2p16.3 Deletion (NRXN1) and NRXN1-Related Disorders

Literature and regulatory information updated: 1 November 2026.

Heterozygous 2p16.3 deletion disrupting the NRXN1 gene (a non-recurrent susceptibility copy number variant, with incomplete penetrance and variable expressivity), heterozygous loss-of-function variants in NRXN1 and, as an autosomal recessive allelic entity, Pitt-Hopkins-like syndrome 2 due to biallelic loss of NRXN1

0. Initial Reference Table

1. Introduction

1.1. From synapse genetics to clinical microarrays

Neurexins were identified in the 1990s as neuronal receptors for alpha-latrotoxin, the neurotoxin of black widow spider venom, and later recognised as presynaptic adhesion molecules that bind postsynaptic neuroligins and take part in the formation and maturation of synapses [1, 30]. Humans have three neurexin genes (NRXN1, NRXN2 and NRXN3); NRXN1, at 2p16.3, is the largest, at about 1.1 Mb, and the one most often altered in patients with neurodevelopmental and neuropsychiatric disorders [9, 18].

The clinical association emerged on three almost simultaneous fronts. In 2007, the Autism Genome Project consortium described NRXN1 deletions in families with ASD, in a study that combined linkage analysis and a search for chromosomal rearrangements [31]. In 2008, Kim and colleagues implicated the gene in two patients with ASD and balanced 2p16.3 rearrangements: in a girl with an insertion ins(16;2)(q22.1;p16.1p16.3) inherited from her father, the gene was disrupted in intron 5, and the father, carrier of the same rearrangement, did not have ASD, which constituted the first demonstration of incomplete penetrance [32]. In the same year, Kirov and colleagues found, in one of 93 patients with schizophrenia studied by array CGH, a 0.25 Mb deletion of the promoter and first exon, absent in 372 controls and present in the affected brother and in the mother without a psychiatric diagnosis, whom the children's psychiatrist described as "strange and neurotic", suggesting subclinical traits such as schizotypal personality [16]; Walsh and colleagues found, among 150 patients with schizophrenia and 268 controls, a 115 kb deletion disrupting NRXN1 in a pair of monozygotic twins with childhood-onset schizophrenia [16]; and Zahir and colleagues described a de novo 320 kb deletion including the alpha promoter and exons 1 to 5 in a boy with mild ID, autistic traits and vertebral anomalies [1].

The widespread adoption of chromosomal microarray as a first-line test in ID, ASD and multiple congenital anomalies, recommended by the ISCA consortium (International Standard Cytogenomic Array) in 2010 on the basis of 21,698 patients and a diagnostic yield of 15–20%, made the 2p16.3 deletion one of the most frequent findings in clinical cytogenomics [26]. The series of Ching (2010), Schaaf (2012), Dabell (2013), Béna (2013) and Lowther (2017) delineated the phenotypic spectrum and showed that the same deletion can be found in people with moderate ID, in people with ASD without ID and in parents without any diagnosis [9, 10, 24, 30, 33]. The case-control studies in schizophrenia by Rujescu (2009), Rees (2014) and the Psychiatric Genomics Consortium (Marshall, 2017) placed the exonic NRXN1 deletion among the eight copy number variant (CNV) loci with genome-wide significance for this disease [23, 34, 35].

1.2. The recessive form

In 2009, Zweier and colleagues described patients with biallelic loss of CNTNAP2 or NRXN1 (homozygosity or compound heterozygosity for deletions and truncating variants) and a picture of severe ID with hyperventilation, autistic behaviour and epilepsy, similar to that of Pitt-Hopkins syndrome caused by TCF4; OMIM designated the two entities Pitt-Hopkins-like syndrome 1 (CNTNAP2) and 2 (NRXN1) [1, 7]. Harrison and colleagues confirmed the phenotype in two sisters with different deletions on each allele [8]. The resemblance to Pitt-Hopkins syndrome was later questioned: in the review by Pavone and colleagues, patients with NRXN1-related disorders did not meet the Pitt-Hopkins consensus criteria, and Orphanet created for the recessive form the entity ORPHA:600663, with a descriptive name that does not mention Pitt-Hopkins [4, 19].

1.3. Nomenclature and entities in this article

Three designations coexist in the literature and in reports. "2p16.3 deletion" describes the cytogenomic alteration, which may range from tens of kilobases to about 5.5 Mb and, when large, include other genes [2, 30]. "Intragenic" or "exonic" NRXN1 deletion describes most cases, in which both breakpoints lie within the gene or in the region immediately upstream [9, 13]. "NRXN1-related disorder" is the expression preferred by patient groups and some reviews, and encompasses deletions and sequence variants [3, 19]. In OMIM, the gene has the number 600565; the associated schizophrenia susceptibility (SCZD17) has the number 621407; the recessive form has the number 614325; and entry 614332 corresponds to the 2p16.3 deletion [1, 15]. The SCZD17 entry, created in November 2025, rests on a single patient with a point variant and on three copy number variant studies, and mapping is by variation in the gene, without indication of an inheritance pattern [16].

This article treats as the main entity the heterozygous deletion involving NRXN1, including exonic deletions, intronic deletions (whose significance is different) and those involving the upstream regulatory region; it includes heterozygous loss-of-function variants in NRXN1, which behave like the deletion; and it covers, as an allelic entity with autosomal recessive inheritance, Pitt-Hopkins-like syndrome 2. The 2p16.3 deletion belongs to the class of neurodevelopmental susceptibility CNVs, together with the 1q21.1, 3q29, 15q11.2, 15q13.3, 16p11.2, 16p13.11 and 22q11.2 deletions and duplications [35, 36].

2. Molecular Genetics and Pathophysiology

2.1. The NRXN1 gene and its isoforms

NRXN1 spans about 1.1 Mb at 2p16.3 (GRCh38 chr2:49,918,503–51,032,132) and is transcribed on the minus strand [1, 5]. OMIM describes 24 exons with very large introns: exon 1, of more than 2 kb, encodes the first LNS domain (laminin/neurexin/sex hormone-binding globulin) and the first EGF (epidermal growth factor) repeat; the last exon encodes the transmembrane domain and the cytoplasmic tail [1]. Two CpG island-rich promoters generate two classes of proteins: the alpha promoter, upstream of exon 1, and the beta promoter, downstream of exon 17 [1].

Neurexin-1-alpha has a signal peptide, three repeats of two LNS domains separated by EGF sequences, an O-glycosylation sequence, a transmembrane domain and a cytoplasmic tail of 55 amino acids; neurexin-1-beta corresponds to the C-terminal half of the alpha form, lacking five of the six LNS domains and the EGF sequences, which are replaced by a short specific sequence [1, 30]. Five canonical alternative splicing sites act on the alpha isoform (only sites 4 and 5 on the beta), and the possible combinations have been estimated at 2,208 alpha and 42 beta transcripts; the expression of these isoforms varies between neuronal types and developmental stages [1, 19]. Expression is almost exclusively neuronal, peaking around birth [18].

The MANE Select transcript is NM_001330078.2 (ENST00000401669.7), which encodes the alpha isoform [5]. The literature has used other reference transcripts, with different exon numberings: Lowther and colleagues used NM_004801.4, with 22 exons [9]; Kastratovic and colleagues described the same deletion as involving "the first two" exons in NM_001330078.2 or "the first three" in NM_001135659.3 [37]; Cosemans and colleagues grouped deletions into exons 1–5 and 6–24 [13]; and Montalbano and colleagues defined "alpha" and "beta" blocks from Ensembl exon identifiers [18]. Comparison of a deletion with published data therefore requires the report to state the transcript and the exons removed, or the genomic coordinates with the genome build.

2.2. Synaptic function

Neurexins are adhesion molecules of the presynaptic membrane that bind neuroligins, leucine-rich repeat transmembrane proteins (LRRTM), cerebellins associated with the GRID1 receptor, dystroglycan and neurexophilins, forming trans-synaptic complexes that align the presynaptic active zone with the postsynaptic density [9, 28, 30]. Neurexin–neuroligin binding is calcium-dependent and is modulated by alternative splicing of both partners [1]. In the presynaptic terminal, neurexins bind the scaffolding protein CASK and take part in coupling voltage-gated calcium channels to the vesicle release machinery [18].

Animal models show that alpha-neurexins are required for calcium-triggered neurotransmitter release: mice lacking all three alpha-neurexins die on the first day of life, and those carrying individual deletions have only modestly reduced postnatal viability [1, 30]. Mice lacking Nrxn1 have a selective decrease in the frequency of spontaneous excitatory postsynaptic currents, with no change in inhibitory currents, and impaired prepulse inhibition [1]. In an allelic series of murine models, homozygous loss of Nrxn1-alpha increased aggression in males, reduced affiliative behaviour in females and altered circadian activity in both sexes [1].

In human stem cell-derived neurons with conditional heterozygous NRXN1 mutations, Pak and colleagues demonstrated decreased neurotransmitter release without changes in neuronal differentiation or synapse formation, accompanied by increased CASK; this was the first direct proof that heterozygosity by itself alters human synaptic function [38]. In cortical organoids, isogenic deletions and deletions from patients with schizophrenia altered ubiquitin-proteasome, splicing and synaptic signalling pathways and caused persistent changes in neuronal networks; the patients' deletions also disrupted the trajectories of early neural progenitors, indicating interaction with the genetic background [39].

2.3. Molecular mechanism of the deletions

Unlike recurrent CNVs mediated by non-allelic homologous recombination between low copy repeats (LCRs), such as those of 22q11.2 or 16p11.2, NRXN1 deletions do not share breakpoints and are not flanked by LCRs [18, 23]. Chen and colleagues mapped 32 breakpoints and found microhomology of 2 to 19 base pairs in 68.8%, suggesting DNA replication errors or microhomology-mediated end joining, with enrichment of short inverted repeats, LTR elements and non-B DNA structures nearby [11]. The Psychiatric Genomics Consortium attributed the locus to the non-homologous end-joining mechanism [35], and Montalbano and colleagues noted that NRXN1 is a late-replicating region, more prone to errors induced by replication stress [18]. The practical result is a deletion "hotspot" with individual breakpoints, concentrated at the 5' end of the gene and in the upstream region, in both patients and controls [9, 11].

2.4. Haploinsufficiency and variant spectrum

ClinGen has assigned NRXN1 a haploinsufficiency score of 3 (sufficient evidence) since the assessment of 8 June 2021, and a triplosensitivity score of 0 [6]. The ClinGen intellectual disability and autism panel classified as "definitive" the validity of the relationship between NRXN1 and complex neurodevelopmental disorder with autosomal dominant inheritance [14]. Proof that loss of function of one allele is the mechanism also comes from point variants: Gauthier and colleagues found in a Hungarian patient with disorganised-type schizophrenia, with normal early psychomotor development and no family history of psychiatric illness, a de novo insertion in exon 22, NM_001330078.2:c.4295_4296insACGG p.(Met1433ArgfsTer29) (historical numbering, = NM_004801.4:c.4205_4206insACGG p.(Met1403ArgfsTer29)), affecting the alpha and beta isoforms; the mutant protein was retained in the cytoplasm and had no synaptogenic activity [1, 40]. In COS-7 cells and cultured hippocampal neurons, the protein did not reach the cell surface and did not bind the partners required for synaptic transmission, which is consistent with haploinsufficiency [16]. OMIM describes this variant as "c.4205insACGG, G1402fs", a published designation that does not correspond to the normalised HGVS description [1, 40].

Rare missense variants in NRXN1 have been described in ASD and schizophrenia, but they are not more frequent in patients than in controls at the aggregate level, and their interpretation is individual [1, 32]. Loss-of-function point variants fall within the same counselling framework as exonic deletions.

2.5. Deletion position, isoforms and penetrance

Most deletions observed in patients involve the alpha promoter and the initial exons. In Lowther's series, 72.7% of the 44 exonic deletions involved at least one of exons 1–4 (NM_004801.4), 22.7% involved only exons from the 5th onwards, and the former were larger (median 242 kb versus 83 kb) [9]. Patients with deletions at the 3' end had a second clinically relevant CNV in 70% of cases, compared with 9.4% for 5' deletions (odds ratio, OR, 7.47; 95% confidence interval, 95% CI, 2.36–23.61), which the authors interpreted as lower relative penetrance of 3' deletions [9]. The same authors proposed that part of the effect of 5' deletions results from loss of the long non-coding RNA AK127244, adjacent to the alpha promoter; deletions involving only AK127244, without NRXN1, segregated with early-onset schizophrenia and affective disorders in a Danish family [9, 41].

Cosemans and colleagues pooled 629 published heterozygous deletions (148 in controls, 341 in probands and 140 in carrier relatives), drawn from 73 publications, and 43 deletions from the Leuven clinical cohort, with exon numbering from NM_001135659.1 (alpha isoform) and NM_001330092.1 (beta isoform) [13]. The exon 1–5 region contained 46.6% of probands' deletions (159/341) and almost all exonic deletions in controls; intron 5 concentrated 69.6% of controls' deletions (103/148); the exon 6–24 region contained 19.9% of probands' deletions (68/341) and was involved only twice in more than 100,000 controls [13]. Estimated penetrance for neurodevelopmental disorders was 32.43% (95% CI 11.47–61.33) for exon 6–24 deletions and 12.59% (95% CI 8.34–18.50) for exon 1–5 deletions; the intervals overlap and, when only controls screened for ASD and for ID or developmental delay were used, the two estimates were similar (31.00% and 23.05%, with very wide intervals) [13]. Probands with exon 6–24 deletions had ID or developmental delay more often than those with exon 1–5 deletions (43/68 versus 63/159; p=0.0013), a difference maintained after excluding deletions that also extended to exons 1–5 (33/54, 61.1%; p=0.0072), and had macrocephaly in 10.3% versus 1.2% (7/68 versus 2/159; p=0.004); there was no difference in epilepsy (15/68 versus 21/159; p=0.113) or in psychiatric diagnoses [13]. In exon 6–24 deletions, the predominant phenotype was ID in 64.2% of probands, and the 34.3% with an isolated psychiatric disorder mainly had schizophrenia (16 of 23); ASD without ID was rare [13]. In exon 1–5 deletions, 42.3% of probands had ID or developmental delay, with or without congenital anomalies, and 44.3% a psychiatric disorder without ID, mainly schizophrenia (53.0% of these) or ASD (31.8%); there was no exon-specific phenotype within this region [13]. In the Leuven cohort, three patients with a de novo deletion of all of exons 1–5 had severe ID with autism, ASD with normal IQ and isolated mild ID, respectively, whereas three patients with a deletion of the same exons inherited from the mother had ASD with borderline or normal IQ and delusional ideas or psychotic episodes with onset at puberty [13]. The two series are not directly comparable, because they use different transcripts and exon groupings; both show that position modifies risk and that no position makes it predictable in an individual.

In the Danish population study, the risk signal for ASD and ADHD was concentrated in deletions of the alpha promoter region (OR for ASD 3.75; 95% CI 1.88–7.47), whereas those of the beta region had a weaker association, significant only for ASD (OR 2.45; 95% CI 1.22–4.90) [18]. Schaaf and colleagues had described, in more C-terminal deletions including the beta isoform, increased head circumference and seizures in 88% of patients, but Béna and colleagues did not confirm the association with macrocephaly [10, 24]. Al Shehhi and colleagues found more relatives with mental health and developmental problems in families with deletions of the beta isoform [42]. Castronovo's review refers to a family in which a deletion of exons 2–4 of the beta isoform (exons 19–22 of the alpha), in a woman with high-functioning autism, anxiety, depression and five first-trimester miscarriages, was transmitted to the four children with autism, anxiety and developmental and speech delay, and not to the unaffected child [21]. The same authors suggest that in-frame deletions of exons 6–17 of the alpha isoform may produce a protein lacking one or more extracellular domains, with a dominant-negative effect, a hypothesis without experimental demonstration [21].

Recent research has added a gain-of-function mechanism. In neurons derived from patients' induced pluripotent stem cells, Fernando and colleagues showed that NRXN1 deletions reduce synaptic activity in glutamatergic neurons but increase it in GABAergic neurons, and that aberrant splicing of the isoforms produced by the deleted allele causes these changes [43]. In 2026, Haskell and Hart expressed in Caenorhabditis elegans four human isoforms from 3' deletions of patients with schizophrenia; two produced more severe behavioural phenotypes than total absence of the nrx-1 orthologue, indicating that some mutant isoforms are actively pathogenic [44]. Both studies are preclinical.

2.6. Intronic deletions and the upstream region

Deletions that do not remove exons are common. In the Trillium laboratory (Ontario), they were found in 0.32% of 6,022 patients and in 0.36% of 15,264 controls (OR 0.88; 95% CI 0.52–1.48), 73.7% were located in intron 5, several were recurrent and 63.2% of patients had a second CNV [9]. The authors concluded that intronic deletions do not appear to increase risk substantially, but that the evidence is insufficient to consider them all benign, citing a severe case with an intron 9 deletion near a known enhancer [9]. In Cosemans' review, 79% of intronic deletions in controls were located in intron 5, 96.9% of intronic deletions in probands were inherited (62/64) and 97.1% of relatives carrying intronic deletions had no phenotype (68/70) [13]. Exact breakpoint determination in the Leuven cohort showed six recurrent deletions covering almost all of intron 5, one of which corresponds to the founder deletion described by Rujescu; in half of the probands with an intron 5 deletion (12/25), another variant explained the phenotype or there was no neurodevelopmental disorder, although nine had moderate or severe ID [13]. Intronic deletions were more frequent in unselected controls (0.280%) than in controls without ASD, ID or schizophrenia (0.055%), suggesting some population effect, without translation into individual clinical significance in the cases analysed [13]. The authors recommend classifying intron 5 deletions as benign and rare deletions of other introns as variants of uncertain significance [13].

A small intronic deletion, described by Rujescu and colleagues in 27 individuals from various European populations (including one homozygote), derives from a single founder [23]. In Denmark, this segregating deletion has a frequency of 0.77 per 1000 (95% CI 0.55–1.06); it was associated with schizophrenia spectrum disorder (OR 2.19; 95% CI 1.15–4.16), but the association did not survive correction for multiple comparisons or restriction to unrelated European individuals (OR 1.8; 95% CI 0.8–3.8), and the background haplotype without the deletion conferred no risk [18].

2.7. Multiple-hit model and modifiers

Incomplete penetrance is explained, in part, by additional variants. A second clinically relevant CNV was found in 22.7% of Lowther's patients with an exonic deletion, in 20% of Béna's and in three of five patients with idiopathic generalised epilepsy in Møller's study [9, 24, 45]. In an Italian trio with a deletion inherited from an unaffected mother, the son with ASD had a greater burden of rare deleterious exonic variants in interacting genes (p = 6.07 × 10⁻⁵) [46]. In Kastratovic's Serbian family, the same 317 kb deletion produced ASD with language delay in two brothers and only obsessive traits and speech delay in childhood in the mother, which the authors interpreted as a multiple-hit model [37]. Somatic mosaicism for NRXN1 CNVs has also been described in patients with schizophrenia [47].

2.8. Biallelic loss: Pitt-Hopkins-like syndrome 2

The recessive form results from the absence of functional neurexin-1. In the first case, an 18-year-old girl, Zweier and colleagues identified compound heterozygosity for a 180 kb deletion of exons 1–4 and the variant NM_001330078.2:c.2936C>G p.(Ser979Ter) (historical numbering, = NM_004801.4:c.2936C>G p.(Ser979Ter), in exon 15) [1, 40]. Harrison's sisters had a 79 kb deletion of exons 20–21, inherited from the mother, and another of 287 kb of the alpha promoter and exons 1–5, inherited from the father, both parents being unaffected; the maternal deletion is expected to cause a frameshift in exon 22 and loss of the alpha and beta isoforms [1, 8]. Homozygous deletions were later described, such as that of exons 2–5 in two siblings [12, 48]. In Drosophila, the orthologues of NRXN1 and CNTNAP2 determine the level of the active zone protein bruchpilot, suggesting a synaptic mechanism shared between the two Pitt-Hopkins-like forms [7].

A female infant with early epileptic encephalopathy, fatal respiratory depression and hypoplasia of the arcuate nucleus of the medulla had heterozygous variants in NRXN1 and NRXN2, inherited from different parents; the authors proposed a digenic combination, supported by the respiratory abnormalities of mice lacking neurexin-1-alpha and -2-alpha, but the hypothesis has not been replicated [49].

2.9. Inheritance pattern

The heterozygous deletion is transmitted in an autosomal dominant manner, with incomplete penetrance and variable expressivity [14, 36]. The proportion of de novo deletions in probands was 36% in Lowther's series (9 of 25 with known origin), 4 of 11 in Ching's, and 37.7% for exonic deletions in Cosemans' review (43/114; 50% for exon 6–24 deletions and 32.5% for exon 1–5 deletions, a non-significant difference), in which the five deletions confined to the beta isoform and almost all intronic deletions were inherited [9, 13, 30]. Among inherited deletions, maternal transmission predominated in Lowther's series (11 of 16), without a formal study of transmission bias [9]. In Cosemans' review, maternal and paternal origin were similar overall (66 and 62 of 133 inherited deletions), but exon 6–24 deletions were inherited from the mother in 13 of 17 (p=0.049); among carrier relatives, 54.2% (26/48) of carriers of exon 1–5 deletions and 16 of 22 carriers of exon 6–24 deletions had a phenotype [13]. Viñas-Jornet and colleagues, after neuropsychological assessment of carrier relatives, found anxiety, borderline intelligence and dysexecutive syndrome in carriers considered unaffected, and suggested that part of the "incomplete penetrance" is unrecognised variable expressivity [50]. Pitt-Hopkins-like syndrome 2 is transmitted in an autosomal recessive manner [1].

3. Clinical Spectrum and Manifestations by System

3.1. Principles for reading frequency data

The percentages reported for the heterozygous deletion depend on how carriers were selected. Clinical series recruit patients referred for microarray because of developmental delay, ID, ASD or malformations, and therefore overestimate ID and underestimate mildly affected carriers [9]. Case-control studies in schizophrenia, epilepsy or Tourette syndrome give odds ratios for a specific disorder, without describing the overall phenotype [23, 45, 51]. Population studies, such as iPSYCH2015 and the UK Biobank, give the least biased risk estimates, but they use registry diagnoses and exclude severe ID from the base population or capture it incompletely [18, 52]. Direct comparison between these sources is not valid, and counselling should always state what type of study a figure comes from.

3.2. Phenotype in clinical series

The following table brings together the main series of patients with a heterozygous exonic deletion.

The systematic review by Castronovo and colleagues (PRISMA method; PubMed and Scopus to 30 November 2018; 57 studies) pooled 51 clinical cohort studies, 30 of them with controls: exonic deletions were present in 0.18% of 166,285 patients and in 0.02% of 162,434 controls (OR 10.3; 95% CI 7.03–15.06), with 0.18% in developmental delay or ID cohorts, 0.36% in ASD cohorts, 0.16% in schizophrenia cohorts and 0.17% in mixed cohorts; the authors did not perform a formal meta-analysis, because the same control samples entered several studies [21]. In clinical populations, the most frequent phenotype was moderate to severe ID (77–92%), followed by ASD (43–70%), ADHD (9–41%), anxiety (6–7%) and schizophrenia (5%); epilepsy occurred in 14–53%, hypotonia in 38–47%, receptive or expressive language delay was constant, mild dysmorphism occurred in 45–71% and congenital anomalies were rare, mainly muscular and cardiac [19, 21]. Most deletions in patients involved the promoter and exons 1–4 of the alpha isoform (NM_004801), including or affecting the non-coding RNA AK127244 [21]. Penetrance estimates from family studies ranged between 46% and 78%, with three limitations pointed out by the authors: lack of phenotypic data on transmitting parents and carrier siblings, predominance of children among index cases, which prevents estimation of adult-onset disorders, and small samples [21]. Undiagnosed carriers may share a phenotype of anxiety, dysexecutive traits such as impulsivity and low to borderline intelligence, without aphasia, amnesia, agnosia or apraxia, which led the authors to recommend neuropsychological and psychiatric assessment of carriers before counselling [21]. The meta-analysis by Guo and colleagues estimated that carriers have a 3.20-fold higher risk of disease than non-carriers [25].

3.3. Development, cognition and language

Speech and language delay is the most constant manifestation and often the first [2, 42]. Brignell and colleagues assessed 21 children from 14 families, aged 1.8 to 17 years, with deletions of 74 to 702 kb, mainly of exons 1–3 or 1–5: 69% had a speech sound disorder, rarely severe; 57% had receptive or expressive language difficulty, without a homogeneous profile; 53% had social language difficulties; and almost all those with language difficulties also had ID or global delay [22]. The conclusion was that there is no characteristic communication phenotype, but that the frequency justifies systematic speech and language assessment [22].

Cognitive level varies from moderate to severe ID in clinical series to normal or borderline intelligence in carriers identified among relatives [24, 50]. Alfieri and colleagues documented below-average cognitive level in all five children assessed and an adaptive deficit in four [54]. Fitzgerald and colleagues compared 21 carriers with 21 age- and sex-matched controls and found a lower mean IQ, poorer spatial working memory and differences in resting-state functional connectivity in the visual and ventral attention networks, without structural differences on magnetic resonance imaging [55]. In the UK Biobank, adult carriers of CNVs associated with neurodevelopmental disorders, analysed as a group, had modest cognitive deficits and lower educational and occupational attainment [52].

3.4. Autism spectrum disorder and ADHD

ASD is the most frequent psychiatric diagnosis in clinical series (31.8–65%) [9, 24]. In the Danish population, exonic deletions were associated with ASD with an OR of 3.05 (95% CI 1.87–4.97) and with ADHD with an OR of 2.01 (95% CI 1.23–3.31), values lower than those of the case-control studies used for comparison (ASD 7.24; ADHD 4.68) [18]. The 2p16.3 deletion is among the most frequent CNVs in ASD cohorts, alongside 16p11.2, 22q11.2, 1q21.1 and 15q13.3 [56].

3.5. Schizophrenia and other psychiatric disorders

Exonic NRXN1 deletions increase the risk of schizophrenia. Rujescu and colleagues found them in 0.24% of 2,977 European patients and in 0.015% of 33,746 controls (OR 8.97; 95% CI 1.8–51.9) [23]. In that study, the total of NRXN1 copy number variants, exonic or not, amounted to 66 deletions and 5 duplications, of 18 to 420 kb, scattered across the gene and without a common breakpoint: 12 deletions and 2 duplications in patients (0.47%) and 49 deletions and 3 duplications in controls (0.15%); the association was significant only for those involving exons (p = 0.0027), and there were no copy number variants in NRXN2 or NRXN3 [16]. Rees and colleagues found 11 exonic deletions in patients (0.16%) and none in controls, with a combined p with the literature of 1.3 × 10⁻¹¹ [34]. In the Psychiatric Genomics Consortium, with 21,094 cases and 20,227 controls, loss at 2p16.3 reached genome-wide significance, with 35 cases and 3 controls (OR 14.4; 95% CI 4.2–46.9) [35]. In the Danish population, by contrast, exonic deletions were not significantly associated with schizophrenia spectrum disorder (OR 1.41; 95% CI 0.69–2.90) or with schizophrenia in the strict sense (OR 1.88; 95% CI 0.81–4.33), which the authors attributed partly to overestimation of risk in selected case-control samples and to the small number of cases [18].

Todarello and colleagues, in North American families with schizophrenia, found deletions in 1.10% of patients, 1.02% of unaffected parents and 0.97% of unaffected siblings, including a de novo deletion in a healthy sibling, and concluded that the deletion may be necessary but is not sufficient [57]. Associations have been described with depression (OR 2.01 in case-control studies; 1.46, not significant, in Denmark) and, in case series, with bipolar disorder and anxiety [9, 18, 50].

3.6. Epilepsy and movement disorders

Møller and colleagues found exonic deletions in 5 of 1,569 patients with idiopathic generalised epilepsy and in 2 of 6,201 controls (p=0.0049), with incomplete segregation in families and a second large CNV in three of the five probands [45]. In Guo's meta-analysis, 81.3% of patients with epilepsy had neurodevelopmental or psychiatric comorbidities, deletions of exons 1–6 were more frequent in epilepsy, and 71.42% of these patients had genetic generalised epilepsy [25]. In Denmark, the association with epilepsy was OR 1.94 (95% CI 1.01–3.73) [18]. Kamgang and Myers described in 2026 the evolution from reflex myoclonic epilepsy of infancy to childhood absence epilepsy in an NRXN1-related disorder [58]. NRXN1 is included in the gene lists for epilepsy with developmental delay [59].

Tourette syndrome was associated with NRXN1 deletions at genome-wide significance (OR 20.3; 95% CI 2.6–156.2) in a European sample of 2,434 cases and 4,093 controls [51]; the association was not replicated in Denmark (OR 1.53; 95% CI 0.65–3.56) [18]. Infantile hypotonia is common in clinical series (38–59%) [10, 24].

3.7. Dysmorphism and malformations

There is no characteristic facial gestalt. Dysmorphic features are mild and non-specific: long face, deep-set eyes, prominent premaxilla, wide mouth [19, 37, 50]. Malformations are infrequent: 9.1% in Lowther's series (two tetralogies of Fallot, one with tracheo-oesophageal fistula and imperforate anus and the other with RAF1-related Noonan syndrome; one cryptorchidism; one small atrial septal defect), ventricular septal defects in two of Ching's patients, and isolated reports of vertebral anomalies, scoliosis, hip dysplasia, omphalocele, pulmonary hypoplasia and congenital diaphragmatic hernia [2, 9, 30]. A causal association of these malformations with the deletion has not been established, and the presence of a major malformation should prompt a search for another cause [9].

3.8. Adult carriers and transmitting parents

Many carriers are identified after a child's diagnosis. In Lowther's series, four of 11 transmitting mothers had a diagnosis (three with mild ID, one with anxiety and depression) [9]. Ching described a "socially awkward" father and a mother with language delay and social difficulties [30]. In Schaaf's series, 8 of the 9 transmitting parents had a history of learning difficulties or neuropsychiatric illness; in Dabell's, eight parents were carriers, some with neurocognitive deficits and others without abnormalities [15]. Unique refers to carrier parents without difficulties, including people with higher education [2]. In population studies, most carriers have no recorded psychiatric diagnosis [18].

3.9. Pitt-Hopkins-like syndrome 2

The biallelic form begins in the first months of life, with severe psychomotor delay, severe ID and absent or almost absent speech [2, 4]. Pavone and colleagues brought together four patients with this diagnosis: severe ID in all, breathing abnormalities and stereotypies in all, constipation in three, ASD, seizures and sleep-wake cycle disturbance in two, normal head circumference and mild dysmorphism (wide mouth, strabismus, protruding tongue with drooling) [1, 19]. Zweier's patient, at 18 years, had normal growth, walking acquired at 2 years, developmental regression after the first years, hyperventilation, autistic behaviour, hyporeflexia of the upper limbs, constipation and mild dysmorphism, without seizures and with normal magnetic resonance imaging [17]. Harrison's sisters had severe epilepsy from infancy, hypotonia, walking acquired at five years by one of them, hyperventilation or breath-holding apnoeas, gastro-oesophageal reflux with poor growth, constipation, precocious puberty, pulmonary stenosis and scoliosis, with normal magnetic resonance imaging [1, 8]. In Guo's meta-analysis, 53.84% of biallelic patients had epilepsy, always with generalised seizures [25]. In the siblings with a homozygous deletion of exons 2–5, one had infantile spasms (West syndrome) and the other ASD, which shows variability even in the recessive form [12].

Castronovo's review brought together 11 patients with biallelic deficiency (10 published and one new), aged 4 to 33 years at last assessment [21]. The core features, present in all or almost all, were severe developmental delay and ID (moderate delay in one case), marked hypotonia, absent expressive language (at most 20–25 words), severe impairment of social interaction, motor stereotypies, sleep-wake cycle disturbance and chronic constipation [21]. In a sizeable minority there were growth delay, strabismus, epileptic seizures (6 of 11, including West syndrome and epileptic encephalopathy), hyperventilation or breath-holding apnoeas, hearing impairment, microcephaly and abnormalities on electroencephalography or magnetic resonance imaging; independent walking was acquired between 2 and 5.5 years, and two patients did not walk [21]. With one exception, the alterations were inherited from apparently healthy parents, although parental neuropsychological assessment was insufficient in most studies; two patients had a deletion and a point variant (nonsense or splicing), and two cases were considered doubtful: a homozygous intronic deletion with parental consanguinity, and a missense variant of uncertain significance in compound heterozygosity with a silent variant [21, 48]. As with heterozygous deletions, most deletions involve the promoter and initial exons of the alpha isoform and spare the beta; the exception, in Harrison's sisters, led to the proposal that loss of the beta isoform is associated with precocious puberty in girls [8, 21]. In these sisters, a paternal 742 kb duplication at 5q35.1 was considered coincidental [17]. Three patients had a deletion including the microRNA miR-8485, located in intron 5, without apparent phenotypic difference [21].

Castronovo's new case was a 6-year-old girl, born to non-consanguineous parents, with a paternal 467 kb deletion (exons 3–14) and a maternal 269 kb deletion (exons 3–4), overlapping in the exon 3–4 region (GRCh37 coordinates) [21]. She had hypotonia and weakness from birth, head control at 8 months and trunk control at 18 months, independent walking at 5.5 years, babbling at 10 months without progression to language, absence of non-verbal communication, hand stereotypies of visual self-stimulation, frequent night waking, insistence on routines, chronic constipation and coeliac disease; she never had seizures, and hearing, electroencephalography, magnetic resonance imaging, fragile X testing and karyotype were normal [21]. Dysmorphic features included a depressed nasal bridge, telecanthus, frontal bossing, low-set ears, a long philtrum, a thin upper lip, brachycephaly and hammer toes; the parents had no neuropsychiatric illness, but there was schizophrenia, depression and mild ID in the maternal family [21].

4. Clinical Recognition and Warning Signs

4.1. When to think of the 2p16.3 deletion

The deletion has no clinical signs that allow it to be specifically suspected; it is almost always identified by microarray or CNV analysis requested for another reason [2, 19]. The situations in which it is found are the general indications for these tests: global developmental delay, isolated speech and language delay, ID, ASD, epilepsy with neurodevelopmental comorbidity and, in some centres, schizophrenia, especially with ID or early onset [25, 26, 60]. The American Academy of Pediatrics recommends standardised ASD screening at 18 and 24 months in primary care and ongoing developmental surveillance, which leads to early identification of the children in whom the genetic aetiology should be investigated [61].

4.2. Special contexts

There are three contexts without an indicating phenotype. In prenatal diagnosis, the deletion may be detected on a microarray requested because of an ultrasound anomaly, which it rarely explains [36, 62]. In family studies, it is found in parents and siblings of a proband [42]. And it may arise as a finding of exome or genome sequencing with CNV analysis [2, 12].

4.3. Warning signs in a known carrier

In a carrier identified before symptoms, as a newborn or young child, Castronovo's review recommends periodic neurobehavioural observations for early detection of warning signs and prescription of early behavioural intervention [21]. The warning signs are those of ASD and language delay (absence of babbling, gestures or words at the expected ages, loss of skills, reduced eye contact and social interaction), episodes suggestive of epileptic seizures, including absences, hypotonia and, in adolescence and adulthood, prodromal symptoms of psychosis or mood changes [21, 45, 61]. Language regression between 18 months and 3 years, as in the case described in section 10, should prompt immediate assessment [37].

4.4. When to suspect biallelic loss

A child with a heterozygous NRXN1 deletion and severe early-onset psychomotor delay, especially with hyperventilation or other breathing abnormalities, early epilepsy, stereotypies, constipation or sleep disturbance, should be investigated for a second variant on the other allele [21, 28]. The same applies to a child with this picture and parental consanguinity or an affected sibling [12]. Emergency situations are those of severe infantile epilepsy and breathing abnormalities, described in the biallelic form and in the case of probable digenic cause with fatal respiratory depression [8, 49].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. There are no clinical criteria

The diagnosis is exclusively molecular. There are no clinical criteria for the heterozygous deletion or for the biallelic form, and the overlap with Pitt-Hopkins syndrome does not allow the consensus criteria of the latter to be used [19]. One conceptual point runs through the whole process: the heterozygous deletion is a susceptibility factor and not, by itself, a diagnosis that explains a patient's entire phenotype, and the report and counselling should reflect this distinction [21, 28].

5.2. First-line tests

Chromosomal microarray is the first-line test for ID, global developmental delay, ASD and multiple congenital anomalies, with a yield of 15–20%, compared with about 3% for conventional karyotyping [26]. In the Oxford Desk Reference, the additional yield of microarray after karyotyping is about 14% in clinically non-specific ID [63]. The recommendations of the American College of Medical Genetics and Genomics (ACMG) for ASD and of the American Academy of Pediatrics include genetic aetiological evaluation of all children with ASD [61, 64]. Since 2021, the ACMG has strongly recommended exome or genome sequencing as a first- or second-tier test in paediatric patients with congenital anomalies or developmental delay/ID [27]; detection of intragenic deletions from exome data depends on coverage and the CNV calling algorithm, and genome sequencing detects them at higher resolution [2, 12].

In schizophrenia, Baker and colleagues advocated microarray as a routine clinical test [60]. In practice, the psychiatric aspects of CNVs are addressed less in counselling than the non-psychiatric manifestations, which Morris and colleagues describe as a care gap [65].

5.3. Resolution and limitations of testing

The ability to detect an intragenic deletion depends on probe density in the gene. In Lowther's study, the OGT and Illumina 4×180K microarrays had about 200–250 probes over NRXN1 (mean spacing 5.5–7.6 kb) and the CytoScan HD about 650, with minimum detection of about 10 kb; the Agilent 244K used by Ching had 130 probes, with detection of deletions from 43 kb [9, 30]. Smaller deletions, point variants and small insertions or deletions are not detected by microarray [21, 26]. Nor are balanced rearrangements that disrupt the gene, such as the insertion described by Kim, because they do not alter copy number [26, 32].

A normal result does not exclude a sequence variant in NRXN1, and a heterozygous deletion does not exclude a second variant on the other allele [21]. About 2% of apparently de novo interstitial CNVs result from a balanced insertion in a parent, which is detected only by targeted fluorescence in situ hybridisation (FISH) or breakpoint PCR, because dosage methods give a normal result in the balanced carrier [63]. Low-level mosaicism is not reliably detected by microarrays [26].

Deletions should be confirmed by a second method when the laboratory requires it: FISH for deletions of at least 100 kb, quantitative PCR or a second microarray for smaller ones [9].

5.4. Classification and reporting

The 2020 joint ACMG/ClinGen standard classifies CNVs into five classes (pathogenic, likely pathogenic, uncertain significance, likely benign and benign) by evidence score; a pathogenic CNV may have reduced or variable penetrance and expressivity, provided the association is documented [66]. For a loss involving NRXN1, a gene with established haploinsufficiency (ClinGen score 3), the categories in section 2 of the standard apply [6, 66]:

Section 5 of the standard slightly penalises inheritance from an apparently unaffected parent (−0.15 in a patient with a non-specific phenotype; −0.30 in a patient with a specific, well-defined phenotype) and scores de novo occurrence and segregation with a consistent phenotype positively [66]. The classification of a CNV is the same for all patients with the same alteration, but the report should explain the clinical significance for the patient concerned [66]. In practice, exonic deletions are classified as pathogenic with reduced penetrance and variable expressivity, as in Kastratovic's case (category 2C-1, with frequency in cases higher than in controls) [9, 37].

The report should state the ISCN nomenclature with the genome build and coordinates, for example arr[GRCh37] 2p16.3(51193626_51510961)x1 mat, the genes and, when the variant is detected by sequencing, the HGVS description with the gene, transcript and exons involved [37, 66]. For NRXN1, it is recommended to cite the MANE transcript (NM_001330078.2) and, when useful for comparison with the literature, the correspondence with NM_004801.4 [5, 9]. Cosemans' interactive table allows a specific deletion to be compared with published deletions in patients and controls [13].

5.5. Complementary assessment after diagnosis

The post-microarray algorithm of the Oxford Desk Reference, adapted from the ISCA consensus, includes consulting databases such as DECIPHER, studying the parents to distinguish de novo from inherited deletion, and attention to incomplete penetrance, unmasking of a recessive allele, imprinting and complex inheritance [36]. Most small CNVs in the genome are normal variation, and knowledge of the consequences of dosage is incomplete, which makes interpretation dependent on databases and the literature [67]. Three steps are specific to NRXN1:

1. Look for other genetic risk factors: a second CNV is present in about a fifth of patients with an exonic deletion, and Castronovo's review recommends always looking for them, because of their possible effect on prognosis; exome or genome sequencing is the usual route [9, 21, 24, 37].
2. Sequence the spared allele when the phenotype is severe and early, of Pitt-Hopkins-like type, to exclude biallelic loss, which microarray alone may not detect [21]. For exon 1–5 deletions, Cosemans recommends exome or genome sequencing when the proband has moderate to severe ID, congenital anomalies or isolated epilepsy, phenotypes that this deletion rarely explains on its own [13].
3. Study the parents, with clinical assessment of the carrier parent, which may reveal mild traits [36, 50].

Baseline clinical assessment, which does not depend on genotype, includes formal assessment of development and language, diagnostic assessment for ASD with standardised instruments, hearing, and electroencephalography when there are suspicious episodes [22, 30, 61]. Neuroimaging is not routine in ASD and is reserved for neurological signs [56].

5.6. Prenatal diagnosis

The American College of Obstetricians and Gynecologists (ACOG) recommends prenatal microarray when there are one or more major structural anomalies and invasive diagnosis [68]. An NRXN1 deletion found in this context rarely explains the ultrasound anomaly [9, 36]. Castronovo's review concludes that incomplete penetrance and pleiotropy do not allow reliable predictions after prenatal detection of an exonic deletion [21]. Some laboratories do not report prenatally CNVs with substantial incomplete penetrance, because the result is uncertain and unlikely to be the cause of the ultrasound findings [36]. In 2026, Chen described the prenatal diagnosis of a familial 2p16.3 deletion involving NRXN1 in which the carriers in the family had no apparent phenotype [62]. The implications for reproductive decision-making are in section 8.

5.7. Practical algorithm

6. Differential Diagnosis

6.1. Framework

The differential diagnosis of the heterozygous deletion arises on two levels. Clinically, it is the differential of developmental delay, ID and ASD, without specific clues. Genetically, once the deletion has been found, the question is whether it explains the picture or whether there is another cause, given that a second finding is common [9, 21]. For the biallelic form, the main differentials are Pitt-Hopkins and Pitt-Hopkins-like 1 syndromes, Rett syndrome and other developmental encephalopathies with breathing abnormalities [7, 19].

6.2. Differential table

6.3. Main differentials

TCF4-related Pitt-Hopkins syndrome has a syndromic facies and a set of international consensus criteria in which an abnormal breathing pattern is a cardinal element; in Pavone's comparison, patients with an NRXN1-related disorder did not meet the Pitt-Hopkins consensus criteria [19]. CNTNAP2-related Pitt-Hopkins-like syndrome 1 has an overlapping picture, and the two genes belong to the neurexin superfamily; in 179 patients with severe ID, recessive CNTNAP2 defects had a frequency of at least 1% [7].

Among susceptibility CNVs, the 2p16.3 deletion is distinguished by the absence of characteristic malformations, by its variable size and position and by the need to interpret the position within the gene [9, 36]. In schizophrenia, the 22q11.2 deletion has the highest odds ratio among the Psychiatric Genomics Consortium loci (OR 67.7, compared with 14.4 for NRXN1), and recurrent LCR-mediated CNVs have fixed breakpoints, unlike NRXN1 deletions [18, 35].

In large 2p16.3 deletions, which may reach about 5.5 Mb, the full gene content of the deletion should be assessed, because other genes in the band may contribute to the phenotype [2, 30].

7. Treatments and Risk Management Strategies

7.1. Principles

There is no specific treatment for the 2p16.3 deletion or for Pitt-Hopkins-like syndrome 2, and there is no clinical guideline dedicated to these conditions; ClinGen has issued no actionability assertion for NRXN1, in the adult or paediatric context, because there is no specific medical intervention that early identification makes possible [28]. Management is that of the manifestations present, according to the general guidelines for ASD, ADHD, epilepsy and psychiatric disorders, with no demonstrated differences for deletion carriers [19, 28]. The clinical value of the genetic diagnosis lies in anticipatory surveillance, early intervention, aetiological explanation and reproductive counselling [21, 65].

7.2. Developmental, language and behavioural intervention

For carriers identified before symptoms, including those diagnosed prenatally or neonatally, Castronovo's review recommends periodic neurobehavioural observations to detect the first warning signs and start early behavioural intervention [21]. Given the frequency of speech (69%) and language (57%) disorders, Brignell and colleagues recommend that speech and language be assessed and that interventions be applied in a targeted and timely manner during childhood [22]. The American Academy of Pediatrics recommends evidence-based behavioural interventions for specific ASD skills and symptoms and shared decision-making with families in choosing interventions [61]. Educational and behavioural interventions are the basis of ASD treatment [56].

In the cases described, the children received speech therapy and special education from the age of two, with progress but with development below that expected for age [37]. Unique emphasises that the deletion does not regress and has no cure, but that the diagnosis allows appropriate monitoring and interventions to be organised [2].

7.3. Treatment of comorbidities

ASD is associated with sleep and feeding disorders, gastrointestinal symptoms, obesity, seizures, ADHD, anxiety and elopement behaviour, which should be sought and treated because they affect functioning and quality of life [61]. Risperidone has some benefit for tantrums, aggression and self-injurious behaviour in children with ASD, and depressive and psychotic disorders may require psychopharmacological treatment [56]. Epilepsy is treated with conventional antiepileptic drugs chosen according to seizure type and epilepsy syndrome; in the heterozygous deletion, genetic generalised epilepsies predominate [25, 28]. There are no data showing a different response to psychotropic or antiepileptic drugs in carriers of NRXN1 deletions; doses follow the labels and guidelines for each indication and are not specific to this condition.

In associated schizophrenia, treatment follows general standards; two adults in Lowther's series, with schizophrenia and ID, had tardive dyskinesia secondary to antipsychotics, an adverse effect to monitor in patients with ID who receive antipsychotics for many years [9].

In Pitt-Hopkins-like syndrome 2, the priorities are control of epilepsy, nutritional support with treatment of gastro-oesophageal reflux and constipation, assessment and treatment of breathing and sleep abnormalities, and orthopaedic surveillance of scoliosis [1, 8, 19].

7.4. Psychiatric genetic counselling

Morris and colleagues describe two counselling scenarios for CNVs associated with psychiatric illness: the person without a psychiatric diagnosis, in whom the goal is to explain the risk and early signs without fatalism, and the person with an established diagnosis, in whom the goal is aetiological explanation and reduction of stigma and guilt [65]. Families want this counselling, and it increases empowerment and self-efficacy [65].

7.5. Regulatory status

7.6. Research and developments in the last twelve months

There are no therapeutic clinical trials, so there are no efficacy results to tabulate; the available data are observational or preclinical, and none has changed clinical practice. Studies with human cells suggest that a targeted treatment, should one ever exist, will have to distinguish deletions acting through loss of function from those producing gain-of-function isoforms, by restoring normal isoforms or eliminating mutant ones [43, 44]. This conclusion is preclinical and there is no trial whatsoever in humans.

7.7. Pregnancy, perioperative care and transition

The sources consulted describe no specific pregnancy complications in carrier women nor anaesthetic risks particular to the deletion; Unique found no association with complications of pregnancy or delivery when the fetus is a carrier [2]. Perioperative care is the general care for people with ASD or ID (preparation, adapted communication) and, in the biallelic form, that imposed by epilepsy, breathing abnormalities and scoliosis [8, 61]. Transition to adulthood requires continuity of psychiatric surveillance, because the risk of schizophrenia and mood disorders manifests mainly from adolescence onwards [20, 23].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

8.2. Penetrance for counselling

The most widely used estimate is that of Kirov and colleagues, who combined frequencies in controls (0.02%), in schizophrenia (0.18%) and in developmental delay, ASD or congenital malformations (0.18%), assuming prevalences of 1% and 4% for these disorders: penetrance of 6.4% (95% CI 2.5–8.3) for schizophrenia, 26% (95% CI 16–80) for developmental delay, ASD or malformations, and 33% (95% CI 18–88) for any of these disorders [20]. Cosemans' estimate, from 174,223 patients and 100,127 controls in the literature, was 14.34% (95% CI 9.94–20.20) for exonic deletions overall, 32.43% (95% CI 11.47–61.33) for exon 6–24 deletions, 12.59% (95% CI 8.34–18.50) for exon 1–5 deletions, 1.10% (95% CI 0.88–1.38) for intronic deletions and 2.78% (95% CI 2.36–3.28) for all deletions; the intervals for the two exonic regions overlap [13]. Estimates from family studies, from 46% to 78%, are higher and limited by the lack of phenotype in carrier parents and siblings and by the predominance of children [21]. These figures are population estimates, with wide confidence intervals, and do not include mild phenotypes such as learning difficulties, anxiety or executive deficit described in carriers assessed in detail [50]. Danish population studies indicate lower relative risks than case-control studies [18].

Counselling should convey that most carriers in the population do not have a serious diagnosis, that the risk of neurodevelopmental disorder is increased, that the phenotype is not predictable and that it may differ between members of the same family, as observed in Kastratovic's family [2, 33, 37]. An undiagnosed carrier parent does not guarantee a carrier child without manifestations, and the converse is also true [32, 57]. For counselling by position, Cosemans proposes: for exon 6–24 deletions, bear in mind the typical association with ID and schizophrenia, even when the proband has a mild phenotype; for exon 1–5 deletions, the most consistent associations are borderline IQ or mild ID, schizophrenia, ASD with or without ID and severe psychiatric disorders with onset or worsening at puberty, and the contribution of the deletion in an individual is difficult to estimate; for intron 5 deletions, classification as benign [13].

8.3. Family study and cascade testing

Study of the parents is necessary to distinguish de novo from inherited deletion and to calculate the recurrence risk [36]. There is no consensus on cascade testing extended to asymptomatic relatives, because of predictive uncertainty [36]. In the Irish series, cascade screening identified 33 additional carriers in 34 families, 16 of whom had an established phenotype [42]. Testing asymptomatic younger siblings allows the anticipatory neurobehavioural surveillance recommended by Castronovo, but has limited predictive value and should be decided with the family; when the sibling has symptoms, the test is diagnostic [21, 65].

8.4. Prenatal and preimplantation diagnosis

Prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic diagnosis are technically possible when the familial deletion is known [2, 36]. The Oxford Desk Reference advises caution with CNVs of variable penetrance and considers that, when the deletion has no apparent phenotype in the parent, there may be no indication for prenatal diagnosis [36]. Castronovo's review concludes that, after prenatal detection, clinical outcome is not reliably predictable [21]. Reproductive decisions should be made in non-directive counselling, with sufficient time, especially when the parent has learning difficulties [36].

In Pitt-Hopkins-like syndrome 2, the 25% risk and the severity of the condition make prenatal and preimplantation diagnosis a clear option for couples in whom both variants have been identified [1, 8].

9. Special Considerations: Population Diversity and Epidemiology

9.1. What each figure measures

The difference between Denmark (0.70 per 1000) and the UK Biobank (0.39 per 1000) was statistically significant; the Danish prevalence was also double that observed in the controls of several case-control studies, and the authors concluded that the higher true frequency in the general population largely explains the lower odds ratios of the population study. The study's limitations are its size (rare deletions), the young age of participants and the use of unconfirmed hospital registry diagnoses [18]. In practice, the exonic NRXN1 deletion is present in about 1 in 1400 to 1 in 3000 people in the general European population, a number that far exceeds that of people with a diagnosis attributable to the deletion [13, 18, 28]. In a North American health system (DiscovEHR cohort, 90,595 participants), pathogenic CNVs and loss-of-function variants in genes for neurodevelopmental neuropsychiatric disorders were together present in more than 1% of people, confirming that this type of variant is common in the adult population [72]. Unique considers it as frequent as better-known syndromes such as Prader-Willi syndrome [2].

9.2. Diversity and founder populations

The segregating intronic deletion described in several European populations derives from a single founder and has a frequency of 0.77 per 1000 in Denmark [18, 23]. Apart from this case, the deletions are non-recurrent and there is no founder effect [23]. The available data are almost all from populations of European ancestry and from North America; there is a small series from Oman, with consanguinity in three of five families [53]. The sources consulted do not include data from Portugal or Brazil, which does not mean that none exist.

9.3. Resources, networks and associations

The European Reference Network for rare congenital malformations and rare intellectual disability (ERN ITHACA) is the relevant network in the European Union [29]. The British association Unique publishes a family guide on 2p16.3 (NRXN1) deletions, updated in 2024 and reviewed by Christian Schaaf, with an easy-read version and translations into French, Dutch and Spanish; in 2024 it had more than 150 member families with a 2p16.3 microdeletion and three with an NRXN1 sequence variant [2]. The NRXN1 Network (nrxn1network.org) is a patient association created by the mother of a child with a deletion, and Simons Searchlight maintains an observational registry of people with genetic alterations and autism traits that includes NRXN1 (NCT01238250) [2, 3, 70].

10. Illustrative Clinical Case

Kastratovic and colleagues described in 2026, in Kragujevac (Serbia), two brothers and their mother with the same NRXN1 deletion [37].

The first son was born by vaginal delivery at 41 weeks, after an uneventful pregnancy. He sat at 6 months, walked at 13 and said his first words at 15–16 months; between 18 months and 3 years there was regression of speech and behaviour. He started speech therapy and special education at 2 years, with progress, but milestones remained below those expected. At 6 years he was referred to the genetics clinic because of uneven development: limited communication, speech intelligible only to his parents, attention sustained mainly through the mobile phone, mild aggression, stereotypies and obsessive behaviours. Weight was 23.8 kg (89th centile), height 120 cm (91st centile) and head circumference 53 cm (83rd centile). Fine motor skills were preserved; eye contact was brief, with anxiety, irritability and traits consistent with ASD. Hearing and electroencephalography were normal, with no seizures. He had an elongated face, prominent ears and chin, bilateral sandal gap between the first and second toes and flat feet [37].

The brother, aged 5 years, had normal motor milestones (sat at 6 months, walked at 13) and first words at 14 months. He had one weekly session of special education and four of speech therapy from 2 years and 8 months. Weight was 16.1 kg (42nd centile), height 100 cm (22nd centile) and head circumference 51 cm (46th centile). He had reduced eye contact, occasional stereotypies, irritability, limited attention and speech in simple sentences, poorly intelligible to strangers; electroencephalography was normal. He had a long face, depressed nasal bridge, wide mouth and bilateral sandal gap between the first toes [37]. The mother, aged 28 years, did not speak until 3 years of age, had no ID or neurological or psychiatric illness, and reported obsessive personality traits [37].

Facial analysis with Face2Gene from frontal photographs did not suggest the diagnosis and proposed fragile X, Angelman and Noonan syndromes [37]. Microarray (Agilent SurePrint G3 4×180K + SNP, GRCh37) identified in both brothers a heterozygous deletion of about 317 kb, arr[GRCh37] 2p16.3(51193626_51510961)x1, which included the 5' upstream region and the first exons specific to the alpha isoform (the first two in NM_001330078.2, the first three in NM_001135659.3); parental testing showed that it had been inherited from the mother. It was classified as pathogenic according to the ACMG/ClinGen standard, with the caveat of reduced penetrance and variable expressivity [37]. The pedigree and the position of the deletion are in figure 2 (https://pmc.ncbi.nlm.nih.gov/articles/PMC13409882/figure/ijms-27-06241-f002/) and the microarray profiles of all three in figure 4 (https://pmc.ncbi.nlm.nih.gov/articles/PMC13409882/figure/ijms-27-06241-f004/) of the article, published under a CC BY 4.0 licence [37].

The authors recommended regular follow-up by child psychiatry, continuation and intensification of speech therapy and special education, and exome sequencing to look for additional variants [37]. This management coincides with the recommendations of Castronovo's review: neurobehavioural surveillance, early intervention and a search for other genetic factors [21]. The case illustrates three points: the deletion was transmitted by a mother with a mild phenotype; the two brothers with the same deletion have different severity; and the facies does not guide the diagnosis.

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The prognosis of the heterozygous deletion ranges from absence of clinical manifestations to moderate to severe ID with ASD and epilepsy, and it is not predictable from the deletion [21, 24]. Estimated overall penetrance is about 33% for developmental disorders or schizophrenia, with 6.4% for schizophrenia [20]. Most carriers in the population have no serious diagnosis, and undiagnosed adult carriers may have modest cognitive deficits [18, 52, 55]. The presence of a second CNV or other rare variants may worsen prognosis and should be sought [21, 46]. Life expectancy does not appear to be reduced [2]. In ASD with severe ID, up to 20% of people develop seizures by adulthood, often with onset in adolescence [56].

In Pitt-Hopkins-like syndrome 2, the prognosis is that of a severe developmental encephalopathy, with severe ID, absent speech and, in about half, epilepsy [19, 25]. In the form with a probable digenic cause described by Rochtus, the breathing abnormalities were fatal in the first year [49].

11.2. Surveillance

The sources consulted do not define a specific surveillance protocol with fixed intervals for the 2p16.3 deletion. The table brings together the applicable general recommendations and the timing at which each source places them.

12. Summary and Final Considerations

The 2p16.3 deletion involving NRXN1 is one of the most frequent CNVs in patients studied by microarray for developmental delay, ID or ASD, and is present in about 1 in 1400 to 1 in 3000 people in the general European population. NRXN1 encodes neurexin-1, a presynaptic adhesion molecule with alpha and beta isoforms and thousands of splice variants; ClinGen recognises haploinsufficiency of the gene and a definitive relationship with complex neurodevelopmental disorder. The deletions are non-recurrent, result from replication errors at an unstable locus and are concentrated in the promoter and initial exons of the alpha isoform.

The phenotype includes speech and language delay, ID of variable degree, ASD, ADHD, generalised epilepsy, Tourette syndrome, schizophrenia and affective and anxiety disorders, without a characteristic facies and with rare malformations. Penetrance is incomplete, about one third for any developmental disorder and about 6% for schizophrenia, depends on the position of the deletion, is modulated by other rare variants and is lower in population studies than in case-control studies. Exclusively intronic deletions, for the most part, do not demonstrably increase risk.

The heterozygous deletion is a susceptibility factor and not a diagnosis that alone explains a clinical picture. Once it has been found, the clinician should classify it according to the ACMG/ClinGen standard, stating the transcript and exons, study the parents, look for other genetic factors and, in the presence of a severe, early picture, sequence the spared allele to exclude Pitt-Hopkins-like syndrome 2, the autosomal recessive form due to biallelic loss, with severe ID, epilepsy, breathing abnormalities, stereotypies and constipation.

There is no specific treatment and there are no therapeutic trials; management rests on anticipatory surveillance of development and behaviour, early language and behavioural intervention and conventional treatment of comorbidities. Counselling should explain the 50% transmission risk in a carrier, the unpredictability of the phenotype, including in prenatal diagnosis, and the variability between members of the same family. Research in human cells has shown isoform-dependent loss- and gain-of-function mechanisms, which may guide future targeted therapies, as yet without clinical translation.

13. References

1. OMIM. Neurexin 1; NRXN1. MIM 600565 [Internet]. Baltimore: Johns Hopkins University; last edited 28-09-2026 [accessed 2026-11-01]. Available from: https://www.omim.org/entry/600565
2. Unique – Rare Chromosome Disorder Support Group. 2p16.3 (NRXN1) deletions. Information guide, version 2.0. Oxted: Unique; 2024 [Internet]. Available from: https://rarechromo.org/media/information/Chromosome%20%202/2p16%203%20(NRXN1)%20deletions%20FTNP.pdf
3. Simons Searchlight. 2p16.3 deletion (NRXN1) [Internet]. Available from: https://www.simonssearchlight.org/research/what-we-study/nrxn1
4. Orphanet. NRXN1-related severe neurodevelopmental disorder-motor stereotypies-chronic constipation-sleep-wake cycle disturbance. ORPHA:600663 [Internet]. Available from: https://www.orpha.net/en/disease/detail/600663
5. NCBI/EMBL-EBI. MANE (Matched Annotation from NCBI and EMBL-EBI) v1.5 summary: NRXN1 MANE Select NM_001330078.2/ENST00000401669.7 [Internet]. Available from: https://ftp.ncbi.nlm.nih.gov/refseq/MANE/MANE_human/current/
6. ClinGen. Dosage Sensitivity Curation: NRXN1 (HGNC:8008) — haploinsufficiency score 3, triplosensitivity score 0; last evaluated 2021-06-08 [Internet]. Clinical Genome Resource; [accessed 2026-11-01]. Available from: https://search.clinicalgenome.org/kb/gene-dosage/HGNC:8008
7. Zweier C, de Jong EK, Zweier M, Orrico A, Ousager LB, Collins AL, et al. CNTNAP2 and NRXN1 are mutated in autosomal-recessive Pitt-Hopkins-like mental retardation and determine the level of a common synaptic protein in Drosophila. Am J Hum Genet. 2009;85(5):655-66. doi:10.1016/j.ajhg.2009.10.004. PMID: 19896112. PMC2775834.
8. Harrison V, Connell L, Hayesmoore J, McParland J, Pike MG, Blair E. Compound heterozygous deletion of NRXN1 causing severe developmental delay with early onset epilepsy in two sisters. Am J Med Genet A. 2011;155A(11):2826-31. doi:10.1002/ajmg.a.34255. PMID: 21964664.
9. Lowther C, Speevak M, Armour CM, Goh ES, Graham GE, Li C, et al. Molecular characterization of NRXN1 deletions from 19,263 clinical microarray cases identifies exons important for neurodevelopmental disease expression. Genet Med. 2017;19(1):53-61. doi:10.1038/gim.2016.54. PMID: 27195815. PMC4980119.
10. Schaaf CP, Boone PM, Sampath S, Williams C, Bader PI, Mueller JM, et al. Phenotypic spectrum and genotype-phenotype correlations of NRXN1 exon deletions. Eur J Hum Genet. 2012;20(12):1240-7. doi:10.1038/ejhg.2012.95. PMID: 22617343. PMC3499754.
11. Chen X, Shen Y, Zhang F, Chiang C, Pillalamarri V, Blumenthal I, et al. Molecular analysis of a deletion hotspot in the NRXN1 region reveals the involvement of short inverted repeats in deletion CNVs. Am J Hum Genet. 2013;92(3):375-86. doi:10.1016/j.ajhg.2013.02.006. PMID: 23472757. PMC3591860.
12. Aksu Uzunhan T, Ayaz A. Homozygous exonic and intragenic NRXN1 deletion presenting as either West syndrome or autism spectrum disorder in two siblings. Clin Neurol Neurosurg. 2022;214:107141. doi:10.1016/j.clineuro.2022.107141. PMID: 35101781.
13. Cosemans N, Vandenhove L, Vogels A, Devriendt K, Van Esch H, Van Buggenhout G, et al. The clinical relevance of intragenic NRXN1 deletions. J Med Genet. 2020;57(5):347-355. doi:10.1136/jmedgenet-2019-106448. PMID: 31932357.
14. ClinGen Intellectual Disability and Autism Gene Curation Expert Panel. Gene-disease validity: NRXN1 – complex neurodevelopmental disorder (MONDO:0100038), autosomal dominant, Definitive; 2019-07-10 [Internet]. Available from: https://search.clinicalgenome.org/kb/gene-validity/CGGV:assertion_14bb833a-eefd-4f02-b112-c74b6866d5be-2019-07-10T100000.000Z
15. OMIM. Chromosome 2p16.3 deletion syndrome. MIM 614332 [Internet]. Last edited: 6 November 2025. Available from: https://www.omim.org/entry/614332
16. Online Mendelian Inheritance in Man. #621407 Schizophrenia 17; SCZD17 [Internet]. Baltimore: Johns Hopkins University; last edited 2026-01-14 [cited 2026-11-01]. Available from: https://www.omim.org/entry/621407
17. OMIM. Pitt-Hopkins-like syndrome 2; PTHSL2. MIM 614325 [Internet]. Available from: https://www.omim.org/entry/614325
18. Montalbano S, Krebs MD, Rosengren A, Vaez M, Hellberg KG, Mortensen PB, et al. Analysis of exonic deletions in a large population study provides novel insights into NRXN1 pathology. NPJ Genom Med. 2024;9(1):67. doi:10.1038/s41525-024-00450-8. PMID: 39695155. PMC11655628.
19. Pavone P, Pappalardo XG, Parano C, Falsaperla R, Corsello A, Parano E, et al. NRXN1-related disorders, attempt to better define clinical assessment. Open Med (Wars). 2024;19(1):20240979. doi:10.1515/med-2024-0979. PMID: 39655047. PMC11627049.
20. Kirov G, Rees E, Walters JT, Escott-Price V, Georgieva L, Richards AL, et al. The penetrance of copy number variations for schizophrenia and developmental delay. Biol Psychiatry. 2014;75(5):378-85. doi:10.1016/j.biopsych.2013.07.022. PMID: 23992924. PMC4229045.
21. Castronovo P, Baccarin M, Ricciardello A, Picinelli C, Tomaiuolo P, Cucinotta F, et al. Phenotypic spectrum of NRXN1 mono- and bi-allelic deficiency: A systematic review. Clin Genet. 2020;97(1):125-137 (published online in 2019). doi:10.1111/cge.13537. PMID: 30873608.
22. Brignell A, St John M, Boys A, Bruce A, Dinale C, Pigdon L, et al. Characterization of speech and language phenotype in children with NRXN1 deletions. Am J Med Genet B Neuropsychiatr Genet. 2018;177(8):700-708. doi:10.1002/ajmg.b.32664. PMID: 30358070.
23. Rujescu D, Ingason A, Cichon S, Pietiläinen OP, Barnes MR, Toulopoulou T, et al. Disruption of the neurexin 1 gene is associated with schizophrenia. Hum Mol Genet. 2009;18(5):988-96. doi:10.1093/hmg/ddn351. PMID: 18945720. PMC2695245.
24. Béna F, Bruno DL, Eriksson M, van Ravenswaaij-Arts C, Stark Z, Dijkhuizen T, et al. Molecular and clinical characterization of 25 individuals with exonic deletions of NRXN1 and comprehensive review of the literature. Am J Med Genet B Neuropsychiatr Genet. 2013;162B(4):388-403. doi:10.1002/ajmg.b.32148. PMID: 23533028.
25. Guo X, Wang C, Long D, Zhang H, Yin S, Peng X, et al. Clinical phenotypic spectrum of NRXN1 microdeletions and their association with epilepsy: A systematic review and meta-analysis. Epilepsia. 2025;66(6):2022-2035. doi:10.1111/epi.18337. PMID: 40126490.
26. Miller DT, Adam MP, Aradhya S, Biesecker LG, Brothman AR, Carter NP, et al. Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies. Am J Hum Genet. 2010;86(5):749-64. doi:10.1016/j.ajhg.2010.04.006. PMID: 20466091. PMC2869000.
27. Manickam K, McClain MR, Demmer LA, Biswas S, Kearney HM, Malinowski J, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-2037. doi:10.1038/s41436-021-01242-6. PMID: 34211152.
28. GeneGuidance. NRXN1 — gene record and clinical curation (version of 26 September 2026).
29. European Reference Network on Rare Congenital Malformations and Rare Intellectual Disability (ERN ITHACA) [Internet]. Available from: https://ern-ithaca.eu
30. Ching MS, Shen Y, Tan WH, Jeste SS, Morrow EM, Chen X, et al. Deletions of NRXN1 (neurexin-1) predispose to a wide spectrum of developmental disorders. Am J Med Genet B Neuropsychiatr Genet. 2010;153B(4):937-47. doi:10.1002/ajmg.b.31063. PMID: 20468056. PMC3001124.
31. Szatmari P, Paterson AD, Zwaigenbaum L, Roberts W, Brian J, Liu XQ, et al. Mapping autism risk loci using genetic linkage and chromosomal rearrangements. Nat Genet. 2007;39(3):319-28. doi:10.1038/ng1985. PMID: 17322880. PMC4867008.
32. Kim HG, Kishikawa S, Higgins AW, Seong IS, Donovan DJ, Shen Y, et al. Disruption of neurexin 1 associated with autism spectrum disorder. Am J Hum Genet. 2008;82(1):199-207. doi:10.1016/j.ajhg.2007.09.011. PMID: 18179900. PMC2253961.
33. Dabell MP, Rosenfeld JA, Bader P, Escobar LF, El-Khechen D, Vallee SE, et al. Investigation of NRXN1 deletions: clinical and molecular characterization. Am J Med Genet A. 2013;161A(4):717-31. doi:10.1002/ajmg.a.35780. PMID: 23495017.
34. Rees E, Walters JT, Georgieva L, Isles AR, Chambert KD, Richards AL, et al. Analysis of copy number variations at 15 schizophrenia-associated loci. Br J Psychiatry. 2014;204(2):108-14. doi:10.1192/bjp.bp.113.131052. PMID: 24311552. PMC3909838.
35. Marshall CR, Howrigan DP, Merico D, Thiruvahindrapuram B, Wu W, Greer DS, et al. Contribution of copy number variants to schizophrenia from a genome-wide study of 41,321 subjects. Nat Genet. 2017;49(1):27-35. doi:10.1038/ng.3725. PMID: 27869829. PMC5737772.
36. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Deletions and duplications (including microdeletions and microduplications); p. 660–665.
37. Kastratovic N, Gazdic Jankovic M, Miletic Kovacevic M, Nikolic S, Pavlovic D, Perovic D, et al. Intrafamilial Variability in NRXN1-Associated Neurodevelopmental Disorders: Clinical and Genetic Insights from a Family Case Study with Literature Review. Int J Mol Sci. 2026;27(14). doi:10.3390/ijms27146241. PMID: 42511584. PMC13409882.
38. Pak C, Danko T, Zhang Y, Aoto J, Anderson G, Maxeiner S, et al. Human Neuropsychiatric Disease Modeling using Conditional Deletion Reveals Synaptic Transmission Defects Caused by Heterozygous Mutations in NRXN1. Cell Stem Cell. 2015;17(3):316-28. doi:10.1016/j.stem.2015.07.017. PMID: 26279266. PMC4560990.
39. Sebastian R, Jin K, Pavon N, Bansal R, Potter A, Song Y, et al. Schizophrenia-associated NRXN1 deletions induce developmental-timing- and cell-type-specific vulnerabilities in human brain organoids. Nat Commun. 2023;14(1):3770. doi:10.1038/s41467-023-39420-6. PMID: 37355690. PMC10290702.
40. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: Accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. PMID: 28967166. PMC5765404.
41. Duong LT, Hoeffding LK, Petersen KB, Knudsen CD, Thygesen JH, Klitten LL, et al. Two rare deletions upstream of the NRXN1 gene (2p16.3) affecting the non-coding mRNA AK127244 segregate with diverse psychopathological phenotypes in a family. Eur J Med Genet. 2015;58(12):650-3. doi:10.1016/j.ejmg.2015.11.004. PMID: 26563496.
42. Al Shehhi M, Forman EB, Fitzgerald JE, McInerney V, Krawczyk J, Shen S, et al. NRXN1 deletion syndrome; phenotypic and penetrance data from 34 families. Eur J Med Genet. 2019;62(3):204-209. doi:10.1016/j.ejmg.2018.07.015. PMID: 30031152.
43. Fernando MB, Fan Y, Zhang Y, Tokolyi A, Murphy AN, Kammourh S, et al. Phenotypic complexities of rare heterozygous neurexin-1 deletions. Nature. 2025;642(8068):710-720. doi:10.1038/s41586-025-08864-9. PMID: 40205044.
44. Haskell D, Hart MP. In vivo dissection of human NRXN1 isoforms reveals gain-of-function pathogenicity of schizophrenia-associated 3' deletions. Hum Mol Genet. 2026;35(18). doi:10.1093/hmg/ddag085. PMID: 42696786. PMC13626207.
45. Møller RS, Weber YG, Klitten LL, Trucks H, Muhle H, Kunz WS, et al. Exon-disrupting deletions of NRXN1 in idiopathic generalized epilepsy. Epilepsia. 2013;54(2):256-64. doi:10.1111/epi.12078. PMID: 23294455.
46. Cameli C, Viggiano M, Rochat MJ, Maresca A, Caporali L, Fiorini C, et al. An increased burden of rare exonic variants in NRXN1 microdeletion carriers is likely to enhance the penetrance for autism spectrum disorder. J Cell Mol Med. 2021;25(5):2459-2470. doi:10.1111/jcmm.16161. PMID: 33476483. PMC7933976.
47. Maury EA, Sherman MA, Genovese G, Gilgenast TG, Kamath T, Burris SJ, et al. Schizophrenia-associated somatic copy-number variants from 12,834 cases reveal recurrent NRXN1 and ABCB11 disruptions. Cell Genom. 2023;3(8):100356. doi:10.1016/j.xgen.2023.100356. PMID: 37601975. PMC10435376.
48. Holmquist P. A boy with dysmorphic features, intellectual disability, and biallelic homozygous deletion in NRXN1. Clin Dysmorphol. 2015;24(2):75-8. doi:10.1097/MCD.0000000000000070. PMID: 25486015.
49. Rochtus AM, Trowbridge S, Goldstein RD, Sheidley BR, Prabhu SP, Haynes R, et al. Mutations in NRXN1 and NRXN2 in a patient with early-onset epileptic encephalopathy and respiratory depression. Cold Spring Harb Mol Case Stud. 2019;5(1). doi:10.1101/mcs.a003442. PMID: 30709877. PMC6371743.
50. Viñas-Jornet M, Esteba-Castillo S, Gabau E, Ribas-Vidal N, Baena N, San J, et al. A common cognitive, psychiatric, and dysmorphic phenotype in carriers of NRXN1 deletion. Mol Genet Genomic Med. 2014;2(6):512-21. doi:10.1002/mgg3.105. PMID: 25614873. PMC4303221.
51. Huang AY, Yu D, Davis LK, Sul JH, Tsetsos F, Ramensky V, et al. Rare Copy Number Variants in NRXN1 and CNTN6 Increase Risk for Tourette Syndrome. Neuron. 2017;94(6):1101-1111.e7. doi:10.1016/j.neuron.2017.06.010. PMID: 28641109. PMC5568251.
52. Kendall KM, Rees E, Escott-Price V, Einon M, Thomas R, Hewitt J, et al. Cognitive Performance Among Carriers of Pathogenic Copy Number Variants: Analysis of 152,000 UK Biobank Subjects. Biol Psychiatry. 2017;82(2):103-110. doi:10.1016/j.biopsych.2016.08.014. PMID: 27773354.
53. AlHousni SS, Idris A, Al-Mamari W, Al-Sayegh A. Clinical Phenotypes Associated with NRXN1 Deletions in Five Children with Autism Spectrum Disorder in Oman. Sultan Qaboos Univ Med J. 2026;26(1):448-454. doi:10.18295/2075-0528.3002. PMID: 42549207. PMC13431230.
54. Alfieri P, Scibelli F, Sinibaldi L, Valeri G, Caciolo C, Novello RL, et al. Further insight into the neurobehavioral pattern of children carrying the 2p16.3 heterozygous deletion involving NRXN1: Report of five new cases. Genes Brain Behav. 2020;19(7):e12687. doi:10.1111/gbb.12687. PMID: 32658356.
55. Fitzgerald J, Molloy CJ, Dinneen T, Feerick NE, O'Sullivan M, O'Conaill R, et al. Evidence of neurocognitive and resting state functional connectivity differences in carriers of NRXN1 deletions. J Neurodev Disord. 2025;17(1):62. doi:10.1186/s11689-025-09625-5. PMID: 41094379. PMC12522947.
56. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Autism and autism spectrum disorders; p. 350–353.
57. Todarello G, Feng N, Kolachana BS, Li C, Vakkalanka R, Bertolino A, et al. Incomplete penetrance of NRXN1 deletions in families with schizophrenia. Schizophr Res. 2014;155(1-3):1-7. doi:10.1016/j.schres.2014.02.023. PMID: 24680031.
58. Kamgang S, Myers KA. Evolution from reflex myoclonic epilepsy of infancy to childhood absence epilepsy in NRXN1-related disorder. Epileptic Disord. 2026;. doi:10.1002/epd2.70409. PMID: 42728855.
59. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Epilepsy in infants and children; p. 410–415.
60. Baker K, Costain G, Fung WL, Bassett AS. Chromosomal microarray analysis-a routine clinical genetic test for patients with schizophrenia. Lancet Psychiatry. 2014;1(5):329-31. doi:10.1016/S2215-0366(14)70308-6. PMID: 26360988.
61. Hyman SL, Levy SE, Myers SM. Identification, Evaluation, and Management of Children With Autism Spectrum Disorder. Pediatrics. 2020;145(1). doi:10.1542/peds.2019-3447. PMID: 31843864.
62. Chen CP. Prenatal diagnosis of a familial 2p16.3 deletion encompassing NRXN1 associated with no apparently phenotypic abnormality in the family carrier members. Taiwan J Obstet Gynecol. 2026;65(5):1044-1046. doi:10.1016/j.tjog.2026.07.009. PMID: 42642129.
63. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Intellectual disability; p. 188–193.
64. Schaefer GB, Mendelsohn NJ. Clinical genetics evaluation in identifying the etiology of autism spectrum disorders: 2013 guideline revisions. Genet Med. 2013;15(5):399-407. doi:10.1038/gim.2013.32. PMID: 23519317.
65. Morris E, Inglis A, Austin J. Psychiatric genetic counseling for people with copy number variants associated with psychiatric conditions. Clin Genet. 2022;102(5):369-378. doi:10.1111/cge.14210. PMID: 35996207.
66. Riggs ER, Andersen EF, Cherry AM, Kantarci S, Kearney H, Patel A, et al. Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen). Genet Med. 2020;22(2):245-257. doi:10.1038/s41436-019-0686-8. PMID: 31690835. PMC7313390.
67. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Chromosomal/genomic imbalance; p. 929–937.
68. Committee on Genetics and the Society for Maternal-Fetal Medicine. Committee Opinion No.682: Microarrays and Next-Generation Sequencing Technology: The Use of Advanced Genetic Diagnostic Tools in Obstetrics and Gynecology. Obstet Gynecol. 2016;128(6):e262-e268. doi:10.1097/AOG.0000000000001817. PMID: 27875474.
69. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. Autismo ou transtorno do espectro autista (TEA); p. 142–145.
70. ClinicalTrials.gov. Online Study of People Who Have Genetic Changes and Features of Autism: Simons Searchlight. NCT01238250 [Internet]. Available from: https://clinicaltrials.gov/study/NCT01238250
71. Friedman AE, Perni M, Millard J, Karanfilovski D, Granato M, Campbell PD. Unique and overlapping behavioral effects of isoform-specific NRXN1 deletions. Dis Model Mech. 2026;19(7). doi:10.1242/dmm.052862. PMID: 42298972. PMC13446563.
72. Shimelis H, Oetjens MT, Walsh LK, Wain KE, Znidarsic M, Myers SM, et al. Prevalence and Penetrance of Rare Pathogenic Variants in Neurodevelopmental Psychiatric Genes in a Health Care System Population. Am J Psychiatry. 2023;180(1):65-72. doi:10.1176/appi.ajp.22010062. PMID: 36475376. PMC10017070.