Clinical analysis of 78 patients with nonclassical 21-hydroxylase deficiency

Retrospectively analyze the clinical characteristics of patients with nonclassical 21-hydroxylase deficiency (NC21OHD) as well as the relationship between the gene mutations and endocrine hormones. In addition, the relationship between different basal 17-hydroxyprogesterone (17OHP) levels and patients’ glucolipid metabolism, hormone levels, pregnancy, and treatment outcomes were examined. Clinical data of 78 females with NC21OHD from January 2012 to July 2022 in the Department of Endocrinology and Metabolism of the Third Affiliated Hospital of Guangzhou Medical University were retrospectively analyzed. Diagnosis was based on the 17OHP level combined with clinical manifestations, imaging, and other endocrine hormones and the cytochrome P450 c21, steroid 21-hydroxylase (CYP21A2) gene. The age at diagnosis of the 78 patients was 29.1 ± 4.2 years; 83.3% (65/78) of the patients had menstrual abnormalities, 70 patients were of childbearing age, and 97.1% (68/70) had a history of infertility with a median time of infertility of 3.6 years. Moreover, 71.8% (56/78) of the patients had polycystic ovaries, 26.9% (21/78) had hyperandrogenemia manifestations on physical examination, 66.7% (52/78) had adrenal hyperplasia, 32.1% (25/78) had combined dyslipidemia, and 41.0% (32/78) had combined insulin resistance. Pathogenic mutations were detected in 78.2% (61/78) of the patients with both CYP21A2 alleles; 14.1% (11/78) of the patients had only one allele and 7.7% (6/78) had no pathogenic mutations. The levels of total testosterone (TT), progesterone (P) (0 min, 30 min), and 17-OHP (0 min, 30 min, 60 min) in the adrenocorticotropic hormone (ACTH) stimulation test varied between the groups. Furthermore, patients with NC21OHD were divided into 17OHP < 2 ng/ml, 2 ng/ml < 17OHP < 10 ng/ml, and 17OHP ≥ 10 ng/ml groups according to their different basal 17OHP levels. The 17OHP ≥ 10 ng/ml group had significantly higher TT, FT4, basal and post-stimulation progesterone, and 17OHP, net value added of 17-hydroxyprogesterone (△17OHP), net value added of 17-hydroxyprogesterone/net value added of cortisol ratio (△17OHP/△F), the incidence of adrenal hyperplasia, and number of gene mutations compared to those of the 17OHP < 2 ng/ml group (P < 0.05). NC21OHD infertile patients who received low-dose glucocorticoids showed a significant increase in pregnancy and live birth rates, and a significant decrease in miscarriage rate (all P < 0.05). Comprehensive analysis is important as NCCAH diagnoses may be false positive or false negative based on clinical characteristics, hormone levels, and gene detection. Females with NC21OHD showed varying degrees of fertility decline; thus, low doses of glucocorticoid treatment for infertile females with NC21OHD can improve fertility and fertility outcomes.


Introduction
Congenital adrenocortical hyperplasia (CAH) is a group of autosomal recessive genetic disorders caused by mutations in genes encoding enzymes associated with cortisol biosynthesis [1]. A deficiency in multiple enzymes causes CAH, with a deficiency of 21 hydroxylase deficiency (21OHD) being the most common defect, accounting for 90-95% of CAH cases. According to the varying degrees of 21OHD, the clinical manifestations can be classified as classical and atypical types (retaining 20-70% of enzyme activity). Among them, the classic type can be subdivided into the classic salt-wasting type (SW) (0% enzyme activity) and the classic simple virilizing type (SV) (1-2% enzyme activity) [2].
Classic patients are generally diagnosed shortly after birth because of their significant clinical manifestations and biochemical abnormalities. Nonclassical congenital adrenocortical hyperplasia (NCCAH) is more than 10 times more likely to occur [3]. Due to the high activity of 21-hydroxylase residue, patients with NCCAH can synthesize sufficient amounts of glucocorticoids (COR) and mineralocorticoids (ALD) to maintain physiological needs. Therefore, these patients lack specific clinical manifestations or biochemical abnormalities, they are born with relatively normal external genitalia, and most are not diagnosed until puberty or adulthood [4,5]. Such patients often exhibit only mild clinical signs and symptoms, with elevated androgen levels, which can be manifested as menstrual abnormalities, hirsutism, acne, and infertility [6]. As a result, NCCAH is easily confused with other diseases that lead to hyperandrogenism, especially polycystic ovary syndrome (PCOS) [7,8]. NC21OHD is a disease to be excluded as a priority. Both NC21OHD and PCOS present with analogous clinical characteristics and high androgen levels and are both associated with obesity, insulin resistance, dyslipidemia, and subfertility [9,10]. PCOS is an exclusionary diagnosis, and NC21OHD needs to be definitively excluded from PCOS diagnosis. As the pathophysiological mechanisms and treatments of the two diseases are completely different, differential diagnosis is important.
NC21OHD is diagnosed by early follicular phase 17OHP measurement (basal and after ACTH stimulation) and gene detection. Basic serum levels of 17OHP below 2.0 ng/ml are generally considered to exclude NC21OHD, whereas levels above 10 ng/ml are diagnosable, and between which they can be confirmed by ACTH-stimulated 17 OHP levels [6,[11][12][13]. Although the efficacy of this diagnostic approach is supported by several studies, it is estimated that between 2 and 11% of patients with NC21OHD are missed based on using the basic 17OHP level < 2.0 ng/ml as the exclusion criteria [7,14,15]. Therefore, the baseline 17OHP threshold for screening for NC21OHD remains controversial [16][17][18]. NC21OHD is caused by mutations in the CYP21A2 gene, with more than 300 different mutant alleles identified thus far, all of which are listed in the Human Gene Mutation database (HGMD) (http:// www. hgmd. cf. ac. uk/ ac/ index. php). The frequency of common point mutations in CYP21A2 is distributed differently across different ethnicities, and the genotype-phenotype correlation varies greatly among different ethnic groups [19]. Therefore, it is important to investigate the CYP21A2 gene profile characteristics in different populations to understand the correlation between the genotype and phenotype. As yet, there has been no CYP21A2 gene spectrum study of adult patients with NC21OHD in southern China. Although most females with NCCAH can conceive naturally [20,21], 10% to 30% of the patients still have fertility problems [7,14]. There is no clear consensus on adult NC21OHD treatment, particularly in the reproductive and perinatal periods [22]. Annual physical examination and hormone measurements (morning 17OHP and androstenedione) are recommended as a minimum by the Endocrine Society, but no guidance is given about specific targets [23].
In this study, the clinical characteristics and endocrine hormones of 78 NC21OHD females were retrospectively analyzed and their reproductive outcomes were tracked.
All patients with NC21OHD were tested for the CYP21A2 gene using Sanger sequencing combined with multiplex ligation-dependent probe amplification (MLPA). The gene spectrum and frequency of CYP21A2 gene mutations in patients with NC21OHD in southern China are presented, and the relationship between the number of gene mutations and endocrine hormones was analyzed. Furthermore, the glucolipid metabolism, hormone levels, pregnancy, and treatment outcomes were compared according to different basal 17OHP levels. Finally, the improvement of fertility (live birth rate, abortion rate) in these patients treated with low-dose glucocorticoids was discussed.

Study population
A total of 81 patients with NC21OHD diagnosed in the Department of Endocrinology and Metabolism of the Third Affiliated Hospital of Guangzhou Medical University from January 2012 to April 2022 were analyzed retrospectively, including three males and 78 females. The data of 78 female patients were analyzed.
The clinical diagnosis of NC21OHD was made mainly based on clinical characteristics, hormonal tests, and imaging. The clinical diagnostic standard of NC21OHD is as follows [11]: (1) menstrual disorders or oligomenorrhea, primary/secondary amenorrhea, or hyperandrogenemia (e.g., hirsutism, acne, clitoral hypertrophy, and alopecia), or infertility in adolescence or adulthood; (2) gynecologic ultrasound indicates polycystic ovarian changes; (3) adrenal CT/MR shows adrenal hyperplasia; (4) progesterone, testosterone, and dehydroepiandrosterone sulfate (DHEAS) may be elevated, ACTH is normal or elevated to varying degrees, and normal or mildly decreased cortisol; and (5) 17OHP collected early in the morning and during the follicular phase, basal 17OHP or ACTH-stimulated 17OHP > 10 ng/ml. Point 5 represents the essential conditions for clinical diagnosis, while points 1-4 represent non-essential conditions. Sanger sequencing combined with MLPA was used for CYP21A2 gene detection in all patients with NC21OHD.
This study was approved by the Ethics Committee of the Third Affiliated Hospital of Guangzhou Medical University (approval number 2022103). Informed consent was obtained from all individual participants included in the study.

Study design
Wit h 'nonclassical 21-hydroxylase def iciency (NC21OHD)' as keyword, using the electronic medical record system of The Third Affiliated Hospital of Guangzhou Medical University, NC21OHD disease cases diagnosed in the Endocrinology and Metabolism Department from January 2012 to April 2022 were screened, and male patients were excluded. To retrospectively analyze the clinical characteristics and endocrine hormone levels of clinically diagnosed female patients according to different number of CYP21A2 gene detection, clinical characteristics, metabolic indices, imaging and hormone levels were further compared and analyzed according to different basal 17OHP levels in the patients who were confirmed by clinical and genetic diagnosis (two mutations detected). Finally, fertility, pregnancy outcome, and fetal status of clinically diagnosed infertility patients were retrospectively analyzed (Fig. 1).

Treatment and follow-up
Females with NC21OHD with primary/secondary infertility who wished to reproduce were treated with dexamethasone before pregnancy at doses ranging from 0.25 to 0.75 mg qd, titrated according to androgen levels. Dexamethasone was discontinued during pregnancy, and most patients were treated with hydrocortisone acetate at a dose of 5-20 mg/ day. For patients with insulin resistance or hyperandrogenism, metformin 1.0-1.5 g/day and spironolactone 20 mg bid-tid were given according to the clinical routine. For patients with subclinical hypothyroidism, LT4 treatment was administered if necessary to maintain normal TSH before and during pregnancy. The pregnancy rate, singlelive birth rate, and abortion rate were compared before and after treatment.
We followed up the fertility and pregnancy outcomes by recording the history of pregnancy and delivery (abortion, preterm delivery, stillbirth, ectopic pregnancy, singleton/ multiple live births); mode of conception; age at delivery; gestational week of delivery; mode of delivery; pregnancy outcome and complications (singleton live birth, gestational diabetes (GDM), placental abruption, premature rupture of membranes, amniotic fluid, cesarean section, preeclampsia, preterm delivery, and twin and multiple births); before and after diagnosis and treatment; and neonatal status (newborn sex, birth weight, and birth length).

Population characteristics
Detailed records of the patient's age, age at menarche, age at diagnosis, menstrual history of marriage, and family history were taken. All patients' anthropometry and laboratory indices were recorded. Anthropometry included height, weight, waist and hip circumference, and blood pressure. During the physical examination, hair distribution, the presence or absence of acne, androgenic alopecia, and acanthosis nigricans and exterior genital development were recorded. Biochemistry and follicular phase hormonal parameters were sampled during the follicular phase of the menstrual cycle.

ACTH stimulation tests
The test was performed in the early morning of the follicular phase. The subjects were given an intravenous injection of 25 U ACTH (manufacturer: Shanghai First Biochemical Pharmaceutical Co., Ltd., specification: 25 U/branch), and it took 0, 30, and 60 min to determine the levels of 17OHP, plasma cortisol, and progesterone, respectively.

Imaging
Gynecological ultrasound scan of the uterus and ovaries, and CT/MR examination of bilateral adrenal glands.

Mutation analysis of the CYP21A2 gene
Peripheral blood was collected and sent to the Guangzhou Kingmed Center Clinical Laboratory for CYP21A2 gene detection. Sanger sequencing and MLPA were used. The sequencing results were analyzed using Sequence Scanner v 1.0 software (Applied Biosystems, Inc., USA). The sequencing results were compared with the CYP21A2 gene reference sequences NM_000500.7 and NG_007941.2 to determine possible gene mutations.

Statistical analysis
Statistical analyses were performed using SPSS 20.0 for Windows (IBM, Armonk, NY, USA). The Shapiro-Wilk method was used to test whether the data were normally distributed. Continuous data with a normal distribution are presented as (X ± S), and the measurement data are expressed as the median and interquartile spacing [M (P25, P75)]. Count data are expressed as frequencies and percentages. Mann-Whitney U test was used to compare the differences between the two groups; the Kruskal-Wallis H test was used to compare the differences between the three groups. The 81 patients with NC21OHD who met the criteria for clinical diagnosis were selected by the electronic medical record systems.
The clinical characteristics and endocrine hormone of 78 female patients were retrospectively analyzed according to having 2 (group A, n=61), 1 (group B, n=11) and 0 (group C, n=6) mutations of CYP21A2 gene.

males excluded.
Group A patients(n=61): divided into three groups group A1 n=9; group A2, n=27; group A3 n=25;) according to different basal 17OHP levels, compared the clinical and metabolic characteristics of each group.
10 patients of non-fertility age or no fertility requirements excluded.
The treatment follow-up, pregnancy outcome and fetal condition of 68 patients were retrospectively analyzed.
Group B and Group C totally 17 patients Fig. 1 Flowchart of patient recruitment detection of CYP21A2 gene mutation in different clinical phenotypes was analyzed by the Pearson χ 2 test. All were bilateral tests, and P values < 0.05 were considered statistically significant.

Hormonal findings
The patients were classified into the following three subgroups according to having 2, 1, and 0 mutations of CYP21A2 gene: A (n = 61), B (n = 11), and C (n = 6). The levels of TT, P (0 min, 30 min) and ACTH-stimulated 17OHP (0 min, 30 min, 60 min) in the three groups were statistically different, among which group A had higher levels than groups B and C (all P < 0.05), while the differences in the remaining hormone levels showed no significant differences between groups (Table 1).

Fertility, pregnancy outcome, and fetal status
Of the 61 patients with NC21OHD confirmed by clinical and genetic diagnosis, 56 patients were diagnosed with primary or secondary infertility in their reproductive years; these patients were treated simultaneously in our fertility center and were actively preparing for pregnancy. Thirty-two patients had primary infertility and 24 had secondary infertility. Among the 24 pregnancies before diagnosis/treatment, the singleton live birth rate was only 12.5% (3/24) and the abortion rate was 83.3% (20/24); after diagnosis and treatment, 51 of the 56 females had become pregnant 58 times, that is, the pregnancy rate was 91.1% (51/56). Among them, the singleton live birth rate was 84.5% (49/58) and the abortion rate decreased to 12.1% (7/58). After low-dose glucocorticoid treatment, the singleton live birth rate increased significantly and the abortion rate decreased significantly (Table 3).
Of the 49 live births after NC21OHD diagnosis and treatment, 38.8% (19/49) were naturally conceived, 36.7% (18/49) were pregnant after ovulation induction, and 24.5% (12/49) were pregnant by in vitro fertilization-embryo transfer. Regarding the delivery mode, 63.3% (31/49) had a vaginal delivery and 26.5% (13/49) had a cesarean section. Regarding gestational complications, gestational diabetes mellitus was present in 34.7% (17/49); premature delivery was present in 14.3% (7/49); subclinical hypothyroidism was present in 10.2% (5/49); too little amniotic fluid was present in 6.1% (3/49); and preeclampsia and placental abruption each accounted for 2.0% (1/49). Among the 49 live births, the mean maternal age at delivery was 30.9 ± 3.4 years; the mean week of gestation at delivery was 38.3 ± 1.8W; the mean birth weight was 3093.5 ± 542.9 g; for boys, the mean length was 49.6 ± 3.70 cm (n = 26) and the mean birth weight was 2783.9 ± 289.0 g. For girls, the mean length was 48.56 ± 1.94 cm (n = 23). The difference in birth weight and length between male and female newborns was not statistically significant (P = 0.055). One male fetus was giant (4350 g); the mother's mode of conception was in vitro fertilization and embryo transfer (IVF-ET), and gestational diabetes was combined during pregnancy.
The neonatal score was not statistically significant (P = 0.440), and no offspring were found with confirmed 21-OHD. Among the 17 female patients with NC21OHD who met the clinical diagnosis but had an unconfirmed genetic diagnosis, five had no need for pregnancy and the remaining 12 were primary/secondary infertile patients. Glucocorticoid therapy was still administered and 11 had 14 pregnancies with a live birth rate of 85.7% (12/14) and an abortion rate of 14.3% (2/14). Table 1 Comparison of hormones between three groups according to having 2, 1, and 0 mutations of CYP21A2 gene *: P < 0.05 refers to statistical differences between the three groups a P < 0.05 refers to statistical differences between group A and group C b P < 0.05 refers to statistical differences between group B and group C

Summary of main results
This study is retrospective study of the clinical characteristics of NC21OHD female patients. We found that the NC21OHD is a heterogeneous disease, often confused with PCOS, and related to infertility. A comprehensive analysis is extremely important for the improvement of infertility in such patients. Low-dose steroid therapy may be beneficial to improve fertility.

Discussion of relevant results and interpretation
PCOS is the main cause of hyperandrogenism, accounting for 50 to 80% of such patients [24]. It is a heterogeneous hyperandrogenic disorder probably of genetic origin, but affected by environmental factors [24]. It is an exclusive diagnosis [25]. NC21OHD is one of the diseases that needs to be excluded, but there are many similarities in clinical manifestations between them. It has been reported [24] that there are no significant differences in age, BMI, WHR, hirsutism incidence, and other clinical features between two diseases. We try to make a summary table of clinical differences to distinguish them from the review articles published in recent years [6,11] on the identification of the two ( Table 4). The prevalence for NCAH due to 21-hydroxylase deficiency has been estimated to range between 1:30 and 1:1000 [26,27]. Among women ascertained by symptoms due to androgen excess, the reported prevalence for NCAH ranges from 0.6 to 9% [28]. A meta-analysis published in 2017 [6], which included all studies published between 1980 and 2015 on screening for women ascertained by symptoms due to hyperandrogenism, estimated the global prevalence of NC21OHD to be approximately 4.2% (95% confidence interval: 3.2-5.4%). However, the exact frequency for NC21OHD has been difficult to accurately establish due to incomplete ascertainment of all cases. NC21OHD can be considered one of the most common autosomal recessive disorders in humans [29], with a similar incidence in males and females. Our study collected 81 patients with NC21OHD, which had a sex disparity of 3/78, similar to the existing literature. Few studies have examined men with NCCAH, which may be related to men not seeking medical help for androgen excess [4,28].The clinical manifestations of NC21OHD in adult females mainly include hypothalamic-pituitary-ovarian dysfunction and hyperandrogenemia (hirsutism, acne, and/ or androgenetic alopecia) [30]. Some studies [7,31] have shown that more than 30-50% of patients with NC21OHD have significant ovulation and menstrual disturbances and often have polycystic ovarian changes (PCOM) [32]. In this study, menstrual disorders (> 80%) and PCOM (> 70%) were found to be common in this group of patients, but the clinical manifestations of hyperandrogenism, such as hirsutism, were not as common as that reported in other studies and were presumably related to the differences in hair growth between ethnic groups. Asian females have relatively less hair, and a m-FG score > 6 as a cutoff point for hirsutism may not be suitable for Asian women, and the clinical assessment of hirsutism score is relatively subjective and generally reliable. Infertility continues to be the leading cause of hospital visits for most patients with NC21OHD, although some data suggest that some patients with NC21OHD have only mildly impaired fertility, not necessarily infertility [13,20,31]. Furthermore, some reports [14,20,21] have found that 10%-30% of females with NCCAH have infertility complaints. In our study, we found a high proportion of patients with primary/secondary infertility, which may be because our center focuses mainly on reproductive endocrine disease, as well as the low demand for medical services from some patients whose fertility is not impaired. Clitoromegaly has been reported in 6-20% of adult women with NCAH [14]. Other clinical symptoms may include anatomical abnormalities of the adrenal cortex (e.g., adrenal hyperplasia) [30] and metabolic disorders (e.g. insulin resistance, disorders of glucolipid metabolism) [31], as well as an increased risk of cardiovascular disease [32]. This study also found a high incidence of glycolipid metabolism disorders and overweight/obesity.
In conclusion, patients with NC21OHD and PCOS have similar clinical presentation and overlapping imaging, making it virtually impossible to distinguish between the two disorders solely on clinical grounds; therefore, the diagnosis relies mainly on early follicular phase 17OHP assay (basal and after ACTH stimulation) and gene detection. In our study, patients with clinically diagnosed NC21OHD were genotyped using a combination of Sanger sequencing and MLPA technology, the most commonly used gold standard for detection of CYP21A2 genotypes [33,34] and copy number variants. As an autosomal recessive genetic disease, 21OHD can be genetically diagnosed only when Table 3 Comparison of pregnancy outcomes before and after diagnosis/treatment of NC21OHD patients confirmed by clinical and genetic diagnosis a, b significant difference between groups (P = 0.00) Pregnancy outcome Before diagnosis/ treatment (n = 24) After diagnosis/ treatment (n = 58) P N (%) N (%) Single live births 3 (12.5%) 49 (84.5% a ) 0.00 Twin live births 0 (0%) 2 (3.84%) -Abortion 20 (84.5% b ) 7 (12.1% b ) 0.00 Ectopic pregnancy 1 (4.16%) 0 (0%) pathogenic mutations are detected in both CYP21A2 alleles [33]. Most individuals with NCCAH are compound heterozygotes, and the phenotype roughly reflects the enzymatic activity encoded by mild mutations [35][36][37]. Compound heterozygotes were detected in 78.2% (61/78) of the patients in this study, and no pathogenic mutations were detected in six patients, with 24 mutation types detected. The two most common mutation types, c.-126C > T (20.59%) and c.-113G > A (19.12%), were located in the promoter region, where they were the most popular variants, reducing 21 hydroxylase activity by up to 20%. These mutations are generally associated with an atypical phenotype. Additionally, the most common mutation types in the coding regions were c.293-13C > G (13.24%) and c.518 T > A (10.9%), which differed from the most commonly reported variants [6] (c.92C > T, c.844G > T). Although c.293-13C > G (I2G) and c.518 T > A (V281L) are the typical mutation loci, the phenotype remains atypical as most patients are compound heterozygotes. The failure to detect pathogenic mutations in two alleles may be due to the mutation site being located in the intron region leading to a false-negative genetic diagnosis. The role of intron region mutations in disease pathogenesis still needs to be further investigated; however, some rare types of CAH can also exhibit clinically similar clinical phenotypes to 21OHD and elevated 17OHP levels, such as 11β-hydroxylase deficiency (11β-OHD) [38] and P450 oxidoreductase deficiency (PORD) [39]. For those with no CYP21A2 mutations detected, further 11β-OHD/POR genetic testing can be performed for differential diagnosis. Our study further compared 61 patients with NCCAH who met both clinical and genetic diagnoses, grouped according to baseline 17OHP levels and found that 41.0% (25/61) of patients had baseline 17OHP > 10 ng/ml, 44.3% (27/61) of patients had baseline 17OHP < 10 ng/ml, and 14.7% (9/61) of patients had baseline 17OHP > 10 ng/ml. If a baseline 17OHP level < 2.0 ng/ml had been used as the exclusion criterion [7,15], 14.7% of patients would have been missed in this study. Therefore, the baseline 17OHP threshold remains controversial. At present, 17OHP is commonly detected by immunoassay, but its sensitivity, specificity, and matrix effect are insufficient and its reliability is poor. Although liquid chromatography combined with mass spectrometry (LC/MS) improves the positive predictive value of the first screening [40], it has not been widely applied in clinical practice due to its high cost and technical complexity. Therefore, it is recommended that each laboratory establishes internal thresholds or indicators that can improve screening sensitivity and specificity. Therefore, in patients with basal 17OHP < 2 ng/ml, it is necessary to integrate clinical manifestation and, if necessary, further improve the examination, including ACTH excitation test and genetic diagnosis, to maximize early diagnosis and avoid missed diagnosis and misdiagnosis. Additionally, we analyzed the effect of different basal 17OHP levels on metabolic and hormonal indicators in this cohort of patients. In terms of metabolism, although hypercholesterolemia is common in patients with NC21OHD, few studies have reported on whether elevated cholesterol is consistent among patients with different baseline levels of 17OHP. Our study found significantly elevated TC levels in patients with basal 17OHP meeting diagnostic criteria compared to those with basal 17OHP < 2 ng/ml. In terms of hormonal indices, FT4, TSH, TT, LH, LH/FSH, P after ACTH stimulation, 17OHP, and Δ17OHP/ΔF were significantly higher in patients with basal 17OHP > 10 ng/ml, while cortisol upregulation after ACTH stimulation (30, 60 min) was significantly lower in patients with baseline 17OHP < 2 ng/ml. The imaging results suggested that the incidence of adrenal hyperplasia was significantly higher in patients with 17OHP > 10 ng/ml. These results suggest that in patients with NC21OHD, the higher the basal 17OHP level, the lower the retention of the 21-hydroxylase activity, and the greater the effect on the hypothalamic-pituitary axis. Thus, the detection rate of CYP21A2 mutations may not reach 100% in patients with clinically diagnosed 21OHD, but clinical and laboratory tests alone may also result in misdiagnosis of 21OHD. Therefore, it is important to combine clinical data with genetic testing, regular follow-up, and comprehensive judgment.
In accordance with the 2018 American College of Endocrinology Clinical Practice Guidelines for patients with CAH, which recommend that adult females with NCCAH with severe hyperandrogenism or infertility require glucocorticoid therapy [12]. Women with NC21OHD who have hyperandrogenemia manifestations but do not have fertility requirements can be treated with oral contraceptives, along with spironolactone if necessary. However, these drugs should not be used during pregnancy [30]. For patients with fertility requirements, some patients may resume ovulation and natural pregnancy after treatment with glucocorticoid alone; however, those who do not ovulate may be given ovulation-promoting therapy or other assisted fertility techniques based on glucocorticoid. In infertile patients with NC21OHD, dexamethasone is used to induce fertility before pregnancy. The goal of treatment is not glucocorticoid replacement [41], but rather to effectively suppress early morning ACTH and 17OHP concentrations to correct irregular menstrual cycles and other symptoms of adrenal androgen excess, such as acne, hirsutism, and infertility. Furthermore, serum 17OHP and androstenedione (A4) are the most commonly used indicators to evaluate the adequacy of glucocorticoid treatment in CAH, and 17OHP and androsterone levels in treated patients are generally recommended to remain above normal or slightly elevated [12]. In this study, dexamethasone was administered to patients of childbearing age with a diagnosis of NC21OHD combined with primary/secondary infertility at a dose of 0.25 to 1.0 mg per night during the preconception stage. We adjusted the dose of dexamethasone according to the testosterone levels and found that the treatment significantly improved pregnancy rates. Dexamethasone is not metabolized by placental 11β-hydroxysteroid dehydrogenase type II, which inhibits the fetal adrenal glands and may have long-term adverse effects on fetal mental development [42]. Therefore, dexamethasone should be avoided during pregnancy. There are still no clear guidelines for post-pregnancy glucocorticoid treatment for patients with NC21OHD. In a retrospective study [15], glucocorticoid treatment did not significantly affect the frequency of miscarriage. Conversely, two other studies suggested an increased risk of miscarriage in patients with NC21OHD, with a significantly lower rate of early miscarriage after treatment with glucocorticoid compared to untreated patients [20,21]. Despite the conflicting results, glucocorticoid treatment during pregnancy is often recommended [6]. Hydrocortisone or prednisone is recommended during pregnancy, both of which are easily metabolized by placental 11β-hydroxysteroid dehydrogenase type 2, with low doses having no clear effects on the fetus. In this study, 83.3% of females of childbearing age who were diagnosed with NC21OHD had miscarriages before glucocorticoid treatment. These patients were treated with low doses of dexamethasone in preparation for pregnancy, and most switched to low doses of hydrocortisone (5-7.5 mg/d) during pregnancy, while A4 levels were regularly tracked during normal deliveries. Moreover, the results showed that there was a significant increase in the singleton live birth rate (85.7%) and a significant decrease in the miscarriage rate (14.3%) of these patients. Generally speaking, the outcome of the children is excellent [43]. Although female fetuses from CAH mother may develop ambiguous genitalia [44], no similar cases were found in this study. NC21OHD patients were still recommended for standardized follow-up and genetic counseling for both husband and wife in order to detect fetal abnormalities early. Finally, it is interesting that, in general, women with NCCAH are more likely to give birth to female children [15,45], but this study presents an opposite answer, with slightly more male fetuses, for unknown reasons.

Limitations
We have only investigated the clinical features and treatment efficacy of NC21OHD female patients in southern China, with a limited sample size and only a single-center findings. Further studies should be considered in combination with multi-centers across China to provide clinicians with evidence for accurate diagnosis and treatment of NC21OHD.

Conclusions
Taken together, although NC21OHD is relatively common, most cases have never been diagnosed, especially in men. Moreover, the clinical manifestations of NC21OHD are diverse and lack specificity; indeed, adolescent and adult females are often misdiagnosed with PCOS. Therefore, it is necessary to improve the diagnosis of NC21OHD in patients suspected to have PCOS who are infertile at childbearing age. However, the diagnosis remains difficult. Relying on clinical manifestations, hormone levels, and gene detection may lead to a false-positive or false-negative diagnosis of NCCAH; therefore, a comprehensive analysis is needed. Moreover, further research is needed to find a better diagnostic threshold for basic 17OHP and to explore more effective hormone indicators for testing. Additionally, to our knowledge, the study is a large sample study on the diagnosis, treatment, follow-up pregnancy outcomes of Chinese infertile NC21OHD patients. We found that females with NC21OHD often exhibit varying degrees of fertility decline. Thus, low-dose glucocorticoid therapy is recommended for infertile females with NC21OHD to improve fertility and fertility outcomes to some extent.