Comparison of clinical characteristics of a pediatric cohort with combined pituitary hormone deficiency caused by mutation of the PROP1 gene or of other origins

The statistical software STATISTICA 10.0 PL was used for the statistical analysis of results. Data were described using arithmetic mean and standard deviation. In order to compare the two data sets, the Student’s t-test for unrelated samples was used, and in the absence of a normal distribution of the data, the Mann–Whitney U test was implemented. To compare categorical values χ² test was used. A two-tailed p-value < 0.05 was considered statistically significant.

The most frequent mutation was the deletion c.301_302del (p.Leu102fs), which was homozygous in 23 patients (genotype c.301_302del/c.301_302del). The second biggest group of patients (14 children) revealed complex heterozygosity with mutations c.301_302del and c.150_151del (p.Gly52fs). In four patients, a homozygous mutation, c.150_151del/c.150_151del, was confirmed. In addition, two less common mutations, c.295C > T (p.Arg99Ter) and c.349 T > A (p.Phe117Ile), were detected in two complex heterozygotes carrying also the c.301_302del mutation. In 25 parents (13 mothers and 12 fathers) of CPHD-PROP1 children, a single mutant PROP1 allele was found. In 10 cases, both parents were carriers, and in 13 cases, the presence of the mutation was confirmed in only one parent (Table 1).

The clinical characteristics of CPHD-PROP1 and CPHD-nonPROP1 are presented in Table 2.

Pregnancy and delivery were uncomplicated in all patients with CPHD-PROP1, with the exception of one premature delivery in the 36th week of gestation. In contrast, in the CPHD-nonPROP1 group, there were 14 cases of abnormal position of the fetus (buttock, foot, or transverse) (Table 3). In two cases, the disorder was related to twin delivery. Additionally, there was one case of maternal gestosis and three cases of premature births, including one in the 30th week of pregnancy. Most of the deliveries were via the birth canal and in eight cases by cesarean section (four in each of the two groups). Despite the birth weight of 95% of the children being normal, ranging from − 2.0 to 2.0 SDS, patients with CPHD-PROP1 had a significantly higher median birth weight compared to CPHD-nonPROP1 (0.21 vs. − 0.29 SDS, p = 0.019). Two newborns with a birth weight above 2 SDS were found in the CPHD-PROP1 group, and two with a birth weight less than − 2 SDS were identified in the CPHD-nonPROP1 group. Hypoglycemia was diagnosed in five CPHD-nonPROP1 patients in the neonatal period: this resulted in further examinations, which led to the diagnosis of hormonal deficits. Additionally, retrospective analysis of the medical records of patients who were diagnosed later in life due to short stature revealed that in a further seven patients (three cases in the CPHD-PROP1 and four cases in the CPHD-nonPROP1 group), transient hypoglycemia was also present in the neonatal period, but no further diagnostic steps were taken during this period.

In all CPHD-PROP1 patients, differential diagnosis was initiated due to short stature. In five patients of the CPHD-nonPROP1 group, diagnosis was made due to hypoglycemia in the neonatal period (before growth failure occurred). With the exception of these five cases, diagnostic procedure was also started due to short stature. No statistically significant difference was observed between CPHD-PROP1 and CPHD-nonPROP1 patients in regard to their mean height at the moment of diagnosis (− 4.1 SDS and − 4.3 SDS, respectively, p = 0.64). The average growth velocity within 3 years preceding growth hormone therapy was lower in the CPHD-PROP1 patients than in the CPHD-nonPROP1 group (− 2.7 vs. − 0.8 SDS, p < 0.001).

All patients in both groups were diagnosed with secondary hypothyroidism. The two groups did not differ in terms of total and free thyroxine concentrations (CPHD-PROP1 and CPHD-nonPROP1 patients: mean 5.5 ± 1.6 vs. 4.7 ± 1.9 µg/dL; p = 0.47 and mean 8 ± 2.3 vs. 7 ± 2.3 pmol/L; p = 0.9, respectively). However, basic TSH concentrations were significantly lower in the CPHD-PROP1 group (mean 1.8 ± 0.8 vs. 2.4 ± 1.6 µIU/mL, p < 0.001), as well as the mean maximal TSH concentration in the TRH test (mean 7.6 ± 14.7 vs. 3.96 ± 11.4 µIU/mL, p = 0.006). The mean age at diagnosis of secondary hypothyroidism in CPHD-PROP1 patients was significantly lower compared to the CPHD-nonPROP1 group (5.2 vs. 8.0 years, p = 0.006).

Symptoms typical of hypogonadism at birth (non-descended testes and/or micropenis) were less frequent in CPHD-PROP1 boys (n = 6/30% vs. n = 12/54%; p = 0.0002). Spontaneous maturation did not occur in any patient. The GnRH test confirmed hypogonadotrophic hypogonadism. At the age typical for puberty, testosterone and estradiol concentrations in both groups were low and corresponded to the values of these hormones in prepubescent patients. The mean basic gonadotropin concentrations in CPHD-PROP 1 and CPHD-nonPROP1 patients were low (mean LH 0.4 ± 0.36 vs. 0.8 ± 0.3 µIU/mL, p = 0.6) and FSH (mean 0.5 ± 0.5 vs. 1.5 ± 0.8 µIU/m L, p = 0.036). The mean LH and FSH concentrations after GnRH stimulation were lower in CPHD-PROP1 patients (mean LH 0.6 ± 0.7 vs. 3.9 ± 1.0 µIU/mL, p = 0.25; mean FSH 0.7 ± 0.3 vs. 2.7 ± 4.0 µIU/mL, p = 0.014).

Concentrations of PRL were significantly lower in CPHD-PROP1 patients (mean 128 ± SD 129 vs. 416.3 ± 439 µIU/mL, p < 0.001). In the CPHD-PROP1 group, prolactin concentration was decreased in 18 of 20 patients (data are not available for the rest of the group). In the CPHD-nonPROP1 group, a decreased prolactin concentration was confirmed in seven of 12 patients (data are not available for the rest of the group due to the retrospective character of the study). The mean age of the diagnosis of prolactin deficiency was lower in the CPHD-PROP1 group compared to the CPHD-nonPROP1 (5.0 vs. 8.0 years, p = 0.006).

In the CPHD-PROP1 group, in those patients who underwent MRI of the pituitary gland, the examination at the time of diagnosis revealed a small pituitary gland in 21 cases (including one case of pituitary stalk interruption and posterior lobe ectopy) and pituitary gland enlargement in eight cases. In nine cases, the result of the examination was normal.

In the CPHD-nonPROP1 group, MRI revealed anterior lobe hypoplasia in 16 patients, pituitary stalk interruption and ectopy of the posterior lobe in 10 patients, anterior lobe hypoplasia and pituitary stalk interruption without posterior lobe ectopy in five patients, enlargement and heterogeneous signal of the pituitary gland in one case, and normal pituitary gland in two patients. In 15 patients of the CPHD-PROP1 group, MRI was repeated after 10 years. In 11 cases, the size of the pituitary gland had decreased, in one case it remained unchanged, and in three cases it increased. The CT scan was the only imaging examination of the pituitary gland in 18 patients, including four CPHD-PROP1 individuals. It suggested pituitary hypoplasia in 14 CPHD-nonPROP1 and in two CPHD-PROP1 patients. Moreover, pituitary microadenoma (4 mm in diameter) and pituitary nodular enlargement with the presence of microcysts and calcifications were described in two CPHD-PROP1 patients aged 8.8 and 18.4 years, respectively.

The estimated global prevalence of CPHD is 1 in 8000 (https://www.ghr.nlm.nih.gov). It may be a disorder of known genetic or of idiopathic etiology. To date, 30 genes have been reported to be involved in the pathogenesis of CPHD [10]. In this study, PROP1 gene variants were found in 43/74 patients (58%) and were, as previously stated by other authors, the most frequent genetic etiology of CPHD [11]. A much higher percentage was found in a Lithuanian cohort (70%) [21]. In comparison to this, in Japan, non-PROP1 mutation was not confirmed in any of the studied patients [22]. PROP1 mutation prevalence is higher in familial rather than in sporadic cases, as was also found in the current study (23/23 100% in familial vs. 20/51 39% in sporadic) [23‐26]. In the study group, we observed significant male predominance in the CPHD-nonPROP1 group. This is consistent with observations made by other authors describing similar cohorts [27]. The origin of this phenomenon remains unclear; however, it has been hypothesized to be due to a role of X-linked recessive genes, sex chromosome-environmental interaction, or an unexplained male susceptibility to perinatal insult [28]. The detailed assessment of the study group in the present study revealed some significant differences between CPHD-PROP1 and CPHD-nonPROP1 patients. The birth weight of most children with CPHD was normal, although patients with CPHD-PROP1 had a higher mean birth weight compared to the CPHD-nonPROP1 group. This is consistent with the observation of other authors [26, 28, 29]. In the present study, among patients with CPHD-PROP1, no cases of abnormal positioning of the fetus at delivery were found. On the contrary, in the CPHD-nonPROP1 group, there were as many as 14 of 37 patients (38%) with abnormal position of the fetus at delivery. The latter was also demonstrated by Diwaker et al., who found that breech presentation was higher in PSIS-CPHD than in POU1F1/PROP1-CPHD (44.4 vs. 5.5%) and concluded that breech presentation in PSIS is likely due to pituitary stalk interruption rather than to hormonal deficiency [27]. This confirms the hypothesis that in the early stage of intrauterine development, factors other than the PROP1 mutations, which result in an alteration of the development of the central nervous system, including the pituitary gland, have a stronger impact on fetal growth and the course of labor [30]. As was previously demonstrated in CPHD-PROP1, TSH and GH deficiencies have a tendency to occur in early childhood, whereas gonadotropin and corticotropin deficiencies manifest later in life [11, 31]. In five patients of the CPHD-nonPROP1 group, diagnosis was made due to hypoglycemia in the neonatal period (before growth failure had occurred) probably because of concomitant ACTH deficiency. In CPHD-PROP1 mutation patients, since ACTH deficiency occurs variably as the patient grows older, symptoms of adrenal failure did not occur in the neonatal period. Another interesting finding may be the fact that severe GH deficiency was found in both groups (in the CPHD-PROP1 group compared to the CPHD-nonPROP1 1.2 vs. 1.08 ng/mL, p = 0.003). The reason for this phenomenon is unknown, although the mechanism of gradual extinction of the activity of the anterior pituitary cells caused by the PROP1 gene mutation may be of importance. Arroyo et al. presented a patient with CPHD-PROP1 with normal height [32]. As a probable cause of this phenomenon, the authors tentatively proposed the presence of GH, although it occurs at low levels in the circulation during childhood and adolescence, the lack of circulating estrogen delaying epiphyseal fusion, resulting in growth beyond the period of normal growth [32]. Despite the absence of a statistically significant difference in mean height at the time of diagnosis in both groups, the CPHD-PROP1 patients presented at a younger age and showed lower growth velocity during the 3 years preceding introduction of growth hormone therapy. Paradoxically, patients in the CPHD-nonPROP1 group were characterized by a better growth rate and thus an older age at the time of diagnosis. This indicates a possible significant role in the growth process of hormonal factors other than the somatotropin axis, most likely of hypothalamic-pituitary origin (the paradox of “growing without growth hormone”) [33‐36]. This issue requires further research. The TSH deficit in the CPHD-PROP1 was diagnosed in the younger cases and was more severe than in the CPHD-nonPROP1 group. As some reports have demonstrated, TSH deficiency in CPHD-PROP1 is also highly variable and has been reported as the first presenting symptom in some cases, while others show delayed TSH deficiency [37, 38]. Symptoms of hypogonadism at birth (non-descended testes and/or micropenis) were found less frequently in boys of the CPHD-PROP1 group. Although PROP1 is essential for the differentiation of gonadotrophs in fetal life, the spectrum of gonadotrophin deficiency ranges from hypogonadism presenting at birth with micropenis and non-descended testes to complete lack of pubertal development or even spontaneous pubertal development with infertility [38‐40]. The most interesting phenomenon in patients with the PROP1 mutation is the occurrence of secondary adrenal insufficiency. It is known that PROP1 is not directly involved in the transcription of genes necessary for the formation of ACTH. Nevertheless, some authors report late onset of ACTH deficiency in patients with PROP1 mutations [34, 41]. This phenomenon has been postulated to be a result of dysfunction of PROP1 in initiating pituitary stem cell migration and differentiation [42]. The etiology of this disorder in patients of the CPHD-PROP1 group remains the subject of hitherto unverified hypotheses. A possible mechanism of this phenomenon could be the lack of important paracrine factors normally produced by the cells surrounding the corticotropes and absent in the pituitary of these patients, or progressive corticotrope apoptosis. However, a similar phenomenon was not confirmed in the mouse model of the disorder. Ward et al. showed that PROP1 in mouse embryos influences the process of migration and metaplasia of the ectodermal to glandular epithelium [43]. Thus, the lack of its influence could adversely affect the morphogenesis and function of the anterior pituitary gland. The pituitary morphology in MRI in CPHD-PROP1 is variable. Most patients have normal pituitary stalk and posterior lobe, while PSIS is typically not observed. The anterior pituitary is normal or enlarged in the early stages of the deficiency and undergoes involution later [44, 45]. The involvement of PROP1 in the abnormal development of the anterior pituitary gland is indirectly supported by the presence of pituitary tumors characterized by an abnormal hypointense signal in T2-dependent MRI scans, which may increase or decrease, leading to hypoplasia of the pituitary gland over time in patients with CPHD-PROP1 [6, 46, 47]. Pituitary enlargement in patients with PROP1 gene inactivating mutations represents cystic hyperplasia of the intermediate pituitary lobe [48, 49]. In the present study, in the CPHD-PROP1 group, MRI of the pituitary gland revealed decreased pituitary diameter in 21 cases and its enlargement in eight. In nine CPHD-PROP1 patients, the anterior pituitary was normal in MRI, which was also found by Bulut et al. in 6/11 CPHD-PROP1 patients [50]. Surprisingly, in one patient, we found a pituitary stalk interruption and ectopy of the posterior lobe. Regarding this particular group of patients, this observation has not, to our knowledge, previously been described in the literature [11, 51], such an MRI finding being much more common in the CPHD-nonPROP1 group. This observation proves that typical pituitary stalk interruption signs in patients with CPHD do not exclude the diagnosis of CPHD-PROP1 and may indicate the coexistence of CPHD causes other than the PROP1 mutation in the same patient. It is also worth emphasizing the variability of MRI images and the lack of correlation between pituitary size and pituitary function in CPHD-PROP1 patients [51].

The main limitation of the current study results from its retrospective nature. Because the data were obtained from archival medical records, there are differences in the methodology of biochemical tests and imaging techniques. For the same reason, the oldest patients lack MRI imaging (this was not available at the time of diagnosis). An undoubted limitation of the study is also the lack of some data that were impossible to obtain due to the retrospective nature of the study and long observation period. Another limitation is the fact that the CPHD-nonPROP1 group consists of highly heterogeneous genetic disorders, i.e., mutations of many transcription factors expressed early in pituitary organogenesis resulting in syndromic hypopituitarism and mutations of POU1F1 (Pit1), which, as with Prophet of Pit1 (PROP1), are expressed at a later stage of pituitary organogenesis and result in non-syndromic hypopituitarism.