2025 Volume 72 Issue 8 Pages 925-935
Recombinant human growth hormone (GH; somatropin) treatment has beneficial effects on body composition in patients with Prader-Willi syndrome (PWS). However, this treatment option is limited to children in most countries and to children with short stature in countries such as the USA and Japan. The aim of this multicohort study was to evaluate the effect of somatropin on body composition and to assess its safety in Japanese pediatric and adult participants with PWS. GH-naïve pediatric participants (n = 6) received somatropin 0.245 mg/kg/week, GH-treated pediatric participants (n = 7) received somatropin 0.084 mg/kg/week, and adult participants (n = 20) received somatropin 0.042 mg/kg/week for 1 month, followed by 0.084 mg/kg/week. The study met its primary endpoint in the adult cohort because the least squares mean (95% CI) of the change from baseline to Month 12 in lean body mass (LBM) (%) was greater than the prespecified efficacy criterion of 0. LBM (%) was higher at 12 months in GH-naïve pediatric participants, while GH-treated pediatric participants showed little deterioration in LBM despite reduced GH dosage. Treatment-emergent adverse events (TEAEs) were experienced by five (83.3%), five (71.4%), and 19 (95.0%) participants in the GH-naïve pediatric cohort, GH-treated pediatric cohort, and adult cohort, respectively. Most TEAEs were mild or moderate in severity. Three participants reported four serious TEAEs, and none were treatment related. Somatropin improved body composition in adult participants, enabled maintenance of body composition in pediatric participants, and demonstrated a favorable safety and tolerability profile in all PWS cohorts. (ClinicalTrials.gov ID: NCT04697381)
Prader-Willi syndrome (PWS) is a representative imprinting disorder resulting from an absence of expression of paternally inherited genes in the chromosome 15q11q13 region [1]. The incidence of PWS is approximately 1 in 15,000 live births [2]. Infants with PWS have severe hypotonia, leading to feeding difficulties. Children with PWS often have delayed motor development, language impairment, intellectual disabilities, and behavioral disorders. The clinical manifestation of PWS includes excessive eating caused by hypothalamic pathology, which affects the satiety signal and can result in severe obesity beginning in childhood.
PWS can also have a negative impact on body composition, and patients typically have an increased fat mass and decreased lean body mass (LBM) [3]. Human growth hormone (GH) is a recommended treatment for both pediatric and adult patients with PWS to help mitigate abnormalities in body composition [4]. Somatropin is a recombinant human GH (r-hGH) with an amino acid sequence that is identical to that of the natural or wildtype human GH [5]. The efficacy and safety of GH treatment have been demonstrated in several randomized clinical trials in children [6, 7] and adults with PWS [8-10]. In a randomized controlled trial, prepubertal children with PWS were randomized to GH treatment or no treatment for the first year; children older than 3 years of age continued for an additional year with the same treatment while all infants received GH treatment in the second year. GH treatment in these children improved body proportions and body fat percentage [7]. In two adult studies, adults with PWS were randomized to receive either GH or placebo for periods ranging from 6 months [8] to 12 months [9], followed by open-label periods where all patients received GH treatment [10, 11]; GH treatment in these adults reduced fat mass and increased LBM. Further, a 2-year crossover study (comparing placebo with GH) in young adults with PWS demonstrated that continuing GH treatment in adulthood enabled maintenance of improved body composition without safety concerns [12].
Somatropin is approved for the treatment of pediatric PWS regardless of stature in European countries, and for children and adults with PWS in New Zealand. In the USA, somatropin is approved for treatment of children with growth failure due to PWS. GH replacement has been approved in Japan since 2002 for “short stature without epiphyseal closure associated with PWS” but at the time the study was conducted, it was not yet approved for improving body composition in pediatric patients or for any use in adult patients with PWS. Moreover, improvement of body composition in patients with PWS had been identified as a high medical need by the Ministry of Health, Labour and Welfare (MHLW) Committee on Unapproved/Off-label used Drugs. However, to date, there have been no clinical studies conducted in Japanese participants with PWS evaluating improvement in body composition as a primary endpoint. The aim of this study was to evaluate the effect of somatropin on body composition and to assess its safety in Japanese participants with PWS following 12 months of treatment.
This multicenter, open-label, multicohort study (ClinicalTrials.gov ID: NCT04697381) included both pediatric and adult participants with PWS across three cohorts (Supplementary Fig. 1). Cohort 1A was chosen to represent younger pediatric patients with PWS who are naïve to GH treatment while Cohort 1B was chosen to represent adolescent patients with PWS currently receiving GH treatment who were about to complete their GH treatment as a means of growth promotion and were switching to a lower GH dose for maintenance of body composition. Participants in Cohort 1B were transitioning from the dose recommended for children to that recommended for adults; part of the rationale for this cohort was to explore the effect of a reduced GH dose on body composition in pediatric patients who were already being treated with GH. In the GH-naïve pediatric cohort (Cohort 1A), participants received somatropin at 0.245 mg/kg/week. This dose was chosen as it was identical to both the approved dose for short stature associated with PWS in Japan as well as the approved dose for pediatric PWS in Europe. Participants in the GH-treated pediatric cohort (Cohort 1B) were already being treated with somatropin (at the approved dose for short stature) before entering the study; during the study, Cohort 1B received somatropin at a reduced dose of 0.084 mg/kg/week, with dose adjustments based on serum IGF-1 level if required but not exceeding 1.6 mg/day. In the adult Cohort (Cohort 2), GH-naïve participants received somatropin at 0.042 mg/kg/week for 1 month, followed by 0.084 mg/kg/week with dose adjustments based on serum IGF-1 level if required but not exceeding 1.6 mg/day. For Cohorts 1B and 2, the dose of 0.084 mg/kg/week was chosen based on the similarities between the clinical characteristics of PWS vs. adult GHD. In Cohort 1B, this dose was approximately three-fold lower than the dose that participants were receiving before enrollment in the study. Adjusting dosage in Cohorts 1B and 2 based on IGF-1 level with a maximum dose of 1.6 mg/day were consistent with the Consensus Guidelines for Recombinant Human Growth Hormone Therapy in PWS [4]. The weekly dose of somatropin was divided into six or seven subcutaneous injections, and all cohorts received somatropin for 12 months. Following the 12-month treatment period, participants could opt to continue to receive treatment for an extension period of 36 months or until regulatory approval, whichever occurred sooner.
ParticipantsParticipants were male or female with a genetically confirmed diagnosis of PWS. They had no plan to initiate a new treatment that would affect their body composition (for the duration of this clinical study), such as gonadal hormone replacement therapy, and were currently on appropriate diet and exercise programs that they were willing to continue throughout the study period. Participants eligible for inclusion in Cohort 1A were naïve to GH treatment and were Tanner stage 1 (testes in males; breasts in females). Participants eligible for inclusion in Cohort 1B had to have received GH treatment for at least 2 years and were receiving GH at the time of study inclusion. Participants in this cohort also had to have a GH dose that had been stable for the previous 6 months, with their most recent dose being higher than 0.084 mg/kg/week. Eligible participants in Cohort 1B were also about to complete GH treatment for short stature (for example, they had met the treatment stopping criteria defined as a height SD score [SDS] more than –2.5 for Japanese adult standards). Eligible participants in Cohort 2 were 18 years or older, had not received GH treatment for at least 1 year, and had a serum IGF-1 level below +2 SDS, adjusted for age and sex.
Participants were excluded from the study if they had uncontrolled diabetes, malignant tumors, severe obesity, or serious respiratory impairment. Participants could also be excluded if they had medical or psychiatric conditions that made study participation inappropriate. Any newly initiated medications/non-drug treatments that could affect body composition, such as hormonal replacement therapy, medications for hyperlipidemia, anti-obesity medication, and surgery accompanied with metal implants were prohibited during the study. GH products other than study intervention and investigational drugs were also prohibited.
Study endpoints and assessments EfficacyThe primary endpoint set for all three cohorts was change from baseline to Month 12 in LBM (%) (measured using dual-energy X-ray absorptiometry [DEXA]) but the primary efficacy criteria was only set for the adult cohort and not for the two pediatric cohorts. The pre-specified efficacy criterion was set to 0 for the adult cohort. Change in LBM from baseline to Month 6 was a secondary endpoint in Cohort 2. Other secondary endpoints included changes from baseline to Month 12 in LBM (%) measured by bioelectrical impedance analysis (BIA), change from baseline to Month 12 in body fat (%) measured by DEXA, and change from baseline to Month 12 in subcutaneous and visceral adipose tissue distribution measured by abdominal computed tomography (CT) scan. Body fat (%) was calculated as follows: 100 – LBM (%). The change in IGF-1 SDS from baseline to each visit was recorded as an exploratory endpoint.
SafetySafety assessments including treatment-emergent adverse events (TEAEs) and serious AEs (SAEs), discontinuations, changes in vital signs, clinical laboratory parameters, and bone maturation were conducted. TEAEs were coded using the Medical Dictionary for Regulatory Activities Version 25.1 and were classified by System Organ Class and Preferred Term.
Statistical analysesThe full analysis set (FAS) consisted of all participants who took at least one dose of study intervention. The efficacy evaluable set (EES) consisted of all participants who took at least one dose of study intervention and had at least one efficacy evaluation. Demographic and baseline characteristics and safety results were based on the FAS; efficacy results were based on the EES. Efficacy was evaluated based on individual data and summary statistics in each cohort for the primary and secondary endpoints. Descriptive statistics included measures of N, counts and percentages for categorical variables, and N, mean, SD, median, range, minimum, and maximum for continuous variables. No formal hypothesis testing was conducted, but two-sided p-values of the Wilcoxon signed-rank test between median values at baseline and each measured time point were calculated post hoc for exploratory purposes. We chose to use non-parametric tests due to uncertainty around whether the data were normally distributed and the small sizes in every cohort. In addition, for Cohort 2, p-values of least squares mean (LSM) of change from baseline based on a mixed-effect model repeated measures (MMRM) were calculated.
For Cohort 2, the LSM and 95% confidence interval (CI) of change from baseline to Months 6 and 12 in LBM (%) measured by DEXA and change from baseline to Months 1, 3, 6, 9, and 12 in LBM (%) measured by BIA were estimated based on a MMRM. The model included fixed effects for visit as a categorical variable and baseline LBM value (%) and interaction between baseline LBM value (%) and visit as continuous variables. In Cohort 2, the prespecified primary efficacy criterion was the point estimate of the LSM of the primary endpoint being above 0. For participants who experienced a major deviation of diet/exercise or another major protocol deviation and/or discontinuation of treatment or study withdrawal, efficacy data after the major deviation and discontinuation were not included in the descriptive statistics and LSM. Safety parameters were descriptively summarized; no imputation was made for missing data.
EthicsThe study was conducted in accordance with the International Council for Harmonisation Guideline for Good Clinical Practice and the Declaration of Helsinki. The protocol was approved by the institutional review board of the participating centers. Each participant provided signed informed consent before any study procedures commenced.
Overall, 33 participants with PWS were enrolled; there were six, seven, and 20 participants in Cohorts 1A, 1B, and 2, respectively. All participants were Asian (Table 1). The percentage of male participants was 66.7%, 57.1%, and 35.0% in Cohorts 1A, 1B, and 2, respectively; the median age in the three cohorts was 6.0, 14.0, and 24.5 years (Table 1). The mean ± SD time between diagnosis of PWS and study initiation was 3.4 ± 2.11, 13.0 ± 4.16, and 19.5 ± 8.00 years for Cohorts 1A, 1B, and 2, respectively. All participants completed the 12-month treatment phase and entered the extension period.
| Cohort 1A (GH-naïve pediatric cohort) (n = 6) |
Cohort 1B (GH-treated pediatric cohort) (n = 7) |
Cohort 2 (Adult cohort) (n = 20) |
|
|---|---|---|---|
| Age, years | 6.0 (2, 10) | 14.0 (11, 19) | 24.5 (18, 38) |
| Male sex, n (%) | 4 (66.7) | 4 (57.1) | 7 (35.0) |
| Asian race, n (%) | 6 (100.0) | 7 (100.0) | 20 (100.0) |
| Ethnicity not Hispanic or Latino, n (%) | 6 (100.0) | 7 (100.0) | 20 (100.0) |
| Weight, kg | 20.2 (10.3–48.4) | 43.4 (36.4–62.5) | 58.5 (36.2–92.9) |
| Height, cm | 105.1 (84.5–131.1) | 154.5 (139.8–162.7) | 151.5 (144.6–169.0) |
| Height SDSa | –1.7 (–2.6, –0.4) | –1.4 (–3.4, 0.2) | –1.5 (–3.5, 0.5) |
| BMIb, kg/m2 | 18.9 (14.2–28.2) | 19.1 (14.4–26.2) | 22.3 (16.0–39.3) |
| Lean body mass (%) (DEXA) | 53.8 (50.5–70.9) | 72.3 (60.5–82.6) | 59.8 (49.3–67.9) |
| Lean body massc (%) (BIA) | 60.9 (47.9–77.7) | 69.0 (64.6–89.3)e | 62.4 (47.4–73.8) |
| Body fatd (%) (DEXA) | 46.2 (29.1–49.5) | 27.7 (17.4–39.5) | 40.2 (32.1–50.7) |
| Body fat (%) (BIA) | 39.1 (22.3–52.1) | 31.0 (10.7–35.4)e | 37.6 (26.2–52.6) |
| Subcutaneous adipose tissue distribution (cm2) | 100.7 (16.6–327.1) | 91.4 (17.9–468.2) | 207.8 (65.8–562.9) |
| Visceral adipose tissue distribution (cm2) | 19.1 (6.3–65.2) | 17.7 (10.5–202.9) | 59.7 (18.5–116.4) |
| IGF-1 SDS | –1.3 (–2.1, 0.2) | 0.4 (–1.4, 2.2) | –2.1 (–4.1, 2.2) |
| Waist circumference (cm) | 61.3 (41.7–90.5) | 67.0 (62.0–87.4) | 83.2 (62.5–122.0) |
BIA, bioelectrical impedance analysis; BMI, body mass index; DEXA, dual-energy X-ray absorptiometry; GH, growth hormone; IGF-1, insulin-like growth factor-1; SDS, SD score.
Values are expressed as median (range) unless otherwise specified.
aCalculated using the mean and SD value of height at age 17 years and 6 months for participants aged >17 years and 6 months.
bBody mass index (kg/m2) = weight (kg)/[height (cm) × 0.01]2.
cLean body mass (%) = lean body mass (kg)/(lean body mass [kg] + body fat [kg]) * 100.
dBody fat (%) = body fat (kg)/(lean body mass [kg] + body fat [kg]) * 100.
eBased on 6 participants
The overall LBM (%) and change from baseline in each cohort measured by DEXA are presented in Table 2. The LBM (%) at Month 12 appeared to be higher than baseline for Cohort 1A (p-value: 0.0313). In Cohort 1B, LBM (%) at Month 12 was not significantly different to baseline (p-value: 0.3125). For Cohort 2, the median (range) change from baseline to Months 6 and 12 in LBM (%) measured by DEXA was 2.2 (–2.0, 7.6; p-value: 0.0006) and 1.8 (–0.4, 9.3; p-value: 0.0001), respectively (Table 2). The LSM (95% CI) change from baseline to Months 6 and 12 in LBM (%) was positive for Cohort 2 (Table 3). The primary efficacy criterion was met because the point estimate of change from baseline was greater than the prespecified efficacy criterion of 0 for the adult cohort.
| Cohort 1A (GH-naïve pediatric cohort) (n = 6) |
Cohort 1B (GH-treated pediatric cohort) (n = 7) |
Cohort 2 (Adult cohort) (n = 20) |
||||
|---|---|---|---|---|---|---|
| Analysis visit | Observed | Change from baseline | Observed | Change from baseline | Observed | Change from baseline |
| Baseline | ||||||
| n | 6 | 7 | 20 | |||
| Median (range) | 53.8 (50.5–70.9) | — | 72.3 (60.5–82.6) | — | 59.8 (49.3–67.9) | — |
| Month 6 | ||||||
| n | 20 | 20 | ||||
| Median (range) | — | NR | — | NR | 64.5 (53.6–70.0) | 2.2 (–2.0, 7.6) |
| p-value* | ||||||
| Wilcoxon signed-rank test | 0.0006 | |||||
| Month 12 | ||||||
| n | 6 | 6 | 6 | 6 | 19 | 19 |
| Median (range) | 60.2 (55.0–73.0) | 3.0 (1.3–13.1) | 66.2 (61.3–79.3) | –2.0 (–5.1, 3.1) | 64.7 (54.6–70.3) | 1.8 (–0.4, 9.3) |
| p-value* | ||||||
| Wilcoxon signed-rank test | 0.0313 | 0.3125 | <0.0001 | |||
* Post hoc test
DEXA, dual-energy X-ray absorptiometry; GH, growth hormone; NR, not reported.
One adult participant had a major deviation of dietary control and one participant in the GH-treated pediatric cohort underwent funnel chest surgery, and therefore the data measured by DEXA at Month 12 were excluded from analyses.
| Lean body mass measurement | Analysis visit | n | LSM estimate (SE); 95% CI; p-value* |
|---|---|---|---|
| DEXA | Month 6 | 20 | 2.38 (0.514); 1.30–3.46; 0.0002 |
| Month 12 | 19 | 3.09 (0.591); 1.85–4.33; <0.0001 |
* Post hoc test
CI, confidence interval; DEXA, dual-energy X-ray absorptiometry; LSM, least squares mean.
LSM and its 95% CI of change from baseline are estimated based on a mixed-effect model for repeated measures. Visit, baseline value, and baseline value-by-visit interaction are fixed effects in the model. p-Value was calculated based on two-sided t test.
One adult participant had a major deviation of dietary control, and therefore the data measured by DEXA (Month 12) were excluded from analyses.
With regard to the methodology for measuring LBM, the median LBM (%) at baseline measured by DEXA was consistent with the measurements by BIA for all 3 cohorts (Table 1). For Cohort 2, the LSM estimates of the change from baseline in LBM (%) over time as measured by BIA (Supplementary Table 1) showed a similar trend to those measured by DEXA at Months 6 and 12 (Table 3). BIA measurements of fat mass (%) showed the same trend as the DEXA measurements of LBM (%) for each cohort; these DEXA and BIA measurements were highly correlated across all cohorts (correlation coefficient: 0.794) (Supplementary Fig. 2).
Change from baseline in body fat was measured by DEXA at Month 12 in all cohorts. In Cohort 2, the LSM (95% CI) of change from baseline to Month 12 in body fat (%) was –3.1 (–4.33, –1.85) (Supplementary Table 2). The median (range) change from baseline to Month 12 in body fat (%) for Cohort 1A, 1B, and 2 was –2.96 (–13.1, –1.3), 2.0 (–3.1, 5.1), and –1.8 (–9.3, 0.4), respectively (Supplementary Table 3).
Adipose tissue distributionIn Cohorts 1A and 2, subcutaneous and visceral adipose tissue distribution (measured by abdominal CT scan) at Month 12 were not significantly different to the values observed at baseline (Table 4). For Cohort 1B, visceral adipose tissue distribution was lower at Month 12 compared with baseline (p-value: 0.0313), although this difference was not large (Table 4). Subcutaneous tissue distribution at Month 12 was not significantly different to baseline for Cohort 1B.
| Cohort 1A (GH-naïve pediatric cohort) (n = 6) |
Cohort 1B (GH-treated pediatric cohort) (n = 7) |
Cohort 2 (Adult cohort) (n = 20) |
||||
|---|---|---|---|---|---|---|
| Analysis visit | Observed | Change from baseline | Observed | Change from baseline | Observed | Change from baseline |
| Subcutaneous adipose tissue, cm2 | ||||||
| Baseline | ||||||
| n | 6 | — | 7 | — | 20 | — |
| Median (range) | 100.7 (16.6–327.1) | 91.4 (17.9–468.2) | 207.8 (65.8–562.9) | |||
| Month 12 | ||||||
| n | 6 | 6 | 6 | 6 | 19 | 19 |
| Median (range) | 95.6 (15.7–667.0) | 7.8 (–24.3, 340.0) | 119.0 (35.9–1002.5) | 17.6 (–1.3, 534.3) | 182.4 (85.2–354.0) | –14.9 (–64.7, 45.6) |
| p-value* | ||||||
| Wilcoxon signed-rank test | 0.3125 | 0.0625 | 0.0799 | |||
| Visceral adipose tissue, cm2 | ||||||
| Baseline | ||||||
| n | 6 | — | 7 | — | 20 | — |
| Median (range) | 19.1 (6.3–65.2) | 17.7 (10.5–202.9) | 59.7 (18.5–116.4) | |||
| Month 12 | ||||||
| n | 6 | 6 | 6 | 6 | 19 | 19 |
| Median (range) | 20.4 (4.3–97.2) | 0.1 (–4.4, 32.0) | 14.9 (8.6–131.3) | –1.6 (–71.6, –0.7) | 47.5 (12.0–134.5) | –3.3 (–46.5, 18.1) |
| p-value* | ||||||
| Wilcoxon signed-rank test | 0.8438 | 0.0313 | 0.4900 | |||
* Post hoc test
CT, computed tomography; GH, growth hormone.
One adult participant had a major deviation of dietary control and one participant in the GH-treated pediatric cohort underwent funnel chest surgery, and therefore the data measured by CT at Month 12 were excluded from analyses.
The mean IGF-1 SDS at baseline was –1.3 in Cohort 1A, 0.5 in Cohort 1B, and –2.0 in Cohort 2. For both Cohort 1A and Cohort 2, the mean IGF-1 SDS appeared to be higher than baseline across post-baseline study visits (Fig. 1). In Cohort 1A, mean IGF-1 SDS was >+2 SDS at Months 3 and 12 (Fig. 1). In Cohort 1B, mean IGF-1 SDS was similar to baseline at all study visits.
Mean values plotted and error bars represent SD.
aWilcoxon signed-rank tests represent comparisons against baseline (post hoc).
GH, growth hormone; IGF-1, insulin-like growth factor-1; SDS, SD scores
The median body mass index (BMI) SDS in both Cohorts 1A and 1B was not significantly different to baseline at all time points (Supplementary Table 4). In Cohort 2, the median BMI SDS was not significantly different to baseline at all time points except for Months 1 and 3, where it appeared to be higher than baseline (Supplementary Table 4).
SafetyAt least one TEAE was experienced by five (83.3%), five (71.4%), and 19 (95.0%) participants in Cohorts 1A, 1B, and 2, respectively. Most TEAEs were mild or moderate in severity. Three participants reported four SAEs (ankle fracture, brain contusion, skull fractured base, and hypertension), and none were related to the study drug. Two (33.3%) participants in Cohort 1A, one (14.3%) participant in Cohort 1B, and seven (35.0%) participants in Cohort 2 had dose reductions or temporarily discontinued from the study due to AEs. There were no deaths reported during the study.
Treatment-related TEAEs were reported in three (50.0%), zero, and nine (45.0%) participants in Cohorts 1A, 1B, and 2, respectively (Table 5); none led to treatment or study discontinuation. In Cohort 1A, no treatment-related TEAEs were reported in more than one participant. Treatment-related TEAEs reported in more than one participant in Cohort 2 were increased glycosylated hemoglobin (n = 3 [15.0%]) and increased IGF (n = 2 [10.0%]).
| Number of participants, n (%) | Cohort 1A (GH-naïve pediatric cohort) (n = 6) |
Cohort 1B (GH-treated pediatric cohort) (n = 7) |
Cohort 2 (Adult cohort) (n = 20) |
|---|---|---|---|
| Any AE | 3 (50.0) | 0 | 9 (45.0) |
| Injection site reaction | 0 | 0 | 1 (5.0) |
| Blood triglycerides increased | 0 | 0 | 1 (5.0) |
| Body height increased | 1 (16.7) | 0 | 0 |
| Glycosylated hemoglobin increased | 1 (16.7) | 0 | 3 (15.0) |
| Insulin-like growth factor increased | 1 (16.7) | 0 | 2 (10.0) |
| Diabetes mellitus inadequate control | 0 | 0 | 1 (5.0) |
| Glucose tolerance impaired | 0 | 0 | 1 (5.0) |
| Type 2 diabetes mellitus | 0 | 0 | 1 (5.0) |
| Arthralgia | 0 | 0 | 1 (5.0) |
| Hypoesthesia | 1 (16.7) | 0 | 0 |
| Snoring | 1 (16.7) | 0 | 0 |
| Hemorrhage subcutaneous | 1 (16.7) | 0 | 0 |
AE, adverse event; GH, growth hormone; TEAE, treatment-emergent AE.
Participants are only counted once per treatment per event. Medical Dictionary for Regulatory Activities Version 25.1 coding dictionary applied. Includes events started on or before the date of Month 12 visit/follow-up.
Overall, seven participants experienced at least one AE of special interest (AESI), all of which were related to glucose metabolism impairment and were treatment related. The most frequently reported AESI was an increase in glycosylated hemoglobin, which was observed in one participant (16.7%) in Cohort 1A and three participants (15%) in Cohort 2. In addition, AESIs of diabetes mellitus inadequate control, glucose tolerance impaired, and type 2 diabetes mellitus were reported by one participant (0.5%) each in Cohort 2. Therefore, there were six instances of glucose metabolism-related AESIs observed in Cohort 2. The participants who reported diabetes mellitus inadequate control and type 2 diabetes mellitus had pre-existing diabetes prior to study commencement; a total of four participants in Cohort 2 had pre-existing diabetes. All AESIs reported were mild or moderate in severity.
One adult male participant reported severe sleep apnea that led to discontinuation from the study during the extension period (following the 12-month main study period), but this AE was not considered to be treatment related. There were no reports of sleep apnea disorder in either Cohort 1A or 1B. Other findings included auxological data, such as height SDS. From baseline to Month 12, mean (SD) height SDS in Cohort 1A increased by 1.2 (0.45). In contrast, there was little change in height SDS in Cohorts 1B and 2 (mean ± SD: 0.1 ± 0.31 and 0.1 ± 0.12, respectively) over the same period. Somatropin did not accelerate bone age; bone maturation at Month 12 was <1.0 in most evaluable participants.
Based on the laboratory values shown in Supplementary Table 5, clinical laboratory test abnormalities observed in >20% of evaluable participants (including participants with abnormal baseline) included the following: elevated alkaline phosphatase (>3.0 × upper limit of normal [ULN]) in 2/6 (33.3%) participants and elevated triglycerides (>1.3 × ULN) in 2/6 (33.3%) participants in Cohort 1A; elevated thyrotropin (>1.2 × ULN) in 2/7 (28.6%) participants in Cohort 1B; and elevated glucose levels (>1.5 × ULN) in 4/5 (80%) participants, elevated triglycerides (>1.3 × ULN) in 11/20 (55%) participants, and elevated low-density lipoprotein cholesterol (>1.2 × ULN) in 7/20 (35%) participants in Cohort 2. Cohort 1A and 2 each had one participant with ALT (>3.0 × ULN). There were no aspartate aminotransferase–related abnormalities reported, and no participants in the study met the criteria for Hy’s Law. The mean ± SD change in glycated hemoglobin from baseline for Cohort 1A, 1B, and 2 was 0.3 ± 0.4, 0.0 ± 0.1, and 0.3 ± 0.3%, respectively.
In this study, we investigated the efficacy of somatropin to improve body composition in two pediatric cohorts (Cohorts 1A and 1B) and one adult cohort (Cohort 2) of Japanese participants with PWS (Graphical Abstract). The primary endpoint in Cohort 2 was met because the LSM (95% CI) of the change from baseline to Month 12 in LBM (%) was greater than the prespecified efficacy criterion of 0 for this cohort. In Cohort 1A, LBM (%) at Month 12 of somatropin treatment was higher than the value at baseline, as measured by DEXA. In Cohort 1B, the LBM (%) at Month 12 was not significantly different from baseline. This study also showed that LBM measurements made using BIA were largely consistent with measurements made using DEXA.
Taken together, the study findings suggest that somatropin treatment improved body composition parameters, specifically LBM (%) in adult participants. In Cohort 1B, there was little deterioration in LBM despite the reduction of GH dosage. The fact that body composition was able to be maintained in this cohort despite the reduction in GH dosage is consistent with two reports of an increase in body fat mass or exacerbation of BMI after cessation of GH treatment in other Japanese studies [13, 14]. The data from this study are consistent with the results of randomized controlled trials in other countries that examined the effect of somatropin in participants with PWS. A 1-year, double-blind, randomized controlled trial of somatropin versus placebo (n = 46) in adult Scandinavian participants with PWS reported that GH treatment increased LBM by 2.25 kg (95% CI: 0.725–3.770; p = 0.005) and decreased fat mass by 4.20 kg (95% CI: –6.40, –2.00; p < 0.001) versus placebo [9]. A 2-year, randomized, double-blind, placebo-controlled crossover study of young adults in the Netherlands investigating the effects of 1 year of somatropin treatment versus 1 year of placebo reported that compared with placebo, somatropin lowered fat mass (–2.9 kg; p = 0.004; relative change, –17.3%) and raised LBM (1.5 kg; p = 0.005; relative change, +3.5%) [12]. Similar results were reported in a study where adult participants with PWS were randomized to somatropin or placebo for 6 months, followed by open-label somatropin for all participants for 12 months [8]. During the randomized phase, somatropin significantly reduced body fat versus placebo. In the somatropin open-label phase, body fat was significantly reduced by 2.5% and LBM was significantly increased by 2.2 kg in participants with a PWS genotype [8]. Studies have also demonstrated the efficacy of somatropin over 2 years. A single-arm study in which adult participants received somatropin for 2 years showed that compared with baseline, DEXA-measured LBM increased by 2.8 kg (95% CI: 1.9–3.6; p < 0.001), whereas fat mass decreased by 3.0 kg (95% CI: 1.1–4.8; p = 0.003) [10].
Patients with PWS can be predisposed to predominantly subcutaneous (rather than visceral) deposition of fat [15]. In this study, Cohort 2 showed a trend toward lower subcutaneous and visceral adipose tissue following treatment with somatropin, consistent with a previous study comparing somatropin with placebo in adults with PWS and further demonstrating the beneficial impact of somatropin on body composition parameters [9]. Similarly, a study in adult participants with PWS demonstrated that 2 years of GH treatment led to a significant decrease in subcutaneous adipose tissue (–12.3%, p = 0.01) and a trend toward lower visceral adipose tissue (–6.5%, p = 0.18) [10]. In the pediatric cohorts (Cohorts 1A and 1B), subcutaneous adipose tissue distribution at Month 12 was not significantly different from baseline; the same was observed for visceral adipose tissue distribution in Cohort 1A. Cohort 1B had lower visceral adipose tissue distribution at Month 12 but this difference was not large (Table 4). Somatropin treatment did not appear to cause a relative increase in visceral adipose tissue in the cohorts investigated, which is important given that a relative increase in visceral adipose tissue can result in visceral obesity and its associated complications. While not surprising for Cohorts 1A and 2, which initiated GH treatment during the study, this result was especially notable in Cohort 1B as the reduced somatropin dose (relative to dosage received prior to study) did not appear to be associated with increased adipose tissue.
Somatropin administration was generally safe and well tolerated in Japanese participants with PWS, and no new safety signals were identified. This finding is consistent with the known safety profile of somatropin, which has been confirmed in several previous studies. No participants withdrew from the study due to TEAEs in the treatment phase, and no treatment-related SAEs were reported. Given the well-established diabetogenic effect of GH treatment [16], TEAEs related to glucose metabolism were of special interest. In this study, seven treatment-related TEAEs related to impaired glucose metabolism were reported, but none led to study or treatment discontinuation. The AEs of glucose tolerance impaired, type 2 diabetes mellitus, and glycosylated hemoglobin increased (GH-naïve pediatric cohort) were not resolved during the study. Taken together and in accordance with previously published data from the KIGS study [17], with careful monitoring of glucose metabolism, somatropin was safe to use in participants with PWS.
In the Cohorts 1A and 2, IGF-1 SDS appeared to increase during somatropin treatment, which is consistent with previous observations in infants and prepubertal children with PWS undergoing somatropin treatment [7]. In particular, mean IGF-1 SDS in Cohort 1A appeared to be higher than for the other cohorts, with levels >+2 SDS at Months 3 and 12 (Fig. 1). As suggested by Festen and coworkers [7], patients with PWS receiving treatment with somatropin should undergo careful IGF-1 monitoring to ensure that these levels do not exceed the normal range.
To our knowledge, this is the first study in Japanese participants with PWS looking at body composition as an endpoint following treatment with somatropin. Over the past 10 years there have been relatively few studies on GH treatment of patients with PWS, and during this time there have been considerable improvements in the treatment and care of these patients. One of the strengths of this study is that it characterizes the effectiveness of somatropin in participants with PWS in the context of the current clinical landscape. For example, current multidisciplinary care combined with early diagnosis and GH treatment has resulted in a decrease in the prevalence of obesity among patients with PWS [18]. Another strength of this study is the fact that the effect of somatropin was studied in both pediatric and adult participants. Further, the three patient groups chosen for the study are broadly representative of patients with PWS who would receive somatropin in real-world clinical practice, thus enabling the evaluation of the impact of treatment on patients with PWS at different stages of life (pediatric patients who have never received somatropin, pediatric patients who are currently receiving somatropin, and patients receiving somatropin as adults). The results from the GH-treated pediatric cohort suggest that continued somatropin treatment is not associated with any new safety issues and is likely to be beneficial in this patient population. This is consistent with a previous study showing that young adults with PWS still derive benefit (in terms of fat mass and LBM) from continued somatropin treatment [12]. Due to the small number of patients with PWS in the population, this study was limited to a single arm. Before this study commenced, a survey questionnaire was sent to major Japanese PWS societies and major patient advocacy groups for PWS to estimate the number of patients likely to meet the key eligibility criteria for this study. The results of this survey confirmed that there were only a limited number of patients in Japan who would likely be eligible for enrollment in this study within a reasonable period of time. These findings, combined with the rarity of PWS as a disease in Japan, meant that pediatric and transition cohorts had a very limited number of study participants. Therefore, no formal hypothesis testing was conducted to validate the efficacy of somatropin, but exploratory post hoc analyses were conducted for several parameters. In contrast, due to the larger sample size in the adult cohorts, a model analysis could be performed to define efficacy.
In conclusion, consistent with previous findings, Japanese participants with PWS benefited from treatment with somatropin over 12 months in terms of improved body composition. Somatropin was generally safe and well tolerated in both pediatric and adult participants.
This study was sponsored by Pfizer.
The authors thank the participating physicians who contributed to this study. Medical writing support was provided by Chu Kong Liew, PhD, of Engage Scientific Solutions and was funded by Pfizer.
Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions and exceptions, Pfizer may also provide access to the related individual de-identified participant data. See https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information.
Masanobu Kawai was a lecturer for Novo Nordisk and Pfizer. Reiko Horikawa has been on advisory boards of Lumos Pharma, Novo Nordisk, OPKO Health, and Pfizer; has received consulting fees from Ascendis Pharma, Lumos Pharma, Novo Nordisk, Pfizer, and Sandoz; has been a recipient of grants from Novo Nordisk and Sandoz; and has been on speaker’s bureaus for Eli Lilly & Company, Novo Nordisk, Pfizer, and Sandoz. Yuko Hoshino and Akifumi Okayama are full-time employees of Pfizer R&D Japan. Takahiro Sato and Nozomi Ebata are full-time employees of Pfizer Japan Inc. Nobuyuki Murakami, Koji Muroya, and Yasuko Fujisawa have no conflicts of interest to declare. Tsutomu Ogata has received honoraria from JCR and Novo Nordisk.
adverse event of special interest
BIAbioelectrical impedance analysis
BMIbody mass index
CIconfidence interval
CTcomputed tomography
DEXAdual-energy X-ray absorptiometry
EESefficacy evaluable set
FASfull analysis set
GHgrowth hormone
IGF-1insulin-like growth factor-1
LBMlean body mass
LSMleast squares mean
MMRMmixed-effect model repeated measures
PWSPrader-Willi syndrome
r-hGHrecombinant human growth hormone
SDSSD score
TEAEtreatment-emergent adverse event
ULNupper limit of normal
