This study summarized the clinical and genetic characteristics of 126 patients, using CHH and DSD phenotypes (micropenis and cryptorchidism) and genetic testing as important cues for pediatric diagnosis of CHH. With an increase in the number of cases, more CHH candidate genes were confirmed in patients. We found that digenic and trigenic variants accounted for 24.2% (23/95) and 3.2% (3/95) of patients, respectively, who underwent genetic testing.
The proportion of gene mutations found in patients in this study was significantly higher than that in patients with CHH reported in the literature. The most common mutant genes were FGFR1 (21.1%), PROKR2 (17.9%), ANOS1 (12.6%), and CHD7 (12.6%) [6, 7, 18–20]. Among them, 10 cases (50.0%) of FGFR1 and 7 cases (41.2%) of PROKR2 mutations were oligogenic mutations. Previous studies reported that FGFR1 mutation may be related to hand and foot malformation in patients with CHH [19]; however, in this study, only one patient with hand and foot malformation had FGFR1 mutation, and there was no obvious phenotypic-genotypic correlation, which may be because more than half of the patients were oligogenic.
Previous literature on HH in adults reported that patients with FGFR1 mutations had a high incidence of cryptorchidism, small testicular size, long treatment time for spermatogenesis, and low sperm concentration [21]. However, in this study, compared with patients with non-FGFR1 mutations, the incidence of cryptorchidism in patients with FGFR1 mutation was not significantly higher, the testicular volume was not small, and the patients with FGFR1 mutations had mainly pure CHH. However, because the diagnosis of pure CHH in this study was based on the T level after the hCG prolongation test, which may be the cause of the difference, further follow-up studies will be conducted.
INHB is a marker of the number of Sertoli cells, and is usually lower than 30 pg/ml in male patients with complete CHH [1, 22]. In some patients with partial CHH, the level of INHB may overlap with that of DP and healthy controls [23, 24]. In this study, the INHB test was performed in 65 patients, including all age groups, of which 45 (69.2%) had an INHB level of > 30 pg/ml, suggesting that the function of testicular Sertoli cells in these patients was still good. After subsequent therapy with GnRH pump or HCG/HMG, it was more likely to promote spermatogenesis, which was consistent with the current spermatogenesis rate of 64–80.3% in CHH patients after treatment [25, 26]. Among the 39 patients with T > 100 ng/dl after the hCG prolongation test, there were still 10 patients whose INHB level was less than 30 pg/ml, suggesting that some of the patients with good response to Leydig cells may still have had poor function of Sertoli cells. Therefore, the evaluation of testicular cell function in patients with HH requires a multi-faceted and multi-index comprehensive evaluation.
However, in this study, one patient with three gene mutations (SEMA3E/CHD7/NSMF) was also diagnosed with dual CHH. The patient was treated with a standard GnRH pump for 6 months, the level of T was 112 ng/dl, and 12 months later, it was observed that the patient had spermatorrhea and good sperm motility and concentration. Therefore, the percentage of patients diagnosed with dual CHH in this study with restored fertility after treatment needs to be further studied. This phenomenon also suggests that there was a false negative response to short-term stimulation in the experiment.
In previous studies of adult cases, patients whose puberty was not induced by hormone therapy at the age of puberty usually showed infertility when GnRH pump therapy was used to stimulate spermatogenic potential in adults. Therefore, for CHH patients and their families, every gene mutation that causes GnRH deficiency should theoretically not be transmitted within the family. Recent studies have shown that a small number of gene mutation sites have strikingly high percentage, such as GNRHR Q106R (44%), GNRHR R262Q (29%) [6, 27–29] and TACR3 W275X (36%) [30–32]. The L173R of PROKR2 accounts for 40% of the CHH population in Europe and the United States, but it is rare in the Asian population [6, 33–36]. The recurrent mutation sites of several genes in this study were PROKR2 (p. W178S, n = 5 in KS and nHH, respectively), ANOS1 (p. V560I, n = 2 in KS), HS6ST1 (p. H63D, n = 3 in KS), and IL17RD (p. P191L and p. S671L, n = 2 in KS, respectively). It was proven that PROKR2 was one of the most common pathogenic genes in CHH, accounting for 17.9% (17/95) in this study, of which W178S accounted for 58.8% (10/17). In another study of Chinese adult CHH patients, PROKR2 mutations accounted for 13.3% (18/135) and W178S accounted for 55.6% (10/18) [37]. Combined with two studies, W178S accounted for 57.1% of PROKR2 mutations in the Chinese CHH population (20/35). Functional analysis showed that the mutant impeded receptor was expressed on the cell surface. The W178S of PROKR2 may be an ancient founder mutation, and it was not eliminated in the Chinese CHH population during evolution. The silencing of its effect on reproduction may be related to oligogenicity, or the mutant may revert during adulthood. However, in this study, the mother of one patient harboring W178S in PROKR2 was a proband, suggesting that the mutation had a wide spectrum and that the patient could undergo germ cell maturation after treatment. Therefore, the complex mechanism of its effect on reproduction needs to be further studied.
Of the five patients with single-gene mutations in PROKR2 (W178S), four were diagnosed with dual CHH. Some patients carried additional gene mutations. One patient with a repetitive mutation site HS6ST1 (H63D) and another with IL17RD (N503S) were also diagnosed with dual CHH, and PROKR2 (W178S) and IL17RD (N503S) were carried by the proband mother simultaneously; however, the results of the HCG test showed that the Leydig cells were dysfunctional. It has been suggested that some patients with dual CHH diagnosed by the hCG test in this study may still recover their reproductive function after standard treatment, and it is possible that Leydig cells had not been stimulated by GnRH for a long time, which may have led to the slow response of the receptor. The curative effects in the patients with the above mutations and the relationship between the curative effect and gene mutations will be further investigated. However, the FGFR1 mutation with the highest proportion of mutations was not found to be recurrent in patients with CHH, suggesting that the mutations of the gene had a greater impact on reproduction. Among the 126 patients in this study, three probands were mothers, and no probands were fathers. It has been reported that 64–80.3% of male CHH patients could produce sperms after GnRH pump or HCG/HMG therapy, but the specific situation may need further observation.
Hypospadias is caused by an abnormal urethral opening closure in the early stage of embryonic development, including the early hormone-independent stage (5–8 weeks) and hormone-dependent stage (8–12 weeks) [38]. There was no significant intersection with the functional time of the HPG axis. Therefore, hypospadias and CHH may be two unrelated diseases. The 2015 CHH consensus also believes that the existence of the hypospadias phenotype can exclude the diagnosis of hypospadias [1].
In addition to our previous report, other HH patients with the hypospadias phenotype have also been reported; for example, Huseyin et al. reported that a 15-year-old boy had no puberty initiation, micropenis, cryptorchidism, and perineal scrotal hypospadias, and his elder sister had no secondary sexual development. Sequencing showed that they carried KISS1R homozygous mutations (p. Y323X) [39]. In 2017, Ji et al. reported that a 28-year-old KS patient presented with no signs of puberty, bilateral cryptorchidism, olfactory loss, and right deafness. Genetic tests found that in addition to nonpathogenic variants (rs808119, rs6185) of ANOS1 and GNRH1, SRD5A2 gene heterozygous variation (c.680G > A) was also present, and T/DHT was 26.40, after the hCG test; thus, it was considered that the patient had KS and 5 α-RD [40]. In Indonesia, 11 patients with 46, XY DSDs were reported to carry pathogenic mutations, including PROKR2, PROK2, WDR11, FGFR1, and CHD7 mutations, and these patients had different degrees of hypospadias [41]. Other studies have found that FGF8, GLI3, CHD7, and other CHH-related genes could cause hypospadias 38. These studies suggest that hypospadias and CHH are not completely separated, and the potential relationship between hypospadias and CHH needs to be further elucidated. There were seven patients with hypospadias in this study, but five carried no hypospadias related genes, one with KS had SALL1 gene mutation (p. M662V), and one with nHH had a SPECClL mutation (p. M232V). SALL1 and SPECC1L gene mutations could cause hypospadias, suggesting that the combined gene mutation may be one of the causes of hypospadias in patients with CHH, but the mechanism of hypospadias in other patients is still not clear.