Clinical practice guidelines for the care of girls and women with Turner syndrome: proceedings from the 2016 Cincinnati International Turner Syndrome Meeting

1.1. Diagnosis and genetics

R 1.1. We recommend considering a diagnosis of TS in phenotypic females with a karyotype containing one X chromosome and complete or partial absence of the second sex chromosome, associated with one or more typical clinical manifestations of TS (⨁⨁⨁⨁).

R 1.2. We recommend against considering a diagnosis of TS in females with one X chromosome and a deletion distal to Xq24 on the other X chromosome, and in women over the age of 50 years with less than 5% 45,X mosaicism (⨁⨁◯◯).

R 1.3. We suggest that the new general surveillance management guideline applies to TS patients with any karyotype (⨁⨁◯◯).

R 1.4. We recommend to consider testing for Turner syndrome (TS) in a female with typical signs (Table 2) (⨁⨁⨁⨁).

R 1.5. We recommend gonadectomy in all female individuals with Y chromosome material identified on standard karyotyping (⨁⨁◯◯).

Guideline development consensus working group

The guidelines project was initiated by the European Society of Endocrinology and the Pediatric Endocrine Society, in collaboration with the European Society for Paediatric Endocrinology, the Endocrine Society, European Society of Human Reproduction and Embryology, the American Heart Association, the Society for Endocrinology, and the European Society of Cardiology. The guideline has been formally endorsed by the European Society of Endocrinology, the Pediatric Endocrine Society, the European Society for Paediatric Endocrinology, the European Society of Human Reproduction and Embryology and the Endocrine Society.

These guidelines were sponsored primarily by ESE, and co-sponsored by PES, ESPE and ES. Furthermore, ESHRE, SfE, and ESC supported their own delegates for the meeting, and additional support was obtained from the AHA, the National Institute of Child Health and Human Development (NICHD grant #1R13HD089663), the National Center for Advancing Translational Sciences, Cincinnati Children’s Hospital Medical Center’s TS Foundation, as well as several advocacy groups (TS Society of the United States, the TS Global Alliance and the Turner Resource Network). The chairs of the consensus working group, Claus H Gravholt and Philippe F Backeljauw, were appointed by the ESE Clinical Committee and PES respectively. Other members of the working and writing group are Niels H Andersen (adult cardiologist), Gerard S Conway (adult endocrinologist), Olaf M Dekkers (epidemiologist), Mitchell E Geffner (pediatric endocrinologist), Angela E Lin (clinical geneticist), Nelly Mauras (pediatric endocrinologist), Karen O Klein (pediatric endocrinologist), Charmian A Quigley (pediatric endocrinologist), Karen Rubin (pediatric endocrinologist), David E Sandberg (clinical psychologist), Theo C J Sas (pediatric endocrinologist), Michael Silberbach (pediatric cardiologist), Viveca Soderstrom-Anttila (fertility specialist), Kirstine Stochholm (adult endocrinologist), Janielle A van Alfen-van der Velden (pediatric endocrinologist) and Joachim Woelfle (pediatric endocrinologist). The working group had two in-person meetings (September 2015, with some of the participants present, and July 2016, where all members were present). Consensus was reached upon discussion; minority positions were taken into account in the rationale behind the recommendations. Each working group included at least one member from Turner syndrome advocacy community. These individuals provided a valuable and nuanced perspective. All participants completed conflict-of-interest forms (Appendix 1, see section on supplementary data given at the end of this article).

A draft of the guideline was submitted for external review and commentary with/without endorsement by the professional societies. All comments and suggestions were discussed and implemented as appropriate by the working/writing group. Responses to the comments are summarized in Appendix 2.

Target group

This guideline document was developed for health care providers of patients with TS, i.e., both primary-care providers (pediatricians, family doctors, internal medicine specialists), as well as specialists, such as specialist pediatricians, geneticists, endocrinologists, cardiologists, fertility doctors and specialists in internal medicine.

Aims

The overall purpose of the guidelines is to provide a practical clinical guideline, with focus on operational recommendations for daily management. We also aimed to address those health care issues not addressed in the previous guidelines.

Summary of methods used for guideline development

The methods used have been described in more detail previously (3). In short, the guidelines used GRADE as a methodological base for four clinical questions. The first step was to define these question(s), followed by a systematic literature search. After including relevant articles, we (1) estimated an average effect for specific outcomes (if possible); and (2) rated the quality of the evidence. Formal evidence syntheses were performed and graded only for these questions.

For the GRADE questions we took into account: (1) quality of the evidence, (2) balance of desirable and undesirable outcomes, (3) values and preferences (patient preferences, goals for health, costs, management inconvenience, feasibility of implementation, etc.) (4). Additional recommendations based on good practice were graded based on expert opinion (5). Evidence tables are provided in Appendix 2 and Supplementary Fig. 1.

All other recommendations were derived from majority consensus of the guideline development working group, but, if members had substantive disagreements, this is acknowledged in the manuscript. For transparency, all recommendations provided are accompanied by a text explaining why specific recommendations were made. The recommendations are worded as recommend (strong recommendation) and suggest (weak recommendation). The quality of evidence behind the recommendations is classified as very low (⨁◯◯◯), low (⨁⨁◯◯), moderate (⨁⨁⨁◯) and strong (⨁⨁⨁⨁) (6). This approach was used for all other recommendations as well. For ‘all other classifications’ not formally submitted for GRADE, the recommendations were proposed by selected members of the working group and accepted by the remaining members of the working group, and as such represent good clinical practice based on the limited available evidence.

Clinical questions, endpoint definitions and eligibility criteria

The guideline panel formulated four clinical questions for which a separate systematic literature search was performed, and for which available evidence was synthesized. For each question, the eligibility criteria, endpoint definition, search strategy and main findings are described below.

What is the effect of growth-promoting treatment in TS? (GRADE question 1)

Short stature, present in most individuals with TS, is treated with GH, with/without oxandrolone (a non-aromatizable androgen), with the goals of increasing adult height. We systematically searched for randomized clinical trials (RCTs) published after 1990 on the effects of GH with or without the addition of oxandrolone. The following outcomes were considered: height, quality of life (QoL), mortality, cardiovascular side effects and masculinization (due to oxandrolone treatment). The following studies were not eligible: non-randomized studies, studies not reporting height, studies only comparing different doses of one drug and crossover trials.

What is the probability of achieving viable pregnancy after oocyte donation in TS? (GRADE question 2)

Usually, TS is accompanied by infertility due to premature ovarian insufficiency. Women with TS can be offered oocyte donation if they desire pregnancy. We searched for studies that reported on the probability of a live birth or viable pregnancy after oocyte donation in TS. Outcomes considered important were live-born children, risk of abortion and complications (e.g., pre-eclampsia and aorta dissection). We also searched for studies that compared the effectiveness of achieving a viable pregnancy with different protocols for oocyte donation.

What are effects of blood pressure treatment on clinical outcomes in TS? (GRADE question 3)

Turner syndrome is often accompanied by primary hypertension, which has been linked to the development of aortic dilation or even AoD – both observed with strikingly increased frequency in TS. Some experts have advocated for stricter blood pressure control in TS individuals. Therefore, two questions were formulated: (1) at what blood pressure threshold should hypertension in TS be treated and (2) what anti-hypertensive treatment is most effective in TS? We searched for studies comparing different blood pressure targets and different blood pressure treatments. Cardiovascular disease and mortality were considered relevant endpoints. Randomized as well as non-randomized studies were considered; cohort studies without control arm and case series were ineligible.

Estrogen replacement in TS (GRADE question 4)

Turner syndrome is usually accompanied by hypergonadotropic hypogonadism and primary or secondary amenorrhea. Most TS individuals will therefore need hormonal replacement therapy (HRT) – first for induction of puberty and later for maintaining secondary sex characteristics, attaining peak bone mass, normalizing uterine growth (for possible pregnancy later) and can today be offered oocyte donation. This leads to the following question: What is the optimal HRT, mainly focusing on dosing throughout adolescence and adulthood?

Description of search and selection of literature

In cooperation with a trained librarian, a search strategy was composed for all four clinical questions. The following databases were searched: PubMed, EMBASE (OVID version), Web of Science, COCHRANE Library, CINAHL, Academic Search Premier and ScienceDirect. The number of articles retrieved, exclusion of articles and final inclusion of eligible articles are shown in the flowchart (Supplementary Fig. 1).

1.1. Diagnosis and genetics of Turner syndrome

R 1.1. We recommend considering a diagnosis of TS in phenotypic females with a karyotype containing one X chromosome and complete or partial absence of the second sex chromosome, associated with one or more typical clinical manifestations of TS (⨁⨁⨁⨁).

R 1.2. We recommend against considering a diagnosis of TS in females with one X chromosome and a deletion distal to Xq24 on the other X chromosome, and in women over the age of 50 years with less than 5% 45,X mosaicism (⨁⨁◯◯).

R 1.3. We suggest that the new general surveillance management guideline apply to TS patients with any karyotype (⨁⨁◯◯).

1.2. Prenatal diagnosis

Sex chromosome abnormalities, in general, can be detected prenatally by chorionic villous sampling or amniocentesis, whether performed for advanced maternal age, abnormal serum markers or the presence of multiple anomalies. Regardless of the indication, test procedure or specific result, genetic counseling by a geneticist or genetic counselor should be provided before and after any prenatal diagnostic procedure. Ultrasonography can play an important role that suggests an increased likelihood of TS. In the first trimester, increased nuchal translucency is common in fetuses with TS, but is also observed in the autosomal trisomy syndromes. The presence of a frank cystic hygroma, however, makes the diagnosis of TS more likely (39). Other ultrasound findings suggestive of TS are coarctation of the aorta and/or left-sided cardiac defects (which should be further imaged by fetal echocardiography), brachycephaly, renal anomalies, polyhydramnios, oligohydramnios and growth retardation (40). Abnormal triple or quadruple maternal serum screening (alpha-fetoprotein, human chorionic gonadotropin, inhibin A and unconjugated estriol) may also suggest the diagnosis of TS (41), but these tests may be perfectly normal together with normal nuchal fold thickness (42). Ultrasound and maternal serum screening are not diagnostic, and karyotype confirmation of TS is obligatory.

Because an amniocentesis karyotype reflects fibroblast analysis, the postnatal outcome and constitutional karyotype of individuals with prenatally diagnosed 45,X are uncertain, especially in patients with mosaicism. Therefore, chromosome analysis should be repeated postnatally in all patients. In general, any of the features of TS may be seen with virtually any of the common chromosome constitutions (2, 10). Although non-mosaic 45,X fetuses with pleural effusion or cystic hygroma often spontaneously abort, these findings are also compatible with delivery of a viable newborn (43).

The recent developments in non-invasive prenatal testing (NIPT) have had implications for diagnosing TS (44). When TS is suspected, the possible interpretations may include TS in the fetus or TS in the mother. Additional testing is offered such as maternal karyotype, amniocentesis, fetal echocardiography and genetic counseling. There is insufficient evidence to recommend NIPT for TS by sequencing or SNP (single-nucleotide polymorphism) array analysis of cell-free fetal DNA (cfDNA) in maternal blood. A recent meta-analysis of 37 studies showed that the detection rate (90%) and positive predictive value (23%) of cfDNA analysis were relatively low (45, 46). In such cases, the investigation of choice continues to be invasive testing for fetal karyotype evaluation.

Pre-implantation genetic screening (PGS) is being used with increased frequency in in vitro fertilization (IVF) programs, and pregnancy rates continue to be studied (47, 48). Such PGS is currently offered to women with recurrent pregnancy loss or repetitive implantation failure. It is also being used as a method of embryo selection to increase pregnancy rates in all women undergoing IVF. Embryos now are being diagnosed with TS, which means that pre-implantation has become an additional time for diagnosis and counseling. If used more extensively in the future, this could have a significant impact on the prevalence of TS.

Decisions regarding pregnancy termination are difficult, and it is critical that the best available information is provided to parents. Physicians and genetic counselors involved in pre- and post-diagnostic counseling need to be fully informed about the prognosis, complications and QoL of individuals affected with TS, as well as of recent advances in management (49). The input of a physician with experience in the long-term follow-up of TS patients will be valuable to put management of the different comorbidities in perspective. The discussion should include variability of features, the possibility of short stature and ovarian failure and their management. It should be emphasized that most individuals with TS have intelligence scores in the normal range, although many specific types of learning disabilities and psychological challenges are common. Of course, the discussion should be tailored to the specific findings of that fetus, because decisions regarding termination are often influenced by the presence and severity of an abnormal phenotype (50, 51). Discussion with families of girls and women with TS can be very helpful. This is often accomplished through contact with TS support organizations.

1.3. Postnatal diagnosis

All individuals with suspected TS should have a standard 20-cell karyotype as recommended by the American College of Medical Genetics, as this will identify at least 10% mosaicism with 95% confidence in the blood (52, 53, 54). If mosaicism is strongly suspected, but not demonstrated with standard karyotype, additional metaphases may be counted or FISH studies performed (26, 52). Although a peripheral blood karyotype is usually adequate, a second tissue, such as skin fibroblasts, buccal mucosa cells or possibly urine for bladder epithelial cells, may be examined if there is a strong clinical suspicion of TS despite a normal blood karyotype or low-level mosaicism.

There are several situations in which a formal karyotype should be repeated: (1) all infants diagnosed prenatally should have a postnatal karyotype to confirm the findings, (2) those diagnosed by a buccal swab only, (3) individuals with a karyotype performed in the distant past or (4) when no original report is available for review.

R 1.5. We recommend gonadectomy in all female individuals with Y chromosome material identified on standard karyotyping (⨁⨁◯◯).

An increased prevalence of germ cell tumors has been noted in TS individuals with Y chromosome sequences. The risk of malignancy is determined by the presence of Y chromosome material and mosaicism (genotype), and the degree of masculinization (phenotype). At least 10% of females with TS harbor Y chromosome sequences. With the availability of newer molecular methods, the detection rate of Y-chromosomal material has increased. Real-time polymerase chain reaction (PCR) with multiple Y-specific probes is more sensitive than FISH and can detect Y mosaicism in up to 12% of TS cases (55, 56, 57, 58). FISH with X and Y centromere probes are recommended to determine the origin of ring or small marker chromosomes. We propose that it is no longer clinically indicated to use FISH with SRY probes to exclude cryptic Y material. The gonadoblastoma locus was mapped adjacent to the centromere to a region that does not include the SRY gene; therefore, SRY probes are not required (59, 60). When virilization is present, it remains essential to test at least two to three tissues to search for cryptic Y material; FISH of buccal cells may detect Y mosaicism that is not detectable in peripheral blood (61, 62).

The rate of gonadoblastoma among TS patients with Y chromosome sequences that were detected by PCR or FISH varied from 4 to 60% in 14 studies, and data on the long-term outcome of this cohort are incomplete. Putting all data together, approximately 10% may develop a gonadoblastoma, although there is considerable variation in risk estimates, possibly related to methodology, sample size and potential selection bias (56, 63). The rate of gonadoblastoma among all TS patients, including those without Y chromosome sequences, is low (approximately 1%). For these reasons, molecular screening to detect Y-chromosomal sequences is currently recommended only in TS individuals with masculine features who are negative for Y material by conventional cytogenetic and FISH analyses. In these individuals, multiple sequences adjacent to the Y centromere should be amplified using PCR.

Chromosomal microarray analysis has revolutionized the study of people with growth disorders, unusual appearance, multiple anomalies and/or neuropsychological differences, but its role in TS is currently considered to be supplemental. X chromosome structural classifications of 187 TS patients using SNP microarray genotypes were comparable to conventional cytogenetics in more than 90% of the cases, and identified two derivative Y chromosomes and 13 large copy-number variants that were not identified by karyotyping (64). Long-term studies are needed to determine the prognostic utility of microarray data before these methods are incorporated into routine clinical practice.

Based on the observation that phenotypical features in TS subjects vary greatly, even when focusing on women with monosomy X, an imprinting effect has been proposed as a potential explanation for the observed variance (65). Most of the studies analyzing parent-of-origin effects in TS, report a predominance of the maternal X, largely due to the non-viability of the karyotype 45,Y and a slight preferential loss of paternal sex chromosomes (66, 67, 68, 69, 70, 71). Several studies analyzed whether X-linked imprinting effects might be associated with distinct phenotypical features, with conflicting results (72, 73, 74). Lack of statistical power to detect subtle differences explains part of the observed heterogeneity (Table 4).

Based on the currently available data, it seems premature to conclude whether the paternal origin of the remaining X might exert an impact on height, growth (response) or the frequency of organ anomalies in TS. Regarding central nervous system (CNS) morphology and function, including certain neuropsychological features, most studies described the differences between individuals who inherited a maternal or paternal X. Some studies described a more severe phenotype in subjects with a maternally derived X chromosome (75), although another study found no convincing effect of genomic imprinting on neurocognitive/socialization function, but rather reported a subtle difference in visual perceptual reasoning with lower scores in subjects with a paternal X chromosome (76).

From a clinical perspective, a genetic work-up to detect the parental origin of the remaining X is currently not indicated in routine care of women with TS.

1.4. Newborn screening

Missed and delayed diagnoses of TS remain a major problem. For girls with TS who are not identified in infancy with lymphedema and webbed neck, diagnosis is often made years after growth failure is apparent, and sometimes when little or no growth potential remains (34, 35, 36, 37, 77). In general, the later GH therapy is initiated, the greater the height deficit and the lower the likelihood of normal adult stature and age-appropriate initiation of therapies for pubertal development. Early diagnosis allows for timely screening and intervention for problems such as strabismus, hearing loss, renal and cardiac abnormalities, hypothyroidism, celiac disease and learning disabilities, thus improving QoL. It may also improve fertility in some individuals with TS by allowing for oocyte or ovarian tissue harvesting before the death of too many follicles. Earlier diagnosis of TS will require greater recognition of the disorder through education and/or population screening.

Ideally, screening for TS would be part of existing newborn screening programs. Karyotyping, considered to be the gold-standard technique for diagnosing TS, has major limitations as a screening tool given its long processing time, high cost and requirement for specialized personnel. However, several molecular methods have been proposed for neonatal screening of TS, of which pyrosequencing and real-time PCR appear to hold the greatest promise (78). Advances in pyrosequencing technology have been rapid, but it remains expensive. In contrast, a recent study estimated that each real-time PCR test for TS detection costs $15 US. All but one patient with TS was detected (albeit only 10 patients with mosaicism were tested) for a detection sensitivity of 95%, and only 0.6% of the newborns required recall for karyotypes (79). Most recently, whole-exome sequencing was shown to accurately diagnose TS, including cases with low-level mosaicism, isochromosome Xq and cryptic Y material (80).

If molecular screening for TS is offered, positive findings will need karyotype confirmation. Like all other disorders diagnosed on newborn screening, infrastructure for follow-up, treatment and support of the newborns diagnosed will need to be developed.

Possible drawbacks to screening include the likelihood that some girls identified with TS will be phenotypically normal, experience minor or no clinical consequences, and, thus, may be needlessly concerned. Other sex chromosome abnormalities, such as Klinefelter syndrome (if both phenotypic males and females are tested), may also be diagnosed and will need appropriate follow-up.

In the United States, nomination of a condition to the Recommended Uniform Newborn Screening Panel requires a high certainty that screening for the targeted condition would lead to a significant net benefit, that screening has high-to-moderate feasibility and that most state screening programs would be able to implement screening within 3 years (81). Studies are needed to quantitate the benefits of newborn diagnoses of TS and the feasibility of molecular techniques for diagnosis needs to be established. We conclude that newborn screening for TS should be considered after additional improvements in methodology.

2.1. Growth-promoting therapies

The goals of growth-promoting therapies are to facilitate attainment of heights during childhood and adulthood that minimize physical restrictions and allow puberty to begin at an age similar to peers. The centerpiece of growth-promoting therapy is GH, which increases HV and results in modest increases in adult stature for most patients (82). However, as most girls will require estrogen replacement to either initiate or complete puberty prior to completion of linear growth, the estrogen route, dosage and escalation tempo will have an impact on pubertal growth (see section on estrogen supplementation) and, therefore, on adult height (AH). AH within the lower range for female population standards is the likely outcome for most women at completion of growth-promoting therapy.

2.2. Efficacy of GH treatment

Despite the plethora of studies of GH treatment in TS, a 2007 Cochrane Center review (82) identified only four trials in which GH treatment was compared in a randomized fashion with a concurrent non-treatment or placebo control for at least 1 year (83, 84, 85, 86), and only a single RCT that followed participants to AH (83). More recently, a 2-year RCT evaluating the impact of GH initiation before age 4 years, and a double-blind, placebo-controlled trial to AH have been published (87, 88). Patients with TS in the North American GH registration trials treated to (near-)AH had average gains vs concurrent control (83), baseline predicted/projected height or historical controls ranging from about 5 to 8 cm over periods ranging from 5.5 to 7.6 years (89, 90, 91); similar results were observed in the more recent placebo-controlled trial (87). These data indicate that height gain of about 1 cm per year of GH therapy is a reasonable expectation. Two European studies using high GH doses at young ages have demonstrated much more dramatic gains of 15–17 cm (mean) vs baseline projected AH (92, 93, 94). A formal GRADE evaluation for the present guideline included 14 randomized studies on the effect of growth-promoting therapies in TS (both GH and oxandrolone) (Appendix 2). Studies were published between 1991 and 2011. Studies displayed large heterogeneity with respect to design, included patients, treatment schedules (dose and duration), follow-up duration and reported endpoints. There is a considerable risk of bias in the included studies. Only five studies were adequately blinded, while the remaining were not. Long-term data were often characterized by considerable loss to follow-up. For example, long-term effects were only studied in approximately 62% of the original study population (95). Such post-randomization loss to follow-up may introduce selection bias – even in a RCT. In addition, the use of a run-in period to exclude non-compliance (96) may hamper the generalizability of the results. In almost all studies, the primary endpoint was not well defined. Finally, many small studies were underpowered (97), and no study reported on mortality.

Most studies reported on HV, expressed as either cm/year or change in height SDS. In studies that enabled a comparison of GH to no treatment or placebo (83, 87, 88, 89), the GH-treated group had a greater growth rates. Additionally, in studies comparing the added value of oxandrolone therapy to GH, this addition of oxandrolone resulted in a greater growth rate (95, 96, 97, 98, 99). AH was taller in GH-treated compared to non-GH-treated patients, with reported differences ranging from 0.5 to 1.5 SDS (83, 87, 88).

If catch-up growth to within the normal range occurred within the first two years of treatment, HV was maintained close to the mean for age, and there was adequate pubertal growth, and then AH within the lower normal range (above about 152 cm or 60 inches) was observed. However, comparison with published ranges for HV and height SD score may be unreliable during the pubertal age range because girls with TS lack the pubertal growth spurt resulting from the synergistic actions of endogenous GH and estrogen. During this period, GH-treated girls with TS may lose ground when height SD score is calculated according to general standards, but height SD score should still continue to increase compared with TS standards.

2.3. Concomitant treatment with the anabolic steroid oxandrolone

R 2.6. We suggest concomitant treatment with oxandrolone from the age of 10 years or older at 0.03 mg/kg/day and maintained below 0.05 mg/kg/day, if the diagnosis of TS (and therefore GH treatment initiation) is delayed, and/or AH outcome is likely to be unsatisfactory with the standard GH dose alone (⨁⨁⨁⨁).

Although well-controlled studies have demonstrated modest synergistic increases in growth response by addition of oxandrolone during treatment with GH (95, 97, 139, 140, 141, 142), the potential for the unwanted effects of delayed breast development and dose-dependent virilization (e.g., clitoromegaly, voice deepening, hirsutism and acne) prompts the need for caution in the use of this therapy. If the decision is made to add oxandrolone, this should not be done until around age 10 years, initiated at a dose of 0.03 mg/kg/day and maintained at no greater than 0.05 mg/kg/day. The GRADE evaluation showed further that AH was approximately 2–5 cm higher in oxandrolone- and GH-treated patients compared to GH-treated only (95, 139, 140). Thus, existing literature show a growth effect of addition of oxandrolone to GH. In one study, virilization was more often reported in the oxandrolone and GH group compared to the GH only group (16 vs 5%) (139). Heterogeneity between studies prevents performing a formal meta-analysis and prohibits a final evidence-based verdict on the optimal growth-promoting treatment schedule in TS.

2.4. Concomitant treatment with childhood ultra-low-dose estrogen

R 2.7. We suggest to not routinely add very-low-dose estrogen supplementation in the prepubertal years to further promote growth (⨁⨁◯◯).

One double-blind, placebo-controlled trial using ultra-low-dose oral ethinylestradiol as a growth-promoting agent during the prepubertal period combined with GH demonstrated a modest synergistic increase in AH, normalization of the timing of thelarche for about one-quarter of the girls, and modest improvements in cognition and memory within specific developmental windows (87, 143, 144, 145).

However, no other independent trials have evaluated this approach, the dosing and administration of childhood estrogen have not been optimized, and long-term safety has not been assessed. The addition of very-low-dose estrogen replacement as a growth-promoting therapy is currently not recommended.

2.5. Sex hormone replacement

R 2.8. We recommend that estrogen replacement should start between 11 and 12 years of age increasing to adult dosing over 2–3 years (⨁⨁⨁◯).

R 2.9. We suggest that low-dose estradiol (E2) is the preferred estrogen and that it be administered by a systemic route and that the transdermal route is preferred (⨁◯◯◯).

R 2.10. We recommend adding progesterone once breakthrough bleeding occurs or after 2 years of estrogen treatment (⨁⨁⨁⨁).

Turner syndrome is usually accompanied by hypergonadotropic hypogonadism and primary or secondary amenorrhea due to gonadal dysgenesis. Approximately one-third of girls with TS have spontaneous thelarche, occurring most often in girls with mosaicism (146, 147, 148). Regular menstrual cycles occur in at most 6% of these subjects (148).

Most patients with TS will therefore need HRT – for induction of puberty and for maintaining female secondary sex characteristics, attaining peak bone mass and normalizing uterine growth.

The optimal estrogen replacement therapy regimen to induce pubertal development is still being evaluated, but, since the previous guidelines (2), studies suggest use of transdermal (TD) preparations as the preferred choice. However, there is no study to date of TD use from initiation of puberty until adulthood. Therefore, we present data in support of TD use based on available studies and theoretical considerations. Theoretical reasons include: a more physiologic route of delivery without the first-pass effect through the liver, thereby avoiding the accumulation of non-physiologic estrogens observed after the oral route (149). The latter route is associated with a pro-coagulable state (150) and increased risk of stroke in the postmenopausal setting (151).

Estradiol (E2) is normally secreted into the systemic circulation; thus, the liver receives the same concentration as other somatic tissues, and a systemic route of replacement estrogen delivery is physiologic (152). In contrast, estrogen given orally reaches the systemic circulation only after absorption into the portal venous system and after metabolism by the liver, thus exposing the liver to a greater amount of estrogen than the rest of the body. Orally administered estrogens are metabolized by the liver to estrone and other metabolites before reaching the systemic circulation. Thus, delivery of E2 into the systemic circulation by TD, transvaginal or parenteral administration more closely mimics normal physiology. A transvaginal route is not recommended for prepubertal girls, and parenteral administration (injection) is not preferred by most patients when a TD option is available.

Timing, route and dose of some recommended estrogen replacement options are depicted in Table 5. The goals of replacement are to mimic the progression of puberty in an average girl and minimize risks. The initiation dose is based on an average progression of puberty and protection of growth potential. Delaying estrogen replacement may be deleterious to bone and uterine health, as observed in anorexia nervosa. It is important to remember that the recommendation is for estrogen replacement of a deficient state, not supplementation of endogenous hormones. Adult dosing is also based on uterine and bone health, as well as symptoms of hypogonadism.

To mimic normal physical and social development, treatment should begin at 11–12 years, as long as gonadotropins are elevated. Luteinizing hormone (LH) and FSH may be measured annually starting at age 11 years or earlier. If gonadotropins are normal for age, observation for spontaneous puberty is appropriate, with future replacement therapy if gonadal failure occurs. Low anti-Müllerian hormone (AMH) and undetectable inhibin B can also be used to predict ovarian failure in TS (153, 154, 155). Using low doses of estrogen is crucial to preserve height potential whether or not GH treatment has already been initiated, because very-low-dose ethinylestradiol (EE), TD E2, oral E2 and systemic (depot) E2 do not interfere with the growth response to GH therapy (87, 94, 152, 156, 157). It should be noted that EE doses ≥3 µg/day administered prior to 14 years may decrease the growth response (140, 158). Therefore, the lower-dose E2 preparations are preferable for initiation of puberty.

Incremental dose increases can occur approximately every 6 months to mimic the normal pubertal tempo until adult dosing is reached over a 2- to 3-year period. This theoretically translates into a 25–100% increase in dose every 6 months using 4–6 dose changes between the initiation and adult doses portrayed in Table 5. However, no studies to date have rigorously studied outcomes in relation to the rate of dose increase for the different preparations. Routine monitoring of serum LH or FSH is not recommended during estrogen treatment as concentrations remain elevated in hypergonadotropic hypogonadal women until higher doses of estrogen are given. Estradiol measurement using an ultrasensitive assay allows titrating dosage, if desired, though E2 concentrations for optimal growth remain to be determined (149). Clinical assessment, patient satisfaction, patient age and often residual growth potential are the primary determinants for a dose increase. If potential for taller stature is still possible, girls may remain on lower estrogen doses longer. If girls are already older at initiation, the duration of time until adult dosing may be shortened.

Studies of the regimens report onset of breast buds within 6 months in most girls (94, 159, 160, 161, 162). The starting doses for depot E2 lead to slightly slower onset and for oral E2 slightly faster onset (156, 159, 160). Each of these regimens results in Tanner stage 4 breasts in approximately 2.25 year, which is similar to the tempo seen in TS girls with spontaneous puberty (1.9 year) (158) and in girls in general (161).

Girls with TS typically have a normal uterus, and progestin must be added once breakthrough bleeding occurs, or after 2 years of treatment, due to the risk of endometrial cancer associated with prolonged unopposed estrogen (163). In adult women, micronized crystalline progesterone, e.g. Prometrium (100–200 mg) or medroxyprogesterone, is preferred based on decreased breast cancer risk, but the data for this are relatively weak (164, 165). However, the risk of breast cancer is low in TS (166, 167, 168), and long-term treatment with HRT does not seem to induce breast cancer (169). Progestin can be added for 10 days each month for withdrawal bleeding, and adult women with TS should preferably continue treatment with 17β-estradiol in combination with a sequential progestational agent. A recent publication showed no increased risk of stroke with progesterone and pregnane- or nortestosterone-derived progestins. However, norpregnane derivatives were found to increase this risk (151). Studies have not been done in TS comparing various progestin options.

Once adult replacement doses are reached, treatment should persist until the risk of continuation outweighs the benefits, around the average age of menopause. Assessment of risk must be individualized for each woman (164, 170). Screening for thromboembolic risk should be performed in girls with a personal or family history. However, routine screening is not recommended, and screening is done only to educate the family on risks, not to postpone estrogen therapy (171).

A randomized controlled trial showed that early, very low-dose, depot E2 monthly injections initiated normal pubertal growth and development in conjunction with GH treatment (156). Transdermal E2 is likely to be preferred by girls over depot injections, although these two modalities have not been directly compared. A fractionated patch dose (one-quarter of a 25-µg patch or about 6.25 µg) applied overnight mimics the normal early morning serum E2 peak (162). Although not recommended by the product labels, two studies report pubertal E2 concentrations and/or breast development with use of a fractionated patch, and include practical guidelines for cutting the patch and application (162, 172). There is also a published opinion that cyclical administration of patches, commencing with the application of a 14–25-µg patch, the lowest doses commercially available, for 1 week monthly, may achieve similar results (156). Because the effect of the weeks without estrogen exposure is unknown, further data are needed before recommendations can be made regarding the best approach for patch application.

The GRADE evaluation of HRT found seven randomized studies on the differences between oral and TD hormone replacement therapies in TS. Studies were published between 1995 and 2013. Two studies (173, 174) were excluded because they did not report sufficient data on differences between routes of administration. Included studies displayed large heterogeneity with respect to design, included patients, choice of treatment (only 3 studies compared oral vs TD E2), treatment schedules (dose and duration), follow-up duration and reported endpoints. Overall, there is a considerable risk of bias in included studies. All studies were non-blinded and several studies included a small number of subjects. Details of the included studies are shown in Supplementary Fig. 1, the GRADE scoring per outcome and evidence synthesis is provided in Appendix 3.

Five studies reported data on HDL cholesterol concentrations (149, 160, 172, 175, 176). TD E2 showed a decrease in HDL plasma concentrations compared to oral estrogen as shown in Appendix 4 (weighted mean difference −8.09; 95% CI −12.7, −3.5). The plasma concentrations of total cholesterol, LDL cholesterol and TG were not different across routes of administration.

In two studies, bone mineral density was studied (149, 160). Both studies showed a somewhat better lumbar Z-score after TD use compared with oral administration (weighted mean difference 0.93; 95% CI 0.68, 1.19). However, sample sizes were too small and follow-up was too short to draw meaningful conclusions regarding bone density.

The type and route of estrogen administration did not affect glucose or insulin concentrations in any study.

Additionally, we included one randomized study that compared low-dose estrogen replacement therapy with high-dose therapy, one study on the effect of timing of administration (morning or evening administration) and one randomized study on the difference between fixed and individualized dose of estrogen therapy (177, 178, 179).

With regard to the dose of estrogen administration, one study compared conjugated and ethinylestradiol and found that some parameters were affected by either low or high dose (hypotrophic endometria, insulin, PTH, 1,25-dihydroxyvitamin D, liver enzymes), some were unaffected by either of them (urinary deoxypyridinoline cross-links) and some required high-dose preparation (FSH, alkaline phosphatase, osteocalcin) (177). A RCT study on the effect of dose of administration of E2 on bone mineral density or anthropometric markers did not find differences during 5 years, although bone-specific alkaline phosphatase was significantly higher in the high dose compared with the low-dose estrogen group (180). A RCT concluded that a low fixed dose of estrogen produces a satisfactory pubertal development not inferior to an individualized dose (178).

Very little evidence is available on the timing of administration. Based only on one RCT, evening administration of oral estrogen together with evening injections of GH may be preferable compared with morning administration (179).

Only one study has directly compared the route of administration in adolescence, oral vs TD, using E2 in 40 girls with TS followed for 1 year (149). The study found no differences in body composition, bone mineralization or plasma lipids when the plasma E2 concentrations were titrated to those of normally menstruating adolescents in both groups. However, oral estrogen was associated with a marked increase in conjugated estrogen precursors, such as estrone sulfate and serum estrogenic bioactivity. This is concerning in the context of the increased thromboembolic risk observed with oral estrogen in epidemiological studies (181, 182, 183).

Current evidence does not indicate that the hepatic effects on lipids or binding proteins cause an appreciable clinical difference between low-dose oral estrogens vs low-dose TD E2. There were no significant differences in glucose (149), insulin tolerance (177, 184), fasting insulin concentration, protein turnover, lipolysis (175), osteocalcin, highly sensitive C-reactive protein, BMI or waist-to-hip ratio (184, 185) between groups with TD vs oral estrogen treatment. Glucagon and insulin levels (during an OGTT) as well as insulin resistance tended to be lower following evening oral E2 administration (0.3–0.5 mg/day) (179). With the exception of one study reporting significantly higher HDL cholesterol after oral E2 (172), there were no significant differences in lipids between groups with different routes of estrogen administration (149, 175, 176, 177). There was also no evidence of liver toxicity from estrogen replacement therapy (186). Liver enzymes were elevated in untreated TS (174, 187, 188, 189) and reduced by exogenous estrogen-progestin administered orally or via the TD route (190). Replacement therapy with E2, by either oral or TD routes, lowers blood pressure (191, 192, 193), although E2 causes salt and water retention (194). This contrasts with EE-containing contraceptives, which, unless containing an anti-mineralocorticoid progestin, raise blood pressure significantly (195). Several studies looking at oral conjugated estrogens (CEE) vs TD E2 replacement in the postmenopausal setting have shown increased thromboembolic risk, especially in the first year of treatment in the oral group, and more pronounced in women with existing risk factors such as obesity (181, 182, 183).

EE exerts dose-related suppression of IGF-I in GH-naïve patients (196, 197). However, data concerning the influence of different routes of estrogen therapy on IGF-I concentration in GH-treated subjects are inconsistent, with some suggesting no effect and others suggesting suppression. Dose–response studies are lacking for most forms of estrogen (149, 172, 174, 175, 179).

Transdermal E2 administration (25–37.5 µg/day) has been reported as better than CEE (0.3–0.45 mg/day) for spine bone mineral density (BMD) in one study of TS girls (160), while no differences were observed in another study in which TD and oral E2 doses were titrated to similar concentrations (149).

Data concerning the influence of different routes of estrogen therapy on uterine volume are still inconclusive as route, dose, age at onset of treatment and duration of treatment all influence uterine growth. One study showed that EE treatment regimen gives rise to satisfactory pubertal induction and maintenance, but 20–30 µg daily failed to induce a fully mature uterus in 50% of the girls (198). Most studies reported that uterine volume was not affected by the type of estrogen used (199, 200), but is related to dose and duration of therapy (201, 202) (see Fertility section for optimal dosing). Although there have been no studies in children, we suggest against CEE in view of thromboembolic and cardiovascular disease risks reported in postmenopausal women (150).

Low-dose oxandrolone (0.03–0.05 mg/kg/day; maximum 2.5 mg/day) may modestly slow pubertal progression in response to estrogen replacement, delay menarche and increase clitoral size slightly, but these effects are minor and/or transient. A reasonable suggestion is that adjunctive treatment with oxandrolone be considered in very short TS girls, as suggested above, and until estrogen therapy is initiated (95, 139, 141, 158, 203).

The young women with TS who reached normal height and had age-appropriate pubertal development reported normal health-related quality of life (HRQoL); satisfaction with breast development (and height) had a positive influence on several HRQoL scales (204). Puberty should be induced at a physiologically appropriate age to optimize self-esteem, social adjustment and initiation of the patient's sex life (205, 206). One study showed that both estrogen use and age of puberty did not influence sexual function in patients with TS (207). Available data indicate that, on average, adult women with TS demonstrate characteristic neurocognitive profiles despite preserved ovarian function or adequate estrogen replacement (208).

3.1. Spontaneous pregnancies

R 3.1. We recommend counseling females with TS that their probability to conceive spontaneously decrease rapidly with age, if at all present, and consideration should be given to offering fertility treatment at a young age (⨁⨁⨁⨁).

Only few TS women will be able to get pregnant with their own eggs, and the few TS individuals that are fertile often enter menopause earlier than normal women. Spontaneous pregnancies occur in 4.8–7.6% of women with TS (18, 28, 209), but the frequency of miscarriages after spontaneous pregnancy was reported to be high: 30.8–45.1% (Supplementary Table 1) (18, 28). A study of 160 spontaneous pregnancies in 74 TS women found that, of the 58% resulting in a live birth, 34% were complicated by a fetal malformation. Of these, two-thirds were TS or Down syndrome, and the remaining etiologies were multifactorial (210). The risk rate of early pregnancy loss in the general population is 8–20% (211).

In a retrospective study of 115 women with a TS karyotype, obstetrical outcomes were compared with those of the general population (212). Most of the pregnancies were conceived naturally, and the TS diagnosis was unknown in 52% at the time of pregnancy. The karyotype of 10 women was 45,X. Pre-eclampsia occurred in 6.3% of women with TS compared to 3.0% in the control group, and AoD occurred in one TS woman (213). Compared with the reference group, women with a TS karyotype gave birth to children at an earlier gestational age and with a lower median birth weight. The cesarean section rate was 35.6% in women with TS karyotype vs 11.8% in the control group. The rate of birth defects did not differ between the two groups (212).

3.2. Assisted reproductive technologies (ART) with autologous oocytes

R 3.2. We suggest that young mosaic TS women with persistent ovarian function should be counseled that oocyte cryopreservation after controlled ovarian hyperstimulation is a possible fertility preservation option (⨁◯◯◯).

R 3.3. We recommend against routine oocyte retrieval for fertility preservation of young TS girls before the age of 12 years (⨁◯◯◯).

In the only published study to date of IVF stimulation in women with TS, Doger et al. reported outcomes of standard IVF in a total of 35 cycles in 22 women with mosaic TS. The clinical pregnancy rate was 8.6%, and live birth rate was 5.7% (214). It is well established that decreased ovarian reserve results in lower IVF pregnancy rates (215, 216). Given that TS women have rapidly decreasing ovarian reserve from a very young age, it is vital to counsel that their chance of pregnancy with autologous oocytes decreases rapidly and that consideration should be given to offering fertility treatment (including standard IVF) at a young age and without unnecessary delay.

3.3. Assisted reproductive technologies with oocyte donation (OD)

R 3.4. We recommend considering OD for fertility, only after thorough screening and appropriate counseling (⨁⨁⨁⨁).

R 3.5. We recommend that management of pregnant women with TS should be undertaken by a multidisciplinary team including maternal–fetal medicine specialists and cardiologists with expertise in managing women with TS (⨁⨁⨁◯).

For most patients with TS, OD is the only way to achieve a viable pregnancy. There have been reports of clinical pregnancies in oocyte recipients with TS since 1990, but the number of patients treated has been small. After GRADE evaluation, 11 studies were included based on eligibility criteria and endpoint definition. Sample size of TS women per study varied from 3 to 30 subjects, totaling 179 patients with either monosomy (45,X) or other variants such as mosaicism with 45,X in one cell line and X deletions. Studies were published between 1990 and 2011. Risk of bias was moderate, mainly due to small sample sizes and the potential of selection bias (i.e. inclusion of patients because of specifically positive outcomes).

The clinical pregnancy rate varied between 16 and 40% (Supplementary Table 2), like other recipients of donated oocytes (18, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227). The pooled proportion (random effects model) of clinical pregnancy per embryo transfer was 28% (95% CI 23–34%) (Supplementary Fig. 2).

Initial studies found high rates of early pregnancy loss in TS females ranging from 16 to 80% (219, 220, 221, 222, 223, 224), likely related to inadequate HRT (228). A structurally abnormal uterus, autoimmune mechanisms and/or diminished endometrial receptivity due to long-term hypoestrogenism may also increase spontaneous abortion rates in women with TS (229).

Previous studies have found that women with TS treated with oral or TD E2 or CEE may attain a normal mature uterine size and configuration (230). Estradiol-based regimens result in better uterine development than contraceptive-based regimens (199).

Women with TS who conceive with OD are at high risk for obstetrical complications (Supplementary Table 3). In TS, the incidence of hypertensive disorders of pregnancy has been reported to be 35–67% in singleton pregnancies and as high as 100% in multiple pregnancies (222, 224, 225, 226, 231, 232, 233). The reported incidence of pre-eclampsia ranged between 18% and 44% in singleton pregnancies and 75–100% in multiples, while the cesarean section rate is as high as 82–100% (222, 224, 225, 226, 231, 232, 233). In comparison, in oocyte recipients without TS, the rate of hypertensive disorders of pregnancy is 13–39%, and the cesarean section rate is 31–85% in singleton pregnancies (234).

In singleton pregnancies in TS, the risk of prematurity has been reported to be 9–33% and that of low birth weight 12–40% (Supplementary Table 4). In many studies, the outcomes of singleton and multiple pregnancies are not reported separately, which make their interpretation difficult. The perinatal mortality rate has been reported between 0 and 11% in published studies (225, 226, 231, 232, 233).

A large Nordic cohort study of 106 women with TS who gave birth after treatment with OD between 1992 and 2011 (233) reported 122 deliveries and a multiple delivery rate of 7.4%. In this cohort, the 45,X karyotype was observed in 44% of the recipients. Ten women had a known cardiac defect before pregnancy, 49% had a pre-pregnancy cardiac evaluation within two years before pregnancy and only 29% had an echocardiogram or CMR during pregnancy. In total, 35% of pregnancies were associated with a hypertensive disorder of pregnancy, including pre-eclampsia in 20.5%. Potentially life-threatening complications occurred in four patients (3.3%), but no maternal deaths were reported. The mean birth weight of the singletons was 3150 g and that of twins is 2200 g. These outcome statistics need to be viewed with caution considering the knowledge that with OD in non-TS women, the risk for hypertension during pregnancy is already twofold to threefold higher than observed after conventional IVF/intracytoplasmic sperm injection (ICSI). Rates of cesarean section, premature birth and low birth weight are also increased compared with conventional IVF/ICSI (234, 235). The higher risk of hypertensive disorder with OD pregnancies in TS women is postulated to be related to abnormal placentation due to maternal immunological reactions from a genetically foreign embryo (236).

3.4. Fertility preservation in Turner syndrome

Fertility preservation is potentially feasible in women with TS, as many girls with TS have ovarian follicles until their late teens, and some women with mosaic TS have follicles for many years thereafter, even though they tend to experience early menopause (237, 238). Oocyte cryopreservation after controlled ovarian hyperstimulation is a possible fertility preservation option in young mosaic TS women with persistent ovarian function. Case reports describe cryopreservation of 8–13 mature oocytes after controlled ovarian hyperstimulation in TS women aged 14–28 years (239, 240, 241). So far, there are no pregnancies reported after oocyte freezing and thawing in TS, as these women are still young and have not attempted pregnancy yet. Vitrification of oocytes at an even younger age, perhaps about 12 years, may be feasible, but so far, there are no reliable data in TS (239, 240, 249).

Biopsy of ovarian cortical tissue is feasible at younger ages, but it requires an operation and anesthesia. In a study of follicular density in TS girls, the youngest girl studied was 8 years old, but she had very small ovaries with no follicles (237). Signs of spontaneous puberty, mosaicism, and normal FSH and AMH concentrations for age and pubertal stage were positive and statistically significant, but not exclusive, prognostic factors to find follicles. To date, there is not enough evidence to recommend routine fertility preservation of young TS girls before the age of 12 years (237, 238).

3.5. Counseling and ethical considerations about fertility preservation or fertility treatment

R 3.6. We suggest that other options for motherhood such as adoption or using a gestational carrier should be mentioned during preconception counseling (⨁◯◯◯).

As women with TS have rapidly decreasing ovarian reserve from a very young age, it is vital to counsel them that their chances to conceive spontaneously decrease rapidly and consideration should be given to offering fertility treatment at a younger age. Counseling regarding fertility aspects should begin at the time of diagnosis to allow the patient and her parent(s)/guardian(s) to have time to consider its implications and the possibility of fertility preservation. Prior to initiating investigation or treatment, it is important for both the physician and patient to consider the ethical implications of proceeding with fertility preservation or fertility treatment. The decision to proceed with treatment should consider many factors, including the apparent increased risk of pregnancy loss or birth defects using autologous oocytes. This may result in grief and hardship, and the patient must be prepared and accepting of this potential outcome. Patients should be counseled about the availability of prenatal genetic diagnosis and pre-implantation genetic diagnosis or screening, although they may not have adequate embryos available from IVF for these procedures. Mothers of daughters with TS might be willing to freeze their own oocytes for future use by the daughter (250). This is an option only in countries allowing egg vitrification and use of a known oocyte donor. Given that such intrafamilial donation comes with unique practical implications determined by different practice rules in different countries, careful additional ethical counseling seems warranted.

Whether conceiving with autologous or donated oocytes, the patient needs to be fully counseled regarding the increased maternal morbidity and mortality. Intensive cardiac screening is recommended before pregnancy, but normal results do not preclude the possibility of maternal death from AoD or aortic rupture (228, 231, 251, 252, 253). More challenging is the patient who has identifiable risk factors for AoD, but who wishes to proceed with fertility treatment. In this scenario, the physician must consider the safety of the patient as well as her reproductive rights. In preconception counseling, other options for motherhood such as adoption or using a gestational carrier should also be mentioned.

3.6. Cardiovascular risks during pregnancy

R 3.7. We suggest that all women with TS should be counseled about the increased cardiovascular risk of pregnancy (⨁◯◯◯).

R 3.8. We recommend imaging of the thoracic aorta and heart with a transthoracic echocardiography (TTE) and CT/CMR within 2 years before planned pregnancy or ART in all women with TS (⨁◯◯◯).

R 3.9. We suggest that ART or spontaneous conception should be avoided in case of an ascending ASI of >2.5 cm/m² or an ascending ASI 2.0–2.5 cm/m² with associated risk factors for AoD, which include bicuspid aortic valve, elongation of the transverse aorta, coarctation of the aorta and hypertension (⨁◯◯◯).

R 3.10. We suggest that women with a history of AoD should be advised against pregnancy. If already pregnant, these women should be followed very closely at a specialist center and deliver by cesarean section (⨁◯◯◯).

R 3.11. We suggest performing transthoracic echocardiography (TTE) in women with TS without aortic dilatation or other risk factors (hypertension, bicuspid aortic valve, coarctation, previous aortic surgery) at least once during pregnancy, at approximately 20 weeks of gestation (⨁◯◯◯).

R 3.12. We suggest that women with TS with an ascending ASI >2.0 cm/m² or any risk factor (hypertension, bicuspid aortic valve, coarctation, previous AoD or surgery) should be monitored frequently, including TTE at 4- to 8-week intervals during pregnancy and during the first 6 months postpartum (⨁◯◯◯).

R 3.13. We suggest that CMR (without gadolinium) should be performed during pregnancy when there is suspicion of disease of the distal ascending aorta, aortic arch or descending aorta (⨁◯◯◯).

R 3.14. We recommend that blood pressure control is strict (135/85 mmHg) in all pregnant women with TS (⨁◯◯◯).

R 3.15. We suggest that during pregnancy prophylactic surgery is reasonable in case of a dilated aorta with rapid increase in diameter (⨁◯◯◯).

In 1997, the first reports of serious cardiac complications and deaths in pregnant women with TS were published (254), followed by additional reports of deaths after AoD related to pregnancy (228, 231, 251, 252, 253). The risk of maternal death from aortic dissection/rupture during pregnancy in women with TS has been estimated to be 2% (231, 255). However, only 38–49% of these women had any cardiac evaluation before embryo transfer. Published studies vary from 0% cardiac evaluation to 100% before fertility treatment (Supplementary Table 3). In the two largest studies to date, including 93 and 106 women, only 38% and 49% respectively had cardiac examinations with echocardiography or CMR imaging before pregnancy (231, 233).

Additional studies have estimated the risk of maternal cardiac complications to be 0–4% in OD pregnancies in women with TS (222, 224, 225, 226, 231, 232, 233). The risk of AoD during a multiple gestation is estimated in the literature to be approximately five times higher than that associated with a singleton pregnancy (232). It is imperative that any woman with TS undergoing treatment with IVF or OD has a mandatory single embryo transfer (233, 255).

4.1. Background

Girls and women with TS face a lifelong heavy burden of congenital and acquired cardiovascular disease, with increased mortality and morbidity (37, 268). Congenital heart disease occurs in approximately 50% of girls with TS, including a high incidence of bicuspid aortic valve, coarctation of the aorta and an aortopathy that can lead to rare but often fatal dissection or rupture of the thoracic aorta. In addition, a generalized arteriopathy is observed (269), and TS alone is an independent risk factor for thoracic aortic dilation (270). While AoD has received more attention since previous international guidelines were published, it has become increasingly apparent that other cardiovascular conditions such as systemic hypertension, ischemic heart disease and cerebrovascular disease (stroke) are the major factors that reduce the lifespan in TS (268). This consensus statement proposes specific aortic dimensions that (1) may warrant consideration for operative intervention, (2) help in decision-making regarding participation in competitive sports and (3) serve to clarify when more frequent health monitoring may be beneficial. The rapidly evolving field of ART is increasing the potential for childbearing for women with TS. Accordingly, it is imperative that reproductive health practitioners and obstetricians understand who might safely attempt pregnancy.

4.2. Medical and operative management of aortic enlargement and aneurysm

R 4.1. We recommend that an infant or child is examined with transthoracic echocardiography (TTE) at the time of diagnosis, even if the fetal echocardiogram or postnatal cardiac examination was normal (⨁⨁◯◯).

R 4.2. We recommend that girls or women with aortic dilatation and/or bicuspid aortic valve be counseled to seek prompt evaluation if they are experiencing acute symptoms consistent with AoD, such as chest, neck, shoulder, back or flank discomfort, particularly if it is sudden in onset and severe (⨁⨁◯◯).

Aortic dissection in Turner syndrome: In TS, AoD occurs in approximately 40 per 100 000 person-years compared to 6 per 100 000 person-years in the general population (271). Most AoD originate in the ascending aorta (Type A), while a smaller percentage (around 10%) originate in the descending thoracic aorta (Type B) (271, 272, 273). It is important to note that for women with TS, AoD appears to occur at smaller ascending aortic diameters than in those with other genetically triggered aortopathies (272, 273, 274). In addition, AoD also occurs at an age (median age 29–35, range 4–64 years) like others with genetically triggered aortopathies (273). Furthermore, in women with AoD, a cardiovascular abnormality such as bicuspid aortic valve, elongation of the transverse aorta, coarctation of the aorta and/or hypertension is common (272, 273, 274, 275). Aortic dilatation and enlargement of the brachiocephalic and carotid arteries may be present even in the absence of structural heart disease, suggesting the presence of an underlying vasculopathy (275, 276, 277). Using ascending ASI (ASI = aortic size index or absolute aortic diameter in cm divided by body surface area (BSA)) has been proposed to better predict risk for AoD in TS (270, 271, 272, 273, 278). The ASI decreases with body growth and age until mid-teenage years and remains relatively stable thereafter making it a useful index beyond the age of 15 years. Limited data suggest that AoD usually occurs above 16 years of age and at an ascending ASI ≥2.5 cm/m². Because the ascending aorta ASI is age dependent, in those patients aged below 16 years, TS-specific Z-scores could be used. In a woman with average BSA for TS, an absolute ascending aortic diameter of >4 cm is equivalent to an ascending ASI of >2.5 cm/m² and corresponds to a TS-specific Z-score of >4 (279), as determined by echocardiography. Although BMI is not consistently a predictor of aortic size (279, 280) and because BSA calculations also include body weight, decisions based upon either Z-score or ASI should be made with caution. This is particularly important in short-statured but obese individuals or those who weigh very little relative to their height. In these cases, an absolute ascending aorta diameter of 4 cm in someone ≥16 years of age may be preferable to ASI when determining AoD risk.

4.3. Cardiac imaging (see monitoring protocols Figs 1 and 2)

R 4.7. We recommend that in adolescents and adults TS cardiovascular screening with TTE and CMR at time of diagnosis is the preferred approach (⨁⨁◯◯).

Suggested monitoring protocol for girls with TS from infancy to 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; TSZ, Turner syndrome specific Z-score of the aorta (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Suggested monitoring protocol for girls with TS from infancy to 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; TSZ, Turner syndrome specific Z-score of the aorta (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

Suggested monitoring protocol for girls with TS from infancy to 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; TSZ, Turner syndrome specific Z-score of the aorta (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

Suggested cardiac monitoring protocol for girls and women with TS above 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; ASI, aortic size index (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Suggested cardiac monitoring protocol for girls and women with TS above 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; ASI, aortic size index (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

Suggested cardiac monitoring protocol for girls and women with TS above 16 years of age. TTE, transthoracic echocardiography; CMR, cardiac magnetic resonance imaging; ECG, electrocardiogram; CoA, coarctation of aorta; BAV, bicuspid aortic valve; HTN, hypertension; ASI, aortic size index (see text for explanation).

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

R 4.8. We recommend that a CMR scan is performed as soon as it is feasible without needing general anesthesia. If an adult or child cannot tolerate a CMR study, a CT scan is a reasonable option (⨁⨁◯◯).

R 4.9. We recommend, in the absence of a bicuspid aortic valve or other significant disease at the initial screening, TTE or CMR surveillance studies should be performed every 5 years in children, every 10 years in adults or prior to anticipated pregnancy (see R 3.8) to evaluate the aorta based on published guidelines (⨁⨁◯◯).

R 4.10. We recommend that, if TS is highly suspected or has been confirmed prenatally, a fetal echocardiogram should be performed (⨁⨁◯◯).

R 4.11. We recommend that diagnosis of a bicuspid aortic valve or a left-sided obstructive lesion in a female fetus or child should prompt a genetic evaluation for TS (⨁⨁◯◯).

R 4.12. We recommend referral to a pediatric cardiologist when congenital heart disease is detected prenatally in a fetus with TS to provide counseling regarding the anatomy and physiology of the specific defect, recommended site and mode of delivery and postnatal multidisciplinary management plan (⨁⨁◯◯).

Because of the high prevalence of congenital and acquired cardiovascular disease in TS, non-invasive cardiac imaging is crucial for diagnosis, management and risk assessment (2, 9, 286, 287). The most common modalities include TTE, CMR and CT (275, 288, 289, 290, 291, 292, 293). TTE is useful in the diagnosis of a bicuspid aortic valve (294) and other congenital heart diseases as well as in the surveillance of aortic dilatation (290). However, the high prevalence of undiagnosed abnormalities such as elongation of the transverse aortic arch, aortic coarctation and partial anomalous pulmonary venous return in TS has led to increased utility of CMR as a screening and surveillance tool (275, 289, 291, 293). CMR has also been shown to be better than TTE in adults with TS for the diagnosis of bicuspid aortic valve (288). CT is another option, but serial surveillance (294) will involve radiation exposure. In addition, both CMR and CT are more sensitive than TTE to changes in aortic size, particularly beyond the aortic root and in adults with TS, who often have limited echocardiographic windows (295, 296).

4.4. Congenital heart disease

Congenital heart disease occurs in 23–50% of individuals with TS and is the most frequent cause of early mortality (21, 297, 298). The incidence is higher in individuals with 45,X compared to X mosaicism or other X structural abnormalities (299, 300). Left-sided obstructive lesions are most common, with a baseline prevalence of 15–30% for bicuspid aortic valve and 7–18% for coarctation (193, 301). Since bicuspid aortic valve is likely to occur 30–60 times more frequently in TS than that in the 46,XX females, it is possible that bicuspid aortic valve in a female may be an independent marker for the TS diagnosis. At least 12.6% of newborn girls with coarctation of the aorta have TS and should be viewed as an independent marker of TS (302). Cross-sectional imaging modalities have unveiled an increased incidence of additional vascular anomalies that might otherwise have gone undetected by TTE, including partial anomalous pulmonary venous connection, left superior vena cava, an elongated transverse arch and dilatation of the brachiocephalic arteries (288, 291). Neck webbing and an increased anterior–posterior thoracic diameter have been shown to be strong predictors of arterial and venous anomalies in TS (303, 304). Additional less frequently occurring anomalies include hypoplastic left heart syndrome, mitral valve anomalies, interrupted inferior vena cava with azygous continuation, cardiac dextroposition, ventricular septal defect, atrioventricular septal defect, pulmonary valve abnormalities, coronary artery anomalies (305, 306) and patent ductus arteriosus (307). Congenital, morphological coronary artery anomalies appear to be common in TS (306). Whether coronary malformations increase mortality risk is unknown. It is important for the cardiothoracic surgeon to be aware of unusual coronary anatomy because it may necessitate modifications of the operative approach.

4.5. Electrocardiogram

R 4.13. We suggest that a resting electrocardiogram (ECG) with QTc measurement should be done in every individual with TS at the time of diagnosis and that Hodge’s may be preferred over Bazett’s formula to estimate QTc (⨁⨁◯◯).

R 4.14. We suggest that 24-h Holter monitoring and exercise testing be considered for risk estimation in women with TS with QTc interval prolongation (QTc >460 ms) (⨁◯◯◯).

R 4.15. We suggest that, in individuals with prolonged QTc, drugs that prolong the QTc should be avoided. If they are deemed necessary, ECG should be performed 1–2 weeks after initiation of QT-prolonging drugs (⨁◯◯◯).

Differences between the ECG in TS and in the general population can be roughly categorized into morphological issues (bundle branch block, T-wave changes and P-wave changes) on the one hand, and time intervals (PR-interval and QT-interval) on the other hand. The reported prevalence of these changes in women and girls with TS is about 50%, which is higher than that in non-TS controls (30%) (308). Some changes such as those in P-wave and QTc-dispersion and heart rate variability in women with TS can be attributed to the underlying characteristic autonomic dysfunction (309). Shortening of the PR-interval (due to accelerated atrioventricular conduction) may be a consequence of excessive sympathetic drive. The clinical relevance of these potential abnormalities may be twofold: (1) right-axis deviation in an individual with TS is correlated with the presence of partially abnormal pulmonary venous return (310) and should trigger further diagnostics in those cases that are not already known and (2) QTc prolongation is associated with an increased risk for arrhythmias or even sudden cardiac death in the general population. Yet, it should be emphasized that there is no published evidence so far for sudden cardiac death related to QTc prolongation in women with TS, although case-based QTc prolongation due to a dose of amiodarone followed by cardiac arrest has been seen (311). Additional uncertainties include the threshold to define QT prolongation, and calculation method to define QTc interval in TS – in view of the increased intrinsic heart rate in many individuals with TS, Hodge’s formula may be preferred over Bazett’s formula, because it takes the higher heart rate into account (312). Whether QTc prolongation should be considered as an intrinsic feature of TS is unclear; a potential correlation with variants in the LQTS genes deserves further investigation (313). Caution with the use of QT-prolonging drugs may be warranted in this population.

4.6. Sports participation

R 4.16. We recommend that the function of the aortic valve and the presence of any other congenital heart disease and/or hypertension should be considered in determining participation recommendations for the athlete with TS and aortic dilation (⨁◯◯◯).

R 4.17. We suggest that, for girls and women with TS ≥16 years with a moderately dilated aorta (ascending ASI ≥2.0 cm/m²), avoidance of intense weight-training should be advised (⨁◯◯◯).

R 4.18. We suggest that, for girls and women with normal aortic size (age <16 years; TS-specific Z-score of <2.5 or age ≥16 years and ASI <2.0 cm/m²), it is reasonable to participate in all sports (⨁◯◯◯).

R 4.19. We suggest that, for girls and women with a mild to moderately dilated aorta (age <16 years old (TS-specific Z-score of 2.5–3), or age ≥16 years (ASI 2.0–2.3 cm/m²)), participation in low and moderate static and dynamic competitive sports may be advised (⨁◯◯◯).

R 4.20. We suggest that girls and women with a moderately to severely dilated aorta, age <16 years (TS-specific Z-score of >3) or age ≥16 years (ASI >2.3 cm/m²) should be advised not to participate in any competitive sports (⨁◯◯◯).

A safe level of exercise is important for a healthy lifestyle in girls and women with TS. Evidence is lacking regarding cardiovascular and aortic risks for competitive athletics with TS. Given the propensity for obesity and the metabolic syndrome in TS, health care professionals should be mindful of the significant benefits of having a ‘heart-healthy’ lifestyle in light of the low risk of AoD in absolute terms (about 40:100 000 patient-years (271)) in this population. Therefore, consideration of the risk for AoD should be tempered by the importance of encouraging individualized levels of physical activity in individuals with TS. In addition, there is no published evidence that collision represents a significant threat for AoD among girls and women with TS without significant aortic enlargement.

4.7. Hypertension

R 4.21. We recommend that in individuals without structural heart disease, annual assessment of blood pressure should be performed and medical treatment thereof should be considered if hypertension is present. We suggest medical treatment to include a beta-blocker, an angiotensin receptor blocker, or both to reduce the risk for AoD in women with TS who are ≥16 years of age for whom their ascending ASI is ≥2.3 cm/m² (⨁⨁◯◯).

R 4.22. We suggest that medical treatment, including a beta-blocker, an angiotensin receptor blocker or both, to reduce dilatation of an enlarged aortic root and/or ascending aorta may be considered for girls with TS who are ≤16 years of age for whom their ascending aorta TS-specific Z-score is ≥3.0 (⨁◯◯◯).

Although individuals with TS are frequently found to have systemic hypertension, and even though hypertension is a known risk factor for progressive cardiac dysfunction and aortopathy, little has been documented about definitions and management of systemic hypertension, specifically for women with TS (281). Individuals with TS have been considered to have systemic hypertension in several studies with frequency as high as 20–40% in childhood (282) and in as many as 60% of adults (191, 314, 315, 316). Systemic hypertension may appear at early ages and continue through adulthood. Systemic hypertension may be the result of renal anomalies that are frequently seen in TS or may be idiopathic. Hypertension can persist after coarctation repair even in those without residual descending aortic pressure gradients and possibly the intrinsic shape of the aorta in individuals with TS without coarctation can be a factor in the etiology of hypertension for some (283).

Several guidelines regarding measurement and ascertainment of systemic hypertension in infants, children, adolescents (317) and adults (318) are available, but none specifically addresses individuals with TS. De Groote et al. (281) suggest a practical approach to hypertension identification and management for girls and women with TS, including a practical algorithm for blood pressure evaluation in adult patients. Left ventricular hypertrophy (increased LV mass) has been identified in TS, even in those who are normotensive (319, 320). This could be an end-organ effect of hypertension that is masked during resting blood pressure assessment or is related to loss of diurnal variation (lack of night-time dipping). Ambulatory blood pressure monitoring (ABPM) has been considered useful in demonstrating abnormal diurnal variation in blood pressure values in women with TS (191, 314, 321, 322).

4.8. The coagulation system

Cerebrovascular accidents (strokes) occur in excess of the general population (268), but whether this is simply related to the increased risk of systemic hypertension or other TS-specific causes is unknown. Disturbances of thrombosis and fibrinolysis are related to thromboembolic stroke. Clotting factors and clotting times may be normal for cohorts with TS when assessed in total, but, on the individual level, many will have increased procoagulant values of clotting and fibrinolytic factors (323). Fibrinogen has been found to be elevated in 65% of females with TS, and proteins C and S were reduced in a large fraction (323). Conversely, clotting factors, fibrinolytic factors and fibrinogen levels, as well as clotting times have also been reported to be within the normal range in other cohorts with TS (315, 324), though high-normal values have been reported for some procoagulant factors (325, 326). The most common mutations associated with thrombus formation are more frequently reported in TS (325, 326). One study showed that factor V Leiden G1691A gene polymorphism heterozygosity is more prevalent in TS (13%) than that in the background population (2%) (325).

Even though the clotting system appears to be excessively activated in some women with TS, outcome data are lacking, and the common denominator has not yet been found. Therefore, no general recommendation can be issued concerning the coagulation system, but awareness about thromboembolic disease in TS will help identify the few women with TS women with coagulation disorders.

5.1. Use of generic and TS-specific tools to meet the transition challenge

R 5.2. We suggest that the pediatric endocrinology team uses or adapts available transition tools to track and document the core elements of transition (⨁⨁◯◯).

R 5.3. We suggest that, irrespective of the health care delivery system, the pediatric and adult health care teams establish a workflow to support a coordinated transition process (⨁⨁◯◯).

R 5.4. We suggest that pediatric endocrinologists and their care teams encourage peer-to-peer (and parent-to-parent) contact with TS support and advocacy organizations to enhance knowledge and confidence, reduce stress and distress and promote the reciprocal sharing of experiences (⨁⨁◯◯).

Many professional societies have endorsed health care transition preparation, planning, and implementation as a key component of quality health care. Generic tools are available at the American Academy of Pediatrics Got Transition Website (www.gottransition.org). An example of condition-specific tools is the TS Pediatric to Adult Care Transition Toolkit, co-produced by the Endocrine Society, Hormone Health Network, Turner Syndrome Foundation and American College of Physicians (ACP) (https://www.acponline.org/system/files/documents/clinical_information/high_value_care/clinician_resources/pediatric_adult_care_transitions/endo_turner/endo_ts_transition_tools.pdf) to help the young adult in transition to achieve optimal self-care as an emerging adult. Each condition-specific set of tools includes, at a minimum, the following three core transition elements:

In addition to the three core elements mentioned previously, the customized set of TS-specific tools includes guiding principles for estrogen therapy, a recommended approach for planning for pediatric practices and a recommended approach for transitioning to an adult practice, which underscores the need for the TS young adult to clarify the roles of each provider on her adult care team.

5.2. Key TS-specific content areas to be addressed during transition

Estrogen therapy and reproductive issues: The impact of TS on puberty should be discussed in simple terms with TS patient by 9–11 years of age, with further explanations during adolescence. The induction of puberty with more physiological hormonal replacement regimens by age 11–12 years provides most girls with TS with documented ovarian insufficiency the psychological benefit of experiencing an appropriate timing of puberty. The scope of available and emerging reproductive options is to be reviewed. Serious health risks of spontaneous or assisted pregnancy need to be discussed. The benefits of long-term estrogen therapy to sustain vascular, bone and psychosexual health, as well as to maintain reproductive organ and tissue integrity are to be emphasized.

Associated needs and lifestyle requirements to ensure optimal health outcomes: Pediatric providers need to articulate clearly to each young adult with TS how, in addition to the need for continued follow-up care for specific conditions identified during childhood, that there is a risk for conditions not yet detected by the time of their transition to adult care. These conditions need to be specified as well as three key preventive measures that can be taken to stay healthy – (1) live a healthy and active lifestyle, (2) have blood pressure carefully monitored and treated if hypertension develops and (3) continue to take sufficient estrogen well into adulthood. It is essential for each young adult with TS to understand how early prevention or identification and treatment of related comorbidities will determine their future health, QoL and longevity.

Cardiovascular health care from childhood to adulthood: Girls with TS are at increased risk of obesity, elevated cholesterol, diabetes, hypertension, stroke and ischemic heart disease. Table 6 is a list of topics that are helpful to discuss. During this transition period, the importance of preparation for adulthood cardiovascular care is emphasized with the aim of ensuring guideline-driven care and a reduction in morbidity and loss to medical follow-up.

Psychosocial, educational and vocational issues to ensure full potential and high quality of life: The typical neurocognitive profile, personality type and difficult social adjustment observed in TS pose psychosocial risk that, if not addressed during transition, can impact QoL. It is also important for pediatric providers to ascertain that educational and career goals align with the abilities of the individual TS young woman and to provide appropriate counseling as needed. See also Section 7.

5.3. Transition care models from adolescent to adult care and guiding principles

Joint pediatric-adult care multidisciplinary programs offer the advantage of providing a more seamless transition and exchange of health care information from pediatric to adult care by sharing similar members of the health care team, a familiar site of care to the transitioning patient, and the same electronic health records. This vertically integrated care model appears more prevalent in Europe where each center serves large populations of TS patients and report low rates of loss to follow-up care when young adults leave the pediatric phase of their care. In the United States, similar models exist in few places.

Structural requirements: The reality is that many TS young adults do not have access to an integrated pediatric-adult multidisciplinary care program. Some have access to adult care providers within the same health care system using the same electronic health records (EHR), while others need to transition to one or more adult providers outside of their health care system (Fig. 3). Regardless of the care model, it does not preclude a quality transition as long as a staged transition process that includes the core elements of transition as detailed above is followed. Identifying an office and clinical ‘transition champion’ to oversee the implementation of transition activities is recommended. It is essential to define which member(s) of the pediatric care team is(are) responsible for implementing each core element and to design an office workflow to support transition. Helping the individual with TS and her family identify an adult endocrinologist and other adult clinicians to best meet their needs is an essential role for the pediatric endocrinology care team to fulfill. Establishing a good partnership with the receiving adult providers is encouraged and needed. Overcoming non-interoperability of health information between pediatric and adult providers using different EHRs will assist in a seamless transition and a printed rather than digital health summary document may be required. In some settings, a brief period of shared or co-management between the pediatric and receiving adult endocrinologist may be beneficial to ensure a successful transition and ascertain that the TS patient successfully transferred to her adult provider(s). Leveraging of telemedicine platforms may open the door to yet another transition strategy in the near future.

Transition models from pediatric to adult care.

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Transition models from pediatric to adult care.

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

Transition models from pediatric to adult care.

Citation: European Journal of Endocrinology 177, 3; 10.1530/EJE-17-0430

• Download Figure
• Download figure as PowerPoint slide

6.1. Comorbidities in Turner syndrome (from childhood to adulthood)

In this section, management recommendations for comorbidities of TS not covered elsewhere will be made. Attention will be given to: (1) the age of subjects, (2) geographical differences in management of issues such as hyperlipidemia and vitamin D deficiency (as occur in normal populations) and (3) cost–benefit ratios. There exists little evidence for many of the following recommendations beyond expert opinion. For a summary of recommended screenings see Table 6.

6.2. The adult clinic

R 6.21. We recommend that girls and women with TS attend specialist inter- or multidisciplinary clinics for health surveillance (⨁⨁◯◯).

The burden of health care issues for adults is sufficiently great to warrant an annual assessment preferably in a multidisciplinary clinic familiar with the natural history of TS. Items for a medical review are listed in Table 8. Recommendations on the frequency of testing are based on the incidence of new complications arising in adults with TS. For instance, hypertension and hypothyroidism affect approximately 25–50% of adults with a prevalence increasing with age.

As women with TS frequently report suboptimal care when they attend multiple specialty clinics without central coordination, a multidisciplinary clinic should strive to provide one-stop-shopping patient-centered care. Because of the multisystem nature of health surveillance, a variety of practitioners should be accessible either on the day of the visit or through a direct referral system for another day. Relevant specialties include endocrinology, gynecology (including fertility), cardiology (preferably with expertise in adult congenital heart disease), genetics, psychology, otolaryngology, audiology, dermatology and gastroenterology (including hepatology). Allied health care providers, such as nutritionists and physical therapists make useful additions to the team.

Clinic participation should also cover lifestyle factors such as exercise, diet and weight control, psychosocial issues including relationships both personal and work related, sexual function and plans for future fertility. A clinic care coordinator can be helpful as the point of contact, be vigilant for psychosocial issues, ensure access to tests that may be required and follow-up on non-attendance.

In common with the general population, obesity is a central factor determining many long-term health outcomes including risk of hypertension, diabetes and steatohepatitis. Therefore, weight management should be a core activity of the adult TS clinic.

Contraception and fertility options should be reviewed regularly taking into account the evidence presented in Section 6. In addition to assisted reproduction technologies, the team should be familiar with the pathways to adoption.

Given that multidisciplinary coordination at a specialty center may occur annually, engagement with a primary provider to care for urgent issues as well as routine preventative adult care is highly encouraged.

7.1. Background

Turner syndrome is associated with a neurocognitive profile, which may negatively impact educational attainment and HRQoL (Table 9). Behavioral health specialists, trained to diagnose or provide interventions for the educational and psychosocial difficulties sometimes associated with TS, include clinical neuropsychologists and clinical psychologists.

Although not yet the standard of care in many countries, regular screenings for developmental or emotional/behavioral problems are increasingly called for by national and international health organizations (461, 462). In the United States, recommended preventive pediatric health care for all children (beginning in the first year) and adolescents includes annual developmental and behavioral screenings (462). Annual mental health screenings are also mandated in adult populations (463). These screenings are optimally part of routine visits to primary-care providers. In settings where such services are not delivered, then specialists caring for girls and women with TS should consider incorporating psychometrically robust screening tools (464) as a feature of the model of care (2). Ideal measures will share the characteristics of brief administration time, simple scoring and validation in multiple languages. An example of such a measure is the Strengths and Difficulties Questionnaire (SDQ) (465).

Following recommendations and guidelines in other chronic pediatric conditions involving threats to neurocognitive function (466, 467), behavioral screenings are optimally augmented by neuropsychological assessments sensitive to the well-documented neurocognitive profile in TS (Table 10). Changes in the social, emotional and educational environment associated with passage through childhood, adolescence and young adulthood, represent both reasonable and necessary indications to conduct a neuropsychological assessment. Accordingly, it is recommended that a neuropsychological evaluation be conducted in early life (preschool), at school entry, at transition to high school and higher education, or at any time that difficulties arise (468). If a neuropsychologist (or otherwise qualified psychologist) is not available to serve as a member of the multidisciplinary team, then effort should be directed at identifying such providers in the community to whom a referral can be made (e.g. school psychologists). It is also recommended that children be referred for occupational, physical and speech therapy in early life or at school entry, as warranted, given their emerging pattern of developmental strengths and weaknesses as well as substantial risk for motor and possible communication difficulties. While there is a growing literature that outcome is better for children who receive early intervention with a host of developmental disorders (469), the intensity or dosage of treatment required to substantially alter outcome is an area of active research (470, 471) and contingent on the deficit being targeted and therapy employed (472).

The following sections review the neurocognitive, psychoeducational and psychosocial domains most likely affected in TS and their potential impact on adaptive function (Table 9). Interventions specifically targeting biological mechanisms related to the core cognitive phenotype of TS have yet to be developed. However, the cognitive and psychosocial challenges associated with TS align with recognized diagnoses, including attention-deficit/hyperactivity disorder (ADHD) (72, 473), specific learning disorders (474), social communication disorder, autism spectrum disorders (475) and developmental coordination disorder (476), and evidence suggests that generic interventions developed for non-TS populations may be adapted with similar positive effects (477) (Table 10).

As in other genetic syndromes, it is critical to note the high degree of individual variability in the somatic, cognitive and psychosocial phenotypes of girls and women with TS; one should never assume the presence (or absence) of a deficit based on karyotype. Furthermore, neurodevelopmental function changes dynamically over the course of development and may significantly improve with appropriate intervention. Nevertheless, awareness of the challenges that TS can present around academic, interpersonal and vocational goals reflects a prudent course of action for clinicians, girls and women with TS and their families.

7.2. Intellectual functioning

The majority of individuals with TS are of average intelligence; nevertheless, approximately 10% experience intellectual disability (9). Verbal reasoning is consistently higher than is perceptual reasoning, likely reflecting specific impairments in executive functions and visuospatial abilities (478). A ring X chromosome is associated with the highest risk for intellectual disability (479, 480). When intellectual disability is present, appropriate support must be provided to optimize adaptive functioning and educational attainment (479, 480).

7.3. Attention, working memory and cognitive control

Deficits in executive functioning occur in a subset of girls and women with TS. These may include problems in switching attention between tasks, planning, inhibiting inappropriate responses and working memory (the transient holding, processing and manipulation of information) (478, 481). Reduced processing speed (the pace at which one takes in information, makes sense of it and begins to respond) is also a frequent finding (478). ADHD is observed in approximately 25% of school-aged girls with TS and is often accompanied by a high rate of hyperactivity symptoms. It is unclear if these symptoms continue in adulthood. Executive impairments are less common in girls with TS without ADHD behaviors and appear to be independent of visuospatial deficits (473). Problems stemming from executive function issues – in children or adults – can be ameliorated via cognitive-behavior therapy (482, 483, 484). For children, this approach commonly includes parent management training and classroom modifications (485, 486). Workplace accommodations for adults may also be beneficial (482, 483).

7.4. Visuospatial, perceptual-motor and sensorimotor development

The most frequently observed cognitive challenges in girls and women with TS are in the visuospatial domain. Visuospatial issues may manifest as difficulties with everyday tasks that involve a visual component (e.g., aspects of driving, finding one’s way around a new place and/or judging distances (445, 487)) and contribute to poor performance in mathematics courses. Studies on motor function are relatively limited, but suggest both general and specific motor impairments, including difficulties in visuo-motor coordination, learning and fine motor function (354, 476, 488). In some cases, a diagnosis of developmental coordination disorder may be appropriate. Children should be referred for academic support if spatial deficits are interfering with academic functioning. Occupational and physical therapies are additional strategies for achieving therapeutic gains (489).

7.5. Speech and language

Average to above-average performance on standardized batteries of receptive and expressive language functioning (490, 491) has been noted in girls with TS. However, weaknesses are reported in language tasks that incorporate executive demands (490) or which rely on spatial language (492). When speech and language issues arise that substantially impact communication or academic performance, referral to a speech and language pathologist is warranted. Verbal strengths can be used to enhance academic and work performance.

7.6. Declarative memory

Long-term memory for verbal information may be preserved or enhanced (493, 494, 495), though some studies report deficits (496, 497). In contrast, object and location memory are reported as less efficient, which may be secondary to accelerated forgetting (498). Visual learning may be improved by describing materials aloud and incorporating verbal mnemonics (499). Strong verbal memory can be used to enhance academic and work performance.

7.7. Social cognition

Girls and women with TS may demonstrate poor facial and emotion recognition (500, 501, 502) and difficulty understanding the mental states of others (503). Some literature suggests that individuals with TS are at greater risk for autism spectrum disorder (ASD) (480, 504), although this remains controversial (505). When symptoms are present, referral to a psychologist, a specialist with training in applied behavior analysis or a psychiatrist is indicated for further evaluation and treatment. Social skills group therapy through a school or community provider may prove helpful. Social skills interventions developed for individuals with ASD, such as the Program for the Education and Enrichment of Relational Skills (PEERS) (506), may be adapted for individuals with TS.

7.8. Educational attainment and professional satisfaction

The academic phenotype of TS includes a mathematics learning disorder (dyscalculia) (507, 508, 509, 510), with an estimated prevalence of about 50% relative to a 6–10% estimate in the general population (511). School-aged children with TS display advanced reading and word recognition (490). Findings regarding phonological processing (512, 513) and reading comprehension are inconsistent (512, 514). When learning issues are present, academic accommodations should be made, including tutoring, empirically supported interventions, extension of time demands and utilizing learning/teaching strategies that capitalize on verbal strengths.

Despite these well-documented learning issues, women with TS show a similar or increased level of educational attainment compared to the general population (206, 515). The employment status of young women with TS is equal (452, 516, 517) or higher (206, 518) than comparison groups; however, retirement occurs earlier (517, 519). Adult women with TS, especially older cohorts, show a lower occupational status than would be expected from their level of education and report less positive/challenging working experiences (519, 520). Women with TS often choose careers in health care, social services and teaching (206, 515, 518), but such a finding should not be viewed as a recommendation. Instead, career guidance ideally includes vocational/career counseling (521).

Some researchers and clinicians have used the label nonverbal learning disability to characterize the neurocognitive phenotype of TS. Yet, the nonverbal learning disability construct is still in the scientifically ‘formative’ stage (522), and neither the Diagnostic and Statistical Manual of Mental Disorders nor the International Classification of Disease and Related Disorders (ICD-10) offer a specific category for nonverbal learning disability. As such, we have opted to describe the pattern of strengths and weaknesses seen in TS rather than use this label until future research is able, to clearly distinguish essential from non-essential features in various risk groups.

7.9. Psychosocial issues

Studies on the incidence of anxiety or depression in girls (473, 475, 523) and women (516, 524, 525) with TS show conflicting data, possibly due to selection bias, disparate outcome measures and small sample sizes. In general, girls (526, 527) and women (205, 524, 528) with TS experience lower self-concept compared to girls/women without TS. During school age, but also continuing in adulthood, teasing appears to be a frequent problem in TS (49, 516, 529, 530). Studies have failed to detect a relationship between the presence of TS stigmata (e.g., cubitus valgus and webbed neck) and social adaptation (526, 531) or psychological well-being (532). Girls and women with TS experience higher degrees of social isolation (452, 516). They report fewer close friendships (523, 525), marry at lower rates (206, 515, 518, 533) and are less likely to cohabitate (515, 516, 517). These observations may arise, due to difficulties in social cognition. In addition, hearing impairment and limited sexual experience (associated with delayed puberty) predict lower self-esteem and poorer psychosocial adaptation (205, 534, 535, 536).

7.10. Impact of hormonal therapies on neurocognition and behavior

Available data indicate that, on average, adult women with TS demonstrate characteristic neurocognitive profiles (intact verbal abilities and impaired visuospatial skills) despite preserved ovarian function or adequate estrogen replacement (208). These cognitive differences are seen in comparison to women with premature ovarian failure of different etiologies (496), suggesting that, at least by adulthood, endogenous or exogenous estrogen has limited impact on modifying cognitive differences that likely arise much earlier in development. However, a trial of low-dose estrogen administration in girls with TS, between the ages of 5 and 8 years, demonstrated modest improvements in verbal and nonverbal memory (145) and replacement in girls aged 10–12 years demonstrated improvement in processing speed and motor function (144), suggesting a possible treatment effect within specific developmental windows. A relationship between age-appropriate somatic sexual development and positive self-concept, social adaptation and perceptions of health has been reported (49, 534, 537). These data support age-appropriate induction of puberty. Evidence supporting androgen administration in youth with TS is conflicting, with one study indicating that low doses of oxandrolone may improve working memory in girls aged 10–14 years (538), while another study found no effect on neurocognition and higher reports of anxiety and depression with oxandrolone administration in adult women with TS (142).

Studies of GH in TS have failed to demonstrate an effect on neurocognition (539), and some research calls into question the long-standing assumption that short stature leads to psychosocial dysfunction (540, 541). However, research that target shorter-than-average children in the general population may not be representative of the experiences of children who present to pediatric endocrinologists with concerns regarding short stature. Furthermore, tools available for assessment of the impact of short stature on QoL in individuals with TS may lack sensitivity and specificity. A systematic review examining psychological and cognitive effects of GH treatment for short stature in a wide range of syndromes, including TS, indicated that 94% of reported outcomes suffered from a high risk of bias (542). In the sole, epidemiologically oriented study of GH effect on psychosocial (e.g., self-esteem, social adjustment and HRQoL) and psychosexual (e.g., sexual milestones and degree of sexual experiences) adaptation of young adult women with TS, neither AH nor estimated height gained through GH treatment had an effect (452, 534).

8.1. Exploring alternatives to hospital clinics

It is estimated that only a minority of women with TS attend any type of health care provider. Increasingly, patients access health information though Internet searches and social media. However good a clinic set up there is, it can only cater to those who attend, and so all teams should have a strategy to engage with a wider population of women with TS. There are several alternatives to face-to-face clinic encounters. Care for women with TS lends itself to telemedicine and the use of online medical records so that disparate specialists can access one set of records. Such a system can reduce the number of tests and ensure that vital data such as cardiac status are available when acutely unwell.

8.2. Role of patient support organizations

The value of peer support for adolescents with chronic conditions is well established. A systematic review of peer-support interventions in adolescents with chronic illness indicated improvement in adolescents’ behavioral and emotional symptoms (543). We recommend that information regarding access to patient support organizations be made available to each individual with TS on a regular basis. Members of each medical care team should take part in combined meetings with patient groups in order to share experience and develop future care pathways. Many women with TS indicate that belonging to a peer-support group has had a positive impact on their lives (544). Parents of young girls, adolescents and women with TS should be advised to contact a local TS peer-support organization to diminish the risk of isolation and to get specific information and support from peers.

Patient support organizations are vital in providing expertise outside of clinic attendances and for the support of careers or partners of individuals with a new diagnosis of TS. Support organizations can also assist in developing a care network of providers on a national basis. In addition, it is important to link patients and families to local, regional and national patient/advocacy TS groups that offer activities geared to adolescents and toward increasing their developmental autonomy.

Patient and family TS advocacy groups have taken the lead in providing educational and networking forums for patients, families, adolescents, young adults and women with TS. Their advocacy efforts have and continue to lead to important advances in research and care for girls and women with TS.

8.3. National registries for Turner syndrome

Several countries have developed registries recording data for women with TS, and information from these sources has greatly expanded our knowledge regarding the natural history of this condition. Elsewhere, registries are at an early phase of development, and it is likely that improved stratification of risk for rare events will be achieved with the development of these resources.

Examples of registries include one from Sweden that has recorded data on morbidity, including cancer and mortality (all causes shown) in all women with a TS karyotype since 1957 (545). In Denmark, the Central Cytogenetic Registry has been valuable in allowing comparison with clinical data with reduced ascertainment bias compared to clinical registries (37, 517). France and the United Kingdom have developed National Centers for Rare Diseases under the umbrella of which specialist expertise for the care of TS can develop.

We recommend that registries recording clinical and psychosocial data from individuals with TS are established at a national level and that data from them be combined in order to determine factors contributing to rare outcomes such as AoD. Ideally, data are contributed by providers and patients (‘stakeholders’).