3.3. Quality of Life
A recent retrospective analysis of patients who were treated with GH during childhood in 11 centers of pediatric endocrinology in Switzerland compiled the information on QoL during adult life. Data of 300 participants showed that QoL results are in relation with the underlying indication for GH treatment. In particular, patients with associated diseases or syndromes scored slightly lower, and former cancer patients scored lower than patients with isolated GHD or idiopathic short stature. Lower physical component summary was associated with lower educational level [49
]. These data emphasize the need to bear in mind the underlying condition when interpreting results of QoL measurements in adults with childhood onset GHD.
QoL improvements have not been reported in all adults or in all assessed socio-psychological domains [50
]. Some studies have shown even worsening in some domains, such as social function and mental health, in adults with childhood-onset GHD [52
]. In general, responses are very heterogeneous and patients with worse QoL before treatment are the best responders [27
]. Therefore, it is possible that patients who have a normal QoL at baseline do not benefit from GH treatment in this regard.
3.5. Adverse Events
Side effects derived from water retention caused by GH occur in 5–18% of patients. They are edema, arthralgias, myalgias, paresthesias, and carpal tunnel syndrome, and usually improve after dose reduction. Elderly, female, and overweight patients have a higher risk of developing these effects. Retinopathy and benign intracranial hypertension are very rare complications of GH treatment. The needs of cortisol and thyroxin may be increased in patients with hypopituitarism [55
]. Special attention should be given to older, heavier, and female patients with GHD because they are more susceptible to adverse events. Monitoring IGF-I concentrations is essential to avoid adverse effects of excess GH treatment. A recent randomized, open-label, clinical trial including 32 adults receiving GH therapy for at least one year showed that although increasing GH dose to achieve IGF-I levels between 1 and 2 Standard Deviation Score (SDS) improved waist circumference and the overall better feeling, safety was not guaranteed with the demonstrated effect on HDL cholesterol in men, and reported myalgia [56
To minimize the impact of adverse effects, it is appropriate to start treatment with low doses and titrate upwards according to clinical response and IGF-I levels, as well as following some general recommendations that are shown in Table 2
3.6. Risk of Hyperglycemia and Diabetes
Treatment with GH reduces insulin sensitivity and elevates blood glucose. In one study, conducted in 90 patients on GH treatment, a significant increase in both glucose and hemoglobin A1c levels was demonstrated. Changes were evident at six months and persisted after two years of treatment [57
]. These data have been corroborated in some studies [58
] and by a meta-analysis of 37 randomized, placebo-controlled, clinical trials, which showed that GH treatment was accompanied by an increase in fasting blood glucose and insulin concentrations, regardless of the dose and duration of treatment [33
]. In a randomized, double-blind, placebo-controlled clinical trial in 166 adult patients with GHD, treatment with GH was associated with a worsening of glucose tolerance with the appearance of diabetes in 4% and carbohydrate intolerance in 20% of patients after 12 months, while in the placebo group only 8% developed intolerance [32
In the Hypopituitary Control and Complications Study (HypoCCS) the prevalence and incidence of diabetes in adults treated with GH has also been analyzed [59
]. Results of an analysis of 2922 patients in the United States and 3709 in Europe, with a mean follow-up of 4.1 years, showed that, in the United States, the incidence rate of diabetes adjusted for age, sex, and BMI is higher in patients treated with GH than in the general population. In France and Germany, the incidence rates were comparable with the reference population, while in Sweden the incidence rate was more than double in patients on GH treatment. A more recent analysis of 5143 patients from the Pfizer International Metabolic Database (KIMS), with 20,106 patient-year follow-up showed that, in southern Sweden, the observed/expected cases ratio was 10.8 the first year of treatment with GH and fell to 1.9 after eight years of treatment, but always stayed above 1. When the incidence of diabetes in patients in the KIMS study was compared with the incidence rates in age-adjusted populations in other European and US regions, the observed/expected cases ratios ranged from 2.11 to 5.22 [60
The Safety and Appropriateness of Growth Hormone Treatment in Europe (SAGhE) was a population-based cohort study that analyzed the long-term mortality and morbidity in adults who were treated with GH during childhood. A recent report by Poidvin et al. [61
] have used the French SAGhE database to evaluate the prevalence of diabetes, including gestational diabetes, in a population-based cohort of patients treated with GH for short stature during childhood, and to compare this prevalence with that of the general population in France. They analyzed a cohort of 5100 children with idiopathic short stature, or short stature in children born short for gestational age. In summary, these authors found no increase in the risk of treated diabetes in subjects receiving GH treatment during childhood, with a mean follow-up of 19 years.
A recent study in 245 patients with adult-onset GHD and more than four years of GH replacement showed that this therapy did no adversely affect glucose homeostasis in the majority of adults patients, although seven patients developed diabetes [62
]. A meta-analysis of 94 randomized controlled trials and open trials did not demonstrate neither an increased frequency of diabetes in the short-term placebo controlled trials or a consistently increased incidence of diabetes during long-term GH replacement. However, the small number of subjects and the absence of adequate control population are a limitation for the interpretation of these results [63
Close monitoring of glucose status is advisable during GH treatment in patients with obesity or a family history of type 2 diabetes, because they are more prone to develop impaired glucose intolerance and diabetes during therapy [64
3.7. Risk of Neoplasia
Epidemiological studies have suggested an association of high circulating IGF-I or GH to cancer incidence [65
], and the incidence of some malignant neoplasia is known to be increased in acromegaly [67
]. Initial studies conducted in children treated with extractive GH showed an increased risk of mortality from cancer and, in particular, from colorectal cancer and Hodgkin’s disease [68
]. The oncogenic risk during treatment in children treated with GH was established in the National Cooperative Growth Study (NGCS) and the Pfizer International Growth Database (KIGS), which evaluated more than 50,000 patients each, with almost 200,000 patient-years each. The standardized incidence rate of cancer was 1.12 (95% confidence interval: 0.75–1.61) in the former [69
] and 1.26 (0.86–1.78) in the latter [70
], thus suggesting that there was no significant increase in the risk of cancer in these children. No greater risk of leukemia than the general population could be demonstrated in these studies. Furthermore, recent studies have not confirmed an increased risk of mortality in children treated with GH, especially when applying sex-specific mortality models adjusting for birth characteristics [71
The risk of second neoplasia in survivors of childhood cancer has been analyzed in depth in the Childhood Cancer Survivor Study (CCSS). In a scrutiny conducted in 2002, this study showed that children treated with GH had a relative risk for cancer of 3.21 (1.88–5.46) in comparison to cancer survivors not treated with this hormone [72
]. This relative risk was even higher, 4.98 (1.95–12.74), in leukemia survivors. In the 2006 analysis [73
], including a further 32-month follow-up, the risk of second neoplasia was reduced to 2.15 (1.33–3.47), but remained significant. Another study in one center, including 49 patients with GHD within a cohort of 310 childhood cancer survivors, showed that GH replacement therapy did not seem to increase the risk of second neoplasia, but the authors suggest the need of a close long-term follow-up in these patients [74
Data available in adult patients suggest that replacement therapy with GH does not increase the rate of recurrence or progression of tumors in the hypothalamic-pituitary area [75
An analysis of the risk of neoplasia was carried out in the SAGhE study database, including 6928 patients treated for idiopathic GHD, neurosecretory dysfunction, idiopathic short stature and small for gestational age, that is, etiologies of GHD without increased risk of neoplasia. Mean time of follow-up after completing therapy was 7.8 years [76
]. In this study, the standardized mortality ratio for bone and cartilage neoplasia reached the striking figure of 5.00 (1.01–14.61). In a group of 6840 adults included in the HypoCCS study, the incidence ratio of neoplasia was 0.88 (0.74–1.04), but it rose to 3.79 (1.39–8.26) when patients younger than 35 years were analyzed, and to 2.74 (1.18–5.42) when analyzing patients with childhood onset GHD [77
]. These results suggested that the overall risk of primary cancer in adult life is not increased; however, there are subgroups with high risk.
Using the HypoCCS database, Child et al. [78
] reported the incidence of primary neoplasia in a cohort of 8418 patients treated with GH, as well as in 3668 GH-treated patients with history of pituitary adenoma and 956 GH-treated patients with history of craniopharyngiomas. Comparison were carried out with cohorts of untreated patients. During a mean follow-up of 4.8 years, they found that no increased risk for all-site cancers, including breast, prostate and colorectal cancers, in GH-treated patients. In addition, GH treatment did not increase the risk of recurrence of pituitary adenoma or craniopharyngioma in this study.
Recent studies have also reported reassuring results. A recent study, including 426 patients with nonfunctioning pituitary adenomas, with 4599 patient-years of follow-up, of whom 207 had used GH therapy and 219 had not received GH, showed a reduced overall mortality in GH-treated patients compared with the general population and found that death due to malignancy was not increased in GH-treated patients [79
]. However, selection bias explaining some of the results cannot be excluded in this study. On the other hand, a recent metaanalysis, including two retrospective and seven prospective studies with a total of 11,191 participants, suggested that GH replacement therapy could reduce risk of cancer in adult with GHD [80
Nevertheless, GH therapy should not be given to patients with active malignancy [64
], and should be prescribed with caution in GHD adult patients with a history of cancer, strong family history of cancer, and advancing age [81
]. Childhood cancer survivors may be at increased risk for secondary neoplasm compared with general population. GH should be used cautiously in this subgroup of patients [81
3.8. Mortality Risk
A study conducted with data from the Dutch National Registry of GH treatment [83
] compared 2229 patients undergoing treatment with GH with a primary control group of 109 patients diagnosed but not treated with GH and a secondary control group of 356 patients treated with GH in whom the treatment had been discontinued. The standardized mortality ratio in relation to the general population was 1.27 (1.04–1.56) for the treatment group. This ratio was 1.29 (1.05–1.59) when patients with acromegaly and Cushing’s disease were excluded, and 1.00 (0.79–1.26) after the exclusion of high-risk patients (craniopharyngioma and other causes). It is noteworthy that the authors found no significant increase in mortality in the two untreated control groups. The authors also found a significant increase in the standardized mortality ratio due to cardiovascular disease in women in the treatment group (2.52 [1.57–4.06]), which persisted after exclusion of high-risk patients.
The evaluation of a group of 13,983 GH-treated patients, with a mean follow-up of 4.9 years, showed that all-cause mortality in these patients was 13% higher than in the general population (standardized mortality ratio of 1.13 [1.04–1.24]). There was no increase in mortality due to cardiovascular disease or cancer [84
]. In the aforementioned SAGhE study, the risk of overall mortality in patients who received GH in childhood increased by 33%. In this study, the standardized mortality ratio for circulatory diseases was 3.07 (1.40–5.83), and rose to 5.29 (1.42–13.55) and 6.66 (1.79–17.05) when deaths due to cerebrovascular disease and intracranial hemorrhage were considered, respectively. On the contrary, the observational GeNeSIS study, evaluating 9504 GH-treated children followed for at least four years found no increase in risk of mortality compared with children in the general population [85
]. This study found an increase in mortality risk for children with history of malignant neoplasia. However, this study is limited to the period of active treatment with GH and without follow-up in adult life.
In summary, as stated in the Endocrine Society clinical practice guideline [6
], although mortality is increased in patients with hypopituitarism and GHD has been implicated in this, there is no data to show that treatment with GH improves the survival of patients.