Vitamin D in children with growth hormone deficiency due to pituitary stalk interruption syndrome

Growth hormone (GH) deficiency (GHD) can be congenital or acquired. Pituitary stalk interruption syndrome (PSIS), a sign of congenital and permanent GHD, is an anomaly of the pituitary gland characterized by a combination of specific findings on magnetic resonance imaging, including an interrupted pituitary stalk, an absent or ectopic posterior pituitary and anterior pituitary hypoplasia [1]. Because of the heterogeneity in clinical, biological and imaging presentations, recent efforts have focused on discriminating patients with multiple hypothalamic-pituitary (HP) deficiencies from those with isolated GHD [2]. Previous studies have shown an association between PSIS and other syndromes and/or malformations, and occurrence of familial forms have led to identification of gene mutations [3‐7].

Humans mainly acquire vitamin D from exposure to sunlight, but it can also be obtained from food (oily fish). Vitamin D is first metabolized to 25-hydroxyvitamin D (25OHD) in the liver and then to its active form, 1,25-dihydroxyvitamin D (1,25(OH)₂D), in the kidneys. Overall, the concentration of serum 25OHD reflects the vitamin D status of an individual. Although debate remains regarding the 25OHD level that defines the optimal level, vitamin D deficiency is considered to be common worldwide in both adults and children [8‐11]. Vitamin D has a well-known role in bone development in utero and in childhood, and it maintains bone health in adults through effective calcium regulation and bone mineralization [12]. Because vitamin D receptors are present in many tissues, vitamin D may also have “non-classical” effects. Indeed, vitamin D deficiency has been associated with obesity in children [13], cardiovascular disease, type 2 diabetes, cancers (colon, prostate or breast) and auto-immune disorders [14‐16].

Recent studies have shown a relationship between vitamin D status and GH and insulin-like growth factor 1 (IGF1). For example, initiation of cholecalciferol treatment increases serum IGF1 levels in children and adults with vitamin D deficiency [17, 18], and Wei et al. and Saggese et al. showed that six and 12 months of GH treatment, respectively, raised 1,25(OH)₂D concentrations in children with GHD [19, 20].

The first objective of the present study was to assess vitamin D status at the initial evaluation before GH treatment in a large series of children with GHD due to PSIS attended by a single endocrinologist. The second objective was to investigate the relationship between 25OHD and 1,25 (OH) ₂D serum levels and patient characteristics, mainly the GH peak and serum IGF1 level.

The mean serum 25OHD concentration was 33.2 ± 18.0 ng/mL. Two (4%), nine (18%), 15 (30%) and 24 (48%) patients had a 25OHD level ≤ 12 ng/mL, between 13 and 19 ng/mL, between 20 and 30 ng/mL, and above 30 ng/mL, respectively (Table 2). Among these, a concentration > 60 ng/mL was found for four patients. Sex and season were the two factors that significantly influenced 25OHD concentrations, with higher levels in boys than in girls and lower levels in the “cold” than in the “sunny” season (Tables 2 and 3). Conversely, no significant difference in 25OHD level was observed according to the isolated or multiple HP character deficiency, age, bone age, height, BMI, growth rate, GH peak level or IGF1 level.

The mean serum 1,25(OH)₂D concentration was 74.5 ± 40.7 ng/L (Table 3). No significant difference in 1,25(OH)₂D level was observed according to sex, isolated or multiple HP character deficiency, season, age, bone age, height, BMI and growth rate. However, we did observe a trend toward higher levels in those with multiple HP deficiency. No significant correlations were observed between serum 1,25 (OH)₂D concentration and age, bone age, height, BMI, growth rate, IGF1 level or 25 OHD level. Nonetheless, significant positive correlations were observed between the GH peak level and serum 1,25 (OH)₂D concentration (Rho = 0.35; p = 0.015) and the 1,25(OH)₂D/25OHD ratio (Rho = 0.29; p < 0.05).

In the present study, we evaluated vitamin D status in a large series of children with GHD due to PSIS, as observed by a single endocrinologist. We performed this study because both GHD [26] and vitamin D deficiency [27] are associated with bone fragility, impaired growth, muscle weakness, cardiac outcomes and metabolic syndrome.

In our population, the mean 25OHD concentration was 33.2 ± 18.0 ng/mL, which appears to be a satisfactory level. As we did not include a control group, we compared this level with published data for healthy French children/adolescents. Among 212 girls aged 11–16.9 years with Tanner stages of 4 and 5 and living in Normandy (northwest of France), which is at the same latitude as Paris, mean 25OHD concentrations were 20.0 ± 6.1 ng/mL at the end of summer and 15.9 ± 7.4 ng/mL between January and May [28]. Alarmingly, 41% of these otherwise healthy girls had a 25OHD < 12 ng/mL during winter, revealing a severe vitamin D deficiency. The European HELENA study evaluated the vitamin D status of 1006 adolescents with a mean age of 14.9 years living in 10 European cities [29]. The participating French city was Lille, which is located in the extreme north of France, and the mean 25OHD level was 20.4 ± 8.8 ng/mL in girls and 24.1 ± 10.8 ng/mL in boys. In a recent multicenter study of vitamin D status in 326 healthy French children aged 6–10 years, it was found that 3.1% of the participants had a 25OHD concentration below 10 ng/mL and that 34.4% had a 25OHD concentration between 10 and 19.9 ng/mL [11]. Although the mean 25OHD values were not indicated in this article, it can be concluded that the vitamin D status of these children was quite similar to what we found in our children with PSIS. In our present study, female sex and winter season (from October to May) were significantly associated with lower serum 25OHD concentrations, which is consistent with other previous published studies in adult and pediatric healthy populations [9, 30].

We then compared the 25OHD concentration measured in the present study in our patients with PSIS to the values reported in recent studies on GHD patients from various countries prior to GH treatment. Saggese et al. evaluated 26 children with GHD [20] and found normal 25OHD concentrations but low 1,25 (OH)₂D levels before GH treatment and a significant increase in 1,25 (OH)₂D levels after 12 months of GH treatment. Ciresi et al. evaluated 80 Sicilian children with GHD [31]. These authors reported significantly higher values in the sunny season (31.1 ± 11.1 ng/mL in June–September) than in the cold season (17.3 ± 5.3 ng/mL in November–February), with 35% having vitamin D insufficiency and 40% vitamin D deficiency. Witkowska-Sedek et al. evaluated 84 children and adolescents with GHD and found low 25OHD concentrations (22.3 ± 6.9 ng/mL) [32], and Savanelli et al. evaluated 41 adult GHD patients and found mean 25OHD concentration of 21.3 ± 12.3 ng/mL, with vitamin D deficiency present in 51% of the patients compared to 14.6% of the controls [33]. In addition, Ameri et al. evaluated 69 adult GHD patients, most with postsurgical hypopituitarism [34], and found that only 6 patients (8.7%) had a serum 25OHD concentration above 30 ng/mL. These authors reported a positive association between vitamin D status and IGF1, and they also found a trend toward higher GH doses in patients with low vitamin D status (25OHD < 15 ng/mL), suggesting that a better vitamin D status may facilitate achieving normal IGF1 values in GHD patients. In a literature review, these authors proposed that assessment of vitamin D status might be an appropriate method of determining the dose of recombinant hGH for treatment of adult patients with GHD [34].

One possible explanation for the higher vitamin D status in our study may be related to the young age of the subjects (mean 5.4 years). Indeed, in France, the French Society of Paediatrics recommends systematic vitamin D supplementation of children up to five years of age. Unfortunately, no guidelines for older children have been implemented; an exception is adolescents, in whom winter supplementation is recommended but rarely performed [35]. Surprisingly, our patients of more than 5 years old, particularly those aged 6–9 years, had a mean 25OHD level similar to that of our younger patients. This may suggest that vitamin D supplementation had been offered to some of our older patients before evaluation in our unit. In support of this is the fact that although BMI is known to influence spontaneous vitamin D status with lower 25OHD concentrations in overweight subjects, we found no difference in 25OHD levels between patients with a BMI < 0 SDS and those with a higher BMI. If the hypothesis that some of our patients received vitamin D supplementation before their assessment is correct, the finding that they have a normal vitamin D status is likely independent of their GHD. However as no information on vitamin D supplementation before our evaluation was available in the medical charts of our patients, we were unable to confirm that their rather good vitamin D status was due to vitamin D supplementation.

As GH appears to increase the level of 1,25 (OH)₂D [19, 20, 36, 37], albeit likely indirectly, through the effects of IGF1, it may be hypothesized that patients with GHD due to PSIS have a low 1,25 (OH) ₂D serum concentration. This was not the case in our patients who had normal levels. We, however, have no explanation for the higher 1,25(OH)2D level in those with multiple HP deficiencies. Regardless, we did find a significant positive correlation between the GH peak level obtained during a pharmacologic stimulation test and the 1,25 (OH) ₂D concentration as well as with the 1,25 (OH) ₂D/25OHD ratio. These findings suggest that even in these patients who had severely impaired GH secretion and low IGF1 levels, an interplay between the GH/IGF1 axis and the vitamin D system still exists.

Our study has limitations: it was retrospective and did not include a control group. However, the strength of our study pertains to the large series of GHD patients with a homogenous diagnosis (i.e., PSIS) and the same protocol of exploration for years, as they were attended by the same endocrinologist. This design allowed avoiding the considerable selection biases that affect several studies on GH deficiency. The patients who were excluded because of unavailable plasma samples might have introduced a bias, but their similarities to the included patients limit this. Another limitation is the lack of information on vitamin D supplementation before the patients were evaluated.

We found quite an acceptable mean vitamin D status in our series of children with GHD due to PSIS. There are currently no available data from a double-blind, placebo-controlled randomized trial of vitamin D supplementation in GH-treated GHD patients showing that vitamin D optimizes bone health, growth and possibly other outcomes. Despite this lack of high-level evidence, we believe that it is mandatory to maintain optimal vitamin D supplementation in GHD patients.