| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatric Endocrinology (D.S., P.C.), Robert Debré Teaching Hospital, 75019 Paris, France; Department of Pediatric Rheumatology (A.-M.P., P.Q.), National Reference Center for Rare Diseases, Necker-Enfants Malades Teaching Hospital, 75015 Paris, France; and Department of Radiology (J.C.R.), Cochin Teaching Hospital, 75014 Paris, France
Address all correspondence and requests for reprints to: Dominique Simon, M.D., Hôpital Robert Debré, Service d’Endocrinologie, 48 Boulevard Sérurier, 75019 Paris, France. E-mail: dominique.simon{at}rdb.aphp.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objectives: The objective of the study was to evaluate the impact on growth and body composition of rhGH started early after glucocorticoid initiation and to record adverse effects in children with JIA.
Design: This is a 3-yr randomized controlled study.
Setting: This study was conducted in a teaching hospital.
Patients: Thirty children, 12–15 months into glucocorticoid therapy for severe JIA, were enrolled.
Intervention: Patients received rhGH (0.46 mg/kg·wk) in daily sc injections (n = 15) or no rhGH therapy (n = 15) for 3 yr.
Main Outcome Measure: Difference in height SD score (SDS) change between the two groups was assessed. Height velocity, body composition, and oral glucose tolerance were evaluated yearly.
Results: Mean height SDS increase was larger with rhGH (+0.37 ± 1.5 SDS) than without (–0.96 ± 1.2 SDS) (P = 0.04). Mean height velocity returned to normal within the first year of rhGH treatment and remained normal thereafter. Mean lean mass increase was greater with rhGH treatment (+7.3 ± 2.9 kg vs. +4.4 ± 2.8 kg; P = 0.03). Fat mass and bone mineralization were not significantly different in the two groups. Fasting serum insulin increased significantly in rhGH-treated patients (5.2 ± 16 mIU/liter) compared with untreated controls (–2.3 ± 5 mIU/liter) (P = 0.04); fasting glycemia was unchanged.
Conclusions: rhGH started early in the course of JIA preserved normal growth velocity and height. Although rhGH was well tolerated, carbohydrate metabolism should be monitored closely.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
To minimize adverse effects, the smallest effective glucocorticoid dosage should be used. Combined use of other disease-modifying or immunosuppressive drugs may reduce the glucocorticoid requirements, and alternate-day administration may provide a better safety profile than daily dosing. Nevertheless, the smallest glucocorticoid dosage that ensures acceptable disease control may induce adverse effects on growth and on muscle and bone development in children. We found previously that 40% of patients with a history of systemic juvenile idiopathic arthritis (JIA) treated with glucocorticoids for at least 2 yr during childhood had a final height SD score (SDS) under –2 SD and that more than 80% had a final height below their target height. Moreover, final height correlated positively to height at discontinuation of glucocorticoid therapy (1). In observational and controlled studies of children receiving glucocorticoid therapy for JIA, recombinant human growth hormone (rhGH) restored normal height velocity, halted the muscle wasting, and improved bone mineralization (2, 3, 4, 5, 6, 7, 8, 9). Considerable interindividual variability in responses to rhGH occurred, however, depending on JIA severity (10). In most studies, disease severity (i.e. JIA subtype), disease activity [i.e. erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) values], and prednisone dosage correlated negatively with the response to rhGH treatment. Cytokine exposure and glucocorticoid treatment duration at rhGH initiation may also influence the response to rhGH, because both adversely affect growth plates and bone metabolism in a dose- and time-dependent manner, potentially leading to incomplete recovery of chondrogenesis and bone growth (11, 12, 13, 14). Thus, starting rhGH treatment in severely growth-retarded children who have had JIA and taken glucocorticoids for many years is unlikely to restore normal height and body composition. Therefore, we investigated whether early initiation of rhGH therapy prevented loss of height, muscle mass, and bone mass in children with JIA.
We designed a randomized controlled study with the primary objective of investigating the effects on linear growth of rhGH therapy started after 12–15 months of glucocorticoid therapy in children with JIA, i.e. as soon as the patients met the criterion for glucocorticoid dependency. Our secondary objectives were to evaluate the effects of rhGH therapy on body composition and bone mineralization as well as its safety.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In this 3-yr, randomized, controlled study conducted in a teaching hospital in France, patients with JIA were allocated at random in a 1:1 ratio to rhGH [Genotonorm (somatropin, rDNA origin); Pfizer, New York, NY], 0.46 mg/kg·wk divided in daily sc injections and adjusted for weight at each visit, or to no rhGH therapy. At the end of the 3-yr study, rhGH therapy was offered to all patients.
Inclusion criteria
Inclusion criteria were as follows: systemic or polyarticular JIA according to International League Against Rheumatism criteria (15) requiring glucocorticoid therapy, ongoing prednisone therapy started 12–15 months earlier, prednisone dosage equal to or greater than 0.2 mg/kg·d at baseline, prepubertal Tanner stage (16), chronological age (CA) of 18 months to 9 yr for girls or 18 months to 11 yr for boys, and height velocity in the previous year equal to or less than –1 SD for chronological age and sex according to normative data obtained in France (17). We excluded patients who had other severe chronic diseases likely to affect growth (e.g. chronic renal failure, nephrotic syndrome, heart failure, or liver failure) or diabetes mellitus.
Randomization procedure
A computer-generated randomization list was provided by Pfizer (Paris, France). Included patients were assigned numbers in chronological order, which were used for allocation at our center by a person who otherwise had no role in the study. The study was not blinded, and the parents were informed of group assignment.
Study parameters
The same pediatric endocrinologist (D.S.) and a pediatric rheumatologist examined all the study patients at baseline and then at 6-month intervals throughout the study. At each visit, the following were recorded: standing height as the mean of three consecutive measurements using a Harpenden stadiometer, body weight, body mass index (weight in kilograms/height in meters squared), height velocity, vital signs, and adverse events. Height, height velocity, and body mass index were expressed as SDS for CA (height SDS/CA, body mass index SDS/CA, and height velocity SDS/CA, respectively) compared with normative data from France (17, 18). Puberty status was graded according to Tanner (16). Bone age was assessed at baseline and yearly on radiographs of the left hand according to Greulich and Pyle (19). Radiographs of the thoracolumbar spine were assessed at baseline and at study completion. JIA was assessed at each visit by the pediatric rheumatologist, who recorded the number of active joints, visual analog scale scores determined by the parents, changes in glucocorticoid dosage, and changes in other antirheumatic drugs.
Laboratory data and body composition assessment
All laboratory tests were done at the same laboratory, on samples obtained after an overnight fast. IGF-I, ESR, CRP, and glycosylated hemoglobin (HbA1c) were assayed at baseline and yearly thereafter. At the same time points, oral glucose tolerance tests (OGTTs) were done, and body composition was measured using dual-energy x-ray absorptiometry (DEXA).
DEXA was performed using a Hologic QDR machine (1000W/892 mef 1990; Hologic, Waltham, MA). Lean mass (kilograms and percent body weight), fat mass (kilograms and percent body weight), and bone mineral content (grams) were evaluated. Because JIA is associated with delayed growth, values were normalized for body weight. Using the same DEXA method in a control population (20), we determined simple linear correlations linking lean mass, fat mass, total bone mineral content, and weight to gender-dependent constants. These correlations, as well as those used to calculate the corresponding SDs, have been reported elsewhere (5). Because growth retardation occurred, predicted values for lean mass and total bone mineral content were calculated based on expected weight for height (WFH). Bone mineral density was measured in the supine position at the lumbar spine and expressed as grams per square centimeter and as SDS/CA (21). Volumetric bone mineral density (grams per cubic centimeter) was calculated as proposed by Kroger et al. (22).
Serum IGF-I levels were measured by RIA (IGF1-RIAACT; CisBio, Gif sur Yvette, France) and expressed as nanograms per milliliter. OGTTs were done after an overnight fast. Plasma glucose levels were determined before and 30, 60, and 120 min after an oral glucose load (1.75 g/kg, up to 75 g). Fasting insulin levels were measured by a specific radioimmunometric method, for which cross-reactivity with intact proinsulin was less than 1% (Bi-Insulin IRMA; Cisbio). Fasting hyperinsulinemia was defined as fasting insulinemia of more than 11 mIU/liter with normal fasting glycemia. Impaired glucose tolerance and diabetes mellitus were defined according to the American Diabetes Association guidelines (23). HbA1c was assayed by HPLC.
Ethical considerations
The study was conducted in accordance with the guidelines in the Declaration of Helsinki, Good Clinical Practice guidelines, and French legislation. Written informed consent was obtained from all patients older than 7 yr of age and from all parents or legal surrogates. The Paris-St. Louis Ethics Committee approved the study.
Statistical analysis
All patients for whom baseline height and at least one postrandomization height value were available were included in the analysis.
The primary outcome measure was the difference between the two groups regarding the mean change in height SDS over the 3-yr study period. Secondary outcome measures used to assess efficacy were the between-group differences in mean changes in body composition, bone age, and IGF-I levels over the 3-yr study period. Safety data included clinical and laboratory variables reflecting JIA control and glucose-tolerance changes over the 3-yr study period.
Changes were compared between the two treatment groups using the nonparametric Wilcoxon signed-rank test. All tests were two-sided; P values < 0.05 were taken as indicating significant differences. Coefficients of correlation were estimated and compared with zero using the nonparametric Spearman test. The proportions of patients who experienced changes in glucose metabolism were determined and then compared between the two groups using Fisher’s exact tests. In addition, exploratory multiple regression analyses were performed to test a possible interaction between treatment and CRP and treatment and prednisone dose on changes in height SDS.
Statistical analysis was performed using SAS software (version 8.2; SAS Institute Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
We included 30 patients (16 females and 14 males) from 1998–2002. A single family declined participation in the study. At baseline (Table 1
), most patients had systemic JIA and started high-dose prednisone therapy shortly after the diagnosis. A marked decrease in height velocity in the year before inclusion was found consistently. Mean height SDS remained in the normal range, in keeping with the short durations of the disease and prednisone therapy, although the value was lower than the mean target height SDS. Plasma CRP levels were not significantly different between the two groups at baseline. Plasma IGF-I levels (nanograms per milliliter) correlated negatively with CRP levels (milligrams per liter) (
= –0.40; P = 0.03) and positively with body mass index ( = 0.57 and P = 0.001 when expressed as kilograms per square meter; and = 0.60 and P = 0.0005 when expressed as SDS/CA).
|
Premature study withdrawal occurred in four patients, two in each group. In two patients (one in each group), the reason for withdrawal was a decision to perform bone marrow transplantation because of severe disease progression; the decision was taken 3 and 18 months after study inclusion, respectively. The two other patients died (one in each group), one after surgery for spinal cord compression 3 months after inclusion and the other during an acute infection 2 yr after inclusion. Glucocorticoid therapy was discontinued during the study in two controls, 8 months after inclusion.
In the rhGH group, all boys remained prepubertal and two girls started puberty at 8.7 yr (3 yr into the study) and 10.8 yr (21 months into the study) of age, respectively. Among the controls, one girl started puberty at 12 yr of age (33 months into the study) and one boy at 13 yr of age (24 months into the study).
Effect of rhGH treatment on growth parameters
Mean change in height SDS over the 3-yr study period was significantly better in the rhGH group (+0.37 ± 1.5 SDS) than in the control group (–0.96 ± 1.2 SDS) (P = 0.04) (Table 2 and Fig. 1). Mean height velocity (centimeters per year and SDS) returned to normal within the first year of rhGH therapy and remained within the normal range thereafter. In the control group, in contrast, mean height velocity (centimeters per year and SDS) remained well below the normal range during the first 2 yr of the study but returned to normal during the third year.
|
|
= –0.64 and P = 0.0003; second year, = –0.45 and P = 0.02; and third year, = –0.59 and P = 0.002). Mean height gain (centimeters per year) also correlated negatively with the mean cumulative prednisone dosage (milligrams per square meter per year) in the second year ( = –0.42; P = 0.04) and third year ( = –0.53; P = 0.02) of the study. No interaction with rhGH treatment was statistically significant. Effect of rhGH treatment on body composition
Treatment with rhGH was well tolerated, and despite the high dosages used, no patients exhibited clinical symptoms of rhGH-induced water retention. Mean lean mass (percent body weight, SDS/WFH) was within the normal range in both groups at baseline, whereas mean fat mass (percent body weight) was above the normal range (Table 3). Over the 3-yr study, mean lean mass (kilograms) increased in both groups, but the increase was significantly larger in the rhGH-treated group than in the controls (7.3 ± 2.9 kg vs. 4.4 ± 2.8 kg, P = 0.03) and was positively correlated with the height gain (centimeters per year) ( = 0.89; P < 0.0001). Lean mass (percent body weight and SDS) remained within the normal range in both groups throughout the study. Changes in fat mass were not significantly different in the two groups when expressed in kilograms but differed significantly at the 3-yr time point when expressed as percent body weight (–1.5 ± 3.5 and 0.6 ± 1.6, respectively; P = 0.03) (Table 3).
|
Three patients in the rhGH-treated group and two controls had dorsal or lumbar vertebral collapse at baseline. In both groups at baseline, mean lumbar spine bone mineral density (area BMD SDS/CA) was below the normal range, but bone mineral content (SDS/WFH) was within the normal range. No significant differences in bone mineralization were noted between the two groups during the 3-year study (Table 3
). At the end of follow-up, two additional patients in the rhGH-treated group had vertebral collapse.
Outcomes of JIA
All the rhGH-treated patients and all but two of the controls received glucocorticoid therapy throughout the 3-yr follow-up. Many patients also took other antirheumatic drugs. At the end of follow-up, two rhGH-treated patients and one control had been receiving etanercept for more than 1 yr; and 12 rhGH-treated patients and seven controls had been on methotrexate for more than 1 yr. The ESR (millimeters per hour), CRP level (milligrams per liter), and prednisone dosage (milligrams per kilogram per day) tended to decrease during the study; however, at none of the time points were significant differences found between the two groups for any of these parameters (Table 2). The clinical disease activity parameters varied similarly in both groups throughout the study (data not shown).
Effect of rhGH treatment on glucose tolerance
Fasting glycemia was not significantly different between the groups over the 3-yr follow-up (Table 4). Fasting insulinemia increased more in the rhGH-treated group than in the control group (mean change, +5.2 ± 16 [-34; 36] and –2.3 ± 5 [-14; 6], respectively; P = 0.04). After 3 yr, fasting hyperinsulinemia was found in a larger proportion of rhGH-treated patients (seven of 15, 47%) than controls (one of 15, 7%) (P = 0.03). Mean HbA1c levels remained within the normal range over the 3-yr follow-up, although at the end of the second year, seven patients given rhGH and one control had HbA1c levels slightly above normal (P = 0.03). Transient glucose intolerance was observed in one control at baseline and in two rhGH-treated patients after 3 yr. These three patients had severe JIA and a prednisone dosage greater than 0.7 mg/kg·d. In an additional patient, asymptomatic diabetes mellitus was detected by OGTT after 2 yr of rhGH therapy, during a JIA relapse requiring 0.9 mg/kg·d of prednisone. No insulin therapy was needed, and glucose tolerance improved after reductions in the rhGH and prednisone dosages, although glucose intolerance was still noted after 3 yr of rhGH therapy.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In previous investigations of rhGH in children receiving chronic glucocorticoid therapy, rhGH was started after the development of severe adverse effects of glucocorticoid therapy. In our study, in contrast, rhGH was given preventively, based on knowledge about linear growth patterns in JIA and growth effects of rhGH in glucocorticoid-treated children. Early, severe, and sustained loss of height occurs during the active phase of JIA, leading to a decrease in final height (1). Conflicting data have been reported concerning GH secretion in JIA patients (4, 10, 24). In earlier studies, rhGH treatment in dosages ranging from 0.15–0.46 mg/kg·wk (5, 7, 10, 25) normalized height velocity in children receiving long-term glucocorticoid therapy; however, catch-up growth was variable, and height gains were smaller than in other conditions treated with rhGH (5, 7, 25). On the other hand, experimental studies reported evidence of GH resistance during sepsis and inflammation (26, 27, 28) and during glucocorticoid therapy (11). Based on these data, we chose to treat our patients with supraphysiological dosages of rhGH. Our patients differed from those in other studies; they were younger, had a shorter history of inflammation and glucocorticoid therapy, and had a normal height SDS at rhGH initiation despite a dramatic decrease in height velocity. In the rhGH-treated group, the rapid normalization of growth velocity during the first year of treatment completely reversed the loss of height, as shown by the comparison with the controls. Thus, early rhGH treatment preserved the height potential during the acute phase of the disease, so that growth remained within the normal channel. In contrast, height loss occurred in the controls, although its severity was attenuated by the normalization of growth velocity in the third year of follow-up. Several factors may explain this phenomenon: discontinuation of glucocorticoid therapy and onset of puberty in two patients each and improvement of the disease allowing prednisone dosage tapering in one patient.
Inflammation, via cytokine release, is a major determinant of linear growth impairment. In a recent study, TNF
antagonist therapy restored growth velocity in patients with JIA while effectively controlling the inflammation (29). Another study, however, found limited efficacy of etanercept in patients with systemic-onset JIA (30). Moreover, etanercept has been associated with severe treatment-limiting side effects (30). The tight link between inflammation and growth impairment is illustrated by the negative correlation between baseline IGF-I and CRP and between mean height gain and CRP during follow-up in our study of patients with predominantly systemic-onset JIA. However, the difference in growth between the two treatment groups cannot be ascribed to differences in inflammation control, because no significant between-group differences were found in ESR, CRP, or prednisone dosage. Thus, the better growth in the intervention group can be ascribed to rhGH treatment.
In children in the first 3 yr of JIA, smaller lean mass gains, larger fat mass gains, and lower bone mineral content were found, compared with healthy children (31). Earlier studies showed that rhGH therapy significantly improved lean mass and bone mass, particularly during puberty (5, 8, 9). Whether treatment with rhGH, which is an anabolic agent, prevents muscle wasting, bone loss, and excess fat accretion had not been investigated previously. In our study, lean mass (percent body weight or SDS/WFH) was within the normal range at baseline and after 1 yr of glucocorticoid therapy and remained normal in both groups during follow-up. The height increase induced by rhGH treatment was associated with a significantly greater gain in lean mass than in the control group. This change in lean mass was in proportion to the change in height in both control and rhGH-treated children. Further analysis is required to investigate whether rhGH therapy may have a height-independent effect on lean mass in children with JIA.
Severe bone loss, occasionally with vertebral crush fractures, was noted at baseline in our patients. This bone loss can be ascribed to a combination of inflammation, glucocorticoid therapy, suboptimal nutritional status, and reduced physical activity. Treatment with rhGH did not seem to improve bone health in our prepubertal JIA children. These results differ from those of earlier studies, in which significant improvements were noted in bone mineralization as assessed by DEXA (5, 8) or by peripheral quantitative computed tomography (9). However, these patients differed markedly from ours; most of them started puberty during the study and received far smaller prednisone dosages (8, 9), two factors that might have enhanced the beneficial effects of rhGH on bone mineralization. To improve bone health in patients with JIA, rhGH treatment may need to be combined with adequate intakes of calories, calcium, and vitamin D as well as with cautious stimulation of physical activity, because greater muscle mass and strength is associated with greater mechanical resistance of bone (32).
The tolerance and safety of rhGH treatment in children with JIA deserve close attention. JIA outcomes were not significantly different with and without rhGH therapy in our 3-yr study. JIA children treated with rhGH may be at increased risk for alterations in carbohydrate metabolism, because rhGH may add to the effects of chronic inflammation and high-dose glucocorticoid therapy in inducing insulin resistance (33, 34, 35, 36, 37). Transient glucose intolerance was noted in two rhGH-treated patients (and one control) and transient diabetes mellitus in one rhGH-treated patient. None of these patients had symptoms of abnormal glucose metabolism, and all had highly active JIA requiring high glucocorticoid dosages. The effects of rhGH administration on glucose metabolism was less severe than expected. However, carbohydrate metabolism should be monitored closely in patients with JIA who are given rhGH treatment, particularly during the acute phase of their disease.
We found that rhGH therapy in patients with JIA, when started before the onset of growth retardation, normalized prepubertal growth, thus preserving height potential. Tolerance and safety were acceptable, although the data indicated a need for closely monitoring the carbohydrate metabolism. In addition, our results on tolerance and safety should be interpreted with caution, given the short follow-up and small number of patients. Our ongoing study will evaluate the long-term impact of early rhGH treatment on growth and body composition in children with JIA.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: A.-M.P., P.Q., and J.C.R. have nothing to declare; D.S. and P.C. received lecture fees from Pfizer.
First Published Online May 8, 2007
Abbreviations: CA, Chronological age; CRP, C-reactive protein; DEXA, dual-energy x-ray absorptiometry; ESR, erythrocyte sedimentation rate; HbA1c, glycosylated hemoglobin; JIA, juvenile idiopathic arthritis; OGTT, oral glucose tolerance test; rhGH, recombinant human GH; SDS, SD score; WFH, weight for height.
Received December 28, 2006.
Accepted April 30, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
of the insulin-like growth factor I messenger ribonucleic acid response to growth hormone in rat hepatocyte primary culture. Endocrinology 138:1078–1084This article has been cited by other articles:
 |
|
 |
|
  G. Mazziotti, A. Giustina, E. Canalis, and J. P. Bilezikian Treatment of glucocorticoid-induced osteoporosis Therapeutic Advances in Musculoskeletal Diseases, October 1, 2009; 1(1): 27 - 34. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Giustina, G. Mazziotti, and E. Canalis Growth Hormone, Insulin-Like Growth Factors, and the Skeleton Endocr. Rev., August 1, 2008; 29(5): 535 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Varma Growth Hormone Treatment for Children on Chronic Corticosteroids AAP Grand Rounds, October 1, 2007; 18(4): 42 - 42. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
