Crossing over: when meiotic recombination impacts linkage-based preimplantation genetic testing for monogenic disorders Authors: Elizabeth J Klein, Mindy S Christianson Published: 2026 Preimplantation genetic testing (PGT) evaluates embryos created by in vitro fertilization (IVF) to reduce the risk that a resulting fetus will be affected by an inheritable genetic condition. Preimplantation genetic testing for monogenic disorders (PGT-M) may be used to test for single-gene inherited conditions when the familial disease-causing variant(s) are known and sample genetic material is available for creation of a test primer (1). Preimplantation genetic testing for monogenic disorders may be performed via direct or indirect genotyping. In direct PGT-M, the embryo is screened for the disease-causing variant itself. In indirect PGT-M, the embryo is screened for nearby polymorphic markers positioned adjacent to the variant of interest. Indirect or linkage-based PGT-M often is favored due to its comparative immunity against allele dropout, the loss of an allele during genetic amplification, which may cause an affected embryo to appear unaffected (1). Nonetheless, linkage-based PGT-M relies on the cosegregation of the genetic marker with the disease-causing variant and thus is vulnerable to meiotic recombination, the exchange of segments between homologous chromosomes that generates new combinations of parental alleles. The likelihood of a recombination event between two loci increases with the genomic distance separating them; consequently, two variants once demonstrated to reside on the same parental chromosome (in cis) may nonetheless be separated during meiosis and segregate independently (2). In this month’s issue of F&S Reports, Iversen et al. (3) present a case of PGT-M screening for two maternal disease-causing variants located in cis on the X chromosome with direct evidence of embryos carrying recombinant maternal haplotypes. The patient is heterozygous for a pathogenic OTC variant (c.788A>G) at Xp11.4 and the recurrent 1.6 Mb STS deletion at Xp22.31. Her prior son inherited both variants in a hemizygous state and died in the neonatal period following a hyperammonemic crisis attributable to ornithine transcarbamylase deficiency, with additional cutaneous findings of X-linked ichthyosis from steroid sulfatase deficiency. Seeking an unaffected pregnancy, the couple pursued IVF with PGT-M for both conditions, using direct genotyping of the OTC variant and linkage-based analysis alone for the STS deletion, which was not amenable to direct detection. Of 13 embryos tested, 9 (69%) screened positive for both variants or neither variant, as expected for two loci in cis, whereas 4 (31%) screened positive for only one of the two, a pattern the investigators attribute to meiotic recombination between OTC and STS. Estimating the intervening distance at approximately 31 centimorgans (cM), the investigators note that the observed frequency (4 of 13, 31%) approximates the predicted recombination probability. The couple transferred a euploid female embryo classified as low-risk for both variants, declined confirmatory prenatal testing, and delivered a healthy female infant. The investigators present this as the first reported human case of meiotic recombination illustrated through PGT-M for two variants demonstrated to be in cis, cautioning that linkage-based screening of distant loci can yield false-negative results. Although the report is clinically compelling and pedagogically valuable, several areas warrant consideration. The first involves the quantitative interpretation of recombination risk. The investigators approximate the 31 Mb interval between OTC and STS as roughly 31 cM, and therefore, an approximately 31% chance of recombination, invoking the familiar convention that 1 cM corresponds to 1 million base pairs (Mb) and to a 1% recombination probability. This convention is only a genome-wide average, and local recombination rates vary widely. The distal short arm of the X chromosome may recombine more than the genomic average, so physical distance in this region understates the true genetic distance and recombination probability (2). Therefore, the close agreement between the predicted 31% and the observed 4 of 13 (31%) should be regarded as approximate rather than exact, particularly given the small number of embryos. A related point concerns the framing of recombination as being especially likely because the conditions are X-linked. Being X-linked does not itself increase the probability of recombination. The elevated crossover risk in this case derives instead from the substantial physical separation of the two loci and from the fact that the relevant meiosis is maternal: recombination occurs more frequently in female than in male meiosis, by roughly 1.6-fold genome-wide (4). Attributing the recombination to X-linkage, rather than to interlocus distance and maternal transmission, risks incorrectly generalizing the finding to closely linked loci, where the co-segregation assumption remains reasonable. A second important limitation is the absence of confirmatory prenatal or postnatal diagnostic testing. The transferred embryo was classified as low-risk for both variants, yet confirmatory prenatal testing was declined, and no postnatal molecular confirmation was reported beyond a reassuring newborn course. Because the central thesis of the report is that linkage-based PGT-M can produce false-negative results, confirmatory testing would have strengthened the argument and excluded an undetected recombination or genotyping error affecting the transferred embryo. The report also has notable strengths. Foremost is the completeness of the family’s genetic evaluation, including chromosomal microarray analysis that defined the breakpoints and gene content of the maternal STS deletion and confirmed the cis configuration of both variants across the proband and her affected son. Crucially, the availability of direct genotyping for the OTC variant allowed the investigators to observe recombination directly rather than infer it, transforming what would otherwise have been a silent source of misclassification into an interpretable finding. If both loci had been assessed by linkage alone, a recombinant affected embryo could have been reported as unaffected without detection. Together, these features render the report a persuasive illustration of an underappreciated but real limitation of linkage-based testing. In summary, the work by Iversen et al. (3) provides a valuable, real-world demonstration that meiotic recombination can disrupt the assumed co-inheritance of variants in cis and yield unanticipated PGT-M results. While the quantitative concordance the investigators report should be interpreted with caution, and neither prenatal nor postnatal confirmatory testing was obtained, the underlying message is clinically important. Several practical implications follow. When technically feasible, direct variant analysis should complement robust linkage analysis using informative markers flanking the locus on both sides to guard against allele dropout and recombination. As the investigators highlight, confirmatory prenatal diagnosis should be offered after transfer, consistent with American Society for Reproductive Medicine recommendations (5). Future research should prospectively track outcomes of PGT-M cycles for two or more linked conditions and evaluate whether a threshold interlocus distance should mandate direct genotyping or confirmatory testing. Furthermore, clinicians should consider the possibility of meiotic recombination when counseling and interpreting PGT-M for multiple conditions on the same chromosome. In complex PGT-M analyses, medical geneticists and genetic counselors should be consulted to ensure families are informed in their reproductive decision making. CRediT Authorship Contribution Statement Elizabeth J. Klein: Writing – original draft, Writing – review & editing. Mindy S. Christianson: Conceptualization, Writing – original draft, Writing – review & editing. Declaration of Interests E.J.K. has nothing to disclose. M.S.C. has nothing to disclose. References