Genomic Landscape and Perinatal Outcomes of Fetal Polydactyly: A Retrospective Cohort Study Integrating CNV‐seq and Trio‐ES

ABSTRACT

Objective

To evaluate the clinical characteristics, genomic landscape, and perinatal outcomes of fetal polydactyly using combined copy number variation sequencing (CNV‐seq) and trio‐exome sequencing (trio‐ES).

Methods

This retrospective cohort study included 44 prenatally confirmed fetuses with polydactyly. Cases were stratified into isolated (n = 21) and non‐isolated (n = 23) groups. Polydactyly was further subclassified according to anatomical distribution, laterality, and duplication axis. Genetic etiology was investigated using CNV‐seq and trio‐ES, and pregnancy outcomes were ascertained through clinical follow‐up.

Results

The overall genetic diagnostic yield was 36.4% (16/44), comprising aneuploidies (n = 6), one pathogenic CNV and monogenic disorders (n = 9). Causative variants were identified in NEK1, EVC2, BBS4, GLI3, TBX3, MYCN, and KIAA0825, with one incidental finding in PIK3CD. The diagnostic yield was significantly higher in the non‐isolated group than in the isolated group (60.9% vs. 9.5%, p < 0.001). Specific anatomical features were associated with markedly increased genetic burden, including concurrent involvement of both upper and lower limbs (100%), bilateral presentation (64.7% vs. 18.5% for unilateral; p < 0.01), and postaxial polydactyly (PAP) compared with preaxial polydactyly (PPD) (65.0% vs. 12.5%; p < 0.001). Pregnancy outcomes differed substantially between phenotypic subgroups, with a live birth rate of 100% in isolated cases versus 34.8% in non‐isolated cases (p < 0.001).

Conclusion

Fetal polydactyly exhibits profound genetic and phenotypic heterogeneity. Although non‐isolated, multi‐limb, bilateral, and postaxial presentations are strong predictors of underlying chromosomal or monogenic disorders, apparently isolated cases still carry a clinically relevant genetic risk. The integration of detailed prenatal sonographic phenotyping with CNV‐seq and trio‐ES improves diagnostic precision, facilitates prognostic assessment, and informs prenatal counseling and long‐term postnatal surveillance.

Summary

• What is already known about this topic?

  • ◦Polydactyly is a common congenital limb malformation with marked phenotypic and genetic heterogeneity.
  • ◦Prenatal sonographic detection of polydactyly remains challenging, and apparently isolated cases have historically been considered to carry a relatively low risk of an underlying genetic disorder.
  • ◦Conventional chromosomal testing may miss monogenic etiologies in fetal polydactyly.

• What does this study add?

  • ◦In 44 prenatally confirmed fetuses with polydactyly, combined CNV‐seq and trio‐ES yielded an overall diagnostic yield of 36.4%.
  • ◦Non‐Nonisolated polydactyly, as well as multi‐limb involvement, bilateral presentation, and a postaxial duplication axis, were associated with a higher genetic diagnostic yield.
  • ◦Apparently isolated polydactyly still carried a clinically relevant genetic risk, supporting comprehensive genomic testing and postnatal surveillance.

1. Introduction

Polydactyly, a congenital limb malformation characterized by the presence of supernumerary digits on the hands and/or feet, is recognized as one of the most prevalent congenital limb malformations [1]. Globally, the birth prevalence of polydactyly is estimated to range from 0.3 to 3.6 per 1000 live births, exhibiting significant variations across distinct ethnicities and geographical regions [2, 3]. Clinically, polydactyly is broadly divided into non‐syndromic and syndromic forms, with the non‐syndromic type accounting for more than 80% of cases [4]. Anatomically, it is further defined according to the position of the supernumerary digit as preaxial polydactyly (PPD; duplicated digit on the radial/thumb side), postaxial polydactyly (PAP; duplicated digit on the ulnar/little finger side), or central polydactyly (duplication of the second, third, or fourth digit) [5].

The etiology of polydactyly is predominantly genetic, originating from molecular disruptions during limb bud formation and anteroposterior patterning between the fourth and eighth weeks of gestation [5]. Central to its pathogenesis is the dysregulation of the Sonic Hedgehog (SHH) signaling cascade, which governs the zone of polarizing activity (ZPA) and determines digit number and identity [1, 6, 7]. Pathogenic variants in GLI3 or the ZRS enhancer, as well as genes involved in primary cilia function, have all been implicated in the development of polydactyly [5, 7]. Environmental factors, including maternal diabetes and first‐trimester teratogenic exposure, also contribute to this risk [3, 4, 8].

The prenatal diagnosis of fetal polydactyly relies predominantly on ultrasound examination; however, its sensitivity remains low. A recent large‐cohort study reported an overall prenatal detection rate of approximately 19.2% [9]. The late first to early second trimester is considered optimal for evaluating fetal polydactyly [10]. This diagnostic challenge is often attributed to the small size of the digital structures, the clenched position of the fetal hands, maternal habitus, and the fact that digit counting is not a mandatory requirement in routine screening guidelines [8, 9]. Once polydactyly is suspected sonographically, prenatal counseling becomes exceptionally complex, as the finding may represent either an isolated benign manifestation with an excellent prognosis or a sentinel sign of severe, multisystem genetic syndromes [1, 10, 11, 12].

Currently, first‐tier genetic testing utilizing chromosomal microarray analysis (CMA) or copy number variation sequencing (CNV‐seq) is routinely performed in fetuses with structural anomalies. However, these methods fail to detect single‐nucleotide variants (SNVs) and small insertions/deletions (indels), which account for a substantial proportion of monogenic etiologies in limb malformations [1, 6]. Although exome sequencing (ES) has emerged as an effective second‐tier diagnostic tool, yielding incremental diagnostic rates of up to 42% in fetuses with skeletal anomalies, its specific application and diagnostic value in prenatal polydactyly remain underexplored [13]. Therefore, the present study aimed to comprehensively evaluate the chromosomal and monogenic etiologies, anatomical distribution patterns, laterality, and perinatal outcomes of fetuses with polydactyly, thereby providing practical guidance for prenatal phenotypic stratification, genetic testing, and prenatal counseling.

2. Materials and Methods

2.1. Study Population and Setting

This study was approved by the Institutional Ethics Committee of the Maternal and Child Health Hospital of Hubei Province (IRB No. 2026‐010‐01). Written informed consent was obtained from both parents. Between October 2021 and December 2025, a total of 12,457 prenatal and fetal tissue samples were referred to our Medical Genetics Center. Among these, 47 pregnancies (0.38%) were identified as suspected fetal polydactyly. Polydactyly was initially detected on routine prenatal ultrasonography and confirmed by a senior sonographer, followed by detailed ultrasound assessment for associated anomalies. After sequential evaluation and postnatal follow‐up, the diagnosis was revised in three cases: one fetus was excluded because polydactyly was not confirmed on third‐trimester ultrasound, and two additional fetuses were excluded because postnatal follow‐up showed no evidence of polydactyly. The final study population therefore comprised 44 fetuses with confirmed polydactyly, representing 0.35% (44/12,457) of the total tested population. Among these 44 confirmed cases, 36 underwent amniocentesis with CNV‐seq, and trio‐ES was performed either concurrently or sequentially after negative CNV‐seq results. In eight cases with complex multisystem anomalies and parental preference for termination of pregnancy (TOP) without prior amniocentesis, fetal tissue was obtained directly for trio‐ES.

2.2. Collection and Analysis of Clinical Features

Clinical features and genetic data were retrospectively collected by reviewing electronic medical records from the Hospital Information System (HIS). The extracted data included maternal age, family history, toxin exposure, gestational age at diagnosis, prenatal ultrasound findings, and genetic testing results. The cohort was initially stratified into isolated and non‐isolated polydactyly groups based on the presence or absence of additional structural anomalies in other organ systems [10]. Independent of this classification, the anatomical distribution of the polydactyly was categorized as affecting the upper limbs only, lower limbs only, or both. Furthermore, based on the location of the supernumerary digit relative to the limb axis, cases were subclassified into PPD and PAP [5].

2.3. Genetic Testing

Genomic DNA was extracted directly from uncultured amniotic fluid, fetal tissue, and peripheral blood of the biological parents using the Qiagen DNA Blood Midi/Mini Kit (Qiagen GmbH, Hilden, Germany). Short tandem repeat (STR) analysis was routinely performed to exclude maternal cell contamination. CNV‐seq was conducted according to the manufacturer's protocol. Qualified libraries were sequenced on the NextSeq 500CN system (Illumina, San Diego, USA), providing a genome‐wide detection resolution of ≥ 100 kb.

ES was performed using the NanoWES platform (Berry Genomics, China) on the NovaSeq 6000 system (Illumina, San Diego, USA), following the methodology previously described by Qin et al. (2025) [14]. Additionally, an in‐house coverage‐based tool, Sprinkle (Berry Genomics, China), was employed for CNV calling from the ES data. The pathogenicity of identified genetic variants was interpreted according to the American College of Medical Genetics and Genomics (ACMG) guidelines [15, 16]. All candidate pathogenic variants were validated via Sanger sequencing. Only variants classified as pathogenic, likely pathogenic, or warm variants of uncertain significance (VUS) with potential clinical relevance to the primary prenatal indications were reported. Incidental findings and secondary findings associated with childhood‐onset diseases were disclosed based on the parents' choices during pre‐test counseling [17].

2.4. Follow‐Up and Statistical Analysis

Perinatal outcomes were retrieved from institutional delivery records or via the Hubei Maternal and Child Health Wisdom Management System and the Hubei Wuhan Maternal and Child health Services Management Information System. Long‐term postnatal outcomes were ascertained through periodic telephone follow‐up. Continuous variables were expressed as mean ± standard deviation (SD), while categorical variables were presented as frequencies and percentages. Statistical differences were analyzed using Fisher's exact test, with a significance threshold set at p < 0.05. All statistical analyses were performed using SPSS software version 30.0 (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Clinical Features

During the study period, 47 fetuses with sonographically suspected polydactyly underwent genetic investigation at our center. Following longitudinal evaluation, the diagnosis was revised in three cases (6.4%). The remaining 44 fetuses formed the final confirmed cohort, yielding a prenatal sonographic confirmation rate of 93.6% (44/47) (Figure 1; Table 1; Table S1). Of these confirmed cases, 21 (47.7%) presented with isolated polydactyly, and 23 (52.3%) with non‐isolated polydactyly. The cohort consisted of 26 male (59.1%) and 18 female (40.9%) fetuses. The median maternal age was 31 years (range: 19–42 years), and the median gestational age at diagnosis was 23 weeks (range: 13–31 weeks). Most cases were detected in the second trimester (n = 38; 86.4%), with fewer cases identified in the first (n = 1; 2.3%) and third (n = 5; 11.4%) trimesters. A positive family history of polydactyly was identified in 5 cases (11.4%), while no history of toxin exposure was reported. Regarding anatomical distribution, polydactyly was confined to the upper limbs in 29 cases (65.9%), to the lower limbs in 10 cases (22.7%), and involved both upper and lower limbs in 5 cases (11.4%) (Figure 2). Bilateral involvement was observed in 17 cases (38.6%), whereas 27 cases (61.4%) were unilateral. In addition, 24 fetuses (54.5%) had PPD and 20 (45.5%) had PAP (Figure 2).

FIGURE 1.

Flowchart of the study design and sequential genetic diagnostic strategy.

TABLE 1.

Clinical characteristics and prenatal ultrasound findings of the 44 fetuses with polydactyly.

Clinical features	Values
Maternal age (years)	31 (19–42)
Family history of polydactyly	Yes: 5 (11.4%)
	No: 39 (88.6%)
Toxin exposure	Yes: 0 (0%)
	No: 44 (100%)
GA at diagnosis (weeks)	23 (13–31)
Upper limb (hand) only	Unilateral: 22
	Bilateral: 7
Lower limb (foot) only	Unilateral: 5
	Bilateral: 5
Both upper and lower limbs affected	Unilateral: 0
	Bilateral: 5
Fetal sex	Male: 26
	Female: 18
Perinatal outcome	Survival: 29
	TOP: 15

Abbreviation: GA, gestational age.

FIGURE 2.

Anatomical distribution and genetic testing results of 44 fetuses with polydactyly. PAP, postaxial polydactyly; PPD, preaxial polydactyly.

3.2. Diagnostic Yield of Genetic Testing

The overall genetic diagnostic yield among the 44 fetuses with confirmed polydactyly was 36.4% (16/44, Figure 3). Among 36 prenatal amniotic fluid samples, CNV‐seq yielded diagnostic results in six cases (16.7%), comprising trisomy 13 (n = 3), trisomy 18 (n = 1), trisomy 21 (n = 1), and one pathogenic copy number variant (CNV) (Figure 3; Table 2). Prenatal trio‐ES identified causative monogenic variants in four additional amniotic fluid samples (11.1%; Cases 1, 20, 23, and 26) (Figure 3; Table 3). Furthermore, in the eight fetal tissue samples obtained following TOP, trio‐ES yielded a high diagnostic rate of 75.0% (6/8), identifying one case of trisomy 13 and five cases of monogenic disorders (Cases 36, 38, 39, 41, and 42) (Figure 3; Table 3).

FIGURE 3.

Diagnostic yield of genetic testing according to sample type and clinical presentation. CNVs, copy number variants; TOP, termination of pregnancy.

TABLE 2.

Chromosomal abnormalities identified by CNV‐seq in prenatal amniotic fluid samples.

Case	MA (yrs)	GA	Ultrasound findings	CNV‐seq results	Classification	Origin	Pregnancy outcome
13	30	18	Poor visualization of CSP, micrognathia, PAP of both hands and both feet, DORV, severe pulmonary artery stenosis, and right aortic arch	Trisomy 13	P	de novo	TOP
16	33	17	Omphalocele, PAP of both feet	Trisomy 13	P	de novo	TOP
18	37	17	NT5.4mm, generalized cutaneous edema, cervical cystic hygroma, and PAP of the left hand	Trisomy 13	P	de novo	TOP
35	19	19	Bilateral choroid plexus cysts, PPD of the right hand, VSD, horseshoe kidney, and echogenic bowel	Trisomy 18	P	de novo	TOP
19	30	22	Nasal bone hypoplasia, thickened nuchal fold, and PPD of both hands.	Trisomy 21	P	de novo	TOP
28	33	17	PAP of right hand	seq[GRCh37]22q11.21q11.21(18,960,000_ 21,460,000) × 1	P	Maternal	Live birth

Abbreviations: CSP, cavum septum pellucidum; DORV, double outlet right ventricle; NT, nuchal translucency; P, pathogenic; PAP, postaxial polydactyly; PPD, preaxial polydactyly; TOP, termination of pregnancy; VSD, ventricular septal defect.

TABLE 3.

Causative monogenic variants identified by trio‐ES in fetuses with polydactyly.

Case	Ultrasound findings	Sample Type	Gene/RefSeq/Variant	Inheritance	Zygosity	Classification	Origin	Disease (OMIM ID)	Outcome
1	PAP of both hands and feet	Amniotic fluid	NM_001145678.3(KIAA0825):c.‐1–2A > T(p.?); NM_001145678.3(KIAA0825):c.2247–2A > G(p.?)	AR	Compound het	LP(PVS1_Moderate + PM2_Supporting + PM3+PP4); LP(PVS1_Strong + PM2_Supporting + PP4)	Pat Mat	Polydactyly, postaxial, type A10	LB
20	PAP of both hands, SUA	Amniotic fluid	NM_005996.4(TBX3):c.1454del(p.Pro485ArgfsTer127)	AD	Het	LP(PVS1_Strong + PM2_Supporting + PS2_Supporting)	de novo	Ulnar‐mammary syndrome	LB
23	Head circumference below ‐2SD, abdominal double bubble sign, and PAP of the right hand	Amniotic fluid	NM_005378.6(MYCN):c.1117C > T(p.Arg373Ter))	AD	Het	P(PVS1_Strong + PS2+PM2_Supporting)	de novo	Feingold syndrome 1	TOP
26	Bilateral lateral ventricles at the borderline (0.99cm/0.94cm), PPD of both hands and both feet	Amniotic fluid	NM_000168.6(GLI3):c.4431dup(p.Glu1478Ter)	AD	Het	P(PVS1+PS2+PS4+ PM2_Supporting)	de novo	Polydactyly, preaxial, type IV	LB
36	PAP of both hands and both feet, VSD, and pulmonary atresia	Fetal tissue	NM_000168.6(GLI3):c.2374C > T(p.Arg792Ter)	AD	Het	P(PVS1+PS3_Moderate,+PS4_Moderate,+PM2_Supporting + PP1)	Pat	Pallister‐Hall syndrome	TOP
38	Short limbs, fibular aplasia, bilateral hyperechogenic kidneys, abnormal posture of both feet, and PAP of the left hand	Fetal tissue	NM_012224.4(NEK1):c.1908dup(p.Val637SerfsTer9); NM_012224.4(NEK1):c.2167C > T(p.Gln723Ter)	AR	Compound het	P(PVS1+ PM2_Supporting + PM3); LP(PVS1+PM2_Supporting)	Pat Mat	Short‐rib thoracic dysplasia 6 with or without polydactyly	TOP
41	Omphalocele, short limbs with fixed posture, bilateral small thoracic volume, bilateral pleural effusion, clubfoot, and bilateral PAP of the feet	Fetal tissue	NM_012224.4(NEK1):c.1907_1908del(p.Lys636SerfsTer9); NM_012224.4(NEK1): c.1414C > T(p.Arg472Ter)	AR	Compound het	P(PVS1+ PM2_Supporting + PM3); LP(PVS1+PM2_Supporting)	Pat Mat	Short‐rib thoracic dysplasia 6 with or without polydactyly	TOP
39	Fetal femur length below 5th percentile, curved and short ribs, altered thoracic configuration, PAP of both hands, ECD, and ARSA	Fetal tissue	NM_147127.5(EVC2): c.1711–2A > C(p.?); NM_147127.5(EVC2): c.1195C > T(p.Arg399Ter)	AR	Compound het	P(PVS1+ PM2_Supporting + PM3); P(PVS1+PM2_Supporting +PM3_Strong)	Pat Mat	Ellis‐van Creveld syndrome	TOP
42	Bilateral ventriculomegaly, bilateral enlarged and hyperechogenic kidneys, PAP of the left hand, and bilateral PAP of the feet.	Fetal tissue	NM_033028.5(BBS4): c.924T > A(p.Tyr308Ter); seq[GRCh38]15q24.1(72,735,113_72,735,182)x1(BBS4:Del exon13)	AR	Compound het	P(PVS1+ PM2_Supporting + PM3); LP(PVS1+ PM2_Supporting)	Pat Mat	Bardet‐Biedl syndrome 4	TOP

Abbreviations: AD, autosomal dominant; AR, autosomal recessive; ARSA, aberrant right subclavian artery; Compound het, compound heterozygous; dn, de novo; ECD, endocardial cushion defect; Het, heterozygous; LB, live birth; LP, likely pathogenic; mat, maternal; P, pathogenic; PAP, postaxial polydactyly; pat, paternal; PPD, preaxial polydactyly; SUA, single umbilical artery; TOP, termination of pregnancy; VSD, ventricular septal defect.

When stratified by sonographic complexity, the non‐isolated polydactyly group (n = 23) exhibited a diagnostic yield of 60.9% (14/23), consisting of aneuploidy (n = 6) and causative monogenic disorders (n = 8) (Figure 3). Conversely, in the isolated polydactyly group (n = 21), underlying genetic abnormalities were detected in only 9.5% of cases (2/21), including one pathogenic CNV and one monogenic disorder (Figure 3). Statistical analysis revealed a significantly higher diagnostic yield in the non‐isolated polydactyly group compared with the isolated group (60.9% vs. 9.5%, p < 0.001, Fisher's exact test).

Finally, the anatomical distribution of the extra digits significantly influenced the diagnostic outcome. The diagnostic yield was 31.0% (9/29) in fetuses with upper‐limb involvement only, 20.0% (2/10) in those with lower‐limb involvement only, and 100% (5/5) in fetuses with concurrent involvement of both upper and lower limbs (Figure 2). Notably, multi‐limb involvement was associated with a remarkably high genetic burden, yielding a diagnostic rate of 100%, which was significantly higher than that observed in fetuses with isolated hand involvement (p < 0.01, Fisher's exact test) and isolated foot involvement (p < 0.01, Fisher's exact test). In addition, bilateral involvement was associated with a significantly higher diagnostic yield than unilateral involvement (64.7% [11/17] vs. 18.5% [5/27]; p < 0.01, Fisher's exact test). When evaluating the axis of duplication, PAP showed a substantially higher overall diagnostic yield than PPD (65.0% [13/20] vs. 12.5% [3/24]; p < 0.001, Fisher's exact test) (Figure 2).

3.3. Pregnancy Outcomes and Follow‐Up

Pregnancy outcomes were successfully ascertained through follow‐up for all 44 fetuses in our cohort. In total, 29 pregnancies were continued to term, while 15 resulted in TOP. Among the 29 live births, postnatal surgical intervention for polydactyly was performed in 16 cases.

In the isolated polydactyly group (n = 21), all pregnancies were continued. In stark contrast, within the non‐isolated polydactyly group (n = 23), eight pregnancies were terminated immediately upon the sonographic detection of complex multisystem anomalies. Among the remaining 15 non‐isolated cases that underwent prenatal genetic testing, six were terminated following the identification of diagnostic genetic abnormalities. Additionally, one case with otherwise negative diagnostic results (Case 5) was incidentally found to harbor a de novo pathogenic variant in the PIK3CD, which is associated with immunodeficiency 14; the parents elected for TOP due to the potential for adverse outcomes. The live birth rate for the isolated polydactyly group was 100% (21/21), whereas the rate for the non‐isolated polydactyly group was 34.8% (8/23). Statistical analysis revealed that the live birth rate of the isolated group was significantly higher than that of the non‐isolated group (p < 0.001, Fisher's exact test).

4. Discussion

In this retrospective cohort, we comprehensively evaluated the genomic landscape, anatomical distribution patterns, and perinatal outcomes of 44 fetuses with confirmed polydactyly. By integrating CNV‐seq and trio‐ES, we achieved an overall diagnostic yield of 36.4% (16/44), which closely aligns with a recent study reporting a diagnostic rate of 35.5% (16/45) in fetuses with upper limb polydactyly [8]. Our findings highlight the profound genetic and phenotypic heterogeneity of fetal polydactyly, demonstrating that the likelihood of an underlying genetic abnormality is strongly influenced by precise phenotypic stratification. While non‐isolated polydactyly was a strong predictor of complex genetic syndromes, with a diagnostic yield of 60.9%, apparently isolated cases still carried a clinically relevant genetic risk of 9.5%. Moreover, specific anatomical features, including concurrent multi‐limb involvement, bilateral presentation, and a postaxial duplication axis, were associated with a substantially increased genetic burden. Together with the markedly divergent perinatal survival rates observed between phenotypic subgroups, these results underscore the importance of combining detailed prenatal sonographic phenotyping with high‐resolution genomic testing to optimize prognostic counseling and pregnancy management.

In the non‐isolated polydactyly group, the overall diagnostic yield was 60.9% (14/23), comprising chromosomal aneuploidies (n = 6) and monogenic disorders (n = 8). Consistent with previous reports, trisomy 13 was the most frequent chromosomal abnormality identified, accounting for 4 of the 14 positive cases [8]. Trio‐ES identified causative variants in eight fetuses with non‐isolated polydactyly (8/23, 34.8%), including recurrent compound heterozygous variants in NEK1 (Cases 38 and 41), heterozygous GLI3 variants (Cases 26 and 36), and pathogenic or likely pathogenic variants in TBX3, MYCN, EVC2, BBS4. These results underscore the central role of ciliopathy‐related genes in fetuses with polydactyly accompanied by visceral or skeletal anomalies.

Notably, two cases within the non‐isolated group were of particular value for prenatal counseling. Case 20, presenting with PAP of both hands and an isolated single umbilical artery (SUA), harbored a de novo c.1454del frameshift variant in TBX3. Pathogenic variants in TBX3 are associated with ulnar‐mammary syndrome (UMS), which exhibits a highly variable clinical spectrum ranging from severe malformations to nearly subclinical features [18]. Similarly, Case 26, characterized by bilateral PPD of both hands and feet alongside borderline ventriculomegaly, carried a de novo c.4431dup variant in GLI3. This specific clinical‐genomic profile closely mirrors Case 61 described by Arduç et al., which also exhibited PPD of both hands and feet, mild ventriculomegaly, and a GLI3 mutation [19]. These findings strongly suggest that when fetal polydactyly is accompanied by other structural anomalies, even isolated soft markers, the risk of underlying genetic aberrations is substantially elevated. Consequently, concurrent genomic investigations at both the chromosomal (e.g., CNV‐seq) and monogenic (e.g., ES) levels are highly recommended to prevent missed diagnoses.

Historically, the genetic diagnostic rate for isolated polydactyly has been reported to be remarkably low. In an earlier retrospective study, Bromley et al. reported that among 12 fetuses with isolated polydactyly, the five that underwent amniotic fluid karyotyping all yielded normal results [11]. More recently, Xiong et al. identified only one case of trisomy 21 and one involving an FMR1 gene abnormality among 157 fetuses with isolated polydactyly [9]. Similarly, in a recent cohort reported by Arduç et al., genetic testing in eight isolated cases revealed a single pathogenic GLI3 variant [19]. In contrast, the present study detected underlying genetic aberrations in 9.5% (2/21) of isolated cases. The lower diagnostic yields reported in previous studies may be attributed to the limited proportion of patients undergoing invasive prenatal testing and the historical lack of comprehensive, high‐resolution genomic diagnostic methodologies.

The genetic abnormalities identified in our isolated group comprised one case with compound heterozygous variants in KIAA0825 (Case 1) and one case with a 22q11.21 microdeletion (Case 28). KIAA0825 is associated with autosomal recessive nonsyndromic postaxial polydactyly type A10, which is typically characterized by isolated limb involvement and generally carries a favorable prognosis following postnatal surgical correction [20]. In contrast, 22q11.2 deletion syndrome is a multisystem disorder characterized by congenital heart defects, immune deficiency, characteristic facial features, neurodevelopmental impairment, and a broad spectrum of skeletal anomalies, including polydactyly [21]. Although the 22q11.21 microdeletion in Case 28 was inherited from an asymptomatic mother, and the one‐year‐old infant currently shows no abnormalities, this genomic disorder is known for its profound variable expressivity and incomplete penetrance. Therefore, long‐term neurodevelopmental follow‐up remains warranted. Together, these findings indicate that apparently isolated polydactyly may occasionally represent the earliest manifestation of a broader syndromic disorder, supporting the use of trio‐ES after a negative first‐tier chromosomal‐level test to improve prognostic counseling and postnatal surveillance.

Beyond the isolated versus non‐isolated distinction, the anatomical extent, laterality, and digit‐axis pattern of polydactyly all appeared to influence the likelihood of an underlying genetic diagnosis. In our cohort, the genetic diagnostic yield reached 100% (5/5) when polydactyly involved both upper and lower limbs, which was significantly higher than the yields observed in fetuses with upper‐limb‐only (31.0%) or lower‐limb‐only (20.0%) involvement. This suggests that multi‐limb presentation is a strong indicator of an underlying systemic genetic disorder. Laterality was also informative, as bilateral cases showed a substantially higher diagnostic yield than unilateral cases (64.7% vs. 18.5%). This observation is consistent with the cohort reported by Arduç et al., in which bilateral cases were more frequently associated with genetic abnormalities than unilateral cases [19]. Moreover, the axis of duplication emerged as an important predictive factor: PAP was associated with a substantially higher overall diagnostic yield than PPD (65.0% vs. 12.5%). This finding is in line with previous reports indicating that PAP is more often seen in syndromic conditions, particularly ciliopathies, and is associated with a broader spectrum of monogenic disorders [2, 5, 6]. Conversely, PPD, especially when isolated and unilateral, more often reflects a localized developmental anomaly with a lower genetic burden [6]. Therefore, during prenatal counseling, concurrent multi‐limb involvement, bilateral presentation, and a postaxial phenotype should prompt heightened suspicion for an underlying genetic etiology and warrant comprehensive genomic testing.

Pregnancy outcomes closely mirrored the genetic findings, with the live‐birth rates being markedly lower in the non‐isolated polydactyly group than in the isolated group (34.8% vs. 100%; p < 0.0001). In the non‐isolated group, 34.8% (8/23) of families elected termination immediately after the detection of complex multisystem anomalies. Among the remaining 15 cases, one harbored a chromosomal aneuploidy and five carried monogenic variants, which directly influenced the decision for TOP. Additionally, one case with otherwise negative diagnostic results (Case 5) was incidentally found to harbor a de novo c.3061G > A pathogenic missense variant in the PIK3CD, which is associated with immunodeficiency 14. Given the anticipated poor prognosis, the parents ultimately opted for TOP. In contrast, all 21 pregnancies in the isolated group were continued to term, achieving a 100% live birth rate, with subsequent surgical correction performed in 12 infants. These divergent trajectories reinforce that isolated polydactyly carries an excellent prognosis when comprehensive genomic testing excludes syndromic associations [9, 12].

Our study also provides valuable insights into the complex genotype‐phenotype correlations associated with specific genes, particularly GLI3. Pathogenic variants in GLI3 are known to cause a wide spectrum of disorders, including Greig cephalopolysyndactyly syndrome (GCPS; OMIM 175700), Pallister‐Hall syndrome (PHS; OMIM 146510), postaxial polydactyly type A1 and B (OMIM 174200) and preaxial polydactyly type IV (OMIM 174700) [22]. According to the GLI3 genotype‐phenotype correlation model proposed by Johnston et al., GCPS is primarily associated with all types of pathogenic variants located 5′ to nucleotide 1998 and 3′ to nucleotide 3481, whereas PHS is exclusively caused by truncating variants situated between nucleotides 1998 and 3481 [23]. Interestingly, in our cohort, a de novo c.4431dup frameshift variant in GLI3 manifested solely as isolated preaxial polydactyly type IV (Case 26), diverging from previous postnatal reports that linked this specific variant to mild GCPS [24]. Moreover, a paternally inherited c.2374C > T nonsense variant in GLI3 presented as PAP, ventricular septal defect and pulmonary atresia (Case 36), exhibiting slight differences from previously described postnatal GCPS phenotypes [24]. These observations highlight the significant phenotypic pleiotropy of GLI3 variants within the prenatal developmental window and effectively expand the established genotype‐phenotype spectrum.

In our initial cohort, 6.4% (3/47) of prenatally suspected polydactyly cases were ultimately confirmed as false positives. Similarly, Arduç et al. recently reported that among 124 pregnancies with a prenatal suspicion of polydactyly, two were rejected during follow‐up ultrasound examinations, and an additional 18 were postnatally confirmed to have normal digits [19]. Such false‐positive diagnoses are primarily attributed to the minute size of the digital structures, transient fetal hand clenching, maternal habitus, and the presence of soft‐tissue artifacts [8]. These factors underscore that a prenatal sonographic suspicion of polydactyly must be interpreted with caution. To mitigate the false‐positive rate, the integration of three‐dimensional (3D) ultrasonography, careful confirmation from multiple imaging planes, and serial longitudinal ultrasound evaluations are highly recommended [9, 12]. Ultimately, postnatal physical examination of the newborn remains the definitive gold standard for diagnostic confirmation [9].

We acknowledge several limitations in the present study. First, as a single‐center tertiary referral cohort, referral and ascertainment biases were unavoidable. Second, advanced genetic testing is generally self‐paid in China, introducing a socioeconomic bias that may have influenced case enrollment. Third, none of the TOP cases in the non‐isolated group underwent fetal autopsy, which limited comprehensive postmortem phenotypic characterization of these fetuses. Lastly, the follow‐up period was insufficient to capture potential subtle late‐onset neurodevelopmental issues associated with conditions such as the 22q11.2 microdeletion, which could potentially affect the accurate diagnostic yields of the isolated and non‐isolated subgroups.

5. Conclusion

In summary, fetal polydactyly demonstrates marked genetic and phenotypic heterogeneity. In our cohort of 44 prenatally confirmed fetuses, non‐isolated polydactyly, together with multi‐limb involvement, bilateral presentation, and a postaxial duplication axis, was strongly associated with an increased likelihood of underlying chromosomal or monogenic disorders. Importantly, apparently isolated polydactyly was not entirely benign, as clinically relevant genetic abnormalities were still identified in 9.5% of cases. These findings support the integration of detailed prenatal sonographic phenotyping with CNV‐seq and trio‐ES to improve diagnostic precision, refine prognostic assessment, and guide prenatal counseling and postnatal surveillance.

Funding

This work was funded by the Hubei Province Health and Family Planning Scientific Research Project (WJ2023M110).

Ethics Statement

This study adheres to the principles set out in the Declaration of Helsinki. Written informed consent was obtained from all participants, and the study was approved by the Institutional Ethics Committee of the Maternal and Child Health Hospital of Hubei Province (IRB No.2026‐010‐01).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Anatomical distribution, laterality, and duplication axis of fetuses with polydactyly.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.