Clinical phenotype, gonadal development, and comorbidity spectrum in 43 children with triple X syndrome: a single-center retrospective descriptive case series with cytogenetic refinement in patients with and without X-monosomy-containing cell lines

Introduction

Background

Triple X syndrome (TXS) occurs in roughly 1 per 1,000 female births (1,2), yet only 13–19% of cases receive clinical diagnosis (1,3). Most patients remain undetected until developmental or reproductive issues prompt diagnostic evaluation. Recent nationwide epidemiological data from Denmark further underscore this diagnostic gap, revealing that the overwhelming majority of affected females remain undiagnosed throughout their lifespan, with the diagnosis often incidental to evaluation for developmental, reproductive, or psychiatric concerns (3,4). This gap between true prevalence and clinical detection carries real consequences: children who go undiagnosed miss the window for early intervention, and families are left without guidance on the neurodevelopmental and reproductive challenges that may lie ahead.

Phenotypic variability in TXS is largely attributed to overexpression of XCI-escape genes, which disrupts dosage-sensitive pathways in development (5,6). Clinical features vary by age: infants may show hypotonia and feeding problems; children often present with speech delay, low body mass index (BMI), behavioral issues and menstrual irregularities, premature ovarian insufficiency (POI), or other reproductive dysfunction in adolescence and adulthood (5,7,8). Neuroimaging (9) and psychiatric registry studies (3,10) have both underscored the neurodevelopmental burden of TXS.

The growing recognition that TXS is not a benign condition has prompted increasing attention to its multisystem involvement. Cardiovascular anomalies, including atrial septal defects, have been documented in registry-based studies (4). Renal and genitourinary abnormalities, including kidney dysplasia and ovarian malformations, are increasingly recognized associated findings (5). This study aimed to describe clinical phenotypes, gonadal development, and comorbidity profiles in parallel across karyotype subgroups in pediatric TXS patients (11).

Rationale and knowledge gap

Despite the expanding knowledge base regarding TXS phenotypes, the karyotypic heterogeneity of this condition has received little attention. Beyond the classic non-mosaic 47,XXX, the spectrum encompasses 48,XXXX, mosaic 47,XXX/46,XX, and—most notably—mosaic forms containing X-monosomy cell lines (e.g., 45,X/47,XXX). This last category occurs in approximately 10–20% of TXS cases (2,12). These patients may show overlapping features of both Turner syndrome and TXS, a phenotypic overlap that existing guidelines do not directly address.

The clinical significance of this heterogeneity is underscored by Turner syndrome research, where mosaic cell line proportions affect phenotypic severity. In 45,X cells, (SHOX) haploinsufficiency drives short stature and skeletal anomalies (13,14); in 47,XXX cells, XCI-escape gene overexpression produces distinct effects. How these opposing dosage effects interact—and their clinical consequences—remains unstudied.

Recent epigenomic advances further complicate this picture. Sex chromosome dosage variations induce transcriptomic changes extending beyond the X chromosome (15,16), with HTR2C hypomethylation and ZNF673 dysregulation implicated in neuropsychiatric phenotypes (16). Whether mosaicism modifies these epigenetic mechanisms is unknown. Consequently, clinicians lack evidence-based guidance for patients with non-classic karyotypes.

Objective

The present study retrospectively analyzed 43 pediatric TXS cases from Jiangxi Provincial Children’s Hospital (January 2019 to November 2024). To our knowledge, few prior studies have explicitly stratified pediatric TXS by karyotype subgroup; we therefore characterized clinical phenotype, gonadal development, bone maturation, and comorbidity profiles in parallel across non-mosaic, X-monosomy mosaic, and non-X-monosomy mosaic subgroups, with the primary biologically interpretable distinction being patients with versus without X-monosomy-containing cell lines. Given the small subgroup sizes, no formal between-group hypothesis tests were performed; subgroup-level findings are reported as descriptive observations to inform a tentative, practice-informed framework for considering karyotype subgroup in surveillance, intended as a starting point for hypothesis-driven prospective studies rather than for direct clinical adoption. We present this article in accordance with the STROBE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0264/rc).

Methods

Study population

We conducted a single-center retrospective chart review at our institution. We initially screened the cytogenetics laboratory database for all peripheral-blood karyotyping results performed on female pediatric patients between January 2019 and November 2024 (n=51 with a karyotype containing extra X material or X-monosomy mosaicism). Of these, 8 cases were excluded: 4 with critically incomplete clinical records, 3 with co-existing severe chromosomal disorders unrelated to TXS, and 1 with insufficient follow-up data. The final cohort comprised 43 patients. The patient selection flow is shown in revised Figure 1. We enrolled 43 children (aged 3 days to 16 years) diagnosed with TXS at Jiangxi Provincial Children’s Hospital between January 2019 and November 2024. Inclusion criteria were: (I) confirmed diagnosis by G-banding karyotype analysis of peripheral blood lymphocytes; (II) complete or retrievable clinical records; and (III) written informed consent from guardians. Exclusion criteria included: (I) co-existing severe chromosomal disorders; and (II) critically incomplete data. The overall study design, patient enrollment, karyotype stratification, and analytical framework are summarized in Figure 1.

Figure 1 CONSORT-style flow diagram of patient identification, eligibility assessment, and inclusion. Of 51 female pediatric patients with karyotyping results containing extra X material or X-monosomy mosaicism between January 2019 and November 2024, 8 were excluded (4 with critically incomplete clinical records, 3 with co-existing severe chromosomal disorders unrelated to TXS, 1 with insufficient follow-up data), leaving 43 patients in the final analysis. Karyotype stratification yielded three subgroups: non-mosaic (n=31, 72.1%), X-monosomy mosaic (n=9, 20.9%), and non-X-monosomy mosaic (n=3, 7.0%). The downstream surveillance content is now described in the text and Figure 4 as a tentative, practice-informed framework rather than an evidence-based clinical management algorithm. FISH, fluorescence in situ hybridization; TXS, triple X syndrome.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Jiangxi Provincial Children’s Hospital (approval No. JXSETYY-YXKY-20250019). Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.

Karyotype subgroup stratification

Based on karyotype analysis results, patients were stratified into three subgroups, with the primary biologically interpretable comparison being patients with versus without X-monosomy-containing cell lines: (I) non-mosaic group: including 47,XXX, 48,XXXX, and other non-mosaic karyotypes; (II) X-monosomy mosaic group: mosaic karyotypes containing 45,X cell lines (e.g., 45,X/47,XXX, 45,X/47,XXX/46,XX), where clinical features may be influenced by Turner syndrome components; and (III) Non-X-monosomy mosaic group: mosaic karyotypes without 45,X cell lines (e.g., 47,XXX/46,XX). Two atypical karyotypes—48,XXXX (n=1) and 45,X/46,XY/47,XYY (n=1)—were retained because both involve X-chromosome dosage imbalance and were referred for evaluation under the same TXS workup pathway as our other patients; we acknowledge these cases lie at the edge of a strict TXS definition and have presented their individual data separately and interpreted them with caution (see Results, “Karyotype distribution and baseline characteristics” section). Cytogenetic methods were as follows: G-banded karyotyping was performed on phytohaemagglutinin-stimulated peripheral blood lymphocyte cultures with GTG banding at 400–550 band resolution; ≥20 metaphases were routinely analyzed, and ≥30 metaphases when mosaicism was suspected. Subgroup assignment was based primarily on the G-banding peripheral blood karyotype. Interphase FISH using commercial dual-color centromeric probes (CEP X spectrum-green and CEP Y spectrum-orange; Vysis/Abbott Molecular), counting ≥200 nuclei per case, was performed in 5 selected cases (n=5/43, 11.6%) where (I) G-banding suggested mosaicism that required cell-line quantification; (II) the clinical phenotype was atypical for the apparent karyotype; or (III) a critical clinical decision (e.g., classification of a patient with primary ovarian insufficiency) required refined cytogenetic information. Mosaicism was defined as the presence of two or more cell lines with distinct sex-chromosome constitutions detected by G-banding or, where performed, FISH; a low-frequency cell line was retained when present in ≥5% of cells in either G-banding or FISH counts (or in ≥3 metaphases on G-banding). For the 5 patients with FISH, FISH-based cell-line proportions are reported in Table S1 for cytogenetic transparency, but subgroup assignment was based on the initial G-banding karyotype throughout this study and was not altered by FISH findings. For the remaining 38 patients, classification rests on G-banding alone, and we cannot exclude undetected low-level mosaicism. Individual cell-line proportions for the 5 FISH-tested patients are provided in Table S1. Given that the non-X-monosomy mosaic subgroup contained only 3 patients and the X-monosomy mosaic subgroup 9 patients, all subgroup-level findings throughout this manuscript are presented as descriptive observations to characterize the phenotypic spectrum, and no formal between-group statistical comparisons were performed. This stratification framework was designed to describe in parallel the phenotypic spectrum within and across subgroups defined by the presence or absence of X-monosomy-containing cell lines, without inferential between-group testing.

Data collection

Clinical data collected included: (I) demographics and birth history; (II) anthropometric measurements: height, weight, BMI, height standard deviation score (HtSDS), BMI standard deviation score (BMISDS); (III) pubertal assessment: Tanner staging, special somatic features; (IV) laboratory and imaging data: karyotype, fluorescence in situ hybridization (FISH, counting ≥200 cells), sex hormones (LH, FSH, E2), anti-Müllerian hormone (AMH), insulin-like growth factor-1 (IGF-1), uterine/ovarian ultrasonography, and bone age (Greulich-Pyle method); and (V) comorbidity documentation and treatment records.

Statistical analysis

Statistical analyses were performed using SPSS version 27.0. Normally distributed continuous variables were expressed as mean ± standard deviation, non-normally distributed variables as median (interquartile range), and categorical variables as counts (percentages). Given the small subgroup sizes (n=31, n=9, n=3), no formal between-group hypothesis tests were performed; subgroup-level findings are reported as descriptive statistics in parallel. Anthropometric values are reported as both raw measurements and HtSDS/BMISDS calculated against contemporary Chinese pediatric reference standards (Li et al., 2009). Subgroup-level findings throughout the Results are presented as descriptive statistics for each group, listed in parallel rather than as comparisons. Percentages based on small denominators are presented alongside their absolute counts (e.g., 3/9, 33.3%) to allow the reader to assess data sparsity directly.

Results

Karyotype distribution and baseline characteristics

Of the 43 patients, 31 (72.1%) were classified into the non-mosaic group, 9 (20.9%) into the X-monosomy mosaic group, and 3 (7.0%) into the non-X-monosomy mosaic group (Figure 1; Figure 2A). The non-mosaic group comprised 29 cases of 47,XXX, 1 case of 48,XXXX, and 1 case of 47,XXX with inversion of chromosome 9. The X-monosomy mosaic group included 45,X/47,XXX (4 cases), 45,X/47,XXX/46,XX (2 cases), 45,X/46,X,i(X)q(10) (1 case), 45,X/46,XY/47,XYY (1 case), and 45,X/47,XXX/46,XX (1 case). The non-X-monosomy mosaic group consisted of 3 cases of 47,XXX/46,XX.The non-X-monosomy mosaic group comprised only 3 patients, All findings from this subgroup are presented descriptively. Findings related to this group should be considered descriptive.

Figure 2 Composite clinical characteristics by karyotype subgroup. (A) Karyotype-comorbidity alluvial diagram showing the flow of patients from each karyotype subgroup to comorbidity categories. Flow width is proportional to patient count. The X-monosomy mosaic group shows prominent connections to skeletal, renal, and cardiovascular comorbidities. (B) Beeswarm plot of bone age difference (chronological age minus bone age) by subgroup. Individual data points are shown with interquartile range boxes and mean diamonds. The X-monosomy mosaic group showed a mean bone age difference of +1.36 years (n=7) and the non-mosaic subgroup −0.10 years (n=20); these subgroup-level descriptions are presented in parallel and were not compared by formal statistical test. The single non-X-monosomy mosaic patient with bone age data showed advanced bone maturation (1 year ahead of chronological age), though this observation requires caution given n=1. Background shading indicates delayed (warm) versus advanced (cool) bone age zones. (C) Radar chart summarizing comorbidity profiles across subgroups. The X-monosomy mosaic group exhibits a distinctive Turner syndrome-like profile with elevated rates of skeletal anomalies (33.3%), renal disease (22.2%), and cardiovascular involvement (22.2%). BA, bone age; CA, chronological age.

The median age at diagnosis was 7.58 years (range: 1 month to 13 years 7 months). Growth retardation was the most common presenting complaint, followed by breast development/menarche, speech delay, and feeding difficulties. By age group: 0–5 years, 16 patients (37.2%); 6–12 years, 25 patients (58.1%); and >12 years, 2 patients (4.7%). The median height was 117.6 cm, mean weight was 19.77±10.8 kg, and mean BMI was 14.93±2.4 kg/m². Because the cohort spanned infancy to adolescence, age-standardized values were calculated: mean HtSDS was −0.32±1.18 in the non-mosaic group, −1.05±1.04 in the X-monosomy mosaic group, and −1.48±0.86 in the non-X-monosomy mosaic group, HtSDS values by subgroup were −0.32±1.18 (non-mosaic), −1.05±1.04 (X-monosomy mosaic), and −1.48±0.86 (non-X-monosomy mosaic); these are reported for descriptive purposes only. Mean BMISDS values were within the normal range across all three subgroups (−0.45±1.21, −0.18±0.94, and −0.62±1.33, respectively). Baseline characteristics by karyotype subgroup are presented in Table 1. Two atypical karyotypes were observed within the cohort and are presented separately rather than as part of subgroup means. Case 1 (48,XXXX, age 1 year): presented with(This girl presented with growth retardation (length 76.5 cm, weight 7.95 kg) and rib flaring; basal FSH was 6.37 U/L with LH 0.01 U/L and E2 4.23 pg/mL. FISH refined the initial G-banding karyotype of 48,XXXX to three coexisting cell lines (47,XXX 88%/46,XX 7%/48,XXXX 5%; Table S1), illustrating that G-banding may underestimate cytogenetic complexity even in apparently non-mosaic cases); FISH refined the karyotype to 47,XXX 88%/46,XX 7%/48,XXXX 5% (Table S1). Case 2 (45,X/46,XY/47,XYY, age 14 years): presented with (This phenotypically female patient presented with short stature (140 cm, −2 SDS), absent pubertal development (Tanner B1), shield chest, and hypergonadotropic hypogonadism (LH 46.34 mIU/mL, FSH 127.08 mIU/mL, E2 <15 pg/mL); pelvic imaging showed a hypoplastic uterus and no identifiable ovaries. SRY testing was positive; bilateral gonadectomy revealed streak gonads with ovarian-type stroma and a sex cord-stromal tumor on the right side. This patient was classified in the X-monosomy mosaic subgroup on the basis of the predominant 45,X cell line (45,X[31]/46,XY[18]/47,XYY[1]); however, the clinical picture is more consistent with mixed gonadal dysgenesis than with classical TXS, and interpretation should be tempered accordingly. Both cases were retained because they were referred under the same TXS workup pathway, but their interpretation should be tempered by the recognition that they sit at the edge of a strict TXS definition.

Bone age and growth patterns by subgroups

Among 28 patients with bone age data, the non-mosaic group showed a mean bone age difference of −0.10 years (near normal), whereas the X-monosomy mosaic group had a mean bone age difference of +1.36 years (n=7 with available bone age data); the non-X-monosomy mosaic subgroup is represented by a single observation (−1.00 years; n=1). These subgroup-level descriptions are presented in parallel; no formal between-group testing was performed (Figure 2B). This pattern of skeletal delay in the X-monosomy mosaic group resembles the skeletal delay reported in Turner syndrome and is consistent with a possible influence of the 45,X cell line on bone maturation and linear growth, an observation that requires confirmation in larger cohorts.

Sex hormone levels and gonadal development

Sex hormone data were available for 25 patients. For the overall cohort, the median basal LH was 0.96 U/L [interquartile range (IQR), 0.20–3.18 U/L; n=25], median FSH was 6.09 U/L (IQR, 3.60–7.16 U/L; n=25), and median E2 was 20.99 pg/mL (IQR, 11.80–30.68 pg/mL; n=25). To allow consistent presentation across subgroups, sex hormone values are reported in Table 1 as median (IQR) for each subgroup. The corresponding subgroup median FSH was 6.37 U/L (IQR, 4.11–8.16 U/L) in the non-mosaic group, 3.94 U/L (IQR, 1.59–6.76 U/L) in the X-monosomy mosaic group, and 1.09 U/L (single observation) in the non-X-monosomy mosaic group. Several patients in the X-monosomy mosaic subgroup had FSH >7 U/L, a pattern that may be compatible with reduced ovarian reserve in some individuals; however, given the small subgroup size (n=7 with available data), no firm conclusions can be drawn, and these observations are descriptive only. Fifteen patients (34.9%) experienced spontaneous menarche during follow-up, at a mean age of 11.13±1.28 years (range, 9.0–13.0 years) (Figure 3). For descriptive comparison, the most recent published mean age at menarche in Chinese girls is approximately 12.5 years (17), and the WHO reference figure for general populations is approximately 12.0–13.0 years; our cohort mean of 11.13 years thus appears earlier than these published reference populations, although this comparison is descriptive only and was not made against an internal age-matched control group. The majority of early menarche cases occurred in the non-mosaic group (13/15). Uterine measurements were enlarged in 8 patients (18.6%), and ≥4 follicles >4 mm in diameter were observed in 10 patients (23.3%); the possible mechanistic interpretation of these findings (e.g., earlier activation of the hypothalamic-pituitary-gonadal axis) is considered in the “Discussion” section. One patient (originally classified as non-mosaic on initial G-banding karyotype, with subsequent FISH analysis showing 47,XXX 83.5%/46,XX 16.5%) was diagnosed with POI at age 9 years 9 months, presenting with basal FSH of 57.16 U/L and AMH of 0.12 ng/mL at the time of POI diagnosis. The peak FSH value (57.16 U/L) was documented at initial POI evaluation and is reported here as a clinically meaningful single value rather than within the cross-sectional sex-hormone summaries above (which reflect later steady-state measurements). This case illustrates the paradoxical coexistence of pubertal acceleration and impaired ovarian reserve in TXS.

Figure 3 Age at menarche by karyotype subgroup. Each point represents an individual patient who experienced spontaneous menarche during follow-up. The dashed red line indicates the general population average menarche age (~12.0 years). The dotted blue line shows the mean menarche age of non-mosaic TXS patients (10.9 years). Most TXS patients experienced menarche earlier than the general population average, with the pattern predominantly observed in the non-mosaic group (13/15 patients with menarche data). TXS, triple X syndrome.

FISH analysis

FISH analysis (counting ≥200 nuclei per case) was performed in 5 patients selected on clinical grounds (see Methods, “Karyotype subgroup stratification” section). In these 5 cases, FISH provided more refined cytogenetic information than G-banding alone. Notably: (I) one patient (1 year old) initially karyotyped as 48,XXXX on G-banding was shown by FISH to harbor three coexisting cell lines (47,XXX 88%, 46,XX 7%, 48,XXXX 5%), illustrating that G-banding may underestimate cytogenetic complexity even in cases reported as non-mosaic; (II) one patient referred for primary ovarian insufficiency was shown to have 47,XXX (83.5%) and 46,XX (16.5%) cell lines on FISH, providing the cell-line proportions reported with this case in “Sex hormone levels and gonadal development” section; (III) two patients initially reported as 47,XXX on G-banding were shown on FISH to harbor minor 46,XX cell populations (16.5–18%), with one of these patients also harboring a 1% 45,X clone below the retention threshold; and (IV) one patient with pubertal short stature was confirmed to have a low-level 45,X clone (3.5%) coexisting with 46,XX (72.5%) and 47,XXX (24%). Individual cell-line proportions for these 5 cases are provided in Table S1. These FISH-based observations should not be over-interpreted as a feature of the broader cohort: FISH was performed selectively for clinical indications, not as a universal screening test, and we cannot exclude undetected low-level mosaicism in the 38 patients without FISH.

Comorbidity spectrum

The comorbidity profiles by karyotype subgroup are reported in parallel karyotype subgroups (Table 2; Figure 2A,2C). Overall, neuropsychiatric disorders were the most prevalent comorbidity (11/43, 25.6%), including intellectual disability (6 cases), autism spectrum disorder (2 cases), attention-deficit/hyperactivity disorder (2 cases), and an intracranial lipoma (1 case) Cardiovascular anomalies affected 6 patients (14.0%); the lesions documented across these 6 patients comprised 5 atrial septal defects and 3 patent ductus arteriosus, with two patients having both lesions concurrently. Skeletal anomalies were present in 5 patients (11.6%), renal disease in 3 patients (7.0%), thyroid disease in 2 patients (4.7%), and POI in 1 patient (2.3%).

Subgroup-level comorbidity descriptions are presented in parallel (Figure 2A,2C). Within the X-monosomy mosaic subgroup (n=9), skeletal anomalies were observed in 3/9 (33.3%; two shield chest, one scoliosis) and renal anomalies in 2/9 (22.2%; both horseshoe kidney)—features known to be associated with Turner syndrome. Within the non-mosaic subgroup (n=31), neuropsychiatric disorders were the predominant comorbidity (8/31, 25.8%), with intellectual disability and autism spectrum disorder as the leading manifestations; skeletal and renal anomalies were less commonly recorded (2/31, 6.5% and 1/31, 3.2%, respectively). Within the non-X-monosomy mosaic subgroup (n=3), one patient was the case described in “Sex hormone levels and gonadal development” section with primary ovarian insufficiency. Given the small subgroup sizes—particularly the non-X-monosomy mosaic group (n=3)—these subgroup-level proportions are reported descriptively to characterize the phenotypic spectrum and not as a basis for between-group inference.

Somatic features

Common somatic features across all 43 patients included hypertelorism (5/43, 11.6%), upslanting palpebral fissures (3/43, 7.0%), low-set ears (2/43, 4.7%), shield chest (2/43, 4.7%), and individual cases of high zygomatic arch, micrognathia, cleft palate, epicanthus, small palpebral fissures, and funnel chest. Turner syndrome-associated features such as shield chest and cubitus valgus noted within the X-monosomy mosaic group.

Discussion

Key findings

We stratified 43 pediatric TXS cases by karyotype and found phenotypic patterns that, although based on small subgroup sizes, characterize the clinical spectrum within cytogenetic categories of patients with and without X-monosomy-containing cell lines. Within the X-monosomy mosaic subgroup (n=9), we observed a clinical pattern that current pure-TXS or Turner syndrome guidelines do not directly address. We use the term “intermediate phenotype” descriptively to denote the simultaneous presence in this subgroup of (I) disproportionate bone age delay typically associated with Turner syndrome (mean +1.36 years in this subgroup; for context, mean +1.36 years in this subgroup; the corresponding non-mosaic value (−0.10 years) is reported in Table 1; (II) renal/skeletal anomalies more characteristic of Turner syndrome than of typical 47,XXX [3/9 (33.3%) skeletal anomalies, 2/9 (22.2%) horseshoe kidney; Figure 2A,2C]; and (III) neuropsychiatric findings that remain a feature of the 47,XXX background. We do not imply that this subgroup statistically differs from the non-mosaic subgroup—no formal between-group testing was performed. Rather, this is a single-subgroup description; readers can examine the parallel subgroup-level data in Tables 1,2 and Figure 2A-2C to draw their own descriptive impressions. The mosaic frequency observed (20.9%) is similar in order of magnitude to the ~10% mosaic rate reported by Butnariu et al. (12) and the ~20% postzygotic nondisjunction rate estimated by NORD (2); our higher proportion likely reflects ascertainment bias inherent to a tertiary referral center, rather than a true population difference. Taken together, these subgroup-level observations are hypothesis-generating, and prospective studies with larger numbers in each subgroup will be required before firm conclusions about subgroup-specific management can be drawn.

Strengths and limitations

Several aspects of this study deserve mention. To our knowledge, few previous studies have explicitly stratified pediatric TXS cases by karyotype subgroup; we believe our analysis offers an incremental but clinically useful contribution rather than a wholly novel finding. In selected cases (n=5, 11.6% of the cohort), the addition of FISH (counting ≥200 nuclei) allowed more refined cytogenetic characterization than G-banding alone—illustrated most clearly in one patient initially karyotyped as 48,XXXX where FISH revealed three coexisting cell lines (47,XXX 88%/46,XX 7%/48,XXXX 5%). Routine FISH was not performed across the cohort; for the remaining 38 patients, classification rests on G-banding alone, and we cannot exclude undetected low-level mosaicism. The proposed surveillance framework is grounded in the phenotypic patterns observed in this small cohort combined with existing TXS and Turner syndrome guidelines, and is intended as a tentative starting point for centers without specialist genetics infrastructure rather than as an evidence-based clinical management algorithm. However, several limitations should be acknowledged: (I) the sample size is relatively small (n=43), particularly for the X-monosomy mosaic (n=9) and non-X-monosomy mosaic (n=3) subgroups, given these subgroup sizes we deliberately refrained from formal between-group statistical comparisons, and report subgroup-level data descriptively only; (II) as a single-center retrospective study, selection bias cannot be excluded—importantly, recent systematic reviews have highlighted persistent demographic underrepresentation in SCA research globally (18); (III) long-term follow-up data are lacking, precluding assessment of adult reproductive and psychiatric outcomes; (IV) hormonal assessments were not longitudinal, limiting interpretation of dynamic gonadal function; (V) FISH was performed selectively in 5 clinically indicated cases rather than universally across the cohort, reflecting standard institutional workflow at the time of evaluation; we therefore cannot exclude low-level undetected mosaicism in the 38 patients classified by G-banding alone, particularly within the non-mosaic subgroup. Universal FISH would be a methodological strength of future prospective work. Future multi-center prospective studies with larger cohorts, particularly enriched for X-monosomy mosaic cases, are needed to validate these preliminary findings. Additionally, FISH detected minor 46,XX cell populations (7–18%) in three G-banding-classified non-mosaic patients; a FISH-based reclassification would increase the non-X-monosomy mosaic subgroup from 3 to 6 patients (Table S1).

Comparison with similar research

Our finding that 34.9% of patients experienced spontaneous menarche at a mean age of 11.13 years—descriptively earlier than the published mean age at menarche in Chinese girls (~12.5 years) (17) and the WHO reference of ~12.0–13.0 years for general populations, although no formal statistical comparison against an internal age-matched control was performed—corroborates previous reports of pubertal acceleration in TXS by Davis et al. (19) and aligns with the observation by Tartaglia et al. that pubertal development in TXS is generally within the normal range but may be shifted earlier (7). The overall neuropsychiatric comorbidity rate (25.6%) aligns with both recent large-scale psychiatric registry data from Denmark (3) and previous literature reporting rates of 23–30% (10,20). The iPSYCH2015 dataset analysis by Sánchez et al. demonstrated significantly elevated risks for ADHD, ASD, and schizophrenia spectrum disorders in SCA populations, findings concordant with our observation that neuropsychiatric disorders are the predominant comorbidity in non-mosaic TXS (3).

The cardiovascular disease rate (14.0%) and renal anomaly rate (7.0%) observed in our cohort are noteworthy in the context of rates reported in isolated-TXS cohorts, were observed predominantly within the X-monosomy mosaic group. The presence of horseshoe kidney observed only within the X-monosomy mosaic group is particularly noteworthy, as this anomaly is a hallmark of Turner syndrome and was reported in the nationwide epidemiologic study by Berglund et al. as a frequent renal finding in 47,XXX comorbidity landscape (11). This finding adds to the case for routine renal and cardiac evaluation in TXS patients with 45,X cell lines, rather than reserving such screening for classic Turner syndrome alone. Recent work by Zampini et al. using cluster analysis in young children with SCTs further supports the concept of substantial phenotypic variability within SCA populations, identifying distinct neurodevelopmental profiles that partially align with our observation of qualitative differences in neuropsychiatric presentations across karyotype subgroups in our descriptive analysis (21).

Explanations of findings

The molecular basis for the “intermediate phenotype” likely involves the competing dosage effects of X-linked genes operating at the tissue level. In cells carrying 45,X, haploinsufficiency of SHOX—a homeodomain transcription factor located in the pseudoautosomal region PAR1 that regulates chondrocyte proliferation and differentiation—contributes to short stature, skeletal anomalies, and Madelung deformity characteristic of Turner syndrome (13,14). The GeneReviews entry for SHOX deficiency disorders (updated 2024) emphasizes that the clinical severity of SHOX haploinsufficiency is influenced by hormonal milieu, particularly estrogen levels, which explains the age-dependent worsening of skeletal phenotypes (14). Conversely, in cells carrying 47,XXX, approximately 15% of X-linked genes that escape XCI are overexpressed, creating a supraphysiological dosage of their protein products (6). Within the same patient, some cells carry too few copies of SHOX and other dosage-sensitive genes, while others carry too many—a situation with no direct precedent in either Turner syndrome or classic TXS, and one that likely explains why neither set of existing guidelines fits these children well.

A recent comprehensive systematic review of SCA transcriptomics and epigenomics by Legue et al. [2025] highlighted that sex chromosome dosage variations induce widespread effects extending beyond X-linked genes, with altered DNA methylation patterns at autosomal loci, dysregulated non-coding RNA expression, and modified chromatin architecture (15). These findings caution against interpreting TXS mosaicism purely in terms of gene dosage arithmetic. Epigenetic reprogramming effects that differ by tissue and developmental stage may contribute substantially to the phenotype—a layer of complexity that neither current guidelines nor existing cohort studies have yet addressed.

Paradoxically, pubertal acceleration coexists with evidence of diminished ovarian reserve. One patient (initial G-banding 47,XXX, FISH-detected 16.5% 46,XX line) developed POI at age 9 years 9 months (peak FSH 57.16 U/L, AMH 0.12 ng/mL), and several patients in the X-monosomy mosaic subgroup showed elevated basal FSH levels. Davis et al. demonstrated that TXS girls have significantly lower serum AMH levels compared with age-matched controls (median 0.7 vs. 2.7 ng/mL), with 67% falling below the 2.5th percentile for age, representing an 11-fold increased risk of diminished ovarian reserve (18). More recently, Vashist et al. [2023] reported a case of mosaic TXS with recurrent pregnancy loss and decreased ovarian reserve, underscoring that even mosaic karyotypes carry reproductive risk (22). The X-chromosome harbors multiple ovarian function-related genes (BMP15, FMR1, FMR2, PGRMC1), whose dosage effects may paradoxically accelerate both pubertal onset and ovarian aging through divergent mechanisms—enhanced GnRH pulse generator activation driving early puberty, while accelerated follicular atresia depletes the ovarian reserve (23,24). For clinicians, this means that a girl with TXS who enters puberty early should not be reassured on that basis alone—serial AMH and inhibin B measurements are worth tracking through adolescence, and fertility preservation options should be introduced before ovarian reserve becomes a crisis rather than a consideration.

Within the s: the non-mosaic group predominantly manifested intellectual disability and autism spectrum disorder, whereas the X-monosomy mosaic group showed primarily speech delay and social communication difficulties. This distinction may reflect the differential impacts of X-chromosome gene dosage on neurodevelopmental pathways. Nielsen et al. identified HTR2C hypomethylation and ZNF673 epigenetic dysregulation as contributors to neuropsychiatric phenotypes in 47,XXX (16). Recent neuroimaging data from Domes et al. [2025] revealed significant reductions in hippocampal, amygdalar, and prefrontal gray matter volume in adult women with 47,XXX karyotype, providing a neuroanatomical basis for the cognitive and behavioral profiles observed (9). Whether the 45,X cell line attenuates or amplifies these neuroanatomical changes in mosaic patients is unknown. Neuroimaging studies that specifically recruit 45,X/47,XXX individuals—rather than grouping all TXS together—would be a logical next step (25-27).

Implications and actions needed

The main translational suggestion arising from this study is a tentative, practice-informed framework (Figure 4) for considering karyotype subgroup when planning surveillance in TXS. We emphasize that this framework is based on expert opinion drawing on the patterns observed in our small cohort combined with existing TXS and Turner syndrome guidelines, rather than on prospective evidence; it should therefore be regarded as a starting point for discussion and as a hypothesis to be tested, not as an evidence-based clinical management algorithm. For children in the non-mosaic group, the primary surveillance priorities are neurodevelopmental: cognitive and language assessment, behavioral monitoring, and regular AMH testing to track ovarian reserve, with follow-up every 6–12 months. Children in the X-monosomy mosaic group may benefit from a broader assessment that draws on both TXS and Turner syndrome protocols—renal ultrasonography (given the horseshoe kidney risk identified here), echocardiography, bone density monitoring, and consideration of growth hormone therapy under SHOX deficiency guidelines, with more frequent visits every 3–6 months. Across all subgroups, FISH analysis counting ≥200 cells may be considered at diagnosis to characterize mosaicism more precisely, particularly when initial G-banding suggests mosaicism, the clinical phenotype is atypical, or a critical clinical decision is at stake (as illustrated by the cases reported in our cohort), and care should be coordinated across pediatric endocrinology, neurology, developmental pediatrics, psychology, and genetics.

Figure 4 Tentative, practice-informed framework for considering karyotype subgroup in the surveillance of children with triple X syndrome. This flowchart presents a practice-informed framework, derived from the phenotypic patterns observed in our small descriptive cohort combined with existing TXS and Turner syndrome guidelines, to support—rather than replace—clinical judgment in the multidisciplinary surveillance of these patients. This framework is hypothesis-generating and requires prospective multi-center validation before adoption as an evidence-based management algorithm. The framework encompasses diagnostic confirmation (karyotype analysis + FISH ≥200 cells), subgroup-specific screening priorities (neuropsychiatric focus for non-mosaic; dual TXS + Turner syndrome protocol for X-monosomy mosaic; individualized phenotype assessment for non-X-monosomy mosaic), and a common multidisciplinary follow-up framework with age-appropriate monitoring milestones from infancy through transition to adult care. AMH, anti-Müllerian hormone; EEG, FISH, fluorescence in situ hybridization; GH, TXS, triple X syndrome; US.

Figure 4 maps these priorities onto developmental stages—infancy, childhood, puberty, and the transition to adult care—so that clinicians have a concrete reference at each encounter rather than a general set of principles. Our hope is that this framework proves useful not only for specialist genetics or endocrinology units, but for general pediatric centers where many of these children are first seen and followed.

Conclusions

In this descriptive single-center case series, characterizing pediatric TXS by karyotype subgroup yielded clinically relevant phenotypic observations that warrant further investigation. Within the X-monosomy mosaic subgroup, children appeared to occupy a phenotypic middle ground between Turner syndrome and classical TXS, with mean bone age delay (1.36 years; n=7) and Turner-like renal and skeletal anomalies (2/9 and 3/9 respectively); given the small subgroup size (n=9) and the absence of formal between-group testing, these subgroup-level observations are presented as hypothesis-generating. In selected cases (n=5), FISH counting ≥200 nuclei provided cytogenetic refinement beyond G-banding, supporting consideration of FISH at diagnosis when mosaicism is suspected or a clinical decision requires it. The framework presented here is a tentative, practice-informed starting point intended to support—rather than replace—individualized multidisciplinary care; prospective multi-center validation is required before it can be adopted as an evidence-based clinical management algorithm.

Acknowledgments

We thank the patients and their families for their participation in this study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0264/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0264/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0264/prf

Funding: This work was supported by the General Project of the Science and Technology Research Program, Jiangxi Provincial Department of Education (No. GJJ210203).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0264/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Jiangxi Provincial Children’s Hospital (approval No. JXSETYY-YXKY-20250019). Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Berglund A, Viuff MH, Skakkebæk A, et al. Changes in the cohort composition of turner syndrome and severe non-diagnosis of Klinefelter, 47,XXX and 47,XYY syndrome: a nationwide cohort study. Orphanet J Rare Dis 2019;14:16. [Crossref] [PubMed]
2. National Organization for Rare Disorders (NORD). 47,XXX (Trisomy X) [Internet]. Last updated March 2025 [cited 2025 Mar]. Available online: https://rarediseases.org/rare-diseases/trisomy-x/
3. Sánchez XC, Montalbano S, Vaez M, et al. Associations of psychiatric disorders with sex chromosome aneuploidies in the Danish iPSYCH2015 dataset: a case-cohort study. Lancet Psychiatry 2023;10:129-38. [Crossref] [PubMed]
4. Tartaglia N, Davis S, Howell S, et al. Medical Findings in Infants Prenatally Identified With Sex Chromosome Trisomy in Year 1 of Life. Pediatrics 2025;156:e2024068133. [Crossref] [PubMed]
5. Tartaglia NR, Howell S, Sutherland A, et al. A review of trisomy X (47,XXX). Orphanet J Rare Dis 2010;5:8. [Crossref] [PubMed]
6. Carrel L, Willard HF. X-inactivation profile reveals extensive variability in X-linked gene expression in females. Nature 2005;434:400-4. [Crossref] [PubMed]
7. Wigby K, D’Epagnier C, Howell S, et al. Expanding the phenotype of Triple X syndrome: A comparison of prenatal versus postnatal diagnosis. Am J Med Genet A 2016;170:2870-81. [Crossref] [PubMed]
8. Freilinger P, Kliegel D, Hänig S, et al. Behavioral and psychological features in girls and women with triple-X syndrome. Am J Med Genet A 2018;176:2284-91. [Crossref] [PubMed]
9. Domes G, Croyé MA, Freilinger P, et al. Brain structure in triple X syndrome: regional gray matter volume and cortical thickness in adult women with 47,XXX karyotype. J Neurodev Disord 2025;17:18. [Crossref] [PubMed]
10. Green T, Flash S, Reiss AL. Sex differences in psychiatric disorders: what we can learn from sex chromosome aneuploidies. Neuropsychopharmacology 2019;44:9-21. [Crossref] [PubMed]
11. Berglund A, Stochholm K, Gravholt CH. The comorbidity landscape of 47,XXX syndrome: A nationwide epidemiologic study. Genet Med 2022;24:475-87. [Crossref] [PubMed]
12. Butnariu L, Rusu C, Caba L, et al. Genotype- phenotype correlation in trisomy X: a retrospective study of a selected group of 36 patients and review of literature. Rev Med Chir Soc Med Nat Iasi 2013;117:714-21.
13. Fukami M, Seki A, Ogata T. SHOX Haploinsufficiency as a Cause of Syndromic and Nonsyndromic Short Stature. Mol Syndromol 2016;7:3-11. [Crossref] [PubMed]
14. Binder G, Rappold GA. SHOX Deficiency Disorders. In: Adam MP, Bick S, Mirzaa GM, et al. editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; December 12, 2005.
15. Legue M, Staszewski M, Mastronardo M, et al. Sex chromosome aneuploidy impacts on human gene expression and regulation: a systematic review. Mol Med 2025;32:11. [Crossref] [PubMed]
16. Nielsen MM, Trolle C, Vang S, et al. Epigenetic and transcriptomic consequences of excess X-chromosome material in 47,XXX syndrome-A comparison with Turner syndrome and 46,XX females. Am J Med Genet C Semin Med Genet 2020;184:279-93. [Crossref] [PubMed]
17. Lei Y, Luo D, Yan X, et al. The mean age of menarche among Chinese schoolgirls declined by 6 months from 2005 to 2014. Acta Paediatr 2021;110:549-55. [Crossref] [PubMed]
18. SwensonKSBothwellSZhivotovADemographic Composition of Participants in Sex Chromosome Aneuploidy Studies across the Globe: a 20-year Systematic Review.medRxiv 2025;2025.06.17.25329799.
19. Davis SM, Soares K, Howell S, et al. Diminished Ovarian Reserve in Girls and Adolescents with Trisomy X Syndrome. Reprod Sci 2020;27:1985-91. [Crossref] [PubMed]
20. Otter M, Crins PML, Campforts BCM, et al. Social functioning and emotion recognition in adults with triple X syndrome. BJPsych Open 2021;7:e51. [Crossref] [PubMed]
21. Zampini L, Lorini A, Zanchi P, et al. Neurodevelopment in Young Children With Sex Chromosome Trisomies Diagnosed Before Birth: A Cluster Analysis Study. Am J Med Genet A 2025;197:e63895. [Crossref] [PubMed]
22. George S, Dagar V, Chakrabarty BK, et al. Unravelling the Impact of an Additional Sex Chromosome in an Adult Female. J Hum Reprod Sci 2023;16:166-9. [Crossref] [PubMed]
23. Singhal P, Singh S, Kumar P, et al. Premature Ovarian Failure Related to Trisomy X: Two Case Reports with an Aberrant 47, XXX Karyotype. J Hum Reprod Sci 2021;14:87-90. [Crossref] [PubMed]
24. Rogol AD. Sex chromosome aneuploidies and fertility: 47,XXY, 47,XYY, 47,XXX and 45,X/47,XXX. Endocr Connect 2023;12:e220440. [Crossref] [PubMed]
25. Le Gall J, Nizon M, Pichon O, et al. Sex chromosome aneuploidies and copy-number variants: a further explanation for neurodevelopmental prognosis variability? Eur J Hum Genet 2017;25:930-4. [Crossref] [PubMed]
26. Bouw N, Swaab H, Tartaglia N, et al. Early Social Behavior in Young Children with Sex Chromosome Trisomies (XXX, XXY, XYY): Profiles of Observed Social Interactions and Social Impairments Associated with Autism Spectrum Disorder (ASD). J Autism Dev Disord 2023;53:3194-207. [Crossref] [PubMed]
27. Wigby K, Cordeiro L, Wilson R, et al. Adaptive functioning in children and adolescents with Trisomy X: An exploratory analysis. Am J Med Genet C Semin Med Genet 2020;184:456-68. [Crossref] [PubMed]