第一部分 嗅觉评估对先天性低促性腺激素性性腺功能减退症患者的鉴别诊断价值

背景

嗅觉作为人类重要的感官功能，在个体交流、情绪活动及感知周围环境的危险等方面有突出作用。嗅觉障碍对自身行为、安全、心理等方面均有不良影响。因此，明确患者嗅觉功能对改善其生活质量有重要意义。先天性低促性腺激素性性腺功能减退症（CHH）患者根据有无嗅觉异常，分为合并嗅觉异常的卡尔曼综合征（KS）和嗅觉正常的CHH（nCHH）。临床中通常根据患者主观自评或嗅神经MRI区分KS和nCHH，但主观自评的嗅觉功能准确率欠佳，而嗅觉通路结构与嗅觉功能存在不匹配的情况。嗅觉功能的心理物理学评估是简单便捷的定量评估嗅觉功能的方法，但由于文化及生活环境的差异，国际上常用的评估工具不完全适用于中国人群。2019年中国科学院心理研究所周雯教授团队开发了中国人群嗅觉识别测试（CSIT），为评估中国人群的嗅觉功能提供了有效的工具。目前CSIT 的检测效能已在帕金森病及精神分裂症患者中得到证实，但尚未应用于CHH患者。

目的

利用中国人群嗅觉功能测试（COFT）工具评估CHH患者的嗅觉功能，比较分析嗅觉主观自评测试（SROS）和嗅神经MRI（MRI-OA）对KS的诊断价值，分析影响CHH患者嗅觉的因素，比较分析nCHH和KS患者治疗疗效及基因变异分布情况。

方法

纳入2020年12月至2021年6月在北京协和医院内分泌科性腺门诊就诊的CHH患者。收集患者的临床资料。采集外周血提取白细胞DNA。采用目标序列捕获联合二代测序的方法筛选存在CHH相关致病基因罕见变异的患者。应用SROS和COFT评估患者嗅觉功能。SROS采用5级评分，分别为“差”、“中下”、“一般”、“中上”及“好”，自评为“好”、“中上”及“一般”为嗅觉功能正常；自评为“中下”及“差”为嗅觉功能异常。COFT由嗅觉阈值测试（CSTT），嗅觉分辨测试（CSDT）和CSIT组成。三个试验总分（TDI）≥29.25分为嗅觉功能正常；TDI＜29.25分为嗅觉功能异常。利用MRI-OA评估CHH患者嗅觉通路结构，并测量嗅球大小（OBV）及嗅沟深度（OSD）。利用统计学方法比较nCHH及KS患者嗅觉功能及治疗结局。

结果

1、临床特点：研究共纳入71例CHH患者，其中69例为男性，2例为女性。患者平均年龄为23.6±5.8岁，体重指数（BMI）为25.1±4.2 kg/m²，平均血清黄体生成素水平为1.15±2.26 U/L，卵泡刺激素水平为2.33±3.65 U/L，总睾酮水平（T）为1.55±1.52 ng/ml，雌二醇（E2）水平为20.31±11.66 pg/ml。其中28例为初诊的患者，其余43例为已治疗的患者。这些患者中吸烟、饮酒和合并慢性鼻炎的比例分别为25%、28%和31%。

2、nCHH和KS患者嗅觉功能比较：根据COFT结果，36例患者为nCHH患者，35例为KS患者。SROS方面，nCHH患者中3例自评嗅觉功能为“好”，12例为“中上”，20例为“一般”，1例为“中下”。KS患者中，1例患者自评为“中上”，4例为“一般”，2例为“中下”，28例为“差”。MRI-OA方面，nCHH患者中67%（18/27例）的患者MRI未见明显异常，22%（6/27例）为单侧嗅球不发育或发育不良，11%（3/27例）为双侧嗅球不发育或发育不良。96%（27/28例）的KS患者MRI表现为双侧嗅球不发育或发育不良，4%（1/28例）未见明显异常。nCHH患者的嗅球体积及嗅沟深度大于KS患者（37.58±20.64 vs 2.40±10.72 mm³，p=0.000；6.82±2.31 vs 2.64±3.07 mm，p=0.000）。

3、嗅觉影响因素分析：TDI评分为因变量，年龄、BMI、随诊时间、T、E2、吸烟、饮酒、鼻炎、SROS、OBV及OSD为自变量进行线性回归分析。结果显示年龄、BMI、随诊时间、T、E2、吸烟、饮酒及鼻炎与嗅觉功能无显著相关性，SROS、OBV及OSD与嗅觉功能呈显著相关。

4、SROS及MRI-OA诊断价值比较：利用COFT为诊断KS的标准，SROS判断嗅觉功能的准确率为91.5%，诊断KS的敏感性、特异性、阳性预测值和阴性预测值分别为85.7%、97.2%、96.8%和87.5%。MRI-OA判断嗅觉功能的准确率为92.7%，诊断KS敏感性、特异性、阳性预测值和阴性预测值分别为96.4%、88.9%、90.0%和96.0%。

5、治疗疗效分析：共26例促生精治疗男性CHH患者纳入分析，其中19例应用人绒毛膜促性腺激素联合人绝经期促性腺激素治疗，7例GnRH脉冲泵治疗，平均随访49.4±29.5月。患者开始治疗中位年龄20.5 (17.8, 26.0)岁，基线睾丸体积1.25 (1.0, 2.0) ml，血清LH水平为0.32 (0.20, 0.89) U/L，FSH水平为0.57 (0.30, 1.80) U/L，总睾酮水平为0.37 (0.29, 0.61) ng/ml。治疗后患者睾丸体积增加8.8±4.7 ml，血清总睾酮水平提升2.6±1.6 ng/ml。67%（14/21）的患者成功生精。平均生精时间是21.1±12.8月。nCHH和KS患者治疗结局无差异。

6、基因变异分析：共47例患者进行二代测序筛查可能的致病基因，其中51.0%（24/47）患者检测到CHH相关基因的罕见变异位点。根据ACMG指南，11个位点被评为致病/可能致病变异，致病变异检出率为23.4%。nCHH患者检出基因包括CHD7、FGFR1、DUSP6、FGF8、HESX1、PROKR2、PROP1及TACR3，KS患者检出基因包括ANOS1、CHD7、FGFR1、FGF17、IGSF10、SEMA3A。

结论

1、中国人群嗅觉功能测试对CHH患者的鉴别诊断具有一定价值，可作为嗅觉测试工具用于临床及科研。

2、嗅觉主观自评测试对判断“异常”嗅觉更可靠。

3、CHH患者存在嗅觉通路结构与功能不匹配的现象，单侧嗅球不发育或发育不良患者的嗅觉功能正常，而少数双侧嗅球不发育或发育不良的CHH患者嗅觉功能也正常，即存在“无明显嗅球的正常嗅觉”现象。

4、本研究中nCHH和KS患者治疗结局无显著差异。

5、本队列中CHH患者罕见变异检出率为51.0%，致病变异检出率为23.4%。nCHH患者中检出多种与GnRH神经元迁移相关的基因致病变异，提示基因检测对于鉴别nCHH和KS的价值有限。

第二部分 NDNF基因罕见变异患者的临床特点及变异位点致病性分析

背景

先天性低促性腺激素性性腺功能减退症（Congenital hypogonadotropic hypogonadism，CHH）是一种罕见的单基因遗传病，临床主要表现为青春不发育和不育。目前已发现40多种的CHH相关致病基因。但仅有不到50%的患者可检测到致病基因变异，提示部分CHH的致病基因尚未明确。2020年Pitteloud团队在240例CHH患者中检测到4例存在神经元衍生的神经营养因子（neuron-derived neurotrophic factor，NDNF）基因罕见变异，并通过体外功能试验证实这些变异位点影响GnRH神经元迁移，提出NDNF基因是CHH患者的致病基因。但2021年的一项日本的研究在61例HH患者中仅检出9个NDNF基因同义变异，软件预测均为良性变异，提出NDNF基因致病性变异是CHH的罕见病因。因此目前NDNF基因是否为CHH患者致病基因尚待进一步研究证实。

目的

本研究将对大样本量CHH患者进行基因变异的筛查，以明确NDNF基因致病性变异的发生率；分析NDNF基因罕见变异患者临床表型，并进行体外功能实验，验证变异位点致病性。

方法

纳入2016年1月至2021年12月在北京协和医院内分泌科性腺门诊就诊的523例CHH患者，收集患者的一般临床资料。采集外周血提取白细胞DNA。采用全外显子组测序的方法，筛选存在NDNF基因罕见变异患者，用Sanger测序方法验证变异位点。利用生物信息学方法分析变异位点致病性。构建野生型及突变型NDNF真核表达载体，行体外功能实验，利用蛋白质印迹法（WB）验证变异对NDNF蛋白表达的影响。

结果

（1）本研究在523例CHH患者中检测到7例患者存在NDNF基因罕见杂合变异，检出率为1.3%。7例患者中4例合并嗅觉障碍，为卡尔曼综合征患者，3例为嗅觉正常的CHH患者。他们平均就诊年龄为20.9±4.9岁，平均基础睾丸体积2.6±1.2 ml，平均血清LH水平为0.38±0.42 U/L，FSH水平为1.07±0.61 U/L，总睾酮水平为0.78±1.17 ng/ml。

（2）7个变异位点包含3个错义变异c.31C>T（p.L11F），c.769T>G（p.F257V）和c.1430A>G（p.K477R），2个移码变异c.1656dupG（p.H553Afs*8）和c.1669dupT（p.Y557Lfs*4）和2个无义变异c.930C>G（p.Y310*）、c.1483A>T（p.K495*）。除c.31C>T外，其他变异位点均位于NDNF基因4号外显子。根据ACMG指南，p.L11F是可能良性变异，p.F257V、p.K477R、p.H553Afs*8和p.Y557Lfs*4为临床意义未明的变异，p.Y310*和p.K495*为可能致病变异。

（3）WB结果显示7个变异在细胞中均有蛋白表达，其中L11F、F257V、K477R在培养基中检测到蛋白。

结论

NDNF基因失活变异是CHH患者致病机制。本研究CHH患者队列中NDNF基因罕见变异检出率约为1.3%，致病/可能致病变异检出率率为0.4%；第4号外显子是NDNF基因变异的热点区域。NDNF基因变异患者临床可表现为先天性低促性腺激素性性腺功能减退症伴或不伴嗅觉功能障碍。

Part 1 The Differential Diagnosis Value of Olfactory Evaluation for Congenital hypogonadotropic hypogonadism

Context

Olfaction is one of the most important sensations of human beings, which impacts individuals’ social skills, relationships, well-being, and quality of life. According to the olfactory function, congenital hypogonadotropic hypogonadism (CHH) is divided into normosmic CHH (nCHH) and Kallmann syndrome (KS), which is characterized by anosmia or hyposmia. The self-reported smell and MRI of the olfactory apparatus (MRI-OA) are the most commonly used methods to evaluate olfactory function in clinical practice. However, the results of self-reported smell are not always reliable and the MRI is not fully concordant with sense of smell. Psychophysical evaluation of the olfactory function has been widely used in clinical and research settings. However, those international used tools may not be suitable for the Chinese due to differences in culture and environment. In 2019, Wen Zhou et al. developed the Chinese smell identification test (CSIT), which provides an effective tool to assess the olfactory function in the Chinese population. The efficacy of CSIT has been verified in Parkinson disease and schizophrenia, but it has not been validated in CHH patients so far.

Objective

To evaluate the olfactory function of CHH patients by the Chinese olfactory function test (COFT). To compare and analyze the diagnostic accuracy of the Self-reported Olfactory Scale (SROS) and MRI-OA for KS and to analyze the influence factors of the olfactory function in CHH patients. To compare the therapeutic effect and genetic background between nCHH and KS patients.

Methods

Patients with CHH were prospectively recruited at Peking Union Medical College Hospital between November 2020 and June 2021. The clinical data of the patients were retrospectively analyzed. Targeted next-generation sequencing was used to investigate the pathogenic gene variants. The COFT and SROS were adopted as subjective tools, and MRI-OA was adopted as the objective tool to evaluate olfactory function in CHH patients. SROS used a 5-point scale and was grouped into five categories, namely “good”, “above average”, “average”, “below average”, and “poor”. Patients in “good”, “above average” and “average” categories were assigned as normal olfactory function, while patients in “below average” and “poor” categories were assigned as abnormal olfactory function. The COFT includes three tests, the Chinese Smell Threshold Test (CSTT), the Chinese Smell Discrimination Test (CSDT), and CSIT. According to the total score of the three tests (TDI), patients were grouped into normal olfactory function (TDI≥29.25) and abnormal olfactory function (TDI<29.25). The olfactory bulb volume (OBV) and the olfactory sulcus depth (OSD) were quantitively evaluated on MRI. The olfactory function of nCHH patients and KS patients were compared, and linear regression models were built to analyze the determinants of olfactory function in CHH patients.

Results

1 Clinical characteristic: A total of 71 CHH patients were evaluated, including 69 male patients and 2 female patients. The average age of these patients was 23.6 ± 5.8 years and the body mass index (BMI) was 25.1±4.2 kg/m². The average serum LH and FSH level was 1.15±2.26 U/L and 2.33±3.65 U/L, respectively. The total testosterone (T) level was 1.55±1.52 ng/ml and the estradiol (E2) level was 20.31±11.66 pg/ml. The percentage of patients with smoking, drinking, and chronic rhinitis was 25%, 28%, and 31%.

2 Comparison of Olfactory function between nCHH and KS patients: Based on the results of COFT, 36 patients were assigned as nCHH, and the other 35 patients were KS. In nCHH group, based on SROS, patients in “good”, “above-average”, “average” and “below-average” categories were 3, 12, 20 and 1, respectively. While in KS group, patients assigned to “above-average”, “average”, “below-average” and “poor” categories were 1, 4, 2 and 28, respectively. For MRI-OA, 67% of nCHH patients showed normal olfactory apparatus, while 22% of nCHH patients showed unilateral olfactory bulb aplasia/hypoplasia [UOA(H)], and 11% of nCHH patients showed bilateral olfactory bulbs aplasia/hypoplasia [BOA(H)]. Ninety-six percent of KS patients had BOA(H), while 4% of KS patients showed normal olfactory apparatus. The nCHH patients had larger OBV and OSD than KS patients (37.58 ± 20.64 VS 2.40 ± 10.72 mm³, p=0.000; 6.82 ± 2.31 VS 2.64 ± 3.07 mm, p=0.000).

3 Influence factors for olfactory function: In linear analysis, the TDI score was used as the dependent variable and other factors, including age, BMI, follow-up time, T, E2, smoking, drinking, rhinitis, SROS, OBV, OSD, were used as independent variables. It showed the larger OBV and OSD were associated with the higher TDI.

4 Comparison of the accuracy of SROS and MRI-OA in diagnosing KS: Based on COFT, the accuracy of SROS in the diagnosis of nCHH and KS was 91.5% with a sensitivity of 85.7%, a specificity of 97.2%, a positive predictive value of 96.8% and a negative predictive value of 87.5%. The MRI-OA predicted the nCHH and KS patients with an accuracy of 92.7%, a sensitivity of 96.4%, a specificity of 88.9%, a positive predictive value of 90.0% and a negative predictive value of 96.0%.

5 The analysis of therapeutic effect: A total of 26 patients were included. Nineteen patients had gonadotropin therapy and 7 patients had pulsatile gonadotropin-releasing hormone therapy. The average follow-up time was 49.4±29.5 months and the median age was 20.5 (17.8, 26.0) years. The basal testicular volume of these patients was 1.25 (1.0, 2.0) ml. The basal level of serum LH, FSH and total testosterone (T) was 0.32 (0.20, 0.89) U/L, 0.57 (0.30, 1.80) U/L and 0.37 (0.29, 0.61) ng/ml. After treatment, the testicular volume was increased by 8.8±4.7 ml and the T level was increased by 2.6±1.6 ng/ml. Fourteen (67%) patients achieved spermatogenesis and the median period of spermatogenesis was 21.1±12.8 months. There was no difference in the therapeutic ending between nCHH and KS.

6 The analysis of gene variants: A total of 47 patients went on gene sequencing tests, while 24 (51.0%) patients were identified rare gene variants. Eleven variants were assigned as pathogenic or likely pathogenic variants. The genes detected in nCHH patients included CHD7, FGFR1, DUSP6, FGF8, HESX1, PROKR2, PROP1 and TACR3, while the genes detected in KS patients included ANOS1, CHD7, FGFR1, FGF17, IGSF10, SEMA3A.

Conclusion

1 The Chinese olfactory function test is valuable for differential diagnosis of CHH patients and can be used as an olfactory test tool for clinical and scientific research.

2 The SROS is more reliable for self-reported "abnormal" olfaction sense.

3 The structure of the olfactory apparatus does not fully match the olfactory function in CHH patients. The olfactory function of CHH patients with UOA(H) is normal. Some patients with BOA(H) also have the normal olfactory function, which is known as “Normal olfaction without apparent olfactory bulbs”.

4 In our cohort, there was no difference in the prognosis between nCHH and KS patients.

5. In our cohort, the incidence of rare variants of CHH patients was 51%, and the incidence of pathogenic variants was 23.4%. Several genes associated with GnRH neuronal migration have been identified in nCHH patients, suggesting that genetic testing cannot distinguish KS from nCHH patients.

Part 2 The Clinical characteristics and pathogenicity analysis of patients with rare variants of NDNF gene

Context

Congenital hypogonadotropic hypogonadism (CHH) is a genetic disease characterized by incomplete or absent puberty and infertility. At present, more than 40 pathogenic genes have been identified of the disease. However, only about 20%-50% of CHH patients had a molecular diagnosis, suggesting that some pathogenic genes of CHH have not been identified. In 2020, Pitteloud et al. reported 4 CHH patients with rare variants of the neuron-derived neurotrophic factor (NDNF) gene, which was detected among 240 CHH patients. And they confirmed that these variants affect the migration of GnRH neurons using in vitro functional studies, indicating that the NDNF gene is the pathogenic gene of CHH. However, in 2021, another study only detected 9 synonymous variants of the NDNF gene in 61 HH patients. All of these were predicted to be benign by the in silico tools and they indicated that NDNF gene variants are rare in CHH patients. Thus, more studies are needed to confirm whether the NDNF gene is the pathogenic gene of CHH.

Objective

This study aims to identify the incidence of NDNF gene pathogenic variants in a large sample of CHH patients, and analyze the clinical phenotype of patients with NDNF gene variants and verify the pathogenicity of the variants in vitro.

Methods

Patients with CHH admitted to Peking Union Medical College Hospital were recruited between 2016.01 and 2021.12. The clinical data of the patients were retrospectively analyzed. Whole exome sequencing was used to investigate the pathogenic gene variants, and Sanger sequencing was used to verify the NDNF gene variants. The pathogenicity of the variants was analyzed using bioinformatics methods. The wild-type and mutant NDNF eukaryotic expression vectors were constructed, and in vitro functional studies were used to analyze the pathogenicity of the variants.

Results

(1) A total of 7 patients with rare heterozygous variants of NDNF gene were detected in 523 CHH patients. The incidence of NDNF gene rare variants was 1.3%. Among the 7 patients, 4 were Kallmann syndrome patients, and 3 were normosmic CHH. The mean age of these patients was 20.9±4.9 years, and the mean basal testicular volume was 2.6±1.2 ml. The mean serum LH level was 0.38±0.42 U/L, FSH level was 1.07±0.61 U/L, and total testosterone level was 0.78±1.17 ng/ml.

(2) These variants include 3 missense variants c.31C>T (p.L11F), c.769T>G (p. F257V) and c.1430A>G (p.K477R), 2 frameshift variants c.1656dupG (p.H553Afs*8) and c.1669dupT (p.Y557Lfs*4) and 2 nonsense variants c.930C>G (p.Y310*), c.1483A>T (p.K495*). These variants were located in exon 4 of NDNF gene except c.31C>T. According to ACMG guideline, p.L11F was likely benign variant. p.F257V, p.K477R, p.H553Afs*8 and p.Y557Lfs*4 were uncertain significance variants. p.Y310* and p.K495* were likely pathogenic variants.

(3) The results of Western Blot showed that all mutants were expressed in whole cell extracts. L11F, F257V and K477R mutants were detected in the conditioned medium.

Conclusion

The loss of function variants of NDNF gene was the pathogenic mechanism of CHH. In our study, the incidence of CHH patients with NDNF gene rare variants is about 1.3%, while the incidence of likely pathogenic variants was 0.4%. Exon 4 is a hot spot of NDNF gene variants. Patients with NDNF gene variants can present with congenital hypogonadotropic hypogonadism with or without olfactory dysfunction.