Karyotyping uses chromosome banding to visualize metaphase chromosomes and evaluate morphology (band pattern, length, and arm ratio) for detecting chromosomal abnormalities. Banding methods include G-, Q-, R-, C-, and N-banding, which are selected based on clinical indications. Among these, G-banding is the most widely used in clinical cytogenetics because it provides stable preparations with good banding resolution and can be analyzed with standard bright-field microscopy, enabling direct identification of BCRs (Karger, 2024).

G-banding is cost-effective and can detect apparent rearrangements, but its practical resolution is approximately 5–10 Mb (Levy and Wapner, 2018). Banding resolution is reported as bands per haploid set (e.g., 300–850) (Arsham et al., 2017). Higher-band karyograms can reveal subtler changes, with 550–700 bands capturing most abnormalities and 700–850 bands generally suitable for SVs > 3–5 Mb. High-resolution preparations can yield elongated or overlapping chromosomes, complicating interpretation. Analysis is labor-intensive, and turnaround is typically 7–14 days because cell culture is required (Saldarriaga et al., 2015). Interpretation depends on experienced cytogeneticists, although artificial intelligence-based tools can assist chromosome segmentation and metaphase image quality control to improve efficiency and reporting consistency (Chebly, 2024).

In the late 1980s, FISH marked a shift toward molecular cytogenetics by replacing radioactive probe labels with fluorescent dyes (Shakoori, 2017). FISH relies on complementary base pairing: fluorescently labeled DNA probes hybridize in situ to target sequences on chromosomes, and signals are visualized by fluorescence microscopy to detect SVs. With advances in probe design and labeling, FISH has expanded into multicolor formats, including multicolor FISH and spectral karyotyping. Nowadays, FISH is widely used in clinical practice for the diagnosis and evaluation of chromosomal aberrations.

Compared with karyotype analysis, FISH has a faster turnaround time because it does not require cell culture and offers higher resolution (up to 50 kb) (Cheng et al., 2025). However, due to probe design constraints and cost, clinical FISH assays are targeted rather than genome-wide and typically interrogate only regions suspected of structural variation. Therefore, FISH is often used as a second-tier diagnostic method to confirm BCRs.

Short-read sequencing is widely used clinically, but short read length limits detection of BCR breakpoints. MPS, originally established as pairwise end sequencing in the 1990s, was developed to significantly enhance the effective length of DNA fragments (Edwards et al., 1990; Roach, 1995; Roach et al., 1995). This method involves fragmenting DNA into predetermined sizes (e.g., an insert size of 3–10 kb), circularizing the fragments, and generating reads from both ends (mates) of the DNA fragments (Korbel et al., 2007). In comparison to conventional paired-end sequencing, this technology can reach similar physical coverage with fewer sequence reads due to its larger insert sizes. While the core chemistry is well-established, recent nuances in refined circularization protocols and advanced bioinformatic algorithms have significantly enhanced its sensitivity. These modern refinements facilitate detection of breakpoint-supporting signals, which are essential for the precise detection of BCRs.

The clinical utility of this approach for identifying BCRs has been demonstrated in several studies. In a prospective prenatal study by Qian et al., MPS provides an additional 25% diagnostic yield, primarily by detecting pathogenic balanced translocations and inversions (Qian et al., 2025). This methodology provides concurrent information on both the precise genomic location and the orientation of rearrangements, which is essential for accurate clinical interpretation. The ability of MPS to resolve BCRs is further supported in reproductive medicine: Dong et al. utilized this technique to uncover cryptic BCRs linked to idiopathic male infertility, while Chau et al. successfully delineated all breakpoint junctions in azoospermia cases where karyotyping failed to provide a molecular explanation (Chau et al., 2022; Dong et al., 2023). Consequently, MPS remains a robust and cost-effective tool for comprehensive BCR assessment, though its performance can be constrained by mapping ambiguity in highly repetitive regions (Madjunkova et al., 2020).

LRS generates reads of several kilobases or longer, improving detection of SVs and breakpoint resolution in complex or repetitive regions. LRS is mainly implemented through PacBio single-molecule real-time (SMRT) sequencing and Oxford Nanopore sequencing (Scarano et al., 2024). SMRT detects fluorescent signals during DNA synthesis in zero-mode waveguides (Eid et al., 2009), whereas nanopore sequencing infers bases from current changes as single-stranded DNA translocates through a nanopore (Jain et al., 2018).

In recent years, studies applying LRS in clinical research have increased rapidly. The ability to generate long reads has substantially improved the detection and breakpoint resolution of human structural variants, including BCRs, enabling broader clinical applications. For example, Chow et al. used nanopore sequencing to differentiate embryos carrying BCRs from those with a normal karyotype in 55 blastocysts from nine patients with cytogenetically confirmed reciprocal translocations. They successfully identified euploid embryos for transfer, offering a method that may be particularly useful when no informative proband is available (Chow et al., 2020). Pei et al. applied a similar strategy in preimplantation genetic testing for structural rearrangement (PGT-SR) for two patients with reciprocal translocation (Pei et al., 2021). LRS has also been used to identify other BCR types, such as inversions, in clinical research settings. Xu et al. performed LRS to investigate somatic structural variations in a cohort of 21 patients diagnosed with colorectal cancer. Notably, they discovered large-scale inversions that are often challenging to detect with short-read sequencing, which can impact the expression and structure of crucial tumor suppressor genes, including APC and CFTR (Xu et al., 2023). Together, these studies highlight the utility of LRS for resolving BCRs across diverse clinical research settings.

Despite its advantages (long reads and no PCR amplification), LRS remains more expensive than short-read sequencing, and base-calling errors can limit its clinical use (Athanasopoulou et al., 2021). With the development of circular consensus sequencing in SMRT sequencing, the accuracy of long high-fidelity reads can achieve 99.8% (Wenger et al., 2019). For ONT, R10.4 chemistry with duplex read types can generate reads with 99.9% accuracy (Kolesnikov et al., 2024). LRS has been reported to detect BCRs across a wide size range, from a few kilobases up to several megabases (Madjunkova et al., 2020; Watson et al., 2022; Xu et al., 2023).

OGM is a novel cytogenetic approach for genome-wide detection of SVs, including balanced rearrangements (Lam et al., 2012). Ultra-high–molecular-weight DNA is fluorescently labeled at specific sequence motifs, linearized in nanochannels, and imaged to generate genome maps, which are then compared with a reference to detect SVs (Levy-Sakin et al., 2019). OGM has emerged as a powerful tool for identifying BCRs. Wang et al. collected samples from nine infertile balanced translocation carriers and used OGM for detection (Wang et al., 2020). The results were consistent with karyotype analysis and provided precise breakpoint intervals for the translocated chromosomes. Interestingly, they identified four new genes associated with male infertility, which may inform treatment strategies. Additionally, a recent study analyzed 154 genes from the 1000 Genomes Project using OGM, revealing 55 gene loci with complex SVs, highlighting its utility for detecting such events (Levy-Sakin et al., 2019). According to Bionano Genomics, OGM can detect balanced translocations >50 kb, inversions >30 kb, and insertions >500 bp (Levy et al., 2025).

OGM still has some limitations. First, it does not achieve base-pair resolution because it relies on long-range mapping of labeled sequence motifs rather than direct nucleotide sequencing; therefore, it cannot provide base-level sequence information or precisely resolve breakpoints and often benefits from integration with LRS or other methods for breakpoint refinement (Uppuluri et al., 2022). Second, OGM has reduced sensitivity in highly repetitive or centromeric regions, where low label density and repetitive sequence regions complicate SV detection and interpretation (Barseghyan et al., 2024). Additionally, OGM is unable to detect Robertsonian translocations. Overall, OGM provides a valuable genome-wide approach for BCR detection, but complementary technologies are often required to accurately define breakpoint locations in clinical settings.

Selecting the most appropriate technology for BCR detection by maximizing diagnostic sensitivity and precision in clinical practice requires a multi-factorial evaluation of the suspected rearrangement type, clinical urgency, and methodological characteristics. As summarized in our proposed workflow (Figure 1B), karyotyping remains the foundational first-tier test for global screening for BCRs in patients with infertility, miscarriage, or abnormal offspring due to its cost-effectiveness and genome-wide overview, a status reaffirmed by recent guidelines from ESHRE for the initial evaluation of recurrent pregnancy loss and infertility (RPL et al., 2023).

For karyotype-positive cases such as apparently balanced translocations or inversions, breakpoint confirmation and refinement may be required. In this setting, FISH provides a low-cost, hypothesis-driven approach to verify targeted regions, whereas targeted LRS can offer single-base resolution to evaluate potential gene disruptions at the junctions.

For karyotype-negative cases where a strong clinical suspicion persists (e.g., unexplained recurrent miscarriage or severe infertility), second-tier testing is necessary to identify cryptic BCRs. In this clinical scenario, MPS, OGM, and LRS provide complementary strengths. MPS is generally the most cost-effective and can localize breakpoints at near–single-base resolution, but performance may drop in highly repetitive loci and it requires more complex library preparation (Dong et al., 2023). OGM provides accurate, genome-wide detection of BCRs and complex rearrangements, but lacks single-base breakpoint resolution and is less informative in highly repetitive regions (Yin et al., 2026). LRS enables single-base breakpoint resolution and is often preferred for suspected repetitive-region breakpoints, although it is typically the highest-cost option. Accordingly, for karyotype-negative cases, MPS can serve as a cost-effective screen when repetitive-region involvement is unlikely; OGM is useful when rapid genome-wide structural overview or complex architecture is suspected; and LRS is recommended when precise breakpoint definition is required (e.g., for assessing gene disruption or positional effects or breakpoint-linked haplotyping for PGT-SR) (Xue et al., 2025).

In assisted reproduction technology (ART), identifying BCRs is clinically important because it directly informs the design of PGT-SR and embryo transfer strategy. When the locations of BCRs are available, embryos can be classified as euploid non-carriers, euploid balanced carriers, or unbalanced. A practical transfer strategy is to prioritize euploid non-carrier embryos when available; if none are available, transfer of a euploid balanced-carrier embryo can be considered after counseling, whereas embryos predicted to be unbalanced are generally not recommended. In a retrospective study of 300 couples carrying BCRs, PGT-SR identified about 23.8% of blastocysts as both euploid and structurally balanced, and the cumulative live birth rate reached about 55.8%, supporting its utility in reducing miscarriage risk and improving outcomes (Ogur et al., 2023). Importantly, precise breakpoint identification can further enhance PGT-SR robustness by enabling haplotyping or breakpoint-spanning assays. Using proband-independent LRS in 10 de novo BCR families, breakpoint-phased SNP haplotyping supported embryo stratification, and prenatal diagnosis confirmed concordance in 4 cases with four healthy live births (Xue et al., 2025).

For recurrent miscarriage, identifying BCRs shifts clinical management from unexplained recurrent miscarriage to a defined chromosomal etiology, enabling structured counseling on options such as attempting natural conception with invasive prenatal diagnosis in future pregnancies, IVF with PGT-SR to avoid unbalanced conceptions, or the use of donor gametes (Motan et al., 2025). Gupta et al. showed that detecting parental balanced translocation carriers can directly guide targeted counseling and reproductive planning (Gupta et al., 2024).

Despite major progress, important limitations remain. BCRs located in highly repetitive regions, especially centromeric and pericentromeric loci, are still difficult to detect and resolve. Segmental duplications, satellite repeats, and long insertional regions are prone to fragmented assemblies and are incompletely represented in reference genomes (Espinosa et al., 2024). Short reads are often shorter than duplicated segments, resulting in ambiguous mapping and misassembly (Schluth-Bolard et al., 2025). Conventional cytogenetic methods also face inherent constraints: karyotyping has limited resolution, and FISH signals can be difficult to distinguish because of increased cross-hybridization in repetitive regions. Although LRS offers the potential to span long repetitive intervals, ultra-long read yield, coverage variability, and unresolved centromeric cores still limit consistent performance. Population-based or graph-based reference representations may help alleviate mapping ambiguity, and increasingly mature telomere-to-telomere and haplotype-resolved references are likely to provide a more robust framework for breakpoint discovery (Kolmogorov et al., 2023).

Beyond technical limitations, a major gap is the current shortage of large-scale, multicenter clinical studies and long-term follow-up data for BCR carriers. Current evidence is frequently derived from small cohorts or case series, limiting robust estimation of recurrence risks across different BCR subtypes and hampering comparative evaluation of testing strategies in real-world workflows (Qu et al., 2023; Eisfeldt et al., 2024; Liu et al., 2024). Although emerging efforts are addressing this, the field has historically lacked large-cohort validation for detecting BCRs in certain PGT-SR workflows and analytical frameworks (Zhang et al., 2025). Therefore, prospective research with standardized phenotyping and long-term outcome tracking, spanning reproductive outcomes and offspring development, remain critical to translate BCR characterization into evidence-based counseling and management (RPL et al., 2023).

In addition to these technical challenges, there is ongoing debate about how best to implement these methods in clinical workflows. The role of conventional karyotyping as first-line tools for detecting BCRs is increasingly questioned. Apparently balanced rearrangements are often associated with submicroscopic imbalances or cryptic complexity at the nucleotide level. Advocates of a genome sequencing-first strategy argue that whole-genome sequencing (WGS) using short-read sequencing can sensitively detect cryptic BCRs, refine breakpoints at base-pair resolution, and uncover unexpected complexity that is invisible to G-banding or FISH, while simultaneously detecting copy-number variants. Studies of patients with apparently balanced rearrangements and rare genetic disorders have shown that short-read WGS substantially increases diagnostic yield and could function as a first-tier test (Schluth-Bolard et al., 2019; Yu et al., 2020; Lowther et al., 2023). However, many cytogeneticists contend that karyotyping remains an irreplaceable first-line technique for BCRs because it offers an inexpensive, unbiased, genome-wide overview. Moreover, current short-read WGS pipelines continue to miss a subset of BCRs, particularly when breakpoints fall in highly repetitive or low-mappability regions (Schluth-Bolard et al., 2025), reinforcing the view that, at least for now, short-read WGS should complement rather than fully replace conventional cytogenetics in many diagnostic workflows.

To manage complex cases or uncertain results, an integrated approach may be helpful, with OGM used to provide genome-wide, long-range structural mapping and to delineate complex rearrangement patterns (Yin et al., 2026). Candidate events can be verified by MPS or LRS to refine breakpoints at nucleotide-level resolution, enabling evaluation of potential gene disruption/fusion or positional effects. For discordant findings or variants of uncertain significance, it may be advisable to seek confirmation using an independent complementary method across platforms and to incorporate family segregation analysis to clarify inheritance and interpretability (Biesecker et al., 2024). If a BCR is suspected to involve highly repetitive regions, LRS may be preferentially considered together with telomere-to-telomere (T2T) or pangenome references, because long reads can span extended repeats and directly capture junction sequences to improve detection and breakpoint resolution (Liao et al., 2023; Wang et al., 2025). In parallel, these improved reference resources provide more complete and population-representative representations of repetitive regions, reducing reference-related mapping bias and further improving breakpoint interpretation in these difficult-to-map regions.

Looking ahead, these advances are expected to shift BCR diagnostics toward routine, sequence-resolved assessment, particularly for rearrangements that have historically been challenging to interpret. Continued maturation of T2T assemblies and population-scale pangenome resources should further strengthen breakpoint interpretation by improving read alignment, reducing reference bias, and enabling more consistent characterization across diverse ancestries and genomic backgrounds. At the same time, advances in nanopore duplex sequencing and other high-accuracy, real-time long-read platforms are likely to improve analytical reliability while reducing practical barriers such as turnaround time, throughput constraints, and informatics complexity, making comprehensive, base-pair-level characterization of BCRs increasingly feasible in routine clinical laboratories. Together, these innovations should enable truly genome-wide, sequence-resolved diagnostics for individuals carrying BCRs, facilitating clearer genotype-phenotype correlations and bridging the gap between research applications and clinical implementation.