王若光教授专题之［出生缺陷］（54）22q11.2缺失综合征（Digeorge综合征）
【图文整理编辑】
王桂芹 辽宁省大连市美琳达妇儿医院
李荔 中国中医科学院望
问
李秋红 黑龙江讷河市人民医院：
各位老师，这是胎停育后查的胎儿染色体高通量测序结果，发现22号染色体长臂存在微缺失，查父母染色体发现父亲22号染色体存在相同的微缺失，但父亲无异常表型。请问各位老师，他们夫妻还可以正常备孕吗？
流产物染色体高通量测序
父亲染色体高通量测序
问
王桂芹 大连美琳达妇儿医院
这个病例很有意思，父亲与胎儿染色体相同片段缺失，但男方未发现异常表现，那此次胎停育与22号染色体片段缺失有关系吗？还是请王老师给予专业指导吧！
答
高勇 北京希望组
这个好像没有见过与胎停育之间有关系的报道啊，刚才检索了下文献，也没有找到相关报道。
问
王桂芹 大连美琳达妇儿医院
确实不能认为是胎停育的绝对因素，否则男方怎么可能表现为完全正常呢！我现在疑惑的是相同的基因片段异常，表现都会完全一致吗？还是可能会有一定差异？
答
高勇 北京希望组
对于22q11.2这个deletion来说，肯定差异会比较大，属于易感的CNV，表型无法通过基因型预测。
答
郭一然 费城儿童医院
22q11.2 deletion syndrome?
答
王若光 若光医学研究中心
这个片段缺失完全覆盖了22q11.2缺失综合征（Digeorge综合征），其表型涉及从体表到心血管、颅脑神经管、泌尿系统、骨骼肢体、皮肤、颜面、唇腭、听力、甲状腺和胸腺免疫等多个方面，最重要最基本的是心血管畸形。
这位父亲的临床表型需要明确，心脏是否存在缺陷及全身各部位细微表型，建议详细收集，并注意验证父亲是否为嵌合体。
22q11.2缺失综合征目前已知的文献显示约有8%的患者死于心脏病，半数患儿死于出生后1个月之内，大多数患儿死于出生后6个月之内。
22q11.2缺失综合征胚胎死亡率及胎儿死亡率是很高的，而父亲可能是较为幸运的表型轻的。
胚胎与父方染色体缺失的片段大小一致，并不能完全排除是复发性流产的原因，因为早期胚胎发育中，心脏发育异常有所差异，程度不等，会导致胚胎淋巴水肿或胚胎水肿的机率增加，会增加胚胎死亡率。如果胚胎存在这一缺陷，且母体平时运动少，伴有代谢综合征，或潜血栓状态，或妊娠期内分泌支持及母胎循环构建存在细微异常时，足以导致胎停。
答
曹旭 胎儿医学
这个缺失很常见，一般都不会建议生下来。如果这位父亲表型正常，下次再查到胎儿是同样的缺失，解释起来要困难多了。一定要好好调查这个家系。
答
魏魏 长治妇幼
可以查父缘亲属看下有无此缺失。
遗传与疾病论坛微信公众号问《22q11微缺失综合征基因检测普及》：
22q11微缺失综合征（22q11 Deletion Syndrome, 22q11 DS）是由人类染色体22q11.21-22q11.23区域杂合性缺失引起的一类临床症候群，是迄今为止少数已明确病因的先天性疾病之一，其发病率占活产儿的1/ 6000 -1/ 4000。国外22q11 DS的流行病学研究及其发病率调查，明确了22q11 DS是一种较常见的综合征，其发病率仅次于唐氏综合征。
22q11 DS临床表型复杂多样，多数涉及先天性心脏病、上腭异常、面部不规则以及言语和学习困难，部分患者可能存在免疫缺陷和低钙，少数患有这种疾病的孩子可能有喂养问题、肾脏问题、听力丧失、生长激素不足、自身免疫性疾病、癫痫发作或骨骼畸形。在20世纪90年代初期，22q11微缺失综合征的原因被发现之前，大多数患有此综合征的儿童根据临床症状被诊断为如DiGeorge 综合征、Shprintzen综合征、Caylor心-面综合征、腭-心-面综合征及先天性面容异常综合征等。据报道，90%以上有相似临床表现的患者可检测到22号染色体长臂近端大小约3Mb碱基片段的微缺失，该区域编码约30个基因，任何一个基因缺失均可导致22q11的微缺失。
大多数22q11 DS都是新生突变，但能以常染色体显性遗传的方式遗传，根据孟德尔遗传规律，其后代遗传率50%，约8%-28%的患儿由症状轻微的双亲传递而来，父母往往病情较为轻微，可以成活至生育年龄，而后代则可能表现为严重的综合征，其再发风险远远高于普通人群。此外，该病常累及多个系统，影响患儿的身心健康，造成家庭和社会的沉重负担。因此，早期筛查和诊断可以进一步筛查家族中的可疑病例，避免有严重综合征患儿的出生，指导遗传咨询。早期发现和及时治疗亦有助于提高患儿生存率，发掘其智力潜力，增强社会适应性，改善生活质量，减轻家庭压力及社会的经济负担。
《22q11.2缺失综合征》贺建新，中华实用儿科临床杂志，2018，33(4):285-288.：
Lobdell 于1959年首次在病理解剖中注意到甲状旁腺和胸腺同时缺失。DiGeorge 于1965年开始描述婴儿出现甲状旁腺功能减低、胸腺发育不良和细胞免疫缺陷的组合，被定义DiGeorge综合征。很快，该综合征又扩展为包括特殊面容，先天性心脏病尤其圆锥动脉干异常。高分辨率的细胞遗传学发现90%的患者具有22q11区域1.5～3.0 Mb的杂合缺失。该缺失还与其他表型有关，如腭-心-面综合征、圆锥动脉干面综合征、降口角肌发育不良并室间隔缺损（Cayler syndrome）和先天中线发育异常综合征（Opitz-G/BBB），提示这些异常有共同起源。该区域内的TBX1基因单倍型功能不全也可引起DiGeorge异常（DGA）。DGA还与10p13缺失综合征、眼组织缺损、心脏异常、鼻后孔闭锁、智力迟缓、生殖、耳异常（CHARGE综合征）及糖尿病目前所产婴儿有关。胎儿酒精综合征、孕母维生素A暴露也可引起DGA表型。
1 发病机制和分子异常
目前研究认为发病机制是第三、四咽囊发育缺陷，头神经嵴细胞不能正常移行至此所致。
遗传方式大部分为散发的，来源于新发突变。6%～28%呈常染色体显性遗传，56%为母亲来源，44%为父亲来源。母亲平均怀孕年龄为29.5岁，与健康人群近似。女性22q11.2区域重组率为男性的1.6～1.7倍，该区域减数分裂时重组率增加可能是母源比例高的原因。
最初细胞遗传学研究提示DGA患者存在不同染色体间的移位，或者单体型不平衡移位22pter-q11，或者染色体内缺失del(22)(q11.21-q11.23)，因此，推测DGA的关键区域位于22q11。随着分子生物学的进步，针对细胞核型正常的DGA患者进行分子载量分析和荧光原位杂交，发现大部分患者具有22q11区域的杂合缺失。其他缺失还涉及10p13，18q21.33。也有嵌合的22号染色体单体型报道，患者具有DGA的面容特征、肌张力高、关节伸展受限、所有手指呈弯曲收缩状态。
物理图谱显示缺失位于断裂点关键区域的近端。在经典缺失区域鉴定了4个低拷贝重复（LCRs），具有97%～98%的相似性，直接参与22q11缺失的形成。非人灵长类荧光原位杂交（FISH）分析提示重复事件产生LCRs集聚在2.0亿～2.5亿年前已经出现。
DGS患者边缘区域的单体型重建发现近端染色体间的交换明显增加，另一条正常22号染色体间交换出现率为2/15，与遗传距离一致。用MLHI抗体免疫染色，人类精子减数分裂中75%交换定位于22q的远端，也反映遗传距离。与William综合征不同，FISH分析未发现LCRs附近的染色体逆转。减数分裂 Ⅰ 期异常的染色体间交换事件在22号染色体近端区域可能是缺失的原因，小的缺失更常见于家族遗传。
2 临床表现
与22q11缺失相关的症状包括180余种。单卵双生子研究表明存在个体间和家族内的变异。缺失的大小与临床表型缺乏相关性。DGA最初的三联征包括先天性无胸腺、甲状旁腺缺如及心脏异常。基于此，临床诊断标准为符合下述4项中的3项：先天性心脏病、特征面容、甲状旁腺功能低下或新生儿低钙、缺失或异常的胸腺或T淋巴细胞缺陷。
大样本研究显示临床表现为智力缺陷占92.3%，低钙占64.0%，腭异常占42.0%，心血管异常占25.8%。其他包括肥胖占35.0%，甲状腺功能低下占20.5%，听力缺陷占28.0%，胆石症占19.0%，脊柱侧弯占47.0%，皮肤异常（严重痤疮占23.0%、脂溢性皮炎占35.0%），精神分裂占22.6%。
先天性无胸腺是DGA的标志特征，但完全性DGA仅占所有患者的不到1%。临床表型同严重联合免疫缺陷病（SCID），预后恶劣。部分DGA更常见，免疫特征为部分联合免疫缺陷，临床表现为反复上呼吸道感染，下呼吸道感染少见。6月龄后出现荚膜菌引起的反复窦肺感染。伴T 淋巴细胞减少者易于出现病毒，念珠菌感染或早期感染死亡，尤其伴CD4和CD8同时减少，胸腺输出减少或甲状旁腺减低者。
心脏异常主要包括累及流出道的各种异常，包括Fallot四联症、B型主动脉弓离断、永存动脉干、右主动脉弓及右锁骨下动脉畸形。甲状旁腺发育不良导致低钙，婴儿可出现手足抽搐或惊厥。由于胸腺不发育或发育不良导致T淋巴细胞缺陷，患者感染敏感性增高。
婴儿期可出现小下颌，低耳位伴垂直半径短和耳廓异常，内眦距过宽伴短的睑裂，斜视，人中短，小口，球形鼻，方鼻尖。由于黏膜下裂或腭裂致鼻音重。身材矮小。轻到中度学习困难。精神异常见于一小部分成人患者，包括偏执精神分类和抑郁症。少见特征包括甲状腺低下、唇裂和耳聋。
自身免疫见于各个年龄段，疾病随年龄不同而不同，如青少年类风湿关节炎，特发性血小板减少性紫癜，自身免疫性溶血性贫血，鱼鳞病，白化病，炎症性肠病，成人类风湿关节炎和风湿热。
神经系统可见骶脑脊膜膨出，脊柱裂，交通性脑积水，脊髓脊膜膨出，小头，胼胝体发育不良，小脑扁桃体下疝畸形。大脑影像异常包括外侧裂区多小脑回，程度不同，经常不对称，右侧明显。
精神系统可见侵略性爆发，冷漠,精神特征如妄想，幻觉，痴呆。
泌尿生殖系统可见单侧肾不发育，肾发育不良，肾盂积水，无输尿管，原发闭经，伴血性囊肿的处女膜闭锁。
眼部包括角膜后胚胎环，扭曲的视网膜血管，眼睑悬垂，斜视，上睑下垂，弱视，倾斜的视神经。巩膜角膜弹性层膨出，小眼球，眼前段发育异常，虹膜角膜黏连。
3 实验室检查
3.1 淋巴细胞数量及功能 完全性DGA患者出生后严重T淋巴细胞减少（CD3+<0.05×109/L）。针对丝裂原的增殖反应缺失或极度减低。不典型完全性DGA婴儿可出现克隆性T淋巴细胞群，淋巴细胞数量及增殖功能可变，但原始CD4+T淋巴细胞缺乏［CD3+ CD45RA+CD62L+<0.05×109/L，或<5% CD3+T，或T细胞受体切除环（TRECs）<100/10 000 T淋巴细胞］。一些患者，B淋巴细胞减少是其特征之一，尤其婴儿期，随时间恢复正常。
3.2 体液免疫 完全性DGA患者IgG、IgA和IgM减低（尽管出生后数周内母体残留影响IgG）。部分型DGA患者抗体缺陷谱广泛，经常有轻到中度抗体受损，婴儿期明显。低IgG伴亚类缺陷亦有报道，很多患者最初低的免疫球蛋白会随年龄增长变为正常。针对多糖抗体反应缺陷较常见。
3.3 T淋巴细胞受体β链可变区（TCRVB） 不典型完全性DGA婴儿可表现皮疹和淋巴结大，临床表现类似于伴嗜酸性粒细胞增多的SCID。TCRVB呈单克隆性，同时具有无胸腺表型。
4 诊断
（1）婴儿期低钙是特征性表现，有时间段性，1年内可缓解。血甲状旁腺素（PTH）降低。（2）由于胸腺可位于其他位置或很小，不能凭外科手术、放射线或CT来诊断无胸腺，必须有分子生物学证据，如CD3+ CD45RA+ CD62L+<0.05×109/L，或<5% CD3+T，或TRECs<100/10 000 T 细胞。（3） 标准核型分析除外重要重组，或者单体型不平衡移位22pter-q11，或者染色体内缺失del(22)(q11.21q11.23)。用来源于缺失片段的探针行FISH，优先选择移位断裂点附近的探针。（4）如果没有细胞悬液或新鲜血液做核型分析，可用该区域一系列高变异探针来寻找位点缺失。目前常用的方法如多重连接探针扩增（MLPA）、比较基因组杂交（array-CGH）和拷贝数变异方法（CNV）。
5 治疗
（1）补充钙剂和1,25胆骨化醇。（2）在免疫功能健全确认前，行外科手术需输注辐照的血，避免移植物抗宿主病（GVHD）。（3）完全性或不典型完全性DGA患者需立即转移至专业免疫中心行进一步评估和治疗。启动抗卡氏肺囊虫肺炎、抗病毒、抗真菌的预防治疗和免疫球蛋白替代治疗。干细胞移植后可获得供者T淋巴细胞胸腺后的外周植入，但不能证明持续的T淋巴细胞生成。仅有数例长期存活报道，总存活率低（41%～48%），远低于其他SCID （80%）。原因主要为心脏外科情况和GVHD。目前仅有2家实验室能够进行同种异体胸腺移植，存活率72%，致死的主要原因为感染。主要的远期不良反应是自身免疫病，如自身免疫性甲状腺疾病、1系、2系或3系血细胞减少，还包括肾病综合征和自身免疫性小肠炎。（4）部分性患者主要是对症治疗，随着年龄增长病情会减轻。细菌性窦肺感染需及时治疗。可能需要预防性抗生素，尤其冬季，有的患者可能需常年预防。伴有症状性低丙种球蛋白患者或预防效果不好的患者，可能需要丙种球蛋白替代治疗。活疫苗通常是安全的，建议CD4+T<0.4×109/L时避免接种活疫苗，由于保护性抗体水平维持时间短，应定期监测抗体水平，必要时重复接种疫苗。（5） 腭裂可能在黏膜下，需仔细寻找。环咽肌失功能需要尤其关注。语言治疗，教育辅助可能需要。（6）成年患者需关注精神方面异常。
22q11.2缺失综合征是最常见的染色体缺失综合征，由于咽弓发育异常，导致胸腺、甲状旁腺、心脏动脉圆锥和面容异常。完全性患者需要紧急救治，胸腺移植为治愈方法。部分性患者对症治疗，但心脏异常、环咽肌障碍影响预后。
《22q11.2微缺失综合征》潘虹，北京大学第一医院中心实验室，中华围产医学杂志：
22q11.2微缺失综合征是人类最常见的由于拷贝数变异（copy number variations，CNVs）引起的染色体微缺失综合征，20世纪60年代第1次以表型描述。22q11.2微缺失综合征的遗传病理基础是染色体22q11.2区域有0.7~3 Mb片段缺失，累及多系统，具有表型异质性，是第二位常见的引起先天性心脏病和发育迟缓，以及最常见的腭裂综合征的遗传性疾病。男性和女性均可受累。1990年至2000年早期，由于基因检测方法的限制，活产儿中22q11.2微缺失综合征的发病率为1/6 000~1/3 000。但近年来，随着分子检测技术进步和产前诊断病例的积累，胎儿患病率约为1/1 000，如有超声结构异常，特别是有先天性心脏病的胎儿，患病率可高达1/100。22q11.2微缺失大部分为新生性，家族遗传性少见，如果为家族性，则符合常染色体显性遗传规律。亲缘来源分析提示，缺失来源于母亲略多于父亲，但并不受母亲妊娠年龄的影响，父亲的年龄与发生缺失的频率也无关。
一、分型
22q11.2微缺失综合征主要包括以下亚型：
1. DiGeorge综合征[DiGeorge syndrome，DGS，在线人类孟德尔遗传（Online Mendelian Inheritance in Man，OMIM） 188400]：DGS于1965年由Angelo DiGeorge医生首次报道，婴儿期表现为先天性胸腺和甲状旁腺缺陷，其后又有先天性心脏病，特别是流出道异常而被命名。主要指新生儿期甲状旁腺发育不良引起低钙血症、胸腺发育不良为突出表现和心脏流出道缺陷的病例。
2. 腭-心-面综合征（velo-cardio-facial syndrome，VCFS，OMIM 192430，又称Shprintzen综合征）：主要表现为儿童期腭发育异常导致的说话鼻音重，先天性心脏病，特殊面容及认知、精神、学习和行为异常等。
3. Takao综合征（又称圆锥动脉干异常面容综合征，conotruncal anomaly face syndrome，OMIM 217095）：由日本首先报告，DGS以婴儿期心脏表现为突出，该病因甲状旁腺发育不良引起的低钙血症、胸腺发育不良更为突出。
4. CATCH22：以22q11缺失引起心脏畸形、异常面容、胸腺发育不良引起的T细胞缺陷、腭裂和甲状旁腺发育不良引起低钙血症的英文首字母命名（Cardiac abnormality，Abnormal facies，T cell deficit due to thymic hypoplasia，Cleft palate，Hypocalcemia due to hypoparathyroidism resulting from 22q11 deletion）。
上述4个亚型的命名中各自突出了22q11.2微缺失综合征共同表型谱中的某些特点，文献或临床工作中最常用DGS/VCFS。1996年Daw等[8]报道发现了第2个位于染色体10p13-14的引起DGS/VCFS样表现的位点（DiGeorge syndrome/velo-cardio-facial syndrome complex-2，OMIM 601362），具体致病基因尚不明确。
二、遗传病理机制
染色体22q11.2区域基因组结构复杂，包含至少4个低拷贝重复序列（low copy repeats，LCRs）组成的多重复元件块（several large blocks of LCRs or segmental duplications），序列同源性为95%~97%，这样的结构是生殖细胞在减数分裂时易发生非等位基因同源重组（nonallelic homologous recombination，NAHR）的基础，NAHR引起22q11.2再发性缺失。染色体22q11.2区域结构如图1所示。LCR22A与LCR22D之间的NAHR产生3 Mb片段缺失，占22q11.2缺失的85%；近着丝粒端LCR22A与LCR22B、LCR22C之间的NAHR可产生1.5~2 Mb片段缺失，占22q11.2缺失的5%~10%。这些缺失导致了22q11.2缺失综合征主要临床表现。LCR22B与LCR22D和LCR22C与LCR22D之间的缺失也产生与22q11.2缺失综合征重叠表现，但外显率低于LCR22A与LCR22D之间的缺失，更常见于家族遗传性的病例。在LCR22A和LCR22D之间3 Mb的缺失区域内，包含大约90个已知或预测的基因，含基因剂量敏感基因，单倍剂量不足（haploinsuffciency）影响胚胎早期咽弓、心脏、骨骼肌和脑的形态发生。缺失的大小与表型并无明确的相关性，基因型与表型的关系目前还不十分明确。使用全基因组关联研究已发现，在22q11.2缺失区域外存在引起心脏发育异常的修饰基因。最重要的关键基因是TBX1，属于转录因子家族，编码T盒转录因子。另一个基因是DGCR8，编码DGCR8微前体复合体亚单位，是一种双链RNA结合蛋白，介导微小RNA（microRNA，miRNA）生物发生，miRNA通过与特异的mRNA结合转录抑制或降解调节靶基因。miRNA表达水平的微小变化可影响脑发育和突触可塑性，也与心血管系统和胚胎的其他发育有关。其他相关基因还有编码细胞质接受生长因子信号传递受体蛋白CRKL基因，最新报道其与肾脏缺陷最相关；编码突触体相关蛋白29 kDa基因（synaptosomal-associated protein29，SNAP29）；编码儿茶酚胺氧位甲基转移酶（catechol-O-methyltransferase，COMT）和编码Ran GTP酶结合蛋白1（Ran GTPase-binding protein1，Ranbp1）等基因。
三、临床表现
22q11.2微缺失累及多系统和器官，临床表型谱广且变异程度大，不同年龄段有不同特点。
1.胎儿期：超声检查可见先天性心脏缺陷，其中圆锥动脉干缺陷最常见，其他还有血管异常和左心发育不良。胸腺异常也是重要的产前超声征象。泌尿系统缺陷，神经系统发育问题如神经管缺陷、脑发育异常也有报道。面部异常诊断困难。
2.新生儿期及儿童期：表型变异大，轻重不一，多系统受累的表现与年龄相关。主要表现有行为问题、发育迟缓、学习困难、先天性心脏病、腭缺陷、发音鼻音重、免疫缺陷和特殊面容。80%的病例有先天性心脏病，最常见的类型为圆锥动脉干和主动脉缺陷。圆锥动脉干缺陷包括法洛四联症、肺动脉闭锁、永存动脉干和主动脉弓离断及右心室双流出道。非圆锥动脉干缺陷包括室间隔缺损、房间隔缺损和房室瓣膜异常等。法洛四联症最常见，主动脉弓离断B型特异性最强。甲状旁腺功能不足引起的低钙血症是典型症状之一，16%~70%的患儿出现低钙惊厥，新生儿期更易诊断。婴儿期肌张力低下和韧带松弛非常常见。神经系统常见认知缺陷，智商一般为70~75，数学差，但记忆力良好，视觉空间感障碍。儿童期或青少年期有精神发育相关的问题，如注意力缺陷、焦虑、抑郁和孤独症谱系障碍。大部分患者有发音和语言交流障碍，听力障碍也较常见。继发于胸腺发育不良的免疫缺陷是另一个重要表现，程度不一，主要表现为T淋巴细胞数目或功能的异常，免疫球蛋白减低见于少部分患儿，在婴幼儿期表现为易患肺炎。但随着年龄增加，免疫功能会逐渐增强。患儿常有特殊面容，包括以下特点：长脸，面颊平，眼距增宽，眼上斜，泡状上眼睑或眉弓遮盖上眼睑、内眦赘皮，鼻梁宽且低平，球状鼻或鼻翼发育不良，小下颌，耳廓畸形，小嘴，歪嘴哭面容等，但低龄儿不易识别。肾脏可有结构异常，如发育不全或发育不良、尿道梗阻及膀胱输尿管反流等。其他如消化道表现如胃食道反流、便秘和腹痛等。眼科检查中弱视、屈光不正、斜视较常见。体格发育常常落后。
3.成人期：由于临床表现不典型，成年后仍有一部分患者被诊断，主要因特殊面容、发音异常、甲状旁腺功能减退，以及心脏和精神问题。
四、诊断
22q11.2微缺失综合征的诊断一方面依据上述临床表现，特别是几个主要表现的组合，确诊依靠细胞分子遗传学检测。
常用的检测技术包括荧光原位杂交（fluorescence in situ hybridization，FISH）技术；特异性22q11.2探针TUPLE1，主要检测缺失（定性）；多重连接探针扩增技术（multiplex ligation-dependent probe amplification，MLPA）（SALSA MLPA P250 DiGeorge probemix试剂盒）和细菌人工染色体微珠标记（BACs-on-Beads，BoBs）技术在产前诊断中越来越多地被使用等。染色体微阵列分析技术（chromosome microarray analysis，CMA）不仅可定性分析，还可以定量明确缺失的片段大小。
总之，22q11.2微缺失综合征从首次报道，已经有50余年的历史，对其认识越来越深入，明确诊断对产前诊断、妊娠期指导，以及患儿治疗和康复都有重要意义。
答
谭灏文 良培基因
这例应该是LCR22 A-D的del，我看还有一篇文献，起码没有很重的表型的。
问
王桂芹 大连美琳达妇儿医院
谭老师，那这部分基因片段缺失会导致胎停育吗？
答
谭灏文 良培基因
王老师，查到一篇22q11.2缺失综合征与流产的关系，只能说轻度上升吧，我觉得没有很强证据：
Incidence of the 22q11.2 deletion in a large cohort of miscarriage samples
Melissa K. Maisenbacher,Katrina Merrion,Barbara Pettersen,Michael Young,Kiyoung Paik,Sushma Iyengar,Stephanie Kareht,Styrmir Sigurjonsson,Zachary P. Demko and Kimberly A. Martin
Molecular Cytogenetics 201710:6
Abstract
Background
The 22q11.2 deletion syndrome is the most common microdeletion syndrome in livebirths, but data regarding its incidence in other populations is limited and also include ascertainment bias. This study was designed to determine the incidence of the 22q11.2 deletion in miscarriage samples sent for clinical molecular cytogenetic testing.
Results
Twenty-six thousand one hundred one fresh product of conception (POC) samples were sent to a CLIA- certified, CAP-accredited laboratory from April 2010–-May 2016 for molecular cytogenetic miscarriage testing using a single-nucleotide polymorphism (SNP)-based microarray platform. A retrospective review determined the incidence of the 22q11.2 deletion in this sample set. Fetal results were obtained in 22,451 (86%) cases, of which, 15 (0.07%) had a microdeletion in the 22q11.2 region (incidence, 1/1497). Of those, 12 (80%) cases were found in samples that were normal at the resolution of traditional karyotyping (i.e., had no chromosome abnormalities above 10 Mb in size) and three (20%) cases had additional findings (Trisomy 15, Trisomy 16, XXY). Ten (67%) cases with a 22q11.2 deletion had the common ~3 Mb deletion; the remaining 5 cases had deletions ranging in size from 0.65 to 1.5 Mb. A majority (12/15) of cases had a deletion on the maternally inherited chromosome. No significant relationship between maternal age and presence of a fetal 22q11.2 deletion was observed.
Conclusions
The observed incidence of 1/1497 for the 22q11.2 deletion in miscarriage samples is higher than the reported general population prevalence (1/4000–1/6000). Further research is needed to determine whether the 22q11.2 deletion is a causal factor for miscarriage.
Background
The 22q11.2 deletion is the most common microdeletion in humans, and is responsible for causing the distinct range of features associated with the 22q11.2 deletion syndrome, which can include congenital heart defects, hypocalcemic hypoparathyroidism, T-cell mediated immune deficiency, palate abnormalities, and intellectual disability [1, 2]. The vast majority of 22q11.2 deletions are de novo and are caused by meiotic nonallelic homologous recombination events between low-copy repeats [3]. Although the majority (~90%) of patients share a common 2.0–3.5 Mb deletion, approximately 7% of patients have a smaller 1.5 Mb deletion nested within the common deleted region, and ~3% have rarer atypical deletions occurring outside this region [1, 3].
Population-wide estimates of the frequency of the 22q11.2 deletion have ranged from 1/4000 to 1/6000 [4, 5], although, because none of these studies were prospective, they are subject to both under-ascertainment and referral biases. True population prevalence is therefore believed to be higher [1]. Recent studies performed in prenatal cohorts have indicated a higher prenatal incidence of ≥1/1000 for the 22q11.2 deletion [6, 7, 8]. However, given that these studies were retrospective and involved a substantial percentage of cases referred for invasive diagnostic procedures (e.g., due to ultrasound anomalies), these studies are also subject to ascertainment bias. In the absence of a newborn screening program for the detection of 22q11.2 deletion syndrome, true incidence of the deletion remains unknown.
Recognizing the wide difference between prenatal and postnatal estimates of prevalence, we sought to determine the frequency of the 22q11.2 deletion in a large cohort of miscarriage samples using a single-nucleotide polymorphism (SNP)-based genotyping microarray. In addition to identifying aneuploidies, the SNP array-based method can detect subchromosomal imbalances such as the 22q11.2 deletion at a much higher resolution than traditional karyotyping (>0.5 Mb vs. >10 Mb) [9, 10, 11]. Although chromosomal microarray (CMA)-based approaches have already become routine in the pediatric setting and are increasingly being used for the testing of prenatal and adult samples [8, 12, 13], the use of CMA as a first-line test for the analysis of products of conception (POC) specimens is a relatively new application of this technology [9, 14, 15].
Results
Of 26,101 total POC specimens tested, fetal results were obtained for 22,451 (86%) cases. The remaining 3650 (14%) cases were excluded from the analysis due to having maternal cell contamination (MCC) (n = 3549, 13.6%) or incomplete results (n = 101, 0.4%; Fig. 1). A 22q11.2 deletion was detected in 15 (0.07%) of the 22,451 cases with fetal results, yielding an overall incidence of 1/1497. Two of the 15 cases with deletions were present in a pair of twins; as the twins were dizygotic, we considered the two cases as separate events.
Fifty-eight percent (13,053/22,451) of the POC specimens with fetal results were considered abnormal (i.e., had chromosome abnormalities detectable by traditional karyotyping, >10 Mb; Fig. 1). Of these, a 22q11.2 deletion was identified in three cases (incidence, 1/4351), one with a typical 3.5 Mb deletion and a finding of Klinefelter syndrome (47,XXY), one with a nested 1.5 Mb deletion and a finding of trisomy 15, and the third with a rarer 0.65 Mb deletion and a finding of trisomy 16 (Table 1). Among the 42% (9398/22,451) normal cases (i.e., no chromosome abnormalities detectable by traditional karyotyping), a 22q11.2 deletion was identified in 12 cases (incidence, 1/783; Fig. 1). One of these had an additional finding of maternal heterodisomic uniparental disomy of chromosome 17; no other cases had additional findings detectable by the SNP-based array, including other copy number variants (CNVs) (resolution of >0.5 Mb; Table 1). Of the 12 'normal’ cases that presented with the 22q11.2 deletion, three cases (two in the set of twins) had 0.72 Mb deletions; the remaining nine cases had the common 2.4–3.5 Mb deletion (Table 1, Fig. 2). Each case is schematically represented against the hg18 browser ideogram for chromosome 22. The common A, B, C, and D low-copy repeats (LCRs) for the 22q11.2 deletion are shown. There is one small region (190 Kb) between LCR-B and LCR-C that is common to all 15 cases (Fig. 2). The only gene in this region is SCARF2.
aBased on the presence or absence of chromosome abnormalities that were detectable at the resolution of traditional karyotyping (i.e., ≥10 Mb)
bNon-identical twin gestation
Eighty percent (12/15) of the 22q11.2 deletions identified were present on the maternally inherited chromosome (Table 1). Because mothers were not consented for analysis of their samples outside the scope of fetal results, it is unknown how many deletions on the maternal chromosome were de novoversus inherited deletions. Furthermore, the status of deletion on the paternal chromosome (cases 13,14 and 15) cannot be determined because paternal samples were not analyzed.
There was no significant difference in mean maternal age (32.6 years; range, 19.0–41.3 years; standard deviation [SD], 5.9) for the 15 cases with the deletion compared with the entire analysis cohort (35.0 year; range, 15.0–52.0 year; SD, 5.3) (p = 0.38).
Gestational age information, which was available for 10 of the 15 POC specimens with the 22q11.2 deletion, ranged from 6.4 to 13.6 weeks. Among those, 80% (8/10) of the miscarriages occurred in the first trimester and the remainder occurred early in the second trimester. However, because gestational age at loss was available for less than 20% of the study cohort in total, no inference could be drawn about the relationship between gestational age and presence of the 22q11.2 deletion.
Discussion
Although gross chromosomal imbalances have been shown to be present in approximately two-thirds of all first trimester miscarriages [9, 16], relatively little information on the frequency of the 22q11.2 deletion in miscarriage is available. To gain insight into this question, we reviewed a large dataset of microarray results from POC specimens for the presence of 22q11.2 deletions. Among 22,451 miscarriage samples with fetal results, we observed the 22q11.2 deletion at an overall incidence of 1/1497, which is significantly more frequent than the reported population prevalence of 1/4000–1/6000 [4, 5]. Possible explanations for this discrepancy include the published general population prevalence potentially being an underestimate, and/or that some affected fetuses have major anomalies that lead to fetal demise [17, 18]. For example, it is known, that cardiac defects are the primary cause [87%] of mortality in infants with the deletion [1]. Consistent with this, other recent studies have found a higher frequency of the 22q11.2 deletion in prenatal cohorts and stillbirths (1/233 to 1/946) [6, 7, 8, 13, 19].
The high variability of phenotypic presentation of the 22q11.2 deletion is well known, but it is not fully understood why some individuals with the 22q11.2 deletion are more severely affected than others [20, 21]. Recent evidence suggests a role for additional genetic variants that modify risk for congenital heart defects in some patients with the deletion [22]. Conversely, other studies have found no evidence of an increase in novel genome-wide CNVs in patients with the 22q11.2 deletion [23]. However, there is still the potential for other undetectable mutations and /or environmental factors to impact the severity of cases with the 22q11.2 deletion as was proposed to modify clinical severity of the 16p21.2 deletion and termed a 'two-hit hypothesis’ [24].
In this study, we found the incidence of 22q11.2 deletions in samples with additional abnormalities was significantly lower than its incidence in samples without them (incidence, 1/4351 vs. 1/783; p < 0.01). A previous study comprising a subset of this study’s cohort (the first 2392 samples of 26,101 total) found that the distribution of all subchromosomal CNVs, including the 22q11.2 deletion, was similarly skewed, with a higher proportion of variants present in cytogenetically normal samples [9]. The authors of that study suggested that the higher incidence of copy-number changes in cytogenetically normal samples suggested that these findings “likely contributed to miscarriage causality” [9]. That study also found that the frequency of concurrent double aneuploidies in miscarriage cohorts was lower than was expected based on the reported incidence of each individual anomaly [9]. Presumably more than one significant genetic abnormality leads to selective pressure against implantation, or against continuation of pregnancies beyond 6 weeks. Likewise, if the 22q11.2 deletion were an independent cause of miscarriage, it would likely be observed less frequently in cases with additional chromosomal anomalies.
The combination of the 22q11.2 deletion’s high phenotypic variability, high frequency in stillbirth and prenatal samples and high incidence found in this study suggests that the published general population incidence may be an underestimate and/or that the 22q11.2 deletion could be a causal factor for miscarriage.
Other findings of this study were consistent with previous studies. First, the majority (67%) of 22q11.2 deletions identified in this study had the common LCR-A to LCR-D, 2.0–3.5 Mb-sized, deletion [1]; the remaining deletions ranged in size from 0.65 to 1.5 Mb. Second, 80% (12/15) of 22q11.2 deletions in this study were on the maternally inherited chromosome, similar to previous studies that found a higher proportion of 22q11.2 deletions on the maternal chromosome (76% in prenatally diagnosed cases and 60% in postnatal cases) [17, 23]. Although the exact reason for this skew toward maternal origin is unknown, several factors have been considered, including an increased rate of recombination in this region in female germ cells, a mediating role of the Y chromosome due to homology with the 22q11.2 region, reduced fitness for males with the deletion, and a skewed inheritance pattern in male vs. female offspring [23, 24, 25].
Interestingly, a region containing one known gene, SCARF2, was common to all fifteen 22q11.2 deletion cases. Mutations in SCARF2 are associated with Van Den Ende-Gupta syndrome, an autosomal recessive condition characterized by craniofacial and skeletal malformations [26]. SCARF2 is thought to play a role in cell signaling pathways but it is not well characterized. It may play an important role the development of various organ systems [26].
Approximately 7% of individuals who have a child with a 22q11.2 deletion also have the deletion [1]. Therefore, identification of a 22q11.2 deletion in a miscarriage specimen should prompt evaluation of the parent(s) to determine whether there are clinical and reproductive risks for the parents. Additional medical screening and care is recommended for all individuals with the 22q11.2 deletion, including adults diagnosed later in life [27, 28]. For adults with the deletion, the recurrence risk for each subsequent pregnancy is 50%. Parents with an affected child, but who do not have the deletion themselves, may have a slightly elevated (~1%) recurrence risk in each pregnancy, likely due to gonadal mosaicism [1]. Thus, knowing if a POC specimen has the 22q11.2 deletion can lead to better recurrence risk counseling and future pregnancy management for families.
The results of this study show a higher 22q11.2 deletion incidence in the miscarriage population than literature reports of the live-born population. This implies that either the population prevalence of the 22q11.2 deletion in the live-born population has been underestimated or that the 22q11.2 deletion is a causative factor in miscarriage. Further studies are needed to determine the true incidence of the 22q11.2 deletion syndrome in the live-born and prenatal population. Additionally, the advantages of chromosomal microarray analysis support its use as a first-line test for the analysis of POC samples.
Methods
A retrospective review of 26,101 consecutive fresh POC specimens received by a single Clinical Laboratory Improvement Act (CLIA)-certified, College of American Pathologists (CAP)-accredited laboratory for clinical miscarriage testing over a 6-year period (April 2010–May 2016) was performed to determine the incidence of the 22q11.2 deletion. A maternal blood sample was requested with each specimen. For each patient, information about maternal age, gestational age, egg donor use, and indication for testing was requested, and all samples were de-identified before review. Additional clinical information including previous pregnancy history, method of conception, pregnancy records and family history was not collected by the laboratory. This study was granted a waiver of the requirement for documentation of informed consent by the institutional review board (E&I ID# 15148-01).
Data on all clinical findings were previously published for the first 2392 cases of this data set, which included one case with the 22q11.2 deletion [9]. Additionally, all syndromic CNVs in karyotypically normal samples from 17,424 cases in this data set were presented at the American Society of Reproductive Medicine 2016 meeting. This included 6 cases with the 22q11.2 deletion [29].
Upon receipt at the commercial reference laboratory, POC specimens were processed by separating chorionic villi from maternal decidua using a standardized technique [30]. These, along with the maternal samples, were genotyped using lllumina CytoSNP-12b microarrays, which measures approximately 300,000 SNPs across the genome (roughly one every 10 kb) according to the manufacturer’s instructions (Illumina, Inc., San Diego, CA). After a genomic sample is run on a SNP array the results must pass an in-house quality control test before further analysis is done. Genotyped samples were analyzed for DNA copy number, uniparental disomy (UPD), parental origin of chromosome abnormalities, and maternal cell contamination (MCC) using the previously described proprietary Parental SupportTM algorithm [31]. In short, the allele ratios are calculated for each locus across a chromosome, and the clustering of allele ratios is indicative of the copy number for that chromosome. Comparison of the SNP identities between the maternal and POC data is used to identify maternal cell contamination, the parental origin of aneuploidy and unbalanced chromosome segments. Since parents were not consented for testing of their DNA nor is the father genetic material tested, we are not able to determine whether 22q11.2 deletions are inherited or de novo. Samples were classified as 'abnormal’ or 'normal’ based on the presence or absence of chromosome abnormalities that could be detected at the resolution of traditional karyotyping (i.e., >10 Mb). Coordinates for copy number variants (CNVs) >0.5 Mb were entered into the NCBI36/hg18 genome browser to determine clinical significance based on genes affected and previously reported overlapping CNVs. These coordinates were then converted to the GRCh38/hg38 assembly coordinates using the Lift Genome Annotations tool. To determine the statistical significance of maternal age values and of the incidence of the 22q11.2 deletion, independent two-sample t-tests were performed.