Going back in time: Prenatal presentations of postnatal genetic diagnoses made in a neonatal intensive care unit

Abstract

Objectives:

Prenatal genetic diagnosis can impact care across the perinatal continuum; however, prenatal suspicion for genetic disorders may be complicated by incomplete knowledge of fetal rare-disease phenotypes. Here, we describe the prenatal presentations of a cohort of infants with rare genetic conditions who were diagnosed postnatally in a neonatal intensive care unit (NICU), to characterize prenatal presenting features and evaluate why the diagnosis was not identified prenatally.

Methods:

Retrospective cohort study of infants born over a 7 year period (2017–2023) who were admitted to a Level IV NICU and received a postnatal genetic diagnosis prior to 1 year of age. We identified which of these infants had been imaged prenatally at our Maternal Fetal Care Center (MFCC): an opportunity for prenatal genetic diagnosis. Clinical data were abstracted from the medical record.

Results:

51 cases met inclusion criteria. Nine of the 51 infants were not strongly suspected to have a genetic syndrome prenatally when seen at the MFCC, as evidenced by lack of prenatal genetics consultation and lack of documented suspicion for a genetic etiology. These cases largely had absent or uncertain prenatal phenotypes. In most cases (42/51, 82.4%), prenatal diagnostic testing was not pursued even if offered. Overall, postnatal diagnoses, of which there was one dual diagnosis, were made by karyotype/FISH (11/52, 21.1%), microarray (8/52, 15.4%), gene panel/targeted testing (17/52, 32.7%), or exome sequencing (16/52, 30.8%).

Conclusions:

Our data illustrate the challenges in fetal phenotyping and support a broad approach to prenatal testing to facilitate early genetic diagnosis which may meaningfully impact postnatal care.

Introduction:

Prenatal genetic diagnosis can critically influence perinatal care, providing important information pertinent to medical management in both the pre- and postnatal periods^1–5. However, current understanding of prenatal presentations of genetic disease is limited, as many genetic diagnoses are not suspected or diagnosed until after birth, and comprehensive fetal imaging (including high-resolution ultrasound, fetal magnetic resonance imaging (MRI), and fetal echocardiogram) are not routinely obtained. Even with advanced imaging techniques, phenotypes may not point to a specific genetic diagnosis. Due to systemic challenges in linking pediatric and maternal medical records, including often different obstetrics and pediatric care institutions for mother and newborn, there are limited datasets to understand prenatal phenotypic presentations of postnatally diagnosed genetic disease. Here, we assess a retrospective cohort of infants seen in a large, level IV NICU with a broad catchment and referral base who were also seen prenatally in the Maternal Fetal Care Center (MFCC) at the same hospital.

Fetal samples for genetic testing are typically collected through invasive procedures such as amniocentesis or chorionic villus sampling (CVS). Prenatal genetic diagnostic testing such as chromosomal microarray (CMA), exome sequencing (ES), or genome sequencing (GS) is increasingly employed to aid in prenatal counseling regarding diagnosis and prognosis. Several local and national groups have released statements on implementation of prenatal genomic sequencing (ES/GS),^6,7 however utilization and payor reimbursement remain inconsistent⁸. In a joint position statement in 2018, the International Society for Prenatal Diagnosis (ISPD), Society for Maternal Fetal Medicine (SMFM), and Perinatal Quality Foundation (PQF) recommended prenatal sequencing as a diagnostic test “in the setting of a research protocol” or on a case-by-case basis when a genetic disorder is suspected.⁹ ISPD released an updated position statement in 2022 suggesting that ES is beneficial for pregnancies with a “single major anomaly or multiple organ system anomalies”, either after a non-diagnostic CMA or if a “‘pattern’ [of anomalies] strongly suggests a single gene disorder”.¹⁰

Fetal presentation is generally considered suspicious for a genetic condition based upon available phenotypic information alongside family history and screening results (i.e., cell-free non-invasive prenatal and carrier screening), if available. However, the spectrum of prenatal phenotypes associated with many well-delineated genetic disorders remain poorly defined^11,12. This can be due to incomplete phenotyping, imaging at a gestational age where the phenotype has not emerged, or prenatal phenotype not capturable or uncertain given the employed imaging technologies. For example, most functional (e.g. neurocognitive) changes are not detectable in the prenatal period, though in some cases structural features can predict a neurocognitive diagnosis¹³. In recent studies, prenatal ES implemented for any fetal anomaly after a negative evaluation for chromosomal abnormalities, including an apparently isolated structural anomaly, has obtained an additional diagnostic yield of 10%, and identified a possible diagnosis in an additional ~20% of cases^2,4. Thus, use of broad sequencing such as ES or GS on a case-by-case basis as opposed to systematically in the prenatal period may lead to missed opportunities for prenatal diagnosis.

In order to understand the spectrum of fetal presentations of rare genetic disorders, and explore the potential benefits of expanded prenatal sequencing, we reviewed the diagnostic odyssey of infants who received postnatal genetic diagnoses in our neonatal intensive care unit (NICU) after being seen prenatally in our MFCC. The questions that motivated this work were:

1. For critically-ill infants with a postnatal genetic diagnosis, what, if any, was the identified prenatal phenotype?

2. Which fetal cases may warrant broad genetic workups (including exome/genome sequencing) in the prenatal period?

Methods:

Inclusion criteria for our retrospective observational cohort included: live birth between 2017–2023 (inclusive), admission to the level IV NICU at our institution, genetics consultation within the first six months of life, postnatal genetic diagnosis within the first year of life, and a history of fetal evaluation in our MFCC (Figure 1). Infants with prenatal genetic diagnoses were excluded from our cohort. Prenatal assessment in our MFCC requires referral from an obstetrics provider or prenatal genetic counselor due to a suspicion of a fetal abnormality, concern based on family history (including maternal health conditions that may impact fetal development), or other abnormal screening/testing in pregnancy. The course of evaluation in the MFCC is determined by clinical staff based on the referral indication and may include high-resolution ultrasound, fetal MRI, fetal echocardiogram, image review, and counseling by pediatric subspecialists based on identified fetal anomalies. A nurse or prenatal genetic counselor is typically present during subspecialty counseling to integrate information and provide continuity across visits. Over the course of the study period, the availability of medical geneticists and genetic counselors in the MFCC has increased; medical geneticists are consulted in cases of multiple fetal anomalies, concern for a genetic syndrome based on family history, or abnormal genetic testing and/or screening. After evaluation in the MFCC, several pregnancy management options are available to patients including high-risk pregnancy monitoring, delivery at an institution with a high-level NICU, and termination of pregnancy (as per legal limitations which are constantly evolving). After delivery, infants requiring level IV NICU care may be transferred to our NICU.

Figure 1. Inclusion criteria for study cohort.

A. In our tertiary care center, pregnant patients with high-risk pregnancies are referred for evaluation at the Maternal Fetal Care Center (MFCC). After birth, some infants in need of critical care are admitted to the Boston Children’s Hospital Neonatal Intensive Care Unit (BCH NICU), and some receive a postnatal genetic diagnosis. For the purposes of this retrospective review, we first considered infants with a genetic diagnosis obtained postnatally who were seen in the BCH NICU for inclusion. We further filtered participants based on whether they were seen for prenatal consultation in the MFCC.

B. Participants were included in our cohort (n=51) if they: 1) were admitted to the BCH NICU, 2) were born between 1/1/2017 and 12/31/2023, 3) had a genetics consult within the first 6 months of life, 4) had a postnatal genetic diagnosis within the first 12 months of life, and 5) parent(s) were seen in the Maternal Fetal Care Center (MFCC) for prenatal consultation.

Genetic diagnosis was defined as identification of a clinically significant variant as determined by the clinical team, which included board-certified clinical geneticists and genetic counselors. Variant classification was initially performed by the clinical diagnostic laboratories performing testing. In most cases, these were variants defined as pathogenic or likely pathogenic genetic using ACMG/AMP criteria¹⁴ by the clinical diagnostic laboratory. Cases with variants of unknown significance (VUS) were included if they were determined to be clinically diagnostic by both the clinical team and upon retrospective review.

Manual review of electronic medical records (EMR) was completed to determine inclusion criteria and to abstract additional information, including: gestational age at MFCC consult and delivery, prenatal specialist consults, prenatal imaging studies, prenatal clinical assessment, prenatal and postnatal genetic testing, and genetic diagnosis. Prenatal suspicion for a genetic etiology was determined based on review of clinical documentation from prenatal consultations with pediatric specialists. All cases and variants were reviewed by at least one board-certified clinical geneticist and genetic counselor (M.H.D., M.H.W., M.S., T.P.). For cases without documented prenatal suspicion for a genetic etiology, imaging was re-reviewed with a pediatric radiologist (J.E.) to determine if there were features consistent with the genetic diagnosis. A subset of these cases did have non-diagnostic prenatal genetic testing related to the reason for referral. After repeat imaging and in light of non-diagnostic testing results, it was felt that the etiology in these cases was unlikely to be genetic in origin. This retrospective evaluation was approved by the Institutional Review Board at Boston Children’s Hospital.

Results:

A total of 51 cases were identified meeting inclusion criteria. The median gestational age of prenatal referral was 23 weeks (interquartile range, IQR: 20–29 weeks). Participants’ overall demographics were representative of the BCH NICU population (Table 1). Six infants meeting inclusion criteria (6/51, 11.8%) had VUS as interpreted by the clinical laboratory that were considered clinically significant by both the clinical team and upon retrospective review.

Table 1:

Demographic characteristics of participants (n=51)

GA at MFCC Referral, in weeks GA (median (IQR))	23 (20–29)
GA at delivery, in weeks GA (median (IQR))	37 (36–38)
Proband sex assigned at birth	F	25 (49%)
	M	26 (51%)
Birthing parent reported race	White	28 (54.9%)
	Unknown	14 (27.5%)
	Other	5 (9.8%)
	Black	3 (5.9%)
	Asian	1 (1.9%)
Birthing parent reported ethnicity	American	20 (39.2%)
	Unknown	17 (33.3%)
	Other	14 (27.5%)
Birthing parent primary language	English	44 (86.3%)
	Spanish	3 (5.9%)
	Unknown	2 (3.9%)
	Other	2 (3.9%)

Prenatal Evaluation

A medical geneticist was consulted prenatally for 15/51 (29.4%) of cases. The remaining were seen prenatally by other subspecialists related to the referral request or anomaly observed. In most cases (42/51, 82.4%), diagnostic genetic testing was not pursued prenatally, even if discussed during consultation. Based on available retrospective documentation, prenatal diagnostic testing was discussed in at least 33/51 (64.7%) cases, though the extent of this discussion is not always clearly documented. Reasons documented for not pursuing diagnostic testing included: deferral due to gestational age; declining an invasive procedure; belief that information would not impact pregnancy management; high risk cell-free DNA screening, and other personal beliefs. Of 15 participants who had a prenatal genetics consult, 3/15 (20.0%) opted for prenatal diagnostic testing. Of 36 participants without prenatal genetics consultation at our institution, 6/36 (16.7%) had prenatal diagnostic testing via their referring obstetrical provider.

Of the 9 cases with prenatal diagnostic testing, 6/9 (66.7%) had prenatal karyotype or aneuploidy FISH and 8/9 (88.9%) had a prenatal chromosomal microarray (CMA) performed. In one case microarray was performed both prenatally and postnatally and was non-diagnostic. ES was not performed prenatally in any of the 51 cases we reviewed, though it was discussed and declined in one case. The 9 cases with prenatal invasive testing all had non-diagnostic results, in line with inclusion criteria requiring a postnatal diagnosis.

Genetic evaluation:

When including testing performed prenatally, testing included: FISH and karyotype for 21/51 cases (41.2%); CMA for 28/51 (54.9%), targeted sequencing/analysis for 20/51 (39.2%), and ES for 16/51 (31.4%). The diagnostic yield for each test—defined as the number of individuals with new diagnoses for each test—was: 11/21 (52.4%) for FISH and karyotype; 8/28 (28.6%) for CMA; 17/20 (85%) for targeted sequencing, and 16/16 (100%) for ES (Table 2). ​ Across the cohort, including one dual diagnosis, 11/52 (21.1%) of diagnoses were made by FISH and karyotype; 8/52 (15.4%) of diagnoses were made by CMA; 17/52 (32.7%) were made by targeted sequencing/analysis; and 16/52 (30.8%) were made by ES. All but one of the genetic diagnoses obtained postnatally could have been made by exome sequencing plus microarray or ES with CNV analysis—Case 22, postnatally diagnosed with Beckwith-Wiedemann syndrome due to hypomethylation of IC2, could only have been diagnosed using methylation profiling. In addition, 24/51 (47.1%) of cases could theoretically be identified on current clinically available cell-free DNA screening available in the United States (Table 3).

Table 2:

Diagnostic methodologies and diagnostic yield in our cohort

Methodology	FISH and karyotype	CMA	Targeted sequencing /analysis	Exome sequencing
How many individuals from our cohort received the test? † *	21/51 (41.2%)	28/51 (54.9%)	20/51 (39.2%)	16/51 (31.4%)
Diagnostic yield of test in our cohort	11/21(52.4%)	8/28(28.6%)	17/20 (85%)	16/16 (100%)

^†percentages may not total 100, as each individual could have more than one test

^*Includes prenatal diagnostic tests

Table 3:

Summary of chart review of critically-ill infants with genetic diagnoses seen in the MFCC.

Case	GA at Referral (weeks)	Phenotypic Sex	MFCC Possible Diagnosis/Phenotype Description	Genetic Diagnosis	MIM	Gene Affected	Variant Type	Molecular Diagnosis Technology	Could be detected by current clinically available cfDNA methods?
1	15	F	Concern for Stickler Syndrome; mild micrognathia	Stickler Syndrome, type II	604841	COL11A1	Partial gene deletion	Targeted	Y
2	16	F	Concern for aneuploidy; increased nuchal translucency, absent nasal bone, sacral spine differences, ventricular septal defect	Trisomy 18	NA	Multiple	Extra Chromosome	Karyotype	Y
3	18	F	Concern for genetic syndrome; moderate, symmetric, and apparently isolated ventriculomegaly	Basal cell nevus syndrome I (Gorlin syndrome)	109400	PTCH1	Indel	Targeted
4	19	M	Concern for genetic syndrome; congenital diaphragmatic hernia, left hydronephrotic kidney, right dysplastic kidney, inferior vermian hypoplasia, hypoplasia of the corpus callosum, hypertelorism, single umbilical artery, prominent intracranial vein, mild skin thickening of scalp/thorax	Hyperphosphatasia with impaired intellectual development syndrome 1 (Mabry syndrome)	239300	PIGV	Single Nucleotide Variant	Exome Sequencing
5	19	M	Concern for genetic syndrome possibly LCAM1; ventriculomegaly, largely absent corpus callosum, adduction of both thumbs, suspected rocker-bottom feet, fetal growth restriction, mild micrognathia	Encephalopathy, progressive, early-onset, with brain atrophy and spasticity	617669	TRAPPC12	Single Nucleotide Variant	Exome Sequencing
6	19	F	Concern for genetic syndrome; agenesis of corpus callosum, thinned dysmorphic brainstem, ventriculomegaly, microphthalmia with hypertelorism, nuchal skin thickening	Hydrocephalus, congenital, 3, with brain anomalies	617967	WDR81	Indel	Exome Sequencing
7	19	M	No concern for genetic syndrome; thickening and mild increased T2 signal in the nuchal soft tissue	Shpritzen-Goldberg syndrome	182212	SKI	Single Nucleotide Variant	Exome Sequencing
8	19	M	Concern for genetic condition; right club foot, hands in clenched position, ulnar deviation of 5th digits at metacarpophalangeal joint, flexion of interphalangeal joints	Loeys-Dietz syndrome 2	610168	TGFBR2	Indel	Exome Sequencing
9	20	F	Concern for genetic condition; double outlet right ventricle, likely umbilial hernia, polyhydramnios	22q11 deletion syndrome (1.5Mb)	NA	Multiple	Microdeletion	Microarray	Y
10	20	M	No concern for genetic syndrome; mild dolichocephaly, solitary left kidney, hypospadias	Fraser syndrome 1	219000	FRAS1	Single Nucleotide Variant/Indel	Exome Sequencing
11	20	M	Likely Trisomy 21, cffDNA high risk, increased nuchal translucency	Trisomy 21	NA	Multiple	Extra Chromosome	Karyotype	Y
12	20	M	Suspicious for Osteogenesis Imperfecta; brachycephaly, flattened nose, enlongated philtrum, small chest, bent and fractured long bones	Autosomal recessive OI type VIII	610915	P3H1	Single Nucleotide Variant	Targeted
13	20	M	Concern for genetic syndrome; severe hypogenesis of corpus callosum, likely midline fusion of hypothalamus and midbrain, moderate thinning of the temporal lobes and parasagittal occipital lobes, under sulcation of the sylvian fissures, dysmorphic enlargement of the lateral ventricles, mild prominence of the retrocerebellar CSF space, microcephaly	Rett Syndrome, Congenital Variant	613454	FOXG1	Single Nucleotide Variant	Exome Sequencing
14	20	F	Likely Trisomy 21; cffDNA high risk, complete AV cancal, duodenal atresia or stenosis with mild polyhydramnios	Hyperinsulinism-hyperammonemia syndrome and Trisomy 21	NA/606762	Multiple/GLUD1	Extra Chromosome/Single Nucleotide Variant	Karyotype/Targeted	Y
15	21	F	Concern for genetic syndrome; severe bilateral urinary tract dilation, severe oligohydramnios, mild thickening of dorsal neck soft tissues	Alagille syndrome type 2	610205	NOTCH2	Single Nucleotide Variant	Exome Sequencing
16	21	M	No concern for genetic syndrome; prominence of the retrocerebellar CSF space; moderate right pleural effusion	Noonan Syndrome I	163950	PTPN11	Single Nucleotide Variant	Targeted	Y
17	21	M	No concern for genetic syndrome; at 22w GA: retrognathia; mandibular hypoplasia; inferior vermian hypoplasia; prominence of the cisterna magna; at 25w GA: findings had resolved	TARP Syndrome	311900	RMB10	Single Nucleotide Variant	Exome Sequencing
18	21	F	Concern for genetic syndrome; severe hydroureteronephrosis, moderate micrognathia, cleft secondary palate and glossoptosis, low set ears, mild hypertelorism, mild midface underdevelopment, mild bilateral ventriculomegaly	Partial trisomy 8 with mosaicism	NA	Multiple	Partial mosaic trisomy	Microarray	Y
19	21	M	Concern for genetic condition; Increased nuchal translucency, small ventricular septal defect in the membranous septum	Developmental and epileptic encephalopathy 7	613720	KCNQ2	Single Nucleotide Variant	Targeted
20	21	M	Concern for genetic condition; Right congenital diaphragmatic hernia, liver herniated into right chest	22q11 deletion syndrome (Distal Deletion)	NA	Multiple	Microdeletion	Microarray	Y
21	21	M	Concern for genetic syndrome; Dandy-Walker malformation, partial agenesis of the corpus callosum, thinned and dysmorphic brainstem, rocker-bottom feet, abdominal and hepatic cysts, bladder defect, mild mandibular protrusion	Coffin-Siris Syndrome 4	614609	SMARCA4	Single Nucleotide Variant	Exome Sequencing
22	22	F	No concern for genetic syndrome; normal fetal echo of recipient triplet in twin-twin transfusion syndrome	Beckwith-Wiedemann Syndrome	130650	Multiple	Hypomethylation of IC2	Targeted
23	22	F	Concern for genetic syndrome; closed fetal hands, low-set posteriorly rotated ears, small retrocerebellar space	Trisomy 18	NA	Multiple	Extra Chromosome	Karyotype	Y
24	22	M	Concern for genetic syndrome; moderate ventricular size discerepency, mildly hypoplastic mitral valve, aortic valve, and ascending aorta, moderately hypoplastic aortic arch	14q duplication	NA	Multiple	Microduplication	Microarray	Y
25	23	M	Concern for genetic syndrome; heterotaxy with left-sided stomach, interrupted inferior venous cava with azygous continuation	Heterotaxy, visceral, 7, autosomal	616749	MMP21	Single Nucleotide Variant	Exome Sequencing
26	23	F	Concen for genetic syndrome; polyhydramnios, right clubfoot, enlongated philtrum, conoventricular septal defect, right hydroureteronephrosis, globally abnormal brain MR	Multiple congenital anomalies-hypotonia-seizures syndrome 1	614080	PIGN	Partial gene deletion	Exome Sequencing
27	23	M	Concern for skeletal dysplasia; shortened long tubular bones and pulmonary hypoplasia, possible low position of the conus	Spondyloepiphyseal dysplasia	616583	COL2A1	Single Nucleotide Variant	Targeted	Y
28	24	F	Multiple congenital anomalies, no mention of concern for genetic condition	Trisomy 18	NA	Multiple	Extra Chromosome	Karyotype	Y
29	24	M	Concern for genetic condition; myelomeningocele, L5/S1	Pseudohypoparathyroidism (Albright Hereditary Osteodystrophy)	13580	GNAS1	Single Nucleotide Variant	Exome Sequencing
30	25	M	Concern for genetic condition; agenesis of the corpus callosum, colpocephaly, cerebellar vermis hypoplasia	5.5Mb deletion including ZEB2 (Mowat-Wilson Syndrome)	235730	ZEB2	Microdeletion	Microarray
31	26	M	No concern for genetic syndrome; echogenic kidney, question of unilateral renal agenesis, fetal growth restriction, echogenic bowel, dilated inferior vena cava, and abnormal doppler	Helsmoortel-van der Aa syndrome	615873	ADNP	Indel	Exome Sequencing
32	26	M	Concern for genetic syndrome; interrupted aortic arch and a ventricular septal defect	22q11.2 deletion syndrome (2.8 Mb deletion)	NA	Multiple	Microdeletion	Microarray	Y
33	26	F	Concern for genetic syndrome; bladder exstrophy, left microphthalmia with coloboma	Focal dermal hypoplasia (Goltz syndrome)	305600	PORCN	Single Nucleotide Variant	Targeted
34	27	F	Concern for genetic syndrome; oligohydramnios, growth restriction, complete right cleft lip and palate, bilateral choanal atresia, prominence of extra-axial spaces, short corups callosum, bilateral post-axial polydactyly, abnormal positioning of both feet, solitary small right low lying kidney, mildly hypoloastic aortic valve annulus	Smith-Lemli-Opitz Syndrome	270400	DHCR7	Single Nucleotide Variant	Exome Sequencing
35	27	F	No concern for genetic syndrome; pulmonary atresia with intact ventricular septum	22q11.2 duplication (2.8 Mb)	608363	Multiple	Microduplication	Microarray	Y
36	28	F	No concern for genetic syndrome; mild polyhydramnios; fetus small for GA; inferior vermian hypoplasia	Congenital disorder of glycosylation, type Ig	607143	ALG12	Single Nucleotide Variant/Indel	Exome Sequencing
37	28	M	Concern for Autosomal Recessive Polycystic Kidney Disease; oligohydramnios, echogenic enlarged kidneys bilaterally	Polycystic kidney disease 4, with or without hepatic disease	263200	PKHD1	Single Nucleotide Variant	Exome Sequencing
38	28	M	No concern for genetic syndrome; normal fetal echocardiogram (in the setting of maternal T1DM)	Galactosemia	230400	GALT	Single Nucleotide Variant	Targeted
39	29	F	Concern for genetic syndrome; congenital diapragmatic hernia, inferior vermian hypoplasia, polymicrogyria, cerebral dysgenesis, corpus callosum hypogeneis, abnormal sulcal gyral morphology, small left ventricle, bilateral enlarged kidneys, duplicated collecting systems, postaxial polydactyly, micrognathia, microphthalmia, and micrognathia	Trisomy 13	NA	Multiple	Extra Chromosome	Karyotype	Y
40	30	M	Concern for tuberous sclerosis; large left-ventricular mass with additional small cardiac masses, cortical tubers in brain	Tuberous Sclerosis Complex	613254	TSC2	Single Nucleotide Variant (mosaic)	Targeted	Y
41	31	F	Concern for Trisomy 21; polyhydramnios, absent nasal bone, duodenal atresia, possible coaractation of the aorta with mild ventricular size discrenpancy, mild arch hypoplasia and antegrade diastolic flow; ventricular septal defect	Trisomy 21	NA	Multiple	Extra Chromosome	Karyotype	Y
42	32	M	Concern for genetic syndrome; unilateral complete left cleft lip, alveolus, and secondary palate, esophageal atresia with tracheo-esophageal fistual, borderline lateral ventriculomegaly, slightly shortened phallus, large membranous ventricular septal defect	CHARGE syndrome	214800	CHD7	Indel	Targeted	Y
43	32	M	Concern for Trisomy 21, high risk cffDNA; complete atrioventriucual canal defect	Trisomy 21	NA	Multiple	Extra Chromosome	Karyotype	Y
44	33	F	Concern for genetic condition, especially 22q11.2 deletion syndrome; Tetralogy of Fallot	22q11 deletion syndrome	NA	Multiple	Microdeletion	Microarray	Y
45	33	F	Concern for chromosomal condition; ectopic/low right kidney, small lungs, subtle callosal dysgenesis, micrognathia with cleft secondary palate, right-sided aortic arch	Triploidy (69,XXX)	NA	Multiple	Extra Chromosome Set	FISH	Y
46	33	F	Concern for Autosomal Recessive Polycystic Kidney Disease; bilaterally enlarged cystic kidneys, mild oligohydramnios	Polycystic kidney disease 4, with or without hepatic disease	263200	PKHD1	Single Nucleotide Variant	Targeted
47	34	F	Concern for genetic syndrome; duodenal atresia, polyhydramnios possibly shortened corpus callosum	Feingold syndrome type 1	164280	MYCN	Indel	Targeted
48	35	F	Concern for genetic syndrome; neuroblastoma	Neurofibromatosis type 1	162200	NF1	Single Nucleotide Variant	Targeted
49	35	F	Concern for genetic syndrome; polyhydramnios, mildly short extremities, prominence of subcutaneous integumentary layer, dysgenesis of the corpus callosum	Partial 19q and 20 duplications	NA	Multiple	Microdeletions and duplications	Microarray	Y
50	36	F	Concern for genetic syndrome; left pelvic kidney, uterine-like structure in the left hemipelvis, no scrotal sac or testes, duodenal atresia, possible malrotation, polyhydramnios	XY (Disorder of Sexual Development, phenotypically female)	NA	Multiple	Chromosomal Difference	Karyotype	Y
51	36	F	Concern for genetic syndrome; double outlet right ventricle, mitral stenosis, concern for coarctation of the aorta	CHARGE syndrome	214800	CHD7	Indel	Targeted	Y

Bolded genetic diagnoses without strong suspicion of genetic disorder prenatally

Integrating postnatal diagnosis and prenatal imaging

Nine participants’ charts did not have documentation of clinician concern for a genetic etiology during MFCC evaluation (Table 3). Of these participants, 5/9 (55.6%) received a genetic diagnosis postnatally via ES, 3/9 (33.3%) were solved with targeted testing guided by postnatal phenotypes, and 1/9 (11.1%) received a diagnosis from CMA. Four of these 9 participants (44.4%) had some prenatal genetic testing which was nondiagnostic (karyotype and/or CMA); postnatal diagnosis was subsequently obtained in 3/4 of these cases (75%) by ES and 1/4 (25%) by targeted testing. Diagnoses included rare Mendelian disorders, including ADNP-related syndrome (MIM: 615873), Fraser syndrome (MIM: 219000), TARP syndrome (MIM: 311900), and Congenital Disorder of Glycosylation Type 1g (CDG1g, MIM: 607143). These nine cases fell into two general categories, where either the prenatal phenotype was different from the ‘expected’ phenotype (BWS, Noonan syndrome), or the prenatal phenotype consisted of non-specific or ‘minor’ anomalies (Fraser, ADNP-related, CDG1g, TARP, Shprintzen-Goldberg) at the time of the MFCC visit. A representative case of the latter category, which on retrospective radiology review included findings of protuberant forehead, long philtrum, and mild retrognathia is shown in Figure 2.

Figure 2. Comparing pre- and postnatal phenotypes for a patient with Shprintzen-Goldberg Syndrome.

This family was referred to the MFCC due to a history of concern for fetal cerebellar hypoplasia, Blake’s pouch cyst, and left clubfoot. Prior workup included additional imaging revealing normal positioning of both feet and concern for abnormality of the posterior fossa and a non-diagnostic prenatal microarray. MFCC imaging at 20w3d gestational age demonstrated a structurally normal brain and no evidence of clubfoot. The infant was admitted to the NICU with concern for micrognathia, bilateral club feet, respiratory distress and dysmorphic features (Panel B). Postnatal ES was diagnostic for Shprintzen-Goldberg syndrome. Rereview of prenatal imaging (Panel A) did reveal some subtle features consistent with the genetic diagnosis and postnatal phenotype, including recessed nasal bridge, protuberant forehead, long philtrum, and mild retrognathia.

The prenatal imaging studies from these nine participants were subsequently reviewed with a pediatric radiologist for features that, in retrospect, could be consistent with the known genetic diagnosis. In 4/9 cases, this review demonstrated minor features that could be consistent with the diagnoses (Table 4 and Supplemental Materials).

Table 4:

Retrospective radiology review of cases without documented prenatal suspicion for a genetic condition.

Case	Referring Clinical Diagnosis	GA at Referral (weeks)	Phenotypic Sex	Specialists seen in MFCC	Prenatal phenotype	Postnatal phenotype	Genetic Diagnosis	MIM	Gene Affected	Molecular Diagnosis Technology	Retrospective Radiology Review
7	Fetal cerebellar hypoplasia, possible Blake’s pouch cyst, left club foot	19	M	Neurology	Thickening and mild increased T2 signal in the nuchal soft tissue	Choroid plexus cyst; micrognathia; bilateral club feet; dysmorphic facial features; respiratory distress	Shpritzen-Goldberg syndrome	182212	SKI	Exome Sequencing	20 weeks GA: Recessed nasal bridge, protuberant forehead, long philtrum, and mild retrognathia, along with the other features originally described
10	Solitary left kidney, small cerebellum, possible hypospadias	20	M	Urology, Nephrology, Neurology	Mild dolichocephaly; solitary left kidney; hypospadias	Right eyelid coloboma; simplified ears with fused earlobes; broad nasal bridge; subglottic stenosis; solitary kidney; hand and foot syndactyly; short right thumb.	Fraser syndrome 1	219000	FRAS1	Exome Sequencing	Cannot visualize right kidney but cannot confirm its absence due to lack of compensatory hypertrophy of left kidney
16	Right pleural effusion	21	M	Cardiology, General Surgery	Prominence of the retrocerebellar CSF space; moderate right pleural effusion	Bilateral pleural effusions; retrognathia; low-set, posteriorly-rotated ears; pectus excavatum	Noonan Syndrome I	163950	PTPN11	Targeted	No additional findings
17	Abnormal views on prenatal imaging of brain and heart	21	M	Cardiology, Neurology, General Surgery	At 22w GA: retrognathia; mandibular hypoplasia; inferior vermian hypoplasia; prominence of the cisterna magna; At 25w GA: findings had resolved	Microretrognathia; slanting forehead; broad nasal root; slightly low-set ears; deep sacral dimple	TARP Syndrome	311900	RMB10	Exome Sequencing	22 weeks GA: reveals an abnormal fetal profile, including sloping forehead. 25 weeks GA: fetal MR reveals a persistently abnormal and hypoplastic vermis (with some interval growth), a thickened nuchal fold, and subarachnoid fluid
22	Suspected fetal cardiac anomaly, triplet pregnancy with twin-twin transfusion syndrome	22	F	Cardiology	Triplet gestation; recipient twin in twin-twin transfusion syndrome	Macroglossia; flammeus nevus of glabella and occiput; umbilical hernia	Beckwith-Wiedemann Syndrome	130650	Multiple	Targeted	No additional findings
31	Echogenic kidney, IUGR, echogenic bowel, concern for right renal agenesis vs. pelvic kidney, dilated IVC, abnormal dopplers	26	M	Urology, Cardiology, Maternal-Fetal Medicine	Echogenic kidney; question of unilateral renal agenesis; fetal growth restriction; echogenic bowel; dilated IVC; and abnormal dopplers	Unilateral renal agenesis with hydronephrosis/ hydroureter; aneurysmal septum primum; umbilical vein atresia due to thrombosis that resulted in cardiomyopathy; right sided inguinal hernia; hypoglycemia	Helsmoortel-van der Aa syndrome	615873	ADNP	Exome Sequencing	30w GA: tortuous vessels in the pelvis; lung lesion of unknown etiology; undescended testes (expected to be descended at this gestational age)
35	Pulmonary Atresia	27	F	Cardiology, Maternal Fetal Medicine	Pulmonary atresia with intact ventricular septum	Pulmonary atresia with intact ventricular septum, venous infarct in the left frontal, left temporal and left occipital lobes	22q11.2 duplication (2.8MB)	608363	Multiple	Microarray	No addtional findings
36	Mild polyhydramnios	30	F	General Surgery	Mild polyhydramnios; fetus small for GA; inferior vermian hypoplasia	Microcephaly; short, broad nose; low-set, posteriorly-rotated ears, micrognathia; curved spine/spinal anomaly vs. abnormal position due to hypotonia; tethered cord	Congenital disorder of glycosylation, type Ig	607143	ALG12	Exome Sequencing	30w GA: bilateral enlarged ears; low conus (without visible curvature of the spine); small stomach
38	Possibility of a fetal cardiac problem related to diabetes mellitus type I	28	M	Cardiology	Normal fetal echocardiogram (in the setting of maternal T1DM)	Abnormal Newborn Screening for Galactosemia	Galactosemia	230400	GALT	Targeted	No additional findings

Discussion:

We describe a cohort of infants admitted to the NICU and subsequently diagnosed with rare genetic disorders who had undergone a comprehensive prenatal evaluation in our MFCC. Although a genetic diagnosis can have a profound impact on postnatal care through direction of management, prediction of natural history, and establishing a definitive diagnosis for a family—particularly for infants in the NICU¹⁵ —identifying appropriate cases for prenatal diagnostic genetic testing (e.g. genome-wide sequencing) remains challenging. Our work illustrates that critically-ill NICU infants with genetic disorders may have presented with non-specific fetal anomalies, adding to the evidence that even apparently isolated or ‘minor’ anomalies may warrant a broad genetic workup.¹⁶

Our study also demonstrates that cases with characteristic postnatal phenotypes may not present with imaging findings strongly concerning for a genetic diagnosis. Predicting the likelihood of a genetic diagnosis in the prenatal period is complicated by limited information: specific prenatal phenotypic features of genetic disorders may not be well-described or categorized¹⁷, and subtle findings may be missed on prenatal imaging. In our cohort, nine participants had postnatal genetic diagnoses but fetal imaging at the time of the MFCC visit was not strongly suspicious for a genetic etiology due to the apparent lack of multiple anomalies, prior negative genetic testing results, or the presence of anomalies not immediately suggestive of a genetic syndrome. Notably, the prenatal presentations of the disorders ultimately identified in these cases have not been well-described. Given the genetic and phenotypic heterogeneity of prenatal presentations of genetic disease, there is thus need for both comprehensive fetal phenotypic assessment and prenatal sequencing in parallel^18,19.

Efforts to aggregate prenatal phenotype data with known genetic diagnoses is crucial in building the knowledge base to provide meaningful counseling to patients. Improved understanding of progressions of prenatally-diagnosed diseases and their neonatal presentations may better our ability to prognosticate on perinatal outcomes. This facilitates more accurate anticipatory guidance during prenatal counseling, which can help expectant parents make timely decisions about pregnancy management—e.g., delivering at a tertiary care center instead of a community hospital; pursuing supportive care for a critically ill newborn; or safe and legal termination of pregnancy. In turn, neonatal care is often more tailored and effective when a prenatal genetic diagnosis is suspected or a genetic diagnosis has been obtained²⁰—e.g., medications may be contraindicated; pediatric specialists may be indicated in the delivery room if there is high suspicion for a given disorder; or directed treatments may be indicated from birth or in utero.

Lastly, we show that most diagnoses in our NICU cohort could have been made prenatally by ES with microarray, ES with CNV analysis, or GS alone. Even though a genetic diagnosis was suspected prenatally in the majority of the cases identified, the increased availability of ES in the postnatal compared to the prenatal context greatly aided genetic diagnosis: after non-diagnostic karyotype and chromosomal microarray, ES was diagnostic for all but one neonatal patient for whom it was ordered in this cohort. Many of the diagnoses in our cohort were unique (i.e., only occurred in 1 case), further supporting the need for broad testing methodologies over targeted approaches. Based on the genetic diagnoses obtained postnatally, all but one case in our cohort could, in theory, have been solved by employing ES prenatally. However, ES is a phenotype-driven test; as discussed above, deep fetal phenotyping and aggregation of fetal phenotypes is crucial to fully actualize the potential of ES (and GS) as important and timely methods to capture the majority of genetic etiologies prenatally. While genotype-driven genomic sequencing studies in the absence of phenotype have been performed²¹, comprehensive guidelines for variant reporting in these instances, which include incidental and secondary findings, have not been established.

Together, these points support the broad implementation of clinical exome and genome sequencing in the prenatal setting. Adopting a case-by-case approach for use of ES/GS may lead to missed diagnoses in the prenatal period and delayed time to diagnosis, which can have significant implications for parental decision-making and neonatal management. The benefits of timely genetic diagnosis, including personal, informational, and medical utility, have been demonstrated in prior studies^6,20,22,23. Furthermore, a selective approach to offering prenatal testing might further exacerbate existing inequities in access to genetic testing, as providers may be less likely to offer diagnostic testing to people with marginalized identities²⁴. Widespread implementation of prenatal ES and GS is hindered by barriers including lack of insurance coverage, turnaround time, unclear guidelines for ordering prenatal sequencing, and the need to interpret uncertain or incidental genetic test results. In one study of private payers in the US, most reported offering clinical ES/GS in the pediatric setting, while at the time no payers supported coverage for prenatal ES⁸; historically, payers have expressed skepticism that prenatal ES provides any incremental benefit over prenatal imaging and chromosomal microarray testing. However, literature suggests that prenatal ES may have a 31% diagnostic yield when microarray and karyotype are nondiagnostic for severe phenotypes, and a 10% yield when only minor anomalies are seen³. Postnatally, recent work demonstrates that access to comprehensive genetic testing in many level IV NICUs is restricted by limited access to sequencing technology, availability of appropriate specialists, and/or consulting services in critical timeframes²⁵. Equitable access to genetic services in the perinatal period—both technologies and the providers with training and expertise to offer them—is crucial for optimal care for patients and families. Prenatal care presents an opportunity to broaden access to these services. Simultaneously, it is imperative to better characterize biases in referrals, offerings, and uptake of testing to ensure services are implemented equitably.

Notably, the majority of cases (42/51, 82.4%) did not pursue prenatal diagnostic testing for various reasons as documented in the medical record. Cases reviewed in our study spanned a period of 7 years; technology, insurance coverage, and attitudes regarding diagnostic sequencing evolved significantly throughout that period. Additional reasons for the relatively low rate of prenatal diagnostic testing in our cohort include: family decision to postpone testing to the postnatal period, requirement for a postnatal genetic diagnosis, limited availability of exome sequencing early in the study, pregnancies with a prenatal genetic diagnosis not resulting in live birth, and requirement for transfer of infants to our level IV NICU. Many pregnant patients in our cohort also cited the invasive nature of CVS and amniocentesis as a reason for declining diagnostic testing. Non-invasive screening and testing modalities using cell-free DNA are emerging and are increasingly integrated into clinical practice. Beyond screening for aneuploidies early in gestation, which is already recommended by ACOG for all pregnancies²⁶, non-invasive single gene screening panels assessing more than 25 conditions relevant to prenatal diagnosis are now clinically available^27–29. Almost half of all diagnoses in our cohort could theoretically be detected with currently clinically available cell-free DNA screening in the United States. Furthermore, the feasibility of non-invasive exome sequencing has been demonstrated in several pilot studies^30,31. Such technologies have the potential to capture nearly all diagnoses identified in our cohort. Non-invasive broad sequencing approaches are on the horizon, opening up the possibility of increased prenatal genetic testing uptake and earlier diagnosis.

While providing insight into prenatal presentations of postnatal genetic disease, there are several limitations to our study. First, we do not have access to comprehensive prenatal imaging data for all individuals in the NICU during the collection period, only those which were seen in our MFCC. Ideally, the prenatal presentations of a control set of individuals without a genetic diagnosis could be used as a comparison group. This is currently impractical because we have access only to comprehensive imaging data for those individuals seen in our MFCC, and not all individuals in the NICU had broad genetic testing to rule in or out a genetic diagnosis. Second, our dataset is biased towards pregnancies that had some concerning features that would prompt referral to our MFCC. While we do not have access to prenatal imaging on those individuals in the NICU with a genetic diagnosis who were not seen in the MFCC, it can be hypothesized that fewer prenatal phenotypes were observed in this population. Therefore, our cohort is enriched for cases with some prenatal phenotype, even if in certain cases further workup was less concerning for a genetic etiology. Our cohort is also enriched for postnatal presentations with phenotypes necessitating NICU care, and does not include milder post-natal phenotypes. Third, over the defined study period, sequencing technologies and ordering practices have changed. Therefore, our rates of uptake of diagnostic testing and technology used (e.g. ES) also reflect changing practice over time. Fourth, unless there is a strong reason for monitoring by the MFCC (as opposed to the MFM practice), individuals are typically only seen once in the MFCC unless a new concern arises. Therefore, assessment is made at one moment in time and additional features could certainly develop that were not noted in the medical record. Fifth, while datasets such as those presented are scarce due to the fractionation of maternal and fetal records, the total numbers of cases are nonetheless relatively small for statistical assessments, arguing for systematic and prospective improvements for collection and integration of prenatal phenotype/genotype data. Lastly, our cohort excludes pregnancies that resulted in stillbirth, fetuses whose parents chose termination of pregnancy or palliative care, and infants who did not survive to admission to our level IV NICU. The phenotypic presentations and genetic diagnoses of fetuses and/or neonates with those outcomes may be different, and in some cases more severe, from those identified in our cohort.

Future research is needed that would address these limitations and include assessment of rates of genetic diagnosis in non-anomalous fetuses, prenatal imaging findings of infants in the NICU with broad sequencing and no identifiable genetic diagnosis, guidelines for interpretation of incidental or uncertain findings, and further assessment on care changes given a prenatal genetic diagnosis. Such research relies on data sharing of both genetic and phenotypic information across the community. Structured phenotypes that incorporate prenatal-specific terms, such the Human Phenotype Ontology (HPO)¹⁸, and data structures such as Phenopackets to assess longitudinal phenotype-genotype data can aid in this effort³². With large numbers, quantitative metrics on yield of genetic diagnosis by phenotype could be obtained.

Conclusion:

Numerous factors are crucial for a thorough perinatal genetic evaluation, including detailed phenotypic assessment, family history, and cytogenetic/sequencing information. Our data illustrate important challenges: first, genetic diagnoses can exist in pregnancies with minor, non-specific fetal anomalies; second, prenatal phenotypes can have variable relationships to the definitive phenotypes of genetic disorders characterized postnatally; third, ES plus CMA, ES plus CNV analysis, or GS alone could have identified almost all of these genetic diagnoses prenatally, though obtaining sufficient phenotypic information remains a barrier to diagnosis with prenatal ES or GS. Sequencing and imaging technologies continue to advance and are increasingly accessible, enabling more and earlier genetic diagnoses of conditions not previously described in the perinatal period^33,34. Relying on pediatric phenotypic diagnostic criteria alone can present a challenge to recognizing a diagnosis in the appropriate timeframe for prenatal decision-making, or preparation for delivery and postnatal management. Furthermore, an increasing number of fetal therapies are emerging, for which a genetic diagnosis can help direct care^35,36. As we enter the era of fetal genomic medicine, identifying fetuses with an underlying genetic diagnosis is crucial. Increasing access to genomic technologies and aggregation of perinatal phenotypes with genomic data will enable enhanced counseling and management of genetic disease in the perinatal period and beyond.

What is already known about this topic?

• Rare genetic disorders can present prenatally with expanded, non-classical, or absent phenotypes, and the full spectrum of prenatal phenotypic presentations for many Mendelian disorders remain unknown.

• Certain prenatal phenotypes (for example, those with skeletal or multisystem phenotypes) have amongst the highest genetic yield.

What does this study add?

• This work integrates prenatal imaging data with postnatal genetic diagnoses, a rarity due to the typical fractionation of maternal and neonatal healthcare records.

• This study illustrates that critically-ill NICU infants with genetic disorders can have non-specific fetal anomalies and highlights the value of broad sequencing at high-risk centers in the face of limited or uncertain phenotype data.

Data Availability:

De-identified data may be made available upon reasonable request, contingent upon a data use agreement.