Invasive Prenatal Diagnostics: A Cornerstone of Perinatal Management

Abstract

Since the 1950s, invasive prenatal diagnostics have played an integral role in perinatal management. However, its significance extends beyond detecting genetic abnormalities. This paper comprehensively reviews the indications for amniocentesis and chorionic villus sampling. Additionally, it examines various methods of genomic, infectious, and biochemical analysis, with a particular emphasis on the achievements of the last decade.

1. Introduction to Prenatal Diagnostics

In 1955, Fuchs et al. [1] and others [2,3,4,5,6] were the first to determine the sex of the foetus by examining the centrifuged cells of amniotic membranes. The samples were investigated for either the presence or absence of typical sex chromatin mass in cell nuclei. The mentioned chromatin accumulation was later discovered to be an inactive chromosome X and named Barr body after the discoverer. In subsequent years, intensive research regarding amniotic fluid composition was carried out, resulting in the first successful cell culture [7,8]. This method enabled the demonstration of the foetal karyotype for the first time, marking the beginning of a new era in genetic counselling.

In the 1950s and early 1960s, the only methods used in obstetric imaging were based on Roentgen radiation. The main aim of Pelvimetry was the full assessment of the pelvis [9]. For this purpose, X-ray images of four planes of the pelvis were projected. This was applied antenatally in order to foresee various complications, such as pelvic inadequacy, or intranatally in the case of a failure to progress during a trial of labour. Placentography is an arterial angiography of the placenta. It was crucial in establishing the absence of placenta praevia [10]. However, despite its undeniable usefulness, radiology has raised many concerns regarding the radiation hazard. The primary concern associated with antenatal radiation is the risk of malignancy. In 1956, Giles et al. [11], and later, others [12,13,14], traced an excess number of leukaemia and neoplastic conditions of the CNS, kidneys, and adrenal glands. The majority of reported cases were associated with foetal antenatal exposure to radiation. The previously mentioned complications were the primary reason for the withdrawal of radiologic examinations in pregnant women from everyday clinical practice.

The main milestone that opened the door to prenatal diagnostics was the introduction of the world’s first contact compound scanning machine, the Diasonograph, by Ian Donald et al. [15]. This visionary invention consisted of a probe mounted on a patient’s bed with piezo-electric transducer crystals responsible for transmitting and receiving echoes. The examination involved the slow movement of the probe across the abdomen in order to scan deep tissues from as many angles as possible. The whole process took approximately 2 min, culminating in a cross-sectional picture created on a cathode ray tube. The images presented were crude and lacked grayscale. However, this groundbreaking invention has broadened the horizons for further research [16]. It soon became obvious that an ultrasonographic scan of the foetal head, called cephalography, is useful in predicting foetal weight [17] and assessing intrauterine growth [18]. Moreover, in 1969, Campbell et al. [19] demonstrated, that a precise measurement of the biparietal diameter in the 20th to 30th week of gestation is a reliable method of determining the delivery date in pregnancies of uncertain maturity. Obstetric ultrasonographic examination soon established this as an inherent element, which has since been in use for many years. In 1975, Campbell et al. [20] improved this method by introducing the abdominal circumference measurement at the liver and umbilical vein level as a valuable predictor of foetal weight, as well as in growth-restricted foetuses. At this time, ultrasonography was only reserved for more advanced pregnancies. In 1973, Robinson et al. [21] introduced the first detailed biometry charts of foetal crown-rump length from the 7th to the 16th week of gestation. At that time, a tremendous development in the field of prenatal diagnostics was observed. It was initiated by Campbell et al. [22], who, in 1972, presented a case of prenatally diagnosed anencephaly, followed by elective termination of pregnancy. On this basis, he conducted systematic research regarding neural tube defects. In 1977, he reported [23] the valuable role of combining ultrasound examination with a measurement of alpha-fetoprotein in maternal serum and amniotic fluid in diagnosing spina bifida, anencephaly, encephalocele, and hydrocephalus. This was considered to be the first method of screening for foetal abnormalities in a high-risk group of patients. Alongside the aforementioned discoveries, ultrasonography became widespread in prenatal diagnostics. It was possible only in the late 1970s when the first commercial linear array real-time ultrasonographic scanner was introduced. As the probes were much more compact and moveable, it was possible to adjust the view’s angle and follow the foetus’s movement. Moreover, the machines were much less expensive, thus making ultrasonography more accessible to specialists and researchers. This allowed for visualising and measuring various anatomical structures that were not observed previously, such as the length of the femur. As a result, in 1985, Hadlock et al. [24] introduced an equation for estimating foetal weight and assessing foetal growth, which is still used to this day.

The 1980s were a time of significant advancements in prenatal screening. Until that time, invasive diagnostics for genetic reasons were offered only to women aged 35 and above, who were considered a high-risk group. However, this policy of screening based on maternal age resulted in an invasive testing rate of 5% and a detection rate of trisomy 21 of 30%. These highly unsatisfactory numbers demanded improvement. As a part of a search for new screening methods, Beryl Benacerraf et al. [25] first reported an increased thickness of a nuchal fold in foetuses with trisomy 21 in the second trimester. Later, she described other classic second-trimester sonographic markers, such as pyelectasis and a shortened femur. The ultimate breakthrough was the paper published by Nicolaides et al. in 1992 [26]. They demonstrated a pathological accumulation of fluid in the nuchal region as an early marker of trisomies 21, 18 and 13. Later, the nuchal translucency was proved to be useful in detecting many more foetal abnormalities, such as congenital heart defects [27], diaphragmatic hernias, exomphalos, body stalk anomalies, and much more [28]. In 1998, Matias et al. [29] observed a correlation between abnormal ductus venosus flow measured in 11–14-week gestation and chromosomal abnormalities. The blood flow during atrial contractions was absent or reversed in 90,5% of cases of foetuses with chromosomal defects. In 2001, Cicero et al. [30] observed a phenomenon of absence or hypoplasia of the nasal bone in cases of trisomy 21. This was observed in 73% of chromosomally abnormal foetuses irrespectively of increased nuchal translucency. This finding might have been caused by a delayed ossification of the nasal bone due to alterations within the extracellular matrix. Two years later, Huggon et al. [31] concluded that 83% of foetuses with tricuspid valve regurgitation in an 11–14-week scan proved to have karyotype abnormalities. The chromosomal defect most frequently found in those cases was trisomy 21; however, all types of karyotypic anomalies were seen in association.

As a result of all these findings, in the early 2000s, an original screening program called First-Trimester Screening (FTS) was introduced [32]. The idea was to counsel patients in one-stop clinics for assessment of risk (OSCAR) [33,34]. The proposed screening policy for trisomy 21 and other chromosomal abnormalities was based on maternal age, nuchal translucency, maternal free beta-hCG and PAPP-A. The overall individual risk for trisomy 21 was calculated by combining the primary risk and the likelihood ratio resulting from a measurement of the aforementioned markers. The background risk depended on maternal age and gestation. Patients would be categorised into a high-risk group with a risk cut-off of 1 to 100 with a detection rate of 90% and a false-positive rate of 5%. Moreover, the newly introduced ultrasonographic markers—ductus venosus flow impedance, tricuspid regurgitation, and the absence of nasal bone—proved to improve the screening performance to 93–96% and decrease the false-positive rate to 2.5% [35,36,37]. Since its introduction, the FTS programme has proven to be successful in various validation studies [38,39,40]. Therefore, it is used as a gold standard in many countries.

Apart from assessing the individual risk for trisomy 21, 18, and 13, FTS provides a broad spectrum of clinical implications. Ultrasound examination conducted at 11–14 weeks of gestation enabled the early detection of 60% of major foetal anomalies [41], such as neural tube and brain anomalies, abdominal wall, and cardiac and skeletal defects [42,43]. Along with constant development in ultrasound technology, the first-trimester ultrasound has become a tool for screening major morphological abnormalities. Therefore, the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) eventually published practice guidelines for anatomical assessment in the first trimester in 2013 [44], which were appropriately updated 10 years later [45]. Ultimately, an early anomaly scan provides the patient with extensive information regarding the foetus and may contribute to further invasive diagnostics.

Another milestone in prenatal diagnostics happened in 1997 when Lo et al. discovered cell-free foetal DNA in the plasma and serum of pregnant women through the extraction and amplification of Y chromosomal DNA from a plasma analyte [46]. DNA fragments were thought to originate from the placenta, where cytotrophoblasts and syncytiotrophoblasts go through cycles of fusion and apoptosis, releasing DNA into the bloodstream [47]. Cell-free DNA typically mirrors the genetic makeup of the foetus, as both the foetus and placenta normally originate from the same embryo [48]. The foetal component accounts for around 3–13% of the total cell-free DNA in maternal blood and rises as pregnancy progresses [49]. This finding indicated that maternal plasma or serum DNA may be an invaluable asset for non-invasive prenatal diagnosis. Validation studies conducted in the next decade led to the commercial availability of the first tests in 2012. Non-invasive prenatal testing has rapidly been implemented in aneuploidy screening since then. By late 2017, around 4 million to 6 million pregnant women had undergone cell-free DNA analysis to detect foetal aneuploidy [50]. Despite its high screening success rates, cfDNA is not as accurate as diagnostic testing and may produce false-positive results. Patients should not make pregnancy management decisions, such as termination, purely contingent upon NIPT [51]. False-positive screening test findings can be caused by confined placental mosaicism, vanishing twins, maternal chromosomal abnormalities and copy number variants, maternal malignancy, and maternal transplant recipients. While not prevalent, the aforementioned situations are occasionally observed in clinical practice [52,53,54,55,56]. In addition, the success rate of cfDNA testing in twin pregnancies decreases with maternal weight and is higher in IVF pregnancies compared with spontaneous conceptions. Screening twins with cfDNA testing is possible, although the failure rate is higher, and the detection rate may be lower compared with singletons [57]. Aneuploidy is not the only disease entity that is in dire need of being prenatally detected. Numerous companies providing non-invasive prenatal tests offer the choice to test for microdeletion and microduplication disorders. Screening for microdeletions and microduplications using cfDNA-based technology is a new method that lacks substantial clinical validation and is not endorsed by ACOG or SMFM. The high rate of false-positive and false-negative outcomes in non-invasive prenatal testing for subchromosomal abnormalities makes cfDNA-based screening for these conditions not of clinical benefit presently [58,59,60]. Prenatal screening for such disorders is hindered by a lack of comprehensive prospective clinical trials and insufficient outcome data. Professional recommendations currently do not suggest it for routine screening [61]. No screening procedure is flawless; each has its own strengths and weaknesses. Suciu et al. suggest that NIPT is most cost-effective when used as a tool in the contingent screening approach [62]. Non-invasive testing is favoured by most healthcare professionals and patients because of its safety compared with invasive procedures. With that in mind, NIPT does not provide the capability for whole-exome sequencing (WES), which requires amniocentesis or CVS samples.

Introducing cell-free DNA (cffDNA) into screening for foetal chromosomal disorders perfectly exemplifies the rapid changes in the field of modern genomics and prenatal diagnostics.

Simultaneously with research on cffDNA, methods for investigating the human genome were sought. DNA sequencing was first described in 1977 by Maxam et al. [63] and Sanger et al. [64]. Originally, it required an enzymatic denaturation and synthesis followed by gel electrophoresis of DNA fragments in order to analyse the sequence of the gel bands. This was effective only for short fragments of DNA (up to 1000–1500 base pairs) and limited to sequencing a single fragment at a time. A groundbreaking article, published in 2005 by Margulies et al. [65], introduced a technology that utilised microfabricated high-density picolitre reactors capable of performing simultaneous sequencing reactions in millions of small chambers. The concept of massive parallel sequencing (MPS) made it possible to sequence large genomes, such as the human genome, in much shorter timeframes and at lower costs than previously possible. Since that publication, sequencing technology has been intensively developed, leading to the emergence of various platforms that are now widely used in genetic research, medicine, biotechnology, and other scientific fields.

2. Techniques of Invasive Diagnostics

2.1. Amniocentesis

2.1.1. Genetic Analysis

Upon early detection of a foetal abnormality, invasive diagnostics should be offered to the patient. Foetal tissue can be obtained either by amniocentesis or chorionic villus sampling (CVS). Amniocentesis is a safe procedure that is usually carried out in an outpatient setting. It is performed by the insertion of a needle inside an amniotic cavity with constant ultrasound guidance. Its safety has been proven numerous times [66], but both maternal and foetal complications can occur. Therefore, patients should always be properly counselled and give written consent. After removing a sample of amniotic fluid, it undergoes a genetic analysis. Currently, a variety of tests are available, as described below.

Karyotyping—Historically, amniotic fluid routinely underwent a conventional cell culture. After 2–3 weeks, microscope visualisation of chromosomes was feasible. This method was relatively time-consuming and allowed for only a diagnosis of chromosomal number aberrations and significant structural aberrations (resolution limited to 5–10 Mb), neglecting a substantial group of structural defects. Therefore, quality alternatives were needed.

Quantitative Fluorescence Polymerase Chain Reaction (QF-PCR)—QF-PCR is a rapid method for the quantitative analysis of abnormalities at the chromosomal level. Amplification of repeat sequences of polymorphic loci by PCR with fluorochromic primers allows for the visualisation and quantification of the product by an automated DNA scanner [67]. The results are seen as “peaks” corresponding to the number of alleles of amplified genes present in the examined sample. Individuals with a trisomic karyotype will demonstrate three peaks, suggesting the presence of three copies of the same gene, thus enabling the diagnosis of trisomy. This method was initially introduced as a valuable alternative to a standard cell culture. Despite providing results for abnormalities of the chromosomal copies overnight, it was not accepted as a substitute for a classic cytogenetic analysis. It is unlikely to detect some abnormalities of high clinical significance, such as balanced translocations or chromosomal mosaicism [68]. Nevertheless, it is extremely useful in excluding the contamination of the sample with maternal tissue and is still valuable in this indication.

In-situ Hybridisation (ISH)—This method has been used in prenatal diagnostics since the beginning of the 2000s and is often referred to as “chromosome painting” [69]. After obtaining a sample, the genetic material is denatured to a single-stranded state and hybridised to fluorescent probes specific to certain chromosomal regions [70]. This allows for the detection and visualisation of precisely defined sequences, thus making them countable. Therefore, this genetic method provides a quantitative approach rather than a qualitative one. However, the results are usually available in 48 h, which makes it valuable for urgent clinical situations.

Comparative Genomic Hybridisation (CGH)—This method was first described in 1992 by Kallioniemi et al. [71] and was initially useful in oncology, allowing for the detection of discreet gene rearrangements. Eventually, it found a broad application in genetic and prenatal diagnostics. It is based on a standard FISH technique; however, its conceptual difference is crucial. A sample of DNA extracted from the examined material is labelled with green and red fluorescent markers. Subsequently, the test DNA is hybridised on a slide with reference DNA. The target DNA consists of previously prepared metaphase chromosomes derived from the cells of a karyotypically normal individual [72]. Finally, the imaging software compares the metaphase spreads and calculates the green/red ratio along the length of each chromosome. The changes in the green/red ratio correspond to the areas of genomic gains or losses. Therefore, it is possible to distinguish copy number variations (CNVs) with high sensitivity. The resolution of an array of comparative genomic hybridisation has evolved throughout the years; currently, the theoretical peak resolution is 1–3 kilobases [73]. It is important to note that advanced, non-invasive prenatal screening lacks the capacity to identify microdeletions and microduplications as effectively as array CGH performed on chorionic villi or amniotic fluid cells. Having access to foetal cell material with the means of CVS or amniocentesis is therefore preferred over non-invasive screening methods if the main goal is the detection of microdeletion and microduplication disorders [74].

A large study by Shaffer et al. [75] was the first to demonstrate the usefulness of CGH in prenatal diagnosis and its superiority over traditional karyotyping. Among over 5000 analysed cases, the overall detection rate of clinically significant CNVs was 5.3%, which increased to 6.5% in cases with abnormal ultrasonographic findings. Moreover, 71% of the reported abnormalities were <10 Mb and were not expected to be identified by conventional karyotyping. A prospective cohort study published in 2013 by Hillman et al. [76] presented an excess detection rate of abnormalities by CGH of 4.1% over conventional karyotyping among patients with abnormal foetal ultrasonographic findings. In subsequent years, the utility of CGH has been repeatedly proven by many studies [77,78,79,80]. However, the diagnostic yield depends on the targeted population. In a group of high-risk pregnancies with common aneuploidies excluded with rapid PCR tests, Tanner et al. [81] demonstrated that the overall diagnostic yield of CGH was 15.1% and increased to 20% in the case of multiple structural anomalies. It was lowest in samples taken because of isolated foetal growth restriction and abnormal FTS only (5.7%). According to Egloff et al. [82], the diagnostic yield was lowest among foetuses with isolated increased nuchal translucency and common trisomies excluded (2.7%). However, even in this group of patients, a significant number of pathogenic CNVs were detected. Therefore, CGH is considered to be a standard prenatal diagnostic tool in many countries.

Next-Generation Sequencing—While the introduction of comparative genomic arrays has undoubtedly revolutionised the field of genetic diagnostics, next-generation sequencing is a true achievement of modern genomics. It uses a wide range of tools to read and analyse a whole sequence of the examined genome [83]. With this method, it is feasible to detect single nucleotide variants (SNVs). Whole-exome sequencing (WES) is used to detect point mutations and small insertions–deletions within coding sequences. It reaches an unprecedented resolution. However, its utility for detecting CNVs is limited [84]. On the other hand, whole-genome sequencing (WGS) enables an analysis of both coding and non-coding sequences. With this feature, WGS is expected to outperform WES both in SNV and CNV detection [85,86]. Despite currently being the most advanced method of genomic examination, it is still underused as a first-tier tool. The probable reason for this is a lack of population reference maps of structural variations in national biobanks [87], which impedes clinical interpretation. Recently, it has been suggested that a combination of tools relying on different signal types would increase examination performance [87,88,89].

Even though the sequencing of protein-coding regions of genes plays an important role in postnatal diagnosis, its diagnostic yield of prenatal diagnosis is yet to be defined. Most recent studies have examined the detection rate of WES in cases of foetal structural abnormalities and normal karyotyping and CGH. The results vary greatly from 9.1% to 45.9% [90,91,92,93]. This is probably because of small cohort bias and a major discordance among the genetic causes of different structural anomalies. In 2024, Wang et al. [94] classified foetal structural anomalies into seven categories, according to the impaired system. The highest diagnostic yield was found in foetuses with head and neck abnormalities (40%) and skeletal abnormalities (39.1%). The overall additional diagnostic yield was 23.3%. On the other hand, Di Girolamo et al. [95] and Cao et al. [96] analysed the incremental diagnostic yield of WES among foetuses with increased NT and normal karyotyping and CGH without other structural anomalies. Those studies reported that between 4% and 8% of the analysed foetuses showed pathogenic variants detected exclusively by WES. Very little data are available on the incremental diagnostic yield of WGS over WES. A meta-analysis published in 2024 by Shreeve et al. [97] showed an increased detection rate of pathogenic variants of 1%, which was not statistically significant. However, the turnaround time for WGS was shorter, and it required less DNA. Despite the lack of evidence demonstrating the superiority of WGS over ES, its diagnostic yield will most likely increase with our advancing understanding of pathogenic variants of non-coding regions, and, eventually, it will become the most informative technique.

In response to the increasing accessibility and feasibility of exome sequencing (ES), the National Health Service England rolled out a rapid exome sequencing programme called the R21 pathway. It is undertaken for high-risk pregnancies in which genomic diagnosis could guide the management of the foetus. A sample obtained from CVS or amniocentesis is examined towards a nationally agreed panel of genes. Cases are selected within strict inclusion criteria, which provide optimal diagnostic yield for ES, and these are (1) foetuses with anomalies affecting multiple systems, (2) a suspected skeletal dysplasia, (3) large echogenic kidneys and a normal bladder, (4) major central nervous system anomalies in which a neural tube defect has been excluded, (5) multiple contractures (excluding bilateral isolated talipes), (6) an NT greater than 6.5 millimetres and at least one additional anomaly and in which the CGH is normal, and (7) non-immune foetal hydrops (fluid/oedema in at least two compartments detected at or after routine second-trimester scan) and a normal CGH. This approach guarantees an optimal diagnostic pathway for genetic disorders.

2.1.2. Intra-Amniotic Infections

Intra-amniotic infection (IAI) is proven to be one of the factors leading to preterm delivery (PTD) [98,99,100]. Moreover, there is strong evidence indicating a higher frequency of neonatal brain injury in foetuses exposed to chorioamnionitis [101,102]. Therefore, it appears crucial to eradicate microorganisms in order to prolong pregnancy and improve outcomes in patients with threatened preterm delivery and confirmed IAI. Numerous studies have shown that intraamniotic microbial invasion can be successfully treated with specific antibiotic regimens based on ceftriaxone, clarithromycin, and metronidazole [103,104,105]. However, the evidence is limited, and it does not directly demonstrate improvement in perinatal outcomes since the neonatal death rate and the incidence of bronchopulmonary dysplasia, periventricular leukomalacia, intraventricular haemorrhage, and necrotising enterocolitis were not improved [104]. In 2001, Chalupska et al. [106] proposed the following phenotypes of intrauterine inflammation: intra-amniotic inflammation with confirmed microbial invasion (MI) and sterile intra-amniotic inflammation without MI. A study conducted by Yoneda et al. [104] showed adverse effects of antibiotic therapy among patients with threatened preterm birth and microbe-negative amniotic fluid. At the same time, the authors emphasised the importance of IAI evaluation in management in PTD. Currently, the most credible method of diagnosing IAI is a direct bacterial culture or PCR analysis. However, in recent years, many biomarkers of intra-amniotic inflammation have proven to be valuable. In 2011, Oh et al. [107] first showed that matrix metalloproteinase-9 (MMP-9) and interleukin-6 (IL-6) obtained from amniotic fluid had a higher predictive value of IAI than maternal serum C-reactive protein (CRP). In 2015, Kim et al. [108] presented evidence that rapid tests of MMP-8 in amniotic fluid can identify 42% of spontaneous PTD among asymptomatic pregnant women. In 2017, Myntti et al. [109] analysed the concentration of proinflammatory biomarkers derived from activated and degranulated neutrophils. The authors concluded that the concentration of neutrophil-based amniotic fluid biomarkers, such as MMP-8, MMP-9, IL-6, myeloperoxidase (MPO), elafin, and tissue inhibitor of matrix metalloproteinases-1 (TIMP-1), were associated with IAI and MI. However, the number of inflammatory mediators was also reported simultaneously in maternal serum during IAI [110,111,112,113,114]. Therefore, amniocentesis is not routinely offered to patients with PTD.

2.1.3. TORCH Infections

Cytomegalovirus (CMV), Toxoplasma gondii (T. gondii), rubella virus (RBV), herpes simplex virus (HSV), varicella virus (VZV), and parvovirus B19 are classically referred to as TORCH infections [115]. Congenital TORCH infections are a well-described cause of stillbirths, foetal growth restriction, foetal abnormalities, neurodevelopmental impairment, and neonatal death [116]. Precise guidelines in the case of maternal infection are crucial in detecting or preventing a placental transmission.

Toxoplasmosis—International well-established antenatal screening of pregnant women allows for minimising the frequency of congenital toxoplasmosis. However, because of poor hygienic standards in some developing countries, the incidence rate persists at 1.5 per 1000 live births [116]. An endorsed screening protocol consists of T. gondii IgM and IgG, followed by IgG avidity, enabling more accurate timing of infection. In the case of a confirmed maternal infection, the foetus should be examined for transplacental transmission. In this case, amniocentesis after 18 weeks of gestation should be offered to the patient, followed by a PCR for T. gondii in the amniotic fluid [117]. The risk of transmission increases with gestation at the time of infection and varies from less than 15% at 13 weeks to more than 70% at 36 weeks [118]. In order to minimise the risk of transmission, the patient should start prophylaxis with spiramycin, which is a parasitostatic drug [119]. If a foetal infection is confirmed, the treatment is changed to pyrimethamine, sulfadiazine, and folinic acid (P + S + FA) [119,120]. However, this therapy can be associated with adverse effects. The most principal adverse effect is haematological toxicity and neutropenia, which are reported to occur in more than half of the treated neonates [121]. Therefore, a risk and benefit analysis should always be carried out before starting therapy, and postnatal management should not be neglected.

Cytomegalovirus—CMV is currently the most common factor in viral congenital infections, with a global prevalence of approximately 0.2–2% [122]. It is a major cause of central nervous system abnormalities in foetuses and sensorineural hearing impairment among neonates. Despite intensive awareness campaigns in recent years, there is still an unsettling deficiency of knowledge among healthcare providers. Because most adults exposed to CMV are asymptomatic, it is crucial to underline the importance of screening during pregnancy. Each patient should conduct an assay for anti-CMV antibodies in the first trimester. CMV-specific IgM antibodies are observed in serum up to 6–9 months after an acute phase of infection [123]. Therefore, the anti-CMV IgG avidity test is the most reliable source of information about the timing of exposure [124]. Avidity indicates the affinity of antibodies to bind the antigen, hence reflecting the maturation of the immunological response. Anti-CMV IgG with low avidity detected in mothers before 18–20 weeks of gestation identified foetal transmission with 100% sensitivity [125]. According to Yves Ville [126], ultrasound findings might be observed in no more than 5% of infected foetuses. Currently, amniocentesis and CMV-DNA PCR are integral parts of diagnosing a foetus with infection. The optimal timing of the procedure is 21–22 weeks of gestation [127]. This is because the virus requires 6–9 weeks to replicate and be excreted through urine [128]. This technique has 90–98% sensitivity and 92–98% specificity [128,129,130]. Historically, the topic of screening for CMV in pregnancy was controversial because of the lack of effective treatment. However, at the beginning of 2020, evidence began to emerge regarding the effectiveness of valacyclovir. In 2023, D’Antonio et al. [131] and Chatzakis et al. [132] showed a 65% reduction in the transmission rate in the first trimester and periconceptional period and a 70% decrease in infected neonates. Nevertheless, further clinical trials are needed.

Rubella—Following the Global Vaccine Action Plan (GVAP) introduced by the WHO [133], rubella infection has been declared eradicated in 51% of countries. It resulted in an increase in the percentage of infants vaccinated from 40% to 68%, according to the latest report by Ou et al. [134]. However, as a result of insufficient access to the rubella-containing vaccine (RCV) in low-income countries, more than 32,000 children are born with congenital rubella syndrome (CRS) each year [135]. This number is suspected to be underestimated because of the lack of surveillance of CRS in certain countries. Typically, neonates with CRS are growth-restricted and present symptoms depicted as Gregg Triad, which include the following: sensorineural deafness, cataracts and other ocular abnormalities, and cardiac anomalies [136]. Humans are the only source of infection, and the infectious period lasts from 8 days before the onset of the rash to 8 days after [137]. Maternal screening is based on anti-RBV IgG and IgM examination. Anti-RBV IgM antibodies are usually detected only for 2 months after the immunisation; hence, their presence strongly implies a primary infection. It was reported that some patients have the persistent presence of anti-RBV IgM antibodies; however, the prevalence is unknown because of the lack of population-based studies [138,139,140] Anti-RBV IgG avidity reaches a level considered high after around 13 weeks [141]. Therefore, it can be useful in specifying the risk of transmission. In order to diagnose a foetal infection, amniocentesis and amniotic fluid PCR tests are recommended [142]. Upon isolating the viral DNA in amniotic fluid, the patient should be properly counselled. Termination of pregnancy should be offered because of the high risk of neonatal impairment. Otherwise, pregnancy should be continued with strict ultrasound monitoring and postnatal management. No treatment has yet been proven to limit transplacental transmission.

Parvovirus B19 is a leading cause of rash illness among school-age children [143]. Its prevalence is seasonal and is observed to be highest in late winter and early spring [144]. Teachers and nursery workers are at the greatest risk. If a pregnant woman develops a B19 infection, the risk of transplacental transmission is 30% [145]. Most foetuses exposed to parvovirus B19 do not present any abnormalities [146]. However, it may lead to lethal consequences, such as severe foetal anaemia, myocarditis, congestive heart failure, and nonimmune hydrops foetalis [147]. The reason is its high affinity for human erythroid progenitor cells and cardiac myocytes. This leads to increased erythrocyte apoptosis in the bone marrow and liver, which results in high-output cardiac failure, general oedema, and stillbirth within a median of 3 weeks [148] if not treated. The incidence of prenatal infection is rare; hence, routine screening is not advised. Therefore, diagnosis usually begins with ultrasound findings, such as polyhydramnios, cardiomegaly, pleural effusion, placentomegaly, and ascites. The IgM assay is used to detect a recent infection for about 2–3 months [149]. IgG antibodies are present 10–14 days after contact and last for life [150]. However, definite confirmation of transmission is established with PCR of amniotic fluid [151]. Most mild to moderate cases of foetal anaemia are generally well-tolerated and tend to resolve without invasive intervention [152]. Severe cases usually require treatment. Cordocentesis allows for assessing foetal haematocrit, enhancing the severity of foetal anaemia, and transfusing packed red blood cells [153]. This regimen increases the survival rate of compromised foetuses to 82%, compared with 55% if not transfused. However, transfusions after the 35th week of gestation carry a greater foetal risk than immediate delivery [154].

Other—in 2016, Oliveira et al. [155] first reported the presence of Zika virus DNA in the amniotic fluid of a foetus with severe central nervous abnormalities. Since then, few studies have confirmed an existing incidence of congenital Zika syndrome [156,157,158,159,160]. Despite a lack of accepted guidelines of management in that matter, a study by Pereira et al. [161] published in 2019 showed the probable utility of amniocentesis in Zika virus testing.

2.1.4. Biochemical Analysis

Foetal lung maturation—In 1971, Gluck et al. [162] first introduced the concept of assessing the degree of foetal lung maturation (FLM) with the lecithin/sphingomyelin ratio in amniotic fluid. The idea was to estimate the risk of neonatal respiratory distress syndrome (RDS) in the case of an iatrogenic preterm delivery. Their study showed that the concentration of phospholipids in amniotic fluid, represented by lecithin, reflected alveolar stability. After 35 weeks of gestation, they observed a sharp peak in lecithin concentration, which, if referred to sphingomyelin, was a credible source of information regarding the chance of developing RDS. This finding was consistent with a low incidence of this complication after 34 weeks of gestation. However, further research concerning this topic showed very poor positive and negative predictive value for this examination [163,164]. Therefore, in 2019, the American College of Obstetricians and Gynaecologists (ACOG) provided guidelines to clinicians regarding the utility of FLM testing [165]. Since then, FLM testing is not recommended to guide the timing of delivery.

Rh Disease—One of the first reported applications of amniocentesis was the management of Rh disease. In 1952, Bevis et al. [166], and later, others [167,168], introduced a method to assess the severity of foetal haemolysis associated with maternal immunisation. In the case of detecting maternal antibodies in the Indirect Antiglobulin Test, amniocentesis was carried out in biweekly or weekly intervals. Subsequently, the samples were centrifuged and analysed with spectrophotometric tracing. An increased concentration of the bile pigment was found to be associated with the risk of foetal death. Therefore, labour induction was conducted at 36–38 weeks of gestation [167]. In 1963, Liley [168] reported the first successful intrauterine transfusion. After injecting a contrast material, an X-ray scan, an Amniogram, was taken. Using the gastrointestinal tract as a guide, Liley passed a needle through the maternal abdominal wall and the uterine wall into the foetal peritoneal cavity [169]. Subsequently, in group 0, Rh-negative blood from a donor was transfused. This is considered to be the earliest step of foetal interventions. However, because of its high sensitivity and non-invasiveness, ultrasound Doppler evaluation of median cerebral artery flow velocity is an accepted method of assessing foetal anaemia. Upon suspicion of anaemia, the diagnosis should be confirmed through the procedure of cordocentesis.

Alpha-fetoprotein (AFP)—First described in 1956 by Bergstrand [170], AFP is a foetal serum protein produced by the yolk sac, liver, and gastrointestinal tract mucosa [171]. It is excreted with urine and therefore can be detected in amniotic fluid from the 12th week of gestation to birth in variable concentrations [172]. It has found a wide application in predicting various foetal abnormalities. In the 1970s, it was proved that abnormally high levels of AFP were present in the amniotic fluid of foetuses with neural tube defects [173]. Since ultrasound was not yet commonly used, it received enthusiastic acceptance. In 1989, a study by Wald et al. [174] showed a significantly lower level of AFP in the maternal serum of patients with trisomy 21 pregnancy. This finding allowed for the introduction of a novel second-trimester screening method for Down syndrome called the quadruple test in 2003 [175]. While the usefulness of amniotic fluid-AFP in diagnosing foetal abnormalities is well established, a recent study by Huang et al. [171] indicated that it can be a valuable screening tool for adverse perinatal outcomes. In a group of more than 7000 patients, they showed that a concentration of AFAFP of >2 MoM had a high prediction rate of preterm delivery, stillbirth, foetal growth restriction, and hypertensive disorders. This suggests that routine assessment of AFAFP during amniocentesis conducted based on indications would provide early information regarding pregnancy risk without additional costs or risks.

Foetal Growth Velocity—In 2024, Vrachnis et al. tried to assess various markers of foetal growth velocity in amniotic fluid. In a group of 70 patients who underwent amniocentesis in the early second trimester, based on various indications, they found that amniotic fluid angiotensinogen concentration correlated with gestational age and birth weight [176]. This finding indicates a possible role of the renin–angiotensin system (RAS) in foetal growth velocity. In the same year, using the same group of patients, Maroudias et al. [177] demonstrated that calprotectin, a biomarker of neutrophil granulocytes, is significantly higher in patients with excessive foetal growth. A possible cause of this outcome may be exposure to a low-grade chronic inflammation state of LGA foetuses due to their increased adiposity. This leaves a possible field of intervention to reduce adverse perinatal outcomes.

2.2. Chorionic Villus Sampling (CVS)

Chorionic villus sampling (CVS) is an alternative procedure for invasive prenatal diagnostics. Its superiority over amniocentesis is the timing of its feasibility. It can be conducted from 10 to 14 weeks of gestation [178]. It provides an early genetic diagnosis and assessment of pregnancy management and, eventually, a less traumatic experience for the patient if termination of pregnancy is considered. The idea is to sample chorionic villi with an 18- or 20-gauge needle, from either a transabdominal or transvaginal approach. This way, it is possible to obtain a specimen of trophoblastic cells. In the vast majority of cases, the placental tissue DNA will be the same as that of a foetus. Therefore, the aforementioned methods of genetic analysis, such as CGH or NGS, can be used to assess chromosomal or subchromosomal aberrations. However, the main weakness of this method is the inability to detect cases of confined placental mosaicism (CPM). According to guidelines from the Royal College of Obstetricians and Gynaecologists (RCOG) from 2022 [179] regarding CVS, the overall prevalence of CPM is approximately 1–2%. Therefore, in the absence of structural anomalies, only the examination of amniotic fluid should be considered. However, if mosaicism is observed in amniotic fluid, the reliable way to rule out true foetal mosaicism (TFM) is the karyotyping of foetal lymphocytes collected in cordocentesis [180].

Credible data regarding the procedure-related risk for complications of CVS is limited. In a systematic review of 29 studies by Mujezinovic et al. [66] published in 2007, it was shown that the overall risk of pregnancy loss was similar to that related to amniocentesis, and it was 0.7% within 14 days after the procedure. In an update from 2017 [181], Alfirevic confirmed the previous findings. However, it is worth noting that the risk of CVS will always be at least slightly higher because of a higher rate in 10–13 weeks of gestation. Moreover, Bakker et al. [182] noticed that the learning curve of the CVS procedure is significantly longer and that an operator’s expertise has a considerable influence on the outcome. Other less worrying complications are transient leakage of amniotic fluid, vaginal spotting, and culture failure. The overall incidence is reported to be lower than 0.5% [183].

It is also worth noting, that the aforementioned procedures are also applicable in multiple pregnancies. The technique of amniocentesis and CVS does not differ from singleton pregnancies; however, frequently, two samples are required. Therefore, the procedure-related risk is likely to be higher [179]. The most recent data come from a paper from 2022 by Navaratnam et al. [184]. Their retrospective cohort study assessed the risk of foetal loss following amniocentesis and transabdominal CVS in both DCDA and MCDA twins. Their study indicated a significant increase in miscarriage both in the DCDA and MCDA groups versus uncomplicated control twin pregnancies. However, if only genetically and structurally normal foetuses were considered, the increase in pregnancy loss was not statistically significant. These findings underline a substantial limitation of their study, which was the construction of the control group. The most appropriate control group should be composed of twin pregnancies with similar indications for invasive testing that had declined genetic diagnosis. However, such a group would be challenging to assemble; therefore, their results must be interpreted cautiously.

3. Technological Advances in Prenatal Diagnostics

In recent years, we have observed the rapid development of artificial intelligence (AI). Numerous self-learning algorithms have already been implemented in various aspects of our lives, including medicine. Initially, machine learning (ML) and deep learning (DL) models were found to be extremely useful in the field of radiologic image analysis [185,186,187,188]. The overall performance of ML in analysing MRI images was approximately 90% and was significantly higher than that of clinicians. Those findings captured the attention of obstetricians, raising questions regarding the utility of AI in prenatal diagnostics. As a response, in 2022, Zhang et al. [187] introduced an ML model created by Trisomy21Net to assess the risk of trisomy 21 in first-trimester screening. In their study, they obtained 3303 images of a standard midsagittal presentation used to assess nuchal translucency with FTS standards. Scans were shown to a team of experienced, certified sonographers and simultaneously analysed by AI. Surprisingly, the model had better detective performance for foetuses with trisomy 21. This finding suggests that implementing AI into the prenatal screening algorithm would reduce the risk of bias by examinators’ experience and suboptimal visualisation. A year later, Tang et al. [188] developed an AI model called Pgds-ResNet, a fully automated prenatal screening algorithm for various genetic diseases. This tool achieved sensitivity and specificity in detecting trisomy 21, 18, and 13 comparable to those of experienced sonographers. Analysing the heatmap of the AI framework showed that the most informative features of the foetal face in the context of abnormalities were the nose, jaw, and forehead. This contrasts the region of interest of clinicians, who mostly analyse nuchal translucency. On this basis, Ji et al. [189] introduced a model targeted to analyse the features of foetal faces in the context of various abnormalities. Their study once again confirmed the high ability to detect trisomy 21 and 18 by AI. Moreover, they established a high diagnostic value for certain facial markers, such as the inferior facial angle (IFA), maxilla–nasion–mandible (MNM), facial–maxillary angle (FMA), and frontal space distance (FS). Such discoveries only tap into a fraction of the capabilities of artificial intelligence, thus fuelling imagination for the future. The practice of prenatal genetic testing is continually changing, and old methods are constantly being substituted by novel ones.

4. Summary

Taking into account all of the mentioned aspects, it is worth noting that despite its long history and intensive research for alternative methods, amniocentesis remains an important cornerstone of perinatal management. This can be attributed to its wide-spread availability, cost-effectiveness, and high detection rates of foetal aneuploidy, in addition to the extra benefit of diagnosing congenital infections, foetal microdeletion, and microduplication syndromes, which so far have been unmatched by non-invasive methods. No screening method is flawless, and access to invasive prenatal diagnostic procedures guarantees that all pregnant women, irrespective of their socioeconomic or demographic status, can undergo testing for a wide array of foetal abnormalities. Likewise, early detection of serious genetic conditions is essential for parents and medical professionals to plan postnatal care for an affected infant. Ongoing research indicates that numerous potential applications remain undiscovered and must undergo exhaustive and careful investigation before being integrated into clinical practice.