Highlights
· This study analyzed the clinical characteristics, brain magnetic resonance imaging findings, and genetic features of 43 Korean patients with congenital combined pituitary hormone deficiency over 27 years. It highlights the significance of neonatal presentations of hypopituitarism, underscoring the importance of early recognition and management.
Introduction
Congenital combined pituitary hormone deficiency (cCPHD) is the nonacquired deficiency of more than one hormone produced by the anterior pituitary gland or released from the posterior pituitary gland [1,2]. Pituitary gland development is regulated by the sequential expression of transcription factors and signaling molecules [2-4]. Abnormalities in these interactions caused by mutations in transcription factors or structural hypothalamic/pituitary abnormalities can lead to cCPHD [5].
Clinical manifestations of cCPHD vary depending on the timing and location of transcription factor expression during pituitary gland development. Mutations in the late stages of pituitary development cause only hormonal defects, in contrast to those in the earlier stages of pituitary development, which cause multiple hormonal defects and affect extrapituitary organs [1,2,5]. Several studies have reported genetic abnormalities in 1.3% to 7.3% of cases, underscoring that the identification of genetic abnormalities was limited to a small subset of patients [6-12]. Advances in next-generation sequencing (NGS) techniques have enabled the simultaneous analysis of numerous genes related to the development of the pituitary and hypothalamus. Recent studies have reported using NGS techniques, including whole exome sequencing (WES) and whole genome sequencing, to diagnose cCPHD [11,13].
Patients with cCPHD exhibit a wide range of phenotypes as a result of the various combinations of hormone deficiencies and the degree of combined malformations in other organs such as the eye, nose, and brain. In addition, during follow-up, some patients with a single pituitary hormone deficiency develop additional hormone deficiencies [14]. In the present study, we investigated the clinical manifestations, hormone profiles, radiologic features, and follow-up courses and performed a comprehensive genetic analyses in patients with childhood-onset cCPHD from a single tertiary center in Korea.
Materials and methods
1. Study population
We enrolled patients diagnosed with cCPHD diagnosed before the age of 20 years at Seoul National University Children’s Hospital between January 1994 and December 2021. CPHD was defined as the presence of at least 2 biochemically diagnosed pituitary hormone deficiencies, including deficiencies in growth hormone (GH), thyroid-stimulating hormone (TSH), luteinizing hormone (LH)/follicle-stimulating hormone (FSH), adrenocorticotropic hormone (ACTH), and antidiuretic hormone (ADH). Diagnosis of each hormonal deficiency was based on current guidelines (Supplementary Table 1) [15,16]. Of the 444 patients diagnosed with CPHD at <20 years of age, we enrolled 43 after excluding those with acquired causes of hypopituitarism such as pituitary and hypothalamic neoplasms, pituitary surgery, radiotherapy, trauma, vascular injury, or hypoxia in the perinatal or neonatal periods [2,3] (Supplementary Fig. 1).
2. Clinical data
All participants underwent a comprehensive medical evaluation at the time of initial identification of hormone deficiency and subsequent development of additional hormone deficiencies. The time of initial hormone deficiency identification was considered as the point of diagnosis. Data on age, anthropometrics, initial presenting symptoms, physical findings, perinatal history (e.g., neonatal jaundice, hypoglycemia, cryptorchidism, and/or micropenis in boys), family history of cCPHD, extrapituitary phenotype, and hormonal assays (basal and stimulated hormone levels) were collected. At some point during the clinical course, pituitary magnetic resonance imaging (MRI) was performed to evaluate the size and shape of the pituitary gland and stalk, or other cerebral anomalies. Pituitary stalk interruption syndrome (PSIS) was defined as a combination of hypoplasia or aplasia of the anterior pituitary gland, ectopic pituitary gland, and a thin stalk. Height, weight, body mass index (BMI), and metabolic profiles, including fasting glucose, hemoglobin A1c (HbA1c), and lipid profiles, were investigated at the time of adult GH deficiency (GHD) evaluation in patients who had reached their final adult height. Height, weight, and BMI z-scores were calculated using the 2017 Korean National Growth Charts for Children and Adolescents [17].
Serum HbA1c was measured using the immunoturbidimetric method with reagents containing mouse antihuman HbA1c monoclonal antibodies (IVD lab, Hwaseong, Korea), while serum glucose levels were quantified via an enzymatic method (ultraviolet rate) using commercially available reagents (SICDIA L-GLU, Shin Yang Pharm., Seoul, Korea). Serum total cholesterol (SICDIA TCHO, Shin Yang Pharm), serum triglyceride (clinical chemistry triglyceride reagent kit 7D74-22, Abbott, Wiesbaden, Germany), serum high-density lipoprotein (HDL) cholesterol (clinical chemistry ultra HDL reagent kit 3K33-22, Abbott), and serum low-density lipoprotein (LDL) cholesterol (MUTLIGENT direct LDL reagent kit 1E31-20, Abbott) levels were quantified using commercially available reagents.
Serum levels of cortisol (RIA; Immunotech, Prague, Czech Republic), LH/FSH (LH, FSH; IRMA; Asbach Medical Product, Obrigheim, Germany), GH (IRMA; Isotop, Budapest, Hungary), free T4, TSH (IRMA, Shin Jin Medics Inc., Ilsan, Korea), and T3 (RIA; Isotop) were also measured.
3. Genetic analysis
Genetic analyses were performed in 26 patients using targeted gene sequencing (n=19) or WES (n=7). Targeted gene sequencing comprised sequencing of 50 genes, including pituitary transcription factors (Supplementary Table 2). Six patients underwent WES because of combined developmental delays or multiple anomalies.
Targeted gene panels and WES were performed using NGS. Genomic DNA was extracted from the submitted specimens, and the SureSelectXT Human All Exon 50 Mb kit (Agilent Technologies, Inc., Santa Clara, CA, USA) was used to target the exon regions of genomes. Targeted regions were sequenced using the HiSeq sequencing system (Illumina, San Diego, CA, USA) with 100 paired-end reads. DNA sequences were mapped and analyzed in comparison with the published human genome build (UCSC hg19 reference sequence). Targeted coding exons and splice junctions of known protein-coding RefSeq genes were assessed for average depth of coverage and data quality threshold values. In the targeted gene panel, all reportable sequence variants below the threshold quality were confirmed by Sanger sequencing using a separate DNA preparation. For WES, reads were aligned to the Genome Reference Consortium Human Build 37 using Burrow-Wheeler alignment (v.2.2.0) [18].
4. Statistical analysis
All statistical analyses were performed using IBM SPSS Statistics ver. 27.0 (IBM Co., Armonk, NY, USA). Data are presented as means with standard deviations or medians with interquartile ranges (IQRs) for continuous variables. To compare variables between groups, the Mann-Whitney test was used for continuous variables, while Fisher exact test and linear-by-linear association were used for categorical variables. Statistical significance was indicated by a 2-sided P-value <0.05.
Results
1. Baseline characteristics
Among 43 CPHD patients (24 boys, 55.8%), mean age at diagnosis was 3.2±3.8 years (range, 4 days to 14 years; Table 1). Mean height, weight, and BMI z-scores at diagnosis were -2.4±3.8, -0.9±4.8, and 0.3±1.3, respectively. Short stature was the most frequent initial presentation (n=16, 37.2%), followed by neonatal jaundice (n=6, 14.0%) and neonatal hypoglycemia (n=4, 9.3%). More than half of patients (n=23, 53.5%) had one or more neonatal features suggestive of congenital CPHD, such as neonatal jaundice, cryptorchidism or micropenis, neonatal hypoglycemia, hypotension, and lethargy. The most common neonatal features were jaundice in 15 (34.9%) and cryptorchidism or micropenis in 9 of 24 boys (37.5%). Among 23 patients with neonatal features, only 15 (65.2%) were diagnosed in infancy. Thirty-one patients (72.1%) had an extrapituitary phenotype, with developmental delay being the most prevalent (n=15, 34.9%).
Comparison of patients diagnosed in infancy and later revealed that the later diagnosis group had lower height z-scores at diagnosis (-3.4 vs. -1.1, P=0.054) but a higher GH peak level (0.1 vs. 1.1, P=0.031) compared to the early diagnosis group. There were no significant differences in the prevalence of extrapituitary phenotypes, such as developmental delay or visual disorders, between the 2 groups (Supplementary Table 3).
2. Pituitary hormone deficiency profiles
At initial diagnosis, TSH deficiency was most prevalent (n=28, 65.1%), followed by GH, ACTH, and ADH deficiencies. Eight patients (18.6%) had multiple hormonal deficiencies during initial evaluation. At the final follow-up, GH deficiency was most common (n=42, 97.7%), with all but 1 patient having GH deficiency (Fig. 1A and 1B). ACTH and TSH deficiencies were observed in 34 patients (79.1%), each at final follow-up. Of 31 patients who could be evaluated in terms of pubertal age (>9 years in girls, >10 years in boys), 18 (58.1%) had gonadotropin (LH/FSH) deficiency. ADH deficiency was observed in 6 patients (14.0%). The most prevalent combination of deficiencies was GH, ACTH, and TSH (n=13, 30.2%). Three-quarters of patients (n=33, 76.7%) had 3 or more hormonal deficiencies at final follow-up. Detailed hormone deficiency profiles are presented in Supplementary Table 4.
Median age at diagnosis was 1.0 years (IQR, 0–5) for TSH deficiency, 6 (IQR, 1.6–28) months for ADH deficiency, 4.0 (IQR, 2.0–7.3) years for GH deficiency, 3.0 (IQR, 0–10.3) years for ACTH deficiency, and 13.0 (IQR, 13.0–15.5) years for gonadotropin (LH/FSH) deficiency. One patient was diagnosed with GHD but received no treatment, whereas all other patients with cCPHD underwent hormone replacement therapy for pituitary hormone deficiencies.
3. Brain imaging findings
Brain MRI was performed in 41 patients (95.4%), of whom 34 (82.9%) had 1 or more abnormalities (Table 2). Pituitary abnormalities such as anterior pituitary hypoplasia/aplasia, posterior pituitary ectopic/aplasia, and thin pituitary stalk/stalk aplasia were found in 33 patients (97.1%), of whom 6 (14.6%) showed PSIS. Other findings, such as midline defects, optic nerve hypoplasia, and septo-optic dysplasia, were observed in 11 patients (25.6%). Patients with abnormal brain MRI findings had higher numbers of hormone deficiencies than those in the normal MRI group (P for trend=0.049; Supplementary Table 5). The abnormal MRI group also had lower peak GH levels than the normal MRI group with marginal significance (0.6 ng/mL vs 4.3 ng/mL, P=0.059). There were no significant differences in age or anthropometrics at diagnosis or in the prevalence of extrapituitary phenotypes between the groups.
4. Outcomes at the time of adult GHD evaluation
During the follow-up period, 17 of the 22 patients who reached adult height were reevaluated for adult GHD, with this diagnosis confirmed in all. Final z-score values for adult height, weight, and BMI were -0.8±1.9, and 0.3±1.7, and -1.8±1.6, respectively (Table 3). Ten patients (58.9%) had 1 or more metabolic abnormalities, such as impaired fasting glucose or dyslipidemia and hypertriglyceridemia, and low HDL cholesterol levels were common. None of the patients developed diabetes mellitus or received antidiabetic medications or lipidlowering therapies.
5. Genetic evaluation
Twenty-six patients (60.5%) underwent genetic evaluation. Patients who underwent genetic testing were younger at diagnosis and a higher prevalence of extrapituitary phenotypes, particularly developmental delay, than those who did not undergo testing (Supplementary Table 6). Pathogenic or likely pathogenic variants were identified in 5 patients (19.2%). These mutations included heterozygous variants in POU1F1, HESX1, GLI2, and ROBO1 and compound heterozygous variants in TBC1D32 (Table 4). Three of 5 variants (3/19, 15.8%) were identified from the target gene panel, and 2 (2 of 7, 28.6%) from exome sequencing. All patients except 1 had their first hormone deficit identified in infancy, and 3 were diagnosed within the first month of age. Every patient with genetic variants showed 1 or more extrapituitary symptoms such as developmental delay, dysmorphic face, and optic nerve hypoplasia. The detailed clinical characteristics of the patients are shown in Table 4.
Discussion
In the present study, we evaluated the clinical characteristics and genetics of 43 patients with cCPHD in Korea. In our cohort, while the primary manifestation in most patients was short stature during childhood, half exhibited neonatal features suggestive of hypopituitarism in infancy. Prevalent MRI abnormalities were identified and correlated with number of pituitary hormone deficits. Among 26 patients who underwent genetic evaluation, 5 (19.2 %) had pathogenic variants.
In our cCPHD cohort, mean age at diagnosis was 3.2 years, and most patients were diagnosed at age 1–13 years (53.5%) or <1 year (41.9%). A few studies have reported a similar range of diagnostic ages from 2 to 6 years, although the range varies among studies, likely due to the small sample sizes [11,12,19]. We further observed lower peak GH levels in those diagnosed in infancy than in those diagnosed after infancy, suggesting that early diagnosis may be associated with more severe phenotypes. Wadams et al. [20] also reported that patients with a more severe phenotype (cCPHD and optic nerve hypoplasia) were diagnosed at an earlier age. A recent large cohort study from Denmark further reported a higher number of hormone deficiencies in patients diagnosed at less than 1 year of age than in those diagnosed between ages 1 and 17 years [12].
In our study, more than half of the patients showed neonatal features suggestive of cCPHD during infancy; however, only one-third of the patients were diagnosed during infancy. A previous study similarly reported that 24% of patients with isolated neonatal symptoms, such as hypoglycemia or jaundice, had missed hormonal evaluations, leading to delayed diagnosis of cCPHD [11]. Hypoglycemia, prolonged jaundice, electrolyte abnormalities, micropenis, and cryptorchidism are important indicators of cCPHD in neonates [1]. Despite the ambiguity of these symptoms, it is important not to overlook the early signs of cCPHD.
In the present study, we found that TSHD was the most frequent hormone deficiency at the initial diagnosis, whereas GHD was the most prevalent deficiency in patients at the last follow-up. In previous studies, GHD was the most common deficiency during the disease course, followed by TSHD [7,12,13]. Yet another study reported similar rates of GHD and TSHD [8]. Although most patients (76.7%) eventually presented with 3 or more pituitary hormone deficiencies, only 18.6% had multiple hormone deficiencies at the time of their initial evaluation. A previous study that investigated the characteristics of patients with isolated GHD who later developed CPHD reported that a more severe GHD phenotype, indicated by a lower height standard deviation score at diagnosis or a lower stimulated GH level, was associated with the development of CPHD [14]. It is essential to monitor the symptoms of additional hormone deficiencies, such as growth retardation and delayed puberty, and to perform appropriate hormonal evaluations in such patients.
Similar to previous studies, most patients exhibited abnor malities on brain MRI [6-8,11,12,19,21]. The most common structural pituitary abnormalities in previous studies were an absent or ectopic posterior pituitary gland (60.2% to 72.7% of cases), followed by an absent or hypoplastic anterior pituitary gland and pituitary stalk problems (prevalence of 22.0% to 55.6%), consistent with the findings of this study [11,12]. A recent British study reported that anterior pituitary gland hypoplasia or aplasia was the most common pituitary abnormality (72.7%) [13]. In the present study, patients with abnormal MRI findings demonstrated more pituitary hormone deficiencies than those with normal MRI findings. A previous study reported a higher prevalence of pituitary abnormalities on brain imaging among those diagnosed in infancy (at an early age) compared to those diagnosed at 1–17 years [12]. Although no significant differences were observed in other clinical phenotypes based on MRI findings in the present study, patients with abnormal MRI results could be considered to have more severe phenotypes.
In this study, all patients evaluated for adult GHD were diagnosed with GHD, which supports the current international guidelines suggesting that reevaluation is not necessary in patients with organic causes, such as genetic or structural defects in the hypothalamic-pituitary region [22]. At the time of adult GHD evaluation, our patients demonstrated relatively favorable height outcomes; however, more than half showed mild metabolic abnormalities. Considering the central role of GH in glucose and lipid metabolism [23,24], metabolic abnormalities may be associated with GH treatment status, although detailed treatment profiles during the transition period were not assessed in this study. Several studies have indicated a worsening of metabolic profiles during the transition period associated with GH discontinuation [22,25-27]; however, no studies have directly focused on the metabolic outcomes of childhood-onset cCPHD patients. As the underlying metabolic risk factors may differ between patients with congenital and acquired CPHD, further research is needed to understand the outcomes of cCPHD in adulthood.
In this study, 5 patients (19.2%) carried pathogenic variants of POU1F1, GLI2, HESX1, TBC1D32, and ROBO1 at higher frequencies than those reported previously (range, 1.3%–16.7%) [6-9,13,19,28]. This wide range of detection rates may result from the variable clinical characteristics of the participants, differences in the proportion of familial cases, or the methods used to detect variants. While the decision to perform genetic evaluation was clinician-dependent in this study, upon retrospective review, individuals who underwent genetic evaluation were diagnosed at an earlier age and exhibited a higher prevalence of extrapituitary symptoms, which may also have contributed to an increased positivity rate. Most of the previous studies screened only a few selected genes by direct sequencing of around 1–4 genes, resulting in a positive rate of 1.3% to 3.7% [6-9]. There have been 2 Korean studies in which direct sequencing of the POU1F1, PROP1, LHX3, LHX4, and HESX1 genes was performed in patients with cCPHD. Low frequencies of pathogenic variants, ranging from 3.7% (only 1 patient with a HESX1 variant) to 16.7% (2 patients with PROP1 gene mutations) were reported [6,29]. Currently, more than 30 genes are known to be associated with cCPHD, and additional novel genes are being identified using NGS technology [30]. A recent study from Finland using WES for the genetic investigation for cCPHD patients reported a positivity rate of 15%, which is comparable to this study [11]. Although advances in NGS have improved our understanding of the genetic background of cCPHD, the causes of cCPHD are still undiscovered in the majority of cases, suggesting the need for further research to investigate the pathogenesis of these disorders, focusing not only on genetic factors, but also on epigenetic or environmental factors.
Although PROP1 and POU1F1 are commonly associated with cCPHD, most of our patients had variants in other genes involved in early hypothalamus-pituitary development [31], which can be explained by the high prevalence of extrapituitary phenotypes in our cohort. TBC1D32 and ROBO1, 2 genes recently reported to be associated with cCPHD, were also identified in this study. Patient #2, a previously reported case [18] presenting with multiple pituitary hormone deficiencies and a wide range of extrapituitary symptoms, had compound heterozygous mutations of TBC1D32. TBC1D32 encodes a protein associated with ciliary function that regulates cellular polarity and vital signaling cascades, including Sonic Hedgehog signaling [32]. Patient #5 with a dysmorphic face and multiple hormone deficiencies (GH, TSHD, ACTHD) had a novel heterozygous ROBO1 variant. Mutation of ROBO1 is a recently proven genetic cause of CPHD; the protein encoded by this gene is essential for axonal guidance and branching during embryonic development of the nervous system [4,33]. Mutations in ROBO1 have been reported to be associated with PSIS and craniofacial phenotypes, including ophthalmic abnormalities and dysmorphic face [34,35].
Several limitations of this study should be considered. First, the sample size was limited because of the rarity of cCPHD and the single-center nature of the study. Second, data on neonatal history and anthropometric measurements were missing because of the retrospective study design. Third, genetic investigations were conducted only on a subset of patients using various modalities, making it difficult to generalize the results. Nevertheless, this is the first Korean study to investigate the long-term clinical outcomes of patients with cCPHD and to perform genetic evaluation of these patients.
In conclusion, given that most patients with cCPHD have features indicative of hypopituitarism from the neonatal period, clinicians should be aware of suspicious symptoms during infancy. Early detection is important in the timely and appropriate management of both pituitary and extrapituitary phenotypes associated with cCPHD. Although advances in genetic technology have identified additional candidate genes for cCPHD, the yield of genetic testing remains low and warrants further investigation.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement1
Print
