). It remains unclear whether the increased cardiovascular mortality among patients with acroGHD receiving GH replacement, noted in the present study, is associated with their history of previous GH excess or with subsequent GH replacement. It may also be noted that the present study represents a retrospective analysis of observational data rather than a randomized clinical trial, which is the gold standard for assessing efficacy and safety. In addition, a control population of patients with acroGHD, who were not receiving GH replacement, was not available in the database. Inclusion of such a control population would likely be helpful in order to further elucidate the cardiovascular safety of GH replacement in patients with acroGHD and should be considered in future studies.