The Treatment of Growth Hormone Deficiency in Adults

ADULT GH deficiency is recognized as a specific clinical syndrome (1). Well-controlled clinical studies have shown that GH deficiency in adults is associated with a cluster of cardiovascular risk factors (2). These patients have an increased amount of body fat with an abdominal preponderance (3, 4), an increased prevalence of hypertension (5), impaired fibrinolysis (4), glucose intolerance (6), insulin resistance (7), and premature atherosclerosis (6). This cluster of cardiovascular risk factors associated with GH deficiency might explain the doubling in cardiovascular mortality observed in hypopituitary patients (8, 9). The quality of life is impaired, and there are data suggesting an increased prevalence of disability pensions among these patients (10, 11). In addition, these patients have reduced bone mineral content and density with increased prevalence of fractures (12, 13). Randomized placebo-controlled studies indicate that GH replacement therapy can reverse several of these biological changes (3, 14–18). Although these studies have been of relatively short duration, replacement therapy with GH in adults has now been approved by the regulatory authorities in Europe and the United States. The long-term experience is, however, limited. The optimum goal for GH replacement in adults is to improve and normalize the aberrations associated with adult GH deficiency. One cornerstone in releasing this objective is well-defined normative values for the efficacy measurements that are used, particularly during more prolonged GH treatment when dose adjustment may be needed because of increasing age. No golden standard for the monitoring of GH treatment is available, as a specific marker of its tissue effect is lacking. Previous treatment trials in adult patients with GH deficiency have used doses of GH based on body weight or body surface area, adopted essentially from the experience of treating GH-deficient children, and have, therefore, ignored the presence of an individual responsiveness to GH (19). Some of these trials have reported a high frequency of side effects, mainly associated with fluid retention (3, 17). These side effects were more prone to occur in older patients, and in patients with a higher body mass index (20, 21), and may explain why trials comprising young and mostly lean adults have reported few or no side effects (14, 22). In healthy adults, it is known that fertile women have a higher GH secretion than young men (23) and that GH secretion is inversely associated with increasing age and adiposity (24, 25). These factors should be considered during GH replacement. Instead, GH dose schedules based on body weight or surface area results in higher daily doses of GH in subjects with higher body weight and lower for most women compared with men. In middle-aged women, the reported daily production of GH is about 47μ g/L × 24 h, whereas the mean production in the adult male is 15 μg/L × 24 h. Assuming an availability of sc administered recombinant human GH of 60%, a dose between 0.3–0.6 mg (0.6–1.8 IU) per day should be in agreement with the physiological GH production in adults (26), which is markedly lower than the doses of GH used in previous trials of GH replacement in adults. A highly variable response to GH treatment in adult GH deficiency has recently been documented. The variability was, to some extent, explained by differences in baseline body mass index (BMI), baseline serum levels of the GH binding protein (GHBP), age, and gender. In response to GH treatment, patients with a high BMI experienced a lower reduction in body fat and patients with low baseline levels of GHBP and younger patients displayed the most marked increase in lean body mass and lean to fat ratio (19). An inverse relationship between the baseline concentration of GHBP and changes in body composition in response to GH treatment is in contrast to previous experiences in children (27), where the baseline level of GHBP correlates positively with the response to GH administration in terms of growth velocity and insulin-like growth factor (IGF) I increment. The observation is, however, in line with in vitro experiments that demonstrate a dampening effect by GHBP on the binding of GH to cells and on the GH-dependent IGF-I production (28, 29). A gender difference in the response to GH treatment has been observed. The increase in fat-free mass, total body water, and decrease in body fat is more marked in men than in women (19, 30). In addition, men increase their lipoprotein (Lp) (a) levels more than women (31), and women demonstrate a more marked increase in total bone mineral density and a less pronounced increment in bone formation and bone resorptions markers than men in response to the GH treatment (32). This gender difference is most probably explained by differences in interaction between GH and estrogens and androgens, respectively. Adult GH deficiency is not a single clinical entity because the onset may occur during childhood (CO) or in adult life (AO). GH deficiency occurring in children is mainly idiopathic and is recognized as growth failure. Appropriate therapy with GH, standardized over the past decade, is discontinued when final height is reached. Somatic development is unlikely to be complete at this stage. Thus, height, body weight, BMI, lean body mass, and waist to hip ratio were found to be higher in AO than in CO GH-deficient patients. Serum IGF-I concentration was lower in CO than in AO GH-deficient patients (19, 33). Both groups displayed significant psychosocial distress, but the deviation from normality was greater in AO patients (33). The apparent differences in body composition at baseline between subjects with AO and CO GH deficiency is probably explained by the lower body height in CO (19). Other consequences of GH deficiency have been suggested to be more pronounced in adults with CO, particularly in terms of heart structure and function (34, 35), muscle mass (14, 36), and bone mass (12, 37). This may be an effect of the immediate discontinuation of GH treatment after the final height is reached (38). Whether this per se makes the CO patients more or less responsiveness to GH is not known (19). There is no optimal marker for long-term physiological replacement dose of GH. It is unlikely that GH treatment can be adequately monitored by clinical means alone. Hence, various GH-dependent biochemical markers have been suggested, such as IGF-I. IGFs circulate bound to specific binding proteins [IGF binding proteins (IFGBPs)], which regulate their bioavailability and bioactivity. To date, six such IGFBPs have been characterized. IGFBP-3 circulates in highest concentrations and binds to 90% of circulating IGFs to form 150-kDa ternary complexes with an acid-labile subunit (ALS) (39). Both IGFBP-3 and ALS levels are GH dependent. In a study of adult men in the age range of 20–40 yr with CO GH deficiency, the minimal dose of GH required for normalization of serum IGF-I concentration was 0.22 mg/m2·day and 0.33 mg/m2·day for ALS and IGFBP-3. In patients receiving 0.67 mg/m2·day, mean serum IGF-I concentration rose to abnormally high levels, whereas at this dose mean serum IGFBP-3 and ALS concentrations were not different from normal. The lower sensitivity of IGFBP-3 and ALS to GH doses in the high range was also apparent during long-term treatment (40). A dose of 0.2, 0.4, or 0.6 mg per day was used in 51 AO GH-deficient patients. In female patients, a dose of 0.2 mg/day was insufficient to normalize serum IGF-I after 12 weeks of treatment, whereas doses of 0.4 and 0.6 mg/day were able to normalize serum IGF-I levels in most female GH-deficient patients. In the male patients, a dose of 0.2 IU/day was in almost all patients sufficient to normalize serum IGF-I levels (41). Serum IGF-I may, however, be clinically inadequate in reflecting the GH status of the individual patient because a substantial number of patients with severe GH deficiency have normal serum IGF-I concentrations (42, 43). It should also be recognized that GH delivered by a more continuous pattern increases serum IGF-I concentration more than a pulsatile secretion of GH (44), whereas the pulsatile pattern may in some respects produce a more marked tissue response (45). The usefulness of serum IGF-I is further complicated by the fact that there is a diurnal variation during daily sc evening injections of GH, irrespective of whether the patients have been treated for a short or long period of time. A serum sample obtained in the morning from a GH-deficient subject given GH in the evening would overestimate the true 24-h integrated serum concentrations of IGF-I by about 14–17% (46). During more prolonged treatment over 2 yr, increased sensitivity to GH was seen, as the concentration of IGF-I remained unchanged during simultaneous 20% dose reduction of GH (20). Changes in body composition have been a remarkable consistent finding in all treatment trials with GH in adult GH deficiency. Assessing body composition could, therefore, be a tool for optimizing the replacement dose and to monitor the long-term treatment effects. Bioelectric impedance analysis is a simple technique to assess body water. By applying this method in young adult males with GH deficiency it was found that an average dose of GH of 0.37 mg/m2·day resulted in a rapid normalization of whole body resistance. Doses equal to or higher than 0.67 mg/m2·day resulted in abnormally low resistance, suggesting overhydration (47). An incongruity between the normalization of serum IGF-I concentration and body composition has been observed. In a 2-yr study, no relationship was found between the changes in serum IGF-I concentration and the changes in body composition (20). Moreover, in that study the initial dose of GH was based on the patient’s body weight, but after dose adjustment due to side effects and serum levels of IGF-I above the age and sex adjusted reference values there was no relationship between the dose of GH and body weight, indicating the importance of factors other than body weight for the daily dose of GH. In a study of 60 adult GH-deficient patients an individualized dose schedule was used with a low starting dose (0.17 mg or 0.33 mg per day), independent of body weight. The dose was titrated according to clinical response, age- and sex-adjusted reference values for serum IGF-I, and according to age-, sex-, and body weight-adjusted reference values for body composition. Within a year, the average GH dose was 0.45 mg per day. Compared to a matched group receiving an established GH dose according to body weight (12μ g/kg·day), similar mean response to GH in terms of changes in body composition, glucose homeostasis, Lp(a) and blood pressure was obtained in both treatment groups, with significantly fewer side effects in the group receiving the individualized dose schedule. By using the individualized dose schedule for GH treatment, it was found that women were given more GH per kg of body weight than men and older patients less GH per kg of body weight (48). This is well in accordance with our knowledge about individual responsiveness to GH and physiological GH secretion in healthy adults. The previously noted incongruity between the normalization of serum IGF-I concentration and body composition was also obvious in this trial (48). Consequently, individualized GH doses enabled one third of the patients to normalize both their serum IGF-I and body composition, but overtreatment was also observed. This demonstrates the large variance and lack of precision of an individual measurement used to monitor GH replacement. Although several disadvantages, the normalization of serum IGF-I concentration is probably the most useful marker of GH treatment (40) and may better reflect overtreatment than other serum markers and the measurement of body composition. It has been suggested that lower doses/levels of IGF-I stimulate bone formation more markedly than bone resorption (49). Furthermore, IGF-I may be more potent in enhancing the formation of collagen type I and less effective when it comes to stimulating bone resorption than GH (50). It has, therefore, been suggested that lower doses of GH may generate a more marked anabolic action on bone than higher doses of GH (51). The results of measurements of serum osteocalcin and total body bone mineral content in the trial comparing individualized GH doses with higher doses based on body weight (48) indicate, however, that the anabolic action on bone is no less with higher doses of GH than with lower doses. Different baseline bone mineral density z-scores should, however, be considered when studying the response to GH treatment in terms of bone mineral content and bone mineral density (32). The increased Lp(a) concentration in response to GH might be an effect of overly high doses of GH (31, 52). However, although a lower, individualized and probably more physiologic amount of GH is administered, the increment in Lp(a) is similar to that in patients with higher doses of GH (48). This suggests that the more continuous bell-shaped pattern of GH during sc administration might be of more importance to the increment in Lp(a) than the total amount of GH administered (53). We know little about the importance of the time of GH exposure in terms of responsiveness during more prolonged GH treatment, but these aspects are important when it comes to the aim of achieving a physiologic GH replacement. The results of a large, long-term study (20–50 months) (54) confirm the findings of previous, shorter investigations showing that GH replacement therapy can significantly improve many aspects of quality of life in adults with GH deficiency (10, 17, 55). In addition, the improvements in quality of life that occur shortly after the initiation of GH therapy are sustained in the longer term. Indeed, it seems that some aspects of quality of life continue to improve during long-term therapy. Furthermore, in a retrospective analysis, almost 60% of the patients receiving GH for 12–68 months felt that their quality of life was still improving. An important finding is that almost one third of patients who experience beneficial effects of GH therapy state that such effects have become noticeable only after GH had been administered for at least 6 months. This finding has clear implications for clinical practice: it indicates that, once the decision has been taken to initiate GH therapy in an adult with GH deficiency, the therapy should be continued for a minimum of 6 months, and ideally, for longer before judgments are made regarding its efficacy in improving quality of life and psychological well-being. In the absence of an optimal marker for physiological replacement therapy of GH in adults the importance of clinical judgment is increased. To avoid side effects in the initial phase of treatment, the value of closely monitoring clinical effects, particular in terms of increased extracellular fluid volume becomes evident. This is because normalization of serum levels of IGF-I does not exclude the possibility that the GH dose is too high in an individual patient (40). Furthermore, as stated previously, the delay until the patient clearly experiences benefits of the treatment may be long. The patient should be carefully informed of this and that should be considered during the more prolonged GH replacement therapy in adults. Concern has been expressed about the potential risk of development of neoplasia during GH replacement therapy. This is based in part on the increased risk of neoplasia in acromegalic patient. However, experience from a long-term surveillance study in pediatric patients (56, 57) does not show any increased risk of new or recurrent neoplasia. However, it should be borne in mind, that in contrast to the situation in children, GH deficiency in adults is predominantly secondary to pituitary and peripituitary tumors and their associated therapy. An increased left ventricular wall thickness was observed after 42 months of GH treatment in a small number of patients. These patients received overly high doses of GH, at least during the initial phase of treatment (58). This indicates that the therapeutic window with GH in terms of heart structure and function is probably not as wide as has been thought, particularly during long-term replacement, as changes in the myocardium may progress with prolonged GH administration. However, direct extrapolation of data from acromegalic patients to GH-deficient patients with GH therapy seems inappropriate, as the production rate of GH in acromegaly by far exceeds the ordinary GH replacement dose. Further long-term monitoring of GH-deficient adults receiving GH substitution therapy is clearly needed. In analogy with other hormonal replacement therapy GH treatment should be commenced with a low starting dose, independent of body weight or body surface area. Hormonal replacement should mimic the normal physiology to minimize the risk of side effects in the life-long replacement of adults. We should, therefore, consider individual responsiveness and also be aware of the difference between pattern of GH under normal condition and during sc administration. The safety and monitoring of GH replacement therapy in adults have been addressed in the Growth Hormone Research Society Consensus Guidelines for Diagnosis and Treatment of Adults with GH Deficiency from the Port Stephens Workshop, April 1997 (59). Besides finding better and more accurate biochemical markers for choosing correct GH replacement dose, future research should address the long-term benefits and safety with GH replacement in adults, with special emphasize on incipient risks in terms of cardiovascular disease (31, 58) and of neoplasia, in particular.