|
|
||||||||
CLINICAL STUDY |
UCL St Luc Hospital, B-1200 Brussels, Belgium1 University Hospital, Antwerp, B-2650 Edegem, Belgium2 Sahlgrenska University Hospital, SE-413 45, Göteborg, Sweden3 University Hospital of Wales, Cardiff, United Kingdom4 KIGS/KIMS/ACROSTUDY Medical Outcomes, Pfizer Endocrine Care, SE-112 87 Stockholm, Sweden5 Department of Pharmacy, Uppsala University, Uppsala, Sweden
(Correspondence should be addressed to D Maiter; Email: maiter{at}diab.ucl.ac.be)
| Abstract |
|---|
|
|
|---|
Design: Data from 447 patients, who had received radiotherapy (427 in addition to surgery), and 630 patients, who were operated on but not irradiated for their tumour, were retrieved from Pfizer International Metabolic Database (KIMS) and compared at baseline and 1 and 2 years following the onset of GH replacement.
Results: Irradiated and non-irradiated patients exhibited the expected phenotype of GHD at baseline. However, irradiated patients had a greater impairment in the quality of life (QoL), a higher fat mass, lower high-density lipoprotein cholesterol levels and a lower bone mineral content (BMC) than non-irradiated patients. Treatment with GH induced similar changes in both groups. After 1 year of GH replacement, there was an increase in serum IGF-I and fat-free mass, a reduction in fat mass and an improvement in QoL, all changes being equivalent in irradiated and non-irradiated patients. The lipid profile also improved with the irradiated patients showing a better response. These beneficial effects were maintained and the BMC also increased in both groups by the second year of treatment.
Conclusions: This analysis shows that prior irradiation for pituitary adenoma or craniopharyngioma does not compromise the beneficial effects of GH replacement therapy.
| Introduction |
|---|
|
|
|---|
Little is known about the effects of GH replacement therapy in the subset of patients who have received prior pituitary radiation therapy, and the well-known beneficial effects of GH replacement therapy in GH-deficient adults (1214) might be different in this particular population. Previous cranial irradiation may, indeed, induce cognitive impairment, lower quality of life (QoL) and alterations of hypothalamic-dependent functions, such as the regulation of food intake (5, 15, 16). Such differences in baseline features of irradiated patients might significantly influence their response to GH therapy. We, therefore, analysed retrospectively data from Pfizer International Metabolic Database (KIMS), a pharmaco-epidemiological survey of adult GH-deficient patients, to compare both the baseline characteristics and the response to GH replacement of patients previously irradiated for pituitary adenoma or craniopharyngioma, with those of patients not given radiotherapy.
| Patients and methods |
|---|
|
|
|---|
The study included a total of 1107 adult patients (590 males, 517 females) enrolled in KIMS at the start of their GH replacement therapy. All patients had severe GH deficiency (GHD), defined as a peak GH level less than 3 µg/l on one or more validated tests of GH provocation (insulin tolerance, Arginine, glucagon or GH-releasing hormone tests) and had been treated previously for either pituitary adenoma (n = 904; 487 males) or craniopharyngioma (n = 203; 103 males). Most patients were from western European countries (91%). In total, 477 subjects (43% of total, 238 males) had received radiotherapy only once and at least 1 year (median, 9 years) before the onset of GH treatment (irradiated (IRR) group). Of this group, 427 patients (90%) had been treated with a combination of surgery and radiotherapy, while 50 patients had received irradiation alone. The IRR group was compared with a group of 630 individuals (57% of total, 352 males) who had been operated on for their tumour but had not received irradiation (non-IRR group). Among the 477 IRR patients, the vast majority (n = 460) had been treated with conventional fractionated radiotherapy targeted to the sellar and/or perisellar region. Only 11 patients had received stereotactic radiosurgery and only six patients had been treated with radioactive implants. The data used were restricted to patients who had never been treated with GH (naive; n = 967) or who had been treated previously with GH and had discontinued GH treatment for at least 6 months before enrolment into KIMS (semi-naive; n = 140) and for whom sufficient data were available at the onset or restart of GH treatment (baseline visit).
A longitudinal analysis to evaluate the effects of GH treatment within the same individuals was possible after 1 year of therapy in 458 patients (n = 210, IRR; n = 248, non-IRR) and 2 years in 264 subjects (n = 127, IRR; n = 137, non-IRR). The background and baseline characteristics of these subgroups of patients were not different from those found in the whole study population and baseline differences, described between IRR and non-IRR groups, were also observed when considering only patients with 1- or 2-year follow-up (data not shown).
Methods
Background data consisted of the following items: gender, age at diagnosis, adult- or childhood-onset GHD, estimated duration of GHD, age at entry into KIMS, type of pituitary adenoma and relevant medical history. History concerning the primary tumour included the number and type of surgical procedures, use and timing of radiotherapy, and number and type of other pituitary hormone deficits. Overall medical history included information on smoking, cerebrovascular and cardiovascular disease, peripheral vascular disease, hypertension, diabetes mellitus, fractures and neoplasms.
Baseline characteristics recorded at the initiation of GH treatment included weight, height, body mass index (BMI), waist circumference, hip circumference and systolic and diastolic blood pressure. Bioelectrical impedance analysis (BIA) was used to determine fat mass and fat-free mass (FFM) from equations supplied by the manufacturers, with resistance and reactance as input variables. Although less precise than other laboratory techniques, this non-invasive and rapid method has been proven to be clinically valuable to monitor changes in the body composition during GH therapy (17). Bone mineral content (BMC) was evaluated by dual-energy X-ray absorptiometry.
Plasma serum total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides were measured centrally by standard methods and serum low-density lipoprotein (LDL) cholesterol was calculated (18). Serum insulin-like growth factor-I (IGF-I) was measured centrally by an HCl-extraction RIA (Nichols Institute Diagnostics, Inc., San Juan Capistrano, CA, USA). Intra-, interassay and total coefficients of variation were less than 9% in the concentration range 1251046 µg/l. The assay detection limit was 13.5 µg/l. IGF-I values were adjusted for age and expressed as SDS.
QoL was assessed using the Quality of Life Assessment of GHD in Adults (QoL-AGHDA) (19). A higher score (to a maximum of 25) denotes a poorer QoL. Previous studies have shown that QoL-AGHDA can be considered as a robust measure, well suited to assess QoL in adults with GHD (19, 20).
Treatment with GH was started in all patients at a low dose and titrated against clinical response and IGF-I levels. Follow-up parameters available in 458 patients at 1 year and in 264 patients at 2 years were the daily GH dose and changes in IGF-I SDS, weight, BMI, waist circumference, blood pressure, body composition, serum lipids and QoL-AGHDA score.
Statistical analysis
Data analyses were performed using the Statistical Analysis System (SAS Institute, Inc., Cary, NC, USA). Significant differences between the IRR and non-IRR groups at baseline were determined by unpaired t-tests for normally distributed variables or by the Wilcoxon rank-sum test. The effects of GH treatment were analysed using paired t-tests or multivariate analysis for normally distributed data and by the Wilcoxon rank-sum test in other cases. Comparisons of proportions were performed using the
2 and Fischer exact tests. The analysis of baseline and follow-up data was corrected for gender and the presence of craniopharyngioma. Significance was accepted at a value of P < 0.05 and the data are expressed as means ± S.D. unless stated otherwise.
| Results |
|---|
|
|
|---|
A few background variables were different between the IRR and non-IRR groups (Table 1
). Thus, in the non-IRR group, there were more male patients, the subjects were older at the time of tumour diagnosis, they had a shorter estimated duration of GHD than the IRR patients, and a higher prevalence of complete pituitary hormone deficiency (P < 0.01). The relative proportion of patients with craniopharyngioma was slightly higher in the non-IRR group (21%) than in the IRR group (15%, P < 0.01), while the proportions of the different types of pituitary adenomas were similar between both groups. There was also a slight, but expected, difference in the number of surgical operations. The majority of patients had undergone at least one procedure, apart from a small subset of 50 patients in the IRR group (10%), who had received radiotherapy alone. Repeat surgery, because of recurrent or residual tumour, was carried out in 18% of both IRR and non-IRR patients.
|
Baseline data
Patients with a previous history of irradiation showed an equivalent degree of GHD to the non-IRR population, as demonstrated by equally reduced IGF-I concentrations and a similar GH peak in response to stimulation test (Table 2
). Prior to the start of GH treatment, the two groups also did not differ in terms of systolic and diastolic blood pressure, BMI, waist circumference, FFM and serum total cholesterol, LDL cholesterol and triglycerides. IRR patients, however, had a greater impairment in QoL, as estimated by the QoL-AGHDA score, a higher mean body fat mass, a lower level of HDL cholesterol and a lower BMC than non-IRR patients (Table 2
). No difference in BMI and FFM was observed in either gender between IRR and non-IRR patients (data not shown). IRR female patients had a higher fat mass than non-IRR women (29.7 vs 23.5 kg, P < 0.01), while the difference was not significant in men (22.8 vs 20.6 kg).
|
After 1 year, the dose of GH was similar in both groups (0.40 ± 0.02 mg/day, mean ± S.D.) and GH treatment induced significant and similar increases in serum IGF-I concentrations in both IRR and non-IRR patients, as well as similar improvements in QoL, as reflected by a decrease in the QoL-AGHDA score (Table 3
; Fig. 1
). Levels of total and LDL cholesterol were reduced in both groups; however, the reduction was greater in IRR patients (Table 3
; Fig. 2
). Changes in body composition determined by BIA were studied in a subgroup of 115 patients (n = 55, IRR; n = 60 non-IRR; Table 4
). FFM increased during GH therapy in both IRR and non-IRR patients, whereas body fat mass decreased similarly in both groups. BMI was slightly reduced in the IRR group but not in patients who had not previously received irradiation. There was no significant change in triglycerides, HDL cholesterol, waist circumference and BMC in either group (Tables 3
and 4
, and data not shown).
|
|
|
|
| Discussion |
|---|
|
|
|---|
Before the initiation of GH treatment, IRR and non-IRR patients were similar in terms of the severity of GHD. Repeat surgery (presumably for regrowth of the tumour), which is known to be associated with increased disease severity and mortality in patients with pituitary macroadenomas (22), was performed in a similar number of patients in each group. Since the proportion of patients with craniopharyngioma was slightly higher in the non-IRR group, the data were corrected for the presence of craniopharyngioma, as this condition has previously been shown to cause more profound hypothalamic damage and an increased morbidity and mortality (23). There were also slightly more men in the group of non-IRR patients. Since the gender difference in sensitivity to GH is known women being more resistant and requiring higher GH doses to achieve normal IGF-I concentrations (24) data analysis was also corrected for gender and similar results were obtained in male and female IRR patients. Non-IRR patients had a higher number of pituitary deficits, mainly diabetes insipidus, and a significantly shorter estimated duration of GHD prior to entry into KIMS than IRR patients. None of these differences are sufficient, however, to preclude a comparative analysis of the response to GH treatment between the two groups.
Prior to the start of GH treatment, the patients exhibited the characteristics of severe adult GHD, including reduced IGF-I SDS, FFM, BMC and QoL, increased BMI and fat mass and an abnormal lipid profile. Individuals with a previous history of pituitary irradiation, however, had a greater impairment in their QoL, a higher body fat mass, lower HDL cholesterol levels and a lower BMC than non-IRR patients. The observed differences may have resulted from the longer duration of untreated GHD in the IRR group or from direct detrimental effects of radiotherapy on the brain. It has previously been shown that radiotherapy may have an effect on QoL. Page et al. also reported higher levels of depression and emotional lability in a group of IRR GH-deficient patients compared with patients who had undergone mastoid surgery, but the effects of GHD and irradiation could not be distinguished (25). In contrast, in a small cohort of patients with hypopituitarism and GHD, using a different questionnaire, Wallymahmed and colleagues could not find any significant differences in QoL between those who had received radiotherapy and those who had not received radiotherapy (26).
In the current study, GH replacement therapy with a comparable daily dose of GH (median, 0.40 mg) for 1 year produced a similar improvement in the characteristics of severe GHD in IRR and non-IRR adults. Thus, GH induced similar increases in IGF-I and FFM, a similar reduction in fat mass, and a comparable improvement in QoL in both groups. These positive effects of GH treatment were maintained during the first 2 years of treatment and, by the end of the second year of GH treatment, BMC had also increased equally in both groups. Moreover, 2 years of GH therapy attenuated the baseline differences observed in fat mass, HDL cholesterol and BMC between IRR and non-IRR subjects, but was unable to diminish the relative difference in QoL, which remained worse in IRR than in non-IRR patients throughout the study period, despite an overall improvement in QoL in both groups.
In comparison with patients who had never received radiotherapy, IRR patients showed a better response with respect to the lipid profile, the reduction of total and LDL cholesterol being significantly greater in IRR patients. This difference was maintained after 2 years of GH treatment. GH-deficient patients are known to have higher total and LDL cholesterol than normal controls, while HDL cholesterol often tends to be lower than expected; these abnormalities are reversed by GH treatment (27, 28). The mechanisms behind these GH effects on dyslipidaemia are not fully understood, but are likely to involve a stimulatory effect on hepatic LDL receptors and an increase in LDL clearance rate (29). There is, however, no obvious explanation for the observed effects of irradiation on these mechanisms.
To test our hypothesis, the study required a very large patient population, which is not available in individual clinics and can only be provided by central databases such as KIMS. Accepting the well-known limitations of the retrospective analysis of observational surveys, this analysis clearly shows that prior pituitary irradiation does not significantly compromise the beneficial effects of GH replacement therapy.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Klose, T. Watt, J. Brennum, and U. Feldt-Rasmussen Posttraumatic Hypopituitarism Is Associated with an Unfavorable Body Composition and Lipid Profile, and Decreased Quality of Life 12 Months after Injury J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3861 - 3868. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |