Highlights
· This study characterizes the clinical and genetic profiles of 79 children with simple virilizing congenital adrenal hyperplasia due to 21-hydroxylase deficiency at a tertiary pediatric center. The median age at diagnosis was 4.5 years; common presentations included penile enlargement in males and clitoromegaly with advanced bone age in females.
· Genetic analysis of 43 children identified 21 pathogenic variants and 22 genotypes, with p.I173N being the most frequent.
· The findings underscore delayed diagnosis and support early genotyping, particularly in newborn-screened children, to enable optimized, personalized management.
Introduction
Congenital adrenal hyperplasia (CAH) is a group of autosomal recessive genetic diseases that cause a deficiency of one of the necessary enzymes to synthesize adrenal steroid hormones from cholesterol. Among them, pathogenic variants of the CYP21A2 gene that encodes the 21-hydroxylase (21-OH) enzyme account for 95% of CAH cases, with an incidence rate in most races of about 1 in 10,000 to 1 in 16,000 live births [1,2]. CAH due to 21-hydroxylase deficiency (21-OHD) is divided into 3 primary forms depending on remaining 21-OH enzyme activity: salt wasting (SW) (0% enzyme activity), simple virilizing (SV) (1%–5% enzyme activity), and nonclassical form (20%–50% enzyme activity) [3,4]. The SV form, accounting for about one-third of classical CAH cases, is characterized by excess androgen production leading to ambiguous genitalia in females and progressive virilization in both sexes [5-7].
In Vietnam, there has been a noticeable increase in CAH diagnosis in recent years, likely due to improvements in newborn screening programs and heightened awareness among healthcare providers. Despite these advancements, SV-CAH is underdiagnosed in the absence of SW crises that require immediate medical attention. Ambiguous genitalia in newborn girls and subtle signs of virilization in boys can go unnoticed or be mismanaged without specialized expertise, resulting in delayed diagnosis and complications such as short stature, early puberty, and psychosocial challenges [5-7].
This study aims to describe the clinical features and genetic findings of SV-CAH in children managed at a tertiary pediatric center in Vietnam. By highlighting this population’s genotype-phenotype correlations and unique characteristics, we hope to improve early diagnosis, guide individualized treatment strategies, and contribute to the broader understanding of SV-CAH, particularly in regions with limited resources for endocrine care.
Materials and methods
The data presented here are from 79 Vietnamese children with SV-CAH due to 21-OHD (among a total of 230 children with classic CAH) diagnosed at the Center for Endocrinology, Metabolism, Genetics/Genomic and Molecular Therapy in Vietnam National Children's Hospital, from November 2016 to March 2023.
This was a retrospective cross-sectional analysis. Data were extracted from hospital medical records.
The inclusion criteria for 79 children were (1) genital virilization at birth or progressive virilization after birth in girls or penile enlargement in boys, (2) rapid growth and bone age acceleration or positive newborn screening [8,9], (3) high serum 17-hydroxyprogesterone (17-OHP) level [10] and/or CYP21A2 gene analysis results, and absence of salt wasting for at least the first 3 years of life [1] in children diagnosed at a very young age who received glucocorticoid replacement and supplemental mineralocorticoids.
The study excluded children who had CAH due to deficiencies in other enzymes such as 11-hydroxylase, 3β-hydroxysteroid dehydrogenase, or 17α-hydroxylase or the SW form of CAH due to 21-OHD. The study design and inclusion/exclusion criteria are illustrated in Fig. 1.
Clinical and laboratory characteristics were collected at the time of admission before starting hormone replacement therapy and at the final assessment. The clinical information gathered includes reasons for examination, gender, age at diagnosis, degree of genital virilization according to the Prader scale in girls, penile circumference and length and testicular volume in boys, height, hyperpigmentation, pubic hair growth, acne, height at admission, and central precocious puberty. Laboratory investigations were performed for serum 17-OHP, adrenocorticotropic hormone (ACTH), and testosterone levels and bone age at diagnosis and at the final assessment according to Atlas of Greulich and Pyle.
Genotyping and genetic identification of pathogenic variants of CYP21A2 were performed on 43 children using complete sequencing of the CYP21A2 gene and multiplex ligation-dependent probe amplification (MLPA) analysis. Sequence variants were classified as null, A, and B based on their projected in vitro enzymatic function and in silico simulations [3,11]. Homozygotic children with the variant p.I173N (approximately 2% residual 21-OH enzyme activity) or heterozygotes with the null, A, or B group variants were assigned to group B, with 1%–2% residual enzyme activity. Group D comprised patients with unknown enzymatic consequences and novel, uncharacterized variants. Group C consisted of patients with the nonclassic form; we did not have any cases in this group.
MLPA analysis, a high-throughput and straightforward technique for quantifying gene copy numbers, was performed using the MLPA Kit P050B2 (MRC-Holland) according to the manufacturer's protocol. Products from amplification were analyzed on a 3100-Avant Genetic Analyzer ABI-PRISM (ThermoFisher, USA). The kit contains 5 probes for CYP21A2 (p.P31L, p.G111fs*21, p. I173N, E6cluster, and p.Q318*), 3 CYP21A1P-specific probes, and 2 probes for complement components (C3, C4) and is utilized to detect deletions and duplications of one or more exons of CYP21A2.
To identify missense variants in the CYP21A2 gene, a polymerase chain reaction (PCR) amplification product (100–150 ng DNA) was obtained for each sample. After agarose gel separation, the PCR product was purified with a Gel Purification Kit followed by sequencing using Big Dye Terminator V3.1 (Applied Biosystems, USA). CLC Main Workbench Software was used to analyze the results with the reference CYP21A2 in GeneBank (Accession number NM_000500.9). Unreported variants were confirmed by searching online databases (i.e., Leiden Open Variant Database) and publications on the CYP21A2 gene. The primers used are provided in Supplementary Table 1 [3,11].
1. Statistical analysis
Data were processed using IBM SPSS Statistics ver. 25.0 (IBM Co., USA). Percentages were calculated and compared using the binomial test for qualitative variables, while the Wilcoxon rank-sum test was applied to compare the median age between males and females; the Mann-Whitney U-test was used to analyze the median height at diagnosis and chronological-to-bone age difference within the 2 genders and to compare the median current height and the chronological-to-bone age difference between the 2 groups (diagnosis before or after 3 years of age); and the chi-square test was performed to evaluate the percentage of patients with central precocious puberty. Additionally, mean and standard deviation (SD) were calculated for quantitative variables. A P-value <0.05 was considered statistically significant.
2. Ethical statement
Informed consent was obtained from every child’s parents before inclusion in the study. No identifiable information of the patients was disclosed in any form. This study was reviewed and approved by the ethical committees of Vietnam National Children’s Hospital, Hanoi, Vietnam (420/BVNTW-HĐĐĐ, date 9 Mar 2023) (IRB-VN0137/IRB00011976/FWA00028418).
Results
The cohort with a confirmed diagnosis of SV-CAH caused by 21-OHD comprised 79 patients from 75 families (4 pairs of siblings). Table 1 displays the characteristics and clinical symptoms of the subjects.
Among the 30 boys, the oldest was 13.5 years, and 19 boys had a chronological age of 10.0±1.9 (range, 7.0–13.5) years with a bone age of 14–15 years. Their average height was 144.4±9.50 (median, 147.0; range, 122.7–163.0) cm corresponding to a height deviation of 1.9±1.3 (median, 1.5; range, 0.5–5.15) SDS.
Among the 49 girls, there were 2 adult cases: one was 27 years old with a height of 145 cm (-2.8 SDS-WHO) and the other was 19 years old and 152 cm tall (-1.7 SD WHO). Five girls had a chronological age of 11.6±2.58 (range, 8.5– 15.0) years with a bone age of 14 to 16 years. Their average height was 145.2±6.31 (median, 146.2; range, 137– 154) cm, corresponding to a height deviation of -0.02± 1.74 (median, -0.08; range, -1.1 to 3.0) SDS.
Genotyping of the CYP21A2 gene was performed for 43 children and identified 22 genotypes (Table 2) with 21 variants (Table 3) [12]. The spectrum of pathogenic variants—including missense, large deletions, splicesite, frameshift, and nonsense variants—is shown in Fig. 2. These variants were analyzed genetically through complete sequencing of the gene and MLPA analysis. Among the children, one had a complex genotype with 3 pathogenic variants, and one had only one heterozygous variant. These 2 children and 3 others with compound heterozygosity carrying rare variants were included in genotype group D (Table 2). The child with a single heterozygous pathogenic p.I173N variant was a 5-year-old boy who first presented at 4 years of age with penile enlargement. His symptoms included hyperpigmentation, acne, height deviation of 3.5 SDS, penile length of 7 cm, testicular volume of 3 mL, and advanced bone age of 7 years. Laboratory investigations revealed ACTH: 4.73 pmol/L, 17-OHP: 891 nmol/L, and testosterone: 4.81 nmol/L. Abdominal ultrasound findings were normal. However, genetic analysis identified only one variant, possibly due to the limitations of the testing method in detecting a second pathogenic variant.
The c.439T>G (p.F147V) variant was identified in a 7-year-7-month-old girl presenting with clitoromegaly. Clinical examination revealed clitoromegaly of 3 cm with a urogenital sinus and a single urogenital opening, classified as Prader stage III. Bilateral breast development was at Tanner stage B1, pubic hair was at stage P3, and her height deviation was 3.41 SDS; she demonstrated no hyperpigmentation but notable acne. Her bone age was advanced at 11 years, exceeding her chronological age by 3.5 years. Laboratory results showed normal levels of 17-OHP (1.47 nmol/L), testosterone (0.08 nmol/L), and ACTH (0.33 pmol/L) but a low morning cortisol level of 6.07 nmol/L. Urinary steroid profile and abdominal ultrasound were normal. Genetic testing through MLPA analysis and sequencing identified compound heterozygous variants in the CYP21A2 gene: c.293-13C>G and c.439T>G (p.F147V). Subsequent 17-OHP measurements consistently showed elevated levels, peaking at 284 nmol/L. At 8.5 years, the patient underwent clitoral and vaginal reconstruction surgery. While the c.293-13C>G variant is classified as pathogenic, the c.439T>G (p.F147V) variant is categorized as uncertain significance but is considered a polymorphism by MutationTaster and as possibly damaging by PolyPhen-2, supported by a CADD score of 15.99.
Discussion
1. General characteristics
This study described 79 children with SV-CAH due to 21-OHD from November 2016 to March 2023 including 4 sibling pairs among 13 families. Compared to global studies, our cohort has a significantly higher number of SV-CAH children: Yoo et al. [13] (Korea) identified 11 cases in 18 years, Wang et al. [14] (China) reported 4 SV cases in 19 years, Hou et al. [15] (China) reported 25 SV cases in 6 years, and Chi et al. [16] (Vietnam) reported 44 SV of 212 CAH children in 6 years (2011–2016). The 79 SV children in 7 years in our cohort is a large number for a single center. Riedl’s compilation across 44 centers in Germany and Austria included 538 CAH children, with 98 (18.2%) having the SV form [17] a scale comparable to our findings.
CAH is an autosomal recessive genetic condition and should theoretically demonstrate no sex differences. However, our results found a larger proportion of females (62.0%), likely due to more frequent detection of genital abnormalities at birth. In comparison, males with SV-CAH are underdiagnosed due to subtle clinical manifestations that might not be apparent in the first 18 months of life [18]. These findings align with those of Santos-Silva et al. [19] (Portugal, 2019), where 75% of SV cases were female.
The median age at diagnosis in our study was 4.5 years (1 day–22.3 years), similar to findings from China and Korea but older than patients in European studies. Yoo et al. [13] (Korea, 2013) reported a median age of 7±5 years, Hou et al. [15] (China, 2019) documented an age range from 12 days to 12 years, and Santos-Silva et al. [19] (Portugal, 2019) found a median age of 1.4 years. Umaña-Calderón et al. [20] noted all of their 58 classical CAH cases including 5 SV were diagnosed before the age of 4.
2. Phenotypes
In our cohort, the main reasons for medical attention were large phallus (53.3%) and rapid growth (13.3%) in males and genital abnormality (87.8%) in females (Table 1), aligning with global studies. Santos-Silva et al. [19] reported diagnostic factors of pubarche in 83.3% and newborn screening in 16.7% of males, while 100% of females exhibited genital abnormalities. Zeng et al. [21] (China) found clitoromegaly in 89.62% (in a total of 33 SV-CAH cases) of females, while Umaña-Calderón et al. [20] (Costa Rica) noted hyperpigmentation in all cases.
At diagnosis, most subjects had advanced bone age and rapid growth, potentially leading to short adult height. Lee et al. [22] (Taiwan) observed rapid growth and advanced bone age in all SV-CAH males, while 100% of females had clitoromegaly and 68% had fast growth. In our cohort, one adult, diagnosed at 22 years, had a final height of 148 cm. Zeng et al. [21] reported that 100% of adult CAH cases had short stature. Juan et al. [23] (82 CAH cases, 59 SV) found significantly shorter adult height compared to the general population (-1.9±1.1 SD, P< 0.001).
Treated cases showed better outcomes (-1.7±1.1 SDS) than untreated cases (-2.6±1.0 SD, P<0.05). Early intervention significantly improved final height compared with the late treated or untreated group (P<0.001 and P=0.013, respectively), emphasizing the importance of early diagnosis and treatment. Additionally, the earlytreated group tended to be taller than the late-treated group (P=0.089) [23], further underscoring the importance of timing (Table 4).
Serum 17-OHP, ACTH, and testosterone levels were elevated in our cohort, consistent with previous studies. Santos-Silva et al. [19] reported 17-OHP at 457 and 1,035 nmol/L, ACTH at 56.54 and 108.9 pmol/L, and testosterone at 11.55 and 3.57 nmol/L in males and females, respectively. Liu et al. [24] (China) found similarly elevated hormone levels in 26 CAH infants with 17-OHP at 254 (126–527), nmol/L ACTH at 20.8 (11.76–53) pmol/L, and testosterone at 9 (3.5–28.1) nmol/L. While biochemical variations depend on sample timing and laboratory techniques, persistently high 17-OHP, ACTH, and testosterone levels at diagnosis have diagnostic value.
3. Genotypes
1) Variant types
Our findings align with previous studies highlighting the p.I173N variant as the most common in SV-CAH, though the prevalence varies. Santos-Silva et al. [19] (Portugal, 2019) reported p.I173N in 46.9% of cases, while Zeng et al. [21] (China) found it in 40% of 33 SV-CAH children. Other studies from Costa Rica [20], Cyprus [25], Turkey [26], and Taiwan [22] confirm this variant predominance.
2) Rare variants
The p.R427C variant, first reported in a Russian girl with Prader IV virilization, showed no 21-OH enzyme activity in vitro. Diagnosed with 21-OHD at the age of 6 years, she exhibited signs of precocious puberty (pubic hair, bone age 13 years) and was compound heterozygous for p.I173N/p.R427C [27]. Our study also found this variant in a 7-year-old boy with rapid growth but no family history. His genotype, p.I173N/p.R427C, corresponded to the SV phenotype in group B. The p.R427C variant is rare, with only 4 reports in The Human Gene Mutation Database.
The p.R484W variant is associated with both SV and SW phenotypes. Jiang et al. [28] (2012) described a 14-year-old Chinese girl with severe virilization and 2.9% in vitro enzyme activity (clitoromegaly size of 2.0 cm×1.5 cm, amenorrhea, male bone structure, hirsutism, and acne). Her genotype was compound heterozygous p.R484W/E6 cluster. While p.R484W often correlates with the SV phenotype, it was first identified in a Tunisian person diagnosed with the SW form. Ono et al. [29] (2008) reported a 10-day-old girl with clitoromegaly, hyperpigmentation, and poor weight gain. Despite absence of overt SW symptoms, laboratory findings suggested mild SW, with in vitro 21-OH enzyme activity of 2.00%±0.25% and 1.89%±0.30% for progesterone to deoxycorticosterone and 17-OHP to 11-deoxycortisol conversion, respectively. We identified this variant in a 6-year-old boy with peripheral precocious puberty carrying the compound heterozygous genotype p. R484W/c.293-13C>G.
The p.L107R variant was first described by Soardi et al. [30] (2008) in 2 Brazilian siblings with compound heterozygous genotype p.L107R/p.I173N. The older brother, diagnosed at 37 months, presented with precocious puberty, while the younger sister, diagnosed at 3 months, exhibited clitoromegaly. Both siblings had the SV phenotype. Similarly, this variant was identified in a boy with the same compound heterozygous genotype p.L107R/p.I173N.
3) Genotype-phenotype correlation and positive predictive value
Among 43 children with pathogenic variants in the CYP21A2 gene, 38 had genotypes of 2 identifiable variants with known subgroups. Table 2 shows that 86.8% of these children had group B genotypes, 10.5% had group A, and 2.6% had group null. Details of the cases with the null/A genotype are presented in Table 5. Finkielstain et al. [31] (2011) studied 213 cases from 182 families and found a strong genotype-phenotype correlation. Of these, 110 had SW, 56 SV, and 47 nonclassic forms. Genetic analysis identified 188 variants in SW, 98 in SV, and 78 in nonclassic type. Phenotype prediction accuracy was high across groups: 88.9% for null, 91.5% for A, 85.1% for B, and 97.8% for C.
Chí Dũng et al [4] (2021) reported similar findings on genotype-phenotype correlation in 183 children with CAH. The positive predictive values for genotype groups null, A, B, and C were 99.8%, 96.5%, 90.6%, and 100%, respectively. Compared to Krone’s study [3] (2009), the positive predictive values for the null, A, B, and C groups were 100%, 90%, 74%, and 64.7%, respectively, decreasing from group null to group C. This trend reflects the more significant variability in clinical manifestations associated with less severe variants (group C). The study of Santos-Silva et al. [19] (Portugal) on 212 children across 13 centers reported positive predictive values of 85.3% for group null (n=34), 89.3% for group A (n=28), 80% for group B (n=25), and 95.9% for group C (n=123), with no significant difference between groups (P=0.08).
Since our cohort includes only SV patients, the positive predictive value could not be calculated for each genotype group. However, our results show that genotyping is crucial in guiding the treatment of CAH in children, particularly during the development of newborn screening programs. Early detection and diagnosis through newborn screening combined with genotype analysis allow SV children with group B genotype to safely discontinue fludrocortisone without frequent dose adjustments. This approach reduces costs, minimizes blood sampling, avoids unnecessary long-term medicine use, and lessens the side effects and stress for children and their families.
In conclusion, this study highlights the clinical and genetic spectra of patients with CAH due to 21-OHD at our center. Children with the SV-CAH are often diagnosed late, at a median age of 4.5 years. The B genotype group is observed in 86.8% of children with SV-CAH, serving as a significant predictor of the SV phenotype. Genetic analysis should be prioritized early, especially for children diagnosed through newborn screening programs. Genotype identification is very important to guide treatment for patients with SV-CAH, avoiding overtreatment, especially with the expansion of newborn screening in the world as well as in Vietnam.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement
Print
