A case of autosomal dominant hypophosphatemic rickets

Highlights

● Careful anamnesis is essential for determining the clinical approach to hypophosphatemia.

● Monitoring iron metabolism is necessary for optimizing the therapeutic management of ADHR.

Introduction

Autosomal Dominant Hypophosphatemic Rickets (ADHR) is an extremely rare form of hypophosphatemic rickets characterized by autosomal dominant inheritance (1). Under these conditions, a pathogenic variant of fibroblast growth factor 23 (FGF23) produces an abnormal form of FGF23 that is resistant to degradation. This results in excessive FGF23 levels in the blood, leading to hypophosphatemia. Owing to the nature of the disease, the most pathogenic variants that cause ADHR are missense variants that presumably do not impair FGF23 function. The most commonly reported pathogenic variants are p.Arg176Gln, p.Arg176Trp, p.Arg179Gln, and p.Arg179Trp. Arginine residues at positions 176 and 179 correspond to the consensus sequence RXXR, which is essential for protein degradation.

Additionally, some clinical features vary and their pathological mechanisms remain to be elucidated. For example, the age at onset varies with incomplete penetrance. One possible explanation is that the molecular function of mutant FGF23 proteins depends on the type of missense variant, and accumulation of cases with detailed clinical features is required.

Case Presentation

Informed consent was obtained from the patient’s parents.

A 1-yr and 7-mo-old boy presented to our pediatric general outpatient clinic with the chief complaint of “short stature.” His height was 71.5 cm (height SDS: –3.5 SD) and his body weight was 9.2 kg. The patient exhibited severe genu varum and an unstable gait. Skeletal changes on X-ray were suggestive of rickets, showing metaphyseal splaying, cupping, and fraying (Fig. 1a). Routine laboratory tests revealed hypophosphatemia (phosphate: 1.6 mg/dL) and hyper-alkaline phosphatasemia (alkaline phosphatase: 1,819 IU/L, reference range 140–470). Serum levels of 1,25-dihydroxy-vitamin D, 25-hydroxy-vitamin D, and intact parathyroid hormone were within normal ranges for both age and sex. There was decreased tubular phosphate reabsorption (TmP/GFR: 1.8 mg/dL, reference range 5.31 ± 0.40) without hypercalciuria. Although these findings strongly suggested hypophosphatemic rickets, FGF23 levels were not elevated (FGF23: < 5.0 pg/mL, reference range < 30) (Table 1). Furthermore, the blood level of intact FGF23 was not elevated (7.9 pg/mL) at 3-yr and 7-mo-old. On the screening examination, we noted that the serum level of iron was relatively low (Fe: 48 µg/dL, reference range 50–200) (Ferritin: 23.7 ng/mL, reference range 9.0–275).

Fig. 1.

(a-1, a-2) Radiographs of the patient (1 yr and 7 mo) showed metaphyseal splaying, cupping, and fraying. (b-1, b-2) Radiographs of the patient (2 yr and 7 mo) showed improved skeletal abnormalities associated with rickets. (c) No Looser’s zones was observed on X-rays of the father.

Table 1. Laboratory findings of the patient and father.

The patient’s father was treated with oral vitamin D supplements and orthotic therapy during childhood because of “vitamin D-resistant rickets.” These problems diminished with age, and treatment was not required during adolescence or adulthood. His final height was 171 cm (+ 0.03 SD). The proband’s father revealed mild hypophosphatemia (phosphate: 2.3 mg/dL) and a mild reduction in tubular phosphate reabsorption (TmP/GFR: 2.1 mg/dL, reference range 2.3–4.3) (Table 1). Serum levels of alkaline phosphatase and FGF23 were within the normal range (Table 1), and no Looser’s zones were observed on X-ray (Fig. 1c). None of the other patients had similar clinical features in their family members. In the proband, oral phosphorus and vitamin D supplementation improved the growth retardation and skeletal abnormalities associated with rickets (Fig. 1b, Fig. 2a).

Fig. 2.

(a) Growth chart and treatment course of the patient demonstrating improvement in growth retardation following treatment. (b) Chromatograms from DNA sequencing of the patient and father. We identified the same variant, c.526C>G, in exon 3 of FGF23 gene, as the proband whose genetic analysis was carried out using next-generation panel sequencing. (c) Structure and cleavage sites of FGF23.

Given the family history of rickets and the atypical clinical course, we strongly suspected a genetic cause underlying this condition. Next-generation sequencing panel analysis for rickets (Kazusa DNA Research Institute) revealed a heterozygous pathogenic variant, c.526C>G (NM_020638.3: p.Arg176Gly), in exon 3 of FGF23 gene, indicating ADHR (2). No pathogenic variants were identified in the PHEX (phosphate-regulating endopeptidase X-linked), DMP1 (dentin matrix acidic phosphoprotein 1), ENPP1 (ectonucleotide pyrophosphatase/phosphodiesterase 1), or SLC34A3 (solute carrier family 34 member 3) genes. The same variant in FGF23 was also identified in the patient and father (Fig. 2b).

To date, most pathogenic variants identified in ADHR are missense mutations affecting arginine at position 176 or 179, or deletion mutations involving arginine 176. Arginine residues at positions 176 and 179 are required for proper degradation of FGF23, and variants affecting these sites are likely resistant to degradation.

According to ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), three missense variants (p.Arg176Gln, p.Arg176Trp, and p.Arg176Pro) and one in-frame deletion variant (p.Pro172_Arg176del) affecting arginine at position 176 have been identified. Among the three missense variants, two (p.Arg176Gln and p.Arg176Trp) were judged to be pathogenic, and the last variant (p.Arg176Pro) had uncertain significance. In silico analyses, AlphaMissense (https://alphamissense.hegelab.org/#) and PolyPhen2 (http://genetics.bwh.harvard.edu/pph2/) yielded likely pathogenic and probably damaging, respectively. According to the American College of Medical Genetics and Genomics (ACMG) guidelines (3), the p.Arg176Gly variant met the criteria for Pathogenic Strong (PS1), Pathogenic Moderate (PM1, PM5), and Pathogenic Supporting (PP2, PP3).

Discussion

Based on our experience, we add two findings to the current understanding of ADHR. First, the severity of hypophosphatemia is age-dependent, which may partly explain the incomplete penetrance and variability in clinical features of the disease. Second, a missense variant of FGF23 could potentially affect the FGF23 assay, resulting in lower values than expected.

Although an in vitro assay of p.Arg176Gly FGF23 was not performed, we concluded that the missense variant p.Arg176Gly is the cause of ADHR. First, considering the correlation between genotype and phenotype, the p.Arg176Gly variant is presumed to cause ADHR. This is the second reported case of ADHR with heterozygous p.Arg176Gly, and the proband’s father, who had the same variant, exhibited mild hypophosphatemia. Furthermore, we did not identify any variants in genes that cause hypophosphatemia, such as PHEX, DMP1, ENPP1, and SLC34A3. Consistently, the variant is classified as “pathogenic” based on the ACMG guidelines criteria PS1, PM1, PM5, PP2 and PP3.

Secondly, the inactivation mechanism of FGF23 further supports our hypothesis that the p.Arg176Gly causes FGF23 excess. The full-length FGF23 (intact FGF23) was inactivated by cleavage between Arg179 and Ser180 (Fig. 2c). The RXXR motif located adjacent to the cleavage site is essential for protein inactivation (4). The arginine at position 176 is the first arginine of the RXXR motif, and its substitution is presumed to impair protein inactivation, resulting in a supraphysiological excess of FGF23.

Based on the patient’s history, we presume that, although the clinical manifestations improved in adulthood, episodes during the father’s childhood were indicative of ADHR. Such age-dependent changes in clinical phenotypes could obscure family history, possibly delaying diagnosis. Therefore, careful anamnesis is essential to determine the appropriate clinical approach for hypophosphatemia.

Although identifying the mechanisms underlying improved clinical phenotypes in adulthood was beyond the scope of this study, one possible explanation is the decreased renal phosphate excretion after puberty (5). Another potential explanation is iron insufficiency. Serum iron levels are inversely associated with FGF23 levels. Lower iron levels increase serum FGF23 levels, potentially disrupting the balance of phosphate metabolism (6, 7). Infants and children are at high risk of iron insufficiency, which may partially explain the age-dependent severity of the disease. Indeed, in our case, the blood iron level was relatively low (Fe: 48 µg/dL, Ferritin: 23.7 ng/mL) but recovered by the age of 3 yr and 4 mo (Fe: 119 µg/dL, Ferritin: 34.8 ng/mL). In addition to oral phosphorus and vitamin D supplementation, improvements in iron levels may have contributed to the resolution of clinical phenotypes. Monitoring iron kinetics could be essential for therapeutic approaches to ADHR, especially during infancy and childhood.

The clinical pathogenesis of ADHR typically involves elevated levels of intact FGF23, leading to renal phosphate wasting. In the patient, the intact blood FGF23 level was not detectable at the time of examination (1-yr and 7-mo-old) and was not elevated at 3-yr and 7-mo-old. Interestingly, the FGF23 level in our case was within the normal range. Unfortunately, the only study that previously reported a case of p.Arg176Gly did not include data on FGF23 levels (2), which prevented us from confirming whether this phenomenon is common in individuals with p.Arg176Gly. One possible explanation is that the FGF23 antibody used in our assay was unable to adequately recognize the p.Arg176Gly variant of FGF23. The p.Arg176Gly mutation could alter the 3D structure of the protein, possibly affecting the antigen-antibody reaction in the assay. The measuring system in our case recognized both the C and N terminals of the protein using a sandwich ELISA assay, implying that subtle structural changes in the protein could affect the assay’s accuracy. Although the inability to access the measurements of other assays to detect FGF23 was a limitation of our study, a similar case of ADHR with undetectable FGF23 was previously reported (8).

In conclusion, we present a case of ADHR that provides insight into the clinical features and pathogenesis of this disease. To further understand the pathological mechanisms involved, more cases are required.

Conflict of interests

The authors declare no conflict of interest.