Effectiveness of the Combined Recombinant Human Growth Hormone and Gonadotropin-Releasing Hormone Analog Therapy in Pubertal Patients with Short Stature due to SHOX Deficiency

Abstract

Context: Isolated heterozygous SHOX defects are the most frequent monogenic cause of short stature, and combined therapy with recombinant human GH (rhGH) and GnRH analog (GnRHa) in pubertal patients has been suggested, but there are no data on final height.

Objective: The aim of the study was to analyze adult height after rhGH and GnRHa therapy in patients with SHOX haploinsufficiency.

Patients: Ten peripubertal patients with isolated SHOX defects participated in the study.

Intervention: Five patients were followed without treatment, and five were treated with rhGH (50 μg/kg/d) and depot leuprolide acetate (3.75 mg/month).

Main Outcome Measures: Adult height sd score (SDS) was measured.

Results: All patients followed without treatment had marked downward growth shift during puberty (height SDS, −1.2 ± 0.7 at 11.4 ± 1.4 yr; adult height SDS, −2.5 ± 0.5). Conversely, four of five patients treated with rhGH for 2 to 4.9 yr associated to GnRHa for 1.4 to 5.8 yr improved their height SDS from −2.3 ± 1.3 at 11.8 ± 2.1 yr to a final height SDS of −1.7 ± 1.6. The difference between the mean height SDS at the first evaluation and final height SDS was statistically significant in nontreated vs. treated patients (mean height SDS change, −1.2 ± 0.4 vs. 0.6 ± 0.4, respectively; P <0.001).

Conclusion: A gain in adult height of patients with isolated SHOX defects treated with combined rhGH and GnRHa therapy was demonstrated for the first time, supporting this treatment for children with SHOX defects who have just started puberty to avoid the loss of growth potential observed in these patients during puberty.

A gain in adult height of SHOX patients treated with rhGH and GnRH analog was demonstrated, supporting this treatment for SHOX children to avoid the loss of growth potential during puberty.

Among the recognized causes of short stature, isolated SHOX gene (short stature homeobox-containing gene, NM_000451) defects are the most frequent monogenic cause of short stature. SHOX gene is located in the pseudoautosomal region of the X and Y chromosomes (1) and encodes a cell-specific homeodomain protein. Isolated SHOX haploinsufficiency is observed in 56–100% of patients with Leri-Weill dyschondrosteosis (OMIM 127300) (2,3,4,5) and in 1–14% of children with short stature without any apparent skeletal dysplasia, usually classified as idiopathic short stature (OMIM 300582) (1,5,6,7,8).

The growth deficit due to SHOX haploinsufficiency is estimated to be −2 sd score (SDS) of height (2). Longitudinal studies of children with SHOX defects suggest a relatively well-preserved prepubertal growth rate followed by compromised pubertal growth spurt due to premature growth plate fusion (9,10).

Although the exact physiopathological effect of SHOX gene haploinsufficiency on short stature remains to be determined, few studies have been reported with optimistic responses to treatment with recombinant human GH (rhGH) alone (6,11,12) or with the concomitant use of GnRH analog (GnRHa) (13). A recent study demonstrated a gain in final height of 14 patients with SHOX haploinsufficiency treated with rhGH, comparable with height gain of Turner syndrome patients similarly treated. However, to date, there is no information about long-term treatment and final height of patients treated with combined rhGH and GnRHa therapy. The aim of the present study was to evaluate the adult height of patients with SHOX haploinsufficiency treated with rhGH and GnRHa therapy in comparison with untreated ones.

Patients and Methods

Patients

This study was approved by the local Ethics Committee, and patients or patients’ tutors gave written informed consent. Ten patients (six girls) (Supplemental Tables 1 and 2, published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org) with isolated heterozygous SHOX defects were evaluated at peripubertal ages (11.6 ± 1.7 yr). Molecular defects observed in these patients were previously described: seven patients have complete SHOX deletion (5), one has a cryptic intragenic SHOX deletion involving exons IV–VIa (14), and two have a nonsense mutation (p.Y35X) (5). Patients 1 and 2 are brothers, patients 3 and 10 are siblings, and patients 5 and 6 are first cousins. Five patients were followed without any specific intervention (patients 1–5, Supplemental Table 1), whereas five patients were treated with rhGH/GnRHa (patients 6–10, Supplemental Table 2).

Study protocol

All treated children were evaluated at baseline and every 3–4 months during treatment. Evaluations were performed in the morning and included measurement of standing height, sitting height, weight, and pubertal status. Anthropometric data were converted to SDS using age- and gender-specific norms (15,16). Disproportional short stature was defined as the sitting height:height ratio more than 2 sd values above the mean for age and sex (16). Body mass index was calculated (weight/height²) and expressed as SDS. Bone age radiographs were assessed by two observers based on the method of Greulich and Pyle (17). Predicted adult height was determined by the Bayley and Pinneau method (18). The presence of Madelung deformity was assessed clinically and through radiological imaging. Patients were considered to be at final height if they were growing less than 0.5 cm/yr after an observation period of at least 12 months.

Treatment with rhGH and GnRHa

GnRHa therapy (3.75 mg depot leuprolide acetate sc every 28 d) was initiated just after pubertal onset. In two girls (patients 6 and 8), rhGH was started concomitantly, and in the two other girls (patients 7 and 10), rhGH treatment was started 1.0 and 2.3 yr after GnRHa therapy was initiated. The boy (patient 9) started GnRHa after 0.7 yr of isolated rhGH treatment. Adequate pubertal suppression during treatment with GnRHa was confirmed by clinical findings, as well as by basal and 2 h after depot leuprolide LH levels (19). rhGH was administered sc in a dose of 50 μg/kg/d (0.15 U/kg/d), and the dose was adjusted according to changes in weight every visit. All patients continued rhGH therapy after discontinuation of GnRHa until growth velocity (GV) was less than 2 cm/yr.

Statistical analysis

Differences between groups were tested by t test or Kruskal-Wallis and χ² or Fisher exact test, as appropriate. Statistical analyses were performed using the SIGMAstat statistical software package (Windows version 2.03; SPSS Inc., San Rafael, CA).

Results

Patients without treatment

These five patients (Supplemental Table 1) were clinically evaluated first at peripubertal ages, and all were reevaluated at adult ages. All were not treated because their molecular diagnosis was established at the end of growth (patients 1, 2, and 3) or because of their parents’ decision (patients 4 and 5). These patients had normal height SDS and/or normal height SDS prediction when they started puberty. However, all patients without treatment had a marked downward growth shift during puberty (difference between height SDS in the first evaluation and adult height SDS was −1.2 ± 0.4, ranging from −0.8 to −1.8) (Table 1 and Fig. 1). Their adult height SDS (−2.5 ± 0.5) was similar to the adult height of their affected parent (−2.9 ± 1.0). Patients 1 and 2 developed mild Madelung deformity during puberty. Patients 4 and 5, who had sitting height measured before start of puberty and at final height, did not present significant changes in body disproportion (sitting height:height ratio SDS) during follow-up.

Table 1.

Comparison of patients with isolated SHOX defects treated with combined rhGH and GnRHa therapy and without treatment

	Patients without treatment	Patients treated with rhGH and GnRHa
Height SDS of affected parent	−2.9 ± 1.0	−3.4 ± 1.1
Height SDS of nonaffected parent	−0.6 ± 1.0	−0.4 ± 2.0
Baseline
Age (yr)	11.4 ± 1.4	11.8 ± 2.1
Height SDS	−1.2 ± 0.7	−2.3 ± 1.3
Predicted adult height SDS	−1.4 ± 1.2	−3.0 ± 1.3
Last visit
Duration of GnRHa treatment (yr)		2.8 ± 1.8
Duration of rhGH treatment (yr)		3.8 ± 1.1
Age (yr)	18.1 ± 3.7	16.2 ± 2.3
Height SDS	−2.4 ± 0.6a	−1.7 ± 1.7
Total height SDS change	−1.2 ± 0.4	0.6 ± 0.4b
Sitting height/height SDS for age	3.7 ± 1.2	3.1 ± 0.7

^aP = 0.021 between baseline and last visit height SDS of patients followed without treatment. 

^bP < 0.001 between patients with and without treatment, P < 0.001. 

Figure 1.

Individual height SDS change from the first to the last evaluation in patients with isolated SHOX defects. Male patients were shown as triangles (▴ or ▵), and females were shown as circles (• or ○). Patients followed without any specific treatment are shown as open symbols and dashed lines, whereas patients treated with combined therapy (rhGH and GnRHa) are shown as filled symbols and solid lines.

Patients treated with rhGH and GnRHa

At the start of rhGH therapy, patients in this group (Supplemental Table 2) had more severe growth impairment and/or height SDS prediction than patients followed without treatment. All patients, except patient 6, had improved GV in the first year of rhGH therapy (50 μg/kg/d): basal GV, 5.1 ± 1.7 cm/yr (range, 2.6 to 6.6 cm/yr); and GV during the first year of therapy, 7.0 ± 2.5 cm/yr (range, 3.7 to 9.9 cm/yr).

After combined therapy with rhGH (duration from 2 to 4.9 yr) and GnRHa (ranging from 1.4 to 5.8 yr), four patients presented a marked improvement in height SDS compared with basal height SDS (mean height SDS increase, 0.6 ± 0.4; range, 0.0 to 1.1). All patients had normal and complete pubertal development after the interruption of GnRHa and reached a height above that predicted at the start of therapy. In comparison with patients without treatment (Fig. 1 and Table 1), treated patients grew significantly more during follow-up, resulting in similar final height, despite the fact that treated patients were shorter when they started puberty. Body disproportion was observed in all patients at the start of therapy and did not improve during treatment. Two patients had Madelung deformity at the start of therapy that did not deteriorate during follow-up, and the others did not develop Madelung deformity during treatment (Supplemental Fig. 1).

Discussion

The molecular diagnosis of SHOX defects in children with idiopathic short stature or Leri-Weill dyschondrosteosis has therapeutic implications. Short stature observed in patients with Turner syndrome is partially explained by the haploinsufficiency of SHOX gene (10). Because rhGH therapy in Turner syndrome patients improves GV and, consequently, final height (20), treatment of short stature due to isolated SHOX defects is proposed. One study demonstrated that prepubertal children with isolated SHOX defects treated with rhGH during 2 yr presented similar growth response to that of Turner syndrome patients (11). Recently, the same authors analyzed a subgroup of patients (n = 14; 79% prepubertal at start of therapy) from the original cohort who started rhGH treatment at the age of 10 ± 3 yr and had reached their final height with a height SDS gain of 1.1 ± 0.7 after 4.7 yr of rhGH therapy (21). These findings support the indication of rhGH therapy for prepubertal children with isolated SHOX defects. However, patients who initiate pubertal growth spurt before the beginning of rhGH treatment can have a rapid progression of bone maturation with loss of height potential.

Longitudinal follow-up studies of children carrying SHOX defects suggest a relatively well-preserved prepubertal growth followed by compromised pubertal growth due to premature growth plate fusion, thus causing premature growth arrest (10,22). A downward growth shift during puberty was also observed in our patients with SHOX mutations followed without treatment. This raises concerns regarding the adult height prognosis of these “normal” height children with SHOX defects.

The association of GnRHa with rhGH therapy allows longer treatment duration and could prevent or attenuate Madelung deformity. Ogata et al. (13) studied the only two patients described in the literature treated with combined therapy, but GnRHa was added for 1.1 and 3 yr after a period of isolated rhGH treatment ranging from 2.6 to 6 yr using half of the dose that our patients received (24 μg/kg/d). One patient improved her height SDS from −3.3 to −2.4, whereas the other maintained height SDS throughout treatment. Therefore, the role of combined therapy on height gain was not clear, and adult height of SHOX patients treated with this strategy was not available.

In our study, four of five patients treated with combined rhGH and GnRHa therapy presented an increase in height SDS, whereas all patients followed without treatment had a decrease in height SDS (Fig. 1). Interestingly, two siblings followed in our cohort illustrate the effect of the intervention in patients with the same genetic background: patient 3 without treatment had a marked reduction in height during puberty, whereas his sister, patient 10, treated with rhGH and GnRHa presented an improvement in height SDS, reaching an adult height 0.9 sd greater than her brother. Our study did not rule out that rhGH therapy alone can improve the final height in pubertal patients with isolated SHOX defects. However, it is expected that rhGH therapy alone would be less effective because the short stature in nontreated SHOX adults reflects the loss of growth potential during puberty due to a short pubertal spurt.

Isolated SHOX defect causes not only short stature but also body disproportion and Madelung deformity, which is more severe in females and commonly becomes more evident during puberty (5,10). Interestingly, Turner syndrome patients who also present SHOX haploinsufficiency have significantly lower frequency of disproportionate short stature and Madelung deformity when compared with patients with isolated SHOX defects (23), probably due to hypogonadism present in Turner syndrome. Therefore, it was postulated that the use of GnRHa could prevent these features in children with isolated SHOX defects. However, in our cohort, no significant changes in body proportion, shown as sitting height:height SDS, were observed in patients treated with GnRHa plus rhGH.

In conclusion, we demonstrated for the first time an adult height gain in patients with isolated SHOX defects treated with combined rhGH and GnRHa therapy. Despite the limitations of this preliminary study (retrospective and limited number of patients), our results support this treatment in children with SHOX defects who have just started puberty to avoid the loss of growth potential observed in these patients during puberty.