Reproductive Phenotypes and Genotypes in Men With IHH

Abstract

Context

Isolated hypogonadotropic hypogonadism (IHH) is phenotypically and genetically heterogeneous.

Objective

This work aimed to determine the correlation between genotypic severity with pubertal and neuroendocrine phenotypes in IHH men.

Methods

A retrospective study was conducted (1980-2020) examining olfaction (Kallmann syndrome [KS] vs normosmic IHH [nHH]), baseline testicular volume (absent vs partial puberty), neuroendocrine profiling (pulsatile vs apulsatile luteinizing hormone [LH] secretion), and genetic variants in 62 IHH-associated genes through exome sequencing (ES).

Results

In total, 242 men (KS: n = 131 [54%], nHH: n = 111 [46%]) were included. Men with absent puberty had significantly lower gonadotropin levels (P < .001) and were more likely to have undetectable LH (P < .001). Logistic regression showed partial puberty as a statistically significant predictor of pulsatile LH secretion (R² = 0.71, P < .001, OR: 10.8; 95% CI, 3.6-38.6). Serum LH of 2.10 IU/L had a 95% true positive rate for predicting LH pulsatility. Genetic analyses in 204 of 242 IHH men with ES data available revealed 36 of 204 (18%) men carried protein-truncating variants (PTVs) in 12 IHH genes. Men with absent puberty and apulsatile LH were enriched for oligogenic PTVs (P < .001), with variants in ANOS1 being the predominant PTV in this genotype-phenotype association. Men with absent puberty were enriched for ANOS1 PTVs compared to partial puberty counterparts (P = .002). PTVs in other IHH genes imparted more variable reproductive phenotypic severity.

Conclusion

Partial puberty and LH greater than or equal to 2.10 IU/L are proxies for pulsatile LH secretion. ANOS1 PTVs confer severe reproductive phenotypes. Variable phenotypic severity in the face of severe genetic variants in other IHH genes point to significant neuroendocrine plasticity of the HPG axis in IHH men.

Isolated hypogonadotropic hypogonadism (IHH) is a rare endocrine disorder resulting from deficient secretion and/or action of gonadotropin-releasing hormone (GnRH) and presents clinically as absent or partial pubertal development and infertility (1, 2). IHH is phenotypically heterogeneous in terms of both reproductive phenotypes and a wide range of IHH-associated nonreproductive phenotypes. The phenotypic variability of IHH is paralleled by genotypic heterogeneity, with more than 60 different genes having been identified to underlie IHH, accounting for approximately half of cases (3). There is considerable allelic heterogeneity within IHH genes including variant zygosity (heterozygous vs homozygous vs hemizygous), variant type (protein truncating vs non–protein truncating), and variant location within the gene/protein. Further, many individuals with IHH harbor genetic variants in more than one IHH gene (oligogenicity) and such aggregated mutation burden has been postulated to contribute to the phenotypic heterogeneity (1, 2). Given the genetic complexity, a key challenge for genomic medicine for patients with IHH is to discern precise genotype-phenotype correlations that will inform clinical care and genetic counseling.

Prior genotype-phenotype studies in IHH have sought to identify clinical commonalities among patients with rare variants in a specific gene. However, a substantial majority of these studies have defined genotype-phenotype correlations that primarily relate to nonreproductive features for specific IHH genes (eg, dental/skeletal anomalies and midline defects [FGFR1], sensorineural hearing loss [SOX10, CHD7, IL17RD], and unilateral renal agenesis [ANOS1]) (1, 2, 4). In contrast, genotype-phenotype correlations that specifically relate to the severity of reproductive phenotypes in IHH are sparse and have typically been examined in only a handful of genes. To address this gap, in this report we studied a large cohort of well-characterized men with IHH to examine whether the severity of reproductive phenotype correlates with underlying genotypic severity. We hypothesized that severe genetic defects, defined by protein truncating variants (PTVs) in IHH genes, will result in more severe hypothalamo-pituitary-gonadal (HPG) deficits. We reasoned that understanding the relationship between genetic severity and reproductive phenotypes could provide further insights into the molecular pathomechanisms underlying GnRH deficiency. In addition, if genetic burden and/or genotypic severity correlates with severity of reproductive phenotypes, IHH individuals with severe reproductive dysfunction could be specifically prioritized for future genetic studies to uncover the missing genetic etiology.

Reproductive phenotypes in IHH result from defects in hypothalamic GnRH secretion causing downstream failure of pituitary gonadotropin secretion and consequently, impaired gonadal function. Although IHH can affect both sexes, the prevalence of IHH is more common in men and reproductive phenotypic severity relating to puberty (as determined by testicular volume, TV) is more readily discernible in males (1, 2). Hence, in this report, we specifically restricted our genotype-phenotype analysis to IHH males. Pubertal phenotypes in males with IHH can either be severe, with complete failure to initiate puberty (ie, absent puberty with prepubertal TV, < 4 mL) (5, 6), or milder with partial spontaneous puberty (ie, TV > 4 mL with subsequent stalling of pubertal progression). Biochemical features in IHH patients can also be used to characterize the severity of HPG defects in individuals with IHH. Baseline serum luteinizing hormone (LH) and follicle-stimulating hormone (FSH) levels can be used as quantitative measures of biochemical severity. Likewise, dynamic biochemical studies using every-10-minutes sampling over 12 to 24 hours have been used to determine GnRH-induced LH secretion patterns (pulsatile vs apulsatile patterns) (7). To address our hypothesis, we first defined the following reproductive phenotypic severity frameworks using both clinical and biochemical data: (a) pubertal severity (absent vs partial); (b) baseline biochemical values (undetectable vs detectable gonadotropins); (c) dynamic LH pulse studies (apulsatile vs pulsatile); and d) integrated clinical and biochemical features (eg, absent puberty + apulsatile LH vs partial puberty + pulsatile LH). Genotype-phenotype correlations were then examined across all the aforementioned realms.

Materials and Methods

This study was reviewed and approved by the Mass General-Brigham institutional review board and all participants provided written informed consent prior to initiation of study-related procedures.

Participants

Eligible individuals were adult men (aged 18+ years) with confirmed congenital IHH seen at the Massachusetts General Hospital Reproductive Endocrine Unit (1980-2020). Diagnostic criteria included hypogonadal sex-steroid levels (testosterone [T] < 100 ng/dL [< 3 nmol/L]), low (or inappropriately normal) serum gonadotropin (LH, FSH) levels and no evidence of an underlying cause of hypogonadism (ie, normal iron-binding studies, otherwise normal anterior pituitary function, normal imaging of the hypothalamo-pituitary region [ie, computed tomography or magnetic resonance imaging]) (5).

Phenotyping and Neuroendocrine Profiling

Participants were characterized as having Kallmann syndrome (KS: IHH + anosmia) based on formal smell testing using the University of Pennsylvania Smell Identification Test (UPSIT) (8) or self-reported complete inability to smell (9). UPSIT scores above the fifth percentile indicated a normal sense of smell (ie, normosmic [nHH]). UPSIT values less than or equal to the fifth percentile indicate anosmia, and patients with these scores were categorized as having KS. We categorized individuals as having a “pure” neuroendocrine phenotype if they did not exhibit any additional HH-associated phenotypes (ie, sensory deficits [anosmia, hearing loss], midline defects [cleft lip/palate], skeletal anomalies [syndactyly, clinodactyly, split hand/foot], dental anomalies [extra/missing teeth], neurologic anomalies [synkinesia], or anatomic anomalies [unilateral renal agenesis]). The presence of neonatal evidence of GnRH deficiency (cryptorchidism with/without micropenis) was also assessed in a larger cohort of IHH men (N = 590) with available questionnaires completed by the patient/referring physician. Men with TV greater than or equal to 4 mL by Prader orchidometer (before either exogenous gonadotropins or pulsatile GnRH) were classified as having partial puberty. Men with TV less than 4 mL were categorized as prepubertal (5). Participants underwent detailed neuroendocrine profiling following appropriate treatment-specific washout. Endogenous GnRH-induced LH secretion was assessed via overnight blood sampling every 10 minutes for 12 hours as previously described (6). We employed the modified Santen and Bardin methodology to identify LH pulses (10, 11). Serum LH and FSH were determined from study pools. Men who underwent reversal (12) (ie, pulsatile LH secretion and normalization of serum T levels) were carefully reviewed. Only those men with a neuroendocrine study before reversal were included in the analyses. Those men who had a neuroendocrine study at the time of (or after) reversal were excluded from the analyses.

Hormone Assays

Hormone assay results were retrieved from research records for analyses. As LH and FSH assay methods shifted (1980-2020), we relied on previously described internal validation studies (6) to ensure comparability between prior in-house radioimmunoassays (RIAs) and the 2-site monoclonal nonisotopic system (Axsym; Abbott Laboratories). Briefly, the Axsym FSH calibrators ranged from 1 to 150 IU/L in the Second International Reference Preparation (IRP) 78/549. The second IRP 78/549 is a pituitary-derived international reference preparation equivalent to 2.2 to 263 IU/L in the second IRP 71/223—the international human menopausal gonadotropin reference preparation previously used for the FSH RIA. The FSH method crossover study identified a limit of quantitation of 1.6 IU/L (∼20% coefficient of variation [CV]). The Axsym LH calibrators range from 1 to 150 IU/L (second IRP-80/552), which is equivalent to 6 to 611 IU/L in the second IRP 71/223 (the same human menopausal gonadotropin reference preparation used to calibrate the RIA). The LH method crossover study identified the limit of quantitation of 1.6 IU/L (∼20% CV). The Axsym LH and FSH assays have intra-assay and interassay CVs less than 7% and less than 7.4%, respectively. Gonadotropin levels falling below the level of detection were assigned the level of detection (1.6 IU/L) for analyses. Inhibin B (IB) was measured using a double-antibody enzyme-linked immunosorbent assay (Serotec) as previously described (5). The clinical limit of detection is 15.6 pg/mL, with intra-assay and interassay CVs of 4% to 6% and 15% to 18%, respectively. Serum IB levels falling below the limit of detection were assigned the level of detection (15.6 pg/mL) for analyses.

Genetic Sequencing and Analysis

Peripheral blood samples were collected from participants to perform whole-exome sequencing (Broad Institute). We used GATK best practices (Broad Institute) (13) to align exome sequencing (ES) data to the reference genome (hg19), conducting initial quality control, and performing variant calling algorithms. Single-nucleotide variation (SNV) calling and joint genotyping were performed using GATK components HaplotypeCaller and GenotypeGVCFs. Variant call format (VCF) files were annotated using SnpSift 4.3k and Ensembl VEP release 93. Rare SNVs were defined by a minor allele frequency (MAF) of less than 0.1% in the control database—gnomAD (14). Copy number variants (CNVs) were called from whole-exome sequencing with utilization of the GATK-gCNV pipeline as previously described (15).

Definition of Severity of Genetic Defects in IHH Genes

We included both SNVs/indels and CNVs in our genetic analyses. Genetic severity was defined for each of the 62 IHH-implicated genes (Supplementary Table S1) (16)based on their reported mode of inheritance, underlying variant type and MAF (14). For monogenic variants, we focused our analysis on “severe genetic defects”, that is, PTVs. The following class of variants in autosomal genes were deemed “severe”: deletions (partial, whole-gene, multigenic), intragenic duplications, nonsense, frameshift, and splice-site variants (3). For autosomal genes, SNVs/indels with MAF of less than 0.1% and CNVs with MAF less than 1% in gnomAD database were included. For autosomal recessive genes, homozygous/compound heterozygous variants (MAF < 1%) with at least one PTV allele were considered “severe” variants. For X-linked recessive genes, hemizygous PTVs (MAF < 1%) were considered severe. Since oligogenicity is known to occur in IHH, oligogenic variants were defined as SNVs/indels/CNVs in more than one IHH gene occurring in the same individual of which at least one of the variants was a “severe variant” (see aforementioned definition). Secondary oligogenic variants included both “severe” variants and rare missense variants (MAF < 1% for variants in autosomal and X-linked recessive genes and MAF less than 0.1% for variants in autosomal dominant genes).

Statistical Analysis

Results are reported using descriptive statistics (percentages, range, median, mean, SD). Chi-square tests were used (as appropriate) for between-group comparisons based on diagnosis (KS vs nHH) and pubertal status (absent vs partial). T tests were employed to compare gonadotropin levels between groups. Logistic regression was used to assess LH and FSH as predictors of LH pulsatility and report Nagelkerke pseudo R² as an approximation of the respective model's fit. Receiver operating characteristic curves were plotted to identify LH and FSH thresholds respectively for predicting LH pulsatility while minimizing false-positive and false-negative rates. For rare variant association testing, the total number of alternate and reference alleles across the coding regions of IHH genes were aggregated into a single alternate allele count and reference allele count per group. The alternate allele counts and reference allele counts were then used in a single rare variant burden test between groups using chi-square tests. P values less than .05 were considered statistically significant.

Results

A total of 242 men with IHH who underwent comprehensive reproductive phenotyping were included in the initial phenotypic analyses. A subset of 204 of 242 (84%) men with exome sequencing (ES) data were then examined to decipher genotype-phenotype correlations.

Olfactory, Pubertal, and Neuroendocrine Phenotypes

When the cohort was assessed for olfactory function, 131 of 242 (54%) were found to be anosmic (KS). Notably, anosmic men had statistically significantly higher rates of both cryptorchidism (P = .002) and micropenis (P = .012) (Table 1). Men with KS and their nHH counterparts were similar in terms of rates of absent/spontaneous puberty (ie, TV < 4 mL), apulsatile/pulsatile LH secretion, and rates of undetectable LH/FSH (see Table 1). We examined the cohort to identify the proportion of individuals with a “pure” neuroendocrine phenotype—that is, men with only a neuroendocrine defect without sensory deficits, midline defects, or skeletal, dental, neurologic, or anatomic anomalies. As this retrospective study spanned 1980 to 2020, some of the earlier patients (34/242, 14%) were not systematically characterized for HH-associated phenotypes. In total, 40 of 208 (19%) men exhibited a pure neuroendocrine phenotype without additional HH-associated anomalies.

Table 1.

Pubertal and neuroendocrine characteristics by diagnosis (n = 242)

	Anosmic (KS)	Normosmic (nHH)	Total	P
Diagnosis (n, %)	131 (54%)	111 (46%)	242 (100%)
ȃCryptorchidism	43 (33%)	17 (15%)	60 (25%)	<.005
ȃMicropenis	36 (27%)	15 (14%)	51 (21%)	<.05
Spontaneous puberty (n, %) Absent puberty (TV < 4 mL)	79 (60%)	65 (59%)	144 (60%)	.885
ȃPartial puberty (TV≥4 mL)	52 (40%)	46 (41%)	98 (40%)
LH secretion pattern (n, %) apulsatile	104 (79%)	81 (73%)	185 (76%)	.308
ȃPulsatile LH	27 (21%)	30 (27%)	57 (24%)
undetectable gonadotropins (n, %)
ȃUndetectable LH	85 (65%)	63 (57%)	148 (61%)	.246
ȃUndetectable FSH	73 (56%)	65 (59%)	138 (57%)	.754

Abbreviations: FSH, follicle-stimulating hormone; KS, Kallmann syndrome; LH, luteinizing hormone; nHH, normosmic hypogonadotropic hypogonadism; TV, testicular volume.

Reproductive Phenotypic Spectrum: Pubertal Status, Neuroendocrine Profile, and Their Interrelationships

Pubertal status was assessed in all 242 participants using TV measurements. All 242 men had also undergone every-10-minute blood sampling for 12 to 24 hours to assess baseline LH and FSH levels and for assessment of endogenous GnRH-induced LH pulses. Nearly two-thirds of men (144/242, 60%) exhibited absent puberty as attested by TV less than 4 mL (Table 2). Men with absent puberty had statistically significantly higher rates of micropenis (P = .032) but not cryptorchidism (P = .78) compared to men with partial puberty. Overall, approximately three-quarters (185/242, 76%) of men exhibited an apulsatile LH secretion pattern. Apulsatile and pulsatile groups neither differed in rates of cryptorchidism (51/185 vs 9/57; P = .71) nor micropenis (46/185 vs 5/57, P = .15). Men with absent puberty (TV < 4 mL) were more likely to have undetectable LH (absent puberty: 116/144 [81%] vs partial puberty: 32/98 [33%); P < .001) and undetectable FSH (absent puberty: 106/144 [74%] vs partial puberty: 32/98 [33%); P < .001). Similarly, men with absent puberty had significantly lower LH (1.94 ± 1.15 vs 3.39 ± 2.31; P < .001) and FSH (2.01 ± 1.34 vs 3.35 ± 2.63; P < .001) levels compared to men with partial puberty. In relation to LH pulsatility, men with partial pubertal development (TV≥4 mL) were more likely to exhibit pulsatile LH secretion compared to those with absent puberty (47/98 [48%] vs 10/144 [7%]; P < .001) (see Table 2). These observations demonstrate that pubertal status tracks favorably both with quantitative and qualitative neuroendocrine biochemical deficits (ie, LH levels and pulsatility).

Table 2.

Neuroendocrine profiles and rates of cryptorchidism/micropenis of isolated hypogonadotropic hypogonadism cohort by pubertal status (n = 242)

	Absent puberty (TV < 4 mL) n = 144	Partial puberty (TV≥4 mL) n = 98
Cryptorchidism (n, %)	42 (29%)	18 (18%)
Micropenis (n, %)	40 (28%)a	11 (11%)
Apulsatile LH secretion (n, %)	134/144 (93%)	51/98 (52%)
ȃDetectable LH (n, %)	17/134 (13%)	16/51 (31%)
ȃȃRange, IU/L	1.62-9.10	1.70-4.50
ȃȃMedian, IU/L	2.35	2.60
ȃȃMean ± SD, IU/L	2.89 ± 1.85	2.88 ± 0.88
ȃDetectable FSH (n, %)	27/134 (20%)	17/51 (33%)
ȃȃRange, IU/L	1.61-13.10	1.64-3.90
ȃȃMedian, IU/L	2.30	1.80
ȃȃMean ± SD, IU/L	3.08 ± 2.40	2.07 ± 0.58
pulsatile LH secretion	10/144 (7%)	47/98 (48%)
ȃDetectable LH (n, %)	10/10 (100%)	47/47 (100%)
ȃȃRange, IU/L	2.70-10.20	1.70-12.40
ȃȃMedian, IU/L	3.60	4.00
ȃȃMean ± SD, IU/L	4.56 ± 2.29	4.81 ± 2.52
ȃDetectable FSH (n, %)	9/10 (90%)	45/47 (96%)
ȃȃRange, IU/L	2.00-6.30	1.70-13.60
ȃȃMedian, IU/L	5.05	3.96
ȃȃMean ± SD, IU/L	4.50 ± 1.76	5.12 ± 2.92

Abbreviations: FSH, follicle-stimulating hormone; LH, luteinizing hormone; TV, testicular volume.

P less than .05.

To further examine the interrelationships between pubertal developmental status, gonadotropin levels, and LH pulsatility, we applied logistic regression to examine the utility of using spontaneous pubertal development (ie, partial puberty) and serum gonadotropins (LH/FSH) as predictors of LH pulsatility. Evidence of spontaneous pubertal development is a significant predictor (R² = 0.31, P < .001) of LH pulsatility (OR: 12.35; 95% CI, 5.80-26.27). Qualitatively, having detectable LH (> 1.6 IU/L) is associated with pulsatile LH secretion (ie, 57/57 men with pulsatile LH secretion had detectable LH). Having higher LH was significantly associated with likelihood of having pulsatile LH secretion. Considering gonadotropins as a continuous variable, we observed a statistically significant increased likelihood of having pulsatile LH secretion for every one-unit (1 IU/L) increase in serum LH (R² = 0.61, P < .001, OR: 4.92; 95% CI, 3.19-7.58) and FSH (R² = 0.50, P < .001, OR: 3.38; 95% CI, 2.31-4.96), respectively. Puberty was the most meaningful predictor for LH pulsatility. The model including partial puberty (controlling for each unit increase in LH and FSH) yielded the best fit (R² = 0.71) for predicting LH pulsatility (P < .001, OR: 10.80; 95% CI, 3.64-38.6). We created receiver operating characteristic curves to determine specific LH cutoffs for predicting LH pulsatility that demonstrated that serum LH threshold of 2.10 IU/L provides a 95% true-positive rate with a 15% false-positive rate. Finally, by mapping LH pulsatility across pubertal status (absent and partial puberty), we defined 4 gradients of neuroendocrine deficits (Fig. 1) and performed genotype-phenotype correlations. Of the individuals with absent puberty who exhibited pulsatile LH secretion (see Fig. 1), 2 of 10 had a history of maldescended testes—yet this did not differ from the overall rate of cryptorchidism in the overall group (20% vs 31%; P = .48).

Figure 1.

Gradient of neuroendocrine activity according to degree of spontaneous puberty (n = 242). The schematic depicts the relationship between degree of spontaneous pubertal development (absent [n = 144] vs partial [n = 98]) and pulsatility of gonadotropin-releasing hormone (GnRH)-induced luteinizing hormone (LH) secretion. Circle sizes are proportional to the number of individuals in each quadrant.

Relationship of Inhibin B With Olfactory, Pubertal, and Neuroendocrine Phenotypes

A subset of men (152/242, 63%) had results available for IB. In relation to diagnosis (KS: n = 79, nHH: n = 73), anosmic men exhibited significantly low serum IB levels (median: 26 vs 50 pg/mL; P < .001). Considering individuals with signs of absent minipuberty (ie, cryptorchidism and/or micropenis), IB was significantly lower in men (n = 51/152, 34%) with a history of maldescended testes (27 vs 45 pg/mL; P = .005). Similarly, men with a history of micropenis (n = 31/152, 20%) had significantly low IB levels (22 vs 43 pg/mL; P = .001). Men with absent spontaneous puberty (TV < 4 mL, n = 63 [41%]) had significantly lower IB levels (26 vs 71 pg/mL; P < .001). In addition, those men exhibiting apulsatile LH secretion (n = 118/152, 78%) had significantly lower IB levels (29.5 vs 86 pg/mL; P < .001).

Genotype-reproductive Phenotype Correlations

In total, PTVs were identified in 36 of 204 (18%) individuals (Table 3). The majority of PTVs (28/36) occurred along with secondary variants (oligogenic). PTV burden was then correlated with pubertal status, LH pulsatility, the presence or absence of neonatal signs of GnRH deficiency (ie, micropenis and cryptorchidism) as well as other nonreproductive phenotypes (see “Materials and Methods”) and, finally across gradients of neuroendocrine activity. Men with absent puberty trended toward a higher prevalence of PTVs compared to partial puberty counterparts—yet differences did not reach statistical significance (26/118 [22%] vs 10/86 [11%]; P = .063) (Supplementary Table S1) (16). In addition, burden from a PTV with a secondary variant (oligogenic) did not differ between 2 groups (20/119 [17%] vs 8/86 [9%]; P = .14). Examining PTVs in individual IHH genes revealed that men with absent puberty were enriched for PTVs in ANOS1 compared to men with partial puberty (15/118 [13%] vs 1/86 [1%]; P = .002) (Supplementary Table S1) (16). Following ANOS1 (16/36, 44%), men in the cohort carried PTVs in FGFR1 7/36, 19%), CHD7 (2/36, 5.5%), SOX10 (2/36, 5.5%), POLR3B 2/36, 5.5%), PROKR2 (1/36, 3%), GLI3 (1/36, 3%), GNRHR (1/36, 3%), KISS1R (1/36, 3%), NDNF (1/36, 3%), POLR3A (1/36, 3%), and TCF12 (1/36, 3%).

Table 3.

Clinical diagnosis and genetic variants in 36 isolated hypogonadotropic hypogonadism men harboring protein truncating variants

Gene	Variant(s) (PTV in bold)	Dx	Puberty	NE pulsatility	Nonreproductive phenotypes
ANOS1	ANOS1 c.1201_1207 + 4del11 p.N401Kfs*5 hem	KS	Absent	Apulsatile	High arched palate
ANOS1	ANOS1 c.571C > T p.R191* hem FSHB c.59G > T p.S20I het RAB3GAP1 c.1233_1235del p.L412del het	KS	Absent	Apulsatile	Pectus excavatum, ptosis, face/mouth surgery, high arched palate, hearing loss
ANOS1	ANOS1 c.1392_1405del p.E465Qfs*30 hem	KS	Absent	Apulsatile	Strabismus
ANOS1	ANOS1 c.67_92del p.L23Cfs*77 hem CCDC141 c.1802A > T p.H601L het DCC c.2267G > A p.R756Q het GNRH1 c.229C > T p.R77* het	IHH	Absent	Apulsatile	None (> 5th percentile UPSIT testing)
ANOS1	ANOS1 c.1369C > T p.R457* hem AMH c.136C > G p.P46A het CUL4A c.86T > C p.V29A het PNPLA6 c.1123G > C p.A375P het	KS	Absent	Apulsatile	Eye disorder, flattened nose bridge
ANOS1	ANOS1 c.1801del p.L601Yfs*19 hem SMCHD1 c.1804A > G p.I602V het	KS	Absent	Apulsatile	Crowded teeth, synkinesia
ANOS1	ANOS1 c.1449 + 1G > A hem;  GLI3 c.1285C > G p.P429A het IGSF1° c.64G > A p.A22T het	KS	Absent	Apulsatile	Scoliosis, strabismus, hearing loss, learning disability, seizures, speech impairment
ANOS1	ANOS1 c.1887_1888del p.Y630Pfs*36 hem AXL c.1549G > A p.G517S het PLXNA1 c.1765G > A p.V589M het PNPLA6 c.1484C > T p.P495L het	KS	Absent	Apulsatile	Pes cavus, crowded teeth, face/mouth surgery, missing teeth
ANOS1	400 bp intragenic ANOS1 deletion CHD7 c.856A > G p.R286G het PROP1 c.152G > T p.G51V het	KS	Absent	Apulsatile	Eye movement disorder
ANOS1	ANOS1 c.1449 + 1G > A hem;  CHD7 c.1117C > T p.L373F het IGSF1° c.7354ins het KLB c.2329_2331del p.F777del het	KS	Absent	Apulsatile	Scoliosis, strabismus, hearing loss, learning disability, seizures,
ANOS1	ANOS1 c.814C > T p.R272*hem DCC c.3470A > G p.H1157R het;	KS	Absent	Apulsatile	Bone abnormalities, synkinesia
ANOS1	ANOS1 c.1381delC p.R461Gfs*21 hem AMH c.1666G > A p.E556K het AXL c.257 °C > T p.A857V het FGF8 c.159_164del p.T54_V55del het HS6ST1 c.1121G > A p.S374N het	KS	Absent	Apulsatile	Small hands/feet, crowded teeth, synkinesia, other neurologic conditions
ANOS1	ANOS1 c.127 °C > T p.R424* hem DMXL2 c.6046G > A p.D2016N het RAB3GAP1 c.2938G > A p.A980T het	KS	Absent	Apulsatile	Small hands/feet, high arched palate
ANOS1	ANOS1 c.1891C > T p.R631* hem AMHR2 c.1297C > T p.P433S het	KS	Absent	Apulsatile	Excessive joint mobility, eye movement disorder, crowded teeth, deviated septum, high arched palate, synkinesia, neurologic conditions, kidney problems
ANOS1	ANOS1 c.1A > C p. hem IGSF1° c.2237C > G p.P746R het SEMA3A c.458A > G p.N153S het	KS	Absent	Apulsatile	Camptodactyly, bone abnormalities, crowded teeth, synkinesia
ANOS1	1.5 MB multigenic ANOS1 deletion PNPLA6 c.1484C > T p.P495L het SRA1 c.266C > T p.P89L het	KS	Partial	Pulsatile	Synkinesia
FGFR1	FGFR1 c.1809C > A p.C603* het LEPR c.2246T > C p.L749S het	IHH	Absent	Pulsatile	Scoliosis, foreshortened arm/leg
FGFR1	FGFR1 c.1285-2A > G het FSHB c.70A > G p.T24A het IGSF1° c.1046A > G p.N349S het WDR11 c.1175G > A p.R392Q het	IHH	Absent	Apulsatile	None
FGFR1	5MB FGFR1 multigenic deletion AMH c.295A > T p.T99S het AXL c.5C > T p.A2V het	KS	Absent	Apulsatile	None
FGFR1	FGFR1 c.2038C > T p.Q680* het	IHH	Absent	Apulsatile	Foreshortened arm/leg, colorblindness, nystagmus, cleft lip/palate, face/mouth surgery, gap between teeth, other head/face irregularities, hearing loss
FGFR1	FGFR1 c.1727_1734del p.R576Pfs*77 het OTUD4 c.1458G > T p.K486N het OTUD4 c.909C > A p.S303R het	KS	Absent	Apulsatile	None
FGFR1	FGFR1 c.1037_1038del p.S346Yfs*61 het DCC c.619A > G p.I207V het	KS	Partial	Pulsatile	None
FGFR1	FGFR1 c.1864C > T p.R622* het NSMF c.587G > A p.R196H het	KS	Partial	Pulsatile	Pes cavus, bone abnormalities, facial feature imbalance
CHD7	CHD7 c.7957C > T p.R2653* het PLXNA1 c.3293A > G p.N1098S het RAB3GAP2 c.585A > G p.I195M het SPRY4 c.530A > G p.K177R het;	KS	Absent	Apulsatile	Scoliosis, head/face irregularities, hearing loss, external ear defects, other kidney abnormalities, learning disability, peripheral neuropathy, speech impairment, Asperger syndrome/autism
CHD7	CHD7 c.8228_8247dup p.F2750Lfs*4 het	KS	Partial	Pulsatile	Pectus excavatum, small hands/feet
SOX10	SOX1° c.1038_1039delAC p.P347Tfs*54 het	IHH	Partial	Apulsatile	Pes cavus, hearing loss
SOX10	SOX1° c.1038_1039delAC p.P347Tfs*54 het KLB c.3032T > C p.L1011P het	KS	Partial	Apulsatile	Pes cavus, high arched palate
POLR3B	POLR3B c.497-2A > G het POLR3B c.1244T > C p.M415T het	IHH	Absent	Apulsatile	Amblyopia
POLR3B	POLR3B c.1244T > C p.M415T het POLR3B c.2818-2A > T het TBC1D2° c.1016T > C p.M339T het KL c.56 °C > T p.P187L het	IHH	Partial	Apulsatile	None
PROKR2	PROK2 c.163del p.I55* hom AXL c.1549G > A p.G517S het OTUD4 c.386C > G p.A129G het	KS	Absent	Apulsatile	None
GLI3	GLI3 c.1161del p.P388Qfs*13 het CCDC141 c.4369G > A p.V1457I het GNRHR c.317A > G p.Q106R het SPRY4 c.313G > A p.D105N het	KS	Partial	Pulsatile	Scoliosis, polydactyly, syndactyly, high arched palate, missing teeth, learning disability
GNRHR	GNRHR c.112C > T p.R38* hom	IHH	Absent	Apulsatile	Clinodactyly, neurologic condition
KISS1R	KISS1R c.991C > T p.R331* het KISS1R c.1195T > A p.*399R het AMHR2 c.503C > T p.A168V het PNPLA6 c.1198G > A p.D400N het SQSTM1 c.374A > G p.N125S het	IHH	Absent	Apulsatile	None
NDNF	NDNF c.785del p.S262* het CCDC141 c.1521A > C p.Q507H het DMXL2 c.812C > G p.T271S het FGFR1 c.1447C > A p.P483T het SPRY4 c.722C > A p.S241Y het	KS	Partial	Apulsatile	None
POLR3A	POLR3A c.3013C > T p.R1005C het POLR3A c.2830G > T p.E944* het	IHH	Absent	Pulsatile	Nystagmus, missing teeth, bladder outlet obstruction, depth perception impairment, learning disability, seizures, speech impairment, leukodystrophy
TCF12	TCF12 c.1491dup p.V498Cfs*12 het CCDC141 c.3111A > T p.K1037N het NRP2 c.100 °C > T p.R334C het	KS	Partial	Apulsatile	Missing teeth, intracranial cysts

Abbreviations: Dx, diagnosis according to olfactory function; KS, Kallmann syndrome; NE, neuroendocrine; nHH, normosmic isolated hypogonadotropic hypogonadism; PTV, protein-truncating variant; UPSIT, University of Pennsylvania Smell Identification Test; *, stop codon.

We examined the relationship of the PTV burden with the phenotypic characteristics. The PTV burden (either monogenic or oligogenic) did not differ between men with a “pure” neuroendocrine phenotype compared to men exhibiting nonreproductive phenotypes (Supplementary Table S6) (16). Similarly, the presence of cryptorchidism and/or micropenis was not associated with a PTV burden compared to those with intact minipuberty (Supplementary Table S7) (16). However, when we expanded the analysis to a cohort of IHH men with available information on cryptorchidism/micropenis on questionnaires (N = 590), we found men with a history of cryptorchidism/micropenis were enriched for PTVs (both monogenic and oligogenic) (Table 4).

Table 4.

Cryptorchidism and/or micropenis isolated hypogonadotropic hypogonadism (IHH) men (n = 393) harboring protein truncating variants in IHH genes

Variants in IHH genes	cryptorchidism (n = 151)	No cryptorchidism (n = 242)	P
PTV (n, %)	32 (21%)	21 (9%)	.0007
Oligogenic (≥ 1 PTV)	27 (18%)	16 (7%)	.0008
	micropenis (n = 164)	No micropenis (n = 201)
PTV (n, %)	31 (19%)	23 (11%)	.054
Oligogenic (≥ 1 PTV)	27 (16%)	17 (8%)	.023
	cryptorchidism and micropenis (n = 81)	No cryptorchidism and micropenis (n = 145)
PTV (n, %)	17 (21%)	9 (6%)	.001
Oligogenic (≥ 1 PTV)	15 (18.5%)	6 (4%)	.0006

Abbreviations: IHH, isolated hypogonadotropic hypogonadism; PTV, protein-truncating variant.

We then examined the PTV burden in relationship to serum gonadotropin levels. PTV burden in men with undetectable LH (n = 124), undetectable FSH (n = 113), and undetectable LH/FSH (n = 100) did not differ from those with detectable gonadotropin levels (ie, > 1.6 IU/L) (Supplementary Tables S2-S4) (16). Comparing men with apulsatile vs pulsatile LH secretion, neither rates of overall PTVs (29/152 [19%] vs 7/52 [13%]; P = .40) nor oligogenic PTVs differed between groups (23/152 [15%] vs 5/52 [10%]; P = .36) (Supplementary Table S5) (16). As prior work has identified serum IB levels greater than 70 pg/mL as a predictor of fertility outcome for men with IHH (5), we examined PTV burden according to IB—but no statistically significant differences were noted between groups. In summary, genetic severity (PTV in ANOS1) was correlated with pubertal status (as determined by TV). In this cohort of 204 men with HH, genetic severity (PTV) did not track with biochemical profile, LH pulsatility, cryptorchidism/micropenis, or other nonreproductive characteristics. However, the lack of association may result from the lack of power for genetic analyses in men who underwent dynamic neuroendocrine studies.

We then evaluated PTV burden across the 4 gradients of GnRH deficiency (Table 5)—group A: absent puberty with apulsatile LH secretion (n = 108); group B: partial puberty with apulsatile LH secretion (n = 44); group C: absent puberty with pulsatile LH secretion (n = 10); and group D: partial puberty with pulsatile LH secretion (n = 42). Genetic burden analysis revealed the group with the most severe phenotype (ie, men with absent puberty and apulsatile LH secretion) were not enriched for overall PTVs but showed enrichment for oligogenic PTVs (ie, PTV with secondary variants) (see Table 5). However, the majority of PTVs (overall and oligogenic) implicated ANOS1 (Supplementary Table S8) (16). This observation suggests PTVs in ANOS1 PTVs may be the primary driver of the observed enrichment—and the second variant occurring alongside the ANOS1 PTV does not contribute to reproductive phenotypic severity. This notion is also supported by the observation that oligogenic PTVs resulted in relatively less severe phenotypes, that is, 9% of men with partial spontaneous puberty and pulsatile LH patterns and 9% of men with partial spontaneous puberty and apulsatile LH patterns (see Table 5).

Table 5.

Summary of genetic findings in isolated hypogonadotropic hypogonadism men (n = 204) across a gradient of gonadotropin-releasing hormone deficiency^a

Variants	Apulsatile LH secretion		Pulsatile LH secretion		χ2(3)	P
	A: absent puberty (n = 108)	B: partial puberty (n = 44)	C: absent puberty (n = 10)	D: partial puberty (n = 42)
Variants in IHH loci (n, %)
PTV	24 (22%)	5 (11%)	2 (20%)	5 (12%)	5.71	.126
Standardized residuals	1.65	−1.51	1.07	−1.22
Oligogenic variants (n, %)
Oligogenic with ≥ 1 PTV	20 (18%)	4 (9%)	—	4 (9%)	19.35	<.001
Standardized residuals	3.64	−0.13	−3.54	0.04

Abbreviations: IHH, isolated hypogonadotropic hypogonadism; LH, luteinizing hormone; PTV, protein-truncating variants.

Individual variants are provided in supplemental material.

Discussion

A developmental perspective has previously been applied to describe the phenotypic variability in IHH according to degree of spontaneous puberty (6). In this report, we expand on this developmental framework and by juxtaposing detailed phenotypic information and ES data from a large IHH cohort to examine correlations between genotypic severity and reproductive phenotypes. We show that partial spontaneous pubertal development and serum LH levels greater than or equal to 2.10 IU/L are strong predictors of pulsatile LH secretion. We defined a gradient of neuroendocrine deficits by focusing primarily on PTVs in IHH genes as a marker of genotypic severity and by integrating clinical and biochemical features. Within this gradient, oligogenic PTVs (PTVs with a secondary IHH gene variant) were enriched in IHH men with absent puberty and apulsatile LH—compared to the men with partial puberty and pulsatile LH (P < .001). Analyses of individual IHH genes within the enriched group showed that ANOS1 PTVs were likely to be the primary driver of the observed enrichment and resultant phenotypic severity. In contrast, PTVs in other IHH genes (monogenic or oligogenic) were associated with more heterogeneous reproductive phenotypes. Our results reiterate the reproductive phenotypic severity previously demonstrated for ANOS1 variants (17) and provide both quantitative and qualitative evidence for the considerable resilience of the HPG axis to severe genetic defects in most IHH genes.

Plasticity of Neuroendocrine Function in Isolated Hypogonadotropic Hypogonadism (IHH) in the Face of Genotypic Severity for Most IHH Genes

A spectrum of GnRH-deficient phenotypes has been described in humans ranging from constitutional delay of puberty (CDP) and hypothalamic amenorrhea (HA), which represent milder GnRH deficiency phenotypes, to IHH, representing the most severe end of this spectrum. CDP, HA, and IHH have been shown to have a shared genetic architecture with suggestive evidence that less-severe genetic defects result in CDP/HA whereas more severe defects result in IHH (18). In this study we expand on this genetic severity gradient, hypothesizing that even in IHH, the most severe end of the GnRH-deficiency spectrum, genetic severity may determine reproductive phenotypic severity. Although more than 60 genes have been found to underlie IHH (3), to date only a few studies have specifically examined the correlation between genetic burden/severity with reproductive phenotypic severity within IHH. We observed a positive correlation between pubertal phenotypes (absent vs partial puberty) with genotypic severity, and this association was driven by PTVs in ANOS1 (P = .002) (Supplementary Table S8) (16). Thus, results from the present study support prior observations that ANOS1 variants were associated with more severe reproductive phenotypes compared men with IHH harboring variants in FGFR1 (19, 20).

Barring ANOS1, PTVs in other IHH genes displayed more variable reproductive phenotypes. When examined across a neuroendocrine-deficit gradient, PTVs in IHH genes (occurring alone or with secondary variants) resulted in relatively milder reproductive phenotypes (partial puberty with pulsatile/apulsatile LH secretion) (see Table 5). The relative resilience of the HPG axis is depicted in Fig. 2, showing representative examples of dynamic neuroendocrine studies. While ANOS1 PTVs was sufficient to impart a severe phenotype (absent puberty, apulsatile LH profile; Fig. 2B), IHH individuals harboring an FGFR1 PTV either alone (Fig. 2C) or in combination with a secondary variant (Fig. 2D) exhibit relatively milder neuroendocrine deficits. Thus, the preserved (yet attenuated) biochemical secretory function in IHH individuals—even in the face of PTVs—strongly suggests that the human HPG axis possesses considerable plasticity. Further, the notion of plasticity is consistent with the well-defined phenomenon of reversal of GnRH deficiency wherein the HPG axis is reactivated in a subset (ie, 10%-15%) of men with IHH (12). Such plasticity is particularly evident in IHH patients with TAC3/TACR3 variants who have been shown to have normal/near-normal FSH levels, robust response to GnRH stimulation testing (21), and increased prevalence reversal of the IHH phenotype (22, 23). Taken together with previously reported observations, our results underscore the significant resilience of the HPG axis to genetic defects. The robust plasticity of the human HPG axis is consistent with the relative very low incidence and prevalence of IHH in the general population. In addition, the lack of genetic burden in IHH men with apulsatile vs pulsatile secretion as well as between those with undetectable gonadotropins vs detectable gonadotropins could be explained by 2 different nonexclusive hypotheses: (i) First, the total number of PTVs was relatively low and thus, statistical comparisons most likely lack statistical power to show a significant difference; and (ii), given that the assays used in our study may date from many years ago, it is possible that the sensitivities of the assays are probably insufficient to properly classify patients with severe neuroendocrinological phenotypes.

Figure 2.

Representative luteinizing hormone (LH) secretion patterns from dynamic LH pulse studies. The 4 panels depict representative LH pulse studies. The shaded region shows the mean ± 2 SDs for health controls (30). A, Healthy adult male with normal puberty. B, Apulsatile LH in a Kallmann syndrome (KS) male harboring a deletion in ANOS1. C, Low-amplitude pulse study in a normosmic hypogonadotropic hypogonadism (nHH) male carrying a rare, pathogenic variant in FGFR1. D, Low-amplitude pulse study in a KS male harboring oligogenic genetic architecture (FGFR1 and DCC).

Oligogenicity and Severity of Reproductive Phenotype

The genetic architecture of IHH is complex with several reports of oligogenic inheritance wherein IHH individuals have been noted to harbor genetic variants in more than one gene (3, 24, 25). Yet to date, only a few studies have attempted to evaluate the effect of oligogenicity on reproductive phenotypic severity (26, 27). A prior study of 55 KS males with biallelic and monoallelic rare variants in PROKR2 mutations showed that male patients carrying biallelic mutations in PROK2 or PROKR2 have a less variable and more severe reproductive phenotype compared to monoallelic counterparts (26). To our knowledge, ours is the first study examining the cumulative burden of oligogenic PTVs across 62 IHH genes and its correlation with reproductive phenotypes. Our subgroup analysis showed that IHH men with absent puberty and apulsatile LH secretion were enriched for oligogenic PTVs compared to those with absent puberty with pulsatile LH secretion. However, examining individual genes underlying this enrichment revealed that PTVs in the ANOS1 gene appear to be the major driver for the observed enrichment. Thus, oliogenicity in men with absent puberty and apulsatile LH secretion is likely to be incidental rather than causal for the observed phenotypic severity. In addition, our data show that oligogenic PTVs do not necessarily confer severe phenotypes for most IHH genes (see Table 5, groups B and D; Fig. 2D). However, caution is merited for this assertion as the groupwise comparison across the puberty/LH pulsatility gradient for oligogenic variants were based on small numbers (see Table 5, group B [n = 4] and D [n = 4]) and this observation requires further validation using larger cohorts of patients with oligogenic variants.

Olfactory Phenotype as a Predictor of Reproductive Phenotypic Severity

In the present study, we observed rates of anosmia similar to prior reports (54% vs 46%; P = .20) (6). It has been previously suggested that olfactory dysfunction in IHH patients (ie, KS phenotypes) may be associated with more severe phenotypes compared to their normosmic counterparts (6). We observed that 60% of men with KS exhibited absent puberty (TV < 4 mL), which is significantly lower than a 2002 study reporting prepubertal testes in 26/30 (87%) men with KS (P = .006) (6). However, our reported rate of partial puberty among men with nHH is similar to the 2002 cohort (17/42 [41%] vs 46/111 [41%]; P = .91). The similar rates of absent/partial puberty, undetectable LH/FSH, and apulsatile/pulsatile LH secretion between KS and nHH suggest that olfactory phenotype alone does not provide any substantial insights into HPG axis activity.

Serum Luteinizing Hormone (LH) Levels as a Surrogate for Discerning LH Pulsatility

In addition to ascertaining genotype-phenotype correlations, we used the phenotypic framework in this study to examine the interrelationships between clinical and biochemical markers of HPG axes deficits in men with IHH. Linear regression analyses identify partial spontaneous puberty as a clinical proxy for neuroendocrine activity (ie, pulsatile LH secretion pattern). Individuals with TV greater than or equal to 4 mL were 12 times more likely to have pulsatile LH secretion (R² = 0.31, OR: 12.35; 95% CI, 5.80-26.27)—yet controlling for each one IU/L increase in LH and FSH provided the best fit for partial puberty as a predictor of pulsatile LH secretion (R² = 0.71, OR: 10.8; 95% CI, 3.6-38.6). Moreover, the present analysis demonstrates that serum LH greater than or equal to 2.10 IU/L can be used as a surrogate marker for predicting LH pulsatility (95% true-positive rate, 15% false-positive rate). The findings have relevance for phenotyping in research contexts as frequent sampling studies (ie, blood draws every 10 minutes for 12-24 hours) is resource intensive and not feasible in most research settings.

Concordance Between Pubertal Phenotypes and Biochemical Phenotypes

We found the overwhelming majority (93%) of men with absent puberty exhibited apulsatile GnRH-induced LH secretion—consistent with their frankly hypogonadal T levels. In early puberty, FSH stimulates the proliferation of Sertoli cells and spermatogonia that contribute to the testicular growth (TV ³ 4 mL) that marks the onset of puberty (28). We found that men with absent puberty had significantly lower FSH (P < .0001) and were more likely to have undetectable FSH levels compared to men with some degree of spontaneous puberty (74% vs 33%; P < .0001). This observation highlights the critical role of FSH in inducing testicular growth and development. Further, the observation is clinically relevant because it provides additional support for pretreatment with FSH for induction of testicular growth and spermatogenesis in IHH men before initiating human chorionic gonadotropin therapy (or pulsatile GnRH) (29).

Discordance Between Pubertal Phenotypes and Biochemical Phenotypes

Interestingly, 51 of 98 (52%) men with partial puberty had apulsatile LH secretion. An explanation for this observation is that an ultra-sensitive assay (eg, Delfia) was not used. With greater sensitivity it is possible that LH pulses may have been detected. Similarly, 33% of men with partial puberty had FSH levels at or below the level of detection (1.6 IU/L). This observation is curious as FSH secretion would presumably be required for testicular development (ie, proliferation of Sertoli cells and spermatogonia). However, similar to LH, the use of ultra-sensitive assays could detect very low serum FSH levels and such assays were not used in this study. Another notable observation was the small group of 10 (4%) men with absent puberty who exhibited pulsatile LH secretion. It is curious that, despite pulsatile LH secretion, these participants did not show any pubertal development. We consider that this subgroup of men with IHH may also harbor additional testicular/seminiferous tubule defects that could contribute to the lack of testicular development. A 2010 study examined a group of 90 men with IHH who underwent long-term pulsatile GnRH therapy and identified evidence of hypothalamic, pituitary, and testicular defects (17). Specifically, 10 of 90 (11%) men remained hypogonadotropic and hypogonadal pointing to additional pituitary and testicular defects in addition to the hypothalamic defect of isolated GnRH deficiency. As such, it is plausible that the small subset of men in the present study may also harbor a “dual defect” (ie, hypothalamic and testicular) (17).

Study Strengths and Limitations

Relative strengths of this study include the comprehensive screening for variants in IHH loci and the systematic detailed olfactory, pubertal, and neuroendocrine phenotyping in a sizeable cohort of men with IHH. To our knowledge, this is the first study to juxtapose genetic burden and olfactory, pubertal, and neuroendocrine phenotypes. However, it is worth noting several limitations. In terms of the physiologic investigations, this study involved patients seen at the Massachusetts General Hospital Reproductive Endocrine Unit (1980-2020). During this time, gonadotropin assay methods shifted—yet all results were anchored using standards (as described in ‘Materials and Methods”). As previously noted, ultra-sensitive gonadotropin assays (eg, Delfia) were not used and it is possible that the rates of undetectable gonadotropin levels and LH pulsatility could shift with more sensitive assay methodologies. In terms of genetic investigations, genotype-phenotype correlations are challenging in patient cohorts with limited sample size. However, given the rarity of IHH/KS (31), a sample size of 242 is quite robust. Last, because data were collected from a single center, it is possible that there is a risk of referral bias. As such, it is possible that study findings may not be fully representative of all patients with IHH and caution is warranted in extrapolating findings (including women with IHH).

Conclusions

Detailed neuroendocrine phenotyping using clinical and biochemical evaluation reveals a spectrum of GnRH deficiency in IHH men. We identify partial spontaneous pubertal development and serum LH levels greater than or equal to 2.10 IU/L are strong predictors of pulsatile LH secretion.

Genotype-phenotype correlations provide additional evidence that ANOS1 PTVs are associated with more severe GnRH deficiency as evidenced by high rates of absent puberty and apulsatile LH secretion. In contrast, severe defects in other IHH genes impart more variable phenotypic severity pointing to the considerable resilience of the HPG axis to genetic defects in most IHH genes—even in the face of oligogenic variants. Since the genetic etiology in IHH is still unknown for nearly 60% of IHH patients, more work is needed to fully elucidate and chart the molecular basis of pubertal and neuroendocrine phenotypes in IHH.

Abbreviations

CDP

constitutional delay of puberty

CNV

copy number variant

CV

coefficient of variation

ES

exome sequencing

FSH

follicle-stimulating hormone

GnRH

gonadotropin-releasing hormone

HA

hypothalamic amenorrhea

HPG

hypothalamo-pituitary-gonadal

IB

Inhibin B

IHH

isolated hypogonadotropic hypogonadism

IRP

International Reference Preparation

KS

Kallmann syndrome

LH

luteinizing hormone

MAF

minor allele frequency

nHH

normosmic hypogonadotropic hypogonadism

PTV

protein-truncating variant

RIA

radioimmunoassay

SNV

single-nucleotide variant

T

testosterone

TV

testicular volume

UPSIT

University of Pennsylvania Smell Identification Test

VCF

variant call format

Funding

This work was supported by the NIH Eunice Kennedy Shriver National Institute of Child Health and Human Development (1P50HD104224-01) “Massachusetts General Hospital – Harvard Center for Reproductive Medicine”. AAD was also supported by the NIH/NCATS 1R03TR003533-01. MIS was also supported by the NIH/NICHD F32HD108873-01.

Disclosures

The authors have nothing to disclose.

Data Availability

Deidentified data will be made readily available on request for research purposes to qualified individuals within the scientific community.