Is It Feasible to Screen for Fetal De Novo or Paternally Inherited Pathogenic Single Nucleotide Variants in Maternal Plasma Cell‐Free DNA? A Systematic Literature Review

ABSTRACT

Objective

Monogenic disorders (MDs), often associated with developmental delay, intellectual disability, hypotonia, or dysmorphic facial features, typically go undetected during pregnancy. These disorders are frequently caused by de novo single nucleotide variants (SNVs), which are not currently covered by routine non‐invasive prenatal testing (NIPT). This screening gap limits informed decision‐making in pregnancy and can lead to the unexpected birth of neonates with severe conditions. The aim of this study was to look for evidence of whether de novo SNVs can be detected through NIPT and to assess the possibility of screening for autosomal dominant MDs in cell‐free DNA in maternal plasma.

Methods

A systematic literature review conducted on the 27th of February 2024 identified 12 studies examining NIPT of multiple genes associated with MDs. An additional citation analysis for the four most recent studies that were included in the systematic review was conducted on 10th of April 2025. Four additional studies met our inclusion criteria and were incorporated in the final analysis.

Results

The studies demonstrated that next‐generation sequencing of a gene panel or whole exome could detect pathogenic single nucleotide variants in fetuses with high positive predictive values 98.9% (66.7%–100%).

Conclusion

This review confirms that performing NIPT for de novo and paternally inherited pathogenic variants associated with MDs is technically possible. Ethical considerations, including disorder selection, variant disclosure, and the need for large‐scale implementation studies must be addressed to assess the potential risks and ensure effective and responsible implementation.

Summary.

• What's already known about this topic?

  • ◦Monogenic disorders represent a group of serious congenital disorders for which reliable prenatal detection remains limited and often unavailable.
  • ◦Non‐invasive prenatal diagnosis (NIPD) has been employed to detect specific pathogenic variants prenatally.
  • ◦Several recent studies showed that non‐invasive prenatal testing (NIPT) for monogenic disorders by using targeted NGS gene panel sequencing or whole‐exome sequencing of cfDNA may be feasible.

• What does this review add?

  • ◦Reviewed studies show a merged positive predictive value of 98.9% (range 66.7%–100%) for NIPT for de novo or paternally inherited (likely) pathogenic single nucleotide variants, which is comparable to NIPT for trisomy 21 and 18 and higher when compared to NIPT for other trisomies (trisomy 13 or rare autosomal trisomies) or structural chromosomal abnormalities.
  • ◦Multidisciplinary studies are needed and should include important aspects, such as disorder selection, variant disclosure, and ethical considerations.
  • ◦Implementation studies in general populations are lacking.

1. Introduction

Approximately 60% of severe monogenic disorders (MDs) are dominant disorders, most of which are attributed to de novo pathogenic/likely pathogenic (P/LP) single nucleotide variants (SNVs) [1]. Prenatal detection of these disorders typically relies on fetal anomalies that can be visualized by ultrasound. In current practice, invasive genetic diagnosis is only offered when fetal structural anomalies are detected, and in many centers, exome or genome sequencing is performed to determine if the fetus is affected [2, 3, 4]. However, numerous MDs remain undiagnosed in the prenatal period, as their associated features—such as developmental delay/intellectual disability, hypotonia, or dysmorphic facial features—are not visible on ultrasound. This diagnostic gap leads to the unexpected birth of neonates with severe pediatric conditions [5]. Importantly, if there is no indication for invasive prenatal testing, then these genetic syndromes may still be detectable through non‐invasive prenatal testing (NIPT).

Since the discovery of placental cell‐free DNA (cfDNA) in maternal blood [6], NIPT has become the leading screening method to detect common fetal (sub)chromosomal aberrations [7]. NIPT based on next‐generation sequencing (NGS) enables analysis of the fetal genome and contributes to clinical decisions and management strategies for healthcare providers and parents. Lo and colleagues [8] also showed that sequencing of the whole fetal genome in maternal plasma is technically possible.

Detection of known P/LP variants in cfDNA is already performed in clinical settings. Achondroplasia was the first MD detected by non‐invasive prenatal diagnosis (NIPD) in 2000 [9], and in 2012, it was introduced into clinical practice by the UK National Health Service together with the detection of thanatophoric dysplasia [10]. Since then, the number of autosomal dominant conditions for which NIPD is possible has expanded, with for example, Crouzon syndrome and osteogenesis imperfecta [11, 12]. NIPD has been designed for the detection of specific SNVs, which is usually performed using digital PCR, Sanger sequencing, or targeted NGS, but these approaches were limited to detecting known P/LP SNVs in particular target regions [13, 14]. Although many studies have focused on NIPD, NIPT for MDs is not yet clinically available for the general population.

To be able to offer NIPT for MDs in a general population, the target of such tests would need to be much broader and not limited to a small target region. This study aimed to evaluate existing literature on the detection of de novo and paternally inherited SNVs through NIPT, specifically focusing on studies that examined multiple genes, as the goal was to assess the feasibility of using this approach as a screening test, rather than a diagnostic test for specific disorder, in order to screen for autosomal dominant MDs in cfDNA from maternal plasma in the general population.

2. Methods

2.1. Data Sources and Search Strategy

The methodology of this review is reported according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analysis (PRISMA) checklist (Supporting Information S2). The search was conducted by an experienced information specialist (WMB) in four electronic databases: Medline via Ovid, Embase via Embace.com, Web of Science Core Collection via Clarivate, and the Cochrane Central Register of Controlled Trials via Wiley, from inception until February 27th, 2024. To retrieve relevant studies, variations of the following terms were used: (1) non‐invasive prenatal testing/screening, cell‐free DNA, or maternal blood/plasma screening, combined with (2) monogenic disorders or single nucleotide variants.

Articles that focused on neoplasms or aneuploidy were excluded from the search results, as well as conference abstracts, articles reporting animal studies, and articles in languages other than English. The entire search strategy is given in the Supporting Information S1.

Articles were deduplicated in EndNote with the method as described by Bramer et al. [15]. Covidence software [16] was used to select studies and manage the retrieved articles. After excluding duplicate papers, study titles and abstracts were independently assessed by at least two of four reviewers (K.V., K.E.M.D., S.L., M.I.S.) based on the eligibility criteria. Full‐text screening was again done by K.V. and at least a second reviewer (K.E.M.D., S.L., or M.I.S.). Disagreements during the selection process were resolved by discussing with a third or fourth reviewer (K.E.M.D., S.L., M.I.S.).

Additionally, a citation search was conducted by two reviewers on April 10th, 2025 for four most recent studies [17, 18, 19, 20] included in the systematic review. Citation search was performed using Google Scholar to identify articles that had cited these studies. The title and abstracts of all cited articles were independently screened by two authors (K.V., M.I.S.) based on the predefined criteria. Full text screening and data extraction were done again by K.V. and at least a second reviewer.

2.2. Eligibility Criteria for Study Selection

Studies were eligible if:

1. they covered prospective, retrospective, cross‐sectional, or method validation studies,

2. they used plasma cfDNA from pregnant women,

3. cfDNA testing was designed to detect MDs, and

4. at least two genes were tested in cfDNA.

We excluded:

1. review articles, commentaries, and opinion papers (unless new data was generated), conference abstracts, and case reports,

2. studies on NIPT for chromosomal aneuploidies,

3. studies on prenatal diagnosis using amniotic fluid cells, chorionic villi sampling, or fetal cells in maternal plasma,

4. studies on NIPD for MDs targeted to one gene,

5. studies exclusively on known parental carriers of P/LP variants, and

6. studies on conditions that are not monogenic syndromic disorders: for example, blood group tests, paternity tests, cancer, organ transplantations, and infections.

2.3. Quality Assessment and Data Extraction

Due to the small number of studies, which all described small study samples, the prevalences of single gene disorders were not calculated. All studies focused on highly selected patients, which leads to unavoidable biases and potentially limited generalizability. Quality assessment of the selected publications was performed according to the Newcastle–Ottawa Scale modified for this study (Supporting Information S1: Tables S1a and S1b).

The quality criteria deemed most important for high accuracy were:

1. “Selection” consisted of three key questions designed to evaluate the suitability of the studies. These questions focused on the characteristics of the cohorts of pregnant women, the material, and the design of the NIPT approach.

2. “Comparability” depended on whether the cohort size exceeded 20 cases and if the gene panel included more than 25 genes. The quality of the description of the disease selection process and the sequencing method used in the studies were also assessed.

3. “Exposure/Outcomes” was dependent on the quality of the data description and the reported details of the detected variants and confirmation studies.

The results of the studies could not be used for further meta‐analysis because of the limited number and high heterogeneity of the studied cohorts in methods used. Statistical heterogeneity could not be investigated as the cohorts were unlikely to represent the general population of pregnant women.

Two of the four reviewers independently extracted data (K.V., K.E.M.D., S.L., M.I.S.) with disagreements being resolved by a third or fourth reviewer (K.V., K.E.M.D., S.L., M.I.S.).

3. Results

3.1. Study Selection From Systematic Review

After the additional removal of two duplicates by Covidence, our search resulted in 1240 studies (Figure 1). Based on titles and abstracts, 1190 studies were excluded, because they did not meet the inclusion criteria. Full text was assessed for 50 studies. In total, 36 papers were excluded after full‐text review because they did not meet the inclusion criteria. Studies from Zhang et al. (2019) [21] and Nguyen et al. (2023) [22] were excluded after the data extraction step due to data overlap with other already included publications of these research groups: papers Mohan et al. (2022) [23] and Tran et al. (2023) [24], respectively. Based on the inclusion and exclusion criteria, we finally included 12 studies in our review [24]. All reviewers agreed on the sufficient quality and relevance of these studies. Supporting Information S1: Tables S1a and S1b demonstrate the scores of the quality assessment.

FIGURE 1.

Flowchart illustrating the search strategy and study selection process for non‐invasive prenatal testing (NIPT) for de novo single nucleotide variants using cell‐free DNA. NIPD, non‐invasive prenatal diagnosis; SNV, single nucleotide variant.

3.2. Study Selection From Citation Analysis

From the citation search, 49 new articles were found. Based on the title and abstract screening, 10 studies were selected for full text review, from which four [25, 26, 27, 28] met the inclusion criteria specified for the systematic search and were subsequently included to update and expand the findings of our systematic review.

3.3. Systematic Literature Search Results

3.3.1. Overall Outcomes

The 12 included studies were published between 2016 and 2024, and all investigated NIPT for detecting de novo P/LP SNVs in cfDNA from maternal plasma. Two different sequencing strategies were used. The majority (n = 10) [19, 20, 23, 24, 29, 30, 31, 32, 33, 34] utilized targeted sequencing to detect P/LP SNVs in selected genes in the fetal genome, while two studies [17, 18] assessed the diagnostic accuracy of NIPT for MDs associated with de novo P/LP SNVs using whole exome sequencing (WES). The characteristics of all studies are given in Table 1 and more details in Supporting Information S1: Tables S2a and S2b.

TABLE 1.

Twelve relevant studies [17, 18, 19, 20, 23, 24, 29, 30, 31, 32, 33, 34] retrieved in the systematic search; more details can be found in the Supporting Information S1: Tables S2a and S2b.

Study	Adams 2023	Brand 2023	Dan 2016	Dello Russo 2019	Malcher 2018	Miceikaitė 2023	Mohan 2022	Tran 2023	Wang 2021	Xu 2022	Yan 2020	Zhang 2024
Duo/Trio analysis	Duo/trio	Duo	Trio	Trio	cfDNA	Trio	Trio	Trio	Duo/trio	cfDNA	Trio	cfDNA/duo
No of pregnancies tested for MDs	228	51	5 (2 controls)	125	7	36	2208	13	36	1129	13	1003
GA (weeks)	9.6–37.3 (avg 16.2)	All trimesters, NS	22–28 2/7	10 (only for 1 case)	23–33	10–21	9–40.1 (avg 15.9)	20.5 ± 7.00	11–21.1 (avg 14.3)	11.1–32.7 (avg 15.3)	17.3–36.6 (avg 22.3)	13–32 (avg 22.5)
FF (%)	Not shown	6–51	6.84–30.56 (min 5)	12 (only for 1 case)	7–22	3.72–19	> 4.5	> 4	5.3–30.9	4.1–30.9	8–32	10.6 (12–18 wg) 11.7 (19–24 wg) 17.2 (≥ 25 wg)
Aim of testing	AD‐MDs	AD and AR MDs (CNVs)	de novo P‐SNVs (lethal SDs)	AD, AR MDs	MDs (CNVs, ANEU, chrY)	MDs (CNVs, ANEU)	AD‐MDs	AD‐MDs	AD‐SDs	MDs (CNV, ANEU)	AD‐MDs	AD‐MDs (CNV, ANEU)
Gene panel (no of genes)	30	WES	16	13	497	WES	30	30	5	9	30	75
No of pregnancies with abnormal results (AD‐MDs)	8	1	3	0	5	5	125	13	0	8	8	37
No of pregnancies with P/LP variants detected in cfDNA and confirmed in fetal material	4 AC/CVS 1 PB, 1 CB, (2 not tested)	1	3 AC	0	0 (4 PB, 3 not tested)	5 AC, CVS	12 AC, CVS 4 PoC (10 known affected parents, 8 parental blood, 11 postnatal clinical examination, 23 postnatal blood, 57 not tested)	13 AC, CVS	0	8 AC, CVS	8 invasive tests	36 AC, PoC (1 CB)
No of TP‐ AD P/LP variants in cfDNA	6	1	3	—	4	5	67	13	0	8	8	37
No of FP‐ AD P/LP variants in cfDNA	0	0	0	—	0	0	0	0	0	7 a	0	0
Pregnancy outcomes (positive cases)	5 liveborn 3 TOP	NA	NA	NA	NA	NA	40 liveborn 7 TOP 4 stillbirth 13 postnatal demise 2 in utero demise 59 no info	1 liveborn 12 TOP	NA	NA	NA	4 liveborn 29 TOP 4 no info
Sensitivity (%)	NA	Median 96,3 (de novo/paternal SNVs)	TPR: 92.41–100 PPV: 93.97–100 (all SNVs)	NA	NA	100 PV, 95,12 all SNVs	NA	100	100 (spike‐in samples)	100	100	100
Specificity (%)	NA	NA	NA	NA	NA	NA	NA	NA	100	99.3–100 a	100	100
PPV = TP/(TP + FP) for P/LP variants (%)	100 b	100 b	100 b	NA	100 b	100 b	100 b	100 b	NA	NA	100 b	100 b

Note: Only the first author of each study is given.

Abbreviations: AC, amniocentesis; AD, autosomal dominant; ANEU, aneuploidies; AR, autosomal recessive; Avg, average; CB, cord blood; cfDNA, cell‐free DNA; chrY, chromosome Y; CNV, copy‐number variant; FF, fetal fraction; FP, false positive; GA, gestational age; LP, likely pathogenic variant; MD, monogenic disorder; NA, not assessed; NGS, next‐generation sequencing; No, number; NS, not specified; P, pathogenic variant; PB, peripheral blood; PoC, products of conception; PPV, positive predictive value; SD, skeletal dysplasia; SGD, single‐gene disorders; SNV, single nucleotide variant; TOP, termination of pregnancy; TP, true positive; TPR, true positive rate; WES, whole exome sequencing; wg, week of gestation.

^a The source of the false positive sample is not specified in the paper.

^b Additional calculation.

Six studies specifically focused on detecting P/LP SNVs linked to autosomal dominant disorders [20, 23, 24, 29, 31, 34]. Among these, four studies [20, 23, 24, 29] used the same NGS panel with 30 genes, first described by Zhang and colleagues [21]. This panel consists of 30 genes associated with 25 autosomal dominant disorders with a combined population incidence of 1 in 600.

Wang and colleagues [31] included the FGFR2, FGFR3, COL1A1, COL1A2, and COL2A1 genes in their NGS‐panel. These five genes are associated with 11 common autosomal dominant disorders: thanatophoric dysplasia type I–II, achondroplasia, osteogenesis imperfecta type I–IV, achondrogenesis type II, Apert syndrome, Crouzon syndrome, and Pfeiffer syndrome.

Dan and colleagues [34] prepared a targeted NGS panel consisting of 16 genes to detect lethal skeletal dysplasias, such as achondrogenesis type IA, IB and II, hypochondrogenesis, thanatophoric dysplasia types I and II, short rib‐polydactyly syndrome types I, IIA, IIB, and III, fibrochondrogenesis type I, atelosteogenesis type 2, perinatal osteogenesis imperfecta, and hypophosphatasia.

Six studies investigated not only de novo dominant P/LP SNVs but also other clinically relevant findings as well [17, 18, 19, 30, 32, 33]. Dello Russo and colleagues [33] prepared a small NGS panel comprising 98 amplicons covering 13 genes to detect not only autosomal dominant disorders, but also common autosomal recessive disorders. Brand and colleagues [17] expanded the scope by using WES to identify SNVs and indels associated with both autosomal dominant and recessive disorders, and they also detected CNVs. Five studies reported the detection of various fetal anomalies within a single screening test: both (sub)chromosomal aberrations and P/LP SNVs in genes [17, 18, 19, 30, 32].

Of the 12 studies included in our systematic review, only four [19, 20, 23, 24] reported outcomes for pregnancies with positive findings, such as a number of live births, terminations, or fetal demises. Detailed information about pregnancy outcomes for each study is summarized in Table 1.

3.3.2. The Study Samples

The study samples in each study comprised pregnant women with indications for prenatal invasive testing. All studies (n = 12) focused on a high‐risk population, for example, fetuses with ultrasound anomalies suggestive of MDs or patients with other indications for invasive prenatal testing such as advanced paternal/maternal age, abnormal screening results (e.g., increased nuchal translucency), or family history (e.g., previous child or family member known with a MD) [17, 18, 19, 20, 23, 24, 29, 30, 31, 32, 33, 34].

Sample sizes varied across the studies, with the number of pregnancies tested ranging from five [34] to 2208 [23]. Testing was also conducted in all three trimesters of pregnancy, depending on the study design. Notably, none of the studies included pregnancies before 9 weeks of gestation.

The studies were highly biased due to the selective nature of the study samples. The samples were mainly composed of small samples or focused on specific populations, such as high‐risk pregnancies, which limits the generalizability of the findings. This selection bias can result in skewed data towards more severe phenotypes and does not accurately reflect the prevalence of MDs in the general population.

3.3.3. Trio/Duo Analysis

Even though NIPT studies aim to optimize testing by obtaining results solely from sequencing the cfDNA, most studies (n = 10) also investigated maternal genomic DNA (gDNA) [17, 20, 31] or even paternal [18, 20, 23, 24, 29, 31, 33, 34]. According to the authors, this inclusion enhanced the interpretation of the sequencing results.

3.3.4. Fetal Fraction (FF)

FF data were reported in 11 studies [17, 18, 19, 23, 24, 29, 30, 31, 32, 33, 34]. However, only six studies specified the procedure for determining the FF in the samples. The methods used to determine the FF in the included cfDNA samples consisted of: detection of the Y chromosome [30, 33], fragment size identification [17], or SNP‐based methods [17, 23, 30, 32, 34]. FF varied between studies depending on the study design (Table 1). Most studies showed FF > 4%, but most tested pregnancies were advanced in gestation.

3.3.5. Sensitivity and Specificity of the Findings in cfDNA

An overview of the MDs detected in the cohorts included in this review is given in Table 1 and Supporting Information S1: Tables S2a and S2b. Four studies [19, 24, 29, 30] reported 100% sensitivity, that is, all cases that had been found through invasive methods were also detected from cfDNA. These four studies all employed targeted sequencing, with a panel size ranging from nine genes [30] to 75 genes [19], and NIPT was conducted as early as in the 11th week of gestation [30] and as late as the 37th week of gestation [19].

Dan et al. (2016) [34] calculated a positive predictive value (PPV) based on the number of true positive fetal specific SNVs shared by the cfDNA sample and fetal gDNA sample and false negative variants, which were detected in fetal gDNA but absent in cfDNA. The PPVs ranged from 93.97% to 100% in the five analyzed cases. All P/LP variants, which were detected in three cases with a gestational age ranging from 22 to 28 weeks, were confirmed by sequencing amniotic fluid samples.

In four studies, all negative cases were confirmed as well [19, 29, 30, 31]. Three studies [19, 29, 31] reported 100% specificity, with all negative cases confirmed by invasive tests [19, 29] or WES on umbilical cord blood [31]. Xu et al. (2022) [30] reported 99.8% specificity, with 7 out of 1129 cases being false positives. However, the authors did not specify whether false positives concerned MDs cases or fetuses with (sub)chromosomal abnormalities, which were also included in their screening strategy.

Two studies [20, 23] reported no false‐positive and false‐negative cases, but not all cases were confirmed using additional methods.

Two studies [17, 18] performed whole exome sequencing of cfDNA from pregnancies across all three trimesters, starting as early as the 10th week of gestation [18]. Sensitivity for SNV detection for P/LP SNVs was 100% in the study from Miceikaite et al. (2023) [18], but 95.12% for all SNVs across the coding region. Brand et al. (2023) [17] reported a median sensitivity of 96.3% for de novo or paternally inherited SNVs and 81.9% for de novo or paternally inherited indels. Confirmations were performed using invasive tests [17, 18] or using parental gDNA and cord blood [17].

Detailed information about positive findings, confirmatory testing, sensitivity and specificity for each study is shown in Table 1.

3.3.6. Additional Results From Citation Analysis

Detailed information from each study that met inclusion criteria can be found in Table 2. Three studies [25, 26, 28] focused specifically on NIPT for AD MDs, and one study [27] preformed combined NIPT for aneuploidies and AD MDs as well. All studies used [25, 26, 27, 28] panel approach, and panel sizes ranged from 30 [27, 28] to 64 [26] genes. Only one study [28] was conducted on a low‐risk population. Wang and colleagues [26] showed lower PPV (66.7%) with two false positives compared to other studies [25, 27, 28], that all showed 100% PPV without any false positives. False positive NIPT results in the study from Wang at al. (2024) [26] were detected by NIPT in KMT2D gene (NM_003482.4: c.6714dupC (p.Ser2239Lfs*) and c.12862delC (p.Arg4288Gfs*), both likely pathogenic). Both variants were confirmed as mosaicism in maternal leukocytes (9.1% and 20.5% mosaic) [26].

TABLE 2.

Four additional relevant studies [25, 26, 27, 28] from citation analysis included in the final results.

Study	Adams 2025	Luong 2025	Wang 2024	Zhang 2025
Duo/Trio analysis	Duo/trio	cfDNA a	cfDNA	Trio
No of pregnancies tested for MDs	2745	116	112	535
GA (weeks)	9–40 (avg 11.2)	22.3 ± 4.7	17–30 (avg 24 ± 3)	12.0 and 33.7 (avg 20.8)
FF (%)	3–33 (avg 9.7–10.7)	> 4	4.8–27.1 (avg 15.09)	5.6–24.3 b
Aim of testing	AD MDs	AD MDs (ANEU)	AD MDs	AD MDs
Gene panel (no of genes)	30	30	64	34
No of pregnancies with abnormal results (AD‐MDs)	14 (6 fetal, 8 maternal)	8	6	11
No of pregnancies with P/LP variants detected in cfDNA and confirmed in fetal material	6/6 fetal variant (perinatal diagnostic test)	8	4 AC	10 invasive (1 postnatal)
No of TP‐ AD P/LP variants in cfDNA	6	8	4	11
No of FP‐ AD P/LP variants in cfDNA	0	0	2	0
Pregnancy outcomes (positive cases)	9 well baby nursery 3 TOP 2 neonatal intensive care unit	NA	5 labor induction, 1 retain the fetus	5 normal delivery, 5 TOP, 1 unknown
Sensitivity (%)	NA	88.9	100	100
Specificity (%)	NA	100	98.1	100
PPV = TP/(TP + FP) for P/LP variants (%)	100 c	100	66.7	100 c

Note: Only the first author of each study is given.

Abbreviations: AC, amniocentesis; AD, autosomal dominant; ANEU. aneuploidies; Avg, average; cfDNA, cell‐free DNA; FF, fetal fraction; FP, false positive; GA, gestational age; LP, likely pathogenic variant; MD, monogenic disorder; NA, not assessed; No, number; P, pathogenic variant; PPV, positive predictive value; TOP, termination of pregnancy; TP, true positive.

^a Parental sample collected additionally if positive findings were obtained.

^b For positive cases only.

^c Additional calculation.

Based on the findings from the studies included in our systematic review, the PPV of NIPT for detecting de novo or paternally inherited SNVs using gene panels or WES NGS appears to be 93%–100%. Three studies from citation searches have shown 100% PPV [25, 27, 28], except one study that has shown only 66.7% PPV [26] due to two false positive cases caused by maternal mosaicism.

3.3.7. PPV for P/LP De Novo and/or Paternally Inherited SNVs in Merged Cohors

To analyze the PPV specifically for P/LP de novo and/or paternally inherited SNVs, we conducted additional calculations using true positive and false positive SNV data reported in each study. In all studies included in our systematic search and citation analysis where these calculations were feasible, a merged PPV of 98.9% (range 66.7%–100%) was observed [17, 18, 19, 20, 23, 24, 29, 32, 34]. Detailed information about the number of pregnancies tested, false positives and true positives are shown in Table 3.

TABLE 3.

Merged positive predictive value (PPV) for 13 included studies [17, 18, 19, 20, 23, 24, 25, 26, 27, 28, 29, 32, 34] from systematic literature search and citation analysis.

No of studies included	13
Total no of pregnancies tested for MDs	7072
Total no of TP‐AD P/LP variants in cfDNA	173
Total no of FP‐ AD P/LP variants in cfDNA	2
PPV = TP/(TP + FP) for P/LP variants (%)	98.9%

Abbreviations: AD, autosomal dominant; cfDNA, cell‐free DNA; LP, likely pathogenic; MD, monogenic disorder; No, number; P, pathogenic; PPV, positive predictive value; TP, true positive.

4. Discussion

The aim of this study was to provide a comprehensive summary of the currently published studies on NIPT for P/LP de novo and paternally inherited SNVs associated with autosomal dominant MDs using maternal plasma cfDNA sequencing and to assess the suitability of this method for clinical practice. Based on the reviewed studies, it is evident that NGS‐based NIPT for MDs holds significant potential for clinical practice. Our review indicates that both gene panel sequencing [19, 24, 29, 30, 34] and exome analysis [17, 18] demonstrate high sensitivity and PPVs in high‐risk populations. The merged PPV of NIPT for de novo and paternally inherited P/LP SNVs were 98.9% (range 66.7%–100%). A common limitation across the reviewed studies was the absence of confirmatory testing for all NIPT results, which hindered the accurate determination of true sensitivity and specificity. However, several studies that confirmed both positive and negative NIPT results through invasive techniques or postpartum have 88.9%–100% sensitivity and 98.1%–100% specificity [19, 22, 29].

Interestingly, the PPV of NIPT for chromosomal aberrations in high‐risk populations was reported to be 94% for trisomy 21, 80% for trisomy 18, 67% for trisomy 13, 15% for rare autosomal trisomies (RATs) and 50% for subchromosomal aberrations [35]. Our review indicates that PPVs for NIPT for P/LP SNVs associated with MDs are higher than those for NIPT for trisomy 13, RATs, or subchromosomal aberrations in articles published to date. Therefore, we believe that the reviewed papers show promising results to start a large‐scale clinical implementation study in the general population. The study by Zhang and colleagues [19] focused on both aneuploidies and de novo P/LP SNVs, and all eight false positive results out of 876 tested samples were chromosomal aberrations, while there were no false positive cases for NIPT for MDs. One possible explanation for this discrepancy is the biological phenomenon of chromosomal mosaicism [35, 36]. Mosaicism is less common for SNVs than (sub‐)chromosomal aberrations [37], resulting in a higher predictive power and a higher PPV [17, 18, 19, 20, 23, 24, 25, 26, 28, 29, 30, 31, 32, 33, 34]. However, Wang and colleagues (2024) [26] described two false positive NIPT results of Kabuki syndrome. In the first case, FF was 8.95% and the proportion of variation was 17%, and in the second case FF was 21% and the proportion of variation was 6.6%. Discrepancies in FF and the proportion of pathogenic variants may indicate that the variants could be of maternal origin. An accurate configuration of variant filtering that incorporates FF information can lead to better analytical accuracy [17]. Both the fetal fraction or ultrasound findings may influence the PPV of NIPT for P/LP SNVs; however, a recent study by Adams et al. (2025) [28] suggests high efficacy of such screening for monogenic disorder even early in pregnancies without ultrasound anomalies.

A crucial question, which has not yet been answered, is what the PPV would be in a general population, and also which disorders could and should be included in NIPT‐MD if it was implemented in the general population. Adams et al. (2025) [28] showed very promising results in 3480 low‐risk pregnancies, with 73.6% under 11 weeks of gestation. They confirmed all six fetal NIPT‐positive findings invasively or postnatally. However, not all NIPT results were confirmed, precluding the estimation of specificity. Careful selection of conditions for screening seems to be required, guided by the “SEPH” strategy: Severe outcome, Early onset, Prevalence, and High analytical performance [19]. Several studies [20, 23, 24, 29] in our systematic review used a gene panel first introduced by Zhang et al. (2019) [21]. This panel focused on 30 genes that were associated with 25 autosomal dominant disorders with significant health impacts and a cumulative prevalence of 1 in 600. The capability of this panel to detect P/LP de novo SNVs in fetuses was confirmed in a study with the largest cohort to date, encompassing 2745 cfDNA samples [28], as well as in several smaller studies [20, 24, 27]. This panel is expected to detect over 80% of Noonan spectrum disorders, which are the second most common genetic cause of congenital heart defects and are associated with significant morbidity and mortality [38]. Adams et al. (2025) [28] also raised concerns regarding the inclusion of certain genes in this panel, as some of the genes are associated with clinical heterogeneity (e.g. PTPN11) and milder phenotypes (e.g. COL1A2) and may be difficult to counsel in low‐risk population with apparently healthy fetus as the prediction of the severity of the particular disorder is challenging. Complications may also arise when pregnant women are an undiagnosed carrier with a very mild phenotype (e.g. osteogenesis imperfecta), as decribed by Adams et al. (2025) [28]. Several papers focused on smaller and more specific panels detecting various types of skeletal dysplasias [31, 34], whereas others described a larger NGS‐panel to detect a broader range of MDs [19, 32]. WES‐NGS also appears to be a suitable approach for NIPT of MDs [17, 18]. This approach enables screening for a more significant number of genes, but interpreting WES results would require more time than a smaller gene panel analysis. The efficacy of WES‐based NIPT for MDs was also demonstrated in a study by Qi et al. (2024), in which they accurately identified 9 out of 10 P/LP SNVs. Studies indicate that a broad spectrum of MDs may be detectable using NIPT, particularly with the application of WES. However, further research is essential, especially regarding gene selection, as some genes associated with MDs may be more challenging to track.

NIPT‐MD has the potential to be implemented into existing pipelines for NIPT, and several studies suggest that its inclusion may increase the detection rates and offer a practical and impactful enhancement to conventional methods [18, 19, 27, 32], but further investigation and meta‐analysis studies are needed to assess its clinical value and feasibility in the general population [39]. Zhang et al. (2024) [19] reported a 60.7% increase in overall detection rates when including targeted detection of MDs. They demonstrated the use of NGS to detect various genetic conditions using cfDNA, including both (sub)chromosomal aberrations, and MDs. Among 876 fetuses with structural abnormalities on ultrasound, NIPT identified 55 aneuploidies (56.1%), 6 microdeletions (6.1%), and 37 single‐gene P/LP variants (37.8%), with all positive and negative cases confirmed by diagnostic tests.

The NGS‐based NIPT approach requires robust bioinformatics support and the use of precise tools to accurately distinguish fetal variants from sequencing errors [40]. Based on the studies included in our review, the interpretation of sequencing data is simplified when duo [17, 20, 31] or trio analysis is applied [18, 41], which can lead to better identification and classification of de novo/paternally inherited variants or in the case when pregnant women is an undiagnosed carrier with mild phenotype [28]. However, for the clinical use of NIPT for MDs, it would be more practical to validate this method using sequencing of cfDNA samples alone or through duo analysis (fetus/mother) as in clinical practice, obtaining samples from both parents is not always feasible, and the costs associated with trio analysis would naturally be higher. While our review focuses on detection de novo/paternally inherited SNVs, maternally inherited SNVs as well as other types of variants, such as de novo CNVs and indels inherited both paternally and maternally, also contribute to severe genetic disorders. These variants may be more challenging to detect with NIPT based on sequencing [18].

Expanding NIPT to detect not only chromosomal aberrations across the entire genome but also MDs raises significant ethical concerns not only about the phenotypic and genetic heterogeneity of many MDs but also about society’s readiness for such advanced genetic screening. Addressing the ethical implications of NIPT for MDs is crucial as this technology progresses [42, 43]. There are concerns that many genetic findings may not correlate with abnormal ultrasound results, and recent reports indicate a potential for more significant phenotypic variability than previously assumed [28, 44]. This uncertainty can lead to complex decision‐making scenarios for both expectant parents and healthcare providers [45]. Implementing such comprehensive screening without reliable support systems and societal consensus on managing findings could somewhat complicate the diagnostic process rather than improve screening [42]. Therefore, effective clinical implementation requires close collaboration among all stakeholders and the development of follow‐up guidelines [46, 47]. This emphasizes the need for large‐scale general population studies involving multidisciplinary teams, including laboratory experts, clinicians, and social scientists, to ensure thorough preparation of responsible genomic screening and to reduce potential harm.

5. Conclusion

This review demonstrates that NIPT for MDs is technically feasible, with high sensitivity and PPV observed for both targeted panels and WES approaches and can reliably detect P/LP de novo and paternally inherited SNVs (PPV 99.8%, range 66.7%–100%) associated with MDs. Further work is needed to address ethical considerations, selection of disorders to screen for, and establishing guidelines for data interpretation and reporting disorders of variable penetrance. Furthermore, large implementation studies involving multidisciplinary teams are necessary to evaluate all potential risks and benefits. It is essential to design an appropriate clinical strategy to introduce this screening method to ensure responsible implementation and public acceptance.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

Kristína Valovičová is a PhD student at Masaryk University (Project Code: MUNI/A/1556/2023). All authors declare no conflicts of interest.

Supporting information

Supporting Information S1

Supporting Information S2

Data Availability Statement

Data sharing not applicable—no new data generated, or the article describes entirely theoretical research.