Introduction

Silver-Russell syndrome (SRS, OMIM #180860) is a rare congenital imprinting disorder associated with the loss of methylation in H19/IGF2:IG-DMR at chromosome 11p15.5 (11p15 LOM) or maternal uniparental disomy of chromosome 7 (upd(7)mat) found in 30–60% and in 5–10% of patients, respectively [1–4]. The molecular aetiology remains unknown in about 40% of patients with clinical symptoms of SRS [5]. In 2017, the first international consensus about the diagnosis and management of Silver-Russell syndrome was published [4]. Patients with SRS are characterized by intrauterine and postnatal growth retardation — their growth rate is slow, without catch-up [2, 4, 6, 7]. Children with SRS show a wide spectrum of minor dysmorphic and clinical features: triangular face, small mandible, irregular and crowded teeth, down-turned mouth, low-set and/or posteriorly rotated ears, clinodactyly of the fifth finger and/or syndactyly of 2 to 3 toes, delayed closure of the frontal fontanel, low muscle mass, excessive sweating in early childhood, or spinal deformity [2, 3, 8–13].

In many countries, the treatment of children with SRS using recombinant human growth hormone (rhGH) has been used for many years [14-16]. In Poland, rhGH is assigned for SRS patients within the framework of 2 national programs: for children with growth hormone deficiency (GHD), or for children born small for gestational age (SGA). The latter has been available since 2015. The effectiveness of rhGH therapy for increasing height (SDS) is well documented, i.e. in children with isolated growth hormone deficiency, idiopathic short stature, SGA, or patients with Noonan syndrome, Turner syndrome, and Prader-Willi syndrome [17-19].

Growth hormone treatment has several metabolic effects in children born SGA, such as a decrease in fat mass and an increase in lean body mass [20]. Similar effects were observed in patients with GHD and in patients born SGA [21, 22].

This study aimed to analyse the effect of the administered rhGH on anthropometric parameters and body composition, especially with respect to molecular abnormalities and, additionally, considering gender and rhGH treatment programs, in patients with SRS during 3 years of rhGH therapy.

Material and methods

Patients

Thirty-one patients with SRS: 18 (58.1%) boys and 13 (41.9%) girls were included in this study. All the patients had the following genetically confirmed abnormalities: 11p15 LOM in 23 (74.2%) cases, and upd(7)mat in 8 (25.8%) cases. The patients were diagnosed and followed up in one hospital, and all of them were treated with rhGH. The control group consisted of 16 patients with SGA: 9 (56.3%) boys (SRS excluded). For all patients, birth parameters were obtained from patients’ medical records and were standardized according to gender and gestational age [23]. Relative macrocephaly at birth was defined as a head circumference at birth ≥ 1.5 standard deviation score (SDS) above birth length and/or weight [4], SGA was defined as a birth body weight ≤ –2 SDS [24], and premature birth was defined as birth before 37 weeks of pregnancy. Birth parameters of patients with SRS were compared with the SGA control group, and within the SRS group, according to molecular abnormality. Anthropometric measurements were performed in all SRS and SGA patients at the beginning and during 3 years of rhGH therapy at one-year intervals. Body composition was evaluated in 13 SRS patients, 9 boys and 4 girls (10 with 11p15 LOM and 3 with upd(7)mat), and in 14 SGA patients (7 boys). Only patients who had complete BIA measurements at baseline and after 3 years of rhGH treatment were analysed. The study was performed in accordance with the Helsinki Declaration and approved by the Bioethics Committee of The Children’s Memorial Health Institute. All the subjects (patients’ parents) gave informed consent to participate in the research.

Measurements

Height was measured using the SECA 264 stadiometer; the result was the mean of 3 independent measurements (in the case of body asymmetry, equalization was applied for the shorter limb). Weight was measured using a medical scale (Radwag WPT 100/200 O). Age- and sex-specific height, weight, weight for height, and body mass index (BMI) SDS were calculated using the Polish growth references [25]. Anthropometric parameters (SDS) at baseline and after 3 years of rhGH treatment were compared between the SRS study group and the SGA control group, and also within the SRS group (11p15 LOM vs. upd(7)mat).

Height velocity (Hv) (cm/year and SDS/year) was compared between the SRS group and the SGA control group and, additionally, in SRS subgroups according to molecular abnormality rhGH therapy program and gender.

Growth hormone therapy

Growth hormone therapy for children with SRS is available under 2 national programs for children with SGA, and with GHD. The inclusion criteria for both programs are described in detail in Appendix 1. The recommended doses of rhGH for children with SGA and GHD are 0.48–1.29 IU/kg/week (optimal 0.75 IU/kg/week) and 0.3–1.0 IU/kg/week, respectively [26]. The mean dose in 3 years of rhGH treatment was calculated. The median dose of rhGH in the entire SRS group was 0.63 IU/kg/week. The normal level of growth hormone in the SRS group was measured in 12 patients (10 with 11p15 LOM, 2 with upd(7)mat) who were treated under the program for children with SGA (median of mean dose 0.70 IU/kg/week). Reduced “bursts” of GH in 2 independent tests were indicated in 19 patients (13 with 11p15 LOM, 6 with upd(7)mat) who followed the GHD program. In the SGA control group, the median of the mean dose was 0.97 IU/kg/week. The mean age of all patients with SRS at the beginning of rhGH treatment was 6.6 ± 2.7 years, and in the SGA control group it was 7.6 ± 1.5 years. All SRS and SGA patients were in prepubertal stage at the beginning of the therapy.

Body composition

Body composition was measured at one-year intervals using a noninvasive and safe testing method: bioelectrical impedance (BIA). BIA was performed using a Jawon Medical Contact 357S analyser based on the tetra-polar electrode method with 8 touch electrodes and multi-frequency: 5, 50, 250, and 550 kHz. Lean body mass (LBM), skeletal muscle mass (SMM), total body water (TBW), intracellular body water (ICW), extracellular body water (ECW), and fat mass (FM) (in kg and %) was analysed. BIA was performed in the morning, on an empty stomach, or at least 3 hours after a meal. Inclusion criteria were as follows: ability to maintain a stable standing position for 30 seconds and a minimum bodyweight of 10 kg. For children, Jawon Medical software uses standard FM (%) depending on age range. References used in the software are the standard range of FM (%), which is 15–20% of standard bodyweight for men and 20–30% of standard bodyweight for women.

Molecular analysis

Blood samples were collected from the patients, and genomic DNA was extracted from peripheral blood leukocytes using standard procedures. Methylation-sensitive multiplex ligation-dependent probe amplification (MS-MLPA) with the use of a SALSA MLPA KIT ME030BWS/SRS (MRC-Holland, Amsterdam, Netherlands) was performed according to the instructions of the manufacturer. Raw data were analysed using the GeneMarker software v.1.8 (Soft Genetics LLC, State College, PA, USA). Microsatellite analysis was conducted using markers for chromosome 7: D7S507 (7p21), D7S460 (7p14), D7S663 (7q11), and D7S820 (7q21). Informed consent was obtained from the patients’ parents.

Statistical analysis

To describe the baseline characteristics, descriptive analyses were performed. The Shapiro-Wilk test was used to analyse data normality. The homogeneity of variance was checked using Levene’s test and the Brown-Forsythe test. Differences between the examined groups were analysed using Student’s t-test or the Mann-Whitney U test. The chi-square test was used to compare the number of patients in groups and the frequency of comparison characteristics. The analysis of variance (ANOVA, Kruskal-Wallis test) was performed to compare birth parameters between the 11p15 LOM, upd(7)mat, and SGA groups. Differences in height, weight, weight for height, BMI, and Hv during a period of 3 years of rhGH treatment were analysed using the ANOVA Friedman test. Changes in FM between the beginning and the third year of rhGH treatment were assessed using the Wilcoxon test. p-values < 0.05 were considered to be statistically significant. Statistical analysis was carried out using Statistica 13.3 software.

Results

Birth parameters: SRS vs. SGA

The birth parameters of the study group are presented in Table 1. Prematurely born patients were identified: 7 (22.6%) in the SRS group (including 6 children with body weight ≤ –2 SDS) and 7 (43.8%) in the SGA group. SRS patients presented RM at birth more often than patients with SGA: 23 (74.2%) vs. 5 (33.3%); p = 0.008. Gestational age was higher in the SRS group: 39 vs. 37 weeks of pregnancy, respectively (p = 0.031). Head circumference at birth was higher in children with SRS: –1.0 vs. –1.8 SDS, respectively (p <0.001). The difference between head and chest circumference at birth was higher in the SRS group: 5.0 vs. 3.0 cm, respectively (p = 0.018). Weight and length at birth were lower in SRS patients, but not statistically significantly (ns)

Birth parameters in the SRS group: 11p15 LOM vs. upd(7)mat

Within the SRS group, 23 (100%) patients with 11p15 LOM and 6 (75%) with upd(7)mat were born SGA. The number of prematurely born children was lower in the 11p15 LOM group than in the upd(7)mat group: 3 (13%) vs. 4 (50%), respectively, p = 0.053. The average birth weight SDS was significantly lower in patients with 11p15 LOM than in those with upd(7)mat: –4.2 ± 1.5 SDS vs. –2.7 ± 1.5SDS, respectively (p = 0.034). Birth length (SDS) was lower in the 11p15 LOM group: -1.9 ± 1.4 vs. –1.4 ± 1.2 SDS, respectively (ns). The median of head circumference was –1.0 SDS in both groups. RM at birth was revealed in 17 (73.9%) patients with 11p15 LOM and in 6 (75%) patients with upd(7)mat.

Netchine-Harbison clinical scoring system

Patients with SRS were scored using the Netchine-Harbison clinical scoring system (NH-CSS) proposed in the consensus [4]. Twenty-two patients with 11p15 LOM and 4 patients with upd(7)mat met at least 4 of the NH-CSS criteria, including both RM and protruding forehead (p = 0.01). Body asymmetry was identified significantly more often in the group with 11p15 LOM than in the group with upd(7)mat: (18 [78.3%] vs. 1 [12.5%]; p = 0.002). No statistically significant differences were found between the analysed groups with regard to the features: RM (17 [73.9%] vs. 6 [75%]), SGA (23 [100%] vs. 6 [75%]), postnatal growth failure (23 [100%] vs. 8 [100%]), feeding difficulties and/or low BMI (19 [82.6%] vs. 6 [75%]), and protruding forehead (22 [95.7%] vs. 6 [85.7%]).

Anthropometric parameters

The mean age at the beginning of rhGH treatment was 6.6 ± 2.7 years in the SRS group and 7.6 ± 1.5 years in the SGA group. Age for height, defined as the age at which a patient’s body height corresponds with the 50th percentile, was lower in the SRS group: 4.1 ± 2.2 vs. 5.2 ± 1.3 years, respectively (p = 0.041). At baseline of rhGH therapy, SRS patients had lower height, weight, and weight for height (SDS) than patients in the SGA control group: –3.3 ± 1.2 vs. –2.6 ± 06 SDS (p = 0.012), –2.5 vs. –1.9 SDS (p = 0.037), and –1.7 vs. –1.1 (p = 0.038), respectively. BMI (SDS) was also lower in the SRS group (non significant — ns). No such differences were identified between patients with 11p15 LOM and upd(7)mat. After 3 years of rhGH therapy, SRS patients had still lower height, weight, weight for height, and BMI (SDS) than patients from the SGA control group (ns) (Tab. 2).

All analysed parameters (SDS) significantly increased after a 3-year rhGH treatment, except BMI SDS in patients with upd(7)mat. Figure 1 shows the evolution of anthropometric parameters (SDS) in the entire SRS group, in the SRS subgroups: 11p15 LOM and upd(7)mat), and in the SGA control group.

Figure 1. A. Height standard deviation score (SDS); B. Weight SDS; C. Weight for height SDS; D. Body mas index (BMI) SDS; Changes during recombinant human growth hormone (rhGH) treatment in total Silver-Russell syndrome (SRS) group and according to molecular abnormality (loss of methylation at chromosome 11p15.5 [11p15 LOM] and maternal uniparental disomy of chromosome 7 [upd(7)mat]) and in the small for gestational age (SGA) control group. ANOVA Friedman test: ^ns (not significant); *p < 0.00001; **p < 0.001; ***p < 0.01; ****p < 0.5

Height velocity

The highest increase in Hv (cm/year and SDS/year) was observed in the first year of treatment in all examined groups. In the first year of rhGH treatment, patients with SRS had significantly higher Hv than patients with SGA: 8.8 ± 1.7 vs. 8.0 ± 0.9 cm/year (p = 0.041) and 0.7 ± 0.4 vs. 0.5 ± 0.2 SDS/year (p = 0.033), respectively. In the second and third year of rhGH therapy, these differences were not statistically significant. In the third year of therapy, Hv (SDS/year) was the lowest in both examined groups (Tab. 3). Hv in patients with SRS significantly decreased between the first and third year of rhGH treatment, from 8.8 ± 1.7 to 7.4 ± 1.5 cm/year (p = 0.03) and from 0.7 ± 0.4 to 0.3 ± 0.3 SDS/year (p < 0.0001), respectively. In patients in the SGA control group, Hv also decreased, but not statistically significantly (Fig. 2).

Figure 2. Height velocity (Hv; in cm/year) (A) and standard deviation score (SDS)/year (B) during recombinant human growth hormone (rhGH) therapy in patients with Silver-Russell syndrome (SRS) and small for gestational age (SGA)

Patients with upd(7)mat had higher Hv than patients with 11p15 LOM in all the analysed periods, but the difference was statistically significant only in the second year of rhGH treatment. Hv in patients treated under the GHD program was shown to be slightly higher than in patients treated under the SGA program, in all analysed periods. Taking into account the division by gender, Hv was similar in boys and girls. Patients with SRS received a significantly lower mean dose of rhGH than patients with SGA: 0.63 vs. 0.97 IU/kg/week, respectively (p < 0.001). In all patients with SRS, the mean dose was similar, regardless of the subgroup (Table 3).

Body composition

The mean age at baseline of rhGH treatment was 7.4 ± 3.0 and 7.6 ± 1.4 for patients with SRS and SGA, respectively. At baseline and after 3 years of rhGH treatment, no statistically significant differences were identified between the examined groups within the following parameters: weight, LBM, SMM, TBW, ICW, ECW, FM (kg), and FM (%) (Tab. 4). Individual parameters of body composition expressed in kilograms significantly increased during the 3 years of rhGH therapy in both groups, in relation to the increase in whole body weight (kg), but the percentage of FM decreased significantly in patients with SRS, from 4.2% to 3.0% (p = 0.033), and in patients with SGA, from 7.6% to 6.6% (p = 0.046).

Discussion

Silver-Russell syndrome is known to be associated mainly with epigenetic changes in 2 chromosomes: 11p15 LOM and upd(7)mat. In the remaining individuals, diagnosis is based on clinical manifestations and is therefore more probabilistic. The authors have chosen a confirmed epigenetic status as an inclusion criterion for this study.

In analyses of birth parameters, it was indicated that SDS of birth weight and length were lower in patients with SRS than in the SGA control group. Birth weight and length were also lower in the 11p15 group than in the upd(7)mat patients, which is consistent with research studies by other authors [11, 16, 27, 28]. Head circumference was close to the 3rd percentile for SGA children, but for SRS patients it was in the scope of the standard range. Within the SRS group, median of birth head circumference was the same in both groups, regardless of the type of molecular abnormality, which is similar to results obtained in another study [27]. In the presented study, we indicated that birth chest circumference was significantly lower than birth head circumference in the SRS group than in the SGA control group: 5 cm vs. 3 cm (p=0.018). This feature was not considered in other published studies, so the results cannot be compared, but it seems reasonable to suggest that this feature may be important in screening patients suspected of SRS syndrome.

Numerous studies have shown that treatment with rhGH has a positive effect on patients with intrauterine and postnatal growth retardation, including children with SRS [14–16, 27, 29–33]. However, in the majority of studies, SRS diagnosis was based on clinical assessment [14, 15, 31, 33]. Only a few reports on GH therapy include patients with identified (epi)mutations [16, 27, 34]. The most important factors influencing the final height are growth deficiency, time of treatment commencement, parental height, birth weight and length, bone age, the dose of rhGH, puberty, and the height at the beginning of and during rhGH treatment (earlier commencement and longer duration of therapy give better outcomes). Response to rhGH treatment proved to be best during the first year of the therapy, which was expressed as significantly higher Hv after the first year of rhGH therapy in all examined subgroups. Our result is similar to the result of the KIGS study [30], which suggested that a better height achievement depends particularly on the response to therapy during the first year of GH application. Unfortunately, the number of molecularly confirmed SRS patients in the KIGS study is unknown; all SRS patients were classified based on clinical criteria.

In our study, we indicated that during 3 years of rhGH therapy, height SDS decreased in both SRS patients and the SGA control group. After 3 years of rhGH therapy, patients with SGA had higher height SDS than patients with SRS, but attention must be drawn to the fact that patients in the SGA control group also had higher height SDS at baseline. Finally, the increase in height was similar in both described groups: 1.5 SDS/23.9 cm for SRS patients and 1.3 SDS/23.5 cm for the SGA control group during 3 years. It was observed that after 3 years of rhGH therapy, height SDS was almost the same in the SRS subgroups analysed with respect to molecular abnormality, gender, and therapy program. Although patients with upd(7)mat had lower height SDS at the beginning of rhGH therapy than patients with 11p15 LOM, patients from both groups achieved similar height SDS after 3 years of rhGH treatment. A similar observation was published by Smeets [35], who suggested that there was a trend towards a greater height gain in the groups with upd(7)mat and idiopathic SRS compared to patients with 11p15 LOM, which was also in line with the observation of Binder [36], who reported that during approximately 5 years of rhGH treatment, the epigenetic mutation was not a significant predictor of the height SDS. We observed that weight, weight for height, and BMI also increased during rhGH therapy in SRS patients, which corresponds with other studies [15, 35].

The group analysed in this study was heterogeneous regarding the concentration of GH: GHD was diagnosed in 19 patients. We observed that height increased by 0.7 SDS after the first year of rhGH treatment in both the SGA and the GHD rhGH program group. After 3 years of treatment, an increase in height SDS was slightly higher in the GHD group than in the SGA group (1.6 vs. 1.3 SDS/year), which corresponds to another outcome [37]. The better response to rhGH treatment of GHD patients probably results from the fact that this is a substitution therapy in cases of insufficient growth hormone excretion. Conversely, in the case of SRS patients with SGA but without GHD, growth is stimulated in individuals with an epigenetically disturbed growth process.

GH therapy can favourably influence body composition [38]. In the presented study, the authors used BIA to analyse body composition because this is a safe, reliable [coefficient of variation (CV) 3.5–4%], effective, relatively cheap, and non-invasive method for children, with a short time of examination. In the presented study it was noticed that during 3 years of rhGH treatment, all analysed parameters, such as LBM, SMM, TBW, ICW, ECW, and FM (kg), increased, which is related to increased whole-body mass. To estimate changes in body composition, the percentage of FM was calculated. It was indicated that FM% decreased during rhGH therapy in both SRS and SGA group, which is probably related to the catabolic effect of growth hormone on adipose tissue. Similar scores were published by Willemsen et al. [20]. In 2001, a study indicating that patients with SRS had less body fat (%) than patients from a intrauterine growth retardation (IUGR) group (11.68 ± 7.67 vs. 18.21 ± 8.33%, respectively) was published [39]. Its authors also used the BIA method, but patients with SRS were classified based on clinical criteria, which is why it may be a heterogeneous group. Smeets et al. [35] analysed body composition using DXA and noticed that during GH treatment, LBM seemed to be stable whereas body fat (%) increased in both SRS and non-SRS children.

Recently, outcomes of fat percentage in older patients with SRS, mainly adults, were published. In Lokulo-Sodipe’s [28] and Patti’s [40] studies the age range was from 13.32 to 67.71 years and from 18 to 46 years, respectively. Both studies indicated a trend that the percentage of FM in elder SRS patients increased during ontogenesis, and the fat percentage exceeded standard levels. To our knowledge, no other studies evaluated body composition exclusively in SRS children. In our study, related to younger children, we observed a decrease of FM (%) during 3 years of rhGH therapy. However, regarding reported outcomes in older patients, we see the need for continuous monitoring of fat percentage during all stages of development in SRS patients.

There are several limitations to this study. First, the study presented a relatively short time of observation, which follows from the fact that patients with SRS without GHD gained the opportunity of GH treatment in 2015, and it takes several months to classify patients for therapy. Second, our group was younger than others described in the literature. Third, BIA was performed in 13 patients, but we decided to include only complete data (at baseline and after 3 years of rhGH treatment). On the other hand, we described a relatively large group of SRS patients, from one country, and treated rhGH in a single hospital according to 2 national programs. Because our group is relatively young, we have the opportunity for long-term observation and enlargement of our group in the future.

Conclusion

The authors indicated a positive effect of rhGH therapy for growth of SRS patients. It was demonstrated that height, weight, and BMI (SDS) increased and FM (%) decreased significantly. However, longer-term observation and a larger study group are necessary to draw stronger conclusions, especially on body composition.